Chapitre 1: Introduction générale - BICTEL ULg - Université de Liège

Chapitre 1: Introduction générale - BICTEL ULg - Université de Liège

ACADEMIE UNIVERSITAIRE WALLONIE-EUROPE UNIVERSITE DE LIEGE-FACULTE DE MEDECINE VETERINAIRE DEPARTEMENT DES MALADIES INFECTIEUSES ET PARASITAIRES FUNDA...

13MB Sizes 0 Downloads 0 Views

Recommend Documents

ETUDE DE L'IMPACT D'INTERVENTIONS - BICTEL ULg
PSYCHOLOGIQUES PREVENTIVES SUR LE BURNOUT. DES MEDECINS. Isabelle Bragard. Thèse présentée en vue de l'obtention du t

Bictel Küpper table des matières - BICTEL ULg
Das Fräulein von Scuderi 55. 2.1. Voraussetzungen 55. 2.1.1. Entstehung und Überlieferung 55. 2.1.2. Inhalt. Erste Lektü

Tight Glycaemic Control - BICTEL ULg
I would like to thank Clémentine François, Capucine Lardinois and Charlotte De Bien for their outstanding support and

Chapitre 1
préparé par le boulanger ou le client lui-même », couscoussier « partie supérieure du couscoussier, passoire qui s

Oxidative Stress in the Cochlea: An Update - BICTEL ULg
endothelial growth factor. Am. J. Pathol., 2009, 174, 1692-1704. [95]. Nuttall, A.L. Sound-induced cochlear ischemia/hyp

1 Introduction - De Gruyter
allowing them to add the Imperial double-headed eagle to their coat of arms. ... the papacy in 1689 as Alexander VIII ma

1 Introduction - De Gruyter
Feb 23, 2017 - 'Eat me', water says ' Drink me', thunder says 'Fear me' and a woman says. 'Love me''. [Koffka, 1935, p.

Chapitre 12 - Ville de Témiscaming
des terrains et les premiers à bâtir leurs maisons uni- familiales sur la rue Pine sont George Pinkerton, lames Adams

Chapitre 10 - Ville de Témiscaming
Rolly Violette, Del Viens, Alfred Sauve, E. Clement, Roland Chai-land. .... Freve, Roger Presseauit, Marcel Violette, j'

chapitre 4_caracteristiques de la population - Institut National de la
En ce qui concerne la densité de la population du. Cameroun, le nombre de personnes par km² est de 40 en 2007. Il éta

ACADEMIE UNIVERSITAIRE WALLONIE-EUROPE UNIVERSITE DE LIEGE-FACULTE DE MEDECINE VETERINAIRE DEPARTEMENT DES MALADIES INFECTIEUSES ET PARASITAIRES FUNDAMENTAL AND APPLIED RESEARCH ON ANIMAL HEALTH UNITE DE RECHERCHE EN EPIDEMIOLOGIE ET ANALYSE DE RISQUES APPLIQUEES AUX SCIENCES VETERINAIRES

Contribution à l’épidémiologie de la brucellose bovine en Côte d’Ivoire

Contribution to the epidemiology of bovine brucellosis in Ivory Coast

Moussa SANOGO

THESE PRESENTEE EN VUE DE L’OBTENTION DU GRADE DE DOCTEUR EN SCIENCES VETERINAIRES ORIENTATION MEDECINE VETERINAIRE

ANNEE ACADEMIQUE 2014-2015

The illustration of the title page is a schematic representation of the main routes of brucellosis from livestock to human proposed by Sir David BRUCE (1855-1931) available at the following URL : http://m2002.tripod.com/brucellosis.jpg

Académie Universitaire Wallonie-Europe Université de Liège - Faculté de Médecine Vétérinaire Département des Maladies Infectieuses et Parasitaires Fundamental and Applied Research on Animal and Health Unité de Recherche en Epidémiologie et Analyse de Risques appliquées aux Sciences Vétérinaires (UREAR-ULg) Institut de Médecine Tropicale d’Anvers Département de Sciences Biomédicales Unité de Biostatistique et d’Epidémiologie Laboratoire National d’Appui au Développement Agricole (LANADA) Laboratoire Central Vétérinaire de Bingerville (LCVB)

Contribution à l’épidémiologie de la brucellose bovine en Côte d’Ivoire

Moussa SANOGO

Promoteur:

Prof. Dr. Claude SAEGERMAN Unité de Recherche en Epidémiologie et Analyse de Risques Appliquées aux Sciences Vétérinaires (UREAR-ULg) Fundamental and Applied Research on Animal and Health (FARAH) Département des Maladies Infectieuses et Parasitaires Faculté de Médecine Vétérinaire-Université de Liège Co-promoteur:

Prof. Dr. ir. Dirk BERKVENS Unité de Biostatistique et d’Epidémiologie Département de Sciences Biomédicales Institut de Médecine Tropicale d’Anvers

Wallonia-Europe University Academy University of Liege-Faculty of Veterinary Medicine Department of Infectious and Parasitic Diseases, Fundamental and Applied Research on Animal and Health Research Unit of Epidemiology and Risk Analysis Applied to Veterinary Sciences (UREARULg) Institute of Tropical Medicine of Antwerp Department of Biomedical Sciences Unit of biostatistics and Epidemiology Laboratoire National d’Appui au Développement Agricole (LANADA) Laboratoire Central Vétérinaire de Bingerville (LCVB)

Contribution to the epidemiology of bovine brucellosis in Ivory Coast

Moussa SANOGO

Promoter:

Prof. Dr. Claude SAEGERMAN Research Unit of Epidemiology and Risk Analysis Applied to Veterinary Sciences (UREARULg) Fundamental and Applied Research on Animal and Health (FARAH) Department of Infectious and Parasitic Diseases Faculty of Veterinary Medicine - University of Liege Co-promoter:

Prof. Dr. ir.Dirk BERKVENS Unit of biostatistics and Epidemiology Department of Biomedical Sciences Institute of Tropical Medicine (ITM)

In the memory of my father

To my mother To my lovely wife And to my son and my daughter

ABSTRACT ABSTRACT Bovine brucellosis is an endemic infectious disease which can negatively impact on cattle productivity and welfare as well as on human health. In many developing countries such as Ivory Coast, there is a need for knowledge on the distribution and the frequency of the disease (or evidence of its presence) within the animal population and the possible factors associated with the disease. Information is also needed on species and biovars of Brucella at national and regional scales, on the performance of commonly used diagnostic tests for accurate estimation of the true disease prevalence, and on determination of risk factors associated with the disease. These informations are of key importance to set up and implement appropriate and efficient prevention and control measures against brucellosis. For these reasons, the research presented in this thesis aimed to contribute to a better understanding of the epidemiology of bovine brucellosis in Ivory Coast. The thesis is structured into three main parts. The introduction part includes three chapters. The first chapter presents an overview of the literature on the pathogen causing brucellosis, their characteristics and distribution. The impact and the existing strategies for preventing and controlling brucellosis are discussed with a particular reference to the situation of bovine brucellosis in Ivory Coast. The presence and the importance of the disease were confirmed in the country but the disease is still uncontrolled. In the second chapter (Chapter 2), an insight on statistical, epidemiological principles and concepts applied to achieve the different objectives (Chapter 3) is given, including a discussion on available approaches to estimate diagnostic test characteristics and the true prevalence of a disease. The second part of the thesis includes research on different aspects of the epidemiology of bovine brucellosis in Ivory Coast and West Africa (Chapter 4, 5, 6 and 7). Chapter 4 specifically provides a state-of–the-art knowledge on species and biovars of Brucella reported in cattle from Ivory Coast and all other countries of West Africa, through a review of available literature. From the synthesized literature, Brucella abortus was demonstrated to be the most prevalent species in cattle in West Africa, in line with the known host preference for Brucellae. So far, biovars 3 appeared to be commonly the most isolated in West Africa and was also recently identified in Ivory Coast. However, the presence of B. melitensis and/or B. suis was not reported yet in cattle in this part of Africa.

i

Results on prevailing strains of Brucella in cattle were related with commonly used serological diagnostic tools. Thus, chapter 5 was dedicated to verify their appropriateness and to assess the performance of two serological tests, Rose Bengal Test (RBT) and indirect Enzyme-linked Immunosorbent assay (iELISA). Using a Bayesian approach, the correlationadjusted sensitivity of iELISA was estimated at 96.1 % (Credibility Interval (CrI): 92.7-99.8) whereas that of RBT was 54.9 % (CrI: 23.5-95.1). High correlation-adjusted specificities were found for both tests, 95.0 % (CrI: 91.1-99.6) for iELISA and 97.7 % (CrI: 95.3-99.4) for RBT, respectively. The true prevalence of brucellosis was also estimated using the 1228 tested serum samples to be 4.6 with a 95% credibility interval ranging from 0.6 to 9.5% (Chapter 5 and 6). These results also revealed a good performance for the iELISA, which might consequently be a valuable screening assay under the epidemiological conditions prevailing in Ivory Coast. In Chapter 7, risk factors associated with bovine brucellosis seropositivity were investigated using serological results obtained from 907 serum samples collected from unvaccinated cattle of at least 6 months of age in the savannah-forest region of Ivory Coast. Serum samples were tested using the Rose Bengal test (RBT) and indirect enzyme linked immunosorbent assay (iELISA). The logistic regression analysis indicated that brucellosis seropositivity was associated with age and herd size. Cattle above 5 years of age were found to be more likely seropositive (Odd Ratio (OR) =2.8; 95% Confidence Interval (CI): 1.3, 6.4) compared to cattle under 3 years of age. Similarly, the odds of brucellosis seropositivity for herds with more than 100 cattle was 3.3 (95% CI: 1.2, 8.9) times higher compared to those with less than 50 cattle. The third part presents a general discussion on the overall contribution of the current research (Chapter 8), by highlighting the main results and pointing out their significance. The need for more investigations on the epidemiology of brucellosis, in Ivory Coast and at West African scale, is highlighted. It is neccessary to provide additional knowledge on prevailing field strains of Brucella, on the distribution of the disease and on associated risk factors to implement preventive and control measures. Finally, for more cost-effectiveness and efficiency, the need to strengthen the capabilities of the veterinary services and national laboratories and to consider the control of brucellosis and other zoonotic diseases through a regional, integrated and collaborative perspective is also highlighted.

ii

RESUME La brucellose bovine est une maladie infectieuse endémique qui peut impacter négativement la productivité et le bien-être des bovins ainsi que sur la santé humaine. Dans de nombreux pays en voie de développement tels que la Côte d'Ivoire, les connaissances sur la distribution et la fréquence de la maladie (ou les preuves de sa présence) dans la population animale ainsi que sur les facteurs de risque associés à la maladie restent limitées. La disponibilité d’informations sur les espèces et biovars de Brucella à l'échelle nationale et/ou régionale, sur la performance des tests de diagnostic communément utilisés pour déterminer la prévalence réelle de la maladie, et sur les facteurs de risque est également essentielle. Toutes ces informations sont d'une importance clé pour la définition et la mise en œuvre de mesures de prévention et de contrôle appropriées et efficaces contre la brucellose. Cette thèse vise donc à contribuer à une meilleure connaissance et compréhension de l'épidémiologie de la brucellose bovine en Côte d'Ivoire. Elle est organisée en trois parties principales. La partie introductive comprend trois chapitres. Le premier chapitre présente une revue de la littérature sur les agents pathogènes responsables de la brucellose, sur leurs caractéristiques et sur leur distribution. L'impact ainsi que les stratégies existantes de prévention et de contrôle de la brucellose sont également discutés avec des références à la situation particulière de la brucellose bovine en Côte d'Ivoire. La présence et l'importance de la maladie ont été confirmées dans ce pays, mais elle y reste toujours incontrôlée. Le deuxième chapitre (Chapitre 2) comprend un aperçu des principes et concepts épidémiologiques et des statistiques appliqués dans le cadre de cette thèse pour atteindre les différents objectifs présentés au chapitre trois (chapitre 3). Les différentes approches méthodologiques disponibles pour l'estimation des caractéristiques des tests de diagnostic ainsi que de la prévalence réelle d'une maladie sont aussi discutées. La deuxième partie de cette thèse présente, dans un enchainement logique, les recherches effectuées sur différents aspects de l'épidémiologie de la brucellose bovine en Côte d'Ivoire, avec des références à la situation en Afrique de l'Ouest (Chapitre 4, 5, 6 et 7). Le chapitre 4 présente, à travers une revue de la littérature disponible, un état des lieux des connaissances sur les espèces et biovars de Brucella signalés chez les bovins de Côte d'Ivoire mais aussi chez ceux de tous les autres pays de l'Afrique de l'Ouest. Il en ressort que Brucella abortus a été l’espèce la plus répandue chez les bovins en Afrique de l'Ouest, en conformité avec la préférence d’hôte connue pour les Brucella. A ce jour, le biovar 3 semble être le plus généralement isolé dans les pays d’Afrique de l'Ouest y compris en Côte d'Ivoire où il a été

iii

récemment identifié pour la première fois. Cependant, la présence de B. melitensis et/ou de B. suis n'a pas encore été signalée chez les bovins dans cette partie de l'Afrique. Les résultats sur les souches dominantes de Brucella chez les bovins ont été mis en relation avec des outils de diagnostic sérologique couramment utilisés. Ainsi, le chapitre 5 a été consacré à vérifier leur pertinence et à évaluer la performance de deux tests sérologiques, le test de Rose Bengale (TRB) et l’ELISA indirect (iELISA). En utilisant une approche bayésienne, la sensibilité de iELISA ajusté en prenant en compte la corrélation entre les deux tests, a été estimée à 96,1% (intervalle de crédibilité (ICr): 92,7-99,8), tandis que celle de TRB était de 54,9% (ICr: 23,5-95,1). De hautes valeurs de spécificités ont été trouvées pour les deux tests, respectivement 95,0% (ICr: 91,1-99,6) pour l’iELISA et 97,7% (ICr: 95,399,4) pour le TRB. La prévalence réelle de la brucellose a également été estimée à 4,6% avec un intervalle de crédibilté à 95% entre 0,6 et 9,5% sur la base de 1228 sérums analysés (Chapitres 5 et 6). Ces résultats ont mis en évidence une bonne performance pour l’iELISA, qui pourrait être par conséquent un test de dépistage précieux dans les conditions épidémiologiques de la Côte d'Ivoire. Dans le chapitre 7, les facteurs de risque associés à la séropositivité de brucellose bovine ont été étudiés sur base des résultats sérologiques obtenus de 907 échantillons de sérum prélevés chez des bovins non vaccinés d'au moins 6 mois dans la région intermédiaire entre la savane et la forêt, au centre de la Côte d'Ivoire. Les sérums ont été testés en utilisant le TRB et l’iELISA. L'analyse de régression logistique a indiqué que la séropositivité de la brucellose était associée à l'âge des animaux et à la taille du troupeau d’origine. Les bovins de plus de 5 ans présentaient une plus grande probabilité d'être séropositif (Odd Ratio (OR) = 2,8 (Intervalle de confiance (IC) 95%: 1,3-6,4)) par rapport ceux de moins de 3 ans. De même, la côte de séropositivité à la brucellose pour les troupeaux de plus de 100 bovins était 3,3 (IC 95%: 1,2-8,9) fois plus élevée par rapport à ceux de moins de 50 bovins. La troisième partie de cette thèse présente une discussion générale sur la contribution globale de cette recherche (Chapitre 8). La nécessité d’entreprendre plus d’études sur l'épidémiologie de la brucellose, en Côte d'Ivoire et en Afrique de l'Ouest, a été soulignée. Il est nécessaire de fournir des connaissances supplémentaires sur les souches circulantes de Brucella, sur la distribution de la maladie et sur les facteurs de risque associés pour la prise de mesures de prévention et de contrôle appropriés. Enfin, pour un meilleur rapport coûtefficacité, il est également nécessaire de renforcer les capacités des services vétérinaires et des laboratoires nationaux et d’appréhender la lutte contre la brucellose et les autres maladies zoonotiques dans une perspective régionale, intégrée et concertée.

iv

RESUMEN La brucelosis bovina es una enfermedad infecciosa endémica que puede tener un impacto negativo en la productividad y en el bien estar de los bovinos, así como en la salud humana. En varios países en vías de desarrollo, como Costa de Marfil, son limitados el conocimiento sobre la distribución y la frecuencia de la enfermedad, pruebas de su presencia, en la población animal así como los factores de riesgo asociados. Son igualmente necesarios los conocimientos en las especies y biotipos de Brucella en el país y en la región, la validación de las pruebas de diagnóstico comúnmente utilizadas, y los factores de riesgo. Estas informaciones son de gran relevancia para la puesta en marcha de medidas de prevención o y para el control adecuado y eficaz contra la brucelosis. Esta tesis contribuye a mejorar el conocimiento y comprensión de la epidemiología de la brucelosis bovina en Costa de Marfil. Está estructurada en tres partes principales. La parte introductoria comprende tres capítulos. El primer capítulo presenta un acercamiento a la literatura sobre los agentes patógenos responsables de la brucelosis, sus características y su distribución. Se discute el impacto y las estrategias existentes para prevenir y combatir la brucelosis con referencias a la situación particular de la brucelosis bovina en Costa de Marfil. La presencia y la importancia de la enfermedad han sido confirmadas en dicho país, sin embargo, la enfermedad es aún incontrolada. En el segundo capítulo, capítulo dos, se muestra los principios y conceptos epidemiológicos y estadísticos aplicados en el marco de la presente tesis para lograr los diferentes objetivos, visto en el capítulo tres. Son también discutidos las diferentes aproximaciones metodológicas disponibles para la estimación de las características de las pruebas de diagnóstico, así como, de la prevalencia real de una enfermedad. Basados en investigaciones realizadas, la segunda parte de esta tesis relaciona diferentes aspectos epidemiológicos de la brucelosis bovina en Costa de Marfil y en África del Oeste, presente en los capítulos cuatro, cinco, seis y siete. El capítulo cuatro muestra el estado serológico de los biotipos de Brucella en los bovinos de Costa de Marfil, y de varios países de África del Oeste a través de la literatura disponible. Así, la Brucella abortus demostró ser la especie más extendida en los bovinos de África del Oeste, de acuerdo a la preferencia de hospedador conocido por la Brucella. Actualmente, el biotipo 3 es el más aislado en los países de África del Oeste, incluyendo a Costa de Marfil, país confirmado por primera vez. Sin embargo, la presencia de Brucella melitensis o/y Brucella suis no ha sido aún encontrada en los bovinos de esta parte de África.

v

Los resultados de las cepas de Brucella de los bovinos se relacionaron con las herramientas de diagnóstico serológico comúnmente usadas. De esa manera, el capítulo cinco verifica su pertinencia y evalúa su desempeño en dos pruebas serológicas, la prueba de Rosa de Bengala (RBT) y la de ELISA indirecto (iELISA). Haciendo uso de un enfoque bayesiano, la sensibilidad de iELISA ajustada a la correlación entre dos pruebas, fue de 96,1% (Interval de credibilidad (ICr): 92,7-99,8), mientras que aquella para RBT fue 54,9% (ICr: 23,5-95,1). Así mismo, fueron encontrados fuertes valores de especificidad para las dos pruebas, siendo para iELISA de 95,0% (ICr: 91,1-99,6) y para RBT de 97,7% (ICr; 95,3-99,4). A partir de los 1228 sueros ensayados, la prevalencia real de la brucelosis fue estimada a 4,6% con un interval de credibilidad de 95% entre el 0,6 y el 9,5% (Capítulo cinco y seis). Estos resusltados muestran el buen desempeño de la prueba iELISA para ser usado en las pruebas de despistaje epidemiológico en Costa de Marfil. En el capítulo 7, son evaluados los factores de riesgo asociados con la seropositividad de brucelosis bovina mediante resultados serológicos obtenidos a partir de 907 muestras de suero tomadas de bovinos no vacunados de por lo menos 6 meses de edad en la región límite entre la sabana y la selva, en el centro de Costa de Marfil. Las muestras de suero se analizaron mediante la prueba RBT e iELISA. El análisis de regresión logística indicó que la seropositividad de la brucelosis se asoció con la edad del animal y el tamaño del rebaño. Los bovinos de más de cinco años presentan 2,8 veces más de riesgo (Interval de confidencia (IC) 95%: 1,3 - 6,4), que aquellos de más de tres años de edad. Por otro lado, el riesgo para la brucelosis en hatos de más de cien cabezas de ganado es de 3,3 veces mayor (IC 95%: 1,2 8,9) que aquellos con menos de 50 vacas. La tercera parte, vista en el capítulo ocho de la presente tesis, muestra una discusión general sobre la contribución global de la investigación actual poniendo en evidencia los principales resultados y señalando su importancia. En tal sentido, se hacen necesarios mayores trabajos epidemiológicos sobre la brucelosis tanto en Costa de Marfil como en África del Oeste, a fin de proveer conocimientos adicionales y suficientes relativos al origien de la Brucella circulante, la distribución de la enfermedad y los factores de riesgo asociados orientados a las medidas de prevención o y de control. Para una mejor relación entre el costo y la eficacia, se hace necesaria fortalecer las capacidades de los servicios veterinarios y de los laboratorios nacionales, así como, de concebir la lucha contra la brucelosis y las otras zoonosis sobre una perspectiva regional, integrada y colaborativa.

vi

ACKNOWLEDGEMENTS While completing this thesis, I would like to present my sincere acknowledgements to all the nice, special, grateful, supporting people who assisted me all along this “long journey”. First, I offer my special gratitude to my supervisors Prof. Claude SAEGERMAN and Prof. Dirk BERKVENS for their support during all the process of this PhD thesis. To Prof. SAEGERMAN, I would like to thank him for his motivation, for his confidence in me, for his availability, his pieces of advice, and his suggestions and for all the administrative and scientific works accomplished during this thesis. I really appreciate being your student and thank for pushing me until here. My sincere thanks go next to Prof. BERKVENS, for believing in my capability to contribute to science, for your inputs and all your support since the beginning of this “journey”. Thanks to you, for initiating me to Bayesian philosophy and statistics. Beside my promoters, my sincere thanks go to the members of my doctoral committee, Dr. David FRETIN and Dr. Eric THYS for their valuable scientific comments and inputs, for their encouragement and moral support. I am especially deeply grateful to Dr. Eric THYS, for coaching me during my master program and putting me on the trail for the PhD. Thanks for showing me how to do things, for being available all along these years, for your reactivity and your quick feedbacks and for all your support. I would also like to express my special gratitude to Dr. Emmanuel Nji ABATIH, for the support, for the assistance in data analysis, for cross-checking everything, for all the valuable inputs. Thanks for being always available for me. Specials thanks also to the members of my jury, Prof. Laurent GILLET (President of the Jury), Prof. Niko SPEYBROECK (UCL), Dr. Hein IMBERECHTS (CERVA), Prof. Johan DETILLEUX (ULg), Prof. Annick LINDEN (ULg), Prof. Jean-Luc HORNICK (ULg), Prof. Frédéric ROLLIN (ULg) , for evaluating of this works. My sincere thanks also go the staff of the bacteriology and immunology department of the CERVA in Brussels and especially to Martine MARIN, Christel DESMETS, Patrick MICHEL,

Sylvie

MALBRECQ,

Philippe

VANNOORENBERGHE,

and

Damien

vii

DESQUEPER for assisting in laboratory works. I am grateful to Dr Karl WALRAVENS for opening me the doors of “the BSL3 lab” and to Dr. David FRETIN, his successor for adopting me and for accepting me in their team. I also thank Dr. Mark GOVAERT for his advice while I was at the CERVA. My sincere gratitude also goes to the Belgian Directorate-General for Development Cooperation and Humanitarian Aid (DGD), to the University of Liège (ULg), and to the Institute of Tropical Medicine (ITM) for the financial support during the completion of this thesis. Thank you to all the staff of the Biomedical Sciences Department (former Animal Health Department) of the Institute of Tropical Medicine, especially to Redgi DE DEKEN, Danielle DEBOIS and Nadia EHLINGER. Thank you also to all the members of the Research Unit of Epidemiology and Risk Analysis Applied to Veterinary Sciences (UREAR-ULg). My gratitude also goes to the Laboratoire National d’Appui au Développement Agricole (LANADA), my employer, for all the support and for allowing the completion of this PhD program. I am also grateful to the veterinary services in Ivory Coast for the support during the field works. Special thanks to my fellow PhD students, Gilbert AKODA, Boukary A RAZAC, Mahamat OUAGAL, and Nicolas PRAET, for all their support, encouragement and pieces of advice. Thank you to my colleagues at LANADA, Prof. Ardjouma DEMBELE, Dr. Vazin DEA, Dr.Ahoua TANO, Dr. Yolande AKE-ASSI, Ir. Thérèse ANOMAN ADJO, Dr. Clarisse KOMOIN-OKA, Dr. Alassane TOURÉ, Dr. Affourmou KOUAMÉ, Dr. M’Bétiégué COULIBALY, Prof. Emmanuel COUACY-HYMANN, Dr Thérèse DANHO, Dr. Biego Guillaume GRAGNON, Dr. Félicité BEUDJE, Dr. Seidinan Ibrahim TRAORÉ, and Dr. Marie-Pélagie GBAMELÉ. Special thanks to Dr. Louise Y. ACHI ATSE, for her support and her encouragement since the beginning of this thesis. I would like to acknowledge all the people who participated or facilitated the field works, especially animal owner and local animal health agents especially to Dr KONAN BANNY Jean Pierre, Dr TRAORÉ Adama, Oussou N’GUESSAN, Mamery BAMBA, Alfred ASSI ASSI, Mariana E. TRAORÉ, Kouadio KONAN (“Doyen Konan”), Eloi K. ADJÉ, Gnenema BAMBA My gratitude also goes to all the special and nice people I met during my PhD and during my stays in Belgium, Alexis KOFFI DOUA, Olivette Fanny KANTORE, Evelyne DEVAUD,

viii

Isabel GOMEZ DIEZ, Delphine CARRARA, Caroline HAROUN, Jordan KYONGO, Samuel ODIWORDS, Rafael MANZANEDO, Bezeid OULD EL MAMY, Esther KUKIELKA, Juana BIANCHINI, Fabiana DAL POZZO, Ludovic MARTINELLE, Marie-France HUMBLET and all my Master program fellows. Thank you all for being there when I needed. Thanks you to my friends (all of you) and advisors for the moral and spiritual support: Kabiné DIAKITÉ, Elhadj Dr Yahya KARAMOKO, Dr Bakary CISSÉ, and Dr Vessally KALLO, Dr DIBY Konan Jean-Paul... It was so helpful to have you close. Many thanks for my family, my dad (May he rest in peace), my lovely mother, my brothers, my sisters, my lovely wife and my children. Thank you all!

ix

"[...] Great things can happen to ordinary people if we work hard and never give up." Barak Obama

"You have the responsibility to persist until you succeed" T. Newberry

"[...] Making the most of what we have" T. Newberry

x

TABLE OF CONTENTS ABSTRACT ............................................................................................................................................................I ABSTRACT ........................................................................................................................................................ I RESUME ........................................................................................................................................................... III RESUMEN ......................................................................................................................................................... V ACKNOWLEDGEMENTS .............................................................................................................................. VII TABLE OF CONTENTS .................................................................................................................................... XI LIST OF ABBREVIATIONS .......................................................................................................................... XIV LIST OF TABLES .......................................................................................................................................... XVII LIST OF FIGURES ....................................................................................................................................... XVIII LIST OF ANNEXES ........................................................................................................................................ XIX GENERAL INTRODUCTION ............................................................................................................................ 1 PART ONE: ......................................................................................................................................................... 10 LITERATURE REVIEW ................................................................................................................................... 10 CHAPTER 1: BRUCELLOSIS: ETIOLOGY, IMPORTANCE, PREVENTION AND CONTROL WITH SPECIAL REFERENCES TO THE SITUATION OF THE DISEASE IN CATTLE IN IVORY COAST 12 1.1.

ETIOLOGICAL AGENTS OF BRUCELLOSIS .............................................................................................. 12

1.1.1.

BRIEF HISTORY ON BRUCELLA ............................................................................................................. 12

1.1.2.

TAXONOMY, DESCRIPTION AND CHARACTERISTICS OF BRUCELLA ....................................................... 13

1.2.

IMPORTANCE AND DISTRIBUTION OF BRUCELLOSIS ............................................................................. 17

1.2.1.

SOCIO-ECONOMIC AND PUBLIC HEALTH IMPORTANCE ......................................................................... 17

1.2.2.

DISTRIBUTION OF BRUCELLOSIS .......................................................................................................... 19

1.2.2.1.

