|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Novel Allogeneic Granulocyte-Macrophage Colony-Stimulating Factor–Secreting Tumor Vaccine for Pancreatic Cancer: A Phase I Trial of Safety and Immune ActivationFrom the Departments of Oncology, Surgery, and Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD. Address reprint requests to Elizabeth Jaffee, MD, The Johns Hopkins University, The Bunting-Blaustein Cancer Research Building, Room 4M07, 1650 Orleans St, Baltimore, MD 21231; email ejaffee@ jhmi.edu.
PURPOSE: Allogeneic granulocyte-macrophage colony-stimulating factor (GM-CSF)–secreting tumor vaccines can cure established tumors in the mouse, but their efficacy against human tumors is uncertain. We have developed a novel GM-CSF–secreting pancreatic tumor vaccine. To determine its safety and ability to induce antitumor immune responses, we conducted a phase I trial in patients with surgically resected adenocarcinoma of the pancreas. PATIENTS AND METHODS: Fourteen patients with stage 1, 2, or 3 pancreatic adenocarcinoma were enrolled. Eight weeks after pancreaticoduodenectomy, three patients received 1 x 107 vaccine cells, three patients received 5 x 107 vaccine cells, three patients received 10 x 107 vaccine cells, and five patients received 50 x 107 vaccine cells. Twelve of 14 patients then went on to receive a 6-month course of adjuvant radiation and chemotherapy. One month after completing adjuvant treatment, six patients still in remission received up to three additional monthly vaccinations with the same vaccine dose that they had received originally.
RESULTS: No dose-limiting toxicities were encountered. Vaccination induced increased delayed-type hypersensitivity (DTH) responses to autologous tumor cells in three patients who had received CONCLUSION: Allogeneic GM-CSF–secreting tumor vaccines are safe in patients with pancreatic adenocarcinoma. This vaccine approach seems to induce dose-dependent systemic antitumor immunity as measured by increased postvaccination DTH responses against autologous tumors. Further clinical evaluation of this approach in patients with pancreatic cancer is warranted.
ADENOCARCINOMA OF the exocrine pancreas is the fifth leading cause of cancer death in the United States, with an overall 5-year survival rate of less than 5%.1 Although radiotherapy and chemotherapy, particularly gemcitabine, have produced modest benefits in some patients, no currently available treatment affects 2-year survival for patients with locally advanced and metastatic disease.2-4 Similarly, for patients with localized disease, current adjuvant radiation and chemotherapy have shown only modest benefits after surgical resection.5-7 One promising alternative is vaccine therapy. In particular, autologous granulocyte-macrophage colony-stimulating factor (GM-CSF)– secreting tumor vaccines have cured established tumors in mice and shown promising results in patients with prostate and renal cell carcinoma and melanoma.8-12 In theory, autologous tumor cells would be the best source of immunizing proteins, since they would likely display all of the relevant tumor antigens for inducing antitumor immunity in the patient. However, this approach poses important technical problems. First, an autologous vaccine depends on the availability of adequate numbers of tumor cells, which are rarely available because of the reactive processes that are found infiltrating the tumor cells of many common cancers. Second, an autologous vaccine requires de novo gene transfer for treatment of each patient, which is labor intensive and may cause variable GM-CSF expression levels between different patient vaccines. Third, there is significant expense and time required to certify each patient’s lot of vaccine cells so that they meet Food and Drug Administration guidelines. One way to circumvent these technical obstacles would be to develop an allogeneic vaccine strategy based on a panel of cytokine-expressing tumor cell lines that can be formulated and stored before the initiation of clinical studies. This is a particularly attractive approach for the majority of common cancers for which specific tumor antigens have not yet been identified. Three findings provide the immunologic rationale for an allogeneic vaccine approach. First, many tumor antigens seem to be commonly expressed among different patients’ tumors. for example, approximately half of the human melanoma antigens identified so far have been demonstrated to be shared among at least 50% of other patient melanomas.13,14 Similarly, three of the four candidate pancreatic cancer–associated T-cell antigens that are currently known are also shared by the majority (50% to 90%) of pancreatic adenocarcinomas. These include carcinoembryonic antigen, the mucin MUC-1, and the oncogene product of mutated K-ras.15-17 The gene products of mutated p53, although frequently expressed by pancreatic carcinomas, tend to be a heterogenous group of antigens.17 Second, several studies have shown that the professional antigen-presenting cells (APCs) of the host, rather than the vaccinating tumor cells themselves, are responsible for priming CD4+ and CD8+ T cells, both of which are required for generating systemic antitumor immunity.18,19 Although the antigens recognized by the CD8+ T cells were not known at the time, these studies demonstrate that tumor-specific CD8+ T cells are activated via the cross-priming mechanism. Specifically, APCs attracted to the vaccinating tumor as a result of GM-CSF secretion take up whole cellular proteins released into the tumor’s microenvironment and process them in both the major histocompatibility complex class I and II antigen-processing compartments for presentation to both CD4+ and CD8+ T cells. Third, and most important, the efficacy of allogeneic vaccines has recently been validated in animal models.8 Specifically, allogeneic tumor vaccines induced tumor-specific immune responses that correlated with in vivo tumor rejection in two different mouse tumor models.20,21 These findings provide strong support that the immunizing tumor cells do not need to be HLA compatible with the host to generate tumor antigen-specific immunity. We have previously described methods that can routinely establish in vitro pancreatic adenocarcinoma cell lines from primary pancreatic tumor specimens.22 Two of these lines were stably transfected to express the human cytokine GM-CSF.22 Here we report the first clinical evaluation of an irradiated, allogeneic GM-CSF–transduced cancer vaccine composed of these two allogeneic GM-CSF–secreting pancreatic tumor lines. Specifically, we conducted a phase I trial of 14 patients with stage 1, 2, or 3 pancreatic adenocarcinoma to assess the safety and the induction of systemic antitumor immune responses.
