~

Cancer Immunol Irnmunother (1992) 34:241 - 251

ancer mmunolggy mmunotherapy

© Springer-Verlag 1992

Phenotypic and functional analysis of lymphocytes infiltrating paediatric tumours, with a characterization of the tumour phenotype Licia Rivoltinil, Flavio Arientil, Attilio Orazi2, Graziella Cefal03, Marco Gasparini3, Carlo Gambacorti-Passerinil, Franca Fossati-Bellani3, and Giorgio Parmianil 1 Division of Experimental Oncology D, 2 Division of Pathology and 3 Division of Paediatric Oncology, Istituto Nazionale Tumori, 1-20 133 Milan, Italy Received 27 August 1991/Accepted 10ctober 1991

Summary. Tumour-infiltrating lymphocytes (TIL) of paediatric tumours obtained from 37 lesions of different histotype (12 osteosarcomas, 5 Wilms' tumours, 7 soft-tissue sarcomas, 5 neuroblastomas and 8 miscellaneous) were studied to establish their potential for therapy. Fresh isolated TIL were cultured for the first 2 weeks with low doses of interleukin-2 (IL-2) (20 Cetus U/ml) to select for "tumour-specific" lymphocytes potentially present in the neoplastic lesion, followed by culture with high doses of IL-2 (1000 Cetus U/ml) to achieve TIL expansion. TIL were grown with more than 10-fold expansion in only 9 cases (mean expansion: 58-fold, range 13.5-346). In 17 cases no viable cells were obtained. After 30 days of culture with IL-2 the proliferative ability of TIL declined sharply in the majority of cases and TIL became refractory to any further stimulus, including addition of IL-4, tumour necrosis factor c~ (TNF~) or interferon 7, and activation with OKT3 in solid phase. In 20 out of 37 cases TIL were available for phenotypic and functional analysis. TIL after long-term culture were predominantly CD3 + but 2 cases of osteosarcoma showed a predominance of CD3+TcR 7/8 cells. The CD4/CD8 ratio was more than 1 in 10 cases, without correlation with tumour histology, site of lesion or TIL growth. The number of CD16 + and CD25 ÷ lymphocytes decreased progressively during culture, the lätter concomitantly with a reduction of TIL growth rate. The lytic pattern of TIL against allogeneic and autologous tumour (Auto-Tu) cells was variable, but specific lysis of Auto-Tu was seen in only one case (Wilms' tumour) after culture with TNF~ and irradiated Auto-Tu cells. The immunohistochemical analysis of tumour lesions revealed a limited lymphocyte infiltrate, a low expression of histocompatibility leukocyte antigens (HLA) class I and of the adhesion molecules ICAM1, LFA3, and a significant production of transforming growth factor ~ (TGF~). These data indicate that TIL obtained from paediatric patients are difficult to expand at levels required for immunotherapy Offprint requests to: L. Rivoltini, Division of Experimental Oncology D, Isfituto Nazionale Tumori, Via Venezian 1,1-20 133 Milano, Italy

and lack a significant number of tumour-specific T lymphocytes. A low expression of immunomodulatory molecules on tumour cells or the production of suppressive factors may prevent activation and expansion of TIL in paediatric tumours.

Key words: Tumour-infiltrating lymphocytes - Paediatric tumours - Cytokines - TGF[~

Introduction Childhood malignancies represent a major therapeutic challenge in medical oncology. Despite recent progress in new chemo- and radiotherapy regimens, the long-term outcome for patients affected by advanced paediatric tumours remains poor and little improvement can be expected from conventional treatments. Thus new therapeutic approaches, including adoptive immunotherapy, should be evaluated in the management of these neoplasias. Considering the clinical results obtained in renal cell cancer or in melanoma [22], the use of recombinant interleukin-2 (IL-2) with or without reinjection of lymphokineactivated killer (LAK) cells has been recently studied in children. Although data about phase I - I I trials in paediatric tumours are limited [5, 19, 20] and confined to selected histotypes (i. e., neuroblastoma, embryonal rhabdomyosarcoma and acute leukemia), the IL-2 + LAK regimen does not appear to have a significant therapeutic impact in these patients. More recent studies have focused on the isolation, in vitro activation and expansion of lymphocytes present in the tumour lesion, i.e. tumour-infiltrating lymphocytes (TIL). TIL represent a potential source of tumour-specific T cells to be used with IL-2 in adoptive immunotherapy of cancer. The in vivo therapeutic relevance of these T eell populations has been documented in murine models [23] and in a clinical trial [24].

242 In the course o f T I L studies, h o w e v e r , it was noted that there are different patterns o f t u m o u r recognition and killing b y T I L a m o n g h u m a n tumours [2]. In fact, autologous m a j o r - h i s t o c o m p a t i b i l i t y - c o m p l e x (MHC)-restricted tumour-specific c y t o t o x i c i t y m a y be often o b s e r v e d in IL2-activated T I L f r o m metastatic m e l a n o m a s [12] but less frequently in those f r o m g l i o m a s [14], ovarian cancers [6] and other tumours [15]. In contrast, M H C - u n r e s t r i c t e d cytotoxic activity is u s u a l l y r e p o r t e d in T I L from the m a j o r i t y o f h u m a n cancers, including renal and other u r o l o g i c a l carcinomas, liver c a r c i n o m a and ovarian and n a s o p h a r y n geal tumours [3, 9 - 1 1 , 29]. Thus, the role o f T I L has not been clearly established since they could represent a tumour-specific response o f T l y m p h o c y t e s or s i m p l y a c o m p o n e n t o f a non-specific i n f l a m m a t o r y infiltrate. L i k e w i s e , the i m p a c t o f T I L therapy in controlling the g r o w t h and metastatic spread o f cancer is still to be defined. D e s p i t e the ability to b e e x p a n d e d in the presence o f IL-2, functional studies h a v e shown that freshly isolated T I L are often unable to m e d i a t e antitumour c y t o t o x i c i t y and proliferate p o o r l y to m i t o g e n s [17]. It is p o s s i b l e that the r e d u c e d r e s p o n s i v e n e s s o f T I L m a y be due to suppressive influences o f t u m o u r environment. Alternatively, alterations in t u m o u r e x p r e s s i o n o f i m m u n o r e g o l a t o r y molecules, e.g. M H C or a d h e s i o n molecules, m a y be inv o l v e d in d e t e r m i n i n g a r e d u c e d r e c o g n i t i o n o f t u m o u r cells b y T l y m p h o c y t e s [33, 34]. In an attempt to use T I L for therapeutic use, we have first assessed the in vitro functional and p h e n o t y p i c properties o f such l y m p h o c y t e s o b t a i n e d f r o m c h i l d h o o d m a l i g nancies. Different culture conditions have been studied to i m p r o v e T I L outgrowth, including c o m b i n a t i o n of I L - 2 with t u m o u r necrosis factor c~ (TNF-o0, I L - 4 or interf e r o n 7 (IFN7), and activation with anti-CD3 m A b . M o r e o v e r , the e x p r e s s i o n o f m o l e c u l e s k n o w n to be inv o l v e d in T l y m p h o c y t e activation, i. e. M H C and adhesion molecules, or the p r o d u c t i o n o f i m m u n o s u p p r e s s i v e factors (e. g. transforming growth factor [3, T G F ~ ) has been evaluated on t u m o u r cells f r o m these neoplasias. O u t data show that T I L o b t a i n e d f r o m 37 lesions o f different paediatric patients exhibit a l o w in vitro expansion, even in the p r e s e n c e o f different proliferative stimuli, and with a l m o s t no e v i d e n c e o f specific or preferential recognition o f autologous t u m o u r cells. M o r e o v e r , imm u n o h i s t o c h e m i c a l analysis s h o w e d that t u m o u r cells f r o m these lesions are c h a r a c t e r i z e d b y a low e x p r e s s i o n o f M H C and a d h e s i o n molecules, and TGF[3 release, features that m a y concur in p r e v e n t i n g the interaction and efficient activation o f TIL.

