Faculty Bibliography

BACKGROUND: In the development of an antimelanoma vaccine, a critical factor is the identification of antigens that induce a strong immune response in humans and that are expressed by melanoma cells in vivo. The aim of this study was to identify candidate antigens for such vaccine. METHODS: Sixty-nine patients with surgically resected melanomas (American Joint Commission on Cancer [AJCC] stage III) were immunized with a polyvalent vaccine containing multiple melanoma antigens. Antimelanoma antibodies generated in the patients' sera were used as probes to identify the melanoma antigens that are immunogenic in humans and that are expressed on the tumor tissue in vivo. Such responses were determined by an immunoblotting assay that employed an antigen source prepared from membrane fractions of freshly excised melanoma tissue. RESULTS AND CONCLUSIONS: Vaccine treatment stimulated antibody responses in 35 (51%; 95% confidence interval [CI] = 39%-63%) of 69 sequentially enrolled patients. The antibodies were directed to one or more antigens with molecular masses of 45, 59, 68, 79, 89, 95, and/or 110 kd. The most immunogenic antigens were p110 and p68, which induced responses in 33% (95% CI = 22%-44%) and 25% (95% CI = 15%-35%) of patients, respectively. Both antigens were commonly expressed on different melanomas, but they were absent on autologous normal tissue and on an unrelated allogeneic tumor. All the above antigens are attractive candidates for vaccine construction

PURPOSE: Didemnin B is a novel marine natural product cyclic depsipeptide containing unusual amino acid moieties. This agent demonstrates promising preclinical antitumor activity, including activity against B16 melanoma and against melanoma isolates in the human tumor stem cell assay. METHODS: We conducted a phase II study of Didemnin B, given in Cremophor, at a starting dose of 4.2 mg/m2/IV q 28 days. Patients with measurable metastatic or advanced malignant melanoma were eligible. All patients were previously untreated with chemotherapy and had performance status 0 or 1. Doses were escalated to 4.9 and 5.6 mg/m2 in cycles 2 and 3, respectively. RESULTS: Nineteen patients were entered and treated with a median of one cycle per patient. Eight of these patients went off study for toxicity including 7 with anaphylactoid reactions in the first or second cycle. One patient went off study after 3 cycles with severe myopathy, a newly described toxicity. Two were not evaluable for response and five were considered stable, including one patient with a transient PR of soft tissue disease in the first cycle. Another patient had stable disease for twelve cycles before progressing and one went off study electively after 3 cycles, for a total of 7 patients with stable disease. One patient with a measurable partial remission (PR) and went off study after three cycles due to severe myopathy, a then newly-described toxicity. No hematologic toxicity was seen. Nausea and vomiting were controlled with anti-emetics. CONCLUSIONS: This study was indeterminate with respect to the activity of Didemnin B in melanoma. Signs of activity were seen, particularly in soft tissue masses, though a large number of patients could not be evaluated fully for activity due to the occurrence of anaphylactoid reactions. This study does not preclude a clinically important level of activity for Didemnin B

A critical requirement for cancer vaccines is that they stimulate CD8+ T cell responses. In this study, we tested the ability of a polyvalent melanoma vaccine to induce CD8+ T cell responses to the melanoma associated antigens MAGE-3 and Melan A/MART-1. Fifteen HLA-A2+ patients with resected malignant melanoma were immunized with the vaccine s.c. every 2-3 weeks. CD8+ T cells in peripheral blood reacting to HLA-A2 restricted epitopes on MAGE-3 (FLWGPRALV) and Melan A/MART-1/(AAGIGILTV) were quantitated using a filter spot assay at baseline and following 4 immunizations. Vaccine immunization induced CD8+ T cells reacting to one or both of these peptides in 9 of the 15 (60%) patients. These cells were CD8+ and HLA-A2 restricted, as reactivity was abrogated by monoclonal antibodies (MAbs) to CD8 and class I HLA, but not by anti-CD4. All responding patients remained recurrence-free for at least 12 months (median 15 months, range 12 to >21 months), whereas melanoma recurred within 3-5 months in non-responders. The differences in outcome were unrelated to differences in disease severity or overall immunological competence between responders and non-responders. Our results demonstrate directly that MAGE-3 and Melan A/MART-1 can stimulate CD8+ T cell responses in humans, and suggest that these responses are protective and surrogate markers of vaccine efficacy

