Specific Cellular Immune Responsesin Patients ...

0 downloads 0 Views 962KB Size Report
Khalid M. A. Shelkh,2 Michael L. J. Apuzzo, and Martin H. Weiss. Departmentof Surgery,Divisionof Neurosurgery,Universityof SouthernCaliforniaSchoolof ...

[CANCER RESEARCH 39, 1733-i 738, May 1979) 0008-5472/79/0039-0000$02.0O

Specific Cellular Immune Responsesin Patients with MalignantGliomas1 Khalid M. A. Shelkh,2 Michael L. J. Apuzzo, and Martin H. Weiss Departmentof Surgery,Divisionof Neurosurgery,Universityof SouthernCaliforniaSchoolof Medicine,LosAngeles,California90033


The leukocyte adherence inhibition assay was used to meas ure cell-mediated immunity in 26 patients with malignant glial neoplasms and 41 control subjects. A significant inhibition of leukocyte adherence was observed in 21 of 26 (80%) glioma patients in the presence of a 3 M KCI extract of glioma tissue, as compared to that of normal brain extract. Among the control group, no significant difference in the percentage of nonad herent leukocytes was noted in the presence of either antigen. To study the specificity of the reaction, a 3 M KCI extract of meningioma, pituitary tumor, carcinomas of breast, and lung, melanoma, brain, and heart tissues were used as nonspecific antigens. Such studies revealed significantly lower values of

nonadherent leukocytes. These data indicate that patients with malignant glial neo plasms manifest a cellular immune response to glioma-associ ated antigens which can be measured by the tube leukocyte adherence inhibition assay and that leukocyte adherence in hibition assay may render additional useful information in di agnostic and prognostic evaluation of malignant glial neo plasms. INTRODUCTION

Recently, several studies have demonstrated the presence of immunological response to glioma-associated antigens in patients with malignant glial neoplasms (1—3,5, 12, 13, 16, 21 , 22). Cellular



are generally



be more important in the rejection of solid tumors than are humoral antibodies. Several methods have been utilized to monitor or detect CMI3 responses to cancer antigens, i.e., lymphocyte stimulation tests (19), leukocyte migration inhibi tion assays (4, 11), delayed cutaneous hypersensitivity re sponses to membranes and solubilized extracts of autologous and allogeneic tumor cells (10, 19), and lymphocyte microcy totoxicity assays using tissue culture target cells (9, 15, 23). Lymphocyte cytotoxicity against tumor cells is the most notable and extensively used assay for monitoring the course of dis ease in cancer patientsor for immunodiagnosis.The specificity of the observed reactions and the relative lack of normal reactivity in this assay have been increasingly questioned. In addition, this assay inherits certain technical problems such as types of target cells used, optimal maintenance of cultures, and contamination of tumor cell cultures by normal fibroblasts. As in most of the biological systems, there is always a need

for refinement of the most commonly used assays, and there is even a place for new methodologies. Moreover, because of the heterogeneity of human tumor tissues (in a given kind of tumor), an assay to monitor CMI responses in cancer patients should be relatively simple and should consist of reactions specifically with native, relatively unchanged autologous or allogeneic tu mor antigens. Recently, Halliday et a!. (7, 8) described a simple and rapid assayfor CMI, called LAIassay. This assay has been used in a variety of experimental and clinical tumor systems to study CMI to tumor antigens (6, 7). We report here that a large number of patients with malignant glial tumors demonstrated CMI responses to autologous as well as allogeneic tumor antigens as measured by LAI assays. We also include evidence that the antigens were absent from a considerable range of normal as well as other types of malig nant tissues. MATERIALS

Patients. All 26 patients providing material for these studies were recruited from a consecutive series of patients with malig nant glial tumors treated at the Los Angeles County-University of Southern California Medical Center. All patients had rela tively brief histories (less than 6 months) and had been on methylprednisolone therapy for a period of 2 to 5 days, with a mean of 3 days. To assure uniformity in diagnostic appraisal, all pathological materials were reviewed by the director of the Cajal Neuropathological Laboratories at the Los Angeles County-University of Southern California Medical Center. The degree of malignancy observed was considered to be at the level of either a malignant astrocytoma or glioblastoma multi forme in all cases. A group of 41 control subjects









adherence Inhibition; CNS, central nervous system; MEM, Eagles minimum

essentialmedium;PBL,peripheralbloodleukocytes;NAL.nonadherentleuko cytes; PBS, phosphate-buffered saline. Received September 5, 1978: accepted February 9, 1979.

