Increased non-major histocompatibility complex-restricted ... - Nature

© 2000 Nature America, Inc. 0929-1903/00/$15.00/⫹0 www.nature.com/cgt

Increased non-major histocompatibility complex-restricted lytic activity in melanoma patients vaccinated with cytokine gene-transfected autologous tumor cells Peter Mo ¨ller,1,2 Hjo ¨rdis Mo ¨ller,1–3 Yuansheng Sun,3 Tomislav Dorbic,1,4 Beate M. Henz,2 1,4 Burghardt Wittig, and Dirk Schadendorf3 1

Centrum Somatische Gentherapie e.V. and 4Abt. Molekularbiologie und Biomathematik, Freie Universita ¨t, Berlin, Germany; 2Universita ¨tshautklinik, Virchow Klinikum, Humboldt-Universita ¨t zu Berlin, Berlin, Germany; and 3Clinical Cooperation Unit for Dermatology (DKFZ), University of Heidelberg, Heidelberg, Germany. Genetically modified antitumoral vaccines focus on eliciting or increasing the T-cell-mediated antitumoral response. Little is known about non-major histocompatibility complex-restricted responses. In two phase I studies, we have immunized advanced melanoma patients with either interleukin-7 (IL-7) gene-transfected or IL-12 gene-transfected, autologous, irradiated melanoma cells. To monitor the immune response, peripheral blood mononuclear cells were collected before the first vaccination and 2 weeks after the third vaccination. Spontaneous lytic activity and lymphokine-activated killer (LAK) activity after a 5-day culture in the presence of 1000 U/mL IL-2 against autologous and against allogeneic melanoma cells were measured. In parallel, a precursor cytotoxic T-cell frequency analysis was performed using a 25-day limiting dilution analysis assay. A total of 10 of 14 immunologically evaluable patients demonstrated a marked increase in LAK activity, and 7 of 14 showed increased spontaneous lytic activities against autologous melanoma cells after three vaccinations. Remarkably, two patients with a good clinical performance status (Karnofsky index of ⬎70; Multitest Merieux of ⬎13.4 mm/3) and the highest cytotoxic T-lymphocyte (CTL) response after vaccination showed the only clear decrease in LAK and spontaneous lytic activity. Otherwise, three patients with no detectable CTL response after vaccination demonstrated an increase in LAK activity and the strongest increase in the autologous spontaneous lytic activity. This group of patients was associated with a poor clinical performance status (Karnofsky index of ⬍70; Multitest Merieux of ⬍4 mm/1) and with no clinical response. In conclusion, in accordance with other studies, a good clinical and immunological performance status appears to be the prerequisite for a successful CTL response. However, even strong non-major histocompatibility complex-restricted responses could be generated in patients with reduced clinical performance in vaccination therapies with gene-transfected autologous tumor cells. Cancer Gene Therapy (2000) 7, 976 –984

Key words: Melanoma; cytokine gene transfer; vaccination; non-major histocompatibility complex-restricted lysis.

M

elanoma demands early detection and elimination, particularly in light of its notable resistance to conventional therapies in advanced stages.1,2 However, melanoma is presumed to be one of the most immunogenic of tumors. This is emphasized by occasional cases of spontaneous remission of primary melanoma, apparently the result of an effective cytotoxic T-lymphocyte (CTL) response3 and the partly successful clinical administration of interleukin-2 (IL-2) and tumor-infiltrating lymphocytes.4 Recently, several clinical phase I trials have been initiated or brought to completion using selected major histocompatibility complex (MHC) class I-suitable peptides either alone or in association with Received November 17, 1999; accepted January 29, 2000. Address correspondence and reprint requests to Dr. Dirk Schadendorf, Clinical Cooperation Unit for Dermatooncology (DKFZ), Theodor Kutzer Ufer 1, 68135 Mannheim, Germany.

976

professional antigen (Ag)-presenting cells to generate a tumor-specific T-cell response in vivo. Partially effective immunological and clinical responses have been reported.5– 8 Another novel form of immunotherapy uses cancer vaccines consisting of tumor cells genetically engineered to secrete cytokines.9,10 Compared with high-dose cytokine-based therapies, treatment using such cancer vaccines secreting cytokines in a paracrine manner shows virtually no severe adverse effects. Some cytokines, such as IL-2, IL-7, and IL-12, are capable of generating MHC-restricted antitumoral T-cell responses, which can establish a tumor-specific, protective immune memory. In addition, IL-12 can activate type 1 T helper cell function.11 Moreover, all three cytokines can generate non-MHC-restricted responses, including LAK activity.11–13 In previous reports of IL-7 and IL-12 gene therapy phase I studies, we evaluated the feasibility, safety, toxicity, and the clinical response, together

Cancer Gene Therapy, Vol 7, No 7, 2000: pp 976 –984

¨ LLER, MO ¨ LLER, SUN, ET AL: NK ACTIVITY AND LAK ACTIVITY IN IL-7 GENE THERAPY AND IL-12 GENE THERAPY MO

977

Table 1. Characteristics, DTH-Reactivity and CTL Responses of Patients Receiving Melanoma Cells Transfected for IL-7 Expression (Patients 1–10) Receiving Melanoma Cells for IL-12 Expression (Patients 1b– 6b) Patient characteristics

No.