1991. Table 1 shows the characteristics of these patients. Written informed consent was obtained before patients entered the study.

Materials and methods

Combination of IL-2 with different cytokines. The addition of TNF{x (a gift from EuroCetus), IL-4 (lmmunex Manufacturing Corp., Seattle,, Wash., USA) and 1FN7 (kindly provided by Hoffman-La Roche, Basel, Switzerland) to in vitro culture with IL-2 was at a concentration of 250 U/tal, 80 U/tal and 100 U/ml respectively.

Patients. Paediatric patients were eligible for this study if they had metastatic cancer and had failed all standard forms of therapy or had disease for which no effective therapy was available. Patients with brain metastases or patients taking steroid medications, or testing positive for human immunodeficiency virus (HIV) antibody, were excluded from the study. A group of 37 paediatric patients fulfilling the above-mentioned criteria were included into the study between January 1989 and January

TIL culture. Tumour specimens were received directly from the operating room, minced with scissors into 5-mm3 fragments and digested in an enzyme mixture containing collagenase and hyaluronidase (Sigma Chemical Co., St. Louis, Mo., USA) for 4 24 h. The suspension, containing a mixed population of tumour cells, morlocytes and lymphocytes, was adjusted to a concentration of 106 viable cells/ml in complete medium, consisting of RPMI-1640 medium (MA Bioproducts, Walkersville, Md., USA) with 10% fetal calf serum (FCS). IL-2 (kindly provided by EuroCetus B. V., Amsterdam, The Netherlands) was added at the concentration of 20 U/tal for 2 weeks and 1000 U/ml for the remaining culture time. Cultures were split weekly by harvesting and resuspending cells in fresh media, the viable cell concentration being adjusted to 106/rel. An aliquot of the cell suspension was cultured without IL-2 to allow tumour cell growth, when the cell number was sufficient. Indirect immunoßuorescence analysis. Lymphocytes (2 × 105) were suspended in 0.1 ml/well on 96-well flat-bottomed plates (Costar, Cambridge, Mass., USA), incubated at 4°C for 30 min with the primary monoclonal antibody (mAb), washed twice with phosphate-buffered saline (PBS) and the cells further incnbated at 4 ° C for 30 min with 0.1 ml 1 : 30 dilution of fluoresceinated goat anti-(mouse Ig) (Technogenetics, Trezzano, Italy). Cells were mechanically removed and all samples analysed by a fluorescence-activated cell sorter (EPICS C, Coulter Corporation, Hialeah, Fla.). Background fluorescence (determined by control mAb staining) was subtracted in each case. The following mAbs were used: OKT3, OKT4, and OKT8 (Ortho, Raritan, N. J.), anti-Leul lb (CD16), anti-Leul9 (CD56), anti-IL2 receptor (R) (p55) (CD25) (Becton Dickinson, Calif., USA), anti-y/~ T cell receptor (TcR) (BB3 and A13) [7, 31]. Cytotoxic assay. Target cells were incubated with 100 gCi 51Cr (NEN, Florence, Italy) for 1 h at 37°C and then washed twice with RPMI-1640 medium. A total volume of 200 gl complete medium containing 103 target cells and various numbers of effector cells at final effector: target (E:T) ratios of 100, 50, 25 and 12:1 was placed in round-bottomed microtiter plates and incubated at 37 ° C. After 4 h an aliquot (100 gl) of supernatants was collected and counted in a 7 scintillation counter. All experiments were performed in triplicate and the percentage of 51Cr-release was calculated according to the following formula: (E-S)/(M-S) x 100

where E is the mean 51Crrelease (cpm) in the presence of effector cells, S is the mean 51Cr spontaneously released by the target cells incubated with medium alone, and M is the mean maximum release obtained by adding 100 ~tl of 1% NP-40 detergent (BDH, Poole, UK) to the wells. The results always refer to a final E : T ratio of 100: 1. Only results of experiments in which the value of S/M release is less than 30% are included. Proliferation. Lymphocytes were seeded in 9O-well plates (105 cells/well) and incubated in complete medium in the presence of 1 gCi/well [3H]dT (NEN, Florence, Italy) for 6 h at 37 ° C in a humidified CO2 atmosphere. Cells were subsequently harvested and radioactivity counted in a Ô counter.

Solid-phase activation with anti-CD3 mAb. Solid-phase activation was performed as described by Schoof et al. [25]. Flasks were precoated with OKT3 mAb at 5 - 10 txg/ml in borate-buffered saline, pH 8.5 for 3 h at room temperature. They were washed with PBS and left in complete

243 Table 1. Characteristics of tumour samples and the growth of tumour-infiltrating lymphocytes (TIL) Case

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Tumour a histotype Neuroblastoma Ovarian teratoca. Neuroblastoma Wilms' tumour Burkitt's lymphoma Osteosarcoma Rhabdomyosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Uracus carcinoma Wilms' tumour Ovarian teratoca. Teratocarcinoma Wilms' tumour Rhabdomyosarcoma Osteosarcoma Osteosarcoma Colon carcinoma Wilms' tumour Osteosarcoma Osteosarcoma Osteosarcoma Adrenal carcinoma Osteosarcoma Neurofibrosarcoma Neuroblastoma Lymphangiosarcoma Sinovial sarcoma Hepatoblastoma Wilms' tumour Neuroblastoma Osteosarcoma Melanoma Rhabdomyosarcoma Neuroblastoma Melanoma

Stage

P M P P M M M M P M P P P P P M M P P M M M P P M P P P P P P P M M M M P

Prior therapyb CT S+CT Noneh None S+CT S+CT S+CT S+CT+RT CT CT None CT S+CT S+CT CT S+CT S+CT CT None S+CT CT+RT S+CT+RT S+CT CT S+CT None None S+CT S+CT S+CT None None S+CT S S+CT+RT S+CT+RT S

Weight (g) 5 7 12 8 2.1 ml i 2 1.7 12 3 10 6 2 5.3 3 7 7 3 37 9 7 270 mli 3 7 ND 15 7 5 12 10 1 12 2 9 1 3 1.5 0.4

