Tumor-infiltrating T lymphocytes (TIL-T) are always present in B-cell-derived non-Hodgkin's lymphoma (NHL). In this investigation, we explored the possibility ...
T Lymphocytes From Invaded Lymph Nodes in Patients With B-Cell-Derived Non-Hodgkin’s Lymphoma: Reactivity Toward the Malignant Clone By Marie-Christine Jacob, Marie-Pierre Piccinni, Thierry Bonnefoix, Marie-France Sotto, Pierre Couderc, Jean-Claude Bensa, and Jean-Jacques Sotto Tumor-infiltratingT lymphocytes(TIL-T) are always present in B-cell-derived non-Hodgkin‘s lymphoma (NHL). In this investigation, we explored the possibilitythat collaboration might exist between these cells. TIL-T were isolated from 39 lymph nodes of patients with NHL. In most of the cases, few of them (less than 10%) possessed surface activation receptors CD25 or OKT9. In 80% of the cases, they proliferated in response to recombinant interleukin-2 (rlL2). but the degree of proliferation was often low as compared with control populations.The influence of irradiated autologous malignant cells on the TIL-T proliferation in response to rlL-2 (40U/mL) was also investigated: in 38% of the cases, this proliferation was not modified (group 0). and in 41 % it was higher (group +) and in 21 % it was lower (group -1. The mechanism of this immune
response (specific or not) is not elucidated at present. The definition of these groups was statistically correlated with different parameters of the disease: (1) percentage of TIL-T was higher in group (44% A 17%) than in group 0 (31% A 18%) and group - (24% ? 15%); (2) B-cell proliferation in centrofollicular lymphomas was more frequently nodular or nodular and diffuse in group (83%) and 0 (55%) than in group - (0%);(3) low-grade malignancies in the Working Formulation were more frequent in group (75%) than in group 0 (60%) or group - (12%); (4) favorable prognosis evaluated with the Grenoble cytologic classification was more frequent in group and 0 (87%) than in group - (12%); ( 5 )actuarial survival curves showed a significantly better prognosis for patients in group 0 1990 by The American Society of Hematology.
N
with monoclonal antibodies (MoAbs) of activated TIL-T on histologic slides is often the only element in favor of their reactive ~ h a r a c t e r . ~ . ’Nevertheless, ~.’~ many studies support their implication in the tumoral process: T lymphocytes isolated from lymph nodes with B-cell type N H L are capable of inhibiting immunoglobulin (Ig) secretion by the malignant clone,” secreting lymphokines to which B cells are potentially sensitive (references 18 through 20 and personal observations, T. Bonnefoix, 1988), and developing cytotoxic activities toward malignant cells.2’ They also appear to be responsible for granuloma formation in Hodgkin’s disease (personal observations, M.P. Piccinni, 1989). Our objective is to explore the possible existence of an immunologic response within lymph nodes invaded by B-cell-derived NHL. We first looked for evidence of activation features in isolated TIL-T by studying the presence of interleukin-2 (IL-2)22and t r a n ~ f e r r i nmembrane ~~ receptors as well as for the existence of a response to recombinant IL-2 (rIL-2). In the absence of any other stimulation and a t low concentrations, rIL-2 induces the proliferation of activated T lymphocytes only.24Secondly, we explored direct functional interactions between the tumor and its microenvironment by analyzing the influence of the B-cell clone on the response to rIL-2 of T lymphocytes isolated from the same lymph node.
ON-HODGKIN’S LYMPHOMAS (NHL) constitute a heterogenous group of lymphoid tumors, the majority of which are of the B-cell type. Lymph nodes are invaded by the malignant monoclonal B cells arrested at different stages of differentiation’ together with other cells, which are probably normal, that make up the tumor microenvironment. Among them are T lymphocytes, which are present in much variable proportions from one patient to another. Their role and significance in the pathogenesis of the disease remain unknown. It is obvious that there are often too many of them to be considered as simple residual elements from the normal lymph node structure. They possibly may represent a nonspecific accumulation of blood-borne lymphocytes sequestered in the tumor secondarily to an altered pattern of leukocyte extravasation2 One might also postulate that they proliferate in situ in contact with the malignant clone. It is now well-established that T lymphocytes are capable of proliferating under the influence of autologous stimuli either in contact with other normal lymphoid cell^^.^ (ie, the autologous mixed lymphocyte reaction [AMLR]), or in contact with tumor cells2“ (ie, the mixed lymphocyte tumor culture [MLTC]). Many investigators referred to positive MLTC between peripheral blood T lymphocytes and malignant cells,’-’’ but the demonstration of stimulation within the tumor is rare because of the difficulty in isolating tumor-infiltrating T lymphocytes (TIL-T) in solid tumors.2.’2The identification
+
+
+
+
+.
MATERIALS AND METHODS
Lymphoid Tissue From the Laboratoire de recherche d’immunopathologie tumorale, Service d’h5matologie. h6pital A. Michallon, Grenoble, France. Submitted August 31. 1988:accepted October 30.1989. Supported by ARC, Villejuif,France. Address reprint requests to Marie-Christine Jacob, service d’hbmatologie A, h6pital A . Michallon, CHU. 38040 Grenoble Ckdex. France. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.section 1734 solely to indicate thisfact. 0 1990 by The American Society of Hematology. 0006-4971/90/7505-00II $3.Oo/O 1154
Biopsy specimens were analyzed from different patients with B-cell NHL. The only criteria for selection was the recovery of a sufficient number of lymphocytes after extraction. Thirty-eight lymph nodes and 1 spleen were analyzed. Clinical status is given in Table 1. Sex ratio was 1.8 men for 1 woman and the mean age 61 years (24 to 89 years). Twenty-five patients were studied at the time of diagnosis and 14 after relapse (10 were studied at least 6 months after treatment cessation and four were being treated at the time of biopsy). Lymph nodes 5a 5b and 17a 17b each represent two biopsy samples from the same patient, one at the time of diagnosis and one at the first relapse. Table 1 also defines these lymphomas according to different classifications currently used: the anatomo-clinical cla~sification,~~ the Kiel classification;6 the Working Formulation classification (WF)?’ and the Grenoble cytologic classification.28In Blood, Vol75, No 5 (March 1). 1990: pp 1154-1 162
1155
TIL-T IN B-NHL
Table 1. Characteristics of the 39 NHL Studied Case
No.
