Experimental Brain Tumors - Cancer Research

Experimental Brain Tumors - Cancer Research


8MB Sizes 0 Downloads 15 Views


MAY, 1944


Experimental Brain Tumors V. Behavior in Intraocular Transplants* David Freeman, M.D., and H. M. Zimmerman, M D . (From the Laboratory o/Pathology, Yale University School o/ Medicine, New Haven, Connecticttt)

(Received for publication December 20, 1943) Two considerations prompted the performance of the experiments that will be detailed in this communication. One was the desirability of studying the growth and behavior of transplants of a variety of chemically induced brain tumors in mice from their early inception to maturity. The anterior chamber of the eye was selected as the most favorable site for these experiments because the transparent cornea affords an opportunity to keep the transplants under constant observation, even with the aid of a microscope when necessary. The second consideration concerned itself with the question whether these experimentally induced brain tunmrs are genuinely malignant or, at least, whether they have one of the important characteristics of malignancy; namely, self-sufficiency or autonomy. The concept of autonomous growth as evidenced by the ability of tumor tissue to survive in a foreign environment such as the anterior chamber of the eye has been elaborated by Greene (3, 4). It is perhaps doubtful whether this criterion of malignant tumors is of greater diagnostic value than anaplasia, local invasiveness, distant metastasis, or an impairment of the cocarboxylase enzyme system (1). Previous reports dealing with primary brain tumors induced experimentally (2, 5, 6) have emphasized the local invasive proclivities of these neoplasms and, in the case of certain of the gliogenous tumors, have shown the tendency to metastasize along the spinal fluid pathway. But none of the tumors, especially the gliomas, ever metastasized extracranially, a behavioral characteristic shared by human gliomas as well. To be sure, many of the new growths displayed striking evidence of anaplasia, but it was felt that if these tumors showed the property of autonomous growth in the anterior chamber of the eye their malignant nature would be established beyond reasonable doubt. * This investigation was aided by a grant from The Jane Coffin Childs Fund for Medical Research.

METHODS The brain tumors employed in these transplantation experiments were all' induced with 20-methylcholanthrene in C3H and Bagg albino mice. The animals into whose anterior eye chambers the transplants were made were C3H, dba, and ABC albino mice, as well as guinea pigs. Technic o[ t u m o r implantation in the mouse's eye.--

A suspension of tumor cells was prepared in the usual manner. The finest hypodermic needle, 27 gauge, 88 inch length, fitting the ordinary type Luer tuberculin syringe, was employed. It was found that explants measuring 0.5 mm. in diameter could easily pass through a needle of this gauge. If clumps of tumor cells became packed near the exit opening of the syringe, increased pressure rnerely formed larger compact masses that clogged the needle. In an attempt to avoid this difficulty the opening of the ordinary tuberculin syringe was reamed out to a diameter of 2 ram. However, the same difficulty sometimes occurred within the needle base, where it tapers sharply. By withdrawing the suspension from the needle immediately after each injection and constantly agitating the syringe these difficulties could be avoided. It was necessary to hone the needle to an exceptionally sharp point in order to penetrate the almost spherically convex cornea with minimal trauma. This was done by placing a small oil stone on a dissecting microscope stage and under observation passing the needle, fitted to a small syringe without the plunger, lightly along the stone. The syringe was held at about 15~ to the upper surface of the stone and was twirled between the thumb and first two fingers so that it made a complete revolution with each motion. A stylet was inserted in the needle and was allowed to extend just to the needle tip and no farther. Its presence prevented wobbling due to the bevel of the needle, and uneven honing. About 10 or 12 double passages


Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.