Lymphocytes (%) CL c

TS d

NDg 5 5 1 ND 20 10 29 12 10 95 12 8 12 30 10 9 15 30 5 22 30 18 1.5 25 5 6 19 9 6 26 4 4 76 9 4 40

4 2.5 4 2.5 ND 3 2 13 2 3 12 3 2.5 3 2.5 2 2.5 3 1 2.5 14 16 12 1 13 2 2 8 2 2 12 1 2 LNJ 2 1 11

TIL growthe

Time of culture (days)f

4.2 0 0 0 6 0 8.7 205 0 1.6 100 3 0 0 0 0 1.8 0 0 3.9 40 172 25 0 346 0 0 13.5 0 4 73 0 0 90 0 0 0

42 36 20 45 33 29 42 40 28 48 22 66 76 52 40 54 43 34 63 101 19 56 37 39 103 54 28 65 61 61 81 62 57 90 42 33 22

Diagnosis of tumour specimens from which TIL were obtained. R primary lesion; M, metastasis b S, surgery; CT, chemotherapy; RT, radiotherapy c CL, percentage of lymphocytes in cell suspension, as evaluated by cytological techniques d TS, percentage of lymphocyte infiltrate in tumour specimen as assessed by histological examination

e TIL growth is expressed as fold increase from original lymphocyte concentration f Last day of expansion before culture became non-viable g ND, not done h None, patient was submitted to surgical resection of tumour from which TIL cultures were performed i Cells were obtained from pleural effusion J LN, metastatic lymph node

medium for at least 1 h at room temperature before use. For activation, lymphocytes were resuspended in RPMI-1640 medium with 100 U/ml IL-2, and dispensed onto the coated surface and incubated at 37 ° C. After 72 h the cells were recovered, the surface was washed with PBS and the cells were counted and plated in culture with 100 U/ml IL-2. LAK from healthy donors, after culture with IL-2 for 30 days, were also included as positive control.

developed with aminoethylcarbazole (Sigma Co.). mAb against the following markers were employed: CD3 (Ortho, Raritan, N. J., USA), MHC class I (W6.32, Sera-Lab, Crawley Down, UK), MHC class II (HLADR) (Dl.12, kindly provided by Professor R. Accolla, Verona, ltaly), ICAM1 (CD54-RR1/1), LFA3 (CD58-TS9/A, both kindly provided by Dr. T. Springer, Boston), TGFI3 1 (British Bio-Technology Ltd., Cowley, Oxford, UK).

Immunohistochemistry. Labelling by mAb was carried out on air-dried

MLTC. Mixed lymphocyte/tumour cell culture was performed for cases

acetone-fixed (10 min, BDH Poole, UK) slices obtained from cryopreserved tumour specimens; slices were then incubated overnight with the first mAb and with a non-immune mouse serum for negative control. After two 10-min washes in PBS (pH 7.4), a biotin-conjugated horse anti-(mouse Ig) antibody (dilution 1:200; Vector, Burlingame, Calif., USA) was applied for 45 min, washed twice and labelled with modified avidin-biotin-peroxidasecomplexe (ABC) (Vector) [4]. Slices were then

11, 17, 28, 31. From day 21 of culture, TIL were cultured with 1000 U/ml IL-2, in the presence or absence of 20% irradiated (150 Gys) autologous mmour (Auto-Tu) cells with and without TNFc~ (250 U/ml). TIL were restimulated every 7 - 1 4 days of coculture. At the end of each culture week, lymphocytes were counted or tested for cytotoxic activity, for spontaneous proliferation and for immunophenotype (case 31), as described before.

a

244

A

B

500C

70C

60q

60C

50C

50C

40C

40C

30O

300

2oc

20C

10(;

10C

co

0 'Ir"-

X CD

ù.Q

B

10

20

30

40

50

60

E

70

80

)

10

C

t--

20

30

40

50

60

70

80

Culture time (days)

700

O (.~

600

500 400

300 200 100

0 Culture time (days)

Fig. 1 A - C. Expansion of tumour-infiltrating lymphocytes (TIL) from paediatric mmours during culture with interleukin-2 (IL-2). A TIL from osteosarcoma (8 representative cases); *, TIL were injected in the patient. B TIL from Wilms' mmour (3 cases) and from neuroblastoma (3 cases). C TIL from paediatric melanoma (2 cases) and carcinoma (1 uracus carcinoma, 1 adrenal carcinoma and 2 ovarian teratocarcinomas)

Results

Tumour specimens and TIL culture A group of 37 paediatric mmours (including osteosarcoma, soft-tissue sarcoma, Wilms' tumour, neuroblastoma, carcinoma, melanoma and others) were processed to isolate TIL for clinical application. The characteristics of patients and tumour lesions are presented in Table 1. Tissues were obtained from primary tumours in 21 patients and from metastatic deposits in 16. Ten patients did not receive therapy other than surgery, while 27 were treated with chemo- or radiotherapy. The mean wet weight of specimens processed was 7 + 6 g (mean + SD, range 0.4-37), while the mean number of cells recovered was 14 + 22 x 106/g tis sue (range 0.1 x 106 - 96.6 x 106/g). The greatest difficulty in obtaining sufficient amounts of tissue for the different assays was main]y due to the small size of mmour lesions. The percentage of lymphocytes ranged between 1% and 95% (mean 24 + 22%) in the cell suspensions and between 1% and 16% (mean 4.8+4.6%) in the

tumour tissue, as assessed by histological analysis. Since in the latter analysis we evaluated the intratumour lymphocyte infiltrate, the discrepancy between the percentage of lymphocyte detected in the cell suspension and in the tumour tissue may be due to the presence of peritumour lymphocytes in the former. Freshly isolated TIL were cultured for the first 2 weeks with low doses of IL-2 (20 U/ml) to select for "tumourspecific" lymphocytes (i. e., lymphocytes with high-affinity IL-2 receptor) potentially present in the neoplastic lesion. Subsequently, cultures were expanded with a high doses of IL-2 (1000 U/ml). With this schedule of in vitro activation, TIL were grown with more than 10-fold expansion in only 9 of 37 cultures (24%) (Table 1) obtained from different patients (mean expansion: 58-fold, range 13.5-346). In 17 cases no viable cells were obtained after culture with 1L-2. In the successfully expanded cultures, the gradual disappearance of tumour cells and the progressive increase in lymphocytes was usually seen after the 3rd week of culture with IL-2.