Group 1 2 3 4
7 8 9 10 11 12 13 14 15 Group 0 16 17a 17b 18 19 20 21 22 23 24 25 26 27 28 29
30 31 32 33 34 35 36 37
Clinical Status
D R D D D R R R D R R D R D D
IV IV 111
Lc Lc
IV II I I
D
I
cc/cb N cc/cb N cc/cb N cc/cb N cc/cb N cc/cb N cc/cb ND cc/cb ND cc/cb ND cc/cb ND cc/cb D cc/cb D Ib
R D R D D R D R D D D D D R R
IV IV IV IV IV IV 111 IV IV 111 IV 111 IV 111 IV
Lc Lc Lc Lc cc/cb N cdcb ND cc/cb ND cc/cb ND cc/cb ND cc/cb ND cc/cb D cc cc cb cb
Kial
B-Cell Clone Phenotype
WF
Grenoble
Slg
A A A C
1 1
MK MA nd K *K MA nd GX Ax GX nd
+
5a 5b 6
Group
Classifications
Diagnosis (D) or Relapse (R)
111
111 II IV 111 IV 111
IV
IC
1
C C
1 1 1 1
C
1
C C
1 1 1
C
C C
1
D F F H
3 1
A A A A
1 1 1 1
C
1
C C C C D F E E
1 1 1 3 1 1 1
G G
3
1
3
1
1
CDlO
CD5
CD21
-
-
-
+ +
+ +
+
-
+
-MA MK MX MK GK GK
Localization of LN Biopsy
Retroperitoneal Spleen Cervical Mesenteric Cervical Epitroclear Cervical Cervical Cervical Inguinal Mesenteric Cervical Cervical Inguinal Inguinal Inguinal
MK MK nd K MA MA GX GK MK MK MK MK MA MA MK MK
Cervical Axillary Cervical Splenic Cervical Axillary
MK MK MK MA
Coeliac Coeliac Cervical Axillary Cervical Cervical Cervical Axillary
Cervical Inguinal Cervical Axillary Cervical Axillary Inguinal Axillary
II 111 IV Ill
II IV IV IV
IC
cc/cb D cc/cb D cc/cb D cc/cb D cc cb cb
_-
MK GX MA
a and b: Two samples of the same patient (at diagnosis and at the first relapse). Abbreviations: Lc, lymphocytic; cc, centrocytic; cb, centroblastic; cc/cb, centrocytic centroblastic; IC,immunocytic; Ib, immunoblastic; N, nodular; D, diffuse; ND, nodular and diffuse; nd, not done; Slg, surface immunoglobulin; LN, lymph node. 'No light or no heavy chain immunoglobulin.
the latter, NHL are classified with regard to the morphologic features observed under optic microscope. Three groups having a prognostic value are identified essentially from a statistical analysis. In class 1, cells are small lymphocytes (less than 18 pm) with a low cytoplasmic basophilia and less than 1% mitoses. These NHL are related to a favorable prognosis: their evolution remains locoregional for a long time, up to several years. In class 2, cells have a high nuclear cytoplasmic ratio, a high basophilic cytoplasm, and divide frequently (greater than 1% mitoses). Lymphoid cell size is also relatively small (less than 18 pm). These NHL have a poor prognosis with frequent evolution toward acute lymphoblastic leukemia. In class 3, more than 20% of the cells are large cells (greater
than 18 pm). Nuclear cytoplasmic ratio and basophilia may be high or low. These NHL also have a poor prognosis, and behave like solid tumors with early metastases. The immunologic features of the clone are also given in Table 1. Blood samples from 10 healthy donors and lymph nodes from five cases of benign hyperplasia were included as controls.
T- and B-Cell Separation Procedures Whole lymph nodes were received in our laboratory immediately after extraction. They were gently dissociated with a scalpel in RPMI 1640 medium (Boehringer Mannheim, Germany) and washed
JACOB ET
1156
twice. The suspension was depleted of macrophages by adherence on Sephadex G10 (Pharmacia, Sweden). T and B lymphocytes were then separated by rosetting with aminoethylthiouroniumbromidhydrobromide (AET)-sensitized sheep red blood cells (RBCs). Cells were incubated at 4OC overnight, then separated by centrifugation on Ficoll hypaque density gradients. B lymphocytes were removed from the Ficoll medium interface and washed twice. The degree of purity, evaluated by the presence of surface Igs, always exceeded 90%.T lymphocytes were recovered in the pellet. If the proportion of the rosette-forming cells was inferior to 90%, a second or third centrifugation on Ficoll was performed. RBCs were eliminated by lysis with distilled water. The purity of the suspension, evaluated by a positive reaction with anti-CD3 MoAb, ranged from 90% to 98%. For peripheral blood (collected into sterile heparinized tubes) and spleen samples, a preliminary centrifugation on Ficoll hypaque was made to recover mononuclear cells.