MAY 1979

5 healthy



(3 M) (Table

1 ) of various tumor tissues,


ing 5 gliomas (from Patients F. M., I. J., A. M., H. C., and K.

2Towhomrequests forreprintsshouldbeaddressed. 3 The


uals (with no apparent disease), 4 patients with spontaneous subarachnoid hemorrhages secondary to berry aneurysms, 3 patients with cerebral gunshot wounds, 3 patients with simple skull fractures, 4 patients with meningiomas, 5 patients with pituitary tumors, 6 patients with ocular melanomas, 6 patients with carcinomas of the breast, and 5 orthopedic patients, exclusive of CNS trauma. Tissue Extract (Antigens). Tissue specimens were obtained at the time of surgery, when subtotal excision of the tumors was accomplished.After removalof tissuefor routinepatholog ical evaluation, the specimen was transferred in Hanks' bal anced salt solution. The tissue was either used immediately or stored at —70°. Normal brain and normal human heart tissues were obtained from autopsied material within 6 hr after death.

,This work issupported inpartbytheRobert E.andMay A.Wright FoundationKCI



L.), one meningioma

(F. T.), one pituitary

tumor (A. W.), one

carcinoma of the breast (S. J.), one carcinoma of the lung (J. L.), one melanoma

(J. B.), and normal human brain and heart

tissues were prepared by the modification of the methods of

I 733

K. M. A. Sheikh

et a!.

Table1 Summaryof3 M KG! extract of g!ioma or norma! braintissueFresh gliomaor normalbraintissueWash NH4CIProtein

in PBSBathe


in 0.87%


aliquot [email protected]— and store atTo



homogenizewith ultrafiltrationand 10-mmmembraneICentrifuge



at x g

Dialyze against PBS

for 30 mm*-.Centrifuge

Meltzer et a!. (20). Protein concentrations of the extracts were determined by the method of Lowry et a!. (17). Serum albumin was used as the standard. Leukocyte Preparation. Fifteen-mI samples of heparinized blood, obtained preoperatively from tumor patients and control subjects by venipuncture, were allowed to stand at 37°for 30 to 60 mm. Leukocyte-rich plasma was aspirated and centri fuged at 200 x g for 7 mm. Cells were resuspended in 0.05 M NaCI to lyse erythrocytes,

and 9 ml of MEM



After centrifugation at 200 x g for 5 mm, the leukocytes were resuspended in MEM, washed twice, and finally adjusted to a concentration of 1 x 107/ml.

at 40,000 x g for 60 mmStir

with gliomas admitted to the Neurological Surgery Unit of the Los Angeles County-University of Southern California Medical Center was measured in glass tubes in the presence of 3 M KCI extracts of tumor tissues obtained from the same patients (autologous reaction), as well as in the presence of normal brain extracts. Each sample was run in 6 replicates. The mean percentage of NAL in a given sample was calculated as a difference of mean percentage of NAL in the presence of antigen and PBS alone. The results of such a study are shown in Table 2. The percentage of NAL was consistently signifi cantly high in the presence of tumor extracts as opposed to normal brain extracts. No significant difference in the values of