Age/ sex

1 2 3 4 5 6 7 8 9 10 1b 2b 3b 4b 5b 6b

51/F 63/F 52/M 29/M 83/M 26/M 33/F 48/F 52/F 52/M 37/M 28/M 58/M 49/F 47/M 46/M

DTH reactivity

Karnofsky index B*

Numbers of given cells in three vaccines A†

⬎70 ⬎70 ⬍70 ⬎70 ⬎70 ⬍70 ⬍70 ⬎70 ⬍70 ⬎70 ⬎70 ⬎70 ⬍70 ⬎70 ⬎70 ⬎70

3 ⫻ 10 6 ⫻ 106 5 ⫻ 106 6 ⫻ 106 2.3 ⫻ 107 5 ⫻ 105 3 ⫻ 106 1.5 ⫻ 107 2.2 ⫻ 106 1.1 ⫻ 107 1.0 ⫻ 206 2.1 ⫻ 106 2.8 ⫻ 106 6.1 ⫻ 106 8.4 ⫻ 106 1.2 ⫻ 106

Clinical response A

7

SD SD SD NA‡ SD NA PD MR PD MR SD MR PD PD SD SD

Recall Ags (Multitest Merieux) diameter (mm)/ numbers B A 13.5/3 21.5/3 2.5/1 31/4 4/1 0/0 0/0 18/3 3/1 13/2 14/4 14/3 0/0 4/1 24/3 2/1

11.5/3 16/4 6/2 NA 6/2 NA 0/0 3/1 0/0 8/3 11/3 11/3 0/0 8/1 20/3 26/4

CTL response

Autologous tumor lysates A 0 0 0 NA 0 NA 0 0 0 0 0 0 0 0 (5)¶ (5)¶

Numbers of positive LDA cultures§ measured after 25 days in vitro B A 13(4) 0 3(0) NA NA NA 0 4(3) 0 2(0) 0 0 0 3(2) 5(0) 5(3)

25(7) 0 2(0) NA NA NA 0 79(20) 0 10(3) 0 0 0 8(7) 61(20) 4(2)

*B, before vaccination. †A, 2 weeks after the third vaccination. ‡NA, not applicable. Patient 6 died before the second vaccination. 4 refused further vaccinations. §Positive LDA cultures ⫽ LDA cultures with lytic activity either against autologous melanoma cells only, or against autologous melanoma cells and K562 cells in one 96-well; parentheses indicate LDA cultures with lytic activity against autologous melanoma cells in one 96-well only. ¶, After five vaccinations.

with the CTL response, in patients with advanced melanoma.14,15 Enhanced CTL responses after vaccination could be determined in those patients with a good clinical and immunological performance status. Little has been reported concerning non-MHC-restricted lytic activity in gene therapy studies.16 Until presently, a correlation between MHC-restricted lytic activity and non-MHC-restricted lytic activity in gene therapy studies was not reported. The question of whether and how natural killer (NK) activity (innate immunity) is influenced by a CTL response (adaptive immunity) and vice versa remains to be evaluated. In this context, it is highly relevant to examine the activation of spontaneous lytic activity and the capability to induce LAK activity in correlation to the CTL response, and their clinical (Karnofsky index) and immunological (Multitest Merieux, Leimen, Germany) performance status with regard to the current vaccination therapy, employing cytokine gene-transfected tumor cells. MATERIALS AND METHODS

Patient selection and clinical protocol Advanced melanoma patients either refusing or refractory to available therapies were admitted into two phase I studies, with 10 patients for IL-7 gene therapy and 6 patients

Cancer Gene Therapy, Vol 7, No 7, 2000

for IL-12 gene therapy, respectively.14,15 All participants gave written and informed consent before admission into the study, as required by the Virchow Clinic Institutional Ethical Review Board and in accordance with the Declaration of Helsinki. Treatment was carried out at the Department of Dermatology at the Virchow Clinic in Berlin, Germany. Patient recruitment for IL-7 gene therapy commenced in December of 1994 until study closure in March of 1996; for IL-12 gene therapy, recruitment started in February of 1996 until closure in October of 1996. Patient characteristics are summarized in Table 1 and were published previously in greater detail, including MHC Ags and tumor Ags.14,15 According to the protocol published previously,17 patients were injected subcutaneously on days 1, 8, 15, and 36 with irradiated (100 Gy), autologous, genetransfected melanoma cells in phosphate-buffered saline (PBS) sterile solution and in close proximity to the regional lymph nodes of each extremity. The vaccine preparation was generally divided into equal aliquots of ⬃106 cells. On day ⫺1 or day ⫺2 and on day 29, blood samples were collected to determine immune responses in peripheral blood. The delayed-type hypersensitivity (DTH) skin test (Multitest Merieux) was performed on day ⫺2 and on day 34. Clinical response and toxicity were defined according to World Health Organization criteria. A minor response (MR) was defined as a 25–50% decrease in lesions of at least 1 month or as a ⬎50% decrease in lesions during ⬍1 month. Stable disease was defined as a ⬍25% change in size, with no new observable lesions developing over a 6-week period.