245 Table 2. Phenotypic analysis of TIL from paediatric tumoursa Tumour histotype

Percentage of cells positive for

Case

CD4/CD8 ratio

CD3

CD4

CD8

CD56

CD16

CD25

8 10 17 21 22 23 25

70 94 53 62 80 87 90

51 33 36 53 8 1 8

28 76 20 30 13 62 24

13 26 34 26 24 73 12

12 8 21 23 9 5 14

36 15 12 32 41 30 58

Wilms' tumour

12 20 31

69 51 88

18 25 56

44 11 28

73 22 20

20 31 16

2 39 41

0.49 2.27 2

Neuroblastoma

3 27

75 78

33 43

22 35

27 15

10 26

6 11

1.65 1.22

5

79

8

73

29

24

11

0.10

Ovarian teratocarcinoma

2 13

70 62

33 50

20 19

23 82

29 9

27 16

1.65 2.6

Osteosarcoma

Burkitt' s lymphoma

1.8 0.43 1.8 1.76

~/]~b 0.20

~t[~b

Uracus carcinoma

11

50

18

42

45

55

31

0.42

Rhabdomyosarcoma

7 16

77 41

25 8

65 35

51 92

5 12

18 5

0.38 0.22

Lymphangiosarcoma

28

92

12

79

85

20

25

0.15

Melanoma

34

46

31

14

12

21

49

2.2

a Phenotypic analysis was performed after 20 days of in vitro culture with interleukin-2 (IL-2) b These lymphocytes were positive (52% for case 23 and 61% for case 25) for 7/8 T cell receptor, as assessed by A.13 and BB3 mAbs

Table 3. Cytotoxic activity of TIL from paediatric mmours Tumour histotype

Case

Lysis on the target a (%) Daudi

K562

Allo-Tub

Auto-Tuc

Osteosarcoma

8 10 17 21 22 23 25

13 52 70 85 44 18 68

62 50 77 79 73 30 56

29 37 27 31 37 8 33

23 50 ND 18 ND ND ND

Wilms' tumour

12 15 20 31

33 33 10 71

81 58 34 79

29 22 1 68

51 ND ND 22

Neuroblastoma

3 27

40 75

71 65

11 29

5 ND

5

34

58

19

15

Ovarian teratocarcinoma

2 13

59 3

41 3

8 12

35 ND

Rhabdomyosarcoma Lymphangiosarcoma

7 16 28

4 24 76

20 76 76

11 54 71

ND ND ND

Melanoma

34

93

78

100

ND

Burkitt' s lymphoma

a Data refer to an effector: target ratio of 50 : 1 b Allo-Tu, allogeneic tumour cells, including fresh or cultured cells of the same or different histotype c Auto-Tu, autologous turnour cells (cultured or cryopreserved)

246 lO0

80

co 6 0 o

4o 0 cL ö~ 2O

0

] 2

[ 4

I 6

Culture time (weeks)

Fig. 2. Phenotypicvariationsof TIL duringculturewithIL-2. Expression of lymphocytemarkers was assessed by indirect immunofluorescence. 0, CD3; O, TcR y/5 (evaluatedin the 2 cases by A.13 and BB3 mAh); *, CD25; D, CD8; ZX,CD4; A, CD16; mm,CD56. Valuesrepresentthe mean of seven different cultures (standard deviations never exceeded 8%) The degree of TIL expansion did not correlate with tumour histology, site of lesion (TIL were grown in 4/21 primary and 5/16 metastatic tumours) or with patient treatment before surgery (TIL grew in 3/10 patients who did not receive any systemic therapy and in 5/16 patients who were treated with chemo- or radiotherapy). A higher frequency of TIL outgrowth was observed in osteosarcomas (7 out 11 cases) in comparison with other tumour histotypes, e.g., neuroblastoma (1 out 5), but these differences did not reach statistical significance. On the other hand, the TIL ability to grow or their expansion in culture did appear to correlate with the percentage of lymphocytes in tumour tissue or in the single-cell suspension. In fact, a more than 10-fold TIL expansion was observed in tumours with a lymphocyte infiltrate >10% (Table 1). In most cases analysed the pattern of in vitro growth was similar (Fig. 1 A - C ) . TIL culture was characterized by a gradual increase in the number of lymphocytes, which peaked after 3 - 4 weeks of expansion. Then the growth rate of TIL sharply decreased and lymphocytes became refractory to any proliferative stimulus (see below). However, cells were expanded until growth ceased and viability decreased, for a mean culture time of 49 ± 21 days (range 19-103). An exception to this typical pattern of in vitro growth was represented by two cases of osteosarcoma (cases 22 and 23), characterized by a long-term expansion.

Immunophenotype analysis In 20 out of 37 cases TILwere available for the analysis of immunological markers. Table 2 provides a general account of TIL immunophenotype by the 20th culture day. The majority of cultures consisted mainly of CD3 + cells (mean 7 0 + 16%, range 41-94%). CD4/CD8 ratio was greater than 1 in 10 cases and less than 1 in the remaining cases, without a clear correlation with tumour histology, site of lesion, treatment before surgery or ability of TIL to expand. In the 2 cases of osteosarcoma (nos. 22 and 23), in which the typical decrease of the in vitro growth was not

observed, a predominance of CD3 + lymphocytes expressing y/8 TcR was found. A minority of CD 16+ lymphocytes was detected in TIL culture (mean 18+ 11%, range 5-55%), while the LAK marker CD56 was significantly expressed in the early phase of in vitro expansion. The proportion of lymphocytes positive for the ~ chain of IL-2 receptor (p55) (CD25) varied in the different cases. However, a direct correlation between CD25 expression and lymphocyte expansion was observed at this time of activation with IL-2, since all cultures endowed with efficient growth were characterized by more than 30% of CD25+ cells. In 7 cases the TIL phenotype was monitored during culture (Fig. 2). In spite of the marked heterogeneity among the different cases, an increase in CD3 + and CD8 + cells was observed, in parallel with a disappearance of NK/LAK marker expression (i. e. CD16 and CD56). CD25 + cells progressively decreased during culture, concomitantly with a reduction of TIL growth rate. An increase of lymphocytes positive for y/8 TcR was also observed in cases 22 and 23.

Cytotoxic activity of expanded TIL in IL-2 Sequential 4-h 51Cr-release assays of cytotoxic activity of TIL cultures were carried out in 20 cases. Results are presented in Table 3; data refer to the 3rd week of culture. All tests were performed within 24 h of assessing the lymphocyte phenotype. The experiments included the cultured targets Daudi and K562 to evaluate LAK and natural killer (NK) activity, one or more allogeneic targets of the same or different histology (allogeneic tumour, Allo-Tu) and cultured or cryopreserved Auto-Tu, the latter being available in 8/20 cases. Each experiment also included lysis by 3-day LAK cells as a control of Auto-Tu lysability. The lytic pattern of TIL against Allo/Auto-Tu targets was heterogeneous, regardless of the histological type, the site of tumour lesion, the expansion rate or the phenotype of TIL. More precisely, the susceptibility to TIL lysis decreased according to the following order: K562 (mean 58 + 22), Daudi (45 + 28), Allo-Tu (32 + 25) and Auto-Tu (27 ± 16). The lytic activity of TIL was rather weak against Auto-Tu cells, although the same tumour cells were efficiently lysed by allogeneic LAK cells (data not shown); however, an Anto-Tu-specific or preferential pattern of lysis by TIL was almost never observed. Moreover, no correlation between lytic activity and TIL phenotype, including the CD4/CD8 ratio, was found. The majority of TIL lost their cytotoxic ability against all targets in the final stages of in vitro expansion, i.e. after a culture time of 4 - 6 weeks (data not shown).