Topoisomerase 1 (topo-1) inhibitors act on the target enzyme by forming 'cleavable complex,' a high molecular weight DNA protein adduct. The formation of such cleavable complexes results in depletion of the Mr 100,000 'free' topo-1 band detectable by Western blot. The objectives of this study were to determine the maximally tolerated dose of prolonged topotecan infusion in previously untreated and minimally pretreated patients. A secondary objective was to measure the effect of prolonged topotecan infusion on topo-1 levels in peripheral blood mononuclear cells (PBMCs) as a pharmacodynamic end point. In a prior Phase I study of 21-day topotecan infusion (H. Hochster et al., J. Clin. Oncol., 12: 553-559, 1994), the maximum tolerated dose for patients treated previously was 0.53 mg/m2/day for 21 days every 28 days. In this study, patients with no prior therapy were treated similarly at 0.7 mg/m2/day for 21 days, and doses were escalated in 0.1 mg/m2/day increments. Patients who had one prior chemotherapy regimen or radiation therapy to a portal of </=20 cm2 were entered at the 0.6 mg/m2/day level. Cohorts of four patients were entered until the maximum tolerated dose was determined. Peripheral blood was sampled weekly to obtain plasma topotecan drug levels and topo-1 levels in PBMCs by Western Blot. For previously untreated patients, the dose-limiting toxicity was myelosuppression at the dose of 0.8 mg/m2/day. Anemia was seen as a cumulative effect. Unexpected nonhematological toxicity was not observed. topo-1 level analysis by Western blot in 11 cycles with weekly measurements showed progressive decrement in the percentage of free topo-1 (compared to baseline value) during weeks 1, 2, and 3. The median percentage of decrease from baseline was 26% (P, not significant; Wilcoxon signed rank test) at week 1, 45% (P = 0.10) at week 2, and 77% (P = 0.016) at week 3. At week 4, off drug treatment, the median percentage of decrease from baseline was only 14%. Additional analysis of free topo-1 level as a function of both area under the curve (P = 0.005) and day of infusion (P = 0.003) demonstrated a significant relationship by regression analysis using a linear mixed effects model. In this Phase I study of topotecan prolonged infusion, hematological toxicity remained dose limiting without evidence of previously described nonhematological toxicity. The recommended Phase II dose is 0.7 mg/m2/day for 21 days every 28 days for previously untreated patients and 0.6 mg/m2/day for those with limited prior therapy. Western blot analysis of noncomplexed topo-1 in PBMCs sampled weekly showed a progressive depletion of free topo-1 over the 21 days of infusion, which reached statistical significance by week 3. Within 1 week of stopping infusion, topo-1 levels return to baseline. A strong correlation of topo-1 level with area under the curve and duration of infusion was demonstrated. These data suggest that prolonged administration of topo-1 inhibitory drugs results in sustained depletion of free topo-1 enzyme as measured by Western Blot analysis, which may be an important consideration in the clinical use of these agents. Direct randomized, comparative trials will be necessary to determine whether such schedules will improve therapeutic index and efficacy

Single agent activity of vinorelbine in previously untreated breast cancer and its predictable toxicity make it ideal for combination with cisplatin, a drug we found to be very active in combination with paclitaxel (J Clin Oncol; 14:1993, 1996), but with a high rate of neurotoxicity limiting duration of therapy. Eligibility included: histologically proven locally advanced or metastatic measurable breast cancer, ECOG performance status (PS) of 2 or less, adequate organ function. Cisplatin was given at a dose of 75 mg/m2 on day 1, with vinorelbine 30 mg/m2 days 1 and 8 of a 21 day cycle; day 8 vinorelbine dose was modified for neutropenia and thrombocytopenia. 24 patients (pts) entered the study, of whom 23 were eligible and 1 too early for response evaluation. 20 pts were treated as first line therapy for advanced disease (10 locally advanced and 10 metastatic). 3 pts were treated as second-line therapy for metastatic disease. Median age was 49 (range 32-67), median ECOG PS = 0. Nine pts had prior adjuvant chemotherapy and 8 pts had prior RT. A total of 91 cycles of chemotherapy were given with a median of 3 per pt. Hematological toxicity included leukopenia gr 3 = 4 pts, gr 4 = 2; neutropenia gr 3 = 4, gr 4 = 7 pts; no gr 3 or 4 thrombocytopenia. Non-heme toxicity included: N/V gr 2 = 6 pts, gr 3 = 1; neuropathy gr 1 = 10; gr 2 = 2; renal gr 2 = 1 pt. Responses seen included 2 CRs and 6 PRs (of 9 evaluable) locally advanced, 1 CR and 5 PRs (of 10) in metastatic disease, and 1 CR + 1 PR (of 3) second line therapy. Overall response rate was 73% (4 CR + 12 PR + 4 SD = 18% CR and 55% PR) of 22 evaluable pts. These data suggest that combined vinorelbine/cisplatin therapy is highly active in locally advanced and metastatic breast cancer without the high incidence of dose-limiting neurotoxicity seen in our prior paclitaxel/cisplatin trial. Accrual is continuing to improve the 95% confidence interval. (C) American Society of Clinical Oncology 1997