LAI.Antigen-induced LAIto glassin patientswithgliomas NAL

was studied by the tube method (6). One-tenth ml (1 x 10°)of PBL was placed in glass tubes. Either 0.1 ml of gliomaextract (specific antigens), normal brain or heart tissue, or other tumor extracts (nonspecific antigens) containing 200 mg of protein was added to each tube. To measure the spontaneous adher ence inhibition of leukocytes to glass, 0.1 ml PBS (0.01 mol, pH 7.2) was added to a separate tube with every test. The final volume of incubation mixture in all the tubes was adjusted to 0.5 ml with the addition of 0.1 ml of normal human pooled blood and 0.2 ml of MEM and then mixed gently by aspiration. The tubes were incubated at 37°horizontally so as to allow the contents of the tubes the maximum exposure to the glass surface. After 2 hr of incubation, the contents of the tubes were shaken gently with a Pasteur pipet, and a count of NAL was made in the hemocytometer. All assays were done in sextupli cate. Percentage of nonadherent cells was obtained by cam paring with those tubes containing buffer alone. Mean per centage of nonadherent cells and S.E.'s were calculated for each sample, and statistical significance of comparison be tween tumor extracts and normal brain extracts was determined by Student t test. p ..

Normal brain antigen

10Pituitary B. C.

Ui 0



genMeningiomas1. Patient

7 ±2a 9±4



of the breast in the presence of

normal brain, extractsHeart and glioma

72.T.M. B. T. 73.W.A.



ocularme!anomas, LA! of PBL of patients with meningiomas, pituitary tumors,





observed difference

reaction frequency of leukocytes to glioma antigens within the

that of glioma extracts. However, the values for NAL were comparatively higher in the presence of meningioma and pitui tary tumors (CNS tumors) as compared to those of non-CNS


(p < 0.05) LAI in the

presence of glioma antigens as opposed to normal brain anti gens (Tables 2 and 3).




10 -







C .C C C


eeC! C .C CC C C








Chart 2. LAI of PBL of control patients in presence of 3 M KC1extract of normal heart, normal brain, and glioma tissue.



5± 3 15±


9± 19±

± S.D.

Table 6

of PBL of g!ioma patients with the presence of 3 M KCI extracts of meningiomas, LA! to ituitary tumors, carcinomas of the breast, melanomasA. carcinoma of the lung, and pg!ass anoma)K.M. Normal brain

25± 34±

6 6

F. 0. (meningioma) 13± 19±

4 6

W. (pituitary tumor)

S. J. (carcinoma of the breast)

J. L. (car cinoma of the lung)

J. B. (mel

8±3 6±3

3±1 11±4

14±4 17±5

12±4 4±2

F.M. C.P.
























11±3a 23± l.J,10±


7° 7

A. M. (glioma)




VOL. 39

Cellu!ar !mmune Response in Ma!ignant G!iomas glioma patient population and of that between control donors

(Table 4). Moreover, a comparison of LAI reactivity of brain

and glioma patients suggests that we are detecting a tumor related phenomenon.

trauma patients to normal brain and glioma antigens revealed

A large number

of studies have indicated

that the LAI assay

an overlap in the degree of responsiveness. It is hard to explain this cross-reactivity in LAI results between malignant gliomas

was useful in detecting cellular sensitization to a specific anti gen in various experimental (6), as well as clinical tumors (6, 7,

and brain trauma patients. However, there are some sugges tions that gliomas may share certain antigenic components

1 8). The present

with normal brain (25). Recently, Thomas (24) used radial immunodiffusion assays and demonstrated circulating brain


not only


this contention


also reveals that the LAI assay can discriminate between glioma patients and those of control subjects, based on the reactions

of glioma-associated antigens. Our studies indicate comparatively high values of NAL in the presence of both glioma and normal brain antigens as com pared to heart antigens

(Tables 3 and 4), which may be

attributed to organ-specific rather than tumor-associated anti gens. The facts (a) that a large number of glioma patients showed significant increase (p < 0.05) in NAL over normal brain antigens and (b) that leukocytes from patients with other CNS cancers, exclusive of gliomas (i.e. , meningioma and pi tuitary tumors), did not show a difference strengthen the con tention that LAI reactivity depicted in glioma patients may be glioma antigen associated. While it is not absolutely certain that the LAI assay measures exclusively



there is considerable


dence that this is the case. According to Grosser et a!. (6), the reactive cell was a circulating

