Cal~cer Letters,.2 (1977} !39-146

.139

© F.l~evler/North-HollandScientific.PublishingLtd;

IMMUNOLOGICAL T U M O U R PROFILE: ORGAN, SPECIFIC:CARCINOMA DIAGNOSIS IN PATLENTS EMPLOYING THE. MACROPI~AGE ELECTROPHORETIC MOBILITY TEST MARTINMULLER,JURGEN IRMSCHER.RAINERFISCHER,GERHARDHE][DL and HEINZGROSSMANN lnstitule o f Patl~o[ogy, Medieal Acaderny 'Car! Gustav Carus' Dresden, Schubertstrasse 15~ 8053 Dres.:le.'z (G.D.R.)

{Received 1! May 1976) (Revised versionreceived 29 July 1976) SUMMARY The macrophage etectrophoretic mobilit~r (MEM) tes~ provides a highly sensitive in vitro technique for the detection of cell-mediated immunity in. man~. The principle involved is the lymphokine-mediated reduction ot the nega~ve surface charge of guinea pig macropt~ages shown by the slowing of the macro phages during cell electrophoresis: Lymphocytes from 162 patients were:tested by MEM, They wereexposed to a battery Of KC1 extracts from normal and malignant human tissues, ~o encephafitogenic protein (EP), to carcinoembryon.ic antigen (CEA), and to thyroglobulin." Variable lymphocyte responsesto EP, CEA and KC1 extracts from differen% Cancers :gave MEM reaction profiles common to patien.ts with carcinomas of the s ~ e Organ Origin, INTRODUCTION The ~IEM test arose from the findin- '",y Field and Caspary [1,5] that lymphecytes from cancer patients, as ~,~ll as frompatients w..th central nervou,,~•system diseases, are sensitized to EP from human brain. The exp]:anation given was the occurrence of cross-reacting basic proteins at the cancer cell surface [2]. This would mean that common basic ~umour-associated protein cross-reacting to EP exist. Few reports confirming these findings ha',,e appeared [8,13,20], and, with some scepticism, we repeated these experfinents

[13]. All tests were performed four times on 76 patients with histologically verified carcinomas and 42 individuals withou t de'~ectable malignancies. The slo~r.'.ngof macrophages was measured simultaneousl:~' in two cytopherometers by two persons. The findings' of ]?~:eldand Caspm7 were; in general; confirmed [13]. In t;he presence of EP, lymphocytes from 75% of the carcinoma patients rhowed

140

a macrophL~ge slowing greater than 10%, where~ lymphocytes from only 4.8% of the controls did so. If the arbitrary 10% cut-off macr.ophage slowing is not used, and all significant slowings are taken to be positive results, then 88.2% of the carchaoma patients and :[9% of the individuals without cancer gave positive MEM results in the presence o f EP [18]. This overlap between cancers and z:on-cancer.s, and the difficulty in drawing conclusions about the o~gan in which the carcinomas develop, ted to our attempt to improve the NIEM test focusing on its diagnostic application. The working hypothesis was that the turnout-host relationship in an individual may result in sensitization to turnout-associated antigens. If exposed in vitro to a battery of tumour-relevant antigens, the lymphocytes from a patient might recognize those antigens contacted during their circulation within the turnoutbearing body. With this concept, unexpected regulus with respect to specificity, selectivity, and diagnostic value wer~ obtained. MATERIALS AND METHODS

Antigens EP was prepm.~ed as in Caspary and Field [3] from human brain. Thyrogiobulin was a kind gift of Professor Field~ Newcastle, UK. The preparation of CEA from liver r,aetastases of an adenocarcinorna of the colon has been described by Grossmann, Heidl and Milller [9 ]. The perch]oric acid extract wm~ further purified by multiple steps including chromatography on Sephadex G-200, Sepharose 4-]3, and Sepharose 6-B. The final product had the specificity of Dr Gold's CEA. (~?he authors thank Dr Gold, Montreal, for testing substances and providing an ~nti-CEA serum.) KCI extracts from s~gically removed cancers and normal tissues were prepared according to Meltzer et al. [18] with slight modifications [19]. All antigenic preparations were used in amounts of 100 ~g per incubation sample (3 ml).

