|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Adjuvant Immunotherapy of Resected, Intermediate-Thickness, Node-Negative Melanoma With an Allogeneic Tumor Vaccine: Impact of HLA Class I Antigen Expression on OutcomeByFrom the Vanderbilt University, Nashville, TN; Southwest Oncology Group Statistical Center and University of Washington, Seattle, WA; Karmanos Cancer Institute, Wayne State University, Detroit, and University of Michigan Comprehensive Cancer Center, Ann Arbor, MI; and University of Southern California and University of California Los Angeles Immunogenetics Center, Los Angeles, CA. Address reprint requests to Publications Specialist, Southwest Oncology Group (SWOG-9035), 14980 Omicron Dr, San Antonio, TX 78245-3217.
PURPOSE: An association between expression of  two of five HLA class I antigens (HLA-A2, HLA-A28, HLA-B44, HLA-B45, and HLA-C3; collectively called M5) and response to an allogeneic melanoma vaccine (Melacine; Corixa Corporation, Seattle, WA) has been described in stage IV melanoma. This study investigated whether class I antigen expression impacted relapse-free survival (RFS) after adjuvant therapy with this vaccine. PATIENTS AND METHODS: We performed class I (HLA-A, HLA-B, and HLA-C) serotyping on patients enrolled onto Southwest Oncology Group Trial 9035, a randomized, observation-controlled, phase III trial of adjuvant Melacine. All patients had clinically node-negative cutaneous melanoma (1.5 to 4.0 mm). Interactions between treatment and class I antigen expression were tested. Analyses involved all serotyped patients and were adjusted for tumor thickness, method of nodal staging, sex, ulceration, and primary tumor site.
RESULTS: HLA typing was performed on 553 (80%) of the 689 enrolled patients (294 vaccinated and 259 observed). Expression of
CONCLUSION: This prospective analysis indicates a highly significant benefit of adjuvant therapy with Melacine among patients expressing
IN 1992, THE SOUTHWEST Oncology Group (SWOG) Melanoma Committee initiated a randomized clinical trial evaluating an allogeneic melanoma vaccine (Melacine; Corixa Corporation, Seattle, WA) as an adjuvant for patients with primary melanoma of 1.5- to 4.0-mm thickness and no clinical evidence of lymph node involvement. This vaccine is a polyvalent lysate of two cell lines derived from melanoma metastases from two different patients.1 As such, it contains many potential antigens that could be recognized by the immune system.2 It remains unknown, however, which if any of these antigens would be important in causing regression of established melanoma or would impart protection from relapse after resection of all known disease. Earlier work by Mitchell et al1,3,4 indicated that this vaccine has the ability to induce cytotoxic T-lymphocyte immune responses to the vaccine itself and, in some cases, autologous tumor. More importantly, there were objective clinical responses to the vaccine in stage IV melanoma patients, including both partial and complete remissions. Although the total numbers of responses were few, the findings suggested that one or more antigens contained within the cell lysate could mediate regression of even bulky disease, and supported the evaluation of this vaccine in the adjuvant setting. It is well established that CD8+ T lymphocytes recognize processed protein antigens (peptides) expressed on the surface of antigen-presenting cells (APCs) in the context of HLA class I antigens.5 It is also known that peptides bind with widely varying specificity to different HLA antigens, and hence not all peptides will be efficiently presented by all patients’ APCs.6,7 Therefore, it is logical to consider the possibility that host HLA class I antigens are of importance to the efficacy of tumor vaccines. It is also possible that the HLA antigens expressed by tumor cell lines that were used to form the vaccine could affect the degree of host immune response, such that patients who matched one or more HLA antigens with the component tumor cell lines would have an improved outcome with vaccine treatment. Mitchell et al8 analyzed the HLA class I expression of 66 patients with advanced melanoma enrolled onto phase I/II Melacine trials. This included 12 patients who the investigator felt demonstrated some evidence of tumor "regression." Most had objective partial and complete remissions of all disease, whereas others had minor clinical responses and/or prolonged stable disease. The analysis explored the role of many HLA class I antigens, including the impact of matching antigens with those in the tumor cell lines in the vaccine. From the outset, their analysis was intended to be hypothesis-generating and exploratory in nature. It revealed an association between antitumor efficacy as defined by Mitchell et al and the host expression of two or three (no patient had > three) of the following class I antigens: HLA-A2, HLA-A28, HLA-B44, HLA-B45, and HLA-C3 (P = .0057). The strongest association with clinical response was related to expression of HLA phenotypes A2 and C3. The phase III clinical trial that was initiated and completed under the auspices of the SWOG is described in detail in the accompanying article.9 The patient eligibility, treatment plan, clinical follow-up, and overall clinical findings in terms of disease-free survival and toxicity from vaccine are presented in detail in that accompanying report. The trial initiated patient enrollment in April 1992 and closed in November 1996. In September 1994, in recognition of the findings of Mitchell et al’s analysis in stage IV patients, the protocol was amended to mandate blood sample collections for serotyping of HLA-A, HLA-B, and HLA-C from all patients who were to be subsequently enrolled onto the trial. In addition, all treating physicians were encouraged to have patients who were previously enrolled onto the trial reconsent to obtain HLA class I typing on them as well. At the time the protocol was amended, we designed several prospective analyses of the relationship between HLA antigen expression and outcome of both treatment and control arm patients in this large cohort of T3N0 (1.5 to 4.0 mm) melanoma patients. We specifically planned to evaluate whether the expression of certain HLA antigens alone or in combination was predictive of treatment benefit. This report describes the results of the prospectively planned analysis of the 553 patients who were successfully HLA class I serotyped among the 689 total patients entered onto the cooperative group, randomized, controlled trial of Melacine in localized, intermediate-thickness cutaneous melanoma (SWOG-9035).