978


Preparation of melanoma cells and maintenance of cell cultures Tumor specimens were collected from patients with advanced melanoma, either during their diagnostic work-up or during palliative treatment. Solid tumor specimens from lymph nodes, cutis, or subcutis were placed in RPMI 1640 medium (Seromed, Berlin, Germany) immediately upon removal. Adjunct non-melanoma-containing tissue was removed as completely as possible by scalpel or scissors, and tumors were subsequently diced. After passing these pieces through a steel mesh with a pore size of 25 ␮m, the cells were washed twice and cultured in complete RPMI 1640 medium supplemented with 20% fetal calf sera (FCS) (Seromed). For vaccination, 2–10 passages were used. The histological and immunocytochemical examination of cytospin preparations of these cultured cells confirmed melanoma cells by a S-100 staining index of ⬎95%. Tumor Ags known to be expressed in melanoma cells and MHC types were evaluated and have been described previously.14,15 The allogeneic melanoma cell line SK Mel-37 was obtained previously from a metastatic lesion of a patient with malignant melanoma at the Memorial Sloan Kettering Cancer Center (New York, NY) and was characterized for its differentiation markers18 and its LAK susceptibility.19

Ballistomagnetic vector system for transfer of expression plasmids into melanoma cells and preparation of the vaccines The full-length human IL-7 cDNA was recombined with a eukaryotic expression plasmid, pRc/CMV (Invitrogen, Heidelberg, Germany), and completely sequenced before use as described previously.17 The two cDNAs coding two subunits of bioactive human IL-12 (kindly provided by G. Trinchieri, Wistar Institute, Philadelphia, Penn) were each separately recombined using NheI (p35) and XbaI (p40) restriction sites, respectively. Again, the eukaryotic expression vector pRc/ CMV served as the acceptor. Both constructs were completely sequenced before use as described previously.17 For gene transfer, the ballistomagnetic vector system involving a twostep procedure was employed as described previously.14 The ballistomagnetic transfer of expression plasmids into 107 to 3 ⫻ l07 cells, seeded on a 10-cm Petri dish, was achieved by the simultaneous delivery of particles from seven particle carrier membranes distributed evenly to cover the entire Petri dish. Cell culture medium was removed from the Petri dish immediately before commencing the ballistomagnetic vector system. After ballistomagnetic transfer, cells were immediately resuspended in 5 mL of PBS and transferred onto a high-gradient magnetic separation column (capacity 3 ⫻ 107 cells, type AS; Miltenyi, Bergisch-Gladbach, Germany), which was prepared according to the supplier’s protocol. An aliquot of the cell suspension was kept for reference (unsorted fraction). This cell suspension was passed through the column, followed by a washing step with 3 mL of PBS. The effluent was collected (negative or nonmagnetic fraction). After removal of the column from the magnetic separator, the retained cells were flushed back to the top of the column. The column was returned to the separator and washed with 3 mL of PBS, and the effluent was collected (wash fraction). At this point, the column was again removed from the separator and eluted with 5 mL of PBS (magnetic fraction). Recovered cells were sedimented at 400 ⫻ g at 4°C for 7 minutes, resuspended in tissue culture medium, and incubated under culture conditions for 24 hours. On the following day, small volumes of genetransfected and magnetically fractioned melanoma cells were

irradiated with 100 Gy. Cells were subsequently detached from flasks by mechanical scraping and ice-cold PBS, washed three times with sterile PBS, counted, and resuspended in a final concentration of 5 ⫻ 106 gene-transfected melanoma cells/mL.

Determination of IL-7 and IL-12 Aliquots of 106 gene-transfected, magnetically fractionated and irradiated melanoma cells were transferred to each well of a 24-well plate (Nunc, Roskilde, Denmark) to determine cytokine secretion in vitro after 24 hours. In addition, serum levels of IL-7 and IL-12 were determined from the serum fraction of blood samples taken 24 hours after vaccinations. For both, an IL-7 enzyme-linked immunosorbent assay (EIAKIT; PerSeptive Diagnostics, Cambridge, Mass) or an IL-12 enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, Minn) were used. The detection range was between 10 and 200 pg/mL, according to the manufacturer’s instructions.

DTH skin testing DTH tests were performed 2 days before the first vaccination and during weeks 6 and 12 of treatment. Two preparations were used each time: one consisting of an autologous melanoma cell lysate and the other consisting of an autologous peripheral blood mononuclear cell (PBMC) lysate as a negative control. The melanoma test preparation was prepared from autologous cells of each patient, starting with 4 ⫻ 106 suspended, mitomycin-inactivated (45 minutes, 37°C) primary melanoma cells. After PBS washings, cells were lysed by three cycles of freeze-thawing. Aliquots of 1.3 ⫻ 106 cell equivalents were stored in 0.3 mL of PBS per tube at ⫺80°C. The autologous PBMC lysate was prepared after separation of peripheral blood via Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) gradient centrifugation, while otherwise employing the same procedures described above for melanoma cells. For DTH testing, 1 ⫻ 106 cell-equivalents were injected intradermally into the forearm. In parallel, a commercially available recall Ag DTH test (Multitest Merieux) was administered on the opposite forearm. A positive skin-test reaction was defined as a ⱖ5 mm diameter induration after 48 hours.