AL
RESULTS
Analysis of Activation Features in T Lymphocytes Isolated From NHL Lymph Nodes
The antibodies used were fluoresceine isothiocyanate (FITC)conjugated goat F(ab’), fragments of anti-human heavy and light chain Ig antibodies (Kallestadt, Minnesota), and nonconjugated mice MoAbs: CD3, OKT9 (Ortho Diagnostic Systems, Australia); CD19, CD20, CD24 (Coulter Clone, France); CD4 (Dako Patts, Denmark); CD8, CD25 (Immunotech, France); and CD5, CDlO (a kind gift from A. Bernard, Nice, France). With the MoAbs, an FITC-conjugated goat anti-mouse Ig antiserum (Immunotech) was used as a second-step reagent. Fluorescence analysis was performed on a cytofluorimeter (FACScan; Becton Dickinson, Mountain View, CA).
Surface markers of activation. In this study we used CD25 and OKT9 MoAbs, respectively, directed against IL-2 and transferrin membrane receptors, as indicators of T-cell In most of the cases the number of positive cells was low: more than 10% (10% to 20%) of positive TIL-T were observed in only 12% of invaded lymph nodes (4 of 34 cases). In healthy donors’ blood and in lymph nodes with benign hyperplasia it never exceeded 10%.With the OKT9 MoAb, more than 20% (20% to 63%) of positive TIL-T were observed in 24% of invaded lymph nodes (8 of 34 cases) while they were less than 15% in all the control specimens (Tables 2 and 3). T-lymphocyte response to rIL-2. T Lymphocytes isolated from the different samples (blood or lymph node, N H L or control) were cultured for 6 days with or without rIL-2. Whatever their origin, no spontaneous proliferation was ever demonstrated. With rIL-2, there was a response in the majority of cases: 80% of lymph nodes invaded by N H L or with reaction hyperplasia and 100% of healthy donor peripheral blood (nonsignificant difference). When the response was present, the intensity was extremely variable, particularly TIL-T in NHL, with values oscillating between 1,464 to 39,399 cpm. Although heterogeneity appeared to be more restrained in the control samples, the smallest number of events should be taken into consideration (Tables 2 and 3).
Proliferation Tests
Phenotype of the rIL-2 Responding TIL
Tests were performed in 96-well flat-bottomed plates (Falcon, Becton Dickinson). On day 1, lo5lymphocytes were seeded alone or with lo5irradiated B cells (3,000 rads). The final volume was 0.2 mL RPMI 1640 (Boehringer) supplemented with 2 mmol/L glutamine, 25 mmol/L Hepes, 10% fetal calf serum (Flow Laboratories, England), 100 IU penicillin, and 100 pg/mL streptomycin (Flow Laboratories). IL-2 (Boehringer) was added at a concentrationof 40 U/mL. All assays were performed in triplicate. On day 6, ’Hthymidine (1 pCi/well, CEA Commissariat-1’Energie Atomique, France) was added to the cultures. Cells were harvested 8 hours later and ’H-thymidine incorporation measured by standard scintillation techniques. Proliferation was considered positive when ’H-thymidine incorporation was superior to two standard deviations above the mean of cultures containing lo5 irradiated B lymphocytes with medium alone (or 1,322 cpm). The influence of irradiated B cells (B*) was evaluated by the ratio of T-lymphocyte proliferationin the presence of B* to T-lymphocyte proliferation alone ([B* + T]/T). The effect of the B-cell clone was statistically signficiant (Student’s t test) for a ratio inferior to 0.7 or superior to 1.3.
Whenever the number of recovered cells was sufficient (seven cases), we analyzed CD3 CD4 and CD8 expression on TIL after a 6-day culture with rIL-2 (40 U/mL). Table 4 shows that CD3+ cells were always greater than 90%, demonstrating mainly the proliferation of T lymphocytes.
Immunostaining
Statistical Analysis A data correlation study was performed using three statistical tests: Student’s t test, corrected x2 test, and Mann and Whitney U test.
Computation of Patient Survival Actuarial survival was estimated according to Schwartz et al, 29 and comparison of the groups was performed according to Byar.”
Influence of the B-Cell Clone on Autologous T-Lymphocyte Proliferation in the Presence of rIL-2 We compared the response to rIL-2 of TIL-T alone or in the presence of the irradiated malignant B cells added in a 1:l ratio. Their influence was taken into consideration ( P < .05) for a variation in TIL-T proliferation of at least f30% ([B* + TI /T < 0.7 or > 1.3). According to this criteria, three groups among the 39 N H L were isolated (Table 2). The B-cell clone had no effect on TIL-T - Group 0. proliferation (15 cases or 38%). In five cases (33%) TIL-T alone did not respond to rIL-2, and in 10 cases (67%) it proliferated with values from 1,464 to 39,399 cpm. In four cases, experiments were performed with an increased number of stimulating B lymphocytes (2 x lo5cells/well instead of lo5). In three cases, TIL-T response to rIL-2 remained the same while in one case it was increased ( x 2). - Group +. The B-cell clone enhanced rIL-2 TIL-T response (16 cases or 41%). This effect was variable from one lymph node to another with a (B* + T ) / T ratio ranging from 1.7 to 17.6. In 19% of the cases, TIL-T alone did not respond to rIL-2, and in 81% of them it proliferated with values ranging from 1,464 to 9,202 cpm.
TIL-T IN B-NHL
1157 Table 2. Characteristics of T Lymphocytes From NHL Lymph Nodes ~~~
Response to rlL-2
Case No.