Clinical Trial Design Eligible patients were 18 years and older with completely resected primary cutaneous melanoma that measured 1.5 to 4.0 mm in thickness or Clark’s level IV (if thickness was unavailable), met all on-study eligibility criteria, were within 56 days of definitive surgery for the treatment of melanoma, and had no evidence of regional or metastatic disease. Patients had a physical examination documenting complete resection of their primary lesion and no regional lymphadenopathy and no signs or symptoms of metastatic disease. Chest x-ray had to be without evidence of metastatic disease. Patients were randomly assigned to either observation or 2 years of adjuvant therapy with a polyvalent, allogeneic melanoma cell lysate vaccine. The randomization was stratified to balance for sex, tumor thickness (1.5 to 3.0 mm, 3.1 to 4.0 mm, or Clark’s level IV thickness unknown), and method of nodal staging (surgical or clinical only). Tumor ulceration, anatomic location of the primary tumor (extremity or nonextremity), and age were recorded as potential prognostic factors but not used in the stratification at the time of randomization. The Melacine vaccine is composed of an allogeneic melanoma cell lysate plus the immunologic adjuvant DETOX (detoxified Freund’s adjuvant, containing mycobacterial cell wall skeleton plus monophosphoryl lipid A; Corixa). The vaccine was administered as two intramuscular injections of cell lysate mixed with DETOX per vaccination. The vaccines were delivered as four 6-month cycles, each cycle consisting of 10 vaccinations (on weeks 1, 2, 3, 4, 6, 8, 12, 16, 20, and 24) followed by a 3-week rest.
Methods for HLA Class I Serologic Phenotyping A selected panel of antisera characterized for HLA class I antigens was used in an antigen/antibody test, with lymphocytes as the target cells. After lymphocytes and antisera were incubated, complement was added and fixed at the sites of antigen/antibody reaction, thus changing the permeability of the lymphocyte membrane. Cells were coated with 5-carboxyl fluorescein diacetate (Molecular Probes, Inc, Eugene, OR) before they were added to the antisera. If an antigen/antibody reaction took place, ethidium bromide contained in the quench was intercalated with the DNA in the cells. Live cells appeared green when the trays were read under fluorescence microscopy and dead cells appeared red. Quench was used to provide a black background against which the green and red fluorescence could be seen easily. Antigen/antibody reactions were evaluated by the change in viability of the target cells (lymphocytes) between the negative control and the antiserum in each well of the microtest tray. A standard system was used to score percentage of cell killing. Antigens were assigned on the basis of the reactivity of the test lymphocytes against a panel of known HLA class I antisera. HLA types were assigned by analyzing the pattern of lymphocyte reaction against multiple, independent antisera for each specificity.