Peripheral blood samples and lymphocyte cultures Blood samples were obtained 1 or 2 days before the first vaccination and 2 weeks after the third vaccination. PBMCs were separated via Ficoll-Hypaque gradient centrifugation, washed twice in PBS, and either resuspended in complete culture medium or cryopreserved in liquid nitrogen in the presence of ⬎50% FCS or human sera. PBMCs isolated from one identical blood sample were employed for the non-MHCrestricted cytotoxicity assays as well as for the limiting dilution analysis (LDA) to determine the CTL responses either before or after vaccination. Spontaneous lytic activity analysis and flow cytometric analysis were performed immediately in fresh PBMCs. Cultures of LAK cells were generated and maintained in complete culture medium consisting of RPMI 1640, 10% FCS, 2 mM glutamine, 100 U/mL penicillin, and 100 ␮g/mL streptomycin (Seromed) in the presence of 1000 U/mL IL-2 at 37°C in an atmosphere of 5% CO2 in air.

LDA assay To estimate changes in the frequency of anti-melanoma lymphocytes, including precursor CTLs in peripheral blood, we used the 25-day LDA method described by Coulie et al,20 with minor modifications.14 Briefly, cryopreserved PBMCs



were seeded in limiting dilutions (1 ⫻ 104 to 625 PBMCs/ well in 1/2 dilutions) in microcultures of 96 V-bottom microwells that were restimulated weekly with 1 ⫻ 104 mitomycin-inactivated autologous melanoma cells in a Tcell medium consisting of RPMI 1640, 10% pooled human AB sera (Sigma, Deisenhofen, Germany), 25 mM N-2hydroxyethylpiperazine-N⬘-2-ethanesulfonic acid buffer, 2 mM glutamine, 100 U/mL penicillin, and 100 ␮g/mL streptomycin in the presence of 20 U/mL IL-2 (Boehringer Mannheim, Mannheim, Germany). The lytic activity of these microcultures was determined via lactate dehydrogenase release assay, using a commercially available detection kit for lactate dehydrogenase (Boehringer Mannheim). The results have been discussed in detail previously.14

Analysis of spontaneous lytic activity and LAK activity by a 3-(4,5-dimethylthiazol-2-yl)-diphenyltetrazolium bromide (MTT) assay The cytotoxicity of fresh PBMCs and LAK cells was determined via a 24-hour MTT-based colorimetric assay. Modified as a cytotoxicity assay for the quantification of cell-mediated cytotoxicity,21,22 this test has compared favorably with the 51Cr release assay in both sensitivity and reproducibility when the long-term cytotoxicity of adherent target cells is measured. Adherent target cells were plated at 1 ⫻ 104 cells/well in 96-well, flat-bottom tissue culture plates (Greiner, Frickenhausen, Germany) 20 hours before the addition of the effector cells. Effector cells were added at effector to target (E:T) ratios ranging from 20:1 to 2.5:1. If possible, quadruplicate wells were used for each E:T ratio. After 24 hours of coculture, effector cells and lysed target cells were removed by washing. Afterward, the remaining target cells were incubated for 4 hours in the presence of MTT (0.5 mg/mL; Sigma). Supernatants were discarded, and formazan crystals were solubilized by adding 20 ␮L of PBS and 150 ␮L of dimethyl sulfoxide (Merck, Darmstadt, Germany) to each well. Absorbance (A) was read at a wavelength of 540 nm on a scanning multiwell spectrophotometer (Titertek Multiscan MCC/340, Meckenheim, Germany). The percentage of lysis was determined as follows: ([Atarget cell ⫺ {Aexperimental ⫺ Aeffector cell}]/Atarget cell) ⫻ 100, where Atarget cell is defined as the absorbance of the target cells without addition of effector cells, Aexperimental is defined as the mean of the experimental absorbance of viable adherent target cells cocultured with the effector cells, and Aeffector cell is defined as the absorbance correcting for those “sticky” effector cells remaining in the well after washing.

Flow cytometric analysis The Ag phenotype of PBMCs obtained 1 or 2 days before the first vaccination and 2 weeks after the third vaccination was evaluated via direct immunofluorescence with the following monoclonal antibodies: anti-CD56 and anti-CD4 (phycoerythrin-labeled) and anti-CD8 and anti-CD3 (fluorescein isothiocyanate-labeled) (Becton Dickinson, Mountain View, Calif). Samples were analyzed by flow cytometry using an Epics XL (Coulter, Krefeld, Germany).

Statistical analysis The statistical significance of data was calculated by a modified Wilcoxon signed-rank test and a Mann-Whitney U test.