DISEASE IN ANIMALS ...................................................................................................................... 19

1.2.2.2.

DISEASE IN HUMAN ........................................................................................................................ 21

1.3.

PREVENTION AND CONTROL OF BRUCELLOSIS ..................................................................................... 23

1.3.1.

DIAGNOSTIC TOOLS ............................................................................................................................. 23

1.3.1.1.

DIRECT DIAGNOSTIC METHODS....................................................................................................... 24

1.3.1.1.1.

BACTERIOLOGICAL EXAMINATION............................................................................................. 24

1.3.1.1.2.

MOLECULAR METHODS.............................................................................................................. 26

1.3.1.1.3.

MALDI-TOF MASS SPECTROMETRY (MALDI-TOF-MS) ......................................................... 26

1.3.1.2.

INDIRECT DIAGNOSTIC METHODS.................................................................................................... 27

1.3.1.2.1.

SEROLOGICAL METHODS............................................................................................................ 27

1.3.1.2.1.1.

SERO-AGGLUTINATION OF WRIGHT (SAW) ............................................................................... 28

1.3.1.2.1.2.

ROSE BENGAL TEST (RBT) ....................................................................................................... 29

1.3.1.2.1.3.

COMPLEMENT FIXATION TEST (CFT) ........................................................................................ 29

1.3.1.2.1.4.

ENZYME LINKED IMMUNOSORBENT ASSAYS ............................................................................. 29

1.3.1.2.1.5.

FLUORESCENCE POLARIZATION ASSAY (FPA) .......................................................................... 30

xi

1.3.1.2.1.6.

BRUCELLA IMMUNOCHROMATOGRAPHIC LATERAL FLOW ASSAYS (LFA) ................................. 30

1.3.1.2.1.7.

MILK TESTING ASSAYS .............................................................................................................. 31

1.3.1.2.2.

CELLULAR METHODS ................................................................................................................. 31

1.3.1.2.2.1.

SKIN DELAYED-TYPE HYPERSENSIBILITY TEST OR SKIN TEST .................................................. 31

1.3.1.2.2.2.

ANTIGEN-SPECIFIC GAMMA INTERFERON PRODUCTION TEST ................................................... 31

1.3.1.3. 1.3.2.

IDENTIFICATION AND TYPING METHODS ......................................................................................... 32 PREVENTION AND CONTROL MEASURES .............................................................................................. 32

1.3.2.1.

COMMUNICATION AND EDUCATION FOR PREVENTION .................................................................... 35

1.3.2.2.

PROPHYLACTIC MEASURES ............................................................................................................. 35

1.3.2.3.

INTERSECTORAL COLLABORATION FOR CONTROL .......................................................................... 36

CHAPTER 2: EPIDEMIOLOGICAL CONCEPTS AND METHODOLOGIES ......................................... 39 2.1.

SYSTEMATIC REVIEW AND META-ANALYSIS ........................................................................................ 39

2.2.

LOGISTIC REGRESSION ANALYSIS ........................................................................................................ 39

2.3.

DETERMINATION OF DISEASE STATUS ................................................................................................. 40

2.4.

ESTIMATION OF DISEASE TRUE PREVALENCE AND PERFORMANCE OF DIAGNOSTIC TESTS ................... 41

2.4.1.

ESTIMATION OF DISEASE TRUE PREVALENCE ....................................................................................... 41

2.4.2.

ASSESSMENT OF PERFORMANCE OF DIAGNOSTIC TESTS....................................................................... 42

2.4.2.1.

INDICATORS OF AGREEMENT BETWEEN TESTS ................................................................................ 42

2.4.2.2.

PERFORMANCE PARAMETERS OF DIAGNOSTIC TESTS ...................................................................... 43

2.4.3.

METHODS FOR ESTIMATING DISEASE TRUE PREVALENCE AND TEST SENSITIVITY AND SPECIFICITY..... 46

2.4.3.1.

ESTIMATION AT INDIVIDUAL-LEVEL ............................................................................................... 46

2.4.3.2.

ESTIMATION AT HERD-LEVEL ......................................................................................................... 47

2.4.3.3.

BAYESIAN VERSUS FREQUENTIST METHODS FOR ESTIMATING THE DISEASE TRUE PREVALENCE AND

DIAGNOSTIC TEST PERFORMANCE ...................................................................................................................... 48

CHAPTER 3: OBJECTIVES OF THE THESIS .............................................................................................. 54 PART TWO: ........................................................................................................................................................ 57 EXPERIMENTAL SECTION............................................................................................................................ 57 CHAPTER 4: ....................................................................................................................................................... 58 DETERMINATION OF SPECIES AND BIOVARS OF BRUCELLA INFECTING CATTLE POPULATION IN IVORY COAST .................................................................................................................. 58 4.1.

INTRODUCTION ................................................................................................................................... 59

4.2.

PREVAILING SPECIES AND BIOVARS OF BRUCELLA IN CATTLE AND THEIR IMPLICATIONS ..................... 59

CHAPTER 5: ....................................................................................................................................................... 70 PERFORMANCE OF DIAGNOSTIC TESTS FOR BOVINE BRUCELLOSIS IN IVORIAN EPIDEMIOLOGICAL CONTEXT ................................................................................................................... 70 5.1.

INTRODUCTION ................................................................................................................................... 71

5.2.

BAYESIAN ESTIMATION OF TRUE PREVALENCE, SENSITIVITY AND SPECIFICITY OF ROSE BENGAL TEST

AND INDIRECT ELISA FOR THE DIAGNOSIS OF BOVINE BRUCELLOSIS IN IVORY COAST ..................................... 71

xii

CHAPTER 6: ....................................................................................................................................................... 80 TRUE PREVALENCE OF BOVINE BRUCELLOSIS IN IVORY COAST ................................................. 80 6.1.

INTRODUCTION ................................................................................................................................... 81

6.2.

PREVALENCE OF BOVINE BRUCELLOSIS IN IVORY COAST .................................................................... 81

CHAPTER 7: ....................................................................................................................................................... 87 RISK FACTORS ASSOCIATED WITH BOVINE BRUCELLOSIS SEROPOSITIVITY IN IVORY COAST ................................................................................................................................................................. 87 7.1.

INTRODUCTION ................................................................................................................................... 88

7.2.

RISK FACTORS ASSOCIATED WITH BRUCELLOSIS SEROPOSITIVITY AMONG CATTLE IN THE CENTRAL

SAVANNAH-FOREST AREA OF IVORY COAST ...................................................................................................... 88

PART THREE: .................................................................................................................................................... 95 GENERAL DISCUSSION .................................................................................................................................. 95 CHAPTER 8: GENERAL DISCUSSION, CONCLUSIONS AND PERSPECTIVES ................................. 97 8.1.

GENERAL DISCUSSION ......................................................................................................................... 97

8.1.1.

IMPORTANCE OF IDENTIFICATION AND TYPING OF PREVAILING STRAINS OF BRUCELLA IN CATTLE ...... 98

8.1.2.

TRUE PREVALENCE, SENSITIVITY AND SPECIFICITY OF SEROLOGICAL ASSAYS FOR THE DIAGNOSTIC OF

BRUCELLOSIS IN IVORY COAST ........................................................................................................................ 103

8.1.3.

RISK FACTORS ASSOCIATED TO THE SEROPOSITIVITY OF BRUCELLOSIS IN CATTLE FROM IVORY COAST . .......................................................................................................................................................... 105

8.2.

CONCLUSIONS, IMPLICATIONS AND PERSPECTIVES ............................................................................ 106

REFERENCES .................................................................................................................................................. 112 ANNEXES .......................................................................................................................................................... 144

xiii

LIST OF ABBREVIATIONS °C AMOS ANADER AP AUC Bayesp BK BSL3 CERVA CFT CFU CI CO2 CrI DIC DNA DOR EDTA ECOWAS FAO FAOSTAT FCFA FN FP FPA FPSR GDP H2S H38 HSe HSp iELISA IFAD Ig ILRI IMT J K LANADA LFA LID

Degree Celsius Abortus-Melitensis-Ovis-Suis Agence Nationale d’Appui au Développement Rural (National Agency for rural development) Apparent Prevalence Area Under the Receiver-Operating-Characteristic (ROC) Curve Bayesian P-value Berkeley Biosafety Level 3 Centre d’Etude et de Recherches Vétérinaires et Agro-chimiques Complement Fixation Test Colony Forming Unit Confidence Interval Carbon dioxide Credibility Interval Deviation Information Criterion Deoxyribonucleic Acid Diagnostic Odd Ratio Ethylene Diamine Tetraacetic Acid Ecomonic Community of West African States Food and Agriculture Organisation Food and Agriculture Organization Statistics Division CFA Franc (West African currency: 1 EURO= 655.957 FCFA) False negative False Positive Fluorescence Polarization Assay False Positive Serological Reaction Gross Domestic Product Dihydrogen sulfide Vaccine strain H38(killed B. melitenis strain 53H38) Herd Sensitivity Herd specificity Indirect Enzyme Linked Immunosorbent Assay International Fund for Agricultural Development Immunoglobulin International Livestock Research Institute Intsitute of Tropical Medicine Youden index Kappa coefficient Laboratoire National d’Appui au Développement Agricole Lateral Flow Assays Livestock in Development

xiv

LPS LR MCMC MIPARH MLSA MLST MLVA MRT Neg NPV OIE OPS OR P Pc PLS PNDL Pneg Po Pos Ppos PPV PSE PVS R R/C RB51 RBT RLPS RTD S S19 SAW Se SLPS SNP SODEPRA Sp Tb TN TP ULg USD USDA VNTR

Lipopolysaccharide Likelihood Ratio Markov Chain Monte Carlo Ministry of animal and fishery production and resources Multilocus Sequence Analysis Multilocus Sequence Typing Multiple Loci Variable number tandem repeat Analysis Milk Ring Test Negative test result Negative Predictive Value World Animal Health Organization O Polysaccarhide Odd Ratio True prevalence Proportion of observed agreement Projet laitier Sud Programme national de Développement Laitier Index of negative agreement Proportion of agreement due to chance Positive test result Index of positive agreement Positive Predictive Value Programme Sectoriel de l’Elevage, Performance of Veterinary Services Rough A brucellaphage active on rough Brucella Vaccine RB51 Rose Bengal Test Rough Lipopolysccharide Routine Test Dilution Smooth Vaccine strain 19 Slow Agglutination of Wright Sensitivity Smooth Lipopolysaccharide Single Nucleotide Polymorphism Société pour le Développement de la Production Animale Specificity Tbilisi True Negative True positive University of Liège United States Dollar (1 USD= 0.788454 EUROS) United States Department of Agriculture Variable Number Tandem Repeat

xv

Wb WGS WHO γ-IFN

Weybridge Whole Genome Sequencing World Health Organization Gamma interferon

xvi

LIST OF TABLES Table I: Nomenclature and characteristics of Brucella species (from Pappas et al., 2005; Whatmore, 2009; Whatmore et al., 2014) ................................................................................ 16 Table II: Studies on prevalence of bovine brucellosis in Ivory Coast, 1970-2008 .................. 20 Table III: Characteristics for differentiating the different Brucella (from OIE, 2009) ............ 25 Table IV: Types of antibodies detected by conventional serological assays for the diagnostic of brucellosis (from Quinn et al., 1999 and Saegerman, 2004)................................................ 28 Table V : Control strategies of brucellosis according to the epidemiological status (adapted from Benkirane, 2001; Saegerman et al., 2010) ....................................................................... 34 Table VI : Contingency table showing results for two diagnostic tests (Test 1 and Test 2) .... 43 Table VII: Contingency table showing results between a reference test and a given imperfect test (Test) .................................................................................................................................. 45 Table VIII: Definitions of commonly used performance indicators of diagnostic test (Glas et al., 2003) ................................................................................................................................... 45 Table IX: The Multiple Loci Variable Number Tandem Repeats analysis (MLVA) profiles showing number of variable tandem repeats (VTR) for latest west African isolates of B. abortus biovar 3 and their closest MLVA neighbour profile (B. abortus biovar 3 strain BCCN 93_26 from in Sudan, B abortus biovar 3 reference strain Tulya from Uganda and B. abortus biovar 6 strain BfR7 from Chad) in the Brucella MLVAbank (from Bankole et al., 2010, Sanogo et al., 2013a and Boukary et al., 2013, Dean et al., 2014) .......................................... 99

xvii

LIST OF FIGURES Figure 1: Map showing the density of poor people keeping livestock in Africa, 2005 (ILRI, 2012) ........................................................................................................................................... 6 Figure 2: Administrative map of Ivory Coast (Source: http://www.nationsonline.org/oneworld/map/cote-dIvoire-administrative-map.htm) ............... 7 Figure 3: Map showing the agro-ecological areas of Ivory Coast .............................................. 8 Figure 4: Cross-border transhumance routes in West Africa and Central Africa (adapted from OECD, 2008) .............................................................................................................................. 9 Figure 5 : Brucella colonies on a solid culture media, showing translucent honey-colored appearance (credit picture: M. Sanogo) .................................................................................... 13 Figure 6: Dispersion of Brucella species according to their preferred host mammal. The dispersion of the various Brucella species is depicted as cones proportional to the number of strains analyzed (adapted from Moreno, 2014). ....................................................................... 15 Figure 7: Main transmission routes of brucellosis from livestock to humans (by Sir David Bruce, 1855-1931) (from http://m2002.tripod.com/brucellosis.jpg) ........................................ 21 Figure 8: Global Incidence of Human Brucellosis (from Pappas et al., 2006) ....................... 22 Figure 9: Schematic summary of the main objectives of the thesis ......................................... 56 Figure 10: Dendrgram showing the relation between the latest isolates of B. abortus biovar 3 in West Africa and also with neighbour reference strains in the Brucella MLVAbank (B. abortus biovar 3 strain BCCN 93_26 from in Sudan, B abortus biovar 3 reference strain Tulya from Uganda and B. abortus biovar 6 strain BfR7 from Chad) (Bankole et al., 2010; Sanogo et al., 2013a; Boukary et al., 2013, Dean et al., 2014). It is build from results of a simple linkage cluster analysis of the number of variable tandem repeats (VNTR) and the dissimilarities between strains is measured through the eucludian distance between VNTRs (L2 dissimilarity measure). ..................................................................................................... 100 Figure 11: Mapping of field strains of Brucella in cattle in Ivory Coast, 2013 (from PiloMoron et al., 1979; Sanogo et al., 2013a; Sanogo et al., 2013b). Each bubble contains information on the name of the locality of origin of the strain (e.g. Eloka), the year of publication ( e.g. 1979), the biovar (e.g. B. abortus 1) and the number of isolates identified (eg. n=2). ................................................................................................................................ 101

xviii

LIST OF ANNEXES Annex 1: Map showing the western part of Africa and the neighbour countries of Ivory Coast ................................................................................................................................................ 145 Annex 2: Differential characteristics of biovars of Brucella species (from OIE, 2009) ........ 146 Annex 3 : A cow with a carpal hygroma ................................................................................ 147

xix

GENERAL INTRODUCTION

1

General introduction

The majority of the world’s poor population (about 75%) works and lives in rural areas. About 600 million of the 1.3 billion of the poor worldwide keep livestock as means to produce food and generate cash income (IFAD, 2001; Thornton et al., 2002; ILRI, 2012). Hence, livestock is of key importance in people's everyday lives in most of the sub-Saharan African countries where a quarter of the world’s poor come from (Figure 1). Livestock contributes to their financial security, their food security and to the development of their agriculture through animal traction and manure (Starkey, 2010; FAO, 2011). The development of livestock production and its productivity are therefore part of the solution for food security and poverty alleviation, especially in low-income areas. Consequently, there is a need to tackle the different constraints to this development, especially the pathological ones. With its negative impact on animal health and productivity, and its threat to human health, brucellosis is one of the pathological constrains to be considered. On a global basis, this disease is among the thirteen animal diseases and syndromes identified as having a significant impact on poor people worldwide and in West Africa (Perry, 2002). Brucellosis is a bacterial infectious disease affecting domestic, wild animals and humans (Maurin et al., 2005). It is one of the most widespread bacterial zoonotic diseases (Corbel, 2006). According to the World Health Organization (WHO), about 500,000 new cases of human brucellosis are reported annually worldwide (Corbel, 1997, Pappas et al., 2006). In animals, brucellosis is responsible for many economic losses because of abortions, decrease in production (particularly reduced milk production), newborn mortality, reproductive disorders, and costs of intervention. With its impact on productivity, it contributes to worsen the deficit of animal protein especially for populations in developing countries, where the needs are continuously increasing. In areas where people’s livelihood heavily depends on livestock, the impact of brucellosis might also exacerbate poverty (Cáceres, 2010). The most common and widespread form of the disease in animals is bovine brucellosis (Akakpo and Bornarel, 1987; Corbel, 1997; McDermott and Arimi, 2002; Bronvoort et al., 2009). Therefore, it is the main concern in sub-Saharan African countries (McDermott and Arimi, 2002) where average prevalence rates ranging between 10.2 and 25.7% were reported (Mangen et al., 2002). In West Africa, the disease (or evidence of its presence) was reported in 12 out of the 14 countries so far (Mangen et al., 2002; Boukary,

2

General introduction

2013) with higher seroprevalences estimated in Senegal, Togo, Mali, Niger, Burkina Faso and Ivory Coast (Mangen et al., 2002). In Ivory Coast1, the disease is considered as one of the dominant pathologies affecting livestock productivity, with negative impact on livestock breeders’ financial security and annual income (Angba et al., 1987; Mangen et al., 2002). Located between 3° to 9° Longitude West and 5° to 11° Latitude North, Ivory Coast is a West African country of 322 462 kilometers of square (Km²) (Annex 1). It is surrounded by Mali and Burkina Faso in the North, Ghana in the East and Guinea and Liberia in the West (Figure 2). Its population is about 21 millions of inhabitants, of which almost a half live in rural areas (FAO, 2014a). Three main agro-ecological areas are encountered (Figure 3): The Guinean zone or forest area, in the south, is the most humid and covers almost the whole forest region with annual rainfall generally above 1,500 mm. The Soudano-Guinean zone (or savannah-forest area) is an area of transition between the forest zone and the north. In this area, annual rainfall varies between 1,200 and 1,500 mm. The Soudanean zone in the northern part of the country is the savannah region with rainfall ranging between 900 and 1,200 mm per year. According to the FAO, there are about 1.6 millions cattle, 3.3 millions small ruminants, and almost 353,000 pigs (FAO, 2014b). Most cattle herds are concentrated in the northern and central part of the country, which is more favorable for livestock breeding with around 85% of the country’s cattle population. These cattle are of four different breeds: The N’Dama, the Baoulé, and the Lagunaire which belongs to the humpless Bos taurus type and the Zebus of the humped Bos indicus type. There are also various crossbred animals (Bos taurus X Bos indicus). Conversely to B. Taurus breeds which are mostly raised in the sedentary system, Zebus and their crossbred are mostly associated with the transhumance2 or semi-transhumance system, with movement of cattle toward the central part or within the northern and central part of the countries. The extensive system is the dominant type of farming in the country.

1

The official name of the country is “Côte d’Ivoire”, but it is popularly named “Ivory Coast” in english. “Ivory Coast” will be used throughout this manuscript. 2 Transhumance is defined as an oscillating, seasonal movement of livestock under the care of herders, following precise routes in order to exploit pastoral resources. It is distinguished from nomadism, which is characterised by more random movements and is followed by the herder’s whole family (OECD, 2008). There are agreements for transhumance between member countries of the Economic Community of West African States (ECOWAS) since 1998 , allowing inter-states animal movements (Anomynous, 1998; Anonymous, 2003b)

3

General introduction

In Ivory Coast, livestock breeding is still a secondary activity compared to agriculture that is practiced by 2/3 of the whole population. According to available figures, livestock breeding accounts for only 4.5 % of the agricultural Gross Domestic Product (GDP) and 2% of the total GDP of the country (Anonymous, 2003a). As a result, the national coverage of needs in meat products (59%) and dairy products (18%) is still insufficient (Anonymous, 2003a). Therefore, many efforts are required to cover these needs. Meanwhile, the country is dependent on imports from neighbouring countries such as Mali and Burkina Faso. To reduce the magnitude of this dependence, many initiatives have been undertaken. Institutions, projects and programs were promoted for the development of livestock production and increased productivity of local breeds through genetic improvement (e.g., Société pour le Développement de la Production Animale (SODEPRA), Agence Nationale pour le Développement Rural (ANADER), Programme National de Développement Laitier (PNDL) , Programme Sectoriel de l’Elevage (PSE), Projet Laitier Sud (PLS) (Anonymous, 1997)). The sustainability of these initiatives also implies tackling the numerous animal diseases of food producing animals, including the endemic and zoonotic ones such as brucellosis. Controlling brucellosis efficiently requires good diagnostic tools and sufficient and reliable information on the epidemiology of the disease. Until now, different aspects of the disease have been investigated throughout years in Ivory Coast before this research (Gidel et al., 1974; Pilo-Moron et al., 1979; Camus, 1980a; Angba et al., 1987; Thys et al., 2005) since first evidences of brucellosis were reported in the 1970s (Böhnel, 1971; Pilo-Moron et al., 1979). However, available information is still scarce or outdated. Therefore, there is a need to update information on the epidemiology of brucellosis, especially on its distribution, the causes (which Brucella spp. are involved), and the factors favoring the spread of the disease. All these preliminary pieces of knowledge are necessary to understand the epidemiology of the disease and to elaborate future preventive and control programs for countries facing brucellosis in West Africa, including Ivory Coast. It is particularily important to consider the regional perspective, knowing the existence of frequent cattle movements between West African countries through transhumance or commercial exchanges (Figure 4). This thesis aims to improve the current knowledge on the status of brucellosis in Ivory Coast and it is structured in three main parts. The introduction part includes three chapters. In Chapter 1, an overview is presented on the disease-causing agents of brucellosis, their characteristics, and their distribution. In addition, the impact, the prevention and the control measures of brucellosis are presented with references to the situation of bovine brucellosis in 4

General introduction

Ivory Coast. The chapter 2 includes a review of statistical methodological approaches for accurate estimation of diagnostic test characteristics and true prevalence of disease. It also brings an insight on statistical, epidemiological principles and concepts applied to achieve the different objectives of the thesis, presented in Chapter 3. The second part of the thesis includes the research contribution to the different aspects of the epidemiology of bovine brucellosis in Ivory Coast and West Africa (Chapter 4, 5, 6 and 7). Finally, the last part presents a general discussion on the overall contribution of the thesis (Chapter 8).

5

General introduction

Figure 1: Map showing the density of poor people keeping livestock in Africa, 2005 (ILRI, 2012)

6

General introduction

Figure

2:

Administrative

map

of

Ivory

Coast

(Source:

http://www.nationsonline.org/oneworld/map/cote-dIvoire-administrative-map.htm)

7

General introduction

Figure 3: Map showing the agro-ecological areas of Ivory Coast

8

Figure 4: Cross-border transhumance routes in West Africa and Central Africa (adapted from OECD, 2008)

9

PART ONE: LITERATURE REVIEW

10

Chapter 1 : Brucellosis : a literarture review

CHAPTER 1: BRUCELLOSIS: ETIOLOGY, IMPORTANCE, PREVENTION AND CONTROL WITH SPECIAL REFERENCES TO THE SITUATION OF THE DISEASE IN CATTLE IN IVORY COAST

11

Chapter 1 : Brucellosis : a literarture review

CHAPTER 1: BRUCELLOSIS: ETIOLOGY, IMPORTANCE, PREVENTION AND CONTROL WITH SPECIAL REFERENCES TO THE SITUATION OF THE DISEASE IN CATTLE IN IVORY COAST 1.1. Etiological agents of brucellosis 1.1.1.

Brief history on Brucella

Brucellosis is an ancient disease. References to what is now known as brucellosis are argued to exist in the history up to about five century before Jesus Christ, with a description of a resembling condition by Hippocrates (Fernando et al., 2003; Cutler et al., 2005). During the 17th and 18th centuries, cases of a mysterious undulant fever were recorded in many areas all over the Mediterranean region, with different local names (e.g., Mediterranean fever, Rock fever, Gibraltar fever, Cyprus fever, Danube fever, Neapolitan fever, Crimean fever, Cartagena fever, Barcelonan fever, Corps disease, undulant fever). In 1859, Dr Jeffrey Alan Marston, an assistant surgeon of the British royal artillery on duty in the island of Malta, contracted a similar illness, also characterized by an undulant fever (Wyatt, 2013). By describing his own case, Dr Marston produced the first detailed clinical description of “Malta fever”. His illness was later associated with brucellosis, after another army surgeon, Captain David Bruce, identified the causal agent of this disease, a small bacterium (designated Micrococcus melitensis and later named Brucella melitenis), isolated from the liver of a British soldier who died from a similar disease (Bruce, 1887). In 1897, about 10 years after the works of Captain Bruce, Prof. Almroth Wright described the first serological diagnostic test for the disease, the sero-agglutination tube test. Meanwhile, a new bacterium designated Bacillus abortus was isolated from repetitive abortive cows by Bernhard Bang, a Danish veterinarian. The first relation between the disease in human and an animal source was made about fifty years after the first clinical description, in 1905, by Dr Themistocles Zammit who associated the disease in humans with unpasteurized goat milk (Zammit, 1905). In 1917, Alice Evans, an American bacteriologist related Bacillus abortus and Micrococcus melitensis and the two bacteria were grouped into a single genus designated Brucella, as a tribute to Captain David Bruce (Meyer and Show, 1920). In sub-saharan Africa, the first references to Malta fever were made in the early 1900s in Senegal and Mauritania (Bourret, 1910). However, evidence of the disease in animals started to be reported in the 1930s especially in West African countries (Akakpo and

12

Chapter 1 : Brucellosis : a literature review

Bornarel, 1987). In Ivory Coast, Böhnel (1971) provided the first evidence of the disease in cattle during a serological survey in the northern part of the country. Some years later, serological evidences in humans were also provided (Gidel et al., 1974). 1.1.2.

Taxonomy, description and characteristics of Brucella

The etiological agents of brucellosis are bacteria members of the genus Brucella. The genus Brucella belongs to the family Brucellaceae within the order Rhizobiales of the class Alphaproteobacteria (Meyer and Shaw, 1920; Godfroid et al., 2011). Brucellae are facultative intracellular bacteria that grow slowly in aerobic conditions at 37°C but some strains may require 5 to 10% carbon dioxide for growth. Phenotypically, Brucellae appear as short rods (0.5-0.7 µm×0.6-1.5µm), non-motile, non-capsulate, small Gramnegative coccobacilli. After three to seven days of incubation on culture plates (Quinn et al., 1999; Alton et al., 1988), Brucellae colonies appear with round (2-4 mm in diameter), pinpoint shape, smooth, rough or mucoid (intermediate) aspect (Corbel and Brinley-Morgan, 1984). Contrarly to rough strains, smooth strains contain an O antigen on the lipopolysaccharide (LPS), a structural component of the outer membrane of the bacteria and appear translucent with a honey-color (Figure 5). Among the known species, so far, only B. canis and B. ovis have a rough shape (Alton et al., 1988).