Group 1 2 3 4 5a 5b 6 7 8 9 10 11 12 13 14 15
% TIL-T
T4/T8 (%I
CD25 (%I
OKT9 (%I
T cmp
T
+ Bo cpm
T
+ B*/T
+ 9 39 40 22 64 42 36 35 80 50 40 61 39 53 39 62
312 17/23 18/16 1013 41/18 32/15 3419 20113 60125 34114 35/10 40124 13/22 49/17 2916 30138
ND 8 7 1 1 12 5 6 3 0 1 1 13 1 1 9
ND 3 16 0 0 50 20 7 13 2 1 3 20 2 0 30
957 f 29 2,748 i 351 4,354 f 746 3,531 ir 493 3,248 f 958 3,083 f 391 1,916 f 291 14,165 f 1,917 1,464 i 141 1,141 f 286 2,854 f 480 9,202 f 561 530 f 97 5,384 i 638 1,634 f 663 1,609 f 203
1,726 5,695 32,136 7,498 8,332 20,630 11,757 27,046 4,184 20,054 36,266 20,976 2,753 9,369 3,204 5,474
Group 0 16 17a 17b 18 19 20 21 22 23 24 25 26 27 28 29
20 26 20 49 20 10 47 47 60 60 26 20 12 45 5
17/14 1118 1618 31/21 1513 915 33/10 23112 48/17 24/37 1916 1018 218 43114 510
1 1 11 1 1 1 10 1 0 0 1 20 0 1 ND
1 1 0 0 0 38 2 2 4 0 0 35 0 2 ND
790 2,964 4,330 12,890 3,623 7,720 1,188 1,704 800 1,737 917 11,957 961 39,399 1,464
f
291 2,515 4,785 11,300 2,852 6,392 658 1,369 789 1,320 1,093 13,164 669 35,629 1,754
Group 30 31 32 33 34 35 36 37
50 45 18 10 10 22 10 25
32/14 30114 1119 314 715 10116 518 20113
1 ND 1 ND 5 7 ND 2
0 ND 30 ND 63 5 ND 3
1,863 5,144 9,737 1,831 3,700 1,656 4,609 2,389
i
305 595 f 237 f 570 f 762 f 1,489 f 383 f 361 f 300 i 113 f 25 f 558 i 305 + 3,034 i 499 f
176
f 930
812 166 i 244 f 177 2 368 f 434 f
f
614 2,282 3,554 551 2,597 880 2,529 1,135
i
192
f 624 f 1,000
418 322 f 2,173 f 1,101 f 632 f 732 f 132 f 4,451 f 1,626 f 199 f 1,429 f 802 i 1,238 i
f
1.8 2.1 7.4 2.1 2.6 6.7 6.1 1.8 2.9 17.6 12.7 2.3 5.2 1.7 2.0 3.4
f 28 i
484
f 172 f f
i f
f f
f i f f
f f
1,800 317 542 336 276 23 251 86 1,118 418 2,574 33
39 346 f 555 f 104 f 502 f 174 f 115 5 24 f f
0.0 1.1 0.9 0.8 0.8 0.8 0.8 1.1 0.9 1.2
0.3 0.4 0.4 0.3 0.7 0.5
0.5 0.5
Three groups were identified on the basis of B-cell influence on TIL-T proliferation with rlL-2: positive (+), null (0).or negative (-). There is no correlation between these groups and CD4lCD8 or with activation markers on TIL-T; a and b (5 or 17) representsamples from the same patient. Abbreviations: ND, not done; B*, irradiated B lymphocytes.
- Group -. The B-cell clone produced a decrease in TIL-T proliferation in the presence of rIL-2 (eight cases or 21%).This effect varied in proportion from 0.3 to 0.7. In all cases, T lymphocytes alone also responded to rIL-2 (range 1,656 to 9,737 cpm). Two patients (cases 5 and 17) sampled on two different occasions remained within the same group at the time of diagnosis and during relapse. In healthy donors’ blood and hyperplastic lymph nodes (Table 3), only groups 0 and + were present: 10% and 9095, respectively, for healthy donors’ blood, 40% and 60% for
hyperplastic lymph nodes. The positive effect of B lymphocytes appeared less heterogeneousin control samples where B lymphocytes are polyclonal than in NHL lymph nodes where they are monoclonal. A ratio of (B* + T)/T > 10 was observed in two cases of NHL and never noted in the control populations. Correlation With Clinical Presentation and Outcome
Patients had been classified into four clinicopathologic stages2’ and their survival evaluated with the actuarial Results showed no difference with the first
1158
JACOB ET AL Table 3. Characteristics of T Lymphocytes From Control Patients
Case No.
Response to rlL-2
% T Lymphocytes in Lymph Nodes
OKT9 (%I
T4/T8 (%)
CD25 (%)
70120 501 15 54/25 2419 64/32
3 1 1 10 0
0 0 1 8 0
1 ND 1 1 3 1 1 1 2 3
0 ND 3 2 15
T cpm
T+B*cpm
T
+ B*/T
Hyperplastic lymph nodes
1 2
85 62 80 56 77
3 4 5
607 f 159 5,069 f 277 3,841 f 825 2,201 f 489 1,220 f 208
-
814 f 389 5,766 f 191 5,566 f 825 3,056 f 321 2,680 f 1,383
1.1 1.4 1.4 2.2
521 733 496 569 1,471 991 1,548 1,759 1,266 74
1.1 1.3 1.6 2.0 2.0 2.1 3.5 3.6 4.3 5.8
Peripheral blood from healthy patients
1 L
3 4 5 6 7 8 9 10
0 1 1 1 3
1,485 7,609 4,285 2,467 5,652 5,555 2,154 9,090 7,891 5,073
f f f
f
k f f
f f
f
933 822 590 733 520 616 803 559 861 97
1,634 9,528 6,680 4,892 11,312 11,574 5,705 32,781 34,164 29,280
f f f
f f f f
f f f
B-cell influence on T-lymphocyte response to rlL-2 was studied in five hyperplastic lymph nodes and peripheral blood from 10 healthy patients. In 12 of 15 cases, 8 lymphocytes increase T-lymphocyte proliferation. In the remaining three cases they had no effect. Abbreviations: ND, not done: 8.
irradiated B lymphocytes.