Patients and Study Design Patient selection. Fourteen patients with pancreatic adenocarcinoma who underwent pancreaticoduodenectomy at The Johns Hopkins Hospital (Baltimore, MD) between August 1, 1997, and March 30, 1998, were recruited. Patient characteristics are listed in Table 1. Patients were considered eligible for this study if they met all of the following criteria: stage 1, 2, or 3 adenocarcinoma of the pancreas; enrollment within 8 weeks of pancreaticoduodenectomy; age 18 or older; ability to give informed consent; no known second malignancies (other than carcinoma-in-situ of the cervix, superficial skin cancer, or superficial bladder cancer); Eastern Cooperative Oncology Group performance status of 0 or 1; no clinical evidence of metastases; no serious autoimmune or allergic disease nor history of any autoimmune disease requiring treatment with systemic corticosteroids; no systemic corticosteroids within 1 month before receiving the vaccine; adequate hematologic, hepatic, and renal function; and human immunodeficiency virus–negative status.
Study design. The study protocol has been published.23 After obtaining informed consent, baseline studies were performed for tumor assessment (computed tomography [CT] scan and CA19-9 serum marker level measurements) and toxicity evaluation (complete blood counts with differential and platelets, complete chemistry profile, absolute eosinophil count, and thyroid-stimulating hormone, antinuclear antibodies, serum amylase, and lipase analyses). Baseline immune parameters included a 4-mm punch skin biopsy, placement of an anergy panel, and delayed-type hypersensitivity (DTH) testing to autologous pancreatic tumor cells, normal pancreas cells, and peripheral-blood lymphocytes (PBLs). The intervention and data collection schedule is diagrammed in Fig 1. Patients received the first vaccination 7 to 8 weeks after pancreaticoduodenectomy. All patients underwent weekly toxicity monitoring that included a complete blood count with differential, including an absolute eosinophil count, and platelets, a complete chemistry profile, and serum amylase and lipase analyses. A 4-mm punch skin biopsy of one of the vaccination sites was performed on day 3 and again on day 7 after vaccination. Serum was collected for future determination of GM-CSF levels before vaccination and at 24, 48, 72, 96, and 168 hours after vaccination. On day 28 after each vaccination, tumors were assessed by CT scan and measurement of CA19-9 biomarker levels, as well as by repeat anergy and DTH testing of autologous pancreatic tumor cells, normal pancreas tumor cells, and PBLs when available. Eligible patients were subsequently enrolled onto an adjuvant radiation and chemotherapy protocol. Tumors were assessed 4 to 8 weeks after completion of adjuvant radiation and chemotherapy. Patients who were still in remission were treated with up to three additional vaccinations given 1 month apart at the same original dose that they received for the first vaccination. Patients who received the three additional vaccinations underwent the same toxicity evaluation, tumor assessment, and immune monitoring as described for the first vaccination period. Skin biopsies and measures of serum GM-CSF levels were repeated after the second vaccination only.
Vaccinations. Two pancreatic tumor cell lines (PANC 10.05 and PANC 6.03) were tested in this study. The vaccine cell lines were developed from histologically confirmed primary pancreatic adenocarcinomas and genetically modified using a plasmid vector to express the human cytokine GM-CSF, as previously described.22 These lines were confirmed to derive from pancreatic adenocarcinomas based on the expression of cytokeratin, and the same k-ras mutation and cytogenetics profile when compared with the original unpassaged tumor.22 Clinical-grade vaccines were prepared at Bioreliance Corporation (Rockville, MD). Vaccine cells were frozen in serum-free media (AIM-V; BioWhittaker, Walkersville, MD) and 10% dimethyl sulfoxide. On the day of vaccination, vials of cells were thawed, washed three times in Hanks’ balanced salt solution, and irradiated with 150 Gy. For the first vaccination, all patients received three simultaneous injections of a single dose of the PANC 10.05 cell line, which had been genetically modified to express 120 ng/106 cells/24 hours of bioactive GM-CSF. The second, third, and fourth vaccinations consisted of equal numbers of the PANC 10.05 cell line and the PANC 6.03 cell line, which had been genetically modified to express 205 ng/106 cells/24 hours of bioactive GM-CSF. This was a dose-escalation study: three patients each received a total dose of either 1 x 107, 5 x 107, or 10 x 107 vaccine cells per vaccination. Five patients received 50 x 107 vaccine cells per immunization. For each patient, the total number of immunizing cells injected during each vaccination period remained the same throughout the series of vaccinations. Based on our prior feasibility studies, each dose was divided into either three (1 x 107, 5 x 107, and 10 x 107 vaccine cells) or six (50 x 107 vaccine cells) 0.6-mL aliquots, and each aliquot was delivered intradermally into three different limbs. Patients receiving the 50 x 107 dose received two aliquots of 0.6 mL per limb. Assessment of toxicities. Toxicities were graded using the National Cancer Institute’s cancer clinical trials common toxicity criteria. Toxicities were identified by medical history, physical examination, and review of the laboratory studies performed. Pharmacokinetic analysis of serum GM-CSF levels. Serum was separated from whole blood by centrifugation at high speed for 10 minutes and frozen in 1-mL aliquots at -80°C until the day of testing. Serum GM-CSF levels for all collection time points were determined by enzyme-linked immunosorbent assay (Quantikine Systems). All samples were run simultaneously. Each assay plate included a World Health Organization GM-CSF control standard supplied by the Biologic Resources Branch, National Cancer Institute (Frederick, MD). Serum levels of GM-CSF were determined using the World Health Organization GM-CSF standard and a linear regression analysis on the Microsoft Excel software (Microsoft, Redmond, WA).