Effects of TNF, IL-4, IFNy, anti-CD3 mAb or stimulation with Auto-Tu on TIL expansion The majority of TIL cultures stopped growing after 30-40 days and a sufficient number of TIL for in vivo treatment could be rarely recovered after culture with IL-2 alone. The effect of combination of IL-2 with different cytokines on TIL expansion has been, therefore, evaluated

247 80

A

OKT3 70

100

A 6O O ,¢-. 5O cD

-

80ù~.~, ~

,

~

~mnOr 60-

40

~i,

o° ao

r" CL E 20 >~

, 0

10

20

~/~

40-

=u

20-

, ~~,~~ ~'~ 30

40

0

,

a b

b

b

50

Gulture time (days) Fig. 3. Effect of stimulation with OKT3 in solid phase on TIL outgrowth in vitro. TIL were stimulated for 3 days with OKT3 in solid phase and then cultured with IL-2 (see Materials and methods) and the growth of these cells was compared to that of TIL cultured with IL-2 alone. - - , OKT3 plus 1L-2; . . . . , IL-2 alone. Lymphokine-activated killer cells (LAK) from a healthy donor, cultured for 30 days with IL-2, have been included as positive control

~.. 100 O~ 80

,o~°

......................................................................................................... .........................................

i.....

2O

50

0 o o

x

57 Culture time (days)

40 3~

64

75

F i g . 5 A , B. C y t o t o x i c i t y o f T I L f r o m W i l m s ' t u m o u r ( c a s e 3 1 ) d u r i n g

E 2G o. O 11~ C 30

33

47

57

64

75

Culture time (days)

Fig. 4. Proliferation of TIL from Wilms' tumour (case 31) during culture in the presence of different stimuli. Different culture conditions were analysed: ((romfront to rear) black columns, IL-2 (1000 U/ml); lightly hatched columns, IL-2 + autologous irradiated tumour cells; grey columns, IL-2 + tumour necrosis factor « (TNFo~;250 U/tal); strongly hatched columns, IL-2 + TNFa + Auto-Tu cells

by the addition of TNFc~ (250 U/ml), IL-4 (80 U/ml) or IFNy (100 U/ml) to IL-2 culture. There was no significant difference in growth among cultures supplemented with different cytokines, including IL-2 + TNFc~, IL-2 + IL-4 or IL-2 + IFNy (data not shown). Moreover, the addition of these cytokines to IL-2 did not modify TIL cytotoxicity against Auto-Tu or Allo-Tu targets (data not shown). In one case only (Wilms' tumour, case 31) the presence of TNF affected TIL proliferation, phenotype and cytotoxicity (see below). Since anti-CD3 mAb acts as a mitogenic stimulus for peripheral blood lymphocytes (PBL) [25], we tested whether the effect of anti-CD3 mAb in solid phase would additionally increase the recovery of TIL from culture with IL-2. Figure 3 shows that the presence of anti-CD3 m A b did not increase the cell number during a culture time of

culmre in the presence of different stimuli. A a, IL-2 (1000 U/ml); b, IL-2 + irradiated Auto-Tu cells. B c, IL-2 + TNFce (250 U/tal); d, IL2 + TNFc~+ irradiated Auto-Tu cells. Targets (fromfront to rear): black columns, Auto-Tu cells; hatched columns, Daudi; white columns, K562; cross-hatched columns, allogeneic tumour cells from solid tumour. *, After day 47 of culture no viable cells were available for functional studies

2 - 3 weeks, although the percentage of CD3 + cells in TIL examined was higher than 80%. The same schedule of activation, however, generated a significant increase of proliferation in PBL from healthy donors (Fig. 3). In four TIL cultures (cases 11, 17, 28 and 31), in which Auto-Tu cells were available in sufficient number, restimulation with irradiated Auto-Tu cells was performed every 14 days. Surprisingly, tumour refeeding did not allow the maintenance of proliferation and Auto-Tu lytic activity, with the exception of case 31 (see below). Thus, combinations of IL-2 with different cytokines or stimulation with anti-CD3 m A b or Auto-Tu cells were unable, in the majority of cases, to promote and sustain an efficient expansion of TIL from either primary or metastatic tumours in paediatric cases.

Analysis o f TIL from Wilms' tumour (case 31) TIL with preferential cytotoxicity against Auto-Tu cells could be derived from a Wilms' tumour (case 31). Since the patient was not a candidate for immunotherapy, differ-

248 Table 4. Phenotype of TIL flora Wilms' tumour after culture with different stimuli Culture time (days)

Percentage of cells positive for

Stimuli

CD3

CD4

CD8

CD16

CD25

CD56

30

IL-2 a IL-2a + TNF c

53 61

9 5

41 52

41 40

9 26

54 55

33

IL-2 a IL-2" + TNF

88 87

56 26

28 50

16 22

41 47

20 46

47

IL_2a, b + TNF

89

65

16

8

61

21

57

IL-2 + TNF

73

57

14

15

36

18

64

IL-2 + TNF

78

77

6

7

31

15

75

IL-2 + TNF

59

64

10

15

7

30

IL-2 was used at 1000 U/tal b After day 47 of culture no cells culmred with IL-2 alone were available for functional studies c TNE tumour necrosis factor a

Table 5. Expression of immunomodulatory molecules on paediatric tumour cells a

Case

Tumour histotype

TIL growth

6 7 11 13 14 15 19 22 24 25 28 30 31

Osteosarcoma Rhabdomyosarcoma Uracus ca. Ovarian teratoca. Teratoma Wilms' tumour Colon ca. Osteosarcoma Adrenal ca. Osteosarcoma Lymphangiosa. Hepatoblastoma Wilms' tumour

. +/_b . +c +c +/_b . +c

HLA I

HLA I1

.

+/-« . + + +

. + -

.

.

.

+ + + .

LFA 3 .

.

TGF~

CD3

-

+

-

+ + + + + _ + _ + + +

_ +/_ +/_ _ +/_

.

.

_ -

+ -

+ + _

+ + _

+

+

. + + + _ _

+

ICAM 1

.