classification but a better survival in group + and 0 than in group - (statistically significant [P< .05] between group + and -) (Table 1 and Fig 1). Correlation With Different NHL Classifications Immunologic classification. Table 5 shows the effect of the B-cell clone according to its immunologic phenotype. There was no relationship with the pan B markers (CD19, CD20, CD24). Differences appeared with the B-cell subset restricted antibodies: CDlO and CD21 were 4 times more frequent in groups + and 0 than in group -, and CD5 about 5 to 3 times more frequent in group 0 than in groups + and - . However, these differences are not statistically significant ( P > .05). The heavy and light chain Ig isotypes on the B-cell clone surface are shown in Table 1. There is no relationship with the influence of the malignant population. Kiel classifcation. Table 1 shows the distribution of lymphomas in each of the three groups previously defined. There is no evident correlation with their more or less differentiated features. However, the particular distribution of centrofollicular N H L should be noted (Table 6): all Table 4. Phenotype of rll-2-Responding TIL-T Day 6 of Culture With rlL-2
Day 0 CD3
CD3
CD4
CD8
98 94 93 90 90 93 95
95 92 94 90 90 86 88
30 38 44 57 3 44 62
62 65 67 32 92 53 43
rlL-2 respondingTIL-T phenotype was examined after a 6-day culture. The positivity with CD3 MoAb indicates that most responding cells are actually T lymphocytes.
nodular or nodular and diffuse proliferations belong to groups + and 0. They represent 83% and 55%, respectively, of the centrofollicular lymphomas of these groups (among them, pure nodular proliferations are almost exclusively found in group +). None are demonstrated in group -. The diffuse proliferations are found in all three groups, where they represent 17% in group +, 45% in group 0, and 100% in group - . The differences are statistically significant ( P < .01). Working formulation classification. Disease prognosis according to the influence of the B-cell clone on TIL-T proliferation is shown in Table 7. N H L of group + are essentially low-grade malignancies (75%) while those of group - are intermediate-grade malignancies (88%). Differences are statistically significant ( P < .02). Group 0 is divided almost evenly between the low-grade malignancies (60%) and the intermediate-grade malignancies (40%), and it is not statistically different from either group + or group -
Grenoble cytologic classification. Eighty-eight percent of the group + and 87% of the group 0 N H L belong to cytologic class 1, whereas the group - N H L appears to be strongly linked to class 3 (88%) (Table 7). The distributions in classes 1 and 3 are significantly different between groups + and - but also between groups 0 and - (P< .Ol), in contrast with the results of the WF. W e show that the Grenoble classification gives the best degree of correlation with that kind of malignant B-cell clone influence. This results essentially from the classification of diffuse centrofollicular lymphomas for which the cytology is more discriminative than the histology. With the WF, all of these lymphomas are intermediate-grade malignancies. In groups + and 0, they do not have the same degree of malignancy as the other NHL, which are low-grade malignancies. With the Grenoble classification, these N H L are either of class 1 or class 3 like
1159
TIL-T IN 6-NHL
% loo
80
Fig 1. NHL patient survival according t o classification in group f, 0, and Survival was studied by the actuarial method. Patients from group and 0 had a better life span
-.
60
--
40
--
20
--
+
between group
+
and
=F;2l8-.-.----.
VO‘
--
.
1
-m-w
..
I
+, class
TIL-T were identified in lymph nodes by two MoAbs, CD3 and CD5. The relative proportions of malignant B cells and T lymphocytes appeared to be highly variable from one patient to another with T-lymphocyte values ranging from 5% to 80%. In group + there are more than 35% TIL-T in 88% of the cases (mean 44% 2 17%), while in group -, there are less than 26% in 75% of the cases (mean 24% f 15). In group 0, the number of T lymphocytes is greater than 35% in 40% of the cases, and inferior to 26% in 60% of the cases (mean 31% 18%). Differences are significant between groups + and - ( P < .05), but not between groups + and 0. DISCUSSION
In lymph nodes invaded by malignant lymphoma, the B-cell clone is always associated in highly variable proportions, most likely with normal T lymphocytes. We explored the significance of these TIL-T in the tumor microenvironment by looking for evidence in favor of their reactivity toward the malignant cells. First, we tried to demonstrate the existence of activated TIL-T by studying the presence of IL-2 and transferrin
Group0 Group
-
CD19
CD20
CD24
CD21
CD5
CDlO
87% (15)* 93% (15) 100%
100% (10) 75% (12) 86% (7)
91% (11) 86% (14) 75% (8)
67% (151 60% (15) 25% (8)
6% (16) 33% (15) 12% (8)
56% (16) 47% (15) 12% (8)
(7)
-0
-0
-m
-m
-m-u
0-0
dead people: 3/14 -0 0
dead people: 618
I
6-cell clone was studied with six MoAbs, specific either for pan B-cell population (CD19, CD20, CD24) or for 6-cell subsets (CD21, CD5, CD 10). Statistical analysis showed no correlation between 6-cell phenotype and the kind of influence (+, 0, - 1 on TIL-T proliferationwith rlL-2. *Number of cases studied is in parentheses.
*. .
Group
months
-
surface receptor^^^*^^ on one hand, and the proliferative response to rJL-2 on the other. Only activated T lymphocytes proliferate at low concentrations of rIL-2.24 CD25 and OKT9 + TIL-T are present in 90% of the cases, but their percentage is very low in most samples. Results are similar when IL-2 receptors are studied on histologic slide^.'^*^^.'^ In 80% of the NHL studied, TIL-T respond to IL-2, but with a great variability from one patient to another. This inconsistency has also been observed in solid tumor^.^.'^ Cell-cycle analysis also confirms the existence of normal proliferating TIL-T in human lymphomas and demonstrates that this fraction is larger than in reactive lymph nodes.” Our results show that there is no unique behavior within NHL, and no significant difference in comparison with control populations. However, other studies performed in the same patients showed greater T-lymphocyte activation in the tumor than in the peripheral blood .2 Secondly, we looked for relationships between TIL-T and malignant B cells. In MLTC there was no proliferation of TIL-T (data not shown). However, in the presence of rIL-2 we demonstrated functional interrelations between B and T cells: the B-cell clone increased the response to rIL-2 in 41% Table 6. Percent of Nodular, Nodular and Diffuse, and Diffuse 0, and Centrofollicular NHL in Groups
+,
Table 5. B-Call Influence on T-cell Proliferation W i t h Regard t o I t s Immunologic Phenotype
+
-0
.o-Group 0
.a- Group +
Percentage of the TIL-T
Group
-
1
-
the other NHL of the group (class 1 in groups 0 and 3 in group -).