Cancer Research

across the stone were found sufficient, each additional stroke tending to spoil the point. Honing with jeweler's lathes and dentist's wheels gave inferior results. Mature mice were preferred because the eyes are larger. Light ether anesthesia was chosen in preference to nembutal since the animals recover more quickly from its effects, and the narcosis can be extended by a few drops on a wisp of cotton. The mouse was placed on a small board on the stage of a dissecting microscope with a small spot of strong light focused on the eye. The left eye was selected for right-handed operators. The animal's head was steadied between the index finger and thumb of the left hand, the vibrissae were clipped, and extremely gentle pressure was made against the base of the occiput with the left middle finger, causing the left eye to bulge outward slightly. Too great pressure caused intracranial trauma. The microscope, with a magnification of about 10 diameters, was focused on the iris. An assistant, having the syringe and suspension prepared, handed the syringe, with the bevel up, to the operator. The point of the needle was introduced at the limbus, on a line with the external canthus. Gentle pressure was made until the entire opening of the needle was within the anterior chamber. While the syringe was steadied by the operator, pressure was made on the plunger by the assistant. The operator controlled the amount of injection by signalling to stop when several fragments of tun-mr entered the anterior chamber. The needle was then deliberately and quickly removed at the same time that the pressure against the head was released. Nothing was gained by injecting a large mass into the eye and distorting normal relations. Because the anterior chamber of the mouse's eye is very shallow, there was difficulty in avoiding the lens and iris. Injection of some of the tumor material into the iris, lens, and even the vitreous body often took place. It was found, however, that this seeding of tumor cells in the various structures was an actual advantage since their extension from the posterior into the anterior chamber by way of the pupil afforded excellent material for study. The injected eye was first examined 24 hours after operation. For this procedure an ophthalmologist's (Gullstrand) slit-lamp and microscope were used. It was not necessary to anesthetize the animal for daily examinations since sufficient manual immobilization was possible for a magnification of about 20 diameters. Higher magnifications necessitated anesthesia. Loss of eyes from infection was not experienced. Slight corneal cloudiness, traumatic cataract, anterior and posterior synechiae, hyphemia, and other changes were occasionally seen, but these changes did not appear to interfere with tumor growth.

Technic o/ tumor implantation in the guinea pig's eye.--The suspensions of the primary mouse brain tumors were prepared for intraocular implantation in the guinea pig in a manner similar to that for the mice. Under nembutal anesthesia the cornea of the left eye was nicked with a scalpel to aid the introduction of an ordinary 27 gauge needle. In general, both larger total amounts and larger individual pieces of brain tumor tissue were introduced into the anterior chambers of the eyes of these animals than into those of the mice. In all other respects this animal species was treated and studied in a manner strictly comparable to that employed for the mice. RESULTS The experimental brain tumors employed in these intraocular transplantation experiments were induced w i t h 20-methylcholanthrene in C3H and Bagg albino mice. They have already been reported in some detail elsewhere (5, 7) and hence reference to their microscopic appearances will be brief in this report. Emphasis, rather, will be placed here on the characteristics of their growth behavior in intraocular transplants. For comparative purposes the appearances of certain extracranial sarcomas will also be described. Finally, details will be given of the intraocular growth of a human glioblastoma multiforme. Astrocytoma.--The material for transplantation into the anterior charnber of the eye was derived from mouse 69 and was described in the paper on tumors produced with methylcholanthrene (5), where its morphologic features were illustrated in Fig. 13. Four C3H and an equal number of ABC albino mice received intraocular transplants of this neoplasm. After a period of quiescence that varied from 10 to 19 days, the tumor began to grow in the anterior chambers of the C3H mice, filling them completely by the 35th day. There was invasion of the posterior chambers through the pupils, and the iris and cornea of each animal were likewise infiltrated by the neoplasm. An outstanding characteristic of this tumor was its ivory white color and the almost complete lack of vascularity and hemorrhage (Fig. 1). Microscopically the tumor cells had the same characteristics as those of the primary neoplasm, even as regards astrocytcs in mitotic division. The tumor transplants behaved quite differently in the ABC albino mice. Within a week all 4 animals showed evidence of active tumor growth, then recession began so that the eye was clear in 9 days in the first mouse, t5 days in the second and third mice, and 32 days in the fourth mouse. Medulloblastoma.--The brain tumor employed in this experiment came from mouse 49 and was illustrated in Fig. 4 of the paper on experimentally induced

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.