Statistical Analysis For the HLA analysis, the database consisted of 553 patients who were HLA typed. The subset of the 383 patients prospectively registered on or after the HLA amendment date of September 1, 1994, was analyzed separately and confirmed the results. All disease-free survival comparisons were performed using Cox model analysis12 adjusting for stratification factors (tumor thickness, nodal staging method, and sex) as well as ulceration and primary site location, which were recognized as prognostic factors in the overall trial results.9 Tests for interaction between treatment type and one or a combination of the HLA antigens were performed. Testing interactions in this setting amounted to testing whether the effect of one factor remained the same in the presence or absence of the other (eg, whether the vaccine treatment effect remained the same in patients with HLA-A2 antigen expression as in patients without HLA-A2 antigen expression). With generally small numbers and possible imbalances between the factorial cells, a nominal value of P = .10 was chosen to minimize the possibility of missing a potentially significant interaction (Green et al, manuscript submitted for publication). In the case of a potentially significant interaction indicating differential vaccine treatment effects between HLA categories (eg, a stronger vaccine effect among HLA-A2–positive patients than among HLA-A2–negative patients) or differential HLA category effects between treatment arms (eg, a stronger HLA-A2–positive v HLA-A2–negative effect among vaccine-treated patients than among observation patients), further specific analyses were performed to identify the actual differences. Of primary interest was the vaccine versus observation comparison within each particular HLA category. Comparisons among HLA categories within a particular treatment were of secondary interest. To adjust for multiple comparisons, a nominal value of P = .01 was used to indicate a significant result for each analysis. The following hypotheses were evaluated using the above prospectively defined analysis strategy: Hypothesis 1. Vaccine-arm patients expressing HLA class I antigens matching the vaccine’s HLA class I antigen types will have improved outcome. The HLA expression pattern of the melanoma cell lines used for the tumor cell lysate has been recently characterized by molecular and serologic analysis (data provided by K. von Eschen, Corixa). This analysis demonstrated that cell line 1 (Mel S) expresses HLA-A28, HLA-A31, HLA-B51, HLA-B60, and HLA-C3, whereas line 2 (Mel D) expresses HLA-A11, HLA-B60, and HLA-C3. We analyzed treatment by the extent of match to the HLA class I phenotype of the vaccine tumor cell lines (HLA-All, HLA-A28, HLA-A31, HLA-B51, HLA-B60, and HLA-C3). The extent of matching was categorized as zero versus one versus two or more matches.
Hypothesis 2.
Vaccine-arm patients expressing two or more of the five HLA antigens (HLA-A2, HLA-A28, HLA-B44, HLA-B45, and HLA-C3) associated with clinical response in a prior study of Melacine in advanced melanoma8 will have improved outcome. We analyzed treatment effect by the extent of patient expression of the HLA class I antigens previously identified by Mitchell et al8 as predicting response in stage IV disease (HLA-A2, HLA-A28, HLA-B44, HLA-B45, and HLA-C3; collectively referred to as M5). The extent of matching was categorized as zero or one versus Hypothesis 3. Patient expression of HLA-A2 and/or HLA-C3 will be of great importance in improving the vaccine treatment efficacy associated with the M5 antigens. We analyzed treatment effect by the extent of patient expression of HLA-A2 and/or HLA-C3, to determine whether their influence was predominant among the M5, as suggested by Mitchell et al.8 The extent of expression was categorized as expression of HLA-A2 and/or HLA-C3 versus expression of neither HLA-A2 nor HLA-C3. Hypothesis 4. Vaccine-treated patients expressing HLA class I antigens known to bind and present melanoma-associated antigens expressed by Melacine (MAGE-1, MAGE-2, MAGE-3, tyrosinase, gp100, MART-1, and TRP-1; Table 1) will have improved outcome. A number of melanoma-associated proteins have been defined, and are known to contain peptide sequences which, when processed, bind to specific HLA class I molecules and induce CD8+ T lymphocytes capable of recognizing melanoma tumors.2,13-20 We analyzed treatment effect by the expression of specific single HLA class I antigens that are known to bind and display peptides from tumor antigens present in the polyvalent melanoma cell lysate (data provided by K. von Eschen, Corixa). The seven HLA antigen types evaluated were HLA-A1, HLA-A2, HLA-A3, HLA-A24, HLA-A31, HLA-B44, and HLA-B45. Each HLA antigen was evaluated independently, even though there is almost certainly some linkage dysequilibrium between HLA-A, HLA-B, and HLA-C antigens in the general population.21
Demographics of the Study Population A total of 553 (80%) of the entire 689 enrolled patients gave consent and underwent HLA class I serotyping. The total was composed of 383 (94%) of the 409 patients entered on or after September 1, 1994, who were all prospectively typed, and 170 (61%) of the 280 patients entered before September 1, 1994, who were retrospectively typed after earlier entry onto the protocol. There were 294 patients in the Melacine-treated arm and 259 in the observation control arm included in this analysis, and their characteristics are listed in Table 2. This analysis seems to have preferentially excluded some patients relapsing within the initial 2 years of enrollment (1992 to 1994) before the institution of HLA phenotyping. Thus, the overall outcome of the 553 patients is slightly better than all 689 patients randomized (Fig 1). To check whether this fact potentially biased the results, all analyses were repeated and all were confirmed in the subset of 383 prospectively typed patients. Therefore, only data on the entire cohort of all 553 HLA-typed patients are presented.