A critical requirement to use tumor antigens as vaccines is that they stimulate CD8+ T cell responses. In this study, we tested the ability of a shed, polyvalent, melanoma antigen vaccine to induce such responses to the melanoma-associated antigens, MAGE-3 and MART-1/Melan-A. Fifteen HLA-A2+ patients with resected malignant melanoma were immunized to the vaccine sc every 2-3 weeks x 4, and monthly thereafter. CD8+ T cells in peripheral blood reacting to HLA-A2 restricted epitopes on MAGE-3 (FLWGPRALV) and/or MART-1/Melan-A (AAGIGILTV) were quantitated directly using a filter spot assay at baseline and following 4 immunizations. Vaccine immunization induced CD8+ T cells reacting specifically to one or both of these antigens in 9 (60%) patients. These cells were CD8+ and HLA-A2 restricted, as reactivity was abrogated by monoclonal antibodies to CD8 and to class I HLA, but not by anti-CD4. The CD8+ T cells were specifically directed to these antigens, as they did not react to the same targets pulsed with a control HLA-A2 restricted peptide recognized by T cells. All responding patients remained recurrence-free during a follow-up of 12-21 months, whereas melanoma recurred within 3-5 months in non-responders. The differences in outcome were unrelated to differences in disease-severity or overall immunological competence between CD8+ T cell responders and non-responders. These results demonstrate that a polyvalent vaccine can stimulate a CD8+ T cell response to MAGE-3 and MART-1/Melan-A in humans, and suggest that the responses are protective and surrogate markers of vaccine efficacy. (C) American Society of Clinical Oncology 1997

PURPOSE: A phase II study of paclitaxel and cisplatin in patients with advanced breast cancer was performed to determine the objective response rate and make further observations about the toxicity of this regimen. PATIENTS AND METHODS: Patients were required to have histologically proven adenocarcinoma of the breast with no more than one chemotherapeutic treatment for advanced disease. Treatment consisted of paclitaxel 200 mg/m2 administered as a 24-hour intravenous (i.v.) infusion followed by cisplatin 75 mg/m2 i.v. Patients received granulocyte colony-stimulating factor (G-CSF) 5 micrograms/kg subcutaneously on day 3 until WBC recovery. Cycles were repeated every 21 days. Patients continued to receive therapy until disease progression or unacceptable toxicity. RESULTS: Forty-four patients entered the trial. Forty-two patients were assessable for response. Nineteen patients (43%) had no prior chemotherapy and 41 had no chemotherapy for metastatic disease. The median number of cycles administered per patient was five (range, one to seven). There were five complete responses (CRs) (11.9%) and 17 partial responses (PRs) (40.5%), with an overall response rate of 52.4% (95% confidence interval [CI], 36.4% to 68.0%). Nine patients had stage III disease. The response rate for this group was 66.7% (95% CI, 33.0% to 92.5%), with three CRs and three PRs. Among 35 patients with stage IV disease, there were two CRs and 14 PRs, with an overall response rate of 48.5% (95% CI, 30.8% to 66.5%). Overall, the median response duration was 10.6 months. Thirty patients (68%) developed transient grade 4 neutropenia. Cumulative neuropathy was the major dose-limiting toxicity. After five cycles of chemotherapy, 96% of patients had at least grade 1 neurotoxicity and 52% had at least grade 2 neurotoxicity. One patient had a toxic death after cycle 1 of therapy. CONCLUSION: The combination of paclitaxel and cisplatin as first-line chemotherapy for women with advanced breast cancer is an active regimen. However, the cumulative neurotoxicity was significant and dose-limiting in the majority of patients