Vemurafenib in patients with BRAFV600 mutation-positive metastatic melanoma: final overall survival results of the randomized BRIM-3 study

Background

The BRIM-3 trial showed improved progression-free survival (PFS) and overall survival (OS) for vemurafenib compared with dacarbazine in treatment-naive patients with BRAFV600 mutation-positive metastatic melanoma. We present final OS data from BRIM-3.

Patients and methods

Patients were randomly assigned in a 1 : 1 ratio to receive vemurafenib (960 mg twice daily) or dacarbazine (1000 mg/m2 every 3 weeks). OS and PFS were co-primary end points. OS was assessed in the intention-to-treat population, with and without censoring of data for dacarbazine patients who crossed over to vemurafenib.

Results

Between 4 January 2010 and 16 December 2010, a total of 675 patients were randomized to vemurafenib (n = 337) or dacarbazine (n = 338, of whom 84 crossed over to vemurafenib). At the time of database lock (14 August 2015), median OS, censored at crossover, was significantly longer for vemurafenib than for dacarbazine {13.6 months [95% confidence interval (CI) 12.0-15.4] versus 9.7 months [95% CI 7.9-12.8; hazard ratio (HR) 0.81 [95% CI 0.67-0.98]; P = 0.03}, as was median OS without censoring at crossover [13.6 months (95% CI 12.0-15.4) versus 10.3 months (95% CI 9.1-12.8); HR 0.81 (95% CI 0.68-0.96); P = 0.01]. Kaplan-Meier estimates of OS rates for vemurafenib versus dacarbazine were 56% versus 46%, 30% versus 24%, 21% versus 19% and 17% versus 16% at 1, 2, 3 and 4 years, respectively. Overall, 173 of the 338 patients (51%) in the dacarbazine arm and 175 of the 337 (52%) of those in the vemurafenib arm received subsequent anticancer therapies, most commonly ipilimumab. Safety data were consistent with the primary analysis.

Conclusions

Vemurafenib continues to be associated with improved median OS in the BRIM-3 trial after extended follow-up. OS curves converged after ≈3 years, likely as a result of crossover from dacarbazine to vemurafenib and receipt of subsequent anticancer therapies.

ClinicalTrials.gov

NCT01006980.

Anti-PD-1 therapy in patients with advanced melanoma and preexisting autoimmune disorders or major toxicity with ipilimumab

Background

Anti-PD-1 antibodies (anti-PD-1) have clinical activity in a number of malignancies. All clinical trials have excluded patients with significant preexisting autoimmune disorders (ADs) and only one has included patients with immune-related adverse events (irAEs) with ipilimumab. We sought to explore the safety and efficacy of anti-PD-1 in such patients.

Patients and methods

Patients with advanced melanoma and preexisting ADs and/or major immune-related adverse events (irAEs) with ipilimumab (requiring systemic immunosuppression) that were treated with anti-PD-1 between 1 July 2012 and 30 September 2015 were retrospectively identified.

Results

One hundred and nineteen patients from 13 academic tertiary referral centers were treated with anti-PD-1. In patients with preexisting AD (N = 52), the response rate was 33%. 20 (38%) patients had a flare of AD requiring immunosuppression, including 7/13 with rheumatoid arthritis, 3/3 with polymyalgia rheumatica, 2/2 with Sjogren's syndrome, 2/2 with immune thrombocytopaenic purpura and 3/8 with psoriasis. No patients with gastrointestinal (N = 6) or neurological disorders (N = 5) flared. Only 2 (4%) patients discontinued treatment due to flare, but 15 (29%) developed other irAEs and 4 (8%) discontinued treatment. In patients with prior ipilimumab irAEs requiring immunosuppression (N = 67) the response rate was 40%. Two (3%) patients had a recurrence of the same ipilimumab irAEs, but 23 (34%) developed new irAEs (14, 21% grade 3-4) and 8 (12%) discontinued treatment. There were no treatment-related deaths.

Conclusions

In melanoma patients with preexisting ADs or major irAEs with ipilimumab, anti-PD-1 induced relatively frequent immune toxicities, but these were often mild, easily managed and did not necessitate discontinuation of therapy, and a significant proportion of patients achieved clinical responses. The results support that anti-PD-1 can be administered safely and can achieve clinical benefit in patients with preexisting ADs or prior major irAEs with ipilimumab.