Figure 5 : Brucella colonies on a solid culture media, showing translucent honey-colored appearance (credit picture: M. Sanogo) Using bacteriological examination, complete identification of Brucella until biovar level is made with a combination of morphological, cultural and biochemical characteristics (Table I). Classification of Brucella into species is dependent on criteria as natural host

13

Chapter 1 : Brucellosis : a literature review

preference, sensitivity to Brucella phages (Tbilisi (Tb), Weybridge (Wb), BK2, R/C) and oxidative metabolic profiles. Requirement of CO2 on primary isolation, H2S production, sensitivity to inhibition by thionin, basic fuchsin and safranin O dyes, and agglutination response to monospecific antisera for the A antigen of B. abortus and for the M antigen of B. melitensis M are used to determine subtypes or biovars (Corbel and Morgan, 1975; Alton et al., 1988; Saegerman et al., 2010; Godfroid et al., 2010). Nevertheless, the classification into subtypes or biovars may be sometimes problematic due to variability of some of the characteristics used for typing, such as sensitivity to dyes (thionine, fuschine, and safranine O), H2S production and CO2 requirement for growth (Acha and Zysfres, 2003). When they are available, DNA-based methods are also useful tools to characterize the different species and biovars of Brucella. They are particularly useful when a high discrimative power is needed and can be used in combination with other identification and typing methods (Adone et al., 2001; Bricker, 2002; Bricker et al., 2003). Various methods have been developed over the time including the Multilocus sequence analysis (MLSA), the whole genome sequencing and the global genome-wide Single nucleotide polymorphism (SNP) analysis (Le Flèche et al., 2006; Yu and Nielsen, 2010; Bankole et al., 2010; Sanogo et al., 2013a, Jiang et al., 2013; Scholz and Vergnaud, 2013). In addition to their high resolution, molecular based methods limit the manipulation of living agent (Le Flèche et al., 2006). Traditionally, species of Brucellae are determined according to their host preference and pathogenicity. Thus, the different species of Brucella and their associated hosts are as follows: B. abortus (cattle), B. melitensis (goats), B. suis (swine), B. canis (dogs), B. ovis (sheep), and B. neotomae (rodent) (Godfroid et al., 2005; Corbel, 2006; Saegerman et al., 2010). In addition to these common species, new strains of Brucella were later described. B. microti were isolated from common vole (Microtus arvalis) and wild red fox (Vulpes vulpes), B. pinnipediae and B. ceti in marine mammals (Ewalt et al., 1994; Foster et al., 1996; Clavareau et al., 1998; Godfroid et al., 2005; Cutler et al., 2005; Scholz et al., 2008; Scholz et al., 2010; Banai and Corbel, 2010; Nymo et al., 2011). Up to date, at least 10 species have been reported as members of the genus Brucella (Godfroid et al., 2011; Scholz and Vergaund, 2013). A summary of known species until now and a description of the different biovars are presented in Table I and Figure 6. Moreover, because of a high phylogenetic homogeny, it was suggested to consider all Brucella as belonging to the same single species namely B. melitensis and the other species of Brucella becoming the biotypes or biovars (Verger et al. 1985, 1987; De Ley et al., 1987; Cutler et al., 2005; Scholz et al., 2008).

14

Chapter 1 : Brucellosis : a literature review

Despite the scientific accuracy of this homogeny, the suggested classification is still not widely adopted due to lack of practicability (Pappas et al., 2005) and because the different species are also considered as different ecotypes (Cohan, 2002). Regarding the host preference-based classification, even if it appears more convenient, it is now apparent that a given host can be infected by different species of Brucella (Cutler et al., 2005). Thus, B. melitensis and B. suis were also reported in cattle in some epidemiological contexts, where cattle have contact with pigs and where cattle and small ruminants are grazing on common pastures as it is the case in Latina America, southern European countries and in the middle East (e.g. Godfroid and Kasbohrer, 2002; Godfroid et al., 2005, Szulowski et al., 2013, Fretin et al., 2013). The presence of B. melitensis in cattle was also documented in North Africa (e.g. Samaha et al., 2008) and in Eastern Africa (e.g. Muendo et al., 2012). In Western Africa, the lack of this kind of report did not preclude the possible presence of other species of Brucella in cattle population.

Figure 6: Dispersion of Brucella species according to their preferred host mammal. The dispersion of the various Brucella species is depicted as cones proportional to the number of strains analyzed (adapted from Moreno, 2014).

15

Chapter 1 : Brucellosis : a literarture review

Table I: Nomenclature and characteristics of Brucella species (from Pappas et al., 2005; Whatmore, 2009; Whatmore et al., 2014) Species

Biovar

Animal Hosts

Human virulence*

Species discrimation

B. melitensis

1-3

Goats, sheep, camels

++++

Fushin, positive; thionine, positive; safranin inhibition, negative; H2S production, negative; urease, positive in 24 hr; CO2 growth, negative;Tbilisi phage lysis, negative; Weybridge phage lysis, negative

B. abortus

1-6, 7**, 9

Cows, camels, yaks, buffalo

++ to +++

Fushin, positive (except biovar 2); thionine, negative (Biovar 1,2 and 4); safranin inhibition, negative; H2S production, positive (except biovar 5) urease, positive in 24 hr; CO2 growth, positive (biovar 1-4);Tbilisi phage lysis, positive; Weybridge phage lysis, positive

B. suis

1-5

Pigs (biovars 1-3), wild hares (biovar 2), caribou (biovar 4), reindeer (biovar 4), rodents (biovar 5)

+

Fushin, negative (except biovar 3); thionine, positive; safranin inhibition, positive; H2S production, positive (biovar 1); urease, positive in 15 min; CO2 growth, negative;Tbilisi phage lysis, negative; Weybridge phage lysis, positive

B. canis

-

Canines

+

Fushin, positive or negative; thionine, positive; safranin inhibition, negative; H 2S production, negative; urease, positive in 15 min; CO2 growth, negative;Tbilisi phage lysis, negative; Weybridge phage lysis, negative

B. ovis

-

Sheep

-

Fushin, negative for some strains; safranin inhibition, negative; H2S production, negative; urease, negative; CO2 growth, postive;Tbilisi phage lysis, negative; Weybridge phage lysis, negative

B. neotomae

-

Rodents

-

Fushin, negative; safranin inhibition, negative; H2S production, positive; urease, positive in 15 min; CO2 growth, negative;Tbilisi phage lysis, positive or negative; Weybridge phage lysis, positive

B. pinnipidialis and B. ceti

-

Minke whales, dolphins, porpoises (pinnipediae), seals (cetaceae)

+

Fushin, positive; thionine positive; safranin inhibition, negative; H 2S production, negative; urease, positive; CO2 growth, negative for B. pinnipidialis and positive for B. ceti ;Tbilisi phage lysis, negative; Weybridge phage lysis, positive for B. pinnipidialis and negative for B. ceti

B. inopinata

-

Unknown but isolated from human

?***

Fushin positive; thionine , positive; H2S production, positive; urease, positive; CO 2 growth, negative;Tbilisi phage lysis, positive or negative; Firenze phage lysis, negative

B. papionis sp. nov.

-

Unknown but isolated from baboon

?

Fushin positive; thionine , positive; H2S production, negative; urease, strongly positive; CO2 growth, negative; Tbilissi phage lysis, partially positive; Weybridge phage lysis, positive; Berkeley phage lysis, positive; Firenze phage lysis, positive;

* Virulence is graded on a scale from no virulence (–) to the highest degree of virulence (++++); **The status of this B. abortus biovar 7 in under review.***Not known

16

Chapter 1 : Brucellosis : a literarture review

1.2. Importance and distribution of brucellosis 1.2.1.

Socio-economic and public health importance

Brucellosis is an infectious disease with both socio-economic and public health importance. When present, the disease may have serious impact on animal production and productivity. It may also represent a severe hazard for human health. Brucellosis is also an important disease because of its potential to be weaponized for bioterrorism as it was the case during the 1950s (Cutler et al., 2012). With more than 500,000 new human cases recorded yearly, brucellosis is a major bacterial zoonotic disease of global importance (Cutler et al., 2005; Pappas et al., 2006; Corbel, 2006). Human brucellosis (or Malta fever or undulant fever) is responsible for an acute to chronic or severe debilitating and disabling disease with a wide range of clinical signs. These clinical signs include an “undulant” fever, sweating, weakness, headache, anorexia, weight loss, pain in joints and generalized pain (McDermott and Arimi, 2002; Dean et al., 2012). Human brucellosis is rarely fatal but tends to be chronic if not treated. Thus, complications such as endocarditis, meningitis (also called neurobrucellosis) and orchitis may occur (Corbel, 2006). The severety of the disease in human depends on the type of Brucella involved and the source of infection. Most severe clinical cases are commonly associated with B. melitensis (Benkirane, 2001; Corbel, 2006). Data on the actual incidence of the disease in humans are scare or lacking especially in sub-Saharan African countries. Available data suggest a higher incidence in low to middle-income countries where effective diagnosis or treatment is lacking or where programs for detecting and preventing infection in both humans and animals are not adequately implemented (Cutler et al., 2005; Corbel, 2006; Dean et al., 2012). In developing countries the infection rate was estimated to be above 10% (USDA and ILRI, 2013) and in the Republic of Chad, an incidence of 34.8 per 100,000 person-year was reported in nomadic communities (Dean et al., 2012). In animals, brucellosis is also recognized as a major pathological constraint to the development of livestock in sub-Saharan African countries (Camus, 1980a; Domenech, 1987; Akakpo, 1987). As a major constraint, brucellosis needs to be especially accounted for in developing countries, where about 70% of rural poor depend on livestock as part of their livelihood (LID, 1998). In addition to its public health importance, brucellosis has a negative impact on animal health and productivity. It primarily affects the reproductive system of the 17

Chapter 1 : Brucellosis : a literature review

host resulting in economic losses on productivity through late term abortions, calf mortality, reduced milk production and infertility (Pilo-Moron et al., 1979; Domenech, 1987; Corbel, 1997). The disease is reported to be responsible for about 20 to 25% of milk yield reduction (Timm, 1982; Acha and Szyfres 2005). A prevalence of about 30% infected cows within a herd is argued to cause a loss of the herd productivity of about 6 % (Domenech et al., 1982). Economic impact of brucellosis may also be indirect through the costs for veterinary interventions, investment for prevention and control measures (including vaccination and compensation), investment for restocking (in countries where culling is practiced) and losses related to consecutive exportation restrictions. Despite these known consequences, estimation of the actual economic impact of the disease in animal remains difficult. Mangen et al. (2002) estimated the losses of the annual value produced per animal between 6 and 10% (Mangen et al., 2002). Similarly, Camus (1980) reported a loss of about 10% of the annual income of cattle breeders in Ivory Coast due to brucellosis (Camus, 1980a). In Latin America, the annual losses related to bovine brucellosis were estimated to approximately 600 million USD3 (Acha and Szyfres, 2003). Similarly, estimation of the burden of the disease in humans is difficult, but it is expected to be high. For an average of 13 days spent in a hospital, Colmenero-Castillo et al. (1989) estimated an overall cost of 8,000 USD4 for human brucellosis per case in Spain. In addition, the total number of work absence days was 102 days per patient (Colmenero-Castillo et al., 1989). For the case of bioterrorist attack with B. melitensis, the economic impact for 100,000 persons exposed is expected to be about 478 million USD5, related to 82,500 human cases of brucellosis requiring extended therapy and leading 413 deaths (Kauffman et al., 1997). Therefore, given the serious consequences of brucellosis for public health and the economy, the disease needs to be considered carefully, especially in low-income countries such as Ivory Coast (McDermott and Arimi, 2002; Mangen et al., 2002; McDermott et al., 1013).

3

About 473 millions euros or about 310 billions FCFA

4

About 6,308 euros or about 4,138,000 FCFA

5

About 377 millions euros or about 248 billions FCFA

18

Chapter 1 : Brucellosis : a literature review

1.2.2.

Distribution of brucellosis

1.2.2.1.

Disease in animals

Considering the wide range of potential animal hosts, brucellosis is one of the most widespread diseases in the world. Since the beginning of the 1900s, animal brucellosis has been commonly reported in sub-Saharan countries, where bovine brucellosis remains the most widespread form (Akakpo and Bornarel, 1987, Corbel; 1997; Mangen et al., 2002; McDermott and Arimi, 2002; Bronvoort et al., 2009). The disease is endemic in most of the African countries (Corbel, 1997; Akakpo and Bornarel, 1987) with different prevalences (McDermott and Arimi, 2002). An overall apparent seroprevalence rate based on Rose Bengal testing was estimated to range from 10.2 to 25.7% in cattle populations of subSaharan Africa (Mangen et al., 2002). In West Africa, evidence of the disease was found in all the countries where research was conducted including Benin, Burkina Faso, The Gambia, Ghana, Guinea, Ivory Coast, Mali, Niger, Nigeria, Senegal, Sierra Leone and Togo (Mangen et al., 2002; Unger et al., 2003). In Ivory Coast, brucellosis is among the dominant pathologies affecting cattle population (Angba et al., 1987). Since the first evidences in this country, the prevalence of the disease has been investigated throughout the years and at different geographical scales. These investigations mainly covered the northern and central regions, the main breeding areas of the country (Table II). Results from a national survey conducted between 1975 and 1977 reported a seroprevalence in cattle of 11.3% (Angba et al., 1987). More recently, Thys et al. (2005) estimated a true prevalence of 3.6 and 4.2% respectively in dairy farms and their neighbours’ traditional farms in peri-urban area of Abidjan (Thys et al., 2005). Recently, using sera collected during the serosurveillance of Rinderpest in Ivory Coast, the true prevalence of the disease in traditional cattle was estimated to range between 5 and 16% in the central part of the country (Sanogo et al., 2008). The different species and biovars of Brucella are distributed heterogenously throughout the world but B. abortus remains the most prevalent worldwide so far (Corbel, 1997; Robinson, 2003; Acha and Szyfres, 2005). The presence of biovars 1 and 6 of B. abortus has also been confirmed in Ivory Coast from hygroma fluid samples (Pilo-Moron et al., 1979). No isolates of B. melitensis have yet been reported in this country neither in cattle nor in the small ruminants, despite evidence of the presence of the disease in small ruminants (Gidel et al., 1974; Chartier, 1982).

19

Chapter 1 : Brucellosis : a literarture review

Table II: Studies on prevalence of bovine brucellosis in Ivory Coast, 1970-2008 Author (s), year of publication

Study area

Type of herd

Number of tested animals

Tests applied

Average infection rate (%)

Anomynous, 1970

Bouaké

Not specified

24

BPAT a

75% (53.3-90.2) b

Böhnel, 1971

Korhogo

Not specified

554

MRT

11.7% (9.2-14.7)

Coulibaly, 1973

Bouaké

Not specified

281

Not specified

23% (18.3-28.5)

Gidel et al., 1974

Korhogo, Bouaké Katiola, Odienné, Man

Traditional and sedentary farms

1327

MRT

42,9 (23,0-51,0)

749

SAW and CFT

15,5 (2,6-25,8)

Traditional and sedentary farms

12.343

SAW, RBT

10,1 (1,0 -39,3)

Sedentary herds

1.214

RBT

28,3 (9,1-37,7) c

Not specified

Not specified

SAW, RBT

11,3 (9,5-14,0)

Dairy farms

244

Traditional farms

137

Pilo-Moron et al., 1979

Camus, 1980a Angba et al., 1987 Thys et al., 2005

d

Korhogo Bouaké, Abengourou Abidjan Korhogo, Boundiali Odienné, Ferkessédougou, Bouna, Touba National survey District of Abidjan (Bingerville, Azaguié )

SAW-EDTA RBT, CFT, iELISA SAW-EDTA, RBT, CFT, iELISA

3,6 (1,2-7,1) 4,3 (1,3-8,8)

Bongouanou, Dimbokro, SAW-EDTA, Tiébissou, Toumodi, Traditional farms 660 8.8 (5.0-16.4) RBT, CFT, iELISA Yamoussoukro a The antigen used was a B. melitensis strain b The prevalence range is presented within brackets c Only cows were used for this estimation d Infection rate presented by these authors are true prevalence MRT(Milk Ring Test) ; SAW (-EDTA)(Slow Agglutination of Wright(with EDTA) ;RBT(Rose Bengal Test ), CFT(Complement Fixation Test) ; iELISA (indirect ELISA) Sanogo et al., 2008 d

20

Chapter 1 : Brucellosis : a literarture review

1.2.2.2.

Disease in human

Brucellosis is among the most neglected zoonotic disease in the world (WHO, 2012; Mableson et al., 2014). As any other zoonotic diseases, the occurrence of brucellosis in human in a given geographical region is related to the infectious status of animals (Godfroid et al., 2005; Saegerman et al., 2010). Infection of a given host by Brucella may occur either directly via ingestion, via inhalation of infected products or indirectly (Figure 7).

Figure 7: Main transmission routes of brucellosis from livestock to humans (by Sir David Bruce, 1855-1931) (from http://m2002.tripod.com/brucellosis.jpg) Even if direct or close contact with aborted material or infected animal is required for transmission of Brucella, indirect transmission is possible through contaminated pasture, vehicles, feed or water (Roop et al., 2003). Most of the time, the transmission of Brucella to human mainly occurs via the consumption of raw animal products and direct contact with infected animals, aborted tissues and discharges (Marcotty et al., 2009; Saegerman et al., 2010). Human brucellosis is mainly an occupational disease, affecting people who have contact with infected animals or their tissues such as farm workers, veterinarian, ranchers, and meat picking employees (OIE, 2009). Therefore, consumers of unpasteurized dairy products and hunters who unknowingly handle infected animals may also get brucellosis (Ocholi et al., 2004; Arimi et al., 2005). To a lesser extent, sexual transmission was also reported (Meltzer et al., 2010) but not confirmed (Ron-Román et al., 2012). So far, human

21

Chapter 1 : Brucellosis : a literature review

brucellosis is known to be caused by B. melitensis, B. suis, B. abortus and to a lesser extent by B. canis (Acha and Szyfres, 2005). A zoonotic potential has also been alleged for newly reported strains in marine mammals but further investigations are needed (Godfroid et al., 2005). Humans are known to be more sensitive to infections caused by B. melitensis and B. suis, especially biovars 1 and 3 for the later (Godfroid et al., 2005; Saegerman et al., 2010). Brucella abortus infections are relatively low pathogenic and usually develops an insidious subclinical form. The incidence of the disease in human is not well known but higher prevalences are reported in the Mediterranean region of Africa, Middle East, Latin America and Asia (Samartino et al., 2005, Pappas et al., 2006; Aznar et al., 2014) (Figure 8).

Figure 8: Global Incidence of Human Brucellosis (from Pappas et al., 2006) In West Africa, knowledge on the actual impact of the disease in humans remains limited (Figure 8). However, even if the actual incidence of the disease is not known, the presence of human brucellosis cannot be excluded since brucellosis stays neglected and under-reported in many African countries. This might be related to the non-specific clinical signs of human brucellosis, which could also be confused with other endemic febrile illness such as malaria and typhoid fever (McDermott and Arimi, 2002). Nevertheless, serological evidence of the presence of Brucella in humans has already been documented for many Western African countries like Benin, Burkina Faso, Ivory Coast, Guinea, Guinea Bissau, Mali, Togo and Nigeria (Pappas et al., 2006; Akakpo et al., 2010, Dean et al., 2013). Available data on human brucellosis in Ivory Coast are scarce. In the 1970s, serological evidence of the presence of the disease in human was reported in northern and central region

22

Chapter 1 : Brucellosis : a literature review

of the country despite a low infection rate in animals (Gidel et al., 1974). In 1985, brucellosis infection in human was estimated to be 6.52 % and 0.45% using respectively the intradermoreaction with Brucella melitine as allergen and serology (Angba et al., 1987). 1.3. Prevention and control of brucellosis 1.3.1.

Diagnostic tools

Brucellae are facultative intracellular bacteria with a special tropism for the host reproductive system and may affect a wide range of hosts including human, domestic and wild animals. When transmission occurs, Brucellae initially invade regional lymph nodes. Then, after a brief bacteremia, they spread to other tissues and organs of the body, with a particular tropism for the reproductive tract (Olsen and Tatum, 2010). This results in an increased excretion of Brucellae during parturition particularly in aborted fetal fluid, vaginal discharges, placenta and milk. Due to the tropism and their proliferation in the reticuloendothelial cells of the reproductive tract, Brucellae induce various clinical signs from birth of a viable but weak calf, placenta retention, metritis, subclinical mastitis, reduced milk production, infertility, orchitis or epididymitis with or without sterility to late term abortion at the first gestation (Acha and Szyfres, 2003). Among these clinical manifestations, abortion is the cardinal clinical sign commonly associated with brucellosis (Acha and Szyfres, 2003; Godfroid et al., 2010). Joint colonization by Brucella may occur resulting in articular and peri-articular hygroma’s. In Africa, the presence of hygroma in a herd is commonly associated with brucellosis (Thienpont et al., 1961; Ocholi et al., 2004; Bankole et al., 2010; Saegerman et al., 2010). However, none of these clinical signs is specific for brucellosis. Therefore, the use of laboratory techniques remains essential to improve the accuracy of the diagnosis. Two main types of laboratory techniques could be used for the diagnosis of brucellosis: on the one hand, tests which allow a direct detection of the presence of Brucella (bacteriology, molecular methods) and on the other hand, tests which detect this presence indirectly through detection of the immune response of the host to Brucella antigens (mainly serological and intradermic methods). Diagnostic tests are useful to determine the disease status at individual level or group of individual within a population of interest. They are also crucial for studying the epidemiology of the disease, to assess the actual impact of disease (Greiner and Gardner, 2000). Diagnostic assays for brucellosis were thoroughly described and reviewed in the OIE manual of

23

Chapter 1 : Brucellosis : a literature review

diagnostic tests and vaccines for terrestrial animals (OIE, 2009) and by several authors (Nielsen, 2002; Saegerman et al., 2010; Godfroid et al., 2010; Yu and Nielsen, 2010). A brief overview of the different methods is presented in the following sections. 1.3.1.1.

Direct diagnostic methods

1.3.1.1.1.

Bacteriological examination

Morphological, staining and cultural characteristics of Brucella may be used for a direct identification of its presence. Despite its lack of specificity, examination of stained smears from abortive material or suspicious organs or fluid can provide valuable information. However, definitive diagnostic is made by culture (Yu and Nielsen, 2010). According to clinical signs, a range of samples is available including fetal membranes, vaginal secretions, milk, semen, arthritis or hygroma fluids, lymph nodes, spleen, uterus, udder, testes, epididymes, joint exudates, abscesses and other tissues of infected cattle. The stomach content, spleen and lungs of aborted fetuses may also be used for bacteriological examination (Alton et al., 1988; Corbel, 2006, OIE, 2009; Godfroid et al., 2010). In case of abortion caused by brucellosis, concentrations of Brucella in fetal fluids or placenta may reach 109 to 1010 colony-forming units (CFUs)/g compared to an estimated minimum infectious dose of 103 to 104 CFU (Olsen and Tatum, 2010; Saegerman et al., 2010). Brucellae may also be spread from infected udders and supramammary lymph nodes into milk at concentrations going from a few hundred up to 2x106 organisms/ml of milk (Corbel, 1988). Therefore milk, mammary glands and associated lymph nodes can be used as samples (Xavier et al., 2009; O’Grady et al., 2014). In Africa, the use of hygroma fluid as sample for Brucella isolation and identification is common (Ocholi et al., 2004; OIE, 2009; Bankole et al., 2010). Isolation of Brucella is considered as the “gold standard” reference test to determine the status of the animal regarding brucellosis (Godfroid et al., 2010; Yu and Nielsen, 2010). Nevertheless, implementation of bacteriological methods is laborious, time consuming, costly and requires enhanced biosafety and biosecurity measures. The risk for human health implies that handling of Brucella be done careful and restricted to laboratories with appropriate containment facilities like biosafety level 3 (BSL3) laboratories (OIE, 2009). Thus, despite their usefulness for the detection of Brucella, the implementation of these methods is limited in developing countries. The main different characteristics used to differentiate among Brucella species using bacteriological examination are summarized in Table III.

24

Chapter 1 : Brucellosis : a literature review

Table III: Characteristics for differentiating the different Brucella (from OIE, 2009)

25

Chapter 1 : Brucellosis : a literarture review

1.3.1.1.2.

Molecular methods

Compared to bacteriological methods, molecular based methods to detect Brucella are considered less fastidious, less time consuming and with reduced risk of manipulation. Even if detection of antibodies produced against Brucella is indicative of the presence of brucellosis, identification and typing of the disease-causing agents provides the ultimate evidence of the actual presence of the disease (Nielsen, 2002). Molecular methods are based on the detection of Brucella DNA and therefore provide actual evidence of the latter (Yu and Nielsen, 2010). Since 1987, many Polymerase Chain Reaction (PCR) based methods were developed throughout the years for the diagnosis and identification of Brucella (Bricker and Halling, 1994; Bricker, 2002; Whatemore, 2009; Yu and Nielsen, 2010; Godfroid et al., 2011, Scholz and Vergnaud, 2013). It includes i.e., Abortus-Melitensis-Ovis-Suis (AMOS) PCR, conventional multiplex PCR, single nucleotide polymorphism (SNP) analysis and multilocus sequence typing (MLST) or multilocus sequence analysis (MLSA) . They can be used either for diagnostic purposes or for biotyping. Molecular methods are useful and convenient to characterize circulating biovars of Brucella and for epidemiological investigations particularly when a higher discriminatory power is needed (Bricker et al., 2003; Cutler et al., 2005; Le Flèche et al., 2006). Molecular methods can also be used as a complementary test for other tests (Adone et al., 2001; Bricker, 2002). Yu and Nielsen (2010) published a broad overview on molecular methods for detection of Brucella, including highly discriminative methods such as Multiple loci Variable number tandem repeats Analysis (MLVA). More discriminative molecular methods are under development with recently available methods like the whole genome sequencing (WGS) and the global genome-wide Single Nucleotide Polymorphism (SNP) analysis. 1.3.1.1.3.

Maldi-TOF Mass Spectrometry (MALDI-TOF-MS)

The Matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF-MS) is recognized as a reliable method for identification of Brucella at genus level from culture plates samples and directly from blood culture bottles (Ferreira et al., 2010; Lista et al., 2011, Kasymbekov et al., 2013). It is considered as a fast, cost-effective and accurate method, which is suitable for the high-throughput identification of bacteria by less-skilled laboratory personnel (Seng et al., 2009). Identification of bacteria is done by comparing the obtained MS spectra to the MS spectra or profiles from a reference library,

26

Chapter 1 : Brucellosis : a literature review

which constitutes the main limiting factor of the enhanced identification of Brucella using this method so far (Lista et al., 2011). 1.3.1.2.

Indirect diagnostic methods

1.3.1.2.1.

Serological methods

When a host is exposed to Brucella, the immune system induces the production of different types of immunoglobulins (Ig) regardless the Brucella species. This immune response is induced by the presence of surface lipopolysaccharides (LPS) on the outer cell membrane of Brucella, which contains the O chain, the immunodominant antigen6 (Alton et al., 1988). In cattle, the first antibody response is the production of IgM at a larger (or less) and persistent amount (2-3 weeks post-exposure) according to the dose of bacteria, the route of infection and the status of the animal infected. However, IgM may disappear after a few months (Godfroid et al., 2002; Saegerman et al., 2004; Godfroid et al., 2010). Then, the production of IgG1 arrives very shortly after the IgM, followed later by IgG2 and IgA (Nielsen, 2002; Yu and Nielsen, 2010). The presence of these different antibodies can therefore be useful to evidence of the presence of Brucella and serological diagnostic tools can be use to detect them. Compared to bacteriological examination, detection of antibodies appears to be a more convenient approach for the diagnosis of brucellosis, since they are less fastidious, easier to implement and more suitable for large-scale investigation. Because the immunodominant epitope on the Brucella LPS is the basis of the serodiagnosis of brucellosis, most of the conventional serological tests may suffer from some limitations since the immunodominant antigen is also present in many other Gram-negative bacteria. Like Brucella, these bacteria induce the production of identical antibodies resulting in cross-reactions or false positive serological reactions (FPSR) when testing. These false positive serological reactions (FPSR) were reported with Gram-negative bacteria such as (e.g.) Yersinia enterocolitica O: 9, Escherichia coli O: 157, Francisella tularensis,

6

The immunodominant O-polysaccharide (OPS) which has been chemically defined as a homopolymer of 4,6-dideoxy-4-formamide-alpha-D-mannose linked via glycosidic linkages is common in Smooth Brucella strains but is lacking in Rough strains (i.e., B. ovis and B. canis). As a result, B. abortus antigen in the form of whole cells, SLPS or OPS is used as antigen for serological detection of the smooth strains while RLPS is commonly used as the main antigen for detection of antibody for the latters (Yu and Nielsen, 2010).

27

Chapter 1 : Brucellosis : a literature review

Salmonella urbana O: 30, Vibrio cholera. The presence of possible FPSR bacteria become an important issue especially in free brucellosis areas, countries with low incidence or in the last stages of an eradication program (Saegerman et al., 2004; Munoz et al., 2005; Cutler et al., 2005; Corbel, 2006; Yu and Nielsen, 2010). In addition, vaccination with strain S19 (also named B19) may also cause interference in serological reactions. Consequently, the application of serological tests has to be related with the incidence of the disease and vaccination status in order to prevent misinterpretation (Corbel, 2006). The use of a given serological test for the diagnosis of brucellosis should therefore take into account of the epidemiological context (Robinson, 2003). A description of the different types of antibodies detected by the different serological assays is presented in Table IV. Different serological methods for the diagnosis of brucellosis have been developed over the years. These include the agglutination tests, complement fixation tests, primary binding tests and hypersensitivity tests (Nielsen, 2002, Cutler et al., 2005). Table IV: Types of antibodies detected by conventional serological assays for the diagnostic of brucellosis (from Quinn et al., 1999 and Saegerman, 2004) Type Sample Blood

of

Type of antibodies

Tests

IgM

IgA

IgG1

IgG2

SAW (with EDTA)

+

-

±

+

RBPT

+

-

+

-

+

-

+

-

iELISA

-

-

+

+

MRT

+

+

±

±

iELISA

-

-

+

+

CFT a

Milk

+/-/± : serological response ; MRT (Milk Ring Test) ; SAW (-EDTA)(Slow agglutination test of Wright (with EDTA) ; RBPT ( Rose Bengal Plate Test), CFT (Complement fixation test) ; iELISA (Indirect Enzyme linked immunosorbent assays); a Detection of IgG1 and/or IgG2 would depend on the conjugate used

1.3.1.2.1.1.