dead peolple: 3/15
0
%/Centrofollicular NHL
Group
+
Group 0 Group
-
N
ND
50%
33%
(6)* 9% (1) 0%
(4) 46% (5) 0%
D
-
% ofCentrofollicular NHLnotal NHL
17%
75%
(2) 45% (5) 100% (7)
(72) 73% (11) 88% (7)
The nodular or diffuse characteristics of the malignant proliferationwas studied in centrofollicular lymphomas. In group 6-cell proliferation was essentially N or ND (87%); in group 0 it was either N, ND (55%) or D (45%); whereas it was always D (100%) in group Abbreviations: N, nodular; D, diffuse; ND, nodular and diffuse. *Number of cases studied is in parentheses.
+, -.
1160
JACOB ET AL
Table 7. Correlation Between B-Cell Influence on TIL-T Proliferation end Prognosis Classifications Grwp
+
Group 0
Group
Low-grade Intermediate-grade High-grade
75 19 6
60 40 0
12 88 0
Class 1 Class 3
88 12
87 13
12 88
-
malignant cells of T-lymphocyte proliferative responses in the presence of different mitogens, allogenic or autologous cells, has been reported elsewhere in the case of lid^^'^^^^-^^ and lymphoid42tumors. Among the mechanisms responsible for this inhibition are the secretion of soluble factors by malignant cells'4s42343 and the induction of suppressor T
lymphocyte^.'.^^ E-cell influence on TIL-T proliferation was compared with two classifications that have an essentially prognostic value: the Working Formulation (WF) (top) and the Grenoble cytological classification (GCC) (bottom). The majority of group NHL were classified in the categories of the lowest malignancy, whereas group - NHL were of higher malig nancy. The degree of malignancy in group 0 was not the same whether it was evaluated with one classification or the other. With the WF, NHL were almost equally low-grade and intermediate-grade malignancies while they were essentially in class 1 of favorable prognosis with the GCC.
+
of the cases, inhibited it in 21% of the cases, and was without effect in 34%. The mechanism of the positive interaction has not been addressed in this work, but it is under investigation. Several hypotheses may explain this T-cell activation: (1) a simple feeder effect; (2) cell-membrane contact, either nonspecific such as HLA class I1 recognition (as in AMLR6), or specific such as Ig idiotype recognition; (3) it may also be mediated by soluble factors such as 11-1,32which is known to be secreted by some malignant B cell^.^' The existence of T-lymphocyte responsiveness to malignant cells (group + ) has also been reported in other cancers. A number of MLTC studies show similar results, but most of them concerned peripheral blood7-"~34 and not TIL-T.*.I2 The absence of a B-cell clone influence (group 0) might mean that B cells are weak stimulants in AMLR, as it is demonstrated in chronic lymphocytic leukemia (CLL);'S~~or that TIL-T are poor responders. Nevertheless, their capacity to proliferate with rIL-2 and mitogens (phytohemagglutinin [PHA], concanavalin A [ConA], and pokeweed mitogen [PWM] ) indicate that their intrinsic proliferative properties are not uniformally low (data not shown). Furthermore, studies provided by Cazzolino et all2 suggested that their might exist a suppressor T-lymphocyte population in this system. The originality of N H L lymph nodes as compared with control populations is the existence of a group within which B cells provide an inhibiting effect on T-cell proliferation. In our observations, we can rule out the hypothesis of rIL-2 fixation on B lymphocytes (which would make it unavailable for T lymphocytes). This phenomenon has been described in CLL37but seems improbable in the majority of our observations because, in 7 of 9 cases, the B lymphocytes were not positive for CD25 antibodies, and in 8 of 9 cases did not proliferate in the presence of rIL-2. Furthermore, we were able to demonstrate, in one patient, that an increase in rIL-2 concentration (80 U/mL) had no influence on the negative effect of the B-cell clone (data not shown). Inhibition by
The significance of these different effects in the pathophysiology of lymphoma remains to be defined. In our study we investigated the existence of relationships between three kinds of B-cell clone influence (stimulating, inhibiting, or null) and disease expression. Our results show that a correlation exists between the high percentage of TIL-T in vivo and evidence of stimulation by the malignant clone in vitro, and inversely between the weak proportion of TIL-T in vivo and the absence or inhibition of their proliferation in vitro (P < .05). Because B-cell lymphomas represent the expansion of B-cell clones arrested in extremely varied stages of maturation, we wondered whether the nature of the B-cell clone (its differentiation level) was related to the effects observed. The response provided by statistical analysis is negative. However, differences appeared in the distribution of the three B-cell subset restricted antigens: CD5, CD10, and CD21. This suggests that by increasing the number of samples one could isolate phenotypic groups linked to a B-cell clone effect. The malignant cell differentiation as determined by the Kiel classification does not show a correlation with the stimulating, inhibiting, or null effect of the clone on T-lymphocyte proliferation. However, nodular proliferations appear to be linked to groups + and 0, while diffuse proliferation to group - . The observation of a relationship between the results of MLTC and the disease prognosis in solid led us to examine this situation in our study. There was no correlation with prognosis as evaluated by clinicopathologic classification, but correlations appeared with other parameters. Two classifications have an exclusively prognostic objective for NHL: the WF,27which is the reference classification, and the Grenoble cytologic classification.28 Thus we show that the positive influence of the clone is associated with a favorable prognosis in 75% (WF) or 88% (Grenoble classification) of the cases. Conversely, the inhibiting influence of the clone is linked to a poor prognosis in 88% of the cases ( W F and Grenoble classifications). The prognosis for group 0 is different according to whether it is evaluated with the W F (60% of low-grade malignancies and 40% of intermediategrade malignancies) or with the Grenoble classification (87% of favorable prognosis and 13% of poor prognosis). The difference, as we have already shown, comes from the distribution of some diffuse centrofollicular lymphomas in cytologic class 1 and of favorable prognosis. To confirm the prognostic values of the groups +, 0, and -, a statistical comparison of the patients survival was made. A better prognosis was effectively shown for groups + and 0, and differences were significant between groups + and (P< .OS). This difference in the disease prognosis is not due
TIL-T IN B-NHL
1161
to an intrinsic T-cell deficiency, as the response to rIL-2 or mitogens is not statistically different between the three groups. These results are in agreement with those obtained from solid tumors, showing positive MLTC in patients without metastases, negative in those with positive MLTC in favorable evolutions, and negative in the pejorative The existence of a correlation between
TIL-T response to rIL-2 in the presence of malignant autologous B cells and the disease expression led us to postulate that these reactive TIL-T might be important in the pathophysiology of the disease. Recent studies on solid tumors suggest that these cells might have some therapeutic application in the future if, as it was s h o ~ n ; ~ tumor.~~ specific cytotoxic activity could be developed from them.