Freemall and Zimmerman--Experimental Brain Tumors. V brain tumors with methylcholanthrene (5). Intraocular transplants were made into 5 C3H, 4 ABC albino, and 3 dba mice. All animals of the first strain showed active tumor growth in an average of 14 days. The anterior chambers became filled with the neoplasm, which invaded and destroyed the irides. The posterior chambers then became filled by extension through the pupils. Ultimately there was invasion of the vitreous body, the cornea, and, in at least one instance, the retina as well. Melanin pigment granules appeared in the solid milky white tumors after the irides were invaded. Delicate blood vessels developed in the transplants after these had reached a considerable size (Fig. 2). Initially, all 4 ABC albino mice showed evidence of growth of the transplants in about 2 weeks, but shortly thereafter the tun'mrs disappeared from the eyes of 2 of the animals. The remaining 2 mice, however, had vigorously growing transplants until the animals were killed in order to permit microscopic study of the neoplasms. The appearance of both intraocular tumors was similar to that in the C3H mice. Only 1 of the dba mice receiving intraocular transplants of the brain tumor showed any signs at all of tumor growth in the anterior chamber. A month elapsed before the transplant increased in size appreciably, but ultimately a solid milky white mass of tissue filled half of the anterior chamber and extended through the pupil into the posterior chamber. The intraocular tumor from ! of the original 5 C3H mice was employed for subtransplantation into 4 C3H and 4 ABC albino mice. One animal of the first group and 2 of the second had takes. All 3 subtransplants were growing actively by the 12th day. In the C3H mouse the tumor again grew as a solid milky white mass; in the ABC albino animals the tumors were finely granular but also milky white in color. The microscopic appearance of all the individual transplants was essentially identical. The tumor cells were somewhat elongated, not unlike fibroblasts, but there was practically no stroma. Many of the cells were in mitotic division. Frequent pseudorosette formations helped identify the transplants as medulloblastoma. Ependymoblastoma.--Two guinea pigs received intraocular transplants from a brain tumor induced in a Bagg albino mouse with methylcholanthrene (mouse 20). This animal was included in a report on the incidence of experimentally induced brain tumors in different strains of mice (7). The tumor was a typical ependymoblastoma with numerous rosettes and was similar to the tumor of this variety illustrated in Fig. 2 of the communication dealing with brain tumors induced with benzpyrene (6). Within 2 days after intraocular transplantation the tumor showed signs of growth in both guinea pigs' eyes. Within 4 days the


flocculent tumor material extended over much of the irides and became implanted on the anterior surface of the lens capsules. Individual tumor nodules began to coalesce and form larger grey masses, which received their vascular supply from delicate vessels arising at the inferior angle of the anterior chamber (Fig. 3). The transplants continued their active growth for 20 days until the eyes were removed for microscopic study. This disclosed a rapidly growing tumor, as indicated by the presence of numerous cells in mitotic division, with many true rosettes. Unclassified cerebral glioma.--The primary brain tumor was derived from mouse 74 and presented considerable diagnostic difficulty, which has been discussed elsewhere (5). This neoplasm was a highly malignant glioma with many features suggestive of a glioblastoma. Its histologic details were portrayed in Fig. 12 of the article mentioned above. Five C3H and 4 ABC albino mice received transplants of this tumor. These transplants grew in only 1 animal of the C3H and in 2 of the ABC albino strain. After 2 weeks of quiescence active growth was recognized by the appearance of discrete milky white nodules, which soon began to fuse and become highly vascularized. All the intraocular structures were invaded by the neoplasm and hemorrhages, some of fair size, began to appear as the eyeball became greatly enlarged. When rupture threatened, the eyes were removed for microscopic study. The original brain tumor had been transplanted subcutaneously in the homologous strain of mice, where it continued to grow through 12 generations of subtransplants. Pieces of this tumor derived from the third generation of subcutaneous transplant were again implanted in the eyes of 3 C3H mice, and 1 animal developed a take. This occurred within 4 days after implantation. Growth proceeded rapidly and went through the stages already described for this tumor (Fig. 4). T h e microscopic appearance of all 4 intraocular transplants of this neoplasm was similar. There was great cellular pleomorphism with many cells in mitosis. The stroma was scant and none of it was of mesodermal origin. Unclassified cerebellar glioma.--The brain tumor that was induced in a C3H mouse (mouse 42) with methylcholanthrene was transplanted into the anterior chambers of the eyes of 4 C3H and 4 ABC albino mice. This was a cerebellar glioma that was difficult to classify since the tumor cells failed to form characteristic patterns. Two of the C3H mice showed evidence of growth in the transplants after 21 days, the tumors first filling the anterior chambers w~th solid milky white material, then invading the posterior chambers, later the vitreous body and lens capsules, and finally