Sero-agglutination of Wright (SAW)

The Sero-agglutination of Wright (SAW) or Slow Agglutination Test (SAT) is the oldest diagnostic test of brucellosis. It is an inexpensive and relatively easy to implement semi-quantitative test. Different types of SAW exist, from the simple SAW to the SAWEDTA where EDTA is added to enhance the specificity of the test by inactivating some non-

28

Chapter 1 : Brucellosis : a literature review

specific IgM (Alton et al., 1988; Corbel, 2006). It is susceptible to false positive reactions by cross-reacting antibodies. 1.3.1.2.1.2.

Rose Bengal Test (RBT)

The Rose Bengal Test (RBT) is an agglutination test between agglutinating IgM immunoglobulins and a colored antigen suspended in an acid buffer (pH 3.65) to prevent reaction of non-specific agglutinins. It can be used for diagnosis at the herd level and particularly for the sero-surveillance of cattle herds in brucellosis free areas. It is a quite sensitive and relatively simple test to apply (OIE, 2009). Although, it is a good screening test for brucellosis, it is unable to distinguish vaccinated from infected animals (Corbel, 2006) and can produce false positive serological reactions. It is a prescribed test for trade by OIE (OIE, 2009). 1.3.1.2.1.3.

Complement Fixation Test (CFT)

The Complement Fixation Test (CFT) is based on the activation of the complement in the presence or not of the antigen-antibody complex. This presence of the complex is detected using a hemolytic system. This test is quite specific even if false-positive reactions may still occur. Despite its status of prescribed test by the OIE for trade (OIE, 2009), it is a difficult and complicated test to implement (Alton et al., 1988). Moreover, it is not well standardized and anti-complementary reactions (due to bacterial contamination of the sera, or other factors) sometimes make the interpretation of the results difficult (Corbel, 2006). 1.3.1.2.1.4.

Enzyme Linked Immunosorbent Assays

The indirect Enzyme Linked Immunosorbent Assays (iELISA) use a purified Smooth LPS coated on a polystyrene matrix as an antigen (Nielsen, 2002; Saegerman et al., 2004; Godfroid et al., 2010). It allows an indirect detection of antibodies against the Brucella antigen using various conjugates with enzymes. Indirect ELISA techniques are considered as very sensitive tests but do not allow to distinguish between post-vaccination and postinfective antibodies (OIE 2009). Therefore, competitive ELISAs were developed to improve the test specificity (Nielsen et al., 1996a; Portanti et al., 2006) but they are less sensitive compared to indirect ELISAs (Nielsen et al., 1995; Samartino et al., 1999a). Competitive ELISAs are helpful to eliminate false positives and to discriminate between post-vaccination

29

Chapter 1 : Brucellosis : a literature review

antibodies due to the vaccine strain S19 and post-infection antibodies (Weynants et al., 1997). These tests are prescribed for trade by the OIE (OIE, 2009). 1.3.1.2.1.5.

Fluorescence Polarization Assay (FPA)

The FPA is a highly sensitive, specific, rapid and easy to implement test (Gall and Nielsen, 2004). This method consist in the blank reading of the diluted sample in a fluorescence polization analyser, addition of an antigen labeled with a fluorochrome and final reading after two minutes of incubation (Samartino et al., 1999b). It is based on the rotational differences between molecules in solution, the smaller rotating randomly at a rapid rate and resulting in a rapid depolarization of light while the larger will rotate slower and depolarize light at reduced rate (Nielsen and Gall, 2001). Depending on the presence or absence of an antigen-antibody complex, the size of the molecules in the solution will increase (or not) and therefore, its rate of rotation will be proportionally reduced (or not). For brucellosis, a small subunit of the O Polysaccarhide (OPS) of B. abortus strain 1119-3 smooth Lipopolysaccharide (sLPS) conjugated with fluorescein isothiocyanate is added to diluted serum, milk or whole blood sample. The test result is obtained by measuring the time of rotation in a given angle with a polarized light after incubation (OIE, 2009). The specificity of the FPA is quite high even in cattle herds vaccinated with strain 19 (Nielsen et al., 1996b; Gall et al., 2000; Gall et al., 2001). It is a prescribed test by OIE for trade (OIE, 2009). 1.3.1.2.1.6.

Brucella

Immunochromatographic

Lateral

Flow

Assays (LFA) The LFA is a rapid test initially developed for the diagnostic of human brucellosis (Smits et al., 2003; Irmak et al., 2004, Franco et al., 2007). It was later adapted for the serodiagnosis of brucellosis in different livestock species including cattle, sheep, goat and pigs (Abdoel et al., 2008). It is a simple test also based on the detection of IgM/IgG against Brucella LPS. As in other serological tests, cross-reaction may also occur. LFA does not require a specific expertise, expensive equipment, electricity and refrigeration, or a specific training. It is also argued to be suitable for remote areas where access to adequate laboratory facilities is problematic or when testing animals from nomadic or other migratory livestock keepers (Abdoel et al., 2008, Bronsvoort et al., 2009).

30

Chapter 1 : Brucellosis : a literature review

1.3.1.2.1.7.

Milk testing assays

Serological diagnostic tests applied to milk are good and practical means for screening in the dairy sector. They can be easily implemented in milk collection centers. In this case, when the test is positive, it implies that all cows which are in production and were milked, be tested using other serological tests. The "milk iELISA" is a very sensitive and specific test. The Milk Ring Test (MRT), an adaptation of the agglutination test for milk, is also a good alternative test in the absence of ELISA because it is very cheap (OIE, 2009). False positive reactions are common with MRT especially in brucellosis free areas (Corbel, 2006). 1.3.1.2.2.

Cellular methods

1.3.1.2.2.1.

Skin Delayed-type Hypersensibility Test or Skin Test

The “Skin Test” or intradermal test with brucellin is based on the hypersensitivity or allergic inflammatory reaction of the host after an intradermal injection with a protein extract of a rough strain of Brucella (Saegerman et al., 1999; Godfroid et al., 2010). This test is based on the specific cell-mediated immunity against Brucella and is able to identify latent carriers and to discriminate false positive serological reactions due to Yersinia enterocolitica O: 9, when associated with other serological tests (Saegerman et al., 1999; Bercovich, 2000). It is highly specific with some disadvantages such as the inability to discriminate between infected and vaccinated animals, and the need for at least two visits with an interval of 2 to 3 days to get the result (Weynants et al., 1995; Cutler et al., 2005). It is prescribed as an alternative test by the OIE (OIE, 2009). 1.3.1.2.2.2.

Antigen-Specific Gamma Interferon Production Test

The antigen-specific Gamma interferon production test is an in vitro test developed in order to improve the diagnostic specificity of bovine brucellosis. It is based on the quantification of the Gamma interferon (γ-IFN) produced in response to antigenic stimulation. It is a delayed-type hypersensitivity test similar to the "skin test" but with a lower specificity. It can be used as a complementary test with others serological assays (Weynants et al., 1995).

31

Chapter 1 : Brucellosis : a literature review

1.3.1.3.

Identification and typing methods

Results of identification and typing of Brucella are useful to have a better knowledge of the epidemiology of the disease to manage disease outbreaks, to identify appropriate antigens and to test and set up efficient preventive and control measures (Crawford et al., 1979; Ica et al., 1998; Saegerman et al., 2010; Godfroid et al., 2010). At national and at regional levels, identification and typing results from infected animals are helpful to assess potential threats for public health since animal hosts are the source of human brucellosis infections. Despite their high genetic relation, application of both bacteriological and molecular typing methods may be used for identification and typing of Brucella (Scholz and Vergnaud, 2013) as discussed above. However, differentiation among species and biovars is sometimes complicated because of the existence of strains showing atypical characteristics (Acha and Zsyfres, 2003; Scholz and Vergnaud, 2013). Handling and biotyping of Brucella also requires facilities, equipment and technical skills that are not always available in diagnostic laboratories in Africa, limiting the availability of data on prevailing strains of Brucella (Samartino et al., 2005). 1.3.2.

Prevention and control measures

Given that brucellosis is a zoonotic disease, there is a correlation between human and animal brucellosis. Prevention of brucellosis in human mainly depends on the control of the disease in the animal hosts (Godfroid et al., 2005; Pappas et al., 2006; Saegerman et al., 2010). Different strategies for controlling brucellosis exist and have been applied in different part of the world (Benkirane, 2001; Godfroid and Kasbohrer, 2002; Ragan, 2002; Samartino, 2002; Poester et al., 2002; Rivera et al., 2002; McDermott and Arimi, 2002; Saegerman et al., 2010, FAO, 2013). The aim of these strategies is to prevent the spread of the infection, to reduce the risk of abortion and to increase the herd or population immunity. Strategies for controlling brucellosis could include measures as appropriate herd management (Samartino et al., 2005), vaccination of the susceptible population, slaughtering of the animals recognized positive to testing (Benkirane, 2001), and increase of public awareness and education of population at risk (Robinson, 2003). All these measures could be applied separately or in combination but need to be backed up by appropriate regulations or legislation. An efficient control strategy need to consider some key elements like the true prevalence of the disease (Table V), livestock management system, organization of the veterinary services, implication of policymakers and communities (stakeholders), availability of resources to sustain control

32

Chapter 1 : Brucellosis : a literature review

measures, and the intersectoral collaboration between veterinary services and public health actors (Benkirane, 2001; Saegerman et al., 2010). In many developed countries, control programs including measures such as test-and-slaughter with compensation for farmers, accreditation and financial incitation for disease-free herds have been successfully applied to control brucellosis (Saegerman et al., 2010). Many countries such as Australia, Belgium, Canada, Cyprus, Denmark, Finland, the Netherlands, New Zealand, Norway, Sweden, and the United Kingdom have been declared free from bovine brucellosis. In developing countries, despite its known endemicity, its socioeconomic impacts and the beneficial effect of possible control measures, resources allocated to control of brucellosis are declining or absent. Most of the time, vaccination is the only mean applied for the control of animal brucellosis in these countries (McDermott and Arimi, 2002).

33

Chapter 1 : Brucellosis : a literarture review

Table V : Control strategies of brucellosis according to the epidemiological status (adapted from Benkirane, 2001; Saegerman et al., 2010) Epidemiological status

Control strategy measures

and

associated Monitoring method

Outcome/ next step

A: High prevalence in animals; high clinical incidence in humans

-

Mass vaccination Support to veterinary services Rational use of available resources Movement control

-

Serology Bacteriology Monitoring the incidence of human cases

B: Moderate prevalence

-

Combined prophylaxis

-

Go to C Counting and identification of animal Serology control Bacteriological monitoring Communication/sensitization/education Intersectoral collaboration with human health services

C: Low prevalence (<1%)

-

Sanitary prophylaxis

-

Monitoring in farms and slaughter houses Serological monitoring Survey in target groups

Reach D

D: Absence of the disease

-

Movement control

-

Monitoring of risk indicators

Preserve this status

Go to B

34

Chapter 1 : Brucellosis : a literarture review

1.3.2.1.

Communication and education for prevention

Communication and education of public at risk are considered as a key component to increase the awareness of the disease, to prevent its spread and to reduce occupational and food-borne risk linked to zoonotic diseases such as brucellosis (Robinson, 2003). Education and sensitizing should be undertaken to prevent the consumption of unpasteurized milk and milk derivatives (Samartino et al., 2005). Populations with cultural habits encouraging the consumption of milk and the use of its products raw or poorly cooked are highly at risk and should be sensitized in priority. Since the disease is likely to be transmitted in a context where people have close contact with the animal host, hygienic and biosecurity measures during handling and disposal of afterbirths- especially in cases of abortion - should also be taught and encouraged, particularly among professionals at higher risk like hunters, farmers, butchers, and veterinarians (Corbel, 2006; Saegerman et al., 2012). 1.3.2.2.

Prophylactic measures

Since Brucellae are facultative intracellular organisms, the effectiveness of antibiotherapy is limited. Furthermore, the use of antibiotics for the treatment of brucellosis would require the use of massive doses, increasing the risk of antibiotic residues and resistance dissemination to humans through the food chain. Implementation of preventive medical measures, e.g. vaccination, is therefore a key component for the prevention/control of brucellosis. Vaccination is used to increase the resistance of susceptible animals to infection, to reduce the expression of clinical signs and to diminish the excretion of Brucella by infected animals (Corbel, 2006). In many countries, it was adopted as the most practical and economical way for controlling animal brucellosis (McDermott and Arimi, 2002; Aznar et al., 2014). In Ivory Coast, vaccination was used between 1978 to 1982 during a control program conducted by the SODEPRA. Females from 1 to 10 years of age were vaccinated at primovaccination. Then, non-pregnant females of one to two years old were vaccinated every year. About 300,000 females have been vaccinated in the north and the centre of the country (Angba et al., 1987) using mostly H38, but also B19 vaccine strains. The campaign led to the reduction of abortion and mortality rate up to more than 37% (Camus, 1980a; Camus, 1980b; Angba et al., 1987). Because of the resurgence of brucellosis, female calves of 4 to 8 months of age and non-pregnant cows in dairy farms were vaccinated again in 1992 using a single

35

Chapter 1 : Brucellosis : a literature review

dose sub-cutaneaous injection as previous years (Camus, 1995). With the privatization of SODEPRA in 1993, vaccination activities were transferred to the private sector and were progressively abandoned, farmers being henceforth asked to pay for vaccination. Currently, there is no official control program or official vaccination against brucellosis in Ivory Coast. When vaccination is applied for the control of brucellosis, there might be some disadvantages such as its possible interference with most diagnostic tests (serological and hypersensitivity). In cattle, the use of S19 vaccine (smooth attenuated strain of B. abortus) is recommended but is not effective in protecting animals against infections with B. melitensis (Corbel, 1997). The RB51 vaccine (rough attenuated strain of B. abortus) also gives satisfaction and seems to interfere very little with serological tests (Schurig et al., 2002). Despite a lower efficiency compared to the S19 strain, RB51 vaccine is preferred over the S19 in several Latin American countries (Corbel, 2006). In addition to medical prophylactic measures, sanitary measures can also be used to prevent the introduction and the spread of the disease in a given population. For brucellosis, these include hygiene, containment and animal movement control. In addition, the use of appropriate and accurate diagnostic tests, allow to identify and eliminate infected or testpositive individuals. Aiming to prevent the spread of the disease, elimination may imply the slaughtering of positive tested animals (test-and-slaughter). The efficiency of these methods depends on the epidemiological context, the availability of sustainable resources and appropriate regulation. In many developed countries, test-and-slaughter was applied in a control strategy for brucellosis, in combination with compensation of farmers, accreditation and financial incentives for disease-free herds (Saegerman et al., 2010, Godfroid et al., 2013). Also known as stamping out method, test-and-slaughter is generally implemented in association with vaccination (Corbel, 1997). In effect, vaccination is first used to prevent or control the infection among infected host, and then it is gradually restricted while test-andslaughter is implemented to eliminate the infection. Availability of an appropriate financial compensation scheme is the main limiting factor of the implementation and the success of any control program including test-and-slaughter policy (Seleem et al., 2010; Godfroid et al., 2013), particularly in low resource countries. The success of the application of this measure is unlikely if the herd level prevalence is more than 2% (Corbel et al., 2006). 1.3.2.3.

Intersectoral collaboration for control

The efficiency of prevention and control measures of zoonotic diseases like brucellosis requires the implication and the collaboration of both animal and human health sector. It is 36

Chapter 1 : Brucellosis : a literature review

expected to ensure joint administrative arrangements, facilitate cross-notification of cases, as well as coordinated investigations, surveillance and prevention/control activities, and public health education programs (Corbel, 2006). Such collaboration should be encouraged at both national and regional level, in order to put together limited resources and capabilities for an efficient control. Thus, emerging concepts such as the “One health approach” can be considered as an opportunity to improve human health and well-being through an integrated management of pathogens as Brucella spp in both humans and domestic animals (Saegerman et al., 2010; Saegerman et al., 2012; Marcotty et al., 2013). This approach is expected to be particularly beneficial in low resource societies where different disciplines could be combined to improve the strength of the surveillance and the control of infectious diseases like brucellosis.

37

CHAPTER 2: EPIDEMIOLOGICAL CONCEPTS AND METHODOLOGIES

38

CHAPTER 2: EPIDEMIOLOGICAL CONCEPTS AND METHODOLOGIES This chapter describes the epidemiological tools used in the thesis. 2.1. Systematic review and meta-analysis Reliable and good quality data or information are essential to support decision-making and for answering urgent questions. Most of the time, such data or information are provided by systematic review or meta-analysis. They gather data or results from several studies into a single synthesis (Montori and Guyatt, 2003, Leeflang et al., 2008). Synthesizing results from several studies can be done in many ways but not all of these are scientifically robust (Honest and Khan, 2002). Systematic reviews allow a synthesis of relevant studies by applying scientific strategies that limit biases (Wright et al., 2007). A systematic review is a review of a clearly formulated question that uses systematic and explicit methods to identify, select, and critically appraise relevant research, and to collect and analyze data from the studies that are included in the review (Moher et al., 2009). Thus, when doing a systematic review, the author(s) should i) address a defined question; ii) conduct a detailed and exhaustive search for relevant studies; iii) include studies of high methodological quality; and iv) use reproducible approaches to assess the limitations in the methodological quality of the studies on which they focus (Montori and Guyatt, 2003). When similar individual studies are summarized, they can be pooled together and analysed statistically using a meta-analysis (Deeks, 2001; Gatsonis and Paliwal, 2006; Wright et al., 2007). Applications of recommended guidelines are useful to ensure good quality of both systematic reviews and meta-analysis (Moher et al., 2009). 2.2. Logistic regression analysis A range of statistical methods is available to analyze data from epidemiological studies according to the objectives. When epidemiological studies are aiming to demonstrate or identify relationships between different factors or variables of interest, regression methods are mostly used (Lewis and Michael, 2013). These methods are helpful to identify and describe potential associations that might exist between variables of interest such as the serological status of an animal (known as dependent response or outcome variable) and the sex and the age of this animal (independent predictive variables or explicative variables). When the outcome or the response variable is dichotomous such as disease status

39

Chapter 1 : Brucellosis : a literature review

(seropositive or seronegative), the logistic regression modeling was applied. It is therefore used to identify statistical associations among variables of interest and to identify variables that might be relevant for disease control (Lewis and Michael, 2013). Logistic regression is among the most important regression techniques in epidemiology (Stephen, 2001). It is the most appropriate modeling approach to describe and to test hypotheses about relationships between a categorical outcome and one or more categorical or continuous predictor variables. Dohoo et al. (2003) provided more discussions on logistic regression modeling. Briefly, estimations in logistic regression modeling are obtained through a Maximum Likelihood Estimation process. In this later approach, the coefficients in the model represent the amount of the logit of the probability of the outcome changes with a unit increase in the predictor variable. Since these coefficients are hard to interpret, they are commonly expressed as odds ratios (OR). When, the predictor variable is continuous, the OR represents the factor by which the odds of outcome variable are increased (or decreased) for each one-unit change in the predictive variable. When, the predictor is a categorical variable, the coefficient for each level of the variable represents the effect of that level compared to the category (i.e. the 'baseline') not included in the model (Dohoo et al., 2003). 2.3. Determination of disease status Determining the status of a given individual regarding a given condition could be helpful for many purposes. Knowledge of a disease status can be used to support decision for diagnosis or for treatment. Identification of the status of an individual in a given population could also be required to assess a new diagnostic test. Determining a disease status implies the use of appropriate diagnostic tools or tests. When the diagnostic test is able to determine the disease status of an individual with 100% accuracy, it is considered as the gold standard test. In practice, gold standard tests are seldom available due to many factors including the biological variability of each individual. Therefore, disease status has to rely on so-called “bronze” test or another test closest to the standard test. As an alternative to the absence of a gold standard test, a combination of tests could also be used to improve the diagnostic performance or obtain a gold standard effect (Black and Craig, 2002). This is often the case with brucellosis for which an unequivocal diagnosis can be made only with the isolation and identification of Brucella (OIE, 2009). However, isolation and identification methods as Brucella culture are not always available or feasible in common diagnostic conditions contrary to serological tests (Nielsen, 2002; Godfroid et al., 2010). In addition, the

40

Chapter 1 : Brucellosis : a literature review

probability of finding

Brucella spp. can decrease one month after the parturition as

previously assessed (Saegerman et al., 2004). When a combination of tests is used, results can be interpreted in series or in parallel according to the objectives of testing. With serial interpretation, only animals that give positive results to both tests are considered positive. Consequently, this increase the diagnostic specificity (Sp) and decrease the diagnostic sensitivity (Se). With parallel interpretation, animals that give a positive result to one of the tests or to both tests are considered positive. Conversely to serial interpretation, parallel testing increases Se and decreases Sp. Depending on the fact that diagnostic tests target the similar biological phenomenon or not, combined tests could be considered as either dependent or independent. The combined tests may be correlated if they measure or target a similar biological phenomenon such as immunoglobulins (Gardner et al., 2000; Dendukuri and Joseph, 2001). When two tests are combined, the presence of a positive dependence would respectively reduce the test sensitivity value in a parallel interpretation scheme and the test specificity value in a serial interpretation scheme compared with values expected if tests were conditionally independent (Gardner et al., 2000). More discussions on test dependence issues are provided in further sections. 2.4. Estimation of disease true prevalence and performance of diagnostic tests 2.4.1.

Estimation of disease true prevalence

Disease prevalence is a key parameter to assess the impact of a disease in the population of interest and for estimating the disease burden (Speybroeck et al., 2012a). To determine the actual level of a disease in a population of interest, the true prevalence needs to be estimated (Dohoo et al., 2003). Accuracy of true prevalence is related to performance parameters of tests to be applied (Ihorst et al., 2007). Assuming that the sensitivity (Se) and the specificity (Sp) of a given diagnostic test are known, and AP being the apparent prevalence resulting from the application of the test in the population of interest, the true prevalence (P) can be determined using the following formula (Rogan and Gladen, 1978):

41

Chapter 1 : Brucellosis : a literature review

The estimation of the disease true prevalence is then straightforward with this formula when a gold standard test is available or fixed and known values are assumed for test characteristics (Dohoo et al., 2003; Berkvens et al., 2006). In practice, fixed and known values assumption may be unrealistic. Moreover, a straightforward application of this formula may result into estimates exceeding 1 (one) or may yield negative values (Spybroeck et al., 2012a; Lewis and Torgerson, 2012). In addition, perfect reference tests are hardly ever available since diagnostic performance of any test is known to be influenced by several endogenous and exogenous factors and should be considered as context-specific parameters (Saegerman et al., 2004; Berkvens et al., 2006; Rutjes et al., 2007). All these elements imply that imperfect tests should be used for disease prevalence estimation. Additionally, the use of multiple imperfect tests for estimation is suggested to reduce misclassification errors. So, appropriate methods and assumptions should be used to get unbiased estimates (EnØe et al., 2000; Berkvens et al., 2006). 2.4.2.

Assessment of performance of diagnostic tests

The performance and the accuracy of all diagnostic assays need to be determined under routine conditions. This includes the estimation of parameters measuring the accuracy and the diagnostic performance of assays. While the reproducibility of a test measures the degree of agreement between test results when the conditions for testing or measurement change (e.g., two operators or laboratory technicians or two laboratories), the repeatability expresses the similarity of the test results in the same conditions (OIE, 2013). The robustness of an assay is another parameter referring to the assay’s capacity to remain unaffected by minor variations (e.g., pH, temperature of reagents, brand of microtiter plates) while using an assay in the single laboratory conditions. During this thesis work, agreement between test results was assessed using indexes of agreement as indicator and sensitivity and specifity as diagnostic test performance parameters. 2.4.2.1.

Indicators of agreement between tests

In order to increase the diagnostic performance, to assess a new diagnostic test or to evaluate test characteristics, two diagnostic tests can be used in combination. Different indicators can be used to assess the agreement between the results of the different tests. The most commonly applied is the kappa coefficient of agreement (K). It is the corrected index of

42

Chapter 1 : Brucellosis : a literature review

the agreement between the results of two diagnostic tests. It is calculated as the ratio of the observed excess over chance agreement to the maximum possible excess over chance. The kappa coefficient is equal to 1 (100%) when there is perfect agreement and it takes the value of zero when the observed agreement is equal to the chance agreement (Dohoo et al., 2003). However, the kappa coefficient is under the influence of the prevalence. Moreover, it was noticed that despite a high concordance between two tests, the kappa coefficient may paradoxally be low (Feinstein and Cicchetti, 1990; Cicchetti and Feinstein, 1990). To solve paradoxes with kappa coefficient, two indexes, e.g., the positive and negative index of agreement were proposed to measure the level of agreement between two tests (Cicchetti and Feinstein, 1990; Graham and Bull, 1998). These indexes represent respectively the observed agreement proportion for positive and negative results. Using the contingency table (Table VI), the two indexes of positive agreement (Ppos) and negative agreement (Pneg) are respectively: p pos 

2a 2a  b  c

and

pneg 

2d 2d  b  c

Where Ppos and Pneg are respectively the indexes of positive agreement and negative agreement; a, b, c and d are given in the contingency table. Table VI : Contingency table showing results for two diagnostic tests (Test 1 and Test 2) Test 1

Test 2

pos

neg

Total

pos

a

b

a+b

neg

c

d

c+d

Total

a+c

b+d

N

pos: positive result; neg: negative result 2.4.2.2.

Performance parameters of diagnostic tests

In both human and animal health, diagnostic tests are useful tools to determine the true disease status of an individual or a group of individuals in a population of interest. For a given disease, the accuracy of information on individual’s status depends on the performance of the applied diagnostic tests. In fact, the performance of a diagnostic test indicates its ability to correctly identify truly diseased from non-diseased individuals when applied in a randomly

43

Chapter 1 : Brucellosis : a literature review

chosen sample from a population of interest (Lewis and Torgerson, 2012). This ability is also an important point when evaluating a new diagnostic test and for implementing disease control programs since a correct classification of herds and individual animals regarding their status is looked-for (Greiner and Gardner, 2000). The actual level of a disease in a population of interest, i.e. the true prevalence, is also an essential parameter (Dohoo et al., 2003). Its estimate is helpful to assess the impact of a disease in the population of interest and to avoid biased estimation of disease burden (Speybroeck et al., 2012a). Accuracy of true prevalence depends on the performance parameters of the tests to be applied (Ihorst et al., 2007). The performance of a diagnostic test may be evaluated through several quantitative parameters including predictive values, likelihood ratios (LR), the area under the Receiver-OperatingCharacteristic (ROC) curve (AUC), the diagnostic odds ratio (DOR), Youden's index (J), and Se and Sp (Greiner and Gardner, 2000; Glas et al., 2003). These parameters are helpful to support decision-making while selecting a diagnostic test for a given context or purpose. Based on the contingency table below (Table VII), a summary of these different parameters of performance and their definitions are given in Table VIII (Glas et al., 2003). The predictive values express the probability of diseased animals (PPV) or nondiseased animals (NPV) among positive and negative results respectively. The likelihood ratios indicate the ratio of the expected result between animals with a disease and animals without that disease. The area under the ROC curve (AUC) is a plot of sensitivity against one minus the used specificity and is applied to measure the discriminative power of a diagnostic test. The Diagnostic Odds Ratio (DOR) is referred as the ratio of the odds of positivity in diseased animals compared to the odds of the same test result in non-diseased animals. Both DOR and the Youden’s index, which is a combination of sensitivity and specificity minus one, are significantly influenced by these two parameters. Among the indicators of performance, sensitivity and specificity are the most employed. Test Se (or Sp) indicates the probability that a truly infected (or non-infected) individual yields a positive (or a negative) test result. Similarly, when the epidemiological unit of concern is the herd, Se corresponds to the probability that an infected herd yields a positive herd-test result, and herd-level Sp (HSp) is the probability that a non-infected herd yields a negative herd-test result (Martin et al., 1992). Positive herd result might refer to the presence of at least one animal testing positive within this herd while negative herd result corresponds to absence of positive animals. However, the threshold number of positive animals that denotes the herd as positive needs to be determined.