The overall results of SWOG-9035 are presented elsewhere.9 SWOG-9035 demonstrated no improvement in relapse-free survival for patients randomized to receive vaccine therapy for 2 years compared with observation. For the subset of 553 patients who underwent HLA serotyping, relapse-free survival adjusting for the design-specified stratification factors (tumor thickness, nodal staging method, and sex) plus anatomic site and ulceration was no better for patients randomized to vaccine treatment than observation (P = .11).
Hypothesis 1.
There was no interaction between treatment (vaccine or observation) and extent of patient match with HLA antigens (zero v one v Hypothesis 2. As shown in Table 3, there was a statistically significant interaction between treatment (vaccine or observation) and expression of two or more of the M5 (HLA-A2, HLA-A28, HLA-B44, HLA-B45, and HLA-C3) antigen types (P = .003). The following additional analyses were therefore performed:
On the basis of these analyses, hypothesis 2 is strongly supported, indicating an improved outcome for the vaccine arm but not the observation arm patients expressing two or more of the M5 antigens. Hypothesis 3. Although the overall expression of the various HLA class I antigens in this melanoma patient cohort is not the focus of this article, it is obvious that the M5 HLA antigens are not all expressed in a similar percentage of the patients (Table 4). Although HLA-A2, HLA-B44, and HLA-C3 are each relatively prevalent (HLA-A2 being the most frequent antigen in this study population as well as in most analyses of the North American white population22), HLA-A28 and B45 are both fairly infrequent.
We first categorized patients as having expressed one or both of HLA-A2 or HLA-C3 versus neither HLA-A2 nor HLA-C3. The test for interaction between treatment and HLA-A2 and/or HLA-C3 expression yielded a significant result (P = .009) (Table 5), prompting subset analyses. Vaccine arm patients who were HLA-A2–positive and/or HLA-C3–positive (n = 178) had 5-year relapse-free survival of 77% compared with (1) a 5-year relapse-free survival of 63% for the vaccine arm patients who were HLA-A2–negative and HLA-C3–negative (n = 116, P = .002); (2) a 5-year relapse-free survival of 64% for the observation patients who were HLA-A2–positive and/or HLA-C3–positive (n = 145, P = .004); and (3) a 5-year relapse-free survival of 66% for the observation patients who were HLA-A2–negative and HLA-C3–negative (n = 114, P = .03) (Table 5 and Fig 4). Among all those who expressed HLA-A2, vaccine arm patients had better relapse-free survival than observation patients (P = .009) (Table 5 and Fig 5A). Among those patients who did not express HLA-A2, vaccine and observation patients had similar relapse-free survival. Among all those who expressed HLA-C3, vaccine arm patients had marginally better relapse-free survival than observation patients (P = .02) (Table 5 and Fig 5B).
On the basis of this analysis, hypothesis 3 was accepted. Expression of either HLA-A2 or HLA-C3 or both was of great importance in improving the vaccine treatment efficacy. A total of 323 of the 553 serotyped patients (58%) were HLA-A2–positive and/or HLA-C3–positive. Hypothesis 4. A number of melanoma-associated proteins have been defined and are known to contain peptide sequences that, when processed, bind to specific HLA class I molecules and stimulate CD8+ T lymphocytes capable of recognizing melanoma tumors.2,13-20,23-25 At least eight specific HLA alleles are known to bind peptides of melanoma-associated antigens documented to be present in the Melacine vaccine. There was no available serotyping of Cw16, and cohorts expressing HLA-A31 and HLA-B45 (5.8% and 1.1% of the study population, respectively) were considered too small for valid analysis as single entities, leaving five HLA antigens available for testing of interaction with treatment. As part of the analysis presented above for hypothesis 3, we found that vaccine arm patients expressing HLA-A2 had superior relapse-free survival compared with observation patients expressing HLA-A2. A similar treatment effect was not evident among HLA-A2–negative patients. None of the other four antigens examined (HLA-A1, HLA-A3, HLA-A24, and HLA-B44) differentially impacted relapse-free survival between vaccine and observation patients (Table 6).
All analyses described above were repeated in the subset of 383 prospectively typed patients. No substantive differences were evident (data not shown).