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.


Cancer Research

rupturing the corneae. Growth proceeded from that stage to form large epithelialized masses that projected from the orbits (Fig. 5). Microscopic examination of the tumors revealed a core of gliogenous tissue surrounded by a scant a m o u n t of connective tissue, which served as the basement membrane for a stratified squa1-nous epithelial covering. As in the original tumor, these transplants failed to form histologic structures that would warrant a diagnosis of more than glioma. T h e other 2 C 3 H mice never showed signs of tumor growth. Three of the ABC albino mice appeared promising from the standpoint of transplant takes after some 3 weeks, but later all 3 tumors regressed rapidly. T h e transplanted material in the fourth animal showed no signs of growth from the start. Cerebellar fibrosarcoma.--The primary brain tumor was induced with methylcholanthrene in the cerebellum of a C 3 H mouse (mouse 145). It arose apparently from the leptomeninges overlying the vermis and invaded both lateral cerebellar lobes. It consisted of elongated, spindle-shaped cells, many in division, and had an a b u n d a n t stroma of mesodermal reticulum. Transplants of this tumor were made into the anterior chambers of the eyes of 15 C 3 H mice and 13 of the tumors were actively growing at the end of 6 days. T h e remaining 2 animals failed to develop takes. In all 13 positive mice the new growths assumed characteristics that distinguished them from any of the gliomas of the previous experiments. T h e tumor became highly vascularized almost at once and large hemorrhages appeared. There was much necrosis of the new growth, which became deeply pigmented in scattered zones because of the release of melanotic pigment from the irides. There also appeared to be a genuine proliferation of melanophores. These cells intermingled with the malignant elements of the fibrosarcoma. W i t h i n 2 weeks the eyeballs were replaced by tumor tissue and shortly thereafter the corneae ruptured. Extensive invasion of the orbital structures then followed, and in several of the mice that survived

the adjacent cranial sinuses were invaded by the tumor. Rhabdomyosarcoma.--The animal (mouse 9) whose tumor was employed for intraocular transplants developed a rhabdomyosarcoma of the scalp following intracerebral implantation of a pellet of methylcholanthrene. The pellet of carcinogen had worked its way out of the craniotomy wound and had come in contact with the temporal muscle, where the neoplasm originated. Nineteen C3H and 6 ABC albino mice were inoculated with this tumor and in all but 2 C3H animals active growth was established by the sixth day. These 2 mice, and 2 more of the same strain in which the tumors regressed after the tenth day, were the only animals that failed to be hosts to large intraocular neoplasms at the termination of this experiment on the 30th day. As in the case of the cerebellar fibrosarcoma described above, this rhabdomyosarcoma quickly developed an extensive vascular supply and soon after became hemorrhagic (Fig. 6). Parts of the tumor appeared gelatinous and parts granular with considerable brownish pigment. The latter was identified as melanin on microscopic examination and was absent only in the ABC albino mice (Fig. 7). The eyeballs were rapidly destroyed by the growing tumor, the corneae ruptured in all the animals that survived more than 3 weeks, and the orbital tissues were extensively infiltrated with tumor tissue. An opportunity presented itself to study still another rhabdomyosarcoma induced in the scalp tissues of a C3H mouse with methylcholanthrene (mouse 60). Intraocular transplants of this neoplasm were made in 10 C3H mice, 6 of which developed takes. In these animals the eyes were completely filled with the growing tumor in from 21 to 30 days. In each, the new growth was characterized by much vascular proliferation and hemorrhage, as well as by an excess of melanin pigment deposition. The appearance of the tumor in the eyes of these animals was remarkably similar to the tumor derived from mouse 9. A first generation subcutaneous transplant of this