979

RESULTS

Characteristics of patients A total of 16 patients suffered from advanced metastatic melanoma and had received various conventional oncological treatment regimens before their treatment with autologous gene-transfected melanoma vaccines. A total of 8 of 10 enrolled patients received melanoma cells transfected for IL-7 expression, and 6 of 6 patients received melanoma cells transfected for IL-12 expression, in a complete regimen of four vaccinations. The clinical results and MHC-restricted CTL responses have been described previously in greater detail14,15 and are briefly summarized in Table 1. In summary, ⬃50% of the melanoma metastases received could be expanded in vitro. Vaccination treatment could be initiated between 2 weeks and 3 months after surgical removal of the metastases. Melanoma cells were ballistomagnetically transfected with expression plasmids. After irradiation (100 Gy) and magnetic fractionation, these melanoma cells showed a mean IL-7 secretion of 700 pg (per 106 cells in 24 hours) and a mean IL-12 secretion of 950 pg (per 106 cells in 24 hours), respectively. The secretion was decreasingly detectable over 2 weeks in vitro and was able to drive T-cell proliferation in a paracrine manner.14,15,23 Vaccination with gene-transfected melanoma cells was feasible and well tolerated. No serious adverse effects were registered. However, neither a complete response nor a partial response was observed in any patient. Parenthetically, it should be noted that two patients treated with IL-7 gene therapy (patients 8 and 10) and one patient treated with IL-12 gene therapy (patient 2b) exhibited a mixed clinical response (disappearance of cutaneous metastases); five patients showed an enhanced CTL response in vitro 2 weeks after the third vaccination (weak: patients 1, 10 and 4b; strong: patients 8 and 5b, respectively). All three patients with MR (patients 8, 10, and 2b) and four of five patients who showed an increased specific CTL response in vitro (patients 1, 8, 10, and 5b) had a good clinical performance status (Karnofsky ⬎ 70) as well as an intact recall reactivity upon skin testing (Multitest Merieux of ⬎13.4 mm/3) before the vaccinations. In addition, the patients with an enhanced specific CTL response received the highest numbers of given vaccine cells (three doses ⱖ 6 ⫻ 106) and also displayed a relatively good clinical outcome (two of three MR and two stable disease), except for one patient (patient 4b), who exhibited a weaker immune performance status (Multitest Merieux ⫽ 4 mm/1) and progressive disease (PD).

Non-MHC-restricted lytic activity To determine alterations in non-MHC-restricted lytic activity and in surface markers of PBMCs during gene therapy, PBMCs from 8 of 10 patients who received melanoma cells transfected for IL-7 expression and from 6 of 6 patients who received melanoma cells for IL-12 expression were obtained 1 day before gene therapy and 2 weeks after their third vaccination. The cytolytic

980


Figure 1. Lytic activity of one representative patient (patient 1). Various E:T ratios of spontaneous lytic activity (a) and LAK activity (b) against autologous melanoma cells are depicted. All MTT assays were performed in quadruplicate wells with a SD of the mean of ⬍10%. B, PBMCs collected 1 day before the first vaccination; A, PBMCs collected 2 weeks after the third vaccination.

activity of non-MHC-restricted lymphocytes (excluding the lytic activity of adherent macrophages, data not shown) was determined from freshly isolated PBMCs and from PBMCs after 5 days of culture in vitro, in the presence of 1000 U/mL IL-2 to generate LAK cells. Measurements of lytic activity were performed against autologous melanoma cells and against one allogeneic melanoma cell line, SK Mel-37, employing a 24-hour MTT cytotoxic assay (Figs 1 and 2).

Spontaneous lytic activity of PBMCs There was no opportunity to repeat cytotoxic assays due to the limited number of patient PBMCs. However, to compare lytic activities, we required lytic activities of at least 13% lysis. Furthermore, a difference in the percentage of lysis between before vaccination (B) and 2 weeks after the third vaccination (A) of ⬎10% was thought to be necessary for the difference to be considered significant, as tested with PBMCs obtained from healthy donors.21 Based on these criteria, an enhancement of spontaneous lytic activity against autologous melanoma cells was detected in the PBMCs of 7 of 12 evaluable patients (patients 1, 7, 9, 1b, 2b, 3b, and 6b), whereas a reduction in spontaneous lytic activity in the PBMCs of 3 of 12 patients (patients 5, 8, and 10) was observed after vaccination. Spontaneous lytic activity against the allogeneic melanoma cell line was found to be enhanced in 6 of 12 patients and increased in 4 of 12 patients after vaccination (Fig 2).

LAK activity More frequent enhancements, together with generally high lytic activities, were obtained from PBMCs after

vaccination versus before vaccination, when PBMCs were cultured in vitro to generate LAK activity. LAK activity before and after vaccination was determined in parallel in the same assay. A total of 10 of 14 completely measurable patients (patients 1, 2, 3, 5, 7, 9, 10, 2b, 3b, and 4b) showed an increase; in contrast, 2 of 12 patients (patients 8 and 5b) showed a reduction in lytic activity after vaccination. Differences in LAK activity in two patients (patients 1b and 6b) were not considered significant based on the criteria used. Increased LAK activity against an allogeneic melanoma cell line was shown in 9 of 12 patients; however, in 3 of 13 completely evaluable patients, decreased IL-2-induced lytic activity was detected (Fig 2). Remarkably, the patients with the only clear reduction in LAK activity after vaccination (patients 8 and 5b) exhibited the highest increase in their CTL responses. Interestingly, these were associated with a good clinical and immunological performance status and with a relatively good clinical response (Table 1, Fig 2).