HLA class I molecules on the surface of host APCs present protein antigens as processed peptides to CD8+ T lymphocytes.2 Therefore, it is logical to postulate that individuals expressing different HLA class I antigens may respond very differently to tumor lysate vaccines composed of various proteins that need to be processed and presented by host APCs in order to induce an effective antitumor response. Melacine is representative of such a tumor lysate vaccine, composed of two melanoma cell lines (Mel S and Mel D) that are known to contain at least seven protein antigens that can stimulate CD8+ HLA class I restricted T-cell responses.13-20,23-25 These antigens include MAGE-1, MAGE-2, MAGE-3, TRP-1, tyrosinase, gp100, and MART-1. Reports of objective responses, although infrequent, in stage IV melanoma in multicenter phase I and II clinical trials led the SWOG to commence a randomized, controlled phase III trial of Melacine as adjuvant therapy for resected, intermediate-thickness, node-negative cutaneous melanoma in 1992.1,3,26,27 This clinical trial showed no benefit for the vaccine arm over observation (P = .51). Early on during the trial, we recognized that HLA phenotyping of patients could potentially define patient groups who might better take advantage of a complex vaccine and hence demonstrate improved outcomes with vaccine treatment. We were also aware of the findings of Mitchell et al,8 who in 1992 showed an association between expression of specific HLA class I antigens and the likelihood of a clinical response in stage IV melanoma patients receiving Melacine. Mitchell et al’s evaluations were retrospective, included only 12 responders in all, and included as "responders" some patients whose responses did not meet established criteria for partial or complete remissions. Nonetheless, they served to generate hypotheses that could be and were tested prospectively in our trial of earlier stage patients receiving the same vaccine as adjuvant therapy.
The results of our analysis show remarkable similarity to those reported by Mitchell et al and emphasize the benefit of vaccine within a patient population defined by expression of specific HLA class I antigens including HLA-A2, HLA-A28, HLA-B44, HLA-B45, and HLA-C3. The expression of HLA-A2 and HLA-C3 seemed to be the dominant antigens in this effect. Our analysis was prospective, with a restricted number of well-defined hypotheses including an evaluation of the validity of the findings of Mitchell et al, and included all 553 HLA-serotyped patients on an intent-to-treat basis. The findings were all confirmed in the subset of 383 patients who were HLA typed prospectively at enrollment onto the protocol. The clinical trial had a well-defined and objective end point (relapse-free survival), and was adjusted for stratification factors (tumor thickness, method of nodal staging, and sex) and recognized prognostic factors (ulceration and primary site location) in a multivariate analysis. When a test for interaction between treatment (vaccine v observation) and specific HLA status was significant and further subset analyses were performed, we only considered P values Our analysis does have several limitations. We did not have molecular HLA class I allele subtyping available, which could provide more powerful and insightful findings. The HLA-A2 allele itself has over 20 molecularly defined subtypes.29 Although most subtypes seem to have a similar spectrum of binding capacity for peptide antigens, others may differ greatly. In addition, HLA class II antigens (DR, DP, and DQ) were not analyzed as part of the present study. Certainly, CD4+ T lymphocytes directed at tumor antigens that are restricted by these HLA class II alleles may be important to effective antitumor immune responses.30-33 Little information is currently available regarding the impact of class II expression from other adjuvant vaccine studies. These limitations notwithstanding, the statistical significance of our results and their prospective and hypothesis-driven nature make them important and powerful findings. The P value of 0.0005 for an advantage in relapse-free survival among vaccine-treated patients expressing at least two of the "Mitchell five" antigens was observed in 175 patients (97 in the vaccine group and 78 in the observation group). This Cox regression model included adjustments for all the stratification and prognostic factors we know to be important. The HLA antigen categorization of zero to one versus two or more of the Mitchell five antigens was predefined by Mitchell et al, not generated by subset analysis of our data. Furthermore, extending the evaluation to look specifically at the roles of HLA-A2 and HLA-C3 (again previously identified by Mitchell et al) revealed a benefit among those patients in the vaccine arm expressing one or both of HLA-A2 and/or HLA-C3 (P = .002 compared with the HLA-A2–negative/HLA-C3–negative vaccine patients). In the observation arm, expression of HLA-A2 and/or HLA-C3 made no difference in outcome. This