Characteristics of pyrexia in BRAFV600E/K metastatic melanoma patients treated with combined dabrafenib and trametinib in a phase I/II clinical trial

BACKGROUND

Pyrexia is a frequent adverse event with combined dabrafenib and trametinib therapy (CombiDT), but little is known of its clinical associations, etiology, or appropriate management.

PATIENTS AND METHODS

All patients on the BRF133220 phase I/II trial of CombiDT treated at the standard dose (150/2) were included for assessment of pyrexia (n = 201). BRAF and MEK inhibitor-naïve patients (n = 117) were included for efficacy analyses. Pyrexia was defined as temperature ≥38°C (≥100.4(°)F) or related symptoms.

RESULTS

Fifty-nine percent of patients developed pyrexia during treatment, 24% of which had pyrexia symptoms without a recorded elevation in body temperature. Pyrexia was grade 2+ in 60% of pyrexia patients. Median time to onset of first pyrexia was 19 days, with a median duration of 9 days. Pyrexia patients had a median of two pyrexia events, but 21% had three or more events. Various pyrexia management approaches were conducted in this study. A trend was observed between dabrafenib and hydroxy-dabrafenib exposure and pyrexia. No baseline clinical characteristics predicted pyrexia, and pyrexia was not statistically significantly associated with treatment outcome.

CONCLUSIONS

Pyrexia is a frequent and recurrent toxicity with CombiDT treatment. No baseline features predict pyrexia, and it is not associated with clinical outcome. Dabrafenib and metabolite exposure may contribute to the etiology of pyrexia. The optimal secondary prophylaxis for pyrexia is best studied in a prospective trial.

Phase III randomized, open-label, multicenter trial (BRIM3) comparing BRAF inhibitor vemurafenib with dacarbazine (DTIC) in patients with V600EBRAF-mutated melanoma

LBA4

Background: About 50% of melanomas have an activating V600EBRAF mutation which led to the hypothesis that inhibition of the mutated BRAF kinase may be of clinical benefit. Phase I and II trials with vemurafenib (previously PLX4032/RO5185426), an orally available inhibitor of oncogenic BRAF kinase, showed response rates (RR; CR+PR) >50% in V600EBRAF- mutated melanoma patients (pts). We conducted a phase III trial to determine if vemurafenib improved overall survival (OS) and progression-free survival (PFS) in melanoma pts with V600EBRAF mutation. Methods: Pts with previously untreated, unresectable stage IIIC or stage IV melanoma that tested positive for V600EBRAF mutation by the cobas 4800 BRAF V600 Mutation Test were randomized (1:1) to vemurafenib (960 mg po bid) or DTIC (1,000 mg/m2, IV, q3w). Randomization was stratified by PS, stage, LDH, and geographic region. Pts were assessed for tumor responses after weeks 6, 12, and then q9 weeks. Co-primary endpoints were OS and PFS on the intent-to-treat population; secondary endpoints included RR, response duration, and safety. Final analysis was planned at 196 deaths. Results: 675 pts were enrolled at 103 centers worldwide between Jan and Dec 2010. Treatment cohorts were well-balanced. At the pre-planned interim analysis (50% of deaths needed for final analysis), the hazard ratios for OS and PFS were 0.37 (95% CI 0.26 to 0.55; p<0.0001) and 0.26 (95% CI 0.20 to 0.33; p<0.0001), respectively, both in favor of vemurafenib. The confirmed RR was 48.4% and 5.5% to vemurafenib and DTIC, respectively, among the 65% of pts evaluable for RR to date. Benefit in OS, PFS, and RR was seen in all subgroups examined. Due to these data, the DTIC cohort has been allowed to cross over to vemurafenib. At the time of data analysis, 66% of vemurafenib pts and 25% DTIC pts were still on treatment. The most common toxicities of vemurafenib were: diarrhea, rash, alopecia, photosensitivity, fatigue, arthralgia, and keratoacanthoma/skin squamous cell carcinoma. Conclusions: Vemurafenib is associated with significantly improved OS and PFS compared to DTIC in pts with previously untreated, V600EBRAF-mutated metastatic melanoma.

A phase I study to evaluate safety and antitumor activity of biweekly BMS-936558 (Anti-PD-1, MDX-1106/ONO-4538) in patients with RCC and other advanced refractory malignancies

331

Background: Programmed death-1 (PD-1), a T-cell inhibitory receptor, may suppress antitumor immunity. BMS-936558, a fully human PD-1 blocking antibody, has shown antitumor activity and manageable toxicity after biweekly dosing (Sznol, ASCO 2010, #2506). This report provides an update on safety and antitumor activity with special emphasis on RCC.