Lymphocy tes Human lymphocytes were prepared according to Hughes and Caspary [12] from defibrinated blood. Patients scheduled for biopsy or surgery were selected to give blood before diagnosis or therapy. A histologically verified diagnosis was thus assured.

Macrophages Guinea pigs (inbm¢~ s'~rain Albino/Jena) were injected intr.aperitoneally with 20 ml sterile liquid paraffin. The peritoneal cells were harvested with Hanks' solution without heparin after. 10--14 days. The cells from several animals were collected sepaxately, washed in Hanks', and irradiated with 200 r.ad (6o Co sotr~ce). After the admixed lymphocy'~tes had been thus inactivated, the peritoneal cells from several animals could b.e pooled without the disadvantage of a possibte mixed cell reaction. The pooled macrophage suspension:s were washed a~ain in Hanks' and made up in Eag~'e's Minimal Essential Medium.

141 Incubation The incubation samples contained the followhug cells in 3 ml Eagle's m e d i u m (pH 7.3): ( ! ) 1 0 ~ macrophages + I06 l y m p h o c y t e s + 100:ug antigen (test incubation); (2) 107 macr0phages • 106 lymphocytes; (3) 10 ~ mac,:ophages + 100 gg antigen; and (4)110 T macrophages. (Incubations 2--4 were controls.) Using an average of 10 antigens per lymphocyte sample, tests were performed with the lymphoeytes from one patient. Mkxtures were incubated for at hmst 90 rain at 23°C in a water bath. Stabilization o f pH b y addition of 2--4 drops of 2%. NaI-ICO3 in samples where there was more than a 3-h wait before measuring was advantageous. Measurement The electrophoretic mobility was measured by two persons s i m u l a t e o ~ l y in two cytopherometers (OPTON, Oberkochen, GFR:) at pH 7.1--7.3 and 23:C. All readings were taken in the forward stationary plane. The current applied was 13.5 mA giving a voltage of about 200. Medium sized macrophages containing paraffin droplets were timed across one grid square of the eyepiece. Measurements were made of 25--40 ceils per sample in both directions e f the applied current with a maximum tolerated time differer~ce o f 20%. A test was considered to wox:k well if the controls gave results with n o t more than ± 3% difference. Th,e time~ measured for the single cells of one sample were averaged to give the results in seconds. The percentage of slowing was calculated as follows:

(Mean time of test mixture) -- (mean time of control) (mean time cif control)

x I00

The results were computed (Hewlett--Packard) with the t-test program and slowing was considered to be significant i f p < 0.5. Most of the significmat results had a p < 0.001. '[he slowing was estimated as follows: < 5% = nothiug, 5--20% = weak results, > 10% = strong positive results. RES ULTS The results from six representative patient groups (146 individu~'ds) are pre,~ented in Fig. 1. Because of the bruited space, only the most important of the antigenic preparations employed are listed. Lymphocytes from all individua~ reacted strongly to the thyroglobulin (15.0--18.9% slowing). Whereas more than 90% of the c'-mcer patients gave a positive reaction i;o EP, the response to EP was negligible in the group without malignancies (6 patients with multiple sclerosis were not included, since there were no controls with respect to their expected response to EP [1] ). In general, no group of patients showed significant reaction to KC1 extracts of normal tissue. In contrast, the KCI extracts from carcinomas of a single organ induced

142

MEM RESULTS OF 146 PATIENTS (~±s~)

Patients

Without

/\

BreaSt

Ca Kidney

Ca Stomach

Ca Colon Re~tum Ca Lung

Ca Antigens :