latter finding refutes the possibility that HLA-A2 and HLA-C3 patients simply have a better outcome independent of treatment administered. Assuming these results are convincing, then what is their meaning? How do we use the findings to shed light on future vaccine development? HLA-A2 and HLA-C3 likely bind and present numerous melanoma-derived peptides. Many tumor-derived peptides that are capable of binding and being presented by HLA-A2 have already been identified among melanoma tumors.13,34-37 These include six of the seven melanoma-associated protein antigens known to be in Melacine. Furthermore, a MAGE-1 peptide has recently been shown to be presented by HLA-C3 to CD8+ T cells.25,38 The fact that the most dramatic effect of the vaccine was seen not with one, but rather when two or more of the key HLA antigens were present, implies a role for a minimum of several peptides binding to one or more HLA class I antigens. It is likely that the presentation of tumor peptides by more than a single HLA antigen will lead to activation of a larger repertoire of effector T cells and lessen the likelihood for tumor escape by antigen or HLA loss. Although Melacine has been shown to express MAGE-1, MAGE-2, MAGE-3, tyrosinase, gp-100, TRP-1, and MART-1, and any or all of these could play a role in the effectiveness of the vaccine in the HLA-A2/HLA-C3–expressing patients, hitherto undefined antigens presented by either HLA-A2 or HLA-C3 could also be responsible. In this regard, it is important to note that other investigators using other vaccines, cytokines (eg, interleukin-2), or tumor-infiltrating lymphocytes have not reported the same results.39-41 Hoon et al41 administered polyvalent whole cell irradiated melanoma vaccines to 669 stage I to IV melanoma patients whose HLA expression was analyzed. They reported a benefit for vaccine treatment in HLA-A25–positive patients, as well as a favorable outcome of patients whose HLA expression matched the HLA phenotype expressed by the three melanoma cells lines that composed the vaccine. In their study, no benefit was observed in the subset of patients expressing HLA-A2 or HLA-C3. To further investigate our findings, one could study and compare patients from the observation and Melacine arm who express HLA-A2 or HLA-C3 for in vitro CD8+ T-cell responses to peptides derived from melanoma proteins within Melacine known to bind to these HLA antigens. Those who were vaccinated and are still disease-free would be predicted to have a much better response to important antigens. However, the most definitive approach would involve the conduct of a phase III adjuvant trial with Melacine in only HLA-A2 and/or HLA-C3 patients. Although HLA-A2–positive and/or HLA-C3–positive patients represented close to 60% of all HLA-typed patients in SWOG-9035, it is not clear whether such a study could be conducted. The surgical care for cutaneous melanoma has changed since SWOG-9035 was conceived and began accrual. Most patients are now staged with sentinel lymph node biopsy. Hence, today’s intermediate-thickness, clinically lymph node–negative patient likely has a superior prognosis, and such a study would need to enroll much larger numbers of patients than were entered onto SWOG-9035. A trial of Melacine in higher risk (node-positive patients) would likely involve comparison against high-dose interferon alfa-2b for a year, not observation or placebo, and may include combinations with high-dose interferon alfa-2b, which could complicate any analysis of the impact of HLA class I expression. In the meantime, our results certainly justify continued investigation of allogeneic, polyvalent melanoma vaccines in the adjuvant therapy of cutaneous melanoma. We must learn more about what antigens within Melacine are potentially presented by HLA-A2 and HLA-C3 antigens. Identifying these antigens and developing new approaches to effectively deliver them may be critical to producing more effective and better defined tumor vaccines. HLA phenotyping should be a component of all future clinical trials designed to investigate protein-based vaccines in the adjuvant therapy of cancer.
Supported in part by Public Health Service Cooperative Agreement grant nos. CA38926, CA32102, CA20319, CA27057, CA58723, CA22433, CA58861, CA04920, CA46113, CA46441, CA37981, CA7648, CA04919, CA35281, CA58686, CA35176, CA35090, CA16385, CA35119, CA35262, CA12644, CA46136, CA14028, CA45450, CA42777, CA45377, CA35178, CA46282, CA35192, CA58416, CA96429, CA35431, CA13612, CA45807, CA74647, CA58415, CA58348, CA12213, CA45560, CA58882, CA76447, CA46368, CA35261, CA52386, CA76462, CA67663, CA52654, and CA63845 from the National Cancer Institute, Department of Health and Human Services, Bethesda, MD. We thank the SWOG Operations Office and Statistical Center, particularly Dana Sparks, Susan Myers, Claire Chapdu, and Camille White. We are very grateful to Corixa Corporation, especially Ken Von Eschen and Mac Cheever, for their support of the HLA typing on this trial and insights they provided.