44

Chapter 1 : Brucellosis : a literature review

Table VII: Contingency table showing results between a reference test and a given imperfect test (Test) Reference test

Test

Diseased

Not diseased

Positive

TP

FP

Negative

FN

TN

Where TP, FP, FN, TN are respectively the True positive, the False positive, the False negative and the True negative.

Table VIII: Definitions of commonly used performance indicators of diagnostic test (Glas et al., 2003) Test performance Parameters

Formula

Definition

Accuracy

(TP+TN)/(TP+TN+FP+FN)

Proportion of correctly identified subjects

Sensitivity (Se)

TP/(TP+FN)

Proportion of positive test results among diseased (true positive rate)

Specificity (Sp)

TN/(TN+FP)

Proportion of negative test results among the “healthy” (true negative rate)

Positive predictive value (PPV)

TP/(TP+FP)

Proportion of diseased among subjects with a positive test result

Negative predictive value (NPV)

TN/(TN+FN)

Proportion of non-diseased among subjects with a negative test result

Likelihood ratio of a positive test result (LR+)

Se/(1-Sp)

Ratio of a positive test result among diseased to the same result in the “healthy”

Likelihood ratio of a negative test result (LR-)

(1-Se)/Sp

Ratio of a negative test result among diseased to the same result in the “healthy”

Youden’s index (J)

Se +Sp-1

Difference between the true positive rate and the false positive rate 45

Chapter 1 : Brucellosis : a literature review

2.4.3.

Methods for estimating disease true prevalence and test sensitivity and

specificity 2.4.3.1.

Estimation at individual-level

As mentioned above, diagnostic tests are useful to detect the presence or evidence of the presence of an infection or a disease. This ability of a diagnostic test to detect a condition of interest is crucial for selecting appropriate control strategies. So, the performance of a diagnostic test is of key importance. However, since this performance might be influenced by several variables, appropriate methodologies are needed to get better estimates of performance parameters. Estimation of test performance parameters and the true prevalence are two mathematically identical situations, even if the parameter of interest might change according to study objectives (Lewis and Torgerson, 2012). When a gold standard test is available, the true status of an epidemiological unit of interest regarding a disease can be determined. As a result, performance parameters and true prevalence can be easily deduced from the Rogan-Gladen equation, putting into relation apparent prevalence (AP) and true prevalence ( ) with test sensitivity (Se) and specificity (Sp), as described previously (Rogan and Gladen, 1978). However, assuming a test is a “gold standard” test would mean that no classification errors exist and no false positive or negative result is possible. In practice, such a perfect reference test is hardly ever available since diagnostic performance of any test is known to be under the influence of several endogenous and exogenous factors (Rutjes et al., 2007). As an alternative to the absence or the unavailability of a gold standard test (Black and Craig, 2002), the use of multiple imperfect tests was suggested to facilitate estimation. With multiple tests, misclassification errors are reduced and expected to be lower compared to the application of a single imperfect test. Over the years, several authors have attempted to provide options or solutions to get better and unbiased estimates of the disease prevalence and test Se and Sp in different settings and particularly in absence of a gold standard reference test. These solutions were inspired by frequentist and Bayesian concepts, the two statistical approaches through which inference to the population is made (see section 2.4.3.3.).

46

Chapter 1 : Brucellosis : a literature review

2.4.3.2.

Estimation at herd-level

Test properties and prevalence estimation could also be considered at group or herd level. Indeed, for many diseases, control programs include groups of individuals or herd testing. Like for individual testing, herd level test performance parameters are crucial. Christensen and Gardner (2000) and Martin et al. (1992) discussed the evaluation of diagnostic tests at herd level. Assuming known individual test characteristics and a cut-off of at least one animal testing positive for a herd to be considered positive, herd Se (HSe) and Sp (HSp) are computed as in Thrusfield et al. (2005). Dohoo et al. (2003) provided a general formula covering the cases with more than one positive animal to consider the herd as positive. Group or herd level test characteristics were shown to be under the influence of different factors. These factors include individual level test Se and Sp, sample size, threshold number (or the percentage of positive tests that denote the herd, or group as test positive) and the within-herd prevalence (Martin et al., 1992). The threshold number of positivity or minimum within-herd prevalence is usually determined according to the epidemiology of the disease or specific national or international rules (Wagner and Salman, 2004). Usually, the presence of one animal testing positive within a herd would be sufficient to classify it as positive but more than one positive result could be necessary for some diseases (Pfeiffer, 2002). Herd size is also known to have a significant influence on herd test performance (Pfeiffer, 2002). Herd-level test performance estimation is comparatively less complicated when herd size in the population of interest is constant. Estimation becomes more challenging when the herd size varies. Then, the procedure for estimating the herd-level test performance needs to account for this variability by weighting estimated values using herd size. As for individual testing, a single test or a combination of test could be used for herd level testing to improve testing performance. Similar requirements as for individual Bayesian modeling also applies when herd is considered, with also the need of good quality prior information7.

When prior knowlegde provides information on the uncertainty of a parameter to be estimated, it is considered as an“informative prior”. Conversely, prior knowledge might be unavailable implying an absence of information on the uncertainty of a parameter to be estimated. This lack of knowledge can be included in the modelling process as a“non-informative prior”. 7

47

Chapter 1 : Brucellosis : a literature review

2.4.3.3.

Bayesian versus frequentist methods for estimating the disease

true prevalence and diagnostic test performance

This section constitutes a personal view published in The Veterinay Journal.

ARTICLE 1: SANOGO M., ABATIH E., SAEGERMAN C. Bayesian versus frequentist methods for estimating disease true prevalence and diagnostic test performance. Vet. J., 2014, 202, 204 207.

48

Chapter 1 : Brucellosis : a literature review

49

Chapter 1 : Brucellosis : a literature review

50

Chapter 1 : Brucellosis : a literature review

51

Chapter 1 : Brucellosis : a literature review

52

Chapter 1 : Brucellosis : a literature review

CHAPTER 3: OBJECTIVES OF THE THESIS

53

Chapter 3: Objectives

CHAPTER 3: OBJECTIVES OF THE THESIS In many developing countries such as Ivory Coast, the development of livestock and the improvement of their health environment is part of the fight against poverty and for food security of populations. Developing livestock production and productivity imposes dealing with many constraints including pathological ones such as brucellosis. In addition to its impact on animal’s health, brucellosis is also one of the widespread zoonotic diseases. In developing countries and more precisely in Africa, this disease is endemic and known to be among the pathologic constrains to the development of livestock. Despite its known negative socio-economic impact and zoonotic potential, the disease is not considered yet as a priority disease and therefore remains neglected, underreported and uncontrolled in many countries (Mableson et al., 2014). In Africa, bovine brucellosis is the most widespread form among animals. In Ivory Coast, this form was recognized as one of the dominant pathologies and is argued to be responsible for the loss of about 10% of the annual income of the livestock breeders (Angba et al., 1987). In this country, investigations were conducted on bovine brucellosis throughout the years to determine its incidence and for a better knowledge of its epidemiology (Gidel et al., 1974; Pilo-Moron et al., 1979; Camus, 1980a; Thys et al., 2005). But similarly to many low-resource countries, these investigations are still few and their results are outdated particularly on the actual distribution of the disease, on the transmission within and across species and the impact on human and animal health, precluding the development of prevention and control strategies (Marcotty et al., 2013). The general objective of the research presented in this thesis is to improve the knowledge on the epidemiology of bovine brucellosis in Ivory Coast permitting future strategic actions. In this respect, different aspects of the disease were studied including prevailing strains causing brucellosis in cattle, performance of diagnostic test for brucellosis, estimation of true disease prevalence and identification of risk factors associated with brucellosis seropositivity in cattle from Ivory Coast (Figure 9). The specific objectives of the research are as follows: 1) Identification and isolation of the causal agent remains the ultimate evidence of the presence of the disease. The demonstration of Brucella as causal agent of brucellosis may

54

Chapter 3: Objectives

be done through various methods including bacteriological and molecular methods. Knowledge on prevailing field strains of Brucella in the particular context of Ivory Coast is useful to elaborate and set up appropriate preventive and control measures against brucellosis. In effect, data on prevailing field strain would be useful and critical to select the appropriate antigen for serological assay, to determine intra-species and interspecies transmission and to assess the potential risk of human infection. Consequently, one of the specific objective of this research was to investigate circulating species and biovars of Brucella associated with cattle in Ivory Coast (Chapter 4 and 5) aiming to provide an overview, to determine their geographical distribution and discuss public health implications (Chapter 4). 2) Diagnostic test are key components for disease-control programs since they are useful for classifying individuals according their serological status. Estimates of performance parameters are also useful to assess the impact of a disease in a given population, through accurate estimates of true prevalence. Thus, since the performance of diagnostic tests are under the influence of the population in which they are applied (including prevailing disease-causing agent), the diagnostic sensitivity, specificity and the true prevalence were estimated for bovine brucellosis in the Ivorian context (Chapter 5 and 6). The performance of the diagnostic tests was also discussed in light of circulating field species and biovars of Brucella, regarding particularly the appropriateness of the antigen used in serological assays. The true prevalence of brucellosis was estimated in Ivory Coast using a Bayesian approach, a statistical methodology allowing the combination of many testing results for accurate estimates. 3) Identification of risk factors related to the presence or the spread of the disease are useful to adjust preventive and control measures. So, another specific objective of this research was to investigate the potential risk factors associated with bovine brucellosis in Ivory Coast (Chapter 7).

55

Chapter 3: Objectives

Identify field species and biovars of Brucella in cattle from Ivory Coast

Assess the performance of diagnostic tests in the Ivorian context

Estimate disease true prevalence in Ivory Coast

Identify risk factors associated with bovine brucellosis in Ivory Coast

Figure 9: Schematic summary of the main objectives of the thesis

56

PART TWO: EXPERIMENTAL SECTION

57

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

CHAPTER 4: DETERMINATION OF SPECIES AND BIOVARS OF BRUCELLA INFECTING CATTLE POPULATION IN IVORY COAST

58

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

4.1. Introduction Identification and characterization of the causative agent of an infectious disease is important to consider for epidemiological studies, management of outbreaks and to identify potential source of human infection (Saegerman et al., 2010; Godfroid et al., 2010). Additionnaly, it allows to obtain data on possible interspecies and intra-species transmission of Brucella. Knowledge of prevailing species and biovars of Brucella infecting the livestock is a crucial prerequisite to the formulation of strategies for the prevention and the control of brucellosis in animal populations (Ocholi et al., 2004). Aiming to contribute to the knowledge on prevailing strains of Brucella in Ivory Coast and in West Africa, a summary and some updates were provided in this thesis. 4.2. Prevailing species and biovars of Brucella in cattle and their implications This section constitutes the following original paper published in Veterinary Microbiology.

ARTICLE 2: SANOGO M., ABATIH E., THYS E., FRETIN D., BERKVENS D., SAEGERMAN C. Importance of identification and typing of Brucellae from West African cattle: a review. Vet. Microbiol., 2013b, 164, 202–211.

59

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

60

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

61

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

62

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

63

Chapter 4: Determination of species and biovars of Brucella infecting cattle in Ivory Coast

64

65

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

66

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

67

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

68

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

69

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

CHAPTER 5: PERFORMANCE OF DIAGNOSTIC TESTS FOR BOVINE BRUCELLOSIS IN IVORIAN EPIDEMIOLOGICAL CONTEXT

70

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

5.1. Introduction Sufficient knowledge on species and biovars of Brucella at national and regional scales are important to set up and implement efficient control measures against brucellosis. From data on circulating field strains, the appropriateness of the antigen used in serological tests can be verified. This appropriateness is of key importance since the detection of the presence or evidence of the presence of an infection or a disease such as brucellosis is dependent on diagnostic tests. The ability of a diagnostic test to detect a condition of interest can be measured through performance indicators such as sensitivity and specificity. Since test sensitivity and specificity are known to be under the influence of several variables, an appropriate methodology is needed to get accurate and unbiased estimates and to get knowledge on the actual impact of the disease among a population of interest. In this chapter, typing results were put in relation with the epitope used in the serological tests applied to assess their appropriateness. Then, a Bayesian approach was implemented to determine the performance of two commonly used diagnostic tests for the diagnostic of bovine brucellosis in Africa, the Rose Bengal Test and the indirect enzyme linked immunosorbent assay. For representativeness, data from two surveys were combined for the analysis as a single population. Indigenous cattle of Bos indicus type, Bos taurus type and their crossbred, more than one year old were included in this study. Hygroma fluid collected from a carpal hygroma was used as sample for biotyping.

5.2. Bayesian estimation of true prevalence, sensitivity and specificity of Rose Bengal Test and indirect ELISA for the diagnosis of bovine brucellosis in Ivory Coast This section constitutes the following original paper published in The Veterinary Journal.

ARTICLE 3: SANOGO M., THYS E., ACHI Y.L., FRETIN D., MICHEL P., ABATIH E., BERKVENS D., SAEGERMAN C. Bayesian estimation of true prevalence, sensitivity and specificity of Rose Bengal test and indirect ELISA for the diagnosis of bovine brucellosis. Vet. J., 2013a, 195, 114-120.

71

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

72

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

73

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

74

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

75

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

76

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

77

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

78

Chapter 5: Performance of diagnostic for the diagnostic of bovine brucellosis in Ivory Coast

79

CHAPTER 6: TRUE PREVALENCE OF BOVINE BRUCELLOSIS IN IVORY COAST

80

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

6.1. Introduction The actual level of a disease in a population of interest, i.e. the true prevalence, is an essential parameter to assess the impact and importance of a disease in the population of interest and to avoid biased estimation of disease burden. Accuracy of true prevalence is related to performance parameters of tests to be applied. Estimation of test performance parameters and the true prevalence are two mathematically identical situations, even if the parameter of interest might change according to study objectives. By definition, estimation of sensitivity and specificity of a diagnostic test requires knowledge of the true disease status of animals on which this test is applied. This status is given by a reference test, which might be a “gold standard” test. In the absence of a “gold standard” test, a combination of available imperfect tests may be used for estimation. An appropriate methodology being needed for accurate estimation, a Bayesian approach was used to estimate the true prevalence of bovine brucellosis in the centre of Ivory Coast.

6.2. Prevalence of bovine brucellosis in Ivory Coast This section constitutes the following original paper published in La Revue d'élevage et de médecine vétérinaire des pays tropicaux.

ARTICLE 4: SANOGO M., CISSE B., OUATTARA M., WALRAVENS K., PRAET N., BERKVENS D., THYS E. Etude la prévalence de la brucellose bovine dans le centre de la Côte d'Ivoire. Rev. Elev. Med. vet. Pays trop., 2008, 61 (3-4), 147-151.

81

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

82

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

83

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

84

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

85

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

86

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

CHAPTER 7: RISK FACTORS ASSOCIATED WITH BOVINE BRUCELLOSIS SEROPOSITIVITY IN IVORY COAST

87

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

7.1. Introduction In addition to knowledge on circulating field strains of Brucella and availability of adequate diagnostic tests, identification of potential risk factors associated with the disease is also useful for developing and implementing preventive and control measures. Such knowledge might be useful to increase the awareness of farmers, and regulating herd management practices with the ultimate aim of to decrease the prevalence of brucellosis among livestock in Ivory Coast.

7.2. Risk factors associated with brucellosis seropositivity among cattle in the central savannah-forest area of Ivory Coast

This section constitutes the following original paper published in Preventive Veterinary Medicine.

ARTICLE 5: SANOGO M., ABATIH E., THYS E., FRETIN D., BERKVENS D., SAEGERMAN C. Risk factors associated with brucellosis seropositivity among cattle in the central savannah-forest area of Ivory Coast. Prev. Vet. Med., 2012, 107(1-2), 51-56.

88

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

89

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

90

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

91

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

92

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

93

Chapter 6: Estimation of true prevalence of bovine brucellosis in Ivory Coast

94

PART THREE: GENERAL DISCUSSION

95

Chapter 8: General discussion, conclusions and perspectives

CHAPTER 8: GENERAL DISCUSSION, CONCLUSIONS AND PERSPECTIVES

96

Chapter 8: General discussion, conclusions and perspectives

CHAPTER 8: GENERAL DISCUSSION, CONCLUSIONS AND PERSPECTIVES 8.1. General discussion The need to ensure a sustainable development of livestock, to fight poverty and to limit the public health impact of neglected zoonotic diseases as brucellosis, imposes to give consideration to these diseases in low income countries. In West African countries including Ivory Coast, bovine brucellosis is known for many years and evidence was already provided on the benefit to implement control measures against this disease (Camus, 1995; Roth et al., 2003). However, there is still a lack of attention for brucellosis, hindering any evidence-based control measures in most countries. In addition to the need for sufficient and reliable data for a better understanding of its epidemiology, updated knowledge on the disease appeared to be essential in Ivory Coast. Indeed, the country has recently suffered from political instability causing the disorganization and inadequate coverage of veterinary services that are in charge of the animal disease control activities. This situation could have favored the emergence of animal diseases, especially zoonotic ones like brucellosis (Roth et al., 2003). Our research aimed to improve the knowledge on the epidemiology of bovine brucellosis in Ivory Coast. Through this general objective, the research intended to generate useful information, which could be used to prevent the spread of the disease and to document national or regional (future) preventive and control plans and strategies against brucellosis, especially in cattle. Therefore, this research includes different contributions intending to cover the main aspects of epidemiology of the disease as defined by Carr et al. (2007): -

The distribution and frequency of bovine brucellosis and evidence of its presence in Ivorian cattle; involving specifically: o The identification and typing of prevailing field strains of Brucella in cattle; o The assessment of field serological diagnosis test for the diagnostic of bovine brucellosis; o The estimation of the true prevalence of the disease in cattle;

-

The determination of association of brucellosis with other factors, including the identification of potential risk factors associated with brucellosis seropositivity in cattle.

97

Chapter 8: General discussion, conclusions and perspectives

The following sections of the thesis present a general discussion of the research with emphasis on the limitations. More details discussions are included within the different articles composing the second part of the thesis. 8.1.1.

Importance of identification and typing of prevailing strains of

Brucella in cattle As recentlty expressed by Godfroid et al. (2013), accurate knowledge on Brucella spp in both human and the different animal species is needed to identify the source of infection and plan appropriate control measures. In this research, a comprehensive summary of species and biovars of Brucella reported in West African countries with their proportion per origin and their geographical distribution was obtained by applying a systematic review approach (See Chapter 4). So far, B. abortus biovar 3 was found to be the most commonly isolated in cattle in the sub-region while the biovar 1 is considered as the most encountered worldwide (Corbel, 1997), and also in the USA (Bricker et al., 2003) or in Latin America (Acha and Szyfres, 2003). B. abortus biovar 3 was also predominantly isolated in both native cattle and buffalo in eastern Africa and China (Timm, 1982, Domenech et al., 1983). Despite the usefulness of the global map provided on what is known so far about the prevailing field strains of Brucella in cattle in West Africa, this review cannot be assumed as exhaustive and representative of the actual situation. However, it provides an insight on the status of field strains at both national and regional level what is useful considering the frequent and uncontrolled cattle movement (transhumance) between countries. On West African scale, data on prevailing Brucella in cattle have been lately gathered in The Gambia (Bankole et al., 2010), in Niger (Boukary, 2013), in Ivory Coast (Sanogo et al., 2013a) and in Togo (Dean et al., 2014). Similar research initiatives need to be encouraged for more updated and extended data on prevailing field strains of Brucella in West Africa. Except for the Ivorian isolate, which appeared to be negative at oxidase test, the five late B. abortus biovar 3 from West Africa showed identical growth characteristics. Using enhanced molecular typing methods, they showed some dissimilarities despites their classification in the same biovar. The isolates from the Gambia and Niger apparently closer genetically, seem to be more distinct from those of Ivory Coast and Togo (Table IX and Figure 10). This provides some indications on the genetic diversity of circulating strains of Brucella in this sub-region (Dean et al., 2014) and the need for further typing results.

98

Chapter 8: General discussion, conclusions and perspectives

Table IX: The Multiple Loci Variable Number Tandem Repeats analysis (MLVA) profiles showing number of variable tandem repeats (VTR) for latest west African isolates of B. abortus biovar 3 and their closest MLVA neighbour profile (B. abortus biovar 3 strain BCCN 93_26 from in Sudan, B abortus biovar 3 reference strain Tulya from Uganda and B. abortus biovar 6 strain BfR7 from Chad) in the Brucella MLVAbank (from Bankole et al., 2010, Sanogo et al., 2013a and Boukary et al., 2013, Dean et al., 2014)

Panel 1

Variable Reference Strain Strain The Togo_isolate Togo_isolate Togo_isolate tandem Strain BCCNa IVC_isolate Niger_isolate BfR7b gambia_isolate 1 2 3 repeats Tulya 93_26 bruce06 3

3

3

3

3

3

3

3

3

bruce08 5

5

5

5

5

5

5

5

5

bruce11 5

5

5

4

3

4

3

3

3

bruce12 11

11

11

11

11

11

11

11

11

bruce42 2

2

2

2

2

2

2

2

2

bruce43 2

2

2

2

2

2

2

2

2

bruce45 3

3

3

3

3

3

3

3

3

bruce55 3

3

3

3

3

3

3

3

3

bruce18 8

6

6

7

8

7

10

8

8

bruce19 -

-

-

21

21

-

41

41

41

bruce21 8

8

8

8

8

8

8

8

8

bruce04 6

6

6

4

6

5

4

4

4

bruce07 5

8

4

5

2

5

2

2

2

bruce09 3

3

3

3

3

3

3

3

3

bruce16 11

7

8

7

12

3

8

5

6

bruce30 5

7

4

3

7

5

4

4

4

Panel 2

c

a

Brucella Culture Collection; bFederal Institute for Risk Assessment; c additional locus comprised in

the MLVA-16 and absent in MLVA-15, Isolates from Ivory Coast (IVC_isolate), Niger (Niger_isolate) , The Gambia (The Gambia_isolate) and Togo (Togo_isolate 1 , 2 and 3).

99

Chapter 8: General discussion, conclusions and perspectives

Figure 10: Dendrgram showing the relation between the latest isolates of B. abortus biovar 3 in West Africa and also with neighbour reference strains in the Brucella MLVAbank (B. abortus biovar 3 strain BCCN 93_26 from in Sudan, B abortus biovar 3 reference strain Tulya from Uganda and B. abortus biovar 6 strain BfR7 from Chad) (Bankole et al., 2010; Sanogo et al., 2013a; Boukary et al., 2013, Dean et al., 2014). It is built from results of a simple linkage cluster analysis of the number of variable tandem repeats (VNTR) and the dissimilarities between strains is measured through the eucludian distance between VNTRs (L2 dissimilarity measure). Regarding Ivory Coast, our study is significant as it was the first report on biovar 3 of B. abortus in the country since the first evidence of brucellosis was made (Figure 11). In this country, only B. abortus biovar 1 and 6 had been isolated from cattle so far. The new biovar was identified from hygroma fluid samples collected from a cow with a carpal hygroma in the central part of the country (Chapter 5). This result stresses the importance and the need to continue the efforts to identify circulating field strains in Ivory Coast and at a broader extent, in other West African countries. Until now, only 18 isolates were reported in Ivory Coast since the first report in the early 1970s (Figure 11). This is far less compared to the

100

Chapter 8: General discussion, conclusions and perspectives

number of isolates in Senegal (n=232), in Nigeria (n=46), and in Togo (n=30) (Chapter 4). All biovars of Brucella reported so far in Ivory Coast (B. abortus biovar 1, 3 and 6) are characterized by the same “A” epitope” used in the applied serological tests (i.e., RBT and iELISA). This provides an indication on the appropriateness and the adequacy of the serological tests used so far. However, there is a need for more investigations and more data on prevailing strains of Brucella to support this assumption.

Figure 11: Mapping of field strains of Brucella in cattle in Ivory Coast, 2013 (from PiloMoron et al., 1979; Sanogo et al., 2013a; Sanogo et al., 2013b). Each bubble contains information on the name of the locality of origin of the strain (e.g. Eloka), the year of publication (e.g. 1979),

the biovar (e.g. B. abortus 1) and the number of isolates

identified (eg. n=2).

101

Chapter 8: General discussion, conclusions and perspectives

The presence of Brucella in cattle being confirmed again almost twenty years after the first isolates, it confirms the existence and the persistence of a potential risk for the human population. Indeed, the risk cannot be precluded in the West African epidemiological context where i) close contact may occur between animals and people, particularily in urban and periurban areas; ii) hygienic conditions are usually poor; iii) customs often favour consumption of raw milk, and iv) where no prevention and/or control strategies are sustainabily implemented. However, more data on human cases are needed to clearly establish the public health importance of the disease. As a starting point, seropositivity among slaughterhouse workers and other high-risk professionals might be investigated combined with isolation and characterization of Brucella. Reporting of the disease in human could also be improved by considering brucellosis as part of the differential diagnosis for patients with fever of unknown origin (FUO), fever being the most common clinical features in human (Franco et al., 2007). The comprehensive review of prevailing strains in the field also reveals the frequent isolation of strains of B. abortus with unusual characteristics in this sub-region of West Africa (in Senegal, Togo, Niger, The Gambia as well as in Ivory Coast). With conventional typing methods, the differences were not always clear for some of these strains, complicating their classification. The existence of these strains should be considered when typing field strains of Brucella in West Africa. This may also justify the need for more typing in the region and, wherever possible, for the application of more accurate discriminative methods (e.g. MLVA) in addition to conventional biotyping (Bankole et al., 2010; Sanogo et al., 2013a, Dean et al., 2014). In Ivory Coast the difference between biovar 6 and 3 being not always clear, the availability of more discriminative methods would have been very useful. Identification and typing of Brucella strains must continue and be maintained. This type of research will also provide information on possible sources of human infection and on transmission pathways between animals and humans. This is a step needed for an appropriate prevention and control of brucellosis (Adone and Pasquali, 2013). Additionally, the introduction of more advanced methods for identification and typing of Brucella such as the Variable number tandem repeat (VNTR) typing or Multilocus Sequence Analysis (MLSA) should be considered in a regional or continental control strategy for cost-effectiveness.

102

Chapter 8: General discussion, conclusions and perspectives

8.1.2.