EP

IColonBreas~ KidneyI~[idr~eyLung Lung I S,~orn. Store, ICokm. Reell. Rect. Ca Ca [issue Ca rissuejCEA Cv Mucos Ca MucosThyr'

Fig. 1. Since not all antiger~ could be tested with one lymphocyte sample, thv number of patients (n) w.ries. Ca = carcinoma; Mueoa = mucosa; EP = encephalitogenic protein; CEA = careinc,embryonic an~,igen;Thyr. = thyroglobulin, * 1 patient who did not react was included in the mean value. ** 15 patient.,;who did not respond to CEA were included in the mean value. strong and specific lymphocyt,~ responses in those patients having a primary carcinoma o f the ~ame organ, regardless o f w h e t h e r metastases had occmTed. 1%eactions in t h e d i f f e r e n t carcinoma groups were as follows: I 8 / 1 9 w o m e n w i t h b~east cancer r e s p o n d e d t o extracts f r o m carcinomas o f the m a m m a r y gIand; 9/9 patients with h y p e m e p h r o i d carcinoma o f the kidney reacted with e x t r a c t s o f this c~mcer type; 30/30 s~omach cancer patients gave selective reactions with extracts from s t o m a c h carcinomas, and 13/28 patients o f this g r o u p s h o w e d strong positive reactions t o CEA. In the colonrectum cancer g r o u p 21/21 individu~,]s responded t o extracts from colon and rectum carcinomas, and 22/22 gave positive results w h e n tested with CEA; 12/13 lung cancer patients responded t o extracts from bronchogenic carch~om~. Single patients w~th ~ t h e r malignant tumours, who were tested along a large antigen s p e c ~ m bu~; n o t listed in Fig. 1, also gave selective reactions o f diagnostic value. Two w o m e n wi~h uterine cervix cancer had a react:ion to an e x t r e e t o f this carchmma Wpe and less o f a reaction 1~o a preparation from e n d o m e t r i u m cancer; 1 case o f malign~mt m e l a n o m a s h o w e d a selective response to a m e l a v o m a extract; 1 patient w i t h a retroperitoneat leiomyosarcoma reacted with an e x t r a c t

f~om a reticulum ceU sarcoma, b u t showed :no response to the carcinoma extracts; I m ~ i ~ a n t :teratoma of the testis gave positive reactions with most of the cancer extra~ts used: Two patients positive to EP and negative to all tumourrelevant antigens employed were proven to have a carcinoma,(gall-bladder, lip) not represented by our extract. Nine lcatients from our test schedule are still under a diagnostic regimen. Their diseases remain to be verified or confu'med. As shown in Fig. i some weak cro:ss-reactions may occur, but will not interfere with the possible diagnostic statements. The g~oup of patients without malignancies (Fig. 1) consists of pregnant and puerperal women~ patients with benign disorders of the uterus or mmmmary gland, patients with gastritis or gasTxointestinal ulcers, appendicitis, diabetic angiopathy, endocarditis, d~matomyositis, sarcoidosis, tuberculosis, multiple sclerosis, emphysema and bronchitis, urticaria pil~mentosa, and two healthy blood donors. These 'controls' gave no reaction profile characteristic of a carcinonq:a. DISCUSSION Though not definitely proven, the MEM results are thought to be mediated[ imm~uaologically [1,4--6]. Employing the syngeneic and allogeneic routine marnmary turnout host systems, we demonstrated recently that the MEM test reveals l~aph node cell sensitizations compatible with the expected specificity of turnout and transplantation immunology [14]. There;!ore we feel this is a highly sensitive technique to detect immunologically specific lymphocyte sensitizations. The findings indicate that hypertonic KCI extracts of human cancer tissue contain turnout-associated antigens shared by carcinomas of the same organ origin, This assumption is underlined by the fact that extracts made from corresponding normal tissue from the same patient were inactive. In addition to these turnout.associated antigens cha~'acteristic of carcinomas of one organ type, weak cross-reactions between different organ cancers indicate that minor common antigenic componer~ts of a broader spectrum of carcinomas may exist. Another interesting finding was the selective reaction to CEA in all of the colon-rectum cancer patients and in 13/28 stomach carcinoma cases tested. To our knowledge, no definite lymphocyte response to CEA l~as yet been described [15,21]. The earlier claim of Gold [7] that there were antibodies to CEA in colonic cancer patients has not been sustained [16,17] Therefore, we feel our findings contlict with the gen-ral view that man does not respond immunoiogically to CEA. The CEA used ~oz our MEM tests and also as a standard for the direct detection of CEA in human sera by radioimmune precipitation [10] has the antigenicity of Dr Gold:s CEA and contains also the normal cress-reactive antigen [22 ]. We did not try to ~eparate the skin reactive ani'igen :.~romour CEA, as described by Hollinshead et el. [11]. Studies to identify ~he corn ponent of the CEA comple.x responsible for the selective MEM findings