DESCRIPTION OF FIGURES 1 TO 8 Fxc,. 1.--Astrocytoma. Transplant in anterior chamber of C3H mouse from primary brain tumor in mouse 69. Appearance 11 days after intraocular implantation. Fie. 2.--Medulloblastoma. Appearance of tumor in C3H mouse 28 days after transplantation from mouse 49. Fro. 3.--Ependymoblastoma. Appearance of tumor in guinea pig 13 days after intraocular transplantation from Bagg albino mouse 20. Fro. 4.--Unclassified cerebral glioma. Transplant in anterior chamber of C3H mouse. Tumor derived from mouse 74. Appearance 12 days after transplantation.

Fro. 5.--Unclassified cerebellar glioma. Extruded neoplasm 3 months after intraocular transplantation in C3H mouse. Primary tumor derived from C3H mouse 42. F1o. 6.--Rhabdomyosarcoma. Transplant on seventh clay of growth in eye of C3H mouse. Origin of primary tumor in animal of homologous strain (mouse 9). FIG. 7.--Rhabdomyosarcoma. Vascularized transplant in ABC albino mouse on 18th day. Primary tumor same as that shown in Fig. 6. Fio. 8.--Human glioblastoma multiforme. Appearance of transplant in anterior chamber of ABC albino mouse on 14th day. Primary tumor in 59 year old man.

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.

FIcs. 1-8

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.

Freeman and Zimmerman--Experimental Brain Tumors. V rhabdomyosarcoma was utilized for subtransplantation intraocularly in 13 C3H mice. Growth of the tumor did not succeed in 2 of the animals, but the remaining 11 mice developed tumors that filled the eyes completely in from 17 to 32 days. Once again the transplanted tissue was characterized by extensive invasiveness, great vascularity, and hemorrhage. Human glioblastoma multilorme.--This tunmr was removed at operation from the right parietal lobe of a white man 59 years of age. It was a typical glioblastoma on microscopic examination (report No. N.P. 941). A portion of this neoplasm was prepared for intraocular transplantation into 4 C3H and an equal number of ABC albino mice. In 3 of the animals of the former strain the transplants were observed to grow well on the fifth day, but in the fourth animal regression began almost at once and was complete in 11 days. In 2 of the mice in which growth was obvious on the fifth day, regression began on the 17th day and was complete in 3 weeks. In the third animal regressive changes began on the tenth day and the anterior chamber was completely free of tumor at the end of 30 days. The tumor transplants were characterized by a milky white, avascular, and nonhemorrhagic appearance, and closely resembled gliogenous tumors induced in mice with methylcholanthrene. Examination of the anterior chambers of the eyes of 2 of the ABC albino mice disclosed regression of the transplants, which began almost at once and was complete in 6 days in one and 14 days in the other animal. In the third mouse the entire anterior chamber was filled with tumor on the ninth day. The tumor had regressed somewhat by the 14th day (Fig. 8) and further still in the next 7 days, at the end of which the eye was removed for microscopic study. The fourth mouse revealed active growth of the transplant during the first 10 days; then regression began and was completed in 21 days. Microscopic study of the tumor in the third animal showed neoplastic cells in a wavy sheet lining the corneal and iris surfaces of the anterior chamber. Regressive changes, however, had proceeded to such a point that the characteristic histologic structure of a glioblastoma was not present in this transplant. COMMENT The expectations that prompted these experiments, namely, the opportunity to observe the growth characteristics of experimentally induced brain tumors by transplantation into the eyes of other animals and to establish that these tumors are capable of autonomous growth, have been fulfilled. An important difference was demonstrated in the appearance of the gliogenous tumors as compared with the sarcomas. Whereas the