True prevalence, sensitivity and specificity of serological assays for the

diagnostic of brucellosis in Ivory Coast Information on the prevailing field strains of Brucella is also important to select the adequate serological diagnosis tests. In addition to information on the actual presence of the disease, serological diagnostic tests are essential to discriminate the status of individuals or group of individuals. Tests are useful for understanding the disease epidemiology and for informing on possible preventive and control programs. However, none of the tests detecting Brucella is perfect and sources of interferences exist with many others Gram negative bacteria due to the presence of similarities with immunodominant antigen used (Saegerman et al., 2004). Vaccination with strain S19 is also responsible for serological cross-reaction (Corbel, 2006). Therefore, these possible sources of interferences should be taken into account while interpreting serological results (Robinson, 2003). In the epidemiological context of Ivory Coast, vaccination is no more officially practiced since 1992 and no official control program exists so far, excluding therefore interference due to vaccination. As described in our literature review, most serological assays commonly used for the diagnosis of bovine brucellosis use the B. abortus 1 antigen derived from the strain Weybridge 99, epitope A. As a consequence, the performance of assays will also depend on the prevailing field strains in the epidemiological context in which they are applied. The performance of two serological tests commonly used for the diagnostic of brucellosis and also prescribed for trade by OIE (Nielsen, 2002; Saegerman et al., 2004; OIE, 2009; Godfroid et al., 2010; Sanogo et al., 2013a) were assessed in the epidemiological context of Ivory Coast. Ideally, the sensitivity and the specificity of a diagnostic test require knowledge on the true disease status of the population in which the test is applied. This implies, in turn, the availability of a “gold standard reference test” which is absent. Therefore, sensitivity and specificity of Rose Bengal Test and indirect Enzyme Linked Immunosorbent Assay were determined in our study using a Bayesian approach (Chapter 5). By offering the possibility to combine prior or expert knowledge on parameters and actual field data in the same model, the Bayesian approach helps to have more accurate and reliable estimates in absence of a gold standard. However, accuracy and validity of Bayesian estimates depend on the availability and the quality of prior information included in the estimation and on the validity of the protocol (e.g. conditional dependence between tests). In developing countries like Ivory Coast where good priors are lacking, their influence has to be checked using a set of prior distributions, as done in this work. An accurate estimation also requires a representative 103

Chapter 8: General discussion, conclusions and perspectives

sample of the target population including all age categories and, ideally the different stages of the disease. In our work, two datasets were combined to improve the representativeness. From a geographical point of view, the representativeness of the aggregated sample for the whole country might be questionable, since only serum collected in cattle herds from the Southern and central regions of the country was used for estimation. However, no association was demonstrated between the origin of sera and the serostatus of cattle included. Moreover, the combined sample included sedentary as well as extensively managed herds with different herd size, age and sex categories, and all types of breed of the country. This allows us to reasonably consider that the aggregate sample was matching the characteristics of the overall cattle population of Ivory Coast. The provided estimates can therefore serve as prior knowledge for future Bayesian estimation of test characteristics and of disease true prevalence in similar conditions in Ivory Coast. In addition to the provision of estimates of test characteristics, the Bayesian approach also delivered an updated estimation on the true disease prevalence in the sample population. Except the latest studies, most of the previous reports on the prevalence of brucellosis in Ivory Coast, reported only apparent prevalences. Since the prevalence is essential to appraise the impact of a disease in a population of interest, the estimation of the disease true prevalence is of key importance to prevent a biased estimation of disease burden (Dohoo et al., 2003; Speybroeck et al., 2012a). Therefore, the usefulness of methodological options such as the Bayesian approach is obvious. Using a combination of three serological tests, the Bayesian estimates in the central savannah-forest area of the country was 8.8% (credibility interval: 5.0-16.4) (Chapter 6). An overall estimate of the true prevalence of brucellosis using aggregated samples from cattle herds from both central and southern parts of the country was 4.6% (credibility interval: 0.6-9.5) (Chapter 5). Even if the sampling strategy for the two datasets used in these estimations needs to be taken into consideration, these results provide useful indications on the presence of Brucella and the spread of the disease in cattle, justifying the attention that should be given to brucellosis in Ivory Coast. In addition to the sampling bias, the accuracy of the first estimates (Chapter 6) is more questionable since the correlation between the tests on seropositive and seronegative cattle was not taken into account in the modeling process. Indeed, the combination of diagnostic tests targeting a similar biological phenomenon -such as immunoglobulins- may result in dependence or correlation between them (Gardner et al., 2000). According to the conjugate used, conditional dependence had to be considered between RBT, detecting the presence of Immunoglobulins

104

Chapter 8: General discussion, conclusions and perspectives

(Ig) IgG1 and iELISA, targeting IgG1 and/or IgG2 (Nielsen, 2002; Saegerman et al., 2004 and 2010; Sanogo et al., 2013a). Moreover, the sensitivity analysis to assess the consistency of estimates was not performed as recommended (Branscum et al., 2005). Therefore, it appears that an appropriate sampling strategy should be designed at the beginning of any study aiming to estimate true prevalence or test characteristics using a Bayesian approach and the dependence between tests and the implementation of a sensitivity analysis on estimates are crucial to facilitate extrapolation of estimates. Since the performance of RBT and iELISA was evaluated in the Ivorian context, they can be used to support decision making for control and serosurveillance. Different testing strategies can be considered. Following a serial interpretation, RBT positive results have to be confirmed by iELISA while following a parallel interpretation the testing scheme will be expected to detect both acute and chronic infection (Saegerman et al., 2004; Godfroid et al., 2013). As demonstrated in this research, the iELISA might also be implemented on its own. However, a combination with other serological tests as the RBT would be a more appropriate strategy. Since most of the serological tests such as RBT and iELISA do not inform on the source of infection, the capabilities of veterinary services for the detection of Brucella need to be improved. In the short-term, identification methods such as the Matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF-MS) might be introduced to identify Brucella at a genus level at the central laboratory, with an enhancement of biosecurity and biosafety measures. 8.1.3.

Risk factors associated to the seropositivity of brucellosis in cattle from

Ivory Coast Many factors including the density of the animal population, herd size, breed, type of production (dairy or beef), type of husbandry system and environmental factors are thought to be associated with brucellosis. In this research, potential risk factors associated to seropositivity in cattle populations in Ivory Coast were also examined (Chapter 7). This is the first work reporting information on risk factors associated with brucellosis in Ivory Coast. Serological data obtained through a survey in the central savannah-forest area of the country were analyzed using a logistic regression. Age of animals and herd size were identified as risk factors associated with seropositivity of brucellosis in Ivory Coast. Specifically, animals above 5 years of age were estimated to be almost 3 times more likely infected than 3 years old cattle. A similar ratio was observed for cattle herds with more than 100 heads compared

105

Chapter 8: General discussion, conclusions and perspectives

to those with less than 50 heads. This information can be used for increasing farmers awareness and regulating herd management practices in order to decrease the seroprevalence of Brucella in animals and consequently prevent human infections. Indeed, education on risk factors associated with brucellosis is essential to limit the spread of the disease. These results were obtained using only data collected in the central soudano-guinean area but could be reasonably extrapolated to the whole country, as there is currently no control strategy against brucellosis at all. In addition to the identification of risk factors at animal level, which was the focus of this work, environmental risk factors need to be investigated as they may also provide information to elaborate prevention and control strategy. The results are in line with the findings of Akakpo and Bornarel (1987) who identified the age of animal, the type of breed and the climate as risk factors for brucellosis. It was also demonstrated that the effect of crowd (e.g. large herd size) together with lower genetic diversity may favor transmission and select fast replicating organisms with major zoonotic potential as Brucella (McDaniel et al., 2013). Boukary et al. (2013) recently also identified the age of animal as individual risk factor in traditional cattle farms from Niger. Together with the acquired knowledge on the prevailing field strains of Brucella (Chapter 4 and 5), the data gained on diagnostic tests performance (Chapter 5), disease true prevalence (Chapter 5), and risk factors (Chapter 7) are helpful to develop and implement preventive and control measures. 8.2. Conclusions, implications and perspectives The development of livestock and the improvement of their health status are an essential part of a pro-poor enhanced food security strategy for the benefit of vulnerable populations in developing countries such as Ivory Coast. This urges to deal with pathologic constrains like brucellosis. Bovine brucellosis is endemic in many sub-Saharan African countries including Ivory Coast. Its impact on animal production and zoonotic potential are currently well known and the benefits of controlling it was also strongly demonstrated in cattle. However, the disease is still considered as a non-priority disease, i.e. suffering from insufficient knowledge on its epidemiology and public health importance. Therefore, gaining more and accurate knowledge on the epidemiology of brucellosis is required to determine the actual impact of the disease.

106

Chapter 8: General discussion, conclusions and perspectives

It will help to convince the decision makers to implement appropriate and sustainable disease preventive and control measures at national level but also at regional level considering the frequent transboundary herd movements (transhumance). The current research confirms the presence of bovine brucellosis in Ivory Coast and contributes to the knowledge of the epidemiology of the disease. This research investigates prevailing field strains of Brucella in cattle in Ivory Coast but also provides information at West African scale. Biovar 3 of Brucella abortus was identified for the first time in Ivory Coast in cattle in this research. Additionally, the performances of Rose Bengal test and iELISA were assessed in the Ivorian epidemiological context, since those tests are of key importance for investigating the epidemiology of the disease as well for planning prevention and control measures. Finally, Estimates of the true prevalence of the disease are now available and some risk factors associated with brucellosis in the country identified for the first time. Initially, this research intended to cover the different agro-ecological areas of the country including the northern part where the density of cattle population and the presence of transhumant herds are expected to influence the disease epidemiology. Finally, only the southern Guinean and the central Soudano-guinean aeras were covered. This was mainly due to the socio-political context prevailing in the country at the moment of the field study. Data obtained in the accessible areas where combined with previous data collected in the same aeras for true prevalence estimation and for diagnostic performance assessement. The rationale supporting this approach and the consequences on the interpretation of the findings are discussed in Chapter 5 and 6. The contribution on prevailing strains of Brucella in cattle was done by combining a prospective and a retrospective approach. In the prospective approach, only one isolate was obtained, which was very few despite the added value that it provides. However, this stresses the need for more investigation in field strains of Brucella in cattle. Even if hygroma fluid was demonstrated to be a useful sample for strain identification (Sanogo et al., 2013a), other samples such as abortive materials and secretions should be considered. By considering exchanges and movements of cattle within Ivory Coast but also between countries of West Africa, the review provided a more extended picture on the prevailing strains. Despite its limitations, our research contributes largely to a better knowledge of the epidemiology of brucellosis in Ivory Coast and in West Africa. Additional investigations are needed to obtain a global picture and a reliable understanding of the disease epidemiology.

107

Chapter 8: General discussion, conclusions and perspectives

This is crucial to provide useful evidences to advocate among decision makers for adequate preventive and control measures against brucellosis. It will therefore be crucial to investigate the frequency and the distribution of the disease and the associated risk factors in other regions of Ivory Coast, especially in the northern part where the density of cattle population is higher and where animal movements (transhumance) might influence the disease epidemiology. Compared to the southern and central areas, the distribution and the frequency of the disease in the North might be higher. Information about the presence of Brucella in the different livestock breeding areas and systems of the country are essential to implement effective and appropriate prevention and control measures. The role of small ruminants as source of infection for cattle and for human also needs to be addressed. This is important to assess the risk of human brucellosis within the country. Furthermore, other susceptible livestock and wildlife species need to be studied in order to obtain a more extended picture of the disease epidemiology at national and at regional level. In addition to the need of future research for a better understanding of the epidemiology of bovine brucellosis in Ivory Coast as well as in West Africa, this research inspired some points, which need to be considered for an efficient and sustainable prevention and control of brucellosis as well as other (zoonotic) diseases: 

The diagnostic and surveillance capacities of veterinary services need to be strengthened to provide valuable epidemiological information, especially on prevailing strains of Brucella. Hence, improvement of veterinary diagnostic laboratory capabilities, veterinary surveillance and quality and organization of veterinary services are fundamental to provide reliable data, gain of confidence in the veterinary services and disease surveillance, and to ensure the efficiency of the preventive and control programs of brucellosis as well as other zoonoses. So far, RBT is routinely used for screening in Ivory Coast. Additional tests such as indirect or competitive ELISA and FPA also need to be assessed and established as confirmatory tests in the central veterinary laboratory in Ivory Coast. There is also a need to upgrade the laboratory facilities and equipements for safe management of samples possibly contaminated by level 3 pathogens as Brucella. In addition, a comprehensive training on biosafety and biosecurity measures and procedures is also important for laboratory workers, for scientists and for all the persons at risk or working with such hazardous pathogens.

108

Chapter 8: General discussion, conclusions and perspectives



Based on updated information provided on the epidemiological status of bovine brucellosis, a pilot control project covering the savannah-forest sedentary cattle herds and dairy herds in Ivory Coast can be suggested to lower the prevalence (down to 2%). This strategy might include prevention and control measures such as surveillance of dairy herds at national scale through milk testing (at least twice per year), annual vaccination of young calves, seromonitoring of herds (with RBT and iELISA or FPA or competitive ELISA), elimination of infected adult animals from herd and serological control before introduction in the herd. Campaigns of public awareness and education aiming to prevent and reduce risks of transmission from animals to humans are also imperative to sensitize on safe herd management practices and to improve notification of cases of abortion and hygroma. Consideration of the risk factors identified so far in the Ivorian context is helpful to prevent the spread of the disease within cattle population and from cattle to human. This control strategy will be progressively expanded to other livestock breeding areas of the country, but need to be backed by sufficient and updated knowledge on the disease epidemiology (e.g., true prevalence), in line the stepwise progressive approach proposed by FAO (FAO, 2013).



The maintenance and the improvement of animal health depend not only on financial issues but also on capacity, quality, competence, transparency, expertise and organization of veterinary services. Hence, the results and recommendations of the OIE Performance of Veterinary Services (PVS)8 assessement are fundamental and need to be considered. Especially, the adequate coverage of the territory with operational veterinary services (both public and private) is required in Ivory Coast after the socio-political crisis of the last ten years. This is of key importance to set up a functional and sustainable surveillance network, for reporting animal diseases and

The OIE Tool for the Evaluation of Performance of Veterinary Services (OIE PVS Tool) is an evaluation tools, developed initially in collaboration with the Inter-American Institute for Cooperation on Agriculture (IICA) and refined by the OIE, aiming to strengthen the veterinary Services by helping them comply with OIE international standards for quality. It is designed to assist VS to establish their current level of performance, to identify gaps and weaknesses in their ability to comply with OIE international standards, to form a shared vision with stakeholders (including the private sector) and to establish priorities and carry out strategic initiatives for improvement. More information are available on: http://www.oie.int/en/support-to-oie-members/pvs-evaluations/oie-pvstool/ 8

109

Chapter 8: General discussion, conclusions and perspectives

disease related events and for designing appropriate and efficient prevention and control measures. In addition to the effective implication and cooperation with community members (e.g., head of community, paraveterinarians, members of cooperatives, animal owners), the introduction of new tools such as internet and mobile technology might contribute to improve the efficiency of the epidemiological surveillance network especially regarding field data collection, notification of cases of abortion and transmission of reports. The commitment of the government is also essential to guarantee the sustainability of such a system (Ouagal et al., 2008; Ouagal et al., 2012). 

The presence of Brucella in most of the West African countries, the existence of cattle movement between countries and the limited resources allocated for disease control in most of African countries are in favour of the creation of a collaborative regional prevention and control strategy to contain brucellosis infection. Such a strategy should adopt the One Health or the Ecohealth principle (Zinsstag, 2013). The approach should take into account the particular ecosystem of West African countries and should ensure more cooperation, and exchange of information and resources between public health and veterinary authorities not only at national level but also at regional level. The One Health approach implies an integrated approach involving both human health and veterinary services for the surveillance of zoonotic diseases such as brucellosis. This approach allows to better understand the epidemiology of zoonotic diseases and induces a more efficient utilization of the limited resources (Saegerman et al., 2010; Dean et al., 2012). Creation of zoonotic disease units should also be promoted to formalize the above-mentioned intersectoral collaboration. A regulatory framework is also needed for a better coordination of control activities in the field within and between countries. In the same spirit, collaboration between researchers, public health and veterinary actors of Ivory Coast and neighbouring countries need to be established and strengthened. The establishment of a reference laboratory at regional level needs to be considered. Finally, the commitment of national authorities and the political support and leadership of regional institutions such as the Ecomonic Community of West African States (ECOWAS) represent a key requirement, beneficial for the sustainability and the development of livestock in the region.

110

Chapter 8: General discussion, conclusions and perspectives

Despite the numerous priorities, more attention and consideration needs to be given to brucellosis as well as other endemic neglected zoonotic diseases, especially in low-income country as Ivory Coast. This is essential to foresee a sustainable development of livestock, to cover the needs of populations in terms of animal protein and to contribute to poverty alleviation.

111

REFERENCES

112

References

ABDOEL T., TRAVASSOS D.I., CARDOSO R., SMITS H.L. Simple and rapid field tests for brucellosis in livestock. Vet. Microbiol., 2008, 130, 312-319. ACHA P.N., SZYFRES B. Brucellosis. In : Zoonoses and Communicable Diseases Common to Man and Animals: Vol 1: Bacterioses and Mycoses. Pan American Health Organization: Washington, 2003, 40-67. ACHA P.N., SZYFRES B. Zoonoses et maladies transmissibles a` l’homme et aux animaux, Vol. I. Bactérioses et Mycoses, OIE: Paris, France, 2005, 382 p. ADONE R., CIUCHINI F., LA ROSA G., MARIANELLI C., MUSCILLO M. Use of polymerase chain reaction to identify Brucella abortus strain RB51 among Brucella field isolates from cattle in Italy. J. Vet. Med. B Infect. Dis. Vet. Public Health, 2001, 48, 107-113. ADONE R., PASQUALI P. Epidemiosurveillance of brucellosis. Rev. sci. tech. Off. int. Epiz., 2013, 32(1), 199-205. AKAKPO A.J. Brucelloses animales en Afrique tropicale. Particularités épidémiologique, clinique et bactériologique. Rev. Elev. Med. vet. Pays trop. , 1987, 40, 307-320. AKAKPO A.J., BORNAREL P. Epidémiologie des brucelloses animales en Afrique tropicale : Enquêtes clinique, sérologique et bactériologique. Rev. sci. tech. Off. int. Epiz., 1987, 6, 981-1027. AKAKPO A.J., BORNAREL P., ALMEIDA J.F. Epidémiologie de la brucellose bovine en Afrique tropicale : enquêtes sérologique en République populaire du Bénin. Rev. Elev. Med. vet. Pays trop., 1984, 37 (2), 133-137. AKAKPO A.J., SALEY M., BORNAREL P., SARRADIN P. Epidémiologie de la brucellose bovine en Afrique tropicale : Enquête sérologique au Niger, identification des deux premières souches de B. abortus biotype 3. Rev. Elev. Med. vet. Pays trop., 1986, 39, 175-179. AKAKPO A.J., TEKO-AGBO A., KONE PH. L'impact de la brucellose sur l'économie et la santé publique en Afrique. In: Recueil des thèmes présentés à l’assemblée mondiale des délégués et aux commissions, 18th Conference of the OIE Regional Commission

113

References

for

Africa,

N'Djamena,

Chad,

23-27

February

2009,

2010,

71-98.

http://www.oie.int/doc/document.php?numrec=3978303. ALTON G.G., JONES L.M. Laboratory techniques in brucellosis, FAO, Animal Health Branch Monograph no 7, Rome. Italie, 1964, 53 p. ALTON G.G., JONES L.M., ANGUS R.D., VERGER J.M. Techniques for the brucellosis laboratory. INRA: Paris, 1988,192 p. ALTON G.G., JONES L.M., PIETZ D.E. La brucellose : technique de laboratoires. FAOOMS: Genève, 1977, 177 p. ALTON G.G., JONES L.M., PIETZ D.E. Laboratory techniques in Brucellosis. World Health Organization: Geneva, 1975, 163 p. ANGBA A., TRAORE A., FRITZ P. Situation de la brucellose animale en Côte d’Ivoire. Rev. Elev. Med. vet. Pays trop. , 1987, 40, 325-329. ANONYMOUS. Étude de faisabilité d’un programme de développement de la production laitière dans le nord de la Côte d’Ivoire. BDPA/CIRAD-IEMVT/FERTILE, MINAGRA-CFD, 1997, 208 p. ANONYMOUS. Décision CEDEAO A/DEC.5/10/98: Réglementation sous régionale relative

à

la

réglementation

de

la

transhumance,

[Online]

(1998),

http://www.gouv.bj/sites/default/files/Decision-A-DEC%205-10-98.pdf, accessed the 31/07/2014. ANONYMOUS. Rapport national sur l’état des ressources zoogénétiques. MIPARH, République de Côte d’Ivoire, 2003a, 80 p. ANONYMOUS. Réglement C/REG.3/01/03 relatif à la mise en œuvre de la reglementation de la transhumance entre les Etats membres de la CEDEAO, [Online] (2003b), http://www.waapp-ppaao.org/projet/classified/Reglement%2BC-REG.3.01.03Conseil%2Bdes%2BMinistres-CEDEAO.pdf, accessed the 31/07/2014. ANONYMOUS. Rapport sur le fonctionnement pour 1970 du laboratoire national d’élevage de Dakar (Sénégal). In: Böhnel, H., 1971. Research on the causes of death among

114

References

calves on the sub-Sudan savannahs of the northern Ivory Coast. Bull. Epizoot. Dis. Afr., 1970, 19(2),143-157. ARCHER K.J., LEMESHOW S. Goodness-of-fit test for a logistic regression model estimated using survey sample data. S.J., 2006, 6, 97-105. ARCHER K.J., LEMESHOW S., HOSMER D.W. Goodness-of-fit tests for logistic regression models when data are collected using a complex sampling design. Comput. Stat. Data Anal. , 2007, 51, 4450-4464. ARIMI S.M., KOROTI E., KANG’ETHE E.K., OMORE A.O., MCDERMOTT J.J. Risk of infection with Brucella abortus and Escherichia coli O157:H7 associated with marketing of unpasteurized milk in Kenya. Acta Trop., 2005, 96, 1–8. AZNAR M.N., SAMARTINO L.E., HUMBLET M.-F., SAEGERMAN C.

Bovine

brucellosis in Argentina and bordering countries: update. Transbound. Emerg. Dis., 2014, 61(2), 121-133. BALE O.O.J., KUMI-DIAKA J. Serological and bacteriological study of bovine Brucellae from livestock investigation and breeding centers in Nigeria. Br. vet. J. , 1981, 137, 256-261. BANAI M., CORBEL M. Taxonomy of Brucella. Open Vet Sci J., 2010, 4, 85-101. BANKOLE A.A., SAEGERMAN C., BERKVENS D., FRETIN D., GEERTS S., IEVEN G., WALRAVENS K. Phenotypic and genotypic characterization of Brucella strains isolated from cattle in the Gambia. Vet. Rec., 2010,166, 753-756. BASSETT T.J., TURNER M.D. “Sudden Shift of Migratory Drift? FulBe Herd Movements to the Sudano-Guinean Region of West Africa.” Hum. Ecol., 2007, 35, 33-49. https://netfiles.uiuc.edu/bassett/shared/Bassett%20pubs/Bassett%20and%20Turner%2 02007.pdf. BENKIRANE A. Epidemiological surveillance and control of brucellosis in ruminants: the example of the region comprising North Africa and the Near East. Rev. sci. tech. Off. int. Epiz., 2001, 20, 757-767.

115

References

BERCOVICH Z. The use of skin delayed-type hypersensitivity as an adjunct test to diagnose brucellosis in cattle: A review. Vet. Q., 2000, 22, 123-130. BERKVENS D., SPEYBROECK N., PRAET N., ADEL A., LESAFFRE E. Estimating disease prevalence in a Bayesian framework using probabilistic constraints. Epidemiology, 2006, 17, 145-153. BERKVENS D.L., SPEYBROECK N., LESAFFRE E., SAEGERMAN C. Comments on "Pouillot, R., Gerbier, G., Gardner, I.A., 2002."TAGS", a program for the evaluation of test accuracy in the absence of a gold standard. Prev. Vet. Med. 53, 67-81". Prev. Vet. Med., 2003, 59(3), 181-183. BLACK M.A., CRAIG B.A. Estimating disease prevalence in the absence of a gold standard. Stat. Med., 2002, 21, 2653-2669. BÖHNEL H. Research on the causes of death among calves on the sub-Sudan savannahs of the northern Ivory Coast. Bull. Epizoot. Dis. Afr., 1971,19(2) ,143-157. BOUKARY A. R. Epidémiologie de la brucellose et de la tuberculose animales dans les milieux urbain, périurbain et rural au Niger. Thèse de Doctorat en sciences vétérinaires,

Université

de

Liège:

Liège,

Belgique,

2013,

185

p.

http://hdl.handle.net/2268/146448. BOUKARY A.R., SAEGERMAN C., ABATIH E., FRETIN D., ALAMBÉDJI BADA R., DE DEKEN R., HAROUNA A.H., YENIKOYE A., THYS E. Seroprevalence and potential risk factors for Brucella spp. infection in traditional cattle, sheep and goats reared in urban, periurban and rural areas of Niger. PLoS ONE, 2013, 8(12): e83175. doi:10.1371/journal.pone.0083175. BOURRET G. La fièvre méditerranéenne en AOF. Bull. Soc. Pathol. Exot., 1910, 13, 490494. BRANSCUM A.J., GARDNER I.A., JOHNSON W.O. Estimation of diagnostic test sensitivity and specificity through Bayesian modeling. Prev. Vet. Med., 68, 145–163. BRICKER B.J. PCR as a diagnostic tool for brucellosis. Vet. Microbiol. , 2002, 90, 435-446.

116

References

BRICKER B.J., EWALT D.R., HALLING S.M. Brucella ‘‘HOOFPrints’’: strain typing by multi-locus analysis of variable number tandem repeats (VNTRs). BMC Microbiol., 2003, 3, 15–29. BRICKER B.J., HALLING S.M. Differentiation of Brucella abortus bv. 1, 2, and 4, Brucella melitensis, Brucella ovis, and Brucella suis bv. 1 by PCR. J. Clin. Microbiol. , 1994, 32, 2660-2666. BRINLEY-MORGAN W.J., MCCULLOUGH N.B. Genus Brucella Meyer and Shaw 1920. In: R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th Ed. The Williams & Wilkins Co.: Baltimore, 1974 , 278-282. BRONSVOORT B.M.D., KOTERWAS B., LAND F., HANDEL I.G., TUCKER J., MORGAN K.L., TANYA V.N., ABDOEL T.H., SMITS H.L. Comparison of a Flow Assay for Brucellosis Antibodies with the Reference cELISA Test in West African Bos indicus. PLoS ONE, 2009, 4(4): e5221. doi:10.1371/journal.pone.0005221. BRUCE D. Note on the discovery of a microorganism in Malta fever. Practitioner, 1887, 39, 161. CÁCERES S.B. Comparative veterinary capacity in western Africa: implications for livestock development. Livestock Research for Rural Development. [Online] (2010), http://www.lrrd.org/lrrd22/10/burg22180.htm , accessed on 12/12/2013. CAMUS E. Brucellosis vaccination in cows in the North of Ivory-Coast - Techniques and Results. Rev. Elev. Med. vet. Pays trop., 1980b, 33, 363-369. CAMUS E. Clinical incidence of bovine brucellosis in northern Ivory Coast. Rev. Elev. Med. vet. Pays trop., 1980a, 33, 263-269. CAMUS E. Evaluation of trypanosomiasis and brucellosis control in cattle herds of Ivory Coast. Agric. Human Values, 1995, 12, 90-94. CARR S., UNWIN N., PLESS-MULLOLI T. Chapter 5: Epidemiological studies designs. In: Introduction to public health and epidemiology, McGraw-hill Open University Press: Berkshire, 2007, p. 59.

117

References

CHANTAL J., THOMAS J.F. Etude sérologique sur la brucellose bovine aux abattoirs de Dakar. Rev. Elev. Med. vet. Pays trop., 1976, 29, 101-108. CHARTIER C. Ovine brucellosis in Ivory Coast: a serological survey. Bull. Anim. Health Prod. Afr., 1992, 40(3), 213-214 CHIMANA H.M., MUMA J.B., SAMUI K.L., HANGOMBE B.M., MUNYEME M., MATOPE G., PHIRI A.M., GODFROID J., SKJERVE E., TRYLAND M. A comparative study of seroprevalence of brucellosis in commercial and small-scale mixed dairy-beef cattle enterprises of Lusaka province and Chibombo district, Zambia. Trop. Anim. Health Prod., 2010, 42(7), 1541-1545. CHRISTENSEN J., GARDNER I.A. Herd-level interpretation of test results for epidemiologic studies of animal diseases. Prev. Vet. Med., 2000, 45, 83–106. CICCHETTI D., FEINSTEIN A. High agreement but low kappa: II. Resolving the paradoxes. J. Clin. Epidemiol., 1990, 43, 551–558. CLAVAREAU C., WELLEMANS V., WALRAVENS K., TRYLAND M., VERGER J.M., GRAYON M., CLOECKAERT A., LETESSON J.J., GODFROID J. Phenotypic and molecular characterization of a Brucella strain isolated from a minke whale (Balaenoptera acutorostrata). Vet. Microbiol., 1998,144, 3267–3273. COHAN F.M. What are bacterial species? Annu. Rev. Microbiol. , 2002, 56, 457–487. COLMENERO-CASTILLO

J.D.,

CABRERA-FRANQUELO

F.P.,

HERNANDEZ-

MARQUEZ S., REGUERA-IGLESIAS J.M., PINEDO-SANCHEZ A., CASTILLOCLAVERO A.M. Repercusión socioeconomic de la brucelosis humana. Rev. Clín. Esp., 1989, 185 (9), 459-463. CORBEL M.J. Brucellosis in Humans and Animals. World health Organization press edition: Genève, 2006, 89 p. CORBEL M.J. Brucellosis, pp. 189-221. In: Laing, J.A. (ed.) Fertility and Infertility in Veterinary Practice. 4th ed. ELBS, Bailliere Tindall, 1988.

118

References

CORBEL M.J. Brucellosis: an overview (1st International conference on emerging zoonosis). Emerg.

Infect.