1. Mitchell MS, Kan-Mitchell J, Kempf RA, et al: Active specific immunotherapy for melanoma: Phase I trial of allogeneic lysates and a novel adjuvant. Cancer Res 48: 5883-5893, 1988 2. Rosenberg SA: A new era for cancer immunotherapy based on the genes that encode cancer antigens. Immunity 10: 281-287, 1999[CrossRef][Medline] 3. Mitchell MS, Boswell WD, Harel W, et al: Active specific immunotherapy for melanoma. J Clin Oncol 8: 856-869, 1990[Abstract] 4. Harel W, Goedegebuure PS, LeMay LG, et al: Melanoma-specific lysis by cloned CD4+ and CD8+ T cells from actively immunized melanoma patients. Vaccine Res 2: 41-53, 1993 5. Germain RN: MHC-dependent antigen processing and peptide presentation: Providing ligands for T lymphocyte activation. Cell 76: 287-299, 1994[CrossRef][Medline] 6. Rammensee HG, Friede T, Stevanoviic S: MHC ligands and peptide motifs: First listing. Immunogenetics 41: 178-228, 1995[Medline] 7. Kubo RT, Sette A, Grey HM, et al: Definition of specific peptide motifs for four major HLA-A alleles. J Immunol 152: 3913-3924, 1994[Abstract] 8. Mitchell MS, Harel W, Groshen S: Association of HLA phenotype with response to active specific immunotherapy of melanoma. J Clin Oncol 10: 1158-1164, 1992[Abstract]
9.
Sondak VK, Liu P-Y, Tuthill RJ, et al: Adjuvant immunotherapy of resected, intermediate-thickness node-negative melanoma with an allogeneic tumor vaccine: I. Overall results of a randomized trial of the Southwest Oncology Group. J Clin Oncol 20: 2058-2066, 2002 10. Terasaki PI: Histocompatibility testing, in Danovitch GM (ed): Handbook of Kidney Transplantation. Boston, MA, Little, Brown, 1991, pp 43-66 11. Hopkins KA: The basic lymphocyte microcytotoxicity tests, in Phelan D, Mickelson E, Noreen H, et al (eds): ASHI Laboratory Manual (ed 3). Lenexer, KS, American Society for Histocompatability and Immunogenetics, Section I.B.1.1-I.B.1.13 12. Cox DR: Regression models and life-tables. J R Stat Soc (B) 34: 187-220, 1972 13. Rivoltini L, Loftus DJ, Barracchini K, et al: Binding and presentation of peptides derived from melanoma antigens MART-1 and glycoprotein-100 by HLA-A2 subtypes: Implications for peptide-based immunotherapy. J Immunol 156: 3882-3891, 1996[Abstract]
14.
Traversari C, van der Bruggen P, Luescher IF, et al: A nonapeptide encoded by human gene MAGE-1 is recognized on HLA-A1 by cytolytic T lymphocytes directed against tumor antigen MZ2-E. J Exp Med 176: 1453-1457, 1992 15. Kang X, Kawakami Y, el-Gamil M, et al: Identification of a tyrosinase epitope recognized by HLA-A24-restricted, tumor-infiltrating lymphocytes. J Immunol 155: 1343-1348, 1995[Abstract]
16.
Luescher IF, Romero P, Kuznetsov D, et al: HLA photoaffinity labeling reveals overlapping binding of homologous melanoma-associated gene peptides by HLA-A1, HLA-A29, and HLA-B44. J Biol Chem 271: 12463-12471, 1996 17. Brichard VG, Herman J, Van Pel A, et al: A tyrosinase nonapeptide presented by HLA-B44 is recognized on a human melanoma by autologous cytolytic T lymphocytes. Eur J Immunol 26: 224-230, 1996[Medline] 18. Robbins PF, el-Gamil M, Li YF, et al: Cloning of a new gene encoding an antigen recognized by melanoma-specific HLA-A24-restricted tumor-infiltrating lymphocytes. J Immunol 154: 5944-5950, 1995[Abstract] 19. Herman J, van der Bruggen P, Luescher IF, et al: A peptide encoded by the human MAGE3 gene and presented by HLA-B44 induces cytolytic T lymphocytes that recognize tumor cells expressing MAGE3. Immunogenetics 43: 377-383, 1996[Medline]
20.
Gaugler B, Van den Eynde B, van der Bruggen P, et al: Human gene MAGE-3 codes for an antigen recognized on a melanoma by autologous cytolytic T lymphocytes. J Exp Med 179: 921-930, 1994 21. MacLean CJ, Martin RB, Sham PC, et al: The trimmed-haplotype test for linkage disequilibrium. Am J Hum Genet 66: 1062-1075, 2000[CrossRef][Medline] 22. Mori M, Beatty PG, Graves M, et al: HLA gene and haplotype frequencies in the North American population: The National Marrow Donor Program Donor Registry. Transplantation 64: 1017-1027, 1997[Medline] 23. Visseren MJ, van der Burg SH, van der Voort EI, et al: Identification of HLA-A*0201-restricted CTL epitopes encoded by the tumor-specific MAGE-2 gene product. Int J Cancer 73: 125-130, 1997[CrossRef][Medline]
24.