Surface marker analysis of PBMCs As seen in Table 2, all eight patients who were vaccinated with melanoma cells transfected for IL-7 expression showed a significant percentage decrease in CD3⫹ cells in the PBMCs 2 weeks after the third vaccination. In most cases, both subsets, CD4⫹ cells and CD8⫹ cells, were also decreased. CD56⫹ cells were not uniformly altered in their percentage of positive cells upon vaccination. CD25 and HLA-DR were not up-regulated 2 weeks after the third vaccination (data not shown). In contrast, PBMC surface markers in the group of patients who were vaccinated with melanoma cells transfected



981

Figure 2. The lytic activity of freshly isolated PBMCs (spontaneous lytic activity) and after 5 days of culture in the presence of 1000 U/mL IL-2 (LAK activity) was measured against autologous melanoma cells and against an allogeneic melanoma cell line (SK Mel-37). a: Patients 1– 8 received melanoma cells transfected for IL-7 expression. b: Patients 1b-6b received melanoma cells transfected for IL-12 expression. Lytic activities were compared between 1 day before vaccination (B) and 2 weeks after the third vaccination (A). All 24-hour MTT assays were performed in quadruplicate or triplicate wells with an SD of the mean of ⬍10%.

for IL-12 expression were not changed over the course of vaccination.

Serum levels of IL-7 and IL-12 Serum levels measured at 1 day postvaccination did not show any distinction from the serum levels before vaccination (data not shown).


DISCUSSION This study demonstrates that non-MHC-restricted lytic activity against autologous and allogeneic melanoma target cells in a larger subset of patients could be augmented by vaccination of patients who received melanoma cells transfected for IL-7 or IL-12 expres-

982


Table 2. Surface Marker Analysis of PBMCs Performed by Flow Cytometry Before (B) and After (A) Vaccination with IL-7 or IL-12 gene-modified autologous melanoma cells Surface markers (%) positive cells) CD3 CD56 B*

CD4

CD8

A†

B

A

B

A

B

A

Patients (IL-7 gene therapy) 1 65 2 49 3 38 5 40 7 40 8 77 9 50 10 45

42 32 24 21 20 69 26 29

12 7 36 42 33 14 4 19

25 8 18 33 20 8 6 25

51 58 15 48 26 53 29 25

31 25 7 22 18 61 10 27

26 10 35 15 23 26 16 16

14 8 24 5 10 7 16 8

Patients (IL-12 gene therapy) 1b 66 2b 75 3b 63 4b ND‡ 5b 64 6b 74

65 75 62 ND 70 65

12 7 11 ND 3 2

6 4 4 ND 5 5

64 49 47 ND 61 32

62 47 44 ND 61 29

13 23 18 ND 15 36

14 22 14 ND 18 29

*B, 1 day before vaccination †A, 2 weeks after the third vaccination ‡ND, not determined.

sion. However, no clinical response was associated with increased lytic activity. An enhancement of spontaneous lytic activity against autologous melanoma cells was found in fresh PBMCs of 7 of 14 patients and a reduction of activity was found in only 3 patients after vaccination. It is presumed that the lytic activity in fresh PBMCs was produced mainly by non-MHCrestricted killer lymphocytes. In addition to the majority of NK cells, other mostly rare lymphoid subsets are also able to mediate a non-MHC-restricted lytic activity that is indistinguishable from activated NK cell killing. These lymphoid subsets include activated T cells expressing the ␥/␦ T-cell receptor and T cells expressing the ␣/␤ T-cell receptor together with the CD56 surface marker.24 Stimulation with high doses of IL-2 and, to a lesser extent, with IL-7 or IL-12 results in a further augmentation of non-MHC-restricted lytic activity in a broader range of susceptible target cells.25 To discern whether lytic potential could be augmented further by additionally stimulating partly preactivated effector cells in vivo with a high dose of IL-2 ex vivo, we cultured PBMCs of patients for 5 days in the presence of 1000 U/mL IL-2 to generate LAK activity. A total of 10 of 14 patients in the autologous E:T system and 9 of 13 completely measurable patients in an allogeneic E:T system showed an increased lytic activity after vaccination in comparison with the activity before vaccination. With the exception of the melanoma cells of one patient (tumor cells of patient 1 lost HLA-A1 expression), the target cells of all of the autologous cells completely expressed MHC class I molecules,14,15 a state which normally makes target cells less susceptible to non-

MHC-restricted killer cells. However, it is likely that different receptors are involved, depending upon the activation state of the effector cells and the availability of the ligands on the target cells, including MHC class I ligands plus non-MHC class I ligands.26 The augmentation of non-MHC-restricted lytic activity after vaccination could have been caused (a) by a local release of IL-7 or IL-12, (b) by the irradiated melanoma cells used as a vaccine, or (c) by the employed plasmid vectors containing short immunostimulatory DNA sequences, which may induce the release of IL-12, interferons, and other cytokines.27 It is likely that a combination of these factors was to some extent responsible for the increase in the lytic activity after vaccination. Negative controls using melanoma cell vaccines containing no plasmid vector or plasmid vectors without cytokine genes were not performed for ethical reasons. Interestingly, at 2 weeks after the third vaccination, an increase in non-MHC-restricted lytic activity or an increase in CTL response was elicited in PBMCs of different patient subgroups. An increase of non-MHC-restricted lytic activity could be detected particularly in patients with a poor clinical and a poor immunological performance status in this study. This was clearly manifested in a subgroup of three patients (patients 7, 9, and 3b) who showed the highest absolute increase of autologous spontaneous lytic activity, yet had a poor clinical (Karnofsky index of ⬍70) and a poor immunological (Multitest Merieux of ⬍4 mm/1) performance status. In addition, it should be noted that this group showed neither CTL responses nor clinical responses. Inversely, the subgroup (patients 8 and 5b) with the