Dis.,

1997,

3,

213-221.

http://www.cdc.gov/ncidod/EID/vol3no2/corbel.htm. CORBEL M.J. International Committee on Systematic Bacteriology. Subcommittee on Taxonomy of Brucella. Minutes of the Meeting, 10 August 1982, Boston, Massachusetts. Int. J. Syst. Bacteriol., 1984, 34, 366-367. CORBEL M.J., BRINLEY-MORGAN W.J. Genus Brucella Meyer and Shaw 1920, 173AL. In: Krieg N.R., Holt J.G. (Eds.), Bergey’s manual of systematic bacteriology, Vol. 1, The Williams & Wilkins Co.: Baltimore, 1984, 377–388. CORBEL M.J., MORGAN W.J. Proposal for minimal standards for descriptions of new species and biotypes of the genus Brucella. Int. J. Syst. Bacteriol., 1975, 25, 83–89. COULIBALY F. La brucellose en Côte d’Ivoire. Communication à l’OIE. In Böhnel, H., 1971. Research on the causes of death among calves on the sub-Sudan savannahs of the northern Ivory Coast. Bull. Epizoot. Dis. Afr., 1973,19(2), 143-157. CRAWFORD R.P., WILLIAMS J.D., HUBER J.D., CHILDERS A.B. Biovars of Brucella abortus and their value in epidemiological studies of infected cattle herds. J. Am. Med. Assoc. , 1979,175, 1274-1277. CUTLER S.J., WHATMORE A.M., COMMANDER N.J. Brucellosis - new aspects of an old disease. J. Appl. Microbiol., 2005, 98, 1270-1281. DE LEY J., MANNHEIM W., SEGERS P., LIEVENS A., DENIJN M., VANHOUCKE M., GILLIS M.

Ribosomal ribonucleic acid cistron similarities and taxonomic

neighborhood of Brucella and CDC group Vd. Int. J. Syst. Bacteriol., 1987, 37, 35– 42. DEAN A.S., BONFOH B., KULO A.E., BOUKAYA G.A., AMIDOU M., HATTENDORF J., PILO P., SCHELLING E. Epidemiology of brucellosis and Q Fever in linked human and animal populations in northern Togo. PLoS ONE, 2013, 8(8):e71501. doi: 10.1371/journal.pone.0071501.

119

References

DEAN A.S., CRUMP L., GRETER H., SCHELLING E., ZINSSTAG J. Global Burden of Human Brucellosis: A Systematic Review of Disease Frequency. PLoS Neg. Trop. Dis., 2012, 6(10), e1865. doi:10.1371/journal.pntd.0001865. DEAN A.S., SCHELLING E., BONFOH B., KULO A.E., BOUKAYA G.A., PILO P. Deletion in the gene BruAb2_0168 of Brucella abortus strains : diagnostic challenges. Clin. Microbiol. Infect., 2014, 20, O550–O553. DEEKS J.J. Systematic reviews in health care: Systematic reviews of evaluations of diagnostic and screening tests. BMJ, 2001, 323, 157-162. DELAFOSSE A., GOUTARD F., THEBAUD E. Epidémiologie de la tuberculose et de la brucellose des bovins en zone périurbaine d’Abéché, Tchad. Rev. Elev. Med. vet. Pays trop., 2002, 55, 5-13. DENDUKURI N., BELISLE P., JOSEPH L. Bayesian sample size for diagnostic test studies in the absence of a gold standard: Comparing identifiable with non-identifiable models. Stat. Med. , 2010, 29, 2688-2697. DENDUKURI N., JOSEPH L. Bayesian approaches to modeling the conditional dependence between multiple diagnostic tests. Biometrics, 2001, 57, 158–167. DOHOO I., MARTIN W., STRYHN H. Screening and diagnostic tests. In: Veterinary Epidemiologic Research. AVC Inc.: Charlottetown, Prince Edward Island, Canada, 2003, 85-120. DOHOO I.R., WRIGHT P.F., RUCKERBAUER G.M., SAMAGH B.S., ROBERTSON F.J., FORBES L.B. A comparison of five serological tests for bovine brucellosis. Can. J. Vet. Res., 1986, 50, 485-493. DOMENECH J. Importance des brucelloses animales en Afrique centrale. Rev. Elev. Med. vet. Pays trop., 1987, 40, 321-324. DOMENECH J., CORBEL M.J., THOMAS E.L., LUCET PH. La brucellose bovine en Afrique centrale : VI. Identification et typage des souches isolées au Tchad et au Cameroun. Rev. Elev. Med. vet. Pays trop., 1983,36, 19–25.

120

References

DOMENECH J., COULOMB J., LUCET P. La brucellose bovine en Afrique centrale. IV. Évaluation de son incidence économique et calcul du coût-bénéfice des opérations d’assainissement. Rev. Elev. Med. vet. Pays trop., 1982, 35 (2), 113-124. DOMENECH J., LUCET P., VALLAT B., STEWARD C., BONNET J.B., BERTAUDIERE L. La brucellose bovine en Afrique centrale. II. Etude épidémiologique : particularités régionales et problèmes de l’élevage semi-intensif. Rev. Elev. Med. vet. Pays trop. , 1980, 33 (3), 277-284. DORNY P., PHIRI I. K., VERCRUYSSE J., GABRIEL S., WILLINGHAM A. L., BRANDT J., VICTOR B., SPEYBROECK N., BERKVENS D. A Bayesian approach for estimating values for prevalence and diagnostic test characteristics of porcine cysticercosis. Int. J. Parasitol. , 2004, 34, 569-576. DOUTRE M.P., FENSTERBANK R., SAGNA F. Etude de la brucellose bovine dans un village de Basse-Casamance (Sénégal). I.- Diagnostic sérologique et bactériologique. Rev. Elev. Med. vet. Pays trop., 1977,30, 345-351. DUNSON D.B. Commentary: practical advantages of Bayesian analysis of epidemiologic data. Am. J. Epidemiol. , 2001, 12, 1222–1226. ENØE C., GEORGIADIS M.P., JOHNSON W.O. Estimation of sensitivity and specificity of diagnostic tests and disease prevalence when the true disease state is unknown. Prev. Vet. Med., 2000,45, 61–81. EUROPEAN FOOD SAFETY AUTHORITY (EFSA) Scientific Opinion of the Panel on Animal Health and Welfare (Question No EFSA-Q-2008-665), adopted on 5 June 2009. Porcine brucellosis (Brucella suis). The EFSA Journal, 2009, 1144, 1-112. EWALT D.R., PAYEUR J.B., MARTIN B.M., CUMMINGS D.R., MILLER G. Characteristics of a Brucella species from a bottlenose dolphin (Tursiops truncatus), J. Vet. Diagn. Invest., 1994, 6, 448–452. EZE E.N. Isolation of Brucellae from the Nigerian livestock and the typing of such isolates. Bull. Anim. Health Prod. Afr., 1978, 26, 29-36.

121

References

FAO (FOOD AND AGRICULTURAL ORGANIZATION). FAOSTAT. Food and Agricultural

Organization

Statistic

Division,

[Online]

(2014a),

http://faostat.fao.org/site/550/DesktopDefault.aspx?PageID=550#ancor, accessed the 24/02/2014. FAO (FOOD AND AGRICULTURAL ORGANIZATION). FAOSTAT. Food and Agricultural

Organization

Statistic

Division,

[Online]

(2014b),

http://faostat.fao.org/site/573/DesktopDefault.aspx?PageID=573#ancor, accessed the 24/02/2014. FAO (FOOD AND AGRICULTURAL ORGANIZATION) A stepwise approach for progressive

control

of

brucellosis

in

animals.

[Online]

(2013),

http://www.fao.org/ag/againfo/programmes/en/empres/news_250113b.html, accessed 2/10/2014. FAO (FOOD AND AGRICULTURAL ORGANIZATION). World livestock 2011. In: Livestock in food security. FAO: Rome, 2011, 13-27. FAYE B., CASTEL V., LESNOFF M., RUTABINDA D., DHALWA J. Tuberculosis and brucellosis prevalence survey on dairy cattle in Mbarara milk basin (Uganda). Prev. Vet. Med., 67, 267-281. FEINSTEIN A., CICCHETTI D. High agreement but low kappa: I. The problems of two paradoxes. J. Clin. Epidemiol., 1990, 43 (6), 543–549. FENSTERBANK R. Brucellose des bovines et des petits ruminants : diagnostic, prophylaxie et vaccination. Rev. sci. tech. Off. int. Epiz. , 1986, 5, 587-603. FERNANDO C.L., ELIAS F.R., ELENA M.V. Brucellose ovine et caprine. In: Lefèvre P. Blancou J.et Chermette R. (Eds). Principales maladies infectieuses et parasitaires du bétail. Editions Médicales Internationale: Paris, 2003, 891-904. FERREIRA L., CASTAÑO S.V., SANCHEZ-JUANES F., GONZALEZ-CABRERO S., MENEGOTTO F., ORDUÑA-DOMINGO A., GONZALEZ-BUITRAGO J.M., MUÑOZ-BELLIDO J.L. Identification of Brucella by MALDI-TOF Mass Spectrometry. Fast and Reliable Identification from Agar Plates and Blood Cultures. PLoS ONE, 2010, 5(12): e14235. doi:10.1371/journal.pone.0014235.

122

References

FISHER R.A. On the mathematical foundations of theoretical statistics. Phil. Trans. R. Soc. A., 1922, 222, 309-368. FOSGATE G.T., ADESIYUN A.A., HIRD D.W., JOHNSON W.O., HIETALA S.K., SCHURIG G.G., RYAN J. Comparison of serologic tests for detection of Brucella infections in cattle and water buffalo (Bubalus bubalis). Am. J. Vet. Res., 2002, 63, 1598-605. FOSTER G., JAHANS K.L., REID R.J., ROSS H.M. Isolation of Brucella species from cetaceans, seals and an otter. Vet. Rec., 1996,138, 583-585. FRANCO M.P., MULDER M., GILMAN R.H., SMITS H.L. Human brucellosis. Lancet Infect. Dis., 2007, 7, 775-786. FREDSLUND J. PHY.FI: fast and easy online creation and manipulation of phylogeny color figures. BMC Bioinformatics , 2006,7, 315. FRETIN D., CZAPLICKI G., QUINET C. Brucellosis, porcine – Belgium (02): serovar 2, wild

boar,

bovine.

[Online]

http://www.promedmail.org/direct.php?id=20120710.1196658,

(2012), accessed

on

12/07/2012. FRETIN D., MORI M., CZAPLICKI G., QUINET C., MAQUET B., GODFROID J., SAEGERMAN C. Unexpected Brucella suis biovar 2 Infection in a dairy cow, Belgium. Emerg. Infect. Dis., 2013, 19(12), 2053-2054. doi: 10.3201/eid1912.130506. GALL D., NIELSEN K. Serological diagnosis of bovine brucellosis: a review of test performance and cost comparison. Rev. sci. tech. Off. int. Epiz. ,2004, 23 , 989-1002. GALL D., NIELSEN K., FORBES L., COOK W., LECLAIR D., BALSEVICIUS S., KELLY L., SMITH P., MALLORY M. Evaluation of the fluorescence polarization assay and comparison to other serological assays for detection of brucellosis in cervids. J. Wildl. Dis., 2001, 37, 110-118. GALL D., NIELSEN K., FORBES L., DAVIS D., ELZER P., OLSEN S., BALSEVICIUS S., KELLY L., SMITH P., TAN S., JOLY D. Validation of the fluorescence

123

References

polarization assay and comparison to other serological assays for the detection of serum antibodies to Brucella abortus in bison. J. Wildl. Dis. , 2000, 36, 469-476. GARDNER I.A. The utility of Bayes’ theorem and Bayesian inference in veterinary clinical practice and research. Aust.Vet. J., 2002, 80, 758–761. GARDNER I.A., STRYHN H., LIND P., COLLINS M.T. Conditional dependence between tests affects the diagnosis and surveillance of animal diseases. Prev. Vet. Med., 2000, 45(1-2), 107–122. GART J.J., BUCK A.A. Comparison of a screening test and a reference test in epidemiologic studies. II. A probabilistic model for the comparison of diagnostic tests. Am. J. Epidemiol., 1966, 83, 593-602. GATSONIS C., PALIWAL P. Meta-analysis of diagnostic and screening test accuracy evaluations: methodologic primer. Am. J. Roentgenol. ,2006, 187(2), 271-281. GELMAN A., CARLIN J.B., STERN H.S., RUBIN D.B. Bayesian data analysis. Chapman & Hall/CRC: London, 2004, 696 p. GELMAN A., RUBIN D.B. Inference from iterative simulation. Stat. sci., 1992, 7, 457-511. GEORGIADIS M. P., JOHNSON W. O., GARDNER I. A., SINGH R. “Correlation-adjusted estimation of sensitivity and specificity of two diagnostic tests”. J. R. Stat. Soc. Series C, 2003, 52 (1), 63–76. GEURDEN T., BERKVENS D., GELDHOF P., VERCRUYSSE J., CLAEREBOUT E. A Bayesian approach for the evaluation of six diagnostic assays and the estimation of Cryptosporidium prevalence in dairy calves. Vet. Res., 2006, 37, 671-682. GIDEL R., ALBERT J.P., MAO G.L., RETIF M. La brucellose en Afrique occidentale et son incidence sur la santé publique. Résultats de dix enquêtes épidémiologiques effectuées en Côte-d'Ivoire, Haute-Volta et Niger de 1970 à 1973. Rev. Elev. Med. vet. Pays trop. , 1974, 27, 403-418. GLAS A.S., LIJMER J.G., PRINS M.H., BONSEL G.J., BOSSUYT P.M.M. The diagnostic odds ratio: a single indicator of test performance. J. Clin. Epidemiol. , 2003, 56, 1129–1135.

124

References

GODFROID J., AL DAHOUK S., PAPPAS G., ROTH F., MATOPE G., MUMA J., MARCOTTY T., PFEIFFER D., SKJERVE E. A “One Health” surveillance and control of brucellosis in developing countries: Moving away from improvisation. Comp. Immunol. Microbiol. Infect. Dis., 2013, 36, 241– 248. GODFROID J., CLOECKAERT A., LIAUTARD J. P., KOHLER S., FRETIN D., WALRAVENS K., GARIN-BASTUJI B., LETESSON J. J. From the discovery of the Malta fever's agent to the discovery of a marine mammal reservoir, brucellosis has continuously been a re-emerging zoonosis. Vet. Res., 2005, 36, 313-326. GODFROID J., KASBOHRER A. Brucellosis in the European Union and Norway at the turn of the twenty-first century. Vet. Microbiol. , 2002, 90, 135–145. GODFROID J., NIELSEN K., SAEGERMAN C. Diagnosis of Brucellosis in livestock and wildlife. Croat. Med. J., 2010, 51, 296-305. GODFROID J., SAEGERMAN C., WELLEMANS V., WALRAVENS K., LETESSON J.J., TIBOR A., MCMILLAN A., SPENCER S., SANNA M., BAKKER D., POUILLOT R., GARIN-BASTUJI B. How to substantiate eradication of bovine brucellosis when aspecific serological reactions occur in the course of brucellosis testing. Vet. Microbiol. , 2002, 90, 461–477. GODFROID J., SCHOLZ H.C., BARBIER T., NICOLAS C., WATTIAU P., FRETIN D., WHATMORE A.M., CLOECKAERT

A.,

BLASCO J.M.,

MORIYON

I.,

SAEGERMAN C., MUMA J.B., AL DAHOUK S., NEUBAUER H., LETESSON J.J. Brucellosis at the animal/ecosystem/human interface at the beginning of the 21st century. Prev. Vet. Med., 2011,102, 118–131. GRAHAM P., BULL B. Approximate standard errors and confidence intervals for indices of positive and negative agreement. J. Clin. Epidemiol., 1998, 51, 763–771. GREINER M., GARDNER I.A. Application of diagnostic tests in veterinary epidemiologic studies. Prev. Vet. Med., 2000, 45(1-2), 43–59. HONEST H., KHAN K.S. Reporting of measures of accuracy in systematic reviews of diagnostic literature. BMC Health Serv. Res., 2002, 2(1), 4.

125

References

HUI S.L., WALTER S.D. Estimating the error rates of diagnostic tests. Biometrics, 1980, 36, 167–171. HUI S.L., ZHOU X.H. Evaluation of diagnostic tests without gold standards. Stat. Methods Med., 1998, 7, 354-370. ICA T., AYDIN F., ERDENLIG S., GULER L., BÜYÜKCANGAZ E. Characterization of Brucella abortus biovar 3 isolates from Turkey as biovar 3b. Vet. Rec. , 2008, 63, 659-661. IFAD (INTERNATIONAL FUND FOR AGRICULTURAL DEVELOPMENT) Rural poverty report 2001: the challenge of ending rural poverty, Oxford university Press:Oxford, [Online] (2001) , http://www.ifad.org/poverty/e_sum.pdf, accessed on 10/01/2014 . IHORST G., FORSTER J., PETERSEN G., WERCHAU H., ROHWEDDER A., SCHUMACHER M. The use of imperfect diagnostic tests had an impact on prevalence estimation. J. Clin. Epidemiol., 2007, 60, 902-910. INTERNATIONAL LIVESTOCK RESEARCH INSTITUTE (ILRI). Mapping of poverty and likely zoonoses hotspots. Zoonoses Project 4. Report to the Department for International Development, UK. ILRI: Nairobi, Kenya, [Online] (2012), 119 pp, www.dfi d. gov.uk/r4d/Output/190314/Default.aspx , accessed on 14 /08/2013. IRMAK H., BUZGAN T., EVIRGEN O., AKDENIZ H., DEMIROZ A.P., ABDOEL T.H., SMITS H.L. Use of a new, simple and rapid diagnostic test, the Brucella-IgM and IgG flow assays in the serodiagnosis of human brucellosis in an endemic area in eastern Turkey. Am. J. Trop. Med. Hyg., 2004, 70, 688–694. IRWIG L., MACASKILL P., GLASZIOU P., FAHEY M. Meta-analytic methods for diagnostic test accuracy. J. Clin. Epidemiol., 1995, 48, 119-130. JAMES A.D. Guide to epidemiological surveillance for Rinderpest. Rev. sci. tech. Off. int. Epiz., 1998,17, 796-809.

126

References

JERGEFA T., KELAY B., BEKANA M., TESHALE S., GUSTAFSON H., KINDAHL H. Epidemiological study of bovine brucellosis in three agro-ecological areas of central Oromiya, Ethiopia. Rev. sci. tech. Off. int. Epiz. , 2009, 28, 933-943. JIANG, H., WANG H., XU L., HU G., MA J., XIAO P., FAN W., DI D., TIAN G., FAN M., MI J., YU R., SONG L., ZHAO H., PIAO D., CUI B. MLVA Genotyping of Brucella melitensis and Brucella abortus isolates from different animal species and humans and identification of Brucella suis vaccine strain S2 from cattle in China. PLoS ONE, 2013, 8(10), e76332. doi:10.1371/journal.pone.0076332. JOHNSON W.O., GASTWIRTH J.L., PEARSON L.M. Screening without a ‘‘gold standard’’: the Hui-Walter paradigm revisited. Am. J. Epidemiol. , 2001,153, 921-924. JONES G., JOHNSON W.O., HANSON T.E., CHRISTENSEN R. Identifiability of models for multiple diagnostic testing in the absence of a gold standard. Biometrics , 2010, 66, 855–863. JOSEPH L., GYORKOS T., COUPAL L. Bayesian estimation of disease prevalence and the parameters of diagnostic tests in the absence of a gold standard. Am. J. Epidemiol., 1995, 141, 263-272. KADOHIRA M., MCDERMOTT J.J., SHOUKRI M.M., KYULE M.N. Variations in the prevalence of antibody to Brucella infection in cattle by farm, area and district in Kenya. Epidemiol. Infect. , 1997, 118, 35-41. KASYMBEKOV J., IMANSEITOV J., BALLIF M., SCHÜRCH N., PANIGA S., PILO P., TONOLLA M., BENAGLI C., AKYLBEKOVA K., JUMAKANOVA Z., SCHELLING

E.,

ZINSSTAG

J.

Molecular

Epidemiology

and

Antibiotic

Susceptibility of Livestock Brucella melitensis isolates from Naryn Oblast, Kyrgyzstan.

PLoS

Negl.

Trop.

Dis

,

2013,

7(2),

e2047.

doi:10.1371/journal.pntd.0002047. KAUFMANN A.F., MELTZER M.I., SCHMID G.P. The economic impact of a bioterrorist attack: are prevention and postattack intervention programs justifiable? Emerg. Infect. Dis., 1997, 3, 83–94

127

References

KUBUAFOR D.K., AWUMBILA B., AKANMORI B.D. Seroprevalence of brucellosis in cattle and humans in the Akwapim-south district of Ghana: public health implications. Acta Trop., 2000, 76, 45-48. KUNDA J., FITZPATRICK J., KAZWALA R., FRENCH N.P., SHIRIMA G., MACMILLAN A., KAMBARAGE D., BRONSVOORT M., CLEAVELAND S. Health-Seeking Behavior of Human brucellosis cases in Rural Tanzania, BMC Public Health, 2007, 7, 315. LE FLECHE P., JACQUES I., GRAYON M., AL DAHOUK S., BOUCHON N.P., DENOEUD F., NOCKLER K., NEUBAUER H., GUILLOTEAU L.A., VERGNAUD G. Evaluation and selection of tandem repeat loci for a Brucella MLVA typing assay. BMC Microbiol. , 2006, 6, 9. LEEFLANG M.M., DEEKS J.J., GATSONIS C., BOSSUYT P.M. Systematic Reviews of Diagnostic Test Accuracy. Ann. Intern. Med., 2008, 149(12), 889-897. LEWIS F., MICHAEL P.W. Improving epidemiologic data analyses through multivariate regression modeling. Emerg. Themes Epidemiol. , 2013,10, 4. doi: 10.1186/17427622-10-4. LEWIS F.I., TORGERSON P.R. A tutorial in estimating the prevalence of disease in humans and animals in the absence of a gold standard diagnostic. Emerg. Themes Epidemiol., 2012, 9, 9. http://www.ete-online.com/content/9/1/9. LID (LIVESTOCK IN DEVELOPMENT). Strategies for improving DFID’s impact on poverty reduction: A review of best practice in the livestock sector. LID (Livestock in Development): Crewkerne, UK, 1998, 71 pp. LIMET J.N., KERKHOFS P., WIJFELS R., DEKEYSER P. Le diagnostic sérologique de la brucellose bovine par ELISA., 1988, 132, 565-575. LISTA F., REUBSAET F.A., DE SANTIS R., , PARCHEN R.R., JONG A.L.D., KIEBOOM J., VAN DER LAAKEN A.L., VOSKAMP-VISSER I.A.I, FILLO S., JANSEN H.-J., VAN DER PLAS J., PAAUW A. Reliable identification at the species level of Brucella isolates with MALDI-TOF-MS. BMC Microbiol., 2011, 11, 267.

128

References

MABLESON H.E., OKELLO A., PICOZZI K., WELBURN S.C. Neglected Zoonotic Diseases - The Long and Winding Road to Advocacy. PLoS Neg. Trop. Dis., 2014, 8(6): e2800. doi:10.1371/journal.pntd.0002800. MANGEN M.J., OTTE J., PFEIFFER D., CHILONDA P. Bovine brucellosis in sub-saharan africa: estimation of sero-prevalence and impact on meat and milk offtake potential. FAO. Livestock Information and Policy Branch, AGAL, Livestock Policy Paper n°8. Rome, Italie, 2002, 53 p. MAQUART M., LE FLECHE P., FOSTER G., TRYLAND M., RAMISSE F., DJØNNE B., DAHOUK S. A., JACQUES I., NEUBAUER H., WALRAVENS K., GODFROID J., CLOECKAERT A., VERGNAUD G. MLVA-16 typing of 295 marine mammal Brucella isolates from different animal and geographic origins identifies 7 major groups within Brucella ceti and Brucella pinnipedialis. BMC Microbiol. , 2009, 9, 145. MARCOTTY T., MATTHYS F., GODFROID J., RIGOUTS L., AMENI G., GEY VAN PITTIUS N., KAZWALA R., MUMA J., VAN HELDEN P., WALRAVENS K., DE KLERK L. M., GEOGHEGAN C., MBOTHA D., OTTE M., AMENU K., ABU SAMRA N., BOTHA C., EKRON M., JENKINS A.,

JORI F., KRIEK N.,

MCCRINDLE C., MICHEL A., MORAR D., ROGER F., THYS E., VAN DEN BOSSCHE P. Zoonotic tuberculosis and brucellosis in Africa: neglected zoonoses or minor public-health issues? The outcomes of a multi-disciplinary workshop. Ann. Trop. Med. Parasitol. , 2009, 103 (5), 401–411 . MARCOTTY T., THYS E., CONRAD P., GODFROID J., CRAIG P., ZINSSTAG J., MEHEUS F., BOUKARY A.R., BADÉ M.A., SAHIBI H., FILALI H., HENDRICKX S., PISSANG C., VAN HERP M., VAN DER ROOST D., THYS S., HENDRICKX D., CLAES M., DEMEULENAERE T., VAN MIERLO J., DEHOUX J-P., BOELAERT M. Intersectoral collaboration between the medical and veterinary professions in low-resource societies: The role of research and training institutions. Comp. Immunol. Microbiol. Infect. Dis., 2013,36, 233– 239. MARTIN S.W., SHOUKRI M., THORBURN M.A. Evaluating the health status of herds based on tests applied to individuals. Prev. Vet. Med., 1992,14, 33–43.

129

References

MAURIN M. La brucellose à l’aube du 21ème siècle. Med. Maladies Infect., 2005,35, 6-16. MCDANIEL C.J., CARDWELL D.M., MOELLER R.B. Jr,, GRAY G.C. Humans and cattle: a review of bovine zoonoses.Vector Borne Zoonotic Dis., 2013, 14, 1–19. doi: 10.1089/vbz.2012.1164. MCDERMOTT J., GRACE D., ZINSSTAG J. Economics of brucellosis impact and control in low-income countries. Rev. sci. tech. Off. int. Epiz., 2013, 32 (1), 249-261. MCDERMOTT J.J., ARIMI S.M. Brucellosis in sub-Saharan Africa: Epidemiology, control and impact. Vet. Microbiol., 2002, 90, 111-134. MEGERSA B., BIFFA D., NIGUSE F., RUFAEL T., ASMARE K., SKJERVE E. Cattle brucellosis in traditional livestock husbandry practice in Southern and Eastern Ethiopia, and its zoonotic implication. Acta Vet. Scand. , 2011,53, 24. MELTZER E., SIDI Y., SMOLEN G., BANAI M., BARDENSTEIN S., SCHWARTZ E. Sexually transmitted brucellosis in humans. Clin. Infect. Dis., 2010, 51(2):e12–e15. doi: 10.1086/653608. MENTEN J., BOELAERT M., LESAFFRE E. Bayesian latent class models with conditionally dependent diagnostic tests: a case study. Stat. Med., 2008, 27, 4469– 4488. MEYER K.F., SHAW E.B. A comparison of the morphologic, cultural and biochemical characteristics of B. abortus and B. melitensis from cattle. Studies on the genus Brucella nov. gen. J. Infect. Dis., 1920, 27, 173–184. MEYER N., VINZIO S., GOICHOT B. La statistique bayésienne: une approche des statistiques adaptée la clinique. Rev. Med. Interne , 2009, 30(3), 242–249. MOHER D., LIBERATI A., TETZLAFF J., ALTMAN D.G., THE PRISMA GROUP. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Medicine , 2009, 6(7):e1000097. doi:10.1371/journal.pmed.100009. MONTORI V.M., GUYATT G.H. Summarizing studies of diagnostic test performance. Clin. Chem., 2003,49, 1783–1784.

130

References

MORENO E. Retrospective and prospective perspectives on zoonotic brucellosis.Front. Microbiol., 2014, 5, 213. doi: 10.3389/fmicb.2014.00213. MUENDO E., MBATHA P., MACHARIA J., ABDOEL T., JANSZEN P., PASTOOR R., SMITS H.L. Infection of cattle in Kenya with Brucella abortus biovar 3 and Brucella melitensis biovar 1 genotypes. Trop. Anim. Health Prod., 2012, 44, 17–20. MUMA J.B., GODFROID J., SAMUI K.L., SKJERVE E. The role of Brucella infection in abortions among traditional cattle reared in proximity to wildlife on the Kafue flats of Zambia. Rev. sci. tech. Off. int. Epiz., 2007, 26, 721-730. MUMA J.B., SAMUI K.L., SIAMUDAALA V.M., OLOYA J., MATOPE G., OMER M.K., MUNYEME M., MUBITA C., SKJERVE E. Prevalence of antibodies to Brucella spp. and individual risk factors of infection in traditional cattle, goats and sheep reared in livestock-wildlife interface areas of Zambia. Trop. Anim. Health Prod. , 2006, 38(3), 195-206. MUNOZ P.M., MARIN C. M., MONREAL D., GONZALEZ D., GARIN-BASTUJI B., DIAZ R., MAINAR-JAIME R. C., MORIYON I., BLASCO J. M. Efficacy of several serological tests and antigens for diagnosis of bovine brucellosis in the presence of false-positive serological results due to Yersinia enterocolitica O: 9. Clin. Diagn. Lab. Immunol., 2005, 12, 141-151. MUSA M.T., JAHANS K.L. The isolation of Brucella melitensis biovar 3 from a testicular hygroma of a ram in a nomadic flock of sheep and goats in western Sudan. J. Comp. Pathol., 1990, 103, 467-70. MVLABANK. MLVAbank for Bacterial Genotyping: (2010), http://minisatellites.upsud.fr/MLVAnet, accessed 11/01/2011. NEYMAN J., PEARSON E.S. On the use and interpretation of certain test criteria for purposes of statistical inference: Part I. Biometrika, 1928a, 20A, 175-240. NEYMAN J., PEARSON E.S. On the use and interpretation of certain test criteria for purposes of statistical inference: Part II. Biometrika, 1928b, 20A, 263-294.