Methods: An open-label phase I dose escalation study of BMS-936558 was conducted in patients (pts) with treatment refractory metastatic clear-cell renal cell carcinoma (RCC), castrate-resistant prostate cancer (CRPC), melanoma (MEL), non-small cell lung cancer, or colorectal cancer (CRC). Dose escalation continued to 10 mg/kg when an expansion cohort for pts (16) with each tumor type was opened for additional safety and efficacy information. Tumor response (RECIST) was evaluated every 8 weeks. Clinically stable pts with early PD could continue until further PD or clinical deterioration.

Results: 126 pts (18 RCC) were treated with 1, 3, or 10 mg/kg. MTD was not reached. Across all doses, the most common AEs (Any/grade 3-4) were fatigue (45.2%/6.3%) and diarrhea (30.2%/0.8%) while the most common drug-related AEs (Any/grade 3-4) were fatigue (20.6%/0.8%), rash (11.9%/0%), pruritus (11.3%/0%), and diarrhea (10.3%/0.8%). There was no apparent relationship between dose and frequency of AEs. One pt died with sepsis while being treated for drug-related grade 4 pneumonitis. The median number of prior treatment regimens in the RCC cohort was 2 (range 1-6). Of the 18 RCC pts, 16 were treated with 10 mg/kg. The median duration of treatment was 7.6+mo. ORR was 5/16 (31.2%) and SD>4mo was 6/16 (37.5%). The median duration of response was 4.0+ mo (3.7-7.4+ mo). Of the 2 RCC pts treated with 1 mg/kg, 1 obtained a CR (12+ mo) and 1 had SD (21+ mo). For evaluable CRPC pts, 1/15 pts (6.7%) obtained a PR (2+ mo) and 3/15 (20%) had SD>4mo.

Conclusions: BMS-936558 administered biweekly is tolerable and has encouraging antitumor activity in a previously treated patients with RCC. Data on baseline characteristics, long-term toxicity and response duration will be updated at the meeting.

[Table: see text]

A phase III, randomized, double-blind, multicenter study comparing monotherapy with ipilimumab or gp100 peptide vaccine and the combination in patients with previously treated, unresectable stage III or IV melanoma

4

Background: Ipilimumab, a fully human monoclonal antibody against cytotoxic T-lymphocyte antigen-4, demonstrated activity in advanced melanoma. Gp100 vaccine showed immunological and clinical responses, and enhanced clinical activity when combined with other immunotherapy. This phase III study compared efficacy and safety of ipilimumab or gp100 monotherapy and combination.

Methods: Eligible patients (HLA-A*0201+ previously treated adults with unresectable stage III/IV melanoma) were randomized 1:3:1 to ipilimumab (3 mg/kg q3w x 4 doses) + placebo (n=137), ipilimumab + gp100 (peptides 209-217[210M] and 280-288 [288V]; 1mg q3w x 4 doses; n=403), or gp100 + placebo (n=136). There was no maintenance phase. Primary endpoint was comparison of overall survival (OS) between patients who received combination versus gp100 alone; secondary endpoints were all other OS comparisons, best overall response rate (BORR), disease control rate (DCR) to W24, progression-free survival (PFS), and safety.

Results: The study demonstrated statistically significant results for all efficacy endpoints (below). Ipilimumab alone or combined with gp100 resulted in a significant improvement in OS with risk reduction of 32-34% compared to gp100. Significant differences in DCR, BORR, and PFS were observed. Adverse events with ipilimumab were consistent with prior studies: generally mild, immune-related, and medically manageable.

Conclusions: Ipilimumab is the first agent to improve median and long-term OS in a phase III study of previously treated patients with advanced melanoma. Addition of gp100 vaccine to ipilimumab did not improve outcome.