former class of tumors grew in large solid sheets of milky white color, the latter grew in clouds or minute flakes with much vascularity and hemorrhage. As regards these features, the more highly malignant gliomas occupied an intermediate position between the more benign gliomas, like the astrocytomas, and the sarcomas. At first they appeared as solid white masses which, when they increased in size, developed a moderately extensive vascularity and even a few small hemorrhages. With continued intraocular growth the gliomas eventually ruptured the cornea, whereas the sarcomas not only did this but destroyed the whole eyeball and continued to grow by replacing the orbital tissues and invading the adjacent cranial sinuses. Another striking difference between these 2 classes of neoplasms was the dispersion of much uveal melanin pigment in the sarcomas. These experiments have established the fact that intraocular transplants of experimental brain tumors will grow about equally well in homologous and heterologous strains of mice. The failure of certain transplants to take must be ascribed in large measure to faulty technic. Of course, this explanation does not hold for those transplants that regressed, after an initial period of active growth, but such occurrences were not limited to the heterologous strains. Regression was noted with almost equal frequency in the C3H and ABC albino mice. These experiments were not devised to contribute much of value to the problem of tumor regression. Several observations of minor significance require comment. The occurrence of postoperative hemorrhage into the anterior chamber of the eye did not affect adversely the success or rapidity of growth of the transplants, nor did traumatic cataract seem to influence their growth. This can also be said of postoperative adhesions (anterior and posterior synechiae). Tumor cells frequently grew successfully on the inner surface of the cornea and on the lens capsule, but invasion of the lens itself was never observed. The meshwork of the iris angle, Schlemm's canal, and the retina were often invaded by tumor. SUMMARY A technic is described for mice and guinea pigs that permits the intraocular transplantation of brain tumors, those induced with a chemical carcinogen in mice as well as those occurring spontaneously in man. This method of study affords the opportunity of keeping the growing neoplasms under constant observation. It has demonstrated certain characteristics of neoplastic growth behavior and appearance that permit differentiation of gliomas from nongliogenous tumors. It has demonstrated, by the standard of autonomous growth

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.


Cancer Researck

in homologous and heterologous strains of mice, that experimentally induced brain tumors represent true neoplasms.

4. 5.

REFERENCES 1. ABELS, J. C. Personal communication of unpublished data. 2. ARNOLD, H., and ZIMMERMAN, H. M. Experimental Brain Tumors. III. Tumors Produced with Dibenzanthracene. Caneer Research, 3:682-685. 1943. 3. GREEN~, H. S. N. Familial Mammary Tumors in the Rabbit. IV. The Evolution of Autonomy in the Course of Tumor



I)cvelopment as Indicated by Transplantation Experiments. J- Exper. Med., 71:305-324. 1940. GREENE, H. S. N. Heterologous Transplantation of a Human Fibrosarcoma. Cancer Research, 2:649-654. 1942. ZIMME~Ma~, H. M., and ARNOLD, H. Experimental Brain Tumors. I. Tumors Produced with Methylcholanthrene. Cancer Research, 1:919-938. 1941. ZlMMERMAN, H. M., and ARNOLD, H. Experimental Brain Tumors. II. Tumors Produced with Benzpyrene. Am. I. Path., 19:939-955. 1943. ZIIVtMERMAN, H. M., and ARNOLD, H. Experimental Brain Tumors. IV. The Incidence in Different Strains of Mice. Cancer Research, 4:98-101. 1944.

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.

Experimental Brain Tumors. V. Behavior in Intraocular Transplants David Freeman and H. M. Zimmerman Cancer Res 1944;4:273-278.

Updated version

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/4/5/273.citation

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected] To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/4/5/273.citation. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on March 9, 2018. © 1944 American Association for Cancer Research.