Immune Monitoring Studies DTH testing. To evaluate the status of each patient’s cell-mediated immunity before and after treatment, DTH testing was performed using seven common recall antigens (Multitest CMI; Connaught Laboratories, France). Simultaneously, patients were tested for reactivity against autologous, irradiated pancreatic adenocarcinoma cells and normal pancreas cells obtained at the time of pancreaticoduodenectomy. Cells for autologous tumor and normal pancreatic cell DTH testing were prepared from the surgical specimen by enzymatic digestion with collagenase I (Sigma Chemical Co, St Louis, MO) and then washed and frozen in 90% AIM-V serum-free media plus 10% dimethyl sulfoxide (JT Baker, Phillipsburg, NJ) until the day of testing. On the day of testing, cells were thawed, washed three times in Hanks’ balanced salt solution, irradiated to 150 Gy, and injected intradermally at 106 cells/0.2 mL. DTH was measured as bidimensional induration at 48 hours at the site of test antigen administration.10,11,24-29 Patients were tested within 48 hours before vaccination for T-cell anergy to seven common recall antigens using the Multitest CMI (Connaught Laboratories, Swiftwater, PA). DTH tests were performed before the first vaccination and 28 days after each vaccination when sufficient DTH materials were available. Irradiated, autologous PBLs isolated by Ficoll (Pharmacia, Peapack, NJ) separation from whole blood and exposed to the same digestion enzymes as the tumor and normal pancreatic cells were used as control DTH cells. A 4-mm biopsy sample was taken from reactive DTH sites for H&E and immunohistochemical staining, as described above.
Statistical Considerations
Patient Characteristics As shown in Table 1, the mean age of patients enrolled onto the study was 60 years (range, 45 to 77 years). Thirteen of the 14 patients enrolled onto the study had stage 3 (lymph node–positive) adenocarcinoma, and one had stage 2 (lymph node–negative) disease. Of the 13 patients with stage 3 disease, four also were found to have infiltrating carcinoma at a final margin of their pancreaticoduodenectomy specimens. The one patient with stage 2 disease had a large T3 tumor with positive margins. All 14 patients received treatment with the first vaccination before adjuvant radiation and chemotherapy. Six of the 14 patients were still disease-free based on clinical and radiographic evaluation 1 month after completing adjuvant radiation and chemotherapy and were therefore eligible for the three additional vaccinations. One of these six patients (T2 [3 cm], N1 [five of 17+ lymph nodes]) who received the lowest dose of vaccine cells (1 x 107) had already demonstrated a 10-lb weight loss at the time she received the second vaccination and developed intra-abdominal pain consistent with metastases 2 weeks after vaccination. Two patients at dose level 2 (5 x 107 vaccine cells) were eligible for the three additional vaccinations. One of these patients (T2 [2.7 cm], N1 [two of 23+ lymph nodes]) received all three additional vaccinations, each administered 1 month apart, and subsequently developed a small, asymptomatic liver lesion that was identified on CT scan 1 month after completion of the final vaccination. The other patient (T2 [2.5 cm], N1 [two of 17+ lymph nodes]) received two of the three additional vaccinations. He refused the final vaccination because he did not want to continue to travel away from home (he lived about 600 miles away). He developed a single liver nodule 15 months after diagnosis (45 days after completing the third of four vaccinations). One patient (T2 [2.7 cm], N1 [five of 18+lymph nodes] with positive surgical margins) at dose level 3 (10 x 107 vaccine cells) was eligible for the three additional vaccinations. Two weeks after receiving the second vaccination, she developed thrombotic thrombocytopenic purpura related to mitomycin treatment that was a component of her adjuvant chemotherapy regimen, which she had completed 6 weeks earlier. Although this patient demonstrated a DTH reaction to autologous tumor cells after vaccination, she did not go on to receive the final two additional vaccinations. She was disease-free 30 months after diagnosis. Two patients at dose level 4 received all three additional vaccinations. Both patients demonstrated a DTH reaction to autologous tumor cells after vaccination. One patient (T2 [3 cm], N1 [two of 14+lymph nodes]) was disease-free 27 months after diagnosis. The other patient (T3 [3 cm], N0 with positive surgical margins) was disease-free 25 months after diagnosis. Pathologic staging results for each patient is listed with response in Table 2.