Wang RF, Robbins PF, Kawakami Y, et al: Identification of a gene encoding a melanoma tumor antigen recognized by HLA-A31-restricted tumor-infiltrating lymphocytes. J Exp Med 181: 799-804, 1995
25.
Farina C, van der Bruggen P, Boel P, et al: Conserved TCR usage by HLA-Cw*1601-restricted T cell clones recognizing melanoma antigens. Int Immunol 8: 1463-1466, 1996 26. Elliott GT, Mcleod RA, Perez J, et al: Interim results of a phase II multicenter clinical trial evaluating the activity of a therapeutic allogeneic melanoma vaccine (theraccine) in the treatment of disseminated malignant melanoma. Semin Surg Oncol 2: 41-53, 1993 27. Mitchell MS, von Eschen KB: Phase III trial of Melacine melanoma theraccine versus combination chemotherapy in the treatment of stage IV melanoma. Proc Am Soc Clin Oncol 16: 494a, 1997 (abstr 1778) 28. Miller RG: Simultaneous Statistical Inference. New York, NY, Springer-Verlag, 1981
29.
Barouch D, Friede T, Stevanovic S, et al: HLA-A2 subtypes are functionally distinct in peptide binding and presentation. J Exp Med 182: 1847-1856, 1995 30. Pardoll DM, Topalian SL: The role of CD4+ T cell responses in antitumor immunity. Curr Opin Immunol 10: 588-594, 1998[CrossRef][Medline]
31.
Topalian SL, Gonzales MI, Parkhurst M, et al: Melanoma-specific CD4+ T cells recognize nonmutated HLA-DR-restricted tyrosinase epitopes. J Exp Med 183: 1965-1971, 1996
32.
Wang RF, Wang X, Atwood AC, et al: Cloning genes encoding MHC class II-restricted antigens: Mutated CDC27 as a tumor antigen. Science 284: 1351-1354, 1999
33.
Zarour HM, Kirkwood JM, Kierstead LS, et al: Melan-A/MART-1(51-73) represents an immunogenic HLA-DR4-restricted epitope recognized by melanoma-reactive CD4(+) T cells. Proc Natl Acad Sci U S A 97: 400-405, 2000
34.
Coulie PG, Brichard V, Van Pel A, et al: A new gene coding for a differentiation antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanomas. J Exp Med 180: 35-42, 1994 35. Wolfel T, Van Pel A, Brichard V, et al: Two tyrosinase nonapeptides recognized on HLA-A2 melanomas by autologous cytolytic T lymphocytes. Eur J Immunol 24: 759-764, 1994[Medline] 36. Rivoltini L, Loftus DJ, Barracchini K, et al: Binding and presentation of peptides derived from melanoma antigens MART-1 and glycoprotein-100 by HLA-A2 subtypes: Implications for peptide-based immunotherapy. J Immunol 156: 3882-3891, 1996 37. Schneider J, Brichard V, Boon T, et al: Overlapping peptides of melanocyte differentiation antigen Melan-A/MART-1 recognized by autologous cytolytic T lymphocytes in association with HLA-B45.1 and HLA-A2.1. Int J Cancer 75: 451-458, 1998[CrossRef][Medline] 38. van der Bruggen P, Szikora JP, Boel P, et al: Autologous cytolytic T lymphocytes recognize a MAGE-1 nonapeptide on melanomas expressing HLA-Cw*1601. Eur J Immunol 24: 2134-2140, 1994[Medline]
39.
Marincola FM, Venzon D, White D, et al: HLA association with response and toxicity in melanoma patients treated with interleukin 2-based immunotherapy. Cancer Res 52: 6561-6566, 1992 40. Marincola FM, Shamamian P, Rivoltini L, et al: HLA associations in the antitumor response against malignant melanoma. J Immunother 18: 242-252, 1995[Medline]
41.
Hoon DS, Okamoto T, Wang HJ, et al: Is the survival of melanoma patients receiving polyvalent melanoma cell vaccine linked to the human leukocyte antigen phenotype of patients? J Clin Oncol 16: 1430-1437, 1998 Submitted April 16, 2001; accepted January 9, 2002. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
.01 to be "significant." This approach is akin to Fisher’s least significance difference approach to multiple comparisons, wherein a global test is conducted first. Pairwise comparisons follow only when the global test indicates the existence of differential treatment effects between the patient subsets.