highest increase in the specific CTL response exhibited a good clinical (Karnofsky index of ⬎70) and a good immunological (Multitest Merieux of 13.4 mm/3) performance status but no increase in non-MHC-restricted lytic activity upon vaccination. Furthermore, this group showed a superior clinical response, as was also found elsewhere.6,28 Oddly, the non-MHC-restricted lytic activity in this group decreased, thus making it the only group of patients in which LAK activity was clearly reduced. However, some authors have demonstrated a direct correlation between a clinical response and increased non-MHC-restricted lytic activity, as observed in IL-2-treated patients with cancer of the head and neck29 and in IL-2-treated patients after chemotherapy.30 Conversely, other investigators found no significant correlation at all.31 One reason for the difference in the results obtained by these groups may be the timing of the blood collections. In the present study, the decision was made to collect blood at 14 days postvaccination in accordance with the experience gained in previous attempts to measure CTL responses.32 In this period, NK-mediated cytotoxicity may have shifted to CTL-mediated cytotoxicity in patients who possessed a stronger tendency to develop a CTL response (e.g., patients 8 and 5b). As shown in animal trials, NK responses are replaced or down-regulated by a subsequent CTL response, while this was not the case in athymic nude mice. The number and activity of NK cells in these T-cell-deficient mice are elevated for a prolonged period (14 days) after infection with lymphocytic choriomeningitis virus, but are only elevated for 3– 4 days in normal immunocompetent mice.33 A further cause of the differences in the correlation between clinical response and non-MHC-restricted lytic activity might be that good clinical preconditions are not as important for the development of NK responses as they are for the induction of CTL responses. The observation that the number of CD3⫹ cells in the PBMCs of all of the patients who received melanoma cells transfected for IL-7 expression was clearly increased, but not in the PBMCs of the patients who received melanoma cells for IL-12 expression, may be worth considering. In any case, the fact remains that non-MHC-restricted lytic activity with an increased level of lytic activity 2 weeks after vaccination could be clearly detected even in patients with advanced melanoma accompanied by a reduced Karnofsky index and a defective DTH reactivity in skin testing. However, in view of the small numbers of patients studied, this finding will have to be confirmed with more data. It should also be kept in mind that, in the present setting, even when NK activity was increased, it was not sufficient to have an impact on the clinical course of the disease in these metastatic melanoma patients. Nevertheless, kinetic analyses of non-MHC-restricted lytic activity should be performed concurrently with analyses of CTL responses and evaluated in further studies to determine whether this is an immune marker that could be a useful prognostic tool in future cancer studies.


983

ACKNOWLEDGMENTS We thank Antje Sucker, Helga Kemmer, and Iris Ziglowski for excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (Scha 422/3-2) and by the Centrum Somatische Gentherapie e.V.

REFERENCES 1. Ahmann DL, Creagan ET, Hahn RG, et al. Complete responses and long-term survival after systemic chemotherapy for patients with advanced malignant melanoma. Cancer. 1989;63:224 –227. 2. Schadendorf D, Worm M, Algermissen B, et al. Chemosensitivity testing of human malignant melanoma: a retrospective analysis of clinical response and in vitro drug sensitivity. Cancer. 1994;73:103–108. 3. Mackensen A, Carcelain G, Viel S, et al. Direct evidence to support the immunosurveillance concept in a human regressive melanoma. J Clin Invest. 1994;93:1391–1402. 4. Rosenberg SA. The immunotherapy and gene therapy of cancer. J Clin Oncol. 1992;10:180 –199. 5. Marchand M, Weynants P, Rankin E, et al. Tumor regression responses in melanoma patients treated with a peptide encoded by gene MAGE-3. Int J Cancer. 1995;63:883– 885. 6. Ja¨ger E, Ringhoffer M, Dienes HP, et al. Granulocytemacrophage colony-stimulating factor enhances immune responses to melanoma-associated peptides in vivo. Int J Cancer. 1996;67:54 – 62. 7. Rosenberg SA, Yang JC, Schwartzentruber DJ, et al. Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma. Nat Med. 1998;4:321–327. 8. Nestle FO, Alijagic S, Gilliet M, et al. Vaccination of melanoma patients with peptide or tumor lysate pulsed dendritic cells. Nat Med. 1998;4:328 –332. 9. Colombo MP, Forni G. Cytokine gene transfer in tumor inhibition and tumor therapy: where are we now? Immunol Today. 1994;15:48 –51. 10. Mo ¨ller P, Schadendorf D. Somatic gene therapy and its implications in melanoma treatment. Arch Dermatol Res. 1997;289:71–77. 11. Trinchieri G. Interleukin-12: a cytokine produced by antigen-presenting cells with immunoregulatory functions in the generation of T-helper cells type I and cytotoxic lymphocytes. Blood. 1995;84:4008 – 4018. 12. Grimm EA, Mazumber A, Zhang HZ, Rosenberg SA. Lymphokine-activated killer cell phenomenon: lysis of natural killer cell-resistant fresh solid tumor cells by interleukin-2-activated autologous human peripheral blood lymphocytes. J Exp Med. 1982;155:1823–1841. 13. Alderson MR, Sassenfeld HM, Widmer MB. Interleukin-7 enhances lytic T lymphocyte generation and induces lymphokine-activated killer cells from human peripheral blood. J Exp Med. 1980;145:1983–1990. 14. Mo ¨ller P, Sun Y, Dorbic T, et al. Vaccination with IL-7 gene-modified autologous melanoma cells can enhance the anti-melanoma lytic activity in peripheral blood of patients with a good clinical performance status: a clinical phase I study. Br J Cancer. 1998;77:1907–1916. 15. Sun Y, Jurgovsky K, Mo ¨ller P, et al. Vaccination with IL-12 gene-modified autologous melanoma cells: preclinical results and a first clinical phase I study. Gene Ther. 1998;5:481– 490.