Toxicity No local or systemic dose-limiting toxicities were observed. The toxicities observed are listed in Table 3. In the majority of patients at all dose levels, the most common side effects of this treatment was grade 1 or 2 local skin reactions at the site of vaccination that typically manifested as erythema and induration. These reactions lasted up to 2 weeks after vaccination and were self-limited. However, these reactions increased in size and intensity with each additional vaccination. One patient who received the highest dose of vaccine cells demonstrated a few small ulcerations after the third and fourth vaccinations that were also self-limited. Other local/regional reactions observed included pruritus at the vaccination sites, a recall induration at the vaccination sites that occurred about 2 weeks after vaccination, and self-limited lymphedema of an upper extremity in one patient. There were few systemic side effects, all of which were grade 1 or 2 and self-limited. One patient reported fatigue, two patients reported joint stiffness, two patients reported generalized pruritus of the skin, and one patient reported urticaria. Another patient who received the highest dose of vaccine cells developed a generalized, pruritic, papular rash over his trunk and extremities 2 weeks after the second vaccination. Biopsy of this rash showed acantholytic dermatosis (Grover’s disease). Presumably, the response was precipitated by the GM-CSF production associated with vaccine. This patient’s symptoms resolved within a week without treatment and did not recur after the third or fourth vaccinations. As discussed below, this patient was one of four patients in the study in whom detectable serum levels of GM-CSF developed within the first 96 hours after vaccination. None of the other four patients who received vaccinations at this dose level (50 x 107 vaccine cells) developed similar reactions. There was no evidence of autoimmunity or pancreatitis at up to 39 months of follow-up.
Pharmacokinetics of Serum GM-CSF We measured serum GM-CSF levels just before vaccination and at 24, 48, 72, 96, and 168 hours after vaccination. Serum GM-CSF levels were not detected in any patient receiving the 1 x 107, 5 x 107, or 10 x 107 dose of vaccine cells. However, as shown in Fig 2, four of five patients receiving 50 x 107 vaccine cells had low but detectable serum GM-CSF levels that peaked at 48 hours after vaccination. These serum GM-CSF levels could be detected for up to 96 hours after vaccination but were undetectable by 168 hours after vaccination. The detection of peak serum level at 48 hours after vaccination in this study is similar to what was observed in preclinical studies of both autologous and allogeneic vaccine approaches.21
Immunologic Responses The histopathology of the vaccination sites was evaluated as a measure of the local immune reaction to the vaccine. A prevaccine skin biopsy specimen was compared with specimens taken at one of the vaccination sites on day 3 and again on day 7 after the first vaccination. Overall, there was an increase in the percentage of infiltrating immune cells with increasing dose of vaccine. Specifically, five of 14 patients demonstrated a local inflammatory response similar to that observed in preclinical models9 and in patients who received an autologous GM-CSF–secreting tumor vaccine for the treatment of renal cell carcinomas10 and malignant melanoma.11 These five patients received vaccine cells at either dose level 3 (two patients) or dose level 4 (three patients). Cytokeratin staining for pancreatic tumor cells (not shown) was performed to localize the infiltrate to the site of vaccine cells. H&E staining ( Fig 3, A and B) of the day 3 biopsy specimens of these five patients revealed an infiltration of mononuclear cells and eosinophils. This infiltration stained positive for macrophages with few observed CD3+ T cells on day 3 (Fig 3C), and positive for CD3+ T cells with few observed macrophages on day 7(Fig 3D). The other nine patients demonstrated a much weaker local inflammatory response consisting of less than 30% total infiltration of lymphocytes and macrophages at days 3 and 7 after vaccination. Few eosinophils were seen in the infiltrates.
DTH responses to autologous tumor cells were measured to determine whether the allogeneic GM-CSF–secreting tumor vaccine indeed induced systemic immunity against autologous pancreatic cancer. DTH responses to common recall antigens were also evaluated as a general measure of cellular immune responses in each patient. As shown in Table 2, only two of the 14 subjects had positive CMI DTH scores to common recall antigens that compared with the reported range for healthy volunteers. The other eight subjects had depressed CMI DTH scores below the score range reported for healthy volunteers, patients with localized cancer, and patients with disseminated cancer.24 Four of the eight patients were anergic at the time of first vaccination. Interestingly, there seemed to be no correlation between prevaccination CMI DTH scores and the ability to demonstrate a postvaccination DTH response to autologous tumor cells.