monocyte that recognized


reacted with tumor antigens by cytophilic antitumor antibody bound to the surface of the monocyte. In our study, we did not find any significant difference in LAI results in autologous and allogeneic combinations of PBL and tumor extracts (Table 2). This similarity of results strongly suggests that transplantation antigens may not be the cause of the reported reactions. These findings further confirm our previously reported data that gliomas may share antigenic characteristics (22). Using indirect immunofluorescent anti

body techniques, we demonstrated that, of 21 sera obtained from patients with histologically

proven malignant gliomas, 47%

possessed antibodies reactive with cytoplasmic components of allogeneic tumor cells. In vitro, antigenic cross-reactivity be tween tumors of the same histological types or origin has been reported by other researchers in a variety of clinical tumors, utilizing a variety of immunological


i.e. , lym

antigens in 37% of 76 patients with blunt head injury. In light of these studies, it is tempting to suggest that after severe

brain injury there is a breakdown of the blood-brain barrier. Therefore, some of the damaged brain cells may release brain components into the circulation; under normal conditions this would not occur. These antigenic components may in turn provoke host immune responses.

In conclusion, the results presented in this report establish the fact that the LAI assay depicts specific anti-tumor activity

to glioma antigens in patients with gliomas and that this assay is rapid, specific, and reproducible. The fact is that the LAI assay not only discriminates between gliomas, other cancers, and controls but also obviates the necessity of tissue culture cell lines. Because of the possible variation in the results,

however, the tests have to be done repeatedly. This informa tion, coupled with the findings that gliomas may possess cam mon antigens, neurosurgeons

disease. This provides them with an opportunity to better manage glioma patients.

REFERENCES 1. Brem, S. S.. de Tribolet, N., Dohan, F. C., Jr. . and Kornblith, P. I. Demon stration of cell mediated immunity to a human brain tumor: use of an in vitro microcytoxicity assay. Surg. Forum, 23: 428—430,1972. 2. Brooks, W. H., Caidweil, H. D., and Mortara, A. Immune responses in patients with gliomas. Surg. Neurol., 2: 419—423,1974. 3. Coakham, H. B., and Lakshmi, M. S. Tumor-associated surface antigen(s) In human astrocytoma. Oncology (Basel), 3 1: 233—243,1975. 4. Cochrane, A. J., SplIg, W. G., Mackis, A. M., and Thomas, C. E. Postoper ative depression of tumor-directed cell-mediated immunity In patients with malignant disease. Br. Med. J.. 4: 67-70, 1972. 5. Eggers, A. E. Autoradiographic and fluorescence antibody studies of the human host immune response to gliomas. Neurology, 22: 246—250,1972.

6. Glosser,N., Marti, J. H., Proctor,J. W., and Thompson,D. M. P. Tube

phocyte cytotoxicity assays (14, 16), serum cytotoxicity assays (1 2, 1 3), and immunofluorescent

test for antibodies

to tumor


associated antigens (22). Further proof of glioma antigen specificity was sought by reactions of leukocytes from glioma patients with a battery of


antigen extracts obtained from normal brain, gliomas, menin


giomas, pituitary tumors, carcinomas of the breast, carcinomas of the lung, and melanomas (Table 6). In this test, LAI respon siveness to glioma antigens was significantly higher than all other antigens tested. However, a comparatively higher value of LAI in the presence of meningioma and pituitary tumor extracts (CNS tumor) may be attributed to organ-specific com mon antigens. An analysis of LAI responsiveness to normal brain extracts, as well as glioma extracts among various control donors re vealed that patients with brain trauma or brain injury consist ently showed higher values of NAL than did other donors (Table 4; Chart 2). Among 4 of 1 0 patients

with brain trauma,


for NAL ranged between 23 and 36%. These values were significantly higher than those of other control populations MAY 1979

renders the LAI assay as a potential tool for to diagnose gliomas early in the course of the

10. 11. 12. 13. 14.