REFERENCES
1. Foon KA, Todd R F Immunologic classification of leukemia and lymphoma. Blood 68:1,1986 2. Wose BM, Moore M: Human tumor-infiltrating lymphocytes: A marker of host response. Semin Hematol22:27, 1985 3. Gulwani B, Maio M, Imberti L, Melamede RJ, Ferrone S: Immunosuppressive activity of T cell clones generated from autologous lymphocytes cultures. J Immunol 139:2130,1987 4. Rensenkrantz K, Dupont B, Williams D, Flomenberg N: Autocytotoxic and autosuppressor T-cell lines generated from autologous lymphocyte cultures. Hum Immunol 19:189, 1987 5. Santoli D, Radka SF, Igarashi M, Kreider BL, Ferrone S: Autologous B IymphoblastoTd cell lines and long-term cultured T cells as stimulators in the mixed lymphocyte reaction: Analysis of the role of HLA class I1 antigens as stimulatory molecules. J Immunol 137:400, 1986 6. Weksler ME, Moody CE, Kozak RW: The autologous mixed lymphocyte reaction. Adv Immunol31:271, 1981 7. Vanky F, Peterfly A, Book K, Willems J, Klein E, Klein G Correlation between lymphocyte-mediated auto-tumor reactivities and the clinical course. 11. Evaluation of 69 patients with lung carcinoma. Cancer Immunol Immunother 16:17, 1983 8. Vanky F, Willems J, Kreichbergs A, Aparisi T, Andreen M, Brostrom LA, Nilsonner U, Klein E, Klein G: Correlation between lymphocyte-mediated auto-tumor reactivities and clinical course. I. Evaluation of 46 patients with sarcomas. Cancer Immunol Immunother 16:11, 1983 9. Akiyama M, Bean MA, Sadamoto K, Takahashi Y, Brandkovan V: Suppression of the responsiveness of lymphocytes from cancer patients triggered by coculture with autologous tumor-derived cells. J Immunol131:3085,1983 10. De Vries JE, Spits H: Cloned human cytotoxic T lymphocyte (CTL) lines reactive with autologous melanoma cells. I. In vitro generation, isolation and analysis to phenotype and specificity. J Immunol 132510,1984 11. Slovin SF, Lackman RD, Ferrone S, Kiely PE, Mastrangels MJ: Cellular immune response to human sarcomas: Cytotoxic T cell clones reactive with autologous sarcomas. I. Development, phenotype and specificity. J Immunol 137:3042, 1986 12. Cazzolino F, Torcia M, Carossino AM, Giordani R, Sell C, Talini G, Reali E, Navelli A, Pistoia V, Ferrarini M: Characterization of cells from invaded lymph nodes in patients with solid tumors. J Exp Med 166:303,1987 13. Garcia CF, Lowder J, Meeker JC, Bindl J, Levy R, Warnke RA: Differences in “host infiltrates” among lymphoma patients treated with anti-idiotype antibodies: Correlation with treatment response. J Immunol 135:4252, 1985 14. Miescher S, Whiteside TL, Carrel S, Von Fliedner V: Functional properties of tumor-infiltrating and blood lymphocytes in patients with solid tumors: Effect of tumor cells and their supernatants on proliferative responses of lymphocytes. J Immunol 136: 1899,1986 15. Sheibani K, Winberg CD, Van de Velde S, Blaynery DW, Rappaport H: Distribution of lymphocytes with interleukine-2
receptors (TAC Antigens) in reactive lymphoproliferative processes, Hodgkin’s disease and non-Hodgkin’s lymphomas. An immunohistologic study of 300 cases. Am J Clin Pathol 127:27,1987 16. Stranchen JA, Breakstone BA: IL2 receptor expression in human lymphoid lesions. Immunohistological study of 166 cases. Am J Clin Pathol 127:27, 1987 17. Brasiel RM, Sussman E, Baffe ES, Neckersel M, Cossman B: Induction of immunoglobulin secretion in follicular non-Hodgkin’s lymphoma. Role of immunoregulatory T cells. Blood 66:128,1985 18. Ford RJ, Kouttalo NM, Sahasrabuddhe CG, Davis FM, Mehta SR: Growth factor-mediated proliferation in B cell nonHodgkin’s lymphomas. Blood 651335, 1985 19. Laurent G, AI Saati T, Olive D, Laurent JC, Poncelet P, Delsol G: Expression of Tac antigen in B cell lymphomas. Clin Exp Immunol65:354,1986 20. Weiss LM, Michie SA, Mereidos LJ, Strickter JG, Garcia CF, Warnke RA: Expression of Tac antigen by non-Hodgkin’s lymphomas. Am J Clin Pathol65:354, 1986 21. Yssel H, Spits H, De Vries J E A cloned human T cell line cytotoxic for autologous and allogenic B lymphoma cells. J Exp Med 160:239, 1984 22. Tsudo M, Uchiyama T, Takatsuki K, Uchino H, Yodo J: Modulation of Tac antigen on activated human T cells by anti-Tac monoclonal antibody. J Immunol 129592,1982 23. Hercend T, Ritz J, Schlossman SF, Reinherz E L Comparative