The DTH response to unpassaged, irradiated, autologous tumor cells was tested before and 28 days after the first vaccination. These results are presented in Table 2. Ten of the 14 patients treated did not demonstrate significant DTH reactivity (> 5mm) to autologous tumor cells before vaccination. The other four patients demonstrated a prevaccination DTH reaction against autologous tumors that ranged between 6 mm and 12.5 mm. However, only one of these four patients demonstrated a significant change in mean diameter of the DTH reaction (> 5 mm) after vaccination. A significant change in postvaccination DTH activity was not noted in patients treated at dose levels 1(1 x 107 cells) and 2 (5 x 107 cells). Of the patients for whom autologous tumor cells were available, one of three treated at dose level 3 (10 x 107 cells) and two of four treated at dose level 4 (50 x 107 cells) demonstrated a significant change in postvaccination DTH induction (
Biopsies were performed on postvaccination DTH reactions greater than 5 mm in largest diameter (patients no. 2, 7, 8, 10, 13, and 14) and on one prevaccination reaction greater than 5 mm (patient no. 10). The biopsy sites of patients no. 8, 13, and 14 demonstrated the characteristic DTH response that has been associated with the GM-CSF–secreting tumor vaccine approach.10,11 As shown in Fig 5A, these sites consisted of mononuclear cells and perivascular cuffing by lymphocytes. Notably, an intense eosinophil infiltration was also present at the reactive DTH sites. These three patients were disease-free 25 or more months after diagnosis. In contrast, the biopsy sites of patients no. 2, 7, and 10 demonstrated a minimal perivascular lymphocyte infiltrate with few, if any, infiltrating eosinophils (Fig 5B). These three patients developed disease recurrence at 6, 16, and 9 months after diagnosis. The one prevaccination biopsy obtained also demonstrated minimal perivascular lymphocyte infiltration without eosinophils that were positive for the lymphocyte marker CD3 on immunohistochemical staining (data not shown).
Clinical Responses Assessment of efficacy was not a primary objective of this phase I trial. However, as shown in Fig 4, one of the three patients who received 10 x 107 vaccine cells and two of the five patients who received 50 x 107 cells were disease-free at more than 25 months after diagnosis. These are the same patients who demonstrated a greater than 1 cm DTH reaction against autologous tumor cells after vaccination (Fig 4 and Table 2).
These data from a phase I trial of a novel allogeneic GM-CSF–secreting pancreatic tumor vaccine support the following two conclusions. First, the vaccine is safe in patients with pancreatic adenocarcinoma. Second, at high doses, the vaccine provokes DTH responses against autologous tumor, providing strong evidence of antitumor immunity. Most patients experienced only minor local toxicities at the vaccination site as a result of the expected GM-CSF–secreting vaccine’s specific inflammatory response, which has been characterized for autologous GM-CSF–secreting tumor vaccines.10-12 This expected toxicity is also considered an important local response and provides early evidence that the vaccine is initiating the cascade of signals that ultimately will result in systemic antitumor immunity. One patient who received the highest dose of vaccine cells (50 x 107 total cells) developed Grover’s disease, a self-limited primary acantholytic disease that occurs predominantly in patients age 50 or older.32-34 The primary lesions, consisting of discrete papules and papulovesicles, are typically distributed mainly over the chest, back, and thighs and may be intensely pruritic. The pathology of Grover’s disease was recently reviewed by Davis et al at the Mayo Clinic.33 Of the 72 patients included in their report, 63 were men with a mean age of 48 years. Lesions were typically distributed on the trunk and proximal extremities. Although the causes of this syndrome are still unknown, heat, sweat, and extensive sun exposure frequently were exacerbating factors. The only two medications that have been associated with the development of Grover’s disease are sulfadoxine pyrimethamine and recombinant interleukin-4.34 However, low systemic GM-CSF levels were detected in this patient after vaccination. Therefore, it is quite possible that systemic levels of GM-CSF were responsible for the disease in this patient. In the Mayo study, a perivascular lymphocytic infiltrate was associated with eosinophils in 22% of cases.33 Eosinophils were also seen in this patient’s skin biopsies, at the site of the rash, and at the vaccination and DTH sites. Of note, this patient developed a significant peripheral eosinophilia, with up to 25% absolute eosinophils after each vaccination. GM-CSF has been shown to recruit eosinophils and produce a peripheral eosinophilia in mouse and human studies of GM-CSF–secreting tumor vaccines. This is the first report of this reaction in patients receiving a GM-CS–secreting tumor cell vaccine, and it is likely due to the higher dose of GM-CSF–secreting vaccine cells administered in this study. The patient’s Grover’s disease resolved spontaneously after 2 weeks and did not recur after subsequent vaccinations. This patient is one of three patients who was disease-free at approximately 2 years after diagnosis. Significant changes (> 5mm) in postvaccination DTH responses to dissociated autologous pancreatic tumor cells were observed in one of three and two of four patients receiving the 10 x 107 and 50 x 107 vaccine doses, respectively. This finding suggests that there is a dose-dependent antitumor immune response associated with vaccine treatment. DTH measurements of tumor-specific systemic responses have so far been found to be the best correlative measure of clinical response after vaccination in patients with cancer.10-12,24-29 Because the DTH cells were prepared by collagenase digestion, it is possible that this foreign protein contributes to the observed DTH responses. However, unlike in many previously reported studies, the DTH cells used in this study were stored frozen in serum-free media, which makes it unlikely that other foreign proteins contributed to the observed responses. Moreover, all DTH specimens, including the autologous pancreatic tumor cells, normal pancreas cells, and PBLs, tested in each patient were prepared identically and included collagenase. Therefore, it is unlikely that the postvaccination DTH responses were exclusively against collagenase. Four patients did demonstrate a more than 5 mm DTH response to tumor cells before vaccination. However, all four patients also demonstrated similar DTH responses to their PBL control cells. This finding would suggest that some patients mounted small, nonspecific responses to the processed DTH cells. In three of the four patients, the postvaccination DTH reactions either decreased or remained the same size. These three patients have already had tumor recurrence. However, in the fourth