144

are in progress. Whereas the direct C E A detection in h u m a n sera was of little diagnostic va~ueinour handJ~ [I0], the lympihocyte response to C E A showed high sensitivit~ran d selectivily in detecting gastrointestinal carcinomas. There:lore itcan be assumed that lymphocytes from the respective cancer patients can recognize a tumouz-associated determinant at the C E A complex, Xenogenic h y p e r i m m u n e sara to the C E A complex, however, detect a broader spectrum of determfi~ants which are by no means all cancer specific. The response to thyroglobulh~ is a marker for the preserved immu.uocompetence of the purified lymphocytes [5,10,13]. T h o u g h s o m e basic questions remain to be answered, the M E M test, w h e n pefforr-.~edsimultaneously with EP, C E A and a large spectrum of KCI ~tracts from cancers, gives characteristic reaction profiles of high value for carcinoma diagnosis. ,Of 162 patients testa4, 52 cases without malignancies and 99 patients with cancers were in accordance with the diagnostic principle of the 'immunological turnout profile'.In two cases cancers were not detected and 9 cases remain to be verified. Assuming the pessimistic ~,ariantthat these 11 cases give false results, our test would have a failure r a t e o f a b o u t 7%. ACKNOWLEDGI~ M E N T S The a u t h o r s t h a n k Mrs E r i k a Miiller, Mrs Sigrid T h i e m e , Mrs J u t t a Sander, a n d Mrs Sigrid B S h m e for t h e i r ~technical assistance.

REFERENCES 1 Caspary, E.A, ~nd Field, E.J. (I970) Sensitizatio~ of blood lymphocytes go, possible antigan~ in neurological disease~ Ear. Net~ol., 4, ~;57-.266. 2 Caspary, E.A. ~n~ Field.E.J. (1971) Specificlymphocyte sensitizationin cancer: :.~ there a common antigen in human malignant neolt.lasia?Br. Mad. J.,2. 613--617: 3 Caspary, E,A. and Field,E.J. (1965) An cneephali'ogenicprotein of human origin;some chemical and biologicalproperties.Ant/. N.Y. Aend. Sci.,122, 182--198. 4 Caspary, E.A. (1973) A lymph0eyte.sen~itizingturnout-associatedbasicpolypeptlde,its demonstration and properties.In: Immunological Techniques for Detection of Cancer, pp. 71--78. Editor: B.Bj~rklund. Bonniers,Stockholm, Sweden. 5 Field,E,J.and Caspary, E.A, (1970) Lymphocyte sensitization:an in vitrotestfor cancer? Lancet 2, 1337--1341. 6 Field,E,J.and C~spary, E.A. (1970) Circulatingsensitisedlymphoeytes in Grave's disease. Observation~ on its pathogenesis. Lancet 1, 1144--1147. 7 Gold, P. (1967) Circulating antibodie~ against carcinoembryonic antigens of the human digestive system. Cancer, 20, 1663~1667. 8 Goldstene, A.H., Kerr, L. and Iz~ine,W.J. (1973) The macrophage eleetropboretiemigration testin cancer. Clin, Exp. Immunol., 14, 469--472. 9 Grossmaun, H,~ Heid~, G. and Ml~ller, IH. (1975) Zur Isoli~:rungdes cascinoembryonalen Antigens. Acta Biol. Med. Germ., 34, 1347--I357. 10 Heidl, G., Grossmann, H., M~ller, M. and Johannsen, B.A. (i975) Der Nachweis des karzinoembryenalen Antigens (CEA) im Serum yen Patienten durch den Radioimmunpr'azipitationstest. Dr. Gesundh.-Wesen, 30, 1270~1275. 11 Hollinshead, A,G., McWright: C.G., Alford, T.C., CHew, D.H., Gold, P. and Herberman, R.B. (~.972) Separatlon of skin reactive intestinal cancer antigen from the c~wcinoembryonic antigen of Gold. Science, 177, 887~889.