expression of T9, T10 and Ia antigens on activated human T cell subsets. Hum Immunol3:247, 1981 24. Harel-Bellan A, Bertoglio J, Quillet A, Marchiol C, Wakasuki H, Mishall Z, Fradelizi D: Interleukin 2 (IL2) up regulates its own receptor on a subset of human unprimed peripheral blood lymphocytes and triggers their proliferation. J Immunol 136:2463, 1986 25. Carbone PP, Kaplan HS, Musshof K, Smithers DW, Tubiana M: Report of the committee on Hodgkin’s disease staging classification. Cancer 31:1860, 1971 26. Lennert K, Mohri N, Stein H, Koiserling E: The histopathologyof malignant lymphoma. Br J Haematol31:193, 1975 (suppl) 27. The non-Hodgkin’s lymphoma pathologic classificationproject National Cancer Institute sponsored study of classification of non-Hodgkin’s lymphomas. Summary, and description of a working formulation for clinical usage. Cancer 49:2112, 1982 28. Sotto JJ, Caillot P, Sotto MF, Brambilla E, Lachet B, Michallet M, Martin H, Bensa JC, Schaerer R, Seigneurin D, Hollard D: Non-Hodgkin’s malignant lymphomas. A cytologic classification based on statistical analysis and correlated with prognosis. Anal Quant Cytol46, 1982 29. Schwartz D, Flamant R, Lellouch J: L‘essai thtrapeuthique chez I’homme. Paris Flammarion 218,1970 30. Byar D P Analysis of survival data in heterogeneous populations, in Barra JR, Van Custen B, Brodeau F, Romier G (eds): Recent Developments in Statistics: Proceedings of 1976 European Meeting of Statisticians, Grenoble. Amsterdam, The Netherlands, North Holland, 1977, p 51
1162
31. Schackney SE, Skramstad K A dynamic interpretation of multiparameter studies in the lymphomas. Am J Clin Pathol72:756, 1979 32. Mizel SB: Interleukin 1 and T cell activation. Immunol Today 8:330,1987 33. Uggla C, Santelises MA, Rosen A, Mellstedt H, Jondal M: Spontaneous production of interleukin 1 activity by chronic lymphocytic leukemic cells. Blood 70:1851, 1987 34. Herein H, Lemoine C, Weynants P, Vessiere F, Van Pel A, Knuth A, Devos R, Boon T: Production of stable cytolytic T cell clones directed against autologous human melanoma. Int J Cancer 39:390, 1987 35. Ayanlar-Batman 0,Ebert E, Hauptman S: Defectiveinterleukine-2 production and responsiveness by T cells in patients with chronic lymphocytic leukemia of B cell variety. Blood 67:279, 1986 36. Wolos JA, Davey F R Function of lymphocyte subpopulations in chronic lymphocytic leukemia. Activity in the allogenic and autologous mixed lymphocyte reaction. Cancer 45:893, 1980 37. Foa R, Giovarelli M, Jemma C, Fierro MT, Lusso P, Ferrando ML, Lauria F, Forni G: Interleukin 2 (IL2) and interferon gamma production by T lymphocytes from patients with B-chronic lymphocytic leukemia: Evidence that normally released IL2 is absorbed by the neoplastic B cell population. Blood 66:614, 1985 38. Brooks KH, Vitetta ES: Recombinant IL2 but not recombinant interferon gamma stimulates both proliferation and IgM
JACOB ET AL
secretion in a lyl + clone of neoplastic murine B cells (BCLl). J Immunol137:3205,1986 39. Hersey P, Bindon C, Czerniecki M, Spurling A, Wass J, McCarthy WH: Inhibition of interleukin 2 production by factors released from tumor cells. J Immunol 131:2837, 1983 40. Mantovani A, Evans R: Tumor infiltrating leukocytes. Immuno1 Today 6:143,1985 41. Roth JA, Grimm EA, Gupta RK, Ames RS: Immunoregulatory factors derived from human tumors. I. Immunologic and biochemical characterization of factors that suppress lymphocyte proliferative and cytotoxic responses in vitro. J Immunol 128:1955, 1982 42. Hersh EM, Drewinko B: Specific inhibition of lymphocyte blastogenic responses to mitogens by a factor produced by cultured human malignant lymphoma cells. Cancer Res 34215, 1974 43. Fontana A, Hengartner H, De Tribolet N, Weber E: Glioblastoma cells release interleukin 1 and factors inhibiting interleukin 2-mediated effects. J Immunol 132:1837, 1984 44. Wei WZ, Ratner S, Folton AM, Heppner GH: Inflammatory infiltrates of experimental mammary cancers. Biochim Biophys Acta 865:13,1986 45. Anichini A, Fossati G, Parmiani G: Clonal analysis of the cytolytic T cell response to human tumors. Immunol Today 8:385, 1987 46. Rosenberg SR,Spiess P, Lafieniere R: A new approach to the adoptive immunotherapy of cancer with tumor infiltrating lymphocytes. Science 233:1318, 1986