patient, the postvaccination DTH response to tumor cells increased by more than 10 mm in each diameter, whereas the postvaccination response to PBLs remained unchanged. This patient remains disease-free. The fact that none of the patients demonstrated a postvaccination change in DTH response to any of the nontumor DTH reagents further supports our observation that the postvaccination change in DTH responses against autologous tumor cells indicates the induction of clinically relevant antitumor immune responses. In the present study, the patients’ ability to demonstrate DTH responses to autologous tumor cells postvaccination did not correlate with their ability to demonstrate adequate DTH responses to common recall antigens. In fact, only two patients treated on this study had CMI scores within the range reported for normal volunteers. The majority of patients had suppressed CMI scores below those reported for patients with disseminated cancer.24,26 Surgical procedures have been reported to temporarily suppress cellular immune responses.35-37 It is possible that the suppressed CMI scores observed in the majority of patients treated in this study were due to immune suppression from surgical resection of their cancer performed 8 weeks before vaccination. However, surgery alone cannot explain this finding, since most patients continued to demonstrate suppressed CMI indexes even at 12 weeks after vaccination, just before initiating adjuvant radiation and chemotherapy (data not shown). The observed postvaccination DTH responses against autologous tumor cells provide the best evidence that the vaccine generated antitumor immunity. Two additional findings provide further evidence that an antitumor immune response was induced. First, immunohistochemical analysis of vaccination site biopsy specimens revealed immune infiltrates similar to those observed in successful preclinical and clinical studies of the autologous GM-CSF–secreting vaccine approach.9-12 The number of eosinophils and macrophages infiltrating the vaccination site on day 3 after immunization seemed to be vaccine dose–dependent. Second, immunohistochemical staining of reactive DTH biopsy specimens also revealed the characteristic lymphocyte and eosinophilic infiltrate previously observed with the autologous vaccine approach in mice and patients.9-12 This eosinophil infiltrate has been shown to be associated with the presence of GM-CSF when the DTH response of GM-CSF–transduced vaccines was compared with that of nontransduced tumor vaccines in murine tumor models9,38 and in previous autologous GM-CSF–secreting tumor vaccine trials.10 In fact, the effects of whole-cell vaccines engineered to secrete GM-CSF in a paracrine fashion have been well characterized in mouse models.9,39,40 The enhanced immune response is due to the unique ability of GM-CSF to promote the recruitment and differentiation of professional APCs such as dendritic cells at the vaccination site.9,18,39,40 The systemic antitumor immunity produced with GM-CSF–secreting tumor cell vaccines is now known to be due to the activation of both CD4+ and CD8+ T cells by these professional APCs.8,9,18,39,40 In contrast, immunohistochemical staining of the DTH biopsy specimens from patients who did not demonstrate a significant postvaccination change revealed only the minimal perivascular lymphocyte infiltrate also observed in the nonresponders receiving the autologous vaccine in our other clinical and preclinical studies. In this study, we also evaluated systemic GM-CSF levels as an indirect measure of the longevity of vaccine cells at the immunizing site. Preclinical models have shown that up to 2 to 4 days of local GM-CSF production at the site of tumor cells is required to activate a systemic antitumor immune response.9,40 In these models, low but detectable serum GM-CSF levels can be measured as early as 4 hours after vaccination. Serum levels have been reported to peak at 48 hours after vaccination with both the autologous40 and allogeneic21 vaccine approaches, although the peak levels are about two-fold lower after allogeneic vaccination. In the first reported study testing a GM-CSF–secreting autologous renal tumor vaccine in patients, serum GM-CSF levels were not detected at any time between day 0 and day 7 of vaccination, even after a dose of 4 x 107 cells, which was the highest dose achieved in that study. However, in the present study, four out of five patients who received the highest dose of vaccine cells demonstrated detectable GM-CSF serum levels that peaked 48 hours after vaccination and that remained detectable for 96 hours in three of the four patients. Detection of GM-CSF in serum is likely due to total vaccine cell number, since the highest dose of vaccine cells administered in this study was more than 10-fold greater than the doses administered to other patients receiving a GM-CSF–secreting whole-cell vaccine.10-12 The pancreatic tumor cell vaccine has been genetically modified to secrete GM-CSF levels similar to those that have been reported in these other vaccine studies.22 As mentioned above, only one of the four patients with GM-CSF detected in their serum experienced systemic toxicity (Grover’s disease) that may be attributed to systemic GM-CSF levels. The detectable peak serum levels in this patient were still more than 40-fold lower than the levels expected to cause commonly reported side effects associated with the administration of systemic GM-CSF.41,42 To our knowledge, this is the first cytokine-secreting vaccine study to evaluate a measure of longevity of vaccine cells at the immunizing site. These data, together with data from preclinical models, would suggest that the allogeneic vaccine cells produce local concentrations of GM-CSF for a long enough period of time to induce a tumor-specific systemic immune response. Although this was a phase I trial, and so was not designed to test clinical benefit, we were nonetheless encouraged by the observed survival rates. The mean disease-free survival for this vaccine study was 13 months, although three patients were still disease-free at 25 to 30 months. In comparison, the mean reported disease-free survival time for all patients receiving adjuvant radiation and chemotherapy after pancreaticoduodenectomy is 6 to 9 months, with about 30% of patients disease-free at 1 year.5,7 In conclusion, we have identified a dose of 50 x 107 GM-CSF–secreting allogeneic pancreatic vaccine cells for phase II testing. This dose seems safe and may be associated with the induction of antitumor immune responses. Measurements of DTH responses to autologous tumor cells should be further evaluated as an in vivo marker of the induction of immune responses. DTH responses may also represent an intermediate marker of disease-free survival. Further studies are required to determine whether these promising effects on immune activation will translate into a true clinical benefit for patients with pancreatic adenocarcinoma. A phase II study is currently being designed.