131

References

NICOLETTI P. The epidemiology of brucellosis in animals. Dev. Biol. Stand., 1984, 56, 623628. NIELSEN K. Diagnosis of brucellosis by serology. Vet. Microbiol. , 2002, 90, 447-459. NIELSEN K., GALL D. Fluorescence polarization assay for the diagnosis of brucellosis: a review. J. Immunoass. Immunochem., 2001, 22, 183–201. NIELSEN K., GALL D., JOLLEY M., LEISHMAN G., BALSEVICIUS S., SMITH P., NICOLETTI P., THOMAS, F.. A homogeneous fluorescence polarization assay for detection of antibody to Brucella abortus. J. Immunol. Methods , 1996b, 195, 161168. NIELSEN K.H., KELLY L., GALL D., BALSEVICIUS S., BOSSE J., NICOLETTI P., KELLY W. Comparison of enzyme immunoassays for the diagnosis of bovine brucellosis. Prev. Vet. Med., 1996a, 26, 17-32. NIELSEN K.H., KELLY L., GALL D., NICOLETTI P., KELLY W. Improved competitive enzyme immunoassay for the diagnosis of bovine brucellosis. Vet. Immunol. Immunopathol., 1995, 46, 285-291. NYMO I.H., TRYLAND M., GODFROID J. A review of Brucella infection in marine mammals, with special emphasis on Brucella pinnipedialis in the hooded seal (Cystophora cristata). Vet. Res., 2011, 42, 93. O’HAGAN A. Bayesian statistics: principles and benefits. Wageningen UR Frontis Series, 2004, 3, 31-45. [online], http://library.wur.nl/frontis/bayes/03_o_hagan.pdf, accessed on 12/08/2013. OCHOLI R.A., EZEOKOLI C.D., AKEREJOLA O.O., SAROR D.I. Use of the enzymelinked immunosorbent assay for screening cattle for Brucella antibodies in Nigeria. Vet. Q., 1996, 181, 22-24. OCHOLI R.A., KWAGA J.K.P., AJOGI I., BALE J.O.O. Abortion due to Brucella abortus in sheep in Nigeria. Rev. sci. tech. Off. int. Epiz., 2005, 24, 973-979. OCHOLI R.A., KWAGA J.K.P., AJOGI I., BALE J.O.O. Phenotypic characterization of Brucella strains isolated from livestock in Nigeria. Vet. Microbiol., 2004, 103, 47-53.

132

References

O'GRADY D., BYRNE W., KELLEHER P., O'CALLAGHAN H., KENNY K., HENEGHAN T., POWER S., EGAN J., RYAN F. A comparative assessment of culture and serology in the diagnosis of brucellosis in dairy cattle. Vet. J.., 2014, doi: 10.1016/j.tvjl.2014.01.008. OIE (WORLD ANIMAL HEALTH ORGANIZATION). Chapter 1.1.5. Principles and methods of validation of diagnostic assays for infectious diseases. In: Manual of Diagnostic Tests and Vaccines for Terrestrial Animals, Paris, France, 2013, 1-16. http://www.oie.int/fileadmin/Home/fr/Health_standards/tahm/1.01.05_VALIDATION .pdf , accessed 02/06/2014. OIE (WORLD ANIMAL HEALTH ORGANIZATION). Chapter 2.4.3. Bovine Brucellosis. In: Manual of Diagnostic Tests and Vaccines for Terrestrial Animals, Paris, France, 2009,

1-35,

http://www.oie.int/fileadmin/Home/eng/Health_standards/tahm/2.04.03_BOVINE_B RUCELL.pdf, accessed 26/02/2013. OLSEN S., TATUM F. Bovine brucellosis. Vet. Clin. North Am. Food Anim. Pract., 2010, 26, 15–27. OLSEN S.C., HENNAGER S.G. Immune responses and protection against experimental Brucella suis biovar 1 challenge in non-vaccinated or B. abortus strain RB51vaccinated cattle. Clin. Vaccine Immunol., 2010, 17, 1891-1895. ORGANIZATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT (OECD) Livestock and regional market in the Sahel and West Africa: potentials and challenges:

Paris,

France,

2008,170

p.

[online]

http://www.oecd.org/swac/publications/41848366.pdf. OSTERMAN B., MORIYON I. International Committee on Systematics of Prokaryotes Subcommittee on the taxonomy of Brucella. Minutes of the meeting, September, 17th 2003, Pamplona, Spain. Int. J. Syst. Evol. Microbiol., 2006, 56, 1173–1175. OUAGAL

M.,

BERKVENS

D.,

HENDRIKX

P.,

FECHER-BOURGEOIS

F.,

SAEGERMAN C. Estimation du coût d’un réseau d’épidémiosurveillance des maladies animales en Afrique centrale : le cas du réseau tchadien. Rev. sci. tech. Off. int. Epiz., 2012, 31 (3), 809-819. 133

References

OUAGAL M., HENDRIKX P., BERKVENS D., NCHARE A., CISSE B., AKPELI P.Y., SORY K., SAEGERMAN C. Les réseaux d’épidémiosurveillance des maladies animales en Afrique francophone de l’Ouest et du Centre. Rev. sci. tech. Off. int. Epiz., 2008, 27 (3), 689-702. PAPPAS G., AKRITIDIS N., BOSILKOVSKI M., TSIANOS E. Brucellosis. N. Engl. J. Med., 2005, 352, 2325–36. PAPPAS G., PAPADIMITRIOU P., AKRITIDIS N., CHRISTOU L., TSIANOS E.V. The new global map of human brucellosis. Lancet Infect. Dis. , 2006, 6, 91-99. PEPE M.S., JANES H. Insights into latent class analysis of diagnostic test performance. Biostatistics, 2007, 8(2), 474–484. PERRY B. Chapter 7-Animal disease impact on the poor: study results. In: Investing in Animal Research to Alleviate Poverty. Nairobi, International Livestock Research Institute,

2002,

67–78.

http://ilri.org/infoserv/Webpub/fulldocs/InvestAnim/Book1/media/PDF_chapters/B1_ 7.pdf PFEIFFER D.U. Veterinary Epidemiology-An Introduction. Royal Veterinary College, Hawkshead Lane, United Kingdom, 2002, pp. 47-51. PILO-MORON E., PIERRE F., KOUAME J.B. Brucellose bovine en Côte d’ivoire. Epidémiologie. Rev. Elev. Med. vet. Pays trop., 1979, 32, 325-333. POESTER F.P., GONCALVES V. S. P., LAGE A. P. Brucellosis in Brazil. Vet. Microbiol. , 2002, 90, 55-62. PORTANTI O., TITTARELLI M., DI FEBO T., LUCIANI M., MERCANTE T., CONTE A., LELLI R. Development and validation of a competitive ELISA kit for the serological diagnosis of ovine, caprine and bovine brucellosis. J. Vet. Med. B Infect. Dis. Vet. Public Health, 2006, 53, 494-498. POUILLOT R. Response from R. Pouillot, G. Gerbier and I.A. Gardner to D.L. Berkvens et al. (2002). Prev. Vet. Med., 2003, 59 (3), 185-187.

134

References

POUILLOT R., GERBIER G., GARDNER I. A. “TAGS”, a program for the evaluation of test accuracy in the absence of a gold standard. Prev. Vet. Med., 2002, 53, 67–81. POUILLOT R., LESCOAT PH., GARIN-BASTUJI B., REPIQUET D., TERRIER P., GERBIER G., BENET J.J., SANAA M. Risk factors for false-positive serological reactions for bovine brucellosis in Saône-et-Loire (France). Prev. Vet. Med., 1998, 35, 165–179. PRAET N., DORNY P., SAEGERMAN C., MARCOTTY T., BERKVENS D. Estimation de la prévalence d’une maladie et des caractéristiques des tests diagnostiques par une approche bayésienne. Rev. Epid. San. Anim. , 2006, 49, 113-130. PUTT S.N.H., SHAW A.P.M., WOODS A.J., TYLER L., JAMES A.D. The Epidemiological Approach to Investigating Disease Problem. In Veterinary epidemiology and economics in Africa: A manual for use in the design and appraisal of livestock health policy. Veterinary Epidemiology and Economics Research Unit, Department of Agriculture, University of Reading: Reading, Berkshire, England, 1988, [online] http://www.fao.org/wairdocs/ilri/x5436e/x5436e06.htm#4.

the

epidemiological

approach to investigating disease problems. accessed on 24/06/2014. QU Y., HAGDU A. A model for evaluating the sensitivity and specificity for correlated diagnostic tests in efficacy studies with an imperfect reference test. J. Am. Statist. Assoc., 1998, 93, 920-928. QUINN P.J., CARTER M.E., MARKEY B., CARTER G.R. Brucella species. In: Clinical Vet. Microbiol.. Mosby International Limited: Edinburgh, 1999, 261-267. RAGAN V.E. The Animal and Plant Health Inspection Service (APHIS) brucellosis eradication program in the United States. Vet. Microbiol., 2002, 90, 11-18. RENOUX G. La classification des Brucella. Remarques à propos de l'identification de 2,598 souches. Ann. Inst. Pasteur , 1952, 82, 289–298. RIVERA S.A., RAMIREZ M. C., LOPETEGUI I. P. Eradication of bovine brucellosis in the 10th Region de Los Lagos, Chile. Vet. Microbiol. , 2002, 90 , 45-53.

135

References

ROBINSON A. Guidelines for Coordinated Human and Animal Brucellosis Surveillance. FAO Animal production and health paper: Rome,

Italia, 2003, 46 p.

http://www.fao.org/fileadmin/user_upload/drought/docs/Brucellosis.pdf. ROGAN W.J., GLADEN B. Estimating prevalence from the results of a screening test. Am. J. Epidemiol., 1978,107, 71-76. RON-ROMÁN J., SAEGERMAN C., MINDA-ALUISA E., BENÍTEZ-ORTÍZ W., BRANDT J., DOUCE R. First Report of Orchitis in Man Caused by Brucella abortus biovar 1 in Ecuador. Am. J. Trop. Med. Hyg. , 2012, 87, 524-528. ROOP R.M., GEE J.M., ROBERTSON G.T., RICHARDSON J.M., NG W.-L., WINKLER M.E. Brucella stationary-phase gene expression and virulence. Annu. Rev. Microbiol., 2003, 57, 57–76. ROTH F, ZINSSTAG J, ORKHON D, CHIMID-OCHIR G, HUTTON G, COSIVI O., GARRIN G., OTTE J. Human health benefits from livestock vaccination for brucellosis: case study. Bull. World Health Organ., 2003, 81, 867–876. RUTJES A., REITSMA J., COOMARASAMY A., KHAN K., BOSSUYT P. Evaluation of diagnostic tests when there is no gold standard. Health Technol. Assess., 2007, 11(50), 1–51. SAEGERMAN C. Epidémiosurveillance des évènements rares chez les bovins en Belgique. Thèse de Doctorat en sciences vétérinaires, Université de Liège: Liège, Belgique, 2004, http://www.dmipfmv.ulg.ac.be/epidemiovet/Theses/These%20Saegerman.pdf. SAEGERMAN C., BERKVENS D., GODFROID J., WALRAVENS K. Bovine brucellosis. In: Infectious and parasitic diseases of livestock. Lavoisier et Commonwealth Agricultural Bureau – International (ed.): Paris, France, 2010, 971-1001. SAEGERMAN C., DAL POZZO F., HUMBLET M.-F. Reducing hazards for humans from animals: emerging and re-emerging zoonoses. Italian J. Public Health, 2012, 9(2), 1324. SAEGERMAN C., DE WAELE L., GILSON D., GODFROID J., THIANGE P., MICHEL P., LIMBOURG B., VO T.K.O., LIMET J., LETESSON J.J., BERKVENS D.

136

References

Evaluation of three serum i-ELISAs using monoclonal antibodies and protein G as peroxidase conjugate for the diagnosis of bovine brucellosis. Vet. Microbiol., 2004, 100, 91-105. SAEGERMAN C., THIANGE P., LIMBOURG B., CONOTTE G., PETIT N., THIRY G., BOTTON Y., PELZER P., MULLIER P., GODFROID J., DUFEY J. Etude épidémiologique descriptive et identification de facteurs de risque des réactions sérologiques faussement positives en brucellose bovine dans le sud de la province de Namur (Belgique). In: Proceedings of the VIII Symposium International d’Epidémiologie et d’Economie Vétérinaire, Paris, 8–11 July 1997. Rev. Epid. San. Anim. , 1997, 32, 08.04.1–08.04.3. SAEGERMAN C., VO T.K.O., DEWAELE L., GILSON D., BASTIN A., DUBRAY G., FLANAGAN P., LIMET J.N., LETESSON J.J, GODFROID J. Bovine brucellosis diagnosis by skin test: conditions for its use and evaluation of its performance. Vet. Rec., 1999,145, 214-218. SAMAHA H., AL-ROWAILY.M., KHOUDAIR R.M., ASHOUR H.M. Multicenter study of brucellosis in Egypt. Emerg. Infect. Dis., 2008, 14, 1916–1918. SAMARTINO L., GALL D., GREGORET R., NIELSEN K. Validation of enzyme-linked immunosorbent assays for the diagnosis of bovine brucellosis. Vet. Microbiol. , 1999a, 70, 193-200. SAMARTINO L., GREGORET R., GALL D., NIELSEN K. Fluorescence Polarization Assay: Application to the diagnostic of bovine brucellosis in Argentina. Journal of immunoassay, 1999b, 20 (3), 115-126. SAMARTINO L.E. Brucellosis in Argentina. Vet. Microbiol. ,2002, 90, 71-80. SAMARTINO L.E., GIL A., ELZER P. Capacity building for surveillance and control of bovine and caprine brucellosis. In: Capacity building for surveillance and control of zoonotic diseases. FAO Animal Production and Health Proceedings, no. 7; FAO/WHO/OIE Expert and Technical Consultation on Capacity Building for Surveillance and Control of Zoonotic Diseases, Rome, Italy, 2005, 55-66.

137

References

SANOGO M., CISSE B., OUATTARA M., WALRAVENS K., PRAET N., BERKVENS D. THYS E. Prévalence réelle de la brucellose bovine dans le centre de la Côte d’Ivoire. Rev. Elev. Med. vet. Pays trop., 2008, 61, 147-151. SANOGO M., THYS E., ACHI Y.L., FRETIN D., MICHEL P., ABATIH E., BERKVENS D., SAEGERMAN C. Bayesian estimation of true prevalence, sensitivity and specificity of Rose Bengal test and indirect ELISA for the diagnosis of bovine brucellosis. Vet. J., 2013a, 195, 114-120. SANOGO M., ABATIH E., THYS E., FRETIN D., BERKVENS D., SAEGERMAN C. Importance of identification and typing of Brucellae from West African cattle: a review. Vet. Microbiol., 2013b, 164, 202–211. SANOGO M., ABATIH E., THYS E., FRETIN D., BERKVENS D. SAEGERMAN C. Risk factors associated with brucellosis seropositivity among cattle in the central savannahforest area of Ivory Coast. Prev Vet Med., 2012, 107(1-2), 51-56. SANOGO M., ABATIH, E., SAEGERMAN, C. Bayesian versus frequentist methods for estimating disease true prevalence and diagnostic test performance. Vet. J., 2014, 202, 204-207. SCHELLING E., DIGUIMBAYE C., DAOUD S., NICOLET J., BOERLIN P., TANNER M., ZINSSTAG J. Brucellosis and Q-fever seroprevalences of nomadic pastoralists and their livestock in Chad. Prev. Vet. Med., 2003, 61, 279-293. SCHOLZ H.C., AL DAHOUK S., TOMASO H., NEUBAUER H., WITTE A., SCHLOTER M., KAMPFER P., FALSEN E., PFEFFER M., ENGEL M. Genetic diversity and phylogenetic relationships of bacteria belonging to the Ochrobactrum-Brucella group by recA and 16S rRNA gene-based comparative sequence analysis. Syst. Appl. Microbiol. , 2008b, 31 (1), 1–16. SCHOLZ H.C., HUBALEK Z., SEDLÁCEK I., VERGNAUD G., TOMASO H., AL DAHOUK S., MELZER F., KÄMPFER P., NEUBAUER H., CLOECKAERT A., MAQUART M., ZYGMUNT M.S., WHATMORE A.M., FALSEN E., BAHN P., GÖLLNER C., PFEFFER M., HUBER B., BUSSE H.J., NÖCKLER K. Brucella microti sp. nov., isolated from the common vole Microtus arvalis. Int. J. Syst. Evol. Microbiol., 2008, 58, 375-382. 138

References

SCHOLZ H.C., NOCKLER K., GOLLNER C., BAHN P., VERGNAUD G., TOMASO H., AL DAHOUK S., KAMPFER P., CLOECKAERT A., MAQUART M., ZYGMUNT M.S., WHATMORE A.M., PFEFFER M., HUBER B., BUSSE H.J., DE B.K. Brucella inopinata sp. nov., isolated from a breast implant infection. Int. J. Syst. Evol. Microbiol., 2010, 60, 801-808. SCHOLZ, H.C., VERGNAUD G. Molecular characterisation of Brucella species. Rev. sci. tech. Off. int. Epiz., 2013, 32(1),149-62. SCHURIG G.G., SRIRANGANATHAN N., CORBEL M. J. Brucellosis vaccines: past, present and future. Vet. Microbiol. , 2002, 90, 479-496. SELEEM M.N., BOYLE S.M., SRIRANGANATHAN N. Brucellosis: A re-emerging zoonosis. Vet. Microbiol. , 2010, 140, 392–398 SENG P., DRANCOURT M., GOURIET F., LA SCOLA B., FOURNIER P.E., ROLAIN J.M., RAOULT D. Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Clin. Infect. Dis., 2009, 49,543-551. SHEY-NJILA O., DAOUDA N.E., ZOLI P. A., WALRAVENS K., GODFROID J., GEERTS S. Serological survey of bovine brucellosis in Cameroon. Rev. Elev. Med. vet. Pays trop., 2005, 58, 139-143. SMITS H.L., ABDOEL T.H., SOLERA J., CLAVIJO E., DIAZ R. Immunochromatographic Brucella-specific immunoglobulin M and G lateral flow assays for the serodiagnosis of human brucellosis. Clin. Diagn. Lab. Immunol., 2003, 10, 1141–1146. SPEYBROECK

N.,

DEVLEESSCHAUWER

B.,

JOSEPH

L.,

BERKVENS

D.

Misclassification errors in prevalence estimation: Bayesian handling with care. Int. J. Public Health, 2012a, doi: 10.1007/s00038-012-0439-9. SPEYBROECK N., PRAET N., CLAES F., HONG N.V., TORRES K., MAO S., VAN DEN EEDE P., THI THINH T., GAMBOA D., SOCHANTHA T., THANG N.D., COOSEMANS M., BÜSCHER P., D'ALESSANDRO U., BERKVENS D., ERHART A. True versus apparent malaria infection prevalence: The contribution of a Bayesian approach. PLoS ONE , 2011, 6 (2), e16705.

139

References

SPEYBROECK N., WILLIAMS C.J., LAFIA K.B., DEVLEESSCHAUWER B., BERKVENS D. Estimating the prevalence of infections in vector populations using pools of samples. Med.Vet. Entomol., 2012b, 26, 361–371. SPIEGELHALTER D.J., BEST N.G., CARLIN B.P., VAN DER LINDE A. Bayesian measures of model complexity and fit. J. R. Stat. Soc. Series B, 2002, 64, 583-640. SPIEGELHALTER D.J., THOMAS A., BEST N., LUNN D. WinBUGS Version 1.4 User manual, 2003, http://www.politicalbubbles.org/bayes_beach/manual14.pdf, accessed on 24/06/2014. STAQUET M., ROZENCWEIG M., LEE Y.J., MUGGIA F.M. Methodology for the assessment of new dichotomous diagnostic tests. J. Chronic Dis., 1981, 34, 599-610. STARKEY P. Livestock for traction: world trends, key issues and policy implications. AGAL

working

paper

series.

FAO,

Rome,

Italia,

2010,

http://www.healingharvestforestfoundation.org/uploads/1/7/0/8/17089550/livestockfor-traction.pdf, accessed on 24/06/2014. STEPHEN C.N. Chapter 15: Logistic regression and Cox regression, pp. 295-304. In: Biostatistical methods in epidemiology, John Wiley & Sons, Inc. ISBN: 0-471-369144, 2001. SZULOWSKI K., IWANIAK W., WEINER M., ZLOTNICKA J. Brucella suis biovar 2 isolations from cattle in Poland. Ann. Agric. Environ. Med., 2013, 20(4), 672-675. THIENPONT D., VANDERVELDEN M., FAGARD P., MORTELMANS J. L'hygroma brucellique: l'aspect clinique caractéristique de la brucellose bovine au RwandaBurundi. Rev. Elev. Med. vet. Pays trop., 1961, 14, 257-266. THORNTON P.K., KRUSKA R.L., HENNINGER N., KRISTJANSON P.M., REID R.S., ATIENO F., ODERO A.N., NDEGWA T. Mapping poverty and livestock in the developing world. ILRI (International Livestock Research Institute), Nairobi, Kenya, 2002, 124p. THRUSFIELD M. Diagnostic testing. In: Veterinary Epidemiology. Blackwell Science: Oxford, 2005, 303–330.

140

References

THYS E., YAHAYA M.A., WALRAVENS K., BAUDOUX C., BAGAYOKO I., BERKVENS D., GEERTS S. Etude de la prévalence de la brucellose bovine en zone forestière de la Côte d'Ivoire. Rev. Elev. Med. vet. Pays trop. , 2005, 58, 205-209. TILLER R.V., GEE J.E., LONSWAY D.R., GRIBBLE S., BELL S.C., JENNISON A.V., BATES J., COULTER C., HOFFMASTER A.R., DE B.K. Identification of an unusual Brucella strain (BO2) from a lung biopsy in a 52 year-old patient with chronic destructive pneumonia. BMC Microbiol. , 2010,10, 23. TIMM B.M. Brucellosis. Distribution in Man, domestic and wild animals. Berlin: Springer. 1982, 60 p. TOUNKARA K., MAIGA S., TRAORE A., SECK B.M., AKAKPO A.J. Epidémiologie de la Brucellose Bovine au Mali: Enquête sérologique et isolement des premières souches de Brucella abortus. Rev. sci. tech. Off. int. Epiz., 1994, 13, 777-786. UNGER F., MÜNSTERMANN S., GOU MOU A., APIA C. N., KONTE M., HEMPEN M. A. Risk associated with bovine brucellosis in selected study herds and market places in four countries of West Africa, pp 16-54, Animal Health Working Paper 2. ITC, 2003. USDA

(UNITED

STATES

DEPARTMENT

OF

AGRICULTURE),

ILRI

(INTERNATIONAL LIVESTOCK RESEARCH INSTITUTE). An integrated approach to controlling brucellosis in Africa: Workshop report. ILRI: Nairobi, Kenya, 2013, 32 p. VALLVERDÚ J. The False Dilemma: Bayesian vs. Frequentist. E – LOGOS Electronic journal for philosophy, 2008, 1-17. VERGER J., GRIMONT F., GRIMONT P. GRAYON M. Taxonomy of the genus Brucella. Ann. Inst. Pasteur Microbiol., 1987,138, 235–238 VERGER J., GRIMONT F., PAD G., GRAYON M. Brucella, a monospecific genus as shown by deoxyribonucleic acid hybridization. Int. J. Syst. Bacteriol. ,1985, 35, 292– 295.

141

References

VERGER J.M., GRAYON M. Characteristics of 273 strains of Brucella abortus of African origin. Developments in Biological Standardization , 1984, 56, 63-71. VERGER J.M., GRAYON M., CHANTAL J., AKAKPO J.A. Characteristics of Togo strains of Brucella abortus from cattle. Ann. Rech. Vet., 1982,13, 177-184. VERGER J.M., GRAYON M., DOUTRE M.P., SAGNA F. Brucella abortus d’origine bovine au Sénégal : Identification et typage. Rev. Elev. Med. vet. Pays trop., 1979, 32, 25-32. VERGER M., GARIN-BASTUJI B., GRAYON M., MAHE A.M. La brucellose bovine à Brucella melitensis en France. Ann. Rech. Vet., 1989, 20, 93-102. WAGNER B., SALMAN M.D. Strategies for two-stage sampling designs for estimating herd-level prevalence. Prev. Vet. Med., 2004, 66, 1–17. WALTER S.D., IRWIG L.M. Estimation of test error rates, disease prevalence and relative risk from misclassified data: a review. J. Clin. Epidemiol.,1988, 41, 923–927. WASTLING J.M, AKANMORI B.D.; WILLIAMS D.J.L. Zoonoses in West Africa: impact and control. Parasitol. Today ,1999, 15, 309-310. WHATMORE A.M. Current understanding of the genetic diversity of Brucella,

an

expanding genus of zoonotic pathogens. Infect. Genet. Evol., 2009, 9, 1168-1184. WHATMORE A.M., DAVISON N., CLOECKAERT A., AL DAHOUK S., ZYGMUNT M.S., BREW S.D., PERRETT L.L., KOYLASS M.S., VERGNAUD G., QUANCE C., SCHOLZ H.C., DICK E.J., HUBBARD G., SCHLABRITZ-LOUTSEVITCH N.E. Brucella papionis sp. nov. isolated from baboons (Papio spp.). Int. J. Syst. Evol. Microbiol., 2014, doi:10.1099/ijs.0.065482-0 (In press). WEYNANTS V., GILSON D., CLOECKAERT A., TIBOR A., DENOEL P. A., GODFROID F., LIMET J. N., LETESSON J. J. Characterization of smooth lipopolysaccharides and O polysaccharides of Brucella species by competition binding assays with monoclonal antibodies. Infect. Immun., 1997, 65, 1939-1943.

142

References

WEYNANTS V., GODFROID J., LIMBOURG B., SAEGERMAN C., LETESSON J. J. Specific Bovine Brucellosis Diagnosis Based on In-Vitro Antigen-Specific GammaInterferon Production. J. Clin. Microbiol. , 1995, 33, 706-712. WORLD HEALTH ORGANIZATION (WHO). Zoonoses and veterinary public health. Seven

neglected

endemic

zoonoses.

Basic

facts

http://www.who.int/zoonoses/neglected_zoonotic_diseases/en/

[online] ,

(2012),

accessed

on

24/06/2014. WRIGHT R.W., BRAND R.A., DUNN W., SPINDLER K.P. How to write a systematic review. Clinical Orthopaedics and Related Research , 2007,455, 23-29. WYATT H.V. Lessons from the history of brucellosis. Rev. sci. tech. Off. int. Epiz. , 2013, 32 (1), 17-25 XAVIER M.N., PAIXAO T.A., POESTER F.P., LAGE A.P., SANTOS R.L. Pathological, immunohistochemical and bacteriological study of tissues and milk of cows and fetuses experimentally infected with Brucella abortus. J. Comp. Pathol. , 2009, 140, 149-157. YU W.L., NIELSEN K. Review of detection of Brucella spp. by Polymerase Chain Reaction. Croat. Med. J. , 2010, 51, 306-313. ZAMMIT T. A preliminary note on the examination of the blood of goats suffering from Mediterranean fever. Reports of the Commission on Mediterranean Fever, Part III, Harrison and Sons, London, 1905, p 83. ZINSSTAG J. Convergence of ecohealth and one health. Ecohealth, 2013, 9(4), 371–373.

143

ANNEXES

144

Annexes Annex 1: Map showing the western part of Africa and the neighbour countries of Ivory Coast

145

Annex 2: Differential characteristics of biovars of Brucella species (from OIE, 2009)

146

Annex 3 : A cow with a carpal hygroma

(Credit picture: M. Sanogo)

147