984


16. Arienti F, Sule-Suso J, Belli F, et al. Limited antitumor T cell response in melanoma patients vaccinated with interleukin-2 gene-transduced allogeneic melanoma cells. Hum Gene Ther. 1996;7:1955–1963. 17. Schadendorf D, Czarnetzki BM, Wittig B. Interleukin-7, interleukin-12, and GM-CSF gene transfer in patients with metastatic melanoma. J Mol Med. 1995;73:473– 477. 18. Houghton AN, Real FX, Davis LJ, et al. Phenotypic heterogeneity of melanoma: relation to the differentiation program of melanoma cells. J Exp Med. 1987;164:812– 829. 19. Schadendorf D, Bo ¨hm M, Mo ¨ller P, et al. Interleukin-7 induces differential lymphokine-activated killer cell activity against human melanoma cells, keratinocytes, and endothelial cells. J Invest Dermatol. 1994;102:838 – 842. 20. Coulie P, Somville M, Lehmann F, et al. Precursor frequency analysis of human cytolytic T lymphocytes directed against autologous melanoma cells. Int J Cancer. 1992;50:286 –297. 21. Bo ¨hm M, Mo ¨ller P, Kalbfleisch U, et al. Lysis of allogeneic and autologous melanoma cells by IL-7-induced lymphokine-activated killer cells. Br J Cancer. 1994;70:54 –59. 22. Heo DS, Park JG, Hata K, et al. Evaluation of tetrazoliumbased semi-automatic colorimetric assay for measurement of human anti-tumor cytotoxicity. Cancer Res. 1990;50: 3681–3690. 23. Finke S, Trojaneck B, Mo ¨ller P, et al. Increase of cytotoxic sensitivity of primary human melanoma cells transfected with the interleukin-7 gene to autologous and allogeneic immunologic effector cells. Cancer Gene Ther. 1997;4:260 –268. 24. Hiserodt JC. Lymphokine-activated killer cells: biology and relevance to disease. Cancer Invest. 1993;11:420 – 439. 25. Wiebke EA, Custer MC, Rosenberg SA, Lotze MT. Cytokines alter target cell susceptibility to lysis: evaluation of non-major histocompatibility complex-re-

26.

27. 28.

29. 30.

31.

32.

33.

stricted effectors reveals differential effects on natural and lymphokine-activated killing. J Biol Response Mod. 1990;9:113–126. Pessino A, Sivori S, Bottino C, et al. Molecular cloning of NKp46: novel member of the immunoglobulin superfamily involved in triggering of natural cytotoxicity. J Exp Med. 1998;188:953–960. Sato Y, Roman M, Tighe H, et al. Immunostimulatory DNA sequences necessary for effective intradermal gene immunization. Science. 1996;273:352–354. Barth A, Hoon DSB, Foshag LJ, et al. Polyvalent melanoma cell vaccine induces delayed-type hypersensitivity and in vitro cellular immune response. Cancer Res. 1994; 54:3342–3345. Schantz SP, Sillitoe EJ, Brown B, Campbell B. Natural killer cell activity and head and neck cancer: a clinical assessment. J Natl Cancer Inst. 1986;77:869 – 875. Mitchell MS, Kempf RA, Harel W, et al. Effectiveness and tolerability of low-dose cyclophosphamide and low-dose intravenous interleukin-2 in disseminated melanoma. J Clin Oncol. 1988;6:409 – 424. Schomburg A, Menzel T, Ko ¨rfer A, et al. In vivo and ex vivo antitumor activity in patients receiving low-dose subcutaneous recombinant interleukin-2. Nat Immun. 1992; 11:133–143. Herr W, Wo ¨lfel T, Heike M, et al. Frequency analysis of tumour-reactive cytotoxic T lymphocytes in peripheral blood of melanoma patients vaccinated with autologous tumour cells. Cancer Immunol Immunother. 1994;39:93– 99. Kos FJ, Engleman EG. Immune regulation: a critical link between NK cells and CTLs. Immunol Today. 1996;17: 174 –176.