Supported by grant no. RO1CA71806 from the National Cancer Institute (NCI), Bethesda, MD. E.M.J. received project funding from the Specialized Program of Research Excellence (SPORE) in Gastrointestinal Cancer from the National Cancer Institute (grant no. CA62924) to develop the vaccine approach that was tested in this study. D.L. was supported by training grant no. 5T32 CA09071-20 from the NCI. The investigators thank Christine Weber for her assistant in the clinic, Frederick Brancati, MD, for his careful critique of the manuscript and advice on data analysis, Mary Akinyemi for her assistance with data analysis, and Stephanie Porter for her administrative assistance. The investigators also thank Mr and Mrs Phillip Meyers and Roger Powell for their early support of this vaccine study.
1. Boring CC, Squires TS, Tong T: Cancer statistics. Cancer 41: 19-36, 1991
2.
Burris HA, Moore MJ, Cripps MC, et al: Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreatic cancer: A randomized trial. J Clin Oncol 15: 2403-2413, 1997 3. Lawrence T: Gemcitabine as a radiation sensitizer. Semin Oncol 22: 68-71, 1995[Medline]
4.
Blackstock AW, Bernard SA, Richards F, et al: Phase I trial of twice weekly gemcitabine and concurrent radiation in patients with advanced pancreatic cancer. J Clin Oncol 17: 2208-2212, 1999 5. Gastrointestinal Tumor Study Group: Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer. Cancer 59:2006-2010, 1987 6. Yeo CJ, Abrams RA, Grochow LB, et al: Pancreaticoduodenectomy for pancreatic adenocarcinoma: Postoperative adjuvant chemoradiation improves survival. Ann Surg 225: 621-636, 1997[Medline] 7. Abrams RA, Grochow LB: Adjuvant therapy with chemotherapy and radiation therapy in the management of carcinoma of the pancreatic head. Surg Clin North Am 75: 925-938, 1995[Medline]
8.
Greten TF, Jaffee EM: Cancer vaccines. J Clin Oncol 17: 1047-1060, 1999
9.
Dranoff G, Jaffee EM, Lazenby A, et al: Vaccination with irradiated tumor cells engineered to secrete murine granulocyte/macrophage colony-stimulating factor stimulates potent, specific, and long-lasting anti-tumor immunity. Proc Natl Acad Sci U S A 90: 3539-3543, 1993
10.
Simons JW, Jaffee EM, Weber C, et al: Bioactivity of human GM-CSF gene transduced autologous renal vaccines. Cancer Res 57: 1537-1546, 1997
11.
Soiffer R, Lynch T, Mihm M, et al: Vaccination with irradiated autologous melanoma cells engineered to secrete human granulocyte-macrophage colony-stimulating factor generates potent antitumor immunity in patients with metastatic melanoma. Proc Natl Acad Sci U S A 95: 13141-13146, 1998
12.
Simons JW, Mikhak B, Chang J-F, et al: Induction of immunity to prostate cancer antigens: Results of a clinical trial of vaccination with irradiated autologous prostate tumor cells engineered to secrete granulocyte-macrophage colony-stimulating factor using ex vivo gene transfer. Cancer Res 59: 5160-5168, 1999
13.
Boon T, Van Der Bruggen P: Human tumor antigens recognized by T lymphocytes. J Exp Med 183: 725-729, 1996 14. Rosenberg SA, Kawakami Y, Robbins PF, et al: Identification of the genes encoding cancer antigens: Implications for cancer immunotherapy. Adv Cancer Res 70: 145-177, 1996[Medline]
15.
Brossart P, Heinrich K, Stuhler G: Identification of HLA-A2 restricted T-cell epitopes derived from the MUC-1 tumor antigen for broadly applicable vaccine therapies. Blood 93: 4309-4317, 1999 16. Abrams SI, Khleif SN, Bergmann-Leitner ES, et al: Generation of stable CD4+ and CD8+ T cell lines from patients immunized with ras oncogene derived peptides reflecting codon 12 mutations. Cell Immunol 182: 137-151, 1997[Medline] 17. Laheru DA, Jaffee EM: Potential role of tumor vaccines in GI malignancies. Oncology 14: 245-255, 2000[Medline]
18.