. +

The expression of the different markers was evaluated by immunohistochemistry. The results were scored as negative when fewer than 5% of the cells were stained, and as positive when more than 15% cells were stained

b Lymphocyte expansion has been observed, hut their number decreased after 2 - 3 months of culture ° Exponential growth has been observed d +/_, percentage positive cells between 5% and 15%

ent conditions of in vitro culture, including activation with I L - 2 + TNFo~, I F N 7 a n d a u t o l o g o u s i r r a d i a t e d t u m o u r cells, c o u l d b e t e s t e d to d e t e r m i n e t h e g r o w t h p o t e n t i a l , phenotype and cytotoxicity of TIL generated by different schedules of activation. The proliferative activity and the duration of in vitro g r o w t h o f T I L c u l t u r e d w i t h I L - 2 + TNFü~ s i g n i f i c a n t l y exceeded that of TIL cultured with IL-2 alone, while the addition of Auto-Tu cells did not significantly affect the g r o w t h r a t e (Fig. 4). As for the cytotoxicity, the stimulation with Auto-Tu produced a moderate and transient enhancement of lytic a c t i v i t y (Fig. 5 A ) , w h i l e t h e c o m b i n a t i o n o f I L - 2 + T N F induced a borderline increase of cytotoxicity of TIL against A u t o - T u c e l l s at 3 0 d a y s o f c u l t u r e a n d n o s i g n i f i c a n t i n c r e m e n t o f n o n - M H C r e s t r i c t e d a c t i v i t y (Fig. 5 B). H o w e v e r , t h e a d d i t i o n o f T N F w a s n o t s u f f i c i e n t to p r o v i d e h i g h levels of cytotoxicity in the long-term TIL culture. No decrease of cytotoxic activity was observed after 30-40 days in TIL activated with IL-2+TNF, but not with

IL-2 alone, when irradiated Auto-Tu cells were added. Of n o t e is t h a t T I L s t i m u l a t e d b y t h i s s c h e d u l e s h o w e d a preferential lysis of Auto-Tu targets in long-term culture (Fig. 5 B), s u g g e s t i n g t h e p r e s e n c e o f T - s p e c i f i c l y m p h o cytes. Thus, Auto-Tu cells refeeding increased the longterm cytotoxicity against Auto-Tu cells both in IL-2 and IL-2 + TNF TIL cultures. Analysis of the TIL phenotype of this particular case indicated that cultures with IL-2 or IL-2 + TNF contained a s i m i l a r p e r c e n t a g e o f C D 3 ÷ c e l l s ( T a b l e 4). H o w e v e r , t h e r e was a significant shift in the CD4/CD8 ratio, with a relevant increase of CD8 ÷ cells in TIL activated with I L - 2 + T N F as c o m p a r e d w i t h T I L c u l t u r e d w i t h I L - 2 alone. In addition, the percentage of CD56 ÷ and CD16 ÷ cells was slightly enhanced by TNF whereas that of CD25 (TAC) resulted markedly higher. By contrast, the presence of Auto-Tu cells did not change the TIL phenotype (data not shown). Thus the addition of both TNF and irradiated Auto-Tu c e l l s to T I L c u l t u r e d w i t h I L - 2 w a s n e c e s s a r y to i n d u c e a n d

a

249 Table 6. Association between tumour phenotype and in vitro TIL growth Tumour molecule a

Expressionb

TIL growth c

cyte infiltration and the ability of TIL to be expanded by in vitro culture with IL-2.

4HLA I H L A II LFA 3 ICAM1 TGF[~

+ (5)

4

1

-

(8)

1

7

+ (8)

4

4

-

(5)

2

3

+ (3) -(10)

3 2

0 9

+ (4)

4

0

-

(9)

1

8

+ (8)

0

8

-

3

2

(5)

a Expression of different molecules on tumour cells was assessed by immunohistochemistry b In parentheses the number of cases expressing or not expressing the given molecule ° Number of cases from which TIL could be grown

sustain the growth and lytic activity of T lymphocytes reactive against Auto-Tu cells.

Expression of MHC, adhesion molecules and TGF~ on tumour cells and correlation with TIL expansion The reduced rate of TIL growth obtained in paediatric tumours prompted us to investigate the presence of molecules involved in T-lymphocyte activation or suppression on tumour cells. With this aim, the expression of MHC and adhesion molecules or the production of TGF~ was analysed on 13 cryopreserved tumour samples by immunohistochemistry. The results are reported in Table 5. This analysis first confirmed the paucity of CD3+ cell infiltrates, which was barely detectable in 3 cases only. Moreover in these tumours, whatever their histology, MHC class I and ICAM1/LFA3 adhesion molecules were significantly expressed only on a minority of tumour lesions (38%, 30% and 23%, respectively), while MHC class II was present on 77% of samples. On the other hand, in 10 out of 13 specimens analysed a significant reactivity for anti-TGF~ mAb was observed in the cytoplasm of neoplastic cells. Despite the limitations of a qualitative analysis (i. e. immunohistochemistry) in assessing tumour phenotype, these phenotype patterns were then examined in the light of the ability of TIL to grow in vitro (Table 6). All 5 cases with successful TIL expansion were characterized by a marked expression of MHC and adhesion molecules and the lack of detectable TGFI3. Conversely, reduced amounts of MHC, no expression of ICAM1/LFA3 molecules and a marked positivity for TGFI3 were observed in tumours with negative TIL outgrowth. These data suggest that the pattern of tumour phenotype, including the expression of molecules required for lymphocyte recognition or the production of potentially suppressive factors, may influence the presence of lympho-

Discussion

The lymphocytes present in a tumour lesion could represent a tumour-specific response of T lymphocytes, as well as a component of a non-specific infiltrate due to inflammarion, necrosis or bleeding of the neoplastic tissue. Thus, the role of TIL in controlling the growth and metastatic spread of cancer is still to be defined and can be established only by a precise functional characterization of their tumour specificity at cellular and molecular level. Although isolation and characterization of TIL has been successfully performed in human solid tumours including melanoma [12] and head and neck [10], lung [15], ovarian and renal cancer [3, 6], to our knowledge no data about TIL from childhood malignancies are available in the literature with the exception of a report from Rosenberg's group, who included three cases of paediatric tumours (neuroblastoma and Wilms' tumour) while analysing TIL from nonrenal urological tumours [9]. Our overall success rate in expanding TIL from childhood malignancies to a minimum 10-fold increase was only 24%. Such impaired proliferation was accompanied by a refractoriness to further activation stimuli, including different cytokines, anti-CD3 mAb or Auto-Tu cells. Howerer, we can not rule out the possibility that different factors, including the small amounts of tumour tissue for obtaining TIL, the heavy pretreatment of the majority of patients and the limited availability of Auto-Tu cells for TIL refeeding might have influenced TIL expansion. This behaviour was noted in the majority of TIL cultures analysed, with the exception of two TIL populations deriving from osteosarcomas, in which a high percentage of cells positive for q(/8 TcR was detected. Since Fish et al. [8] have recently shown that y/STcR+ lymphocytes may recognize a homologous heat-shock protein on human tumours, one can speculate that these proteins may be expressed on a paediatric osteosarcomas and implicated in 3,/8 lymphocyte activation. The combination of IL-2 with various cytokines, e.g. IL-2 + TNF [32] or IL-2 + IL-4 [16] or IL-2 + IFNy [26] has been recently used to obtain TIL cultures enriched in Auto-Tu-reactive T lymphocytes in patients with solid tumours. Moreover, TNF(x has been used to overcome a possible inhibitory effect on TIL proliferation mediated by TGFI3 [21], since the latter is known to suppress the generation of cytolytic T lymphocytes from TIL and PBL [19], and may be released by tumour cells along with other, less defined factors [28]. Differences in responsiveness to TNF + IL-2 were recently reported between TIL of primary and metastatic liver tumours, in which the outgrowth of TIL cultured in IL-2 + TNFc~ led to an in vitro enrichment in cells with Auto-Tu cytolysis and increased proliferative rate in lymphocytes from primary but not metastatic lesions [26]. At variance with these data, out results show that combinations of IL-2 and other cytokines, added during the phase of decreasing outgrowth, were unable to sustain, in all but orte cases (case 31), the growth and