145 12 Hughes, D. and Caspary, E.A, (1970) Lymphocyte trans£ormatlon ia vitro measured by trltiated thymldine uptake, Int. Arch. Allergy Appl. Immnnol., 37, 506--531. 13 Irmseher, J., Milller, M.; Fischer, R., Otto, .G. and Strietzel, M. (1975) MakrophagenElektrophorese-Mob!lit~ts-Test (MEM) zur immunoiogischen Diagnose maligner Gesehw~lste. Dt. Gesundh.-Wesen, 30, 687--692. 14 Kotzseh, M., Irmscher, J., Fischer, R., Heidl, G. and M~ller, M. Untersuchmzgen zur immunolog~sehen Spezlfi~t des Makrophagen-Elektrophorese-Mobili~tstests (MEMTests) bei M.~usen mit syngenen und allogenen Mammakarzinom-Transptantatem Aeta BioL Med. German. (in press). 15 Lejtenvi, M.C., Freedman, S.O. and Gold, P. (1971) Response of ly.nphocytes from patients with gastrointestinal cancer to the carcinoembryonic antigen of the human digestive system. Cancer, 28, 115--120. ).6 LoGerfo, P., Herter, F.P. and Bennett, S,J. (1972) Absence of circulating antibodies to carclnoembryonie antigen in patients with gastrointestinal malignancies. Int. J. Cancer, 9, 344--348. 17 MacSwee~, J.M. (1975) The antigenieity of carcinoembryonie antigen in ma~ In~. J. Cancer, 15, 246--252. 18 Meltzer, M.S., Leonhard, E.J., Rapp, H.J. and Brosos, Z?. (1971)Tumor-specific antigens solubilized by hypertonic potassium chloride, J. Nath Cancer Inst. 47, 703-709. 19 M~ller, M., Irmscher, J., Fischer, R. and Grossmann~ H. (1975) Immunologlsches Tumorprofil. Ein neuartiges Prinzip in der Anwendung des Makropl~agen-Elektrophorese-lVJobilitlitstest (MEM) zur differenzierten Karzinomdiagnose. Dt. Gesundh.Wesen, 30, 1836--1842. 20 Pritchard, J.A.V., Moore, J.L., Sutherland~ W.H., Joslin, C.A.F. (1973) Evaluation and development of the macrophage eleetrophnretic mobility (IVIEM)test for malignant diseases. Brit. J. Cancer, 27, 1~9. 21 Straus, E., Vernace, S., Janowitz, H. and Pa~onetto, F. (1975) Migration of peripheral leukocytes in the presence of carcinoembryonie antigen. Studies in patients with chrome inflammatory diseases of the intestine and carcinoma of the colon and pancrena Prec. Soc. Exp. BioL Med., 148, 494--497. 22 Von Kleist, S., Chavanei, G. add Burtin, P. (1972) Identification of an antigen from normal human tissue that crossreaets with the eareinoembryorde antigen. Prec. NatL Acad. Sci. USA, 69, 2492--2494.