Huang A, Golumbek P, Ahmadzadeh M, et al: The role of bone marrow derived cells in presenting MHC class I-restricted tumor antigens. Science 264: 961-965, 1994 19. Huang AY, Bruce AT, Pardoll DM, et al: In vivo cross-priming of MHC class I-restricted antigens requires TAP transporter. Immunity 4: 349-355, 1996[Medline]
20.
Toes RE, Blom RJ, Van Der Voort E, et al: Protective antitumor immunity induced by immunization with completely allogeneic tumor cells. Cancer Res 56: 3782-3787, 1996 21. Thomas MC, Greten T, Pardoll DM, et al: Vaccination with allogeneic tumor cells induces specific antitumor immunity. Hum Gene Ther 9: 835-843, 1998[Medline] 22. Jaffee EM, Schutte M, Gossett J, et al: Development and characterization of a cytokine secreting pancreatic adenocarcinoma vaccine from primary tumors for use in clinical trials. Cancer J Sci Am 4: 194-203, 1998[Medline] 23. Jaffee EM, Abrams RA, Cameron JL, et al: A phase I trial of lethally irradiated allogeneic pancreatic tumor cells transfected with the GM-CSF gene for the treatment of pancreatic adenomacarcinoma. Hum Gene Ther 9: 1951-1971, 1998[Medline] 24. Sokol JE: Measurement of delayed skin test responses. N Engl J Med 29: 501-503, 1995 25. McCune CS, O’Donnell RW, Marquis DM, et al: Renal cell carcinoma treated by vaccines for active specific immunotherapy: Correlation of survival with skin testing by autologous tumor cells. Cancer Immunol Immunother 32: 62-66, 1990[Medline] 26. Oren ME, Herberman RB: Delayed cutaneous hypersensitivity reactions to membrane extracts of human tumor cells. Clin Exp Immunol 9: 45-56, 1977
27.
Hoover HC Jr, Surdyke M, Dangel RB, et al: Delayed cutaneous hypersensitivity to autologous tumor cells in colorectal cancer patients immunized with an autologous tumor cell: Bacillus Calmette-Guerin vaccine. Cancer Res 44: 671-676, 1984
28.
Berd D, Maguire HC Jr, Mastrangelo MJ: Induction of cell-mediated immunity to autologous melanoma cells and regression of metastases after treatment with a melanoma cell vaccine preceded by cyclophosphamide. Cancer Res 46: 2572-2577, 1986 29. Berd DJ, Maguire JC Jr, McCue P, et al: Treatment of metastatic melanoma with an autologous tumor-cell vaccine: Clinical and immunologic results in 64 patients. J Clin Oncol 8: 1858-1867, 1990[Abstract] 30. Storer BE: Design and analysis of phase I clinical trials. Biometrics xx: 925-937, 1989 31. Yeo CJ, Cameron JL, Sohn TA, et al: Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma. Ann Surg 229: 613-624, 1999[Medline] 32. Chalet M, Grover R, Ackerman AB: Transient acantholytic dermatosis: A reevaluation. Arch Dermatol 113: 431-435, 1977[Abstract] 33. Davis MP, Dinneen AB, Landa N, et al: Grover’s disease: Clinicopathologic review of 72 cases. Mayo Clin Proc 74: 229-234, 1999[Medline] 34. Parsons JM: Transient acantholytic dermatosis (Grover’s disease): A global perspective. J Am Acad Dermatol 35: 653-666, 1996[Medline] 35. Slade MS, Simmons RL, Yunis E, et al: Immunodepression after major surgery in normal patients. Surgery 78: 363-372, 1975[Medline] 36. Eck HP, Betzler M, Schlag P, et al: Partial recovery of lymphocyte activity in patients with colorectal carcinoma after curative surgical treatment and return of plasma glutamate concentrations to normal levels. J Cancer Res Clin Oncol 116: 648-650, 1990[Medline] 37. Lennard TWJ, Shenton BK, Borzotta A, et al: The influence of surgical operations on components of the human immune system. Br J Surg 72: 771-776, 1985[Medline]
38.
Hung K, Hyashi R, Lafond-Walker A, et al: The central role of CD4+ T cells in the antitumor immune response. J Exp Med 188: 2357-2368, 1998 39. Steinman RM: Dendritic cells and immune-based therapies. Exp Hematol 24: 859-862, 1996[Medline] 40. Jaffee EM, Thomas MC, Huang AY, et al: Enhanced immune priming with spatial distribution of paracrine cytokine vaccines. J Immunother Emphasis Tumor Immunol 19: 176-183, 1996[Medline] 41. Lieschke GJ, Burgess AW: Granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor (part 1). N Engl J Med 327: 28-35, 1992[Medline] 42. Lieschke GJ, Burgess AW: Granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor (part 2). N Engl J Med 327: 99-106, 1992[Medline] Submitted December 22, 1999; accepted July 14, 2000. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
10 x 107 vaccine cells. These three patients also seemed to have had an increased disease-free survival time, remaining disease-free at least 25 months after diagnosis. 