250 expansion of TIL from either primary or metastatic tumours paediatric patients. Recently Schoof et al. [25] reported on the application of solid-phase anti-CD3 mAb to the induction of growth and cytolytic potential of TIL from a variety of solid tumours. However, our experience with OKT3 activation of TIL from childhood malignancies does not indicate any significant advantage in lymphocyte proliferation for OKT3-stimulated lymphocytes. Restimulation of TIL cultures with Auto-Tu cells has been shown to increase proliferation and Auto-Tu lytic activity [27]. The outgrowth of TIL from paediatric tumours refed with Auto-Tu cells at regular intervals but at time when cell proliferation declined, did not appear to exceed the growth of non-refed cultures. Tumour refeeding, even after the onset of the culture, allowed maintenance of proliferation and Auto-Tu cytotoxicity in one case (case 31) and this effect was markedly potentiated by the addition of TNFc~. We cannot exclude, therefore, the possibility that the reduced expansion of TIL from paediatric patients might be modified throughout restimulation with Auto-Tu cells during the entire period of culture. Altogether our data lead us to conclude that TIL from paediatric patients may rarely be isolated and expanded in IL-2 alone (i. e. without autologous tumour cells) to levels required for an in vivo transfer. The analysis of the TIL phenotype showed a heterogeneity of lymphocyte marker expression in different cultures, without a specific pattern for each type of tumour. By contrast, Balch et al. [2] showed that specific patterns of TIL subpopulations could be found in distinct tumour histotypes, including melanoma, renal cancer, sarcoma and colon carcinoma, in which the predominance of T cytotoxic, T helper lymphocytes or NK cells was similar in tumours of the same lineage. We noted that two cases of osteosarcoma (cases 22 and 23) displayed a high percentage of lymphocytes positive for y/8 TcR and characterized by a marked and prolonged in vitro expansion, although the relationship between this TcR and the proliferative activity needs further examination. TIL from childhood malignancies, exhibited a high natural killer (NK) and LAK activity, while the levels of cytotoxicity against cryopreserved or cultured Auto-Tu cells were generally low, although out data are limited to 8 out of 37 cases, because of the low number of viable neoplastic cells available for functional assays. However, no TIL cultures showed Auto-Tu-specific or preferential lysis, with the exception of TIL from patient 31, in which a marked increase of lytic activity against autologous targets was observed when TNF~ + irradiated Auto-Tu cells were added early during IL-2 culture. We did not observe either a consistent relationship between a particular phenotype and the pattern of lytic activity, or a correlation between percentage of CD8 + lymphocytes and cytolytic potential. These data contrast partially with the observations of Topalian et al. [30], who found an association between cultures predominantly containing CD8 + lymphocytes and high lytic activity. In addition, despite the early decrease of CD16 + cells, a parallel decrease of NK or LAK activity was not seen, suggesting that T lymphocytes are responsible for the MHC-unrestricted activity observed.

Since specific T lymphocytes may recognize a putative tumour antigen in association with class I or II MHC molecules, modifications of the expression of these structures on tumour cells could influence the generation and the lytic activity of T cells [34]. Moreover, the high expression on target cells of adhesion molecules, including ICAM 1, LFA3 or VLA, enhances the susceptibility to lysis by autologous or allogeneic lymphocytic effectors [1, 33]. Conversely, tumour-derived immunoinhibitory factors, such as TGFI], may suppress the in vitro generation of cytolytic T cells from TIL of PBL [18], and could be responsible in vivo for lymphocyte suppression, since many human tumours can produce TGF[3 [28]. Our data indicated that paediatric tumours appeared to be characterized by a low expression of MHC class I and ICAM1/LFA3 adhesion molecules. These results are partially in agreement with data by Mechtersheimer et al. [16], who described a reduced or aberrant expression of MHC on osteosarcoma and soft-tissue sarcoma in the adults. A marked immunoreactivity for TGF[~ was found in the majority of paediatrie tumours analysed. It should also be noted that these tumours were characterized by a weak mononuclear cell infiltrate, with a relatively higher percentage of lymphocytes detected only in tumours with significant TIL expansion in vitro. On the basis of these phenotypic features, therefore, tumours from which TIL were significantly expanded were characterized by a marked expression of MHC class I and/or adhesion molecules, and/of by a lack of TGFI3. Since the expression of different markers was evaluated qualitatively by immunohistochemistry, these data cannot be emphasised and need a further evaluation. Moreover, no statistical analysis could be performed on the correlation between tumour phenotype and in vitro outgrowth of TIL. In conclusion, the results of the present report indicate that TIL from paediatric neoplasms may rarely be isolated and expanded at levels required for in vivo treatment and are unlikely to contain a biologically significant number of tumour-specific lymphocytes. A paucity of expression of immunomodulatory molecules on tumour cells could also be responsible of the reduced recognition of tumour cells by T lymphocytes.

Acknowledgements. We thank Miss Adua Marinoni and Mr. Edoardo Marchesi for the excellent technical help. This work was partially supported by EuroCetus (Amsterdam, NL) and by a grant from Istituto Superioredella Sanitä, Italy-USAProgramon New Therapiesof Tumors (Rome,Italy).

References 1. Anichini A, Mortafini R, Supino R, Parmiani G (1990) Human melanomacells with high susceptibilityto cell-mediatedlysis can be identified on the basis of ICAM-1 phenotype, VLA profile and invasive ability. Int J Cancer 46:508 2. Balch CM, Riley LB, Bae YJ, SalmeronMA, Platsoucas CD, von Eschenbach A, Itoh K (1990) Patterns of human tumor-infiltrating lymphocytesin 120 human cancers. Arch Surg 125:200 3. Belldegrun A, Muul LM, and Rosenberg SA (1988) Interleukin-2 expanded tumor-infiltrating lymphocytes in human renal cancer: isolation, charactefizationand antitumoractivity.CancerRes 48:206