leukocyte adherence inhibition assay for the detection of anti-tumor immu nity. I. Monocyte is the reactive cell. Int. J. Cancer, 18: 39—47,1976. Halliday, W. J., Maluish, A. E., and Isbister, W. H. Detection of anti-tumor cell-mediated immunity and serum blocking factors in cancer patients by the leukocyte adherence inhibition test. Br. J. Cancer, 29: 31—35,1974. Halliday, W. J., Maluish, A. E., and Miller, S. Blocking and unblocking of cell mediated anti-tumor immunity in mice as detected by leukocyte adherence inhibition test. Cell. Immunol., 10: 467—475,1974. Hellström,I., Hellström,K. E., and Sjogren, H. 0. Demonstration of cell mediated immunity to human neoplasms of various histological types. mt. J. Cancer, 7: 1—16,1971. Herberman, R. B. In vivo and In vitro assay of cellular immunity to human tumor antigens. Fed. Proc., 32: 160-164, 1973. House, A. K., Wisniewski, S., and Wooding, T. L. Immunity in colonic tumor patients after operation: determination by leukocyte-migratlon inhibition. Dis. Colon Rectum, 18: 100- 106, 1975. Kornblith, P. L., Dohan, F. C., Wood, W. C., and Whitman, B. 0. Human astrocytoma: serum mediated immunologic response. Cancer, 33: 15121519, 1974. Kornblith, P. L., Pricto, A., and Pollen, D. A. Alterations in glial and mesen chymal tumor cell-membrane resistance with heteroimmune and autologous sera. Ann. N. V. Acad. Scm.,159: 585-590, 1969. Kumar, S., and Taylor, G. Specific lymphocytotoxicity and blocking factors

in tumorsofthe centralnervoussystem.Br.J. Cancer,28: 135—141 . 1973. 15. Levy, N. L., Mahaley, M. S., and Day, E. D. In vitro demonstration of cell mediated immunity to human brain tumors. Cancer Res., 32: 477—482, 1972. 16. Lim, R., and Sluskens, L. Immunological specificity of astrocytoma antigens.


K. M. A. Sheikh

et a!.

Cancer Res., 32: 1665—1670,1972. 21 . 17. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Protein measurements with the Folin phenol reagent. J. Biol. Chem., 193: 265—275, 22. 1951 . 18. Marti, J. H., and Thompson, D. M. P. Anti-tumor immunity in malignant melanoma assay by tube leukocyte adherence inhibition. Br. J. Cancer, 34: 23. 116—1 33, 1976. 19. Mavligit, G. M., Ambus, U., Gutterman, J. U., Hersh, E. M., and McBride, C. M. Antigensolubilizedfromhumansolidtumorslymphocytestimulationand 24.

cutaneousdelayedhypersensitivity. Nat.NewBiol.,243: 188—1 90, 1973. 20. Meltzer, M. S., Leonard, E. J., Rapp, H. J.. and Borson, T. Tumor-specific antigen solubilized by hypertonic potassium chloride. J. NatI. Cancer Inst., 4 7: 703—709,197i .

1 738

Quindlen, E. A., Dohan, F. C., and Kornblith, P. L. Improved assay for cytotoxic antiglioma antibody. Surg. Forum, 25: 464—466,1974. Sheikh, K. M. A., Apuzzo, M. L. J., Kochsiek, K. A., and Weiss, M. H. Malignant glial neoplasms: definition of a humoral host response to tumor associated antigen(s). Vale J. Biol. Med., 50: 397—403,1977. Takasugi, M., Mickey, M. R., and Terasaki, P. I. Studies on specificity of cell-mediated immunity in human tumors. J. NatI. Cancer Inst., 53: 1527— 1538, 1974. Thomas,D. G. T. Antigenrelease intothe peripheralbloodof patientswith

headinjury.IRCS-Med.Sci.Libr.Compend.,3: 26, 1975. 25. Wahlstrom, T., Linder, E.. and Saksela, E. Glial-specific antigens in cell cultures from rabbit brain, human foetal and adult brain, and gliomas. Acta Pathol. Microbiol. Scand. Suppl., 8 1: 768—774,1973.



VOL. 39

Suggest Documents