251 4. Cattoretti G, Berti E, Schiro R, D'Amato L, Valeggio C, Rilke F (1988) Improved avidin-biotin-peroxidase complex (ABC) staining. Histochem J 20:75 5. Favrot MC, Floret D, Negrier S, Cochat P, Bouffet E, Dacheng Z, Franks CR, Bijman T, Brunat-Mentigny M, Philip I, Philip T (1989) Systemic interleukin-2 therapy in children with progressive neuroblastoma after high dose chemotherapy and bone marrow transplantation. Bone Marrow Transplant 4:499 6. Ferrini S, Biassoni R, Moretta A, Bruzzone M, Nicolin A, Moretta L (1985) Clonal analysis of T lymphocytes isolated from ovarian carcinoma ascitic fluid. Phenotypic and functional characterization of T cell clones capable of lysing autologous carcinoma cells. Int J Cancer 36:337 7. Ferrini S, Pigione I, Bottino C, Ciccone G, Tambussi S, Mannoliti S, Moretta L, Moretta A (1989) Monoclonal antibodies which react with the T cell receptor y/8 recognize different substes of CD3 + WT31 T lymphocytes. Eur J Immunol 19:57 8. Fisch P, Malkovsky M, Klein BS, Morrisey LW, Carper SW, Welch W J, Sondel PM (1990) Human V y 9/V 8 2T cells recognize a groEL homolog on Daudi Burkitt's lymphoma cells. Science 250:1269 9. Haas GP, Solomon D, Rosenberg SA (1990) Tumor-infiltrating lymphocytes from nonrenal urological malignancies. Cancer Immunol immunother 30:342 10. Heo DS, Whiteside TL, Johnson JT, Chen KN, Barnes EL, Herberman RB (1987) Long-term interleukin-2-dependent growth and cytotoxic activity of tumor-infiltrating lymphocytes frorn human squamous cell carcinoma of the head and neck. Cancer Res 47:6353 11. Heo DS, Whiteside TL, Kanbour A, Herberman RB (1988) Lymphocytes infiltrating human ovarian tumors: I. Role of Leul9(NKH1)positive recombinant IL-2 activated cultures of lymphocytes infiltrating human ovarian cancers. J Immunol 140:4042 12. Itoh K, Platsoucas CD, Balch CM (1988) Autologous tumor-specific cytotoxic T lymphocytes in the infiltrate of human metastatic melanomas: activation by interleukin-2 and autologous tumor cells, and involvement of the T cell receptor. J Exp Med 168:1419 13. Kawakami Y, Rosenberg SA, Lotze MT (1988) Interleukin-4 promotes the growth of tumor-infiltrating lymphocytes cytotoxic for human autologous melanoma. J Exp Med 168:2183 14. Kuppner MC, Hamon MF, deTribolet N (1988) Immunohistological and functional analysis of lymphoid infiltrates in human glioblastomas. Cancer Res 48:6926 15. Kurnick JT, Kradin RL, Blumberg R, Schneeberger EE, Boyle LA (1986) Functional characterization of T lymphocytes propagated from human lung carcinomas. Clin Immunol Immunopatho138:367 16. Mechtersheimer G, Staudter M, Majdic O, Dorken B, Moldenhauer G, Moller P (1990) Expression of MHC-A,B,C,I]2-microglobulin (~2m), MHC-DR,-DP,-DQ and of MHC-D-associated invariant chain (Ii) in soft-tissue tumors. Int J Cancer 46:813 17. Miescher S, Stoeck M, Qiao L, Barras L, Barrelet L, Von Fliedner V (1988) Preferential clonogenic deficit of CD8-positive T-lymphocytes infiltrating human solid tumors. Cancer Res 48: 6992-6998 18. Mulé J J, Schwarz SL, Roberts AB, Sporn MB, Rosenberg SA (1988) Transforming growth factor-[3 inhibits the in vitro generation of lymphokine-activated killer cells and cytotoxic T cells. Cancer Immunol hnmunother 26:95 19. Nasr S, McKolanis J, Pais R, Findley H, Hnath R, Waldrep K, Ragab AH (1989) A phase I study of interleukin-2 in children with cancer and evaluation of clinical and immunological status during therapy a Pediatric Oncology Group study. Cancer 64:783

20. Pais R, Igrim NB, Garcia ML, Abdel-Mageed A, McKolanis J, Igrim ME, Hnath RS, Ziegler K, Ragab AH (1990) Pharmacokinetics of recombinant interleukin-2 in children with malignancies: a Pediatric Oncology Group study. J Biol Response Mod 9:517 21. Ranges GE, Figari IS, Espevik T, Palladino MA (1987) Inhibition of cytotoxic T cell development by transforming growth factor-~ and reversal by recombinant tumor necrosis factor-c« J Exp Med 166:991 22. Rosenberg SA (1988) Immunotherapy of cancer using interleukin 2: current status and future prospects. Immunol Today 9:58 23. Rosenberg SA, Spiess P, Lafreniere R (1986) A new approach to the adoptive immunotherapy of cancer with tumor-infiltrating lymphocytes. Science 233:1318 24. Rosenberg SA, Packard BS, Aebersold PM, Solomon D, Topalian SL, Toy ST, Simon P, Lotze MT, Yang JC, Seipp CA, Simpson C, Carter C, Bock S, Schwartzentruber D, Wei J, White D (1988) Use of tumor infiltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma. N Engl J Med 319:1676 25. Schoof DD, Selleck CM, Massaro AF, Jung S-F, Eberlein TJ (1990) Activation of tumor infiltrating lymphocyte by monoclonal antibodies directed to the CD3 complex. Cancer Res 50:1138 26. Shimizu Y, Iwatsuki S, Herberman RB, Whiteside TL (1991) Effects of cytokines on in vitro growth of tumor-infiltrating lymphocytes obtained from human primary and metastatic liver tumors. Cancer Immunol Immunother 32:280 27. Skornick Y, Topalian S, Rosenberg SA (1990) Comparative studies of the long-term growth of lyrnphocytes from tumor infiltrates, tumor draining lymph nodes and peripheral blood by repeated in vitro stimulation with autologous tumor. J Biol Response Mod 9:431 28. Sporn MB, Roberts AB, Wakefield LM, Assoian RK (1986) Transforming growth factor-I]: biological function and chemical structure. Science 233:532 29. Takagi S, Chen K, Schwarz R, Iwatsuki S, Herberman RB, Whiteseide TL (1989) Functional and phenotypic analysis of tumor-infiltrating lymphocytes isolated from human primary and metastatic liver tumors and cultured in recombinant interleukin-2. Cancer 63:102 30. Topalian SL, Muul LM, Rosenberg SA (1987) Growth and immunological characteristics of lymphocytes infiltrating human tumors. Surg Forum 37:390 31. Triebel F, Faure F, Mami-Chouaib F, Jistukawa S, Griscelli AL, Genevee C, Roman-Roman S, Hercend T (1988) A novel human V/8 gene expressed predominantly in the Ti/y/A+ fraction of y/8+ peripheral lymphocytes 32. Vaccarello L, Wang YL, Whiteside TL (1990) Substained outgrowth of autotumor-reactive T lymphocytes from human solid tumors in the presence of tumor necrosis factor-c~ and interleukin-2. Hum Immunol 28:216 33. Vanky F, Wang P, Patarroyo M, Klein E (1990) Expression of the adhesion molecule ICAM1 and major histocompatibility complex class I antigens on human tumor cells is required for their interaction with autologous lymphocytes in vitro. Cancer Immunol Immunother 31:19 34. Weber JS, Rosenberg SA (1990) Effects of murine tumor class I major histocompatibility complex expression on anti-tumor activity of tumor-infiltrating lymphocytes. J Natl Cancer Inst 82:755