[Table: see text]

[Table: see text]

PLX4032, a highly selective V600EBRAF kinase inhibitor: Clinical correlation of activity with pharmacokinetic and pharmacodynamic parameters in a phase I trial

9021

Background: PLX4032 is an oral, highly selective inhibitor of oncogenic V600EBRAF kinase currently in phase I trial. V600EBRAF mutation activates Raf/MEK/ERK pathway in multiple tumor types. We evaluated the relationship between PK, PD (pERK, Ki67, FDG-PET), tumor histology and clinical activity following PLX4032 administration in a phase I trial. Methods: In the phase I trial, 6 melanoma pts with V600EBRAF were treated with PLX4032 daily at several dose levels and tumor biopsies (baseline vs. day 15) were assessed histologically and by semi-quantitative IHC analysis (modified H-score) for pERK and Ki67. The first 4 pts received a crystalline formulation of PLX4032; the last 2 pts received a formulation with increased bioavailability. Plasma PK parameters were collected at frequent time points on Days 1, 8 and 15. FDG-PET was performed on Days 1 and 15 on last 2 pts. Results: In the first 4 pts, no histological changes were observed with treatment and all developed disease progression. All had decreased percentage of Ki67 positive nuclei (pre-Rx, range 20–60%, median 45%; post-Rx, range 5–25%, median 12.5%) and 3 of the 4 had decreased pERK levels (pre-Rx, range 50–100, median 60; post-Rx, range 10–40, median 11). Mean PLX4032 AUC0–24h ∼ 126 μM*h was in the range for preclinical tumor stasis but below the threshold for shrinkage. In the last 2 pts, striking tumor necrosis and tumor melanosis was observed in the post-Rx samples. One pt remains on study with a confirmed PR, the other showed a clinical response before disease progression occurred in cycle 2. The percentage of Ki67 positive nuclei declined substantially (pre-Rx, 30% and 50% to post-Rx, 5% and 3%), as did the levels of pERK in the pt with PR (pre-Rx: 70 to post-Rx: 2). Mean PLX4032 AUC0–24h was well above the preclinical threshold in the range of 500 - 1000 μM*h. Both pts had decreased FDG uptake on D15. Conclusions: Clinical activity of PLX4032 treatment correlates with drug exposure levels as measured by AUC0–24h and was associated with histological changes in V600EBRAF positive melanomas on Day 15. Reduction of pERK, along with evidence of reduced proliferation and FDG uptake was observed. Further analysis of PD markers with additional pts at the MTD is planned.

[Table: see text]

A phase II trial of perifosine in patients with advanced renal cell carcinoma (RCC) who have failed tyrosine kinase inhibitors (TKI)

5101

Background: The recently demonstrated activity of inhibitors of TORC1 in RCC has raised the possibility that even greater effects may be achieved by targeting upstream of this pathway. Perifosine is a synthetic alkylphospholipid which inhibits Akt activity and also has cell-dependent effects upon the MAP-kinase pathway. Prior single-agent trials showed disease stabilization/regression in patients (pts) with advanced RCC; however, few pts were previously treated with a TKI. Therefore, we conducted a multi-center phase II trial to determine the safety and efficacy of perifosine in pts with advanced RCC refractory to VEGFR TKI. Methods: Primary objectives were to measure the % of pts progression-free at 12 weeks (wks) and overall progression-free survival (PFS) of perifosine (100 mg qhs). Secondary objectives included overall response rate (> PR), and safety, Eligibility: ECOG PS 0–1, pts with metastatic RCC who have RECIST defined progression on either sunitinib or sorafenib. Prior Rx with immunotherapy and bevacizumab was permitted. Normal organ and marrow function required. Results: From 4/07–10/08, 24 pts were treated at four sites. Median age 67 (range 47–78) and 16 were male; 90% of pts had predominantly clear cell histology. Prior sunitinib = 12; prior sorafenib = 12 (1.5 avg prior Rx). As of 12/08, all 24 pts were evaluable for PFS, response and toxicity as follows in the table . 6/24 pts remain on treatment (range 7 - 84 wks). Therapy was well tolerated with primarily Grade (G) 1 & 2 adverse events. G 3 & 4 events were: dyspnea (8%), hyponatremia (8%), pulmonary embolism (4%) and arthalgia (4%). Conclusions: Perifosine has promising activity in pts with RCC who have failed prior TKI therapy. The favorable toxicity profile suggests potential for combinational therapies with VEGF-targeted agents. Additional studies are under consideration to evaluate perifosine for clinical benefit in pts with previously treated RCC.

[Table: see text]

[Table: see text]

Randomized phase II study of dacarbazine with or without sorafenib in patients with advanced melanoma

8511

Background: Sorafenib (SOR), a potent and selective multi-kinase inhibitor, exerts its anti-tumor and anti-angiogenic effects via inhibition of VEGFR-1, -2, -3, PDGFR-a, -β, and Raf. Dacarbazine (DTIC) is an FDA-approved cytotoxic agent for advanced melanoma. Phase I/II results of SOR + DTIC were encouraging and prompted this randomized phase II study. Methods: This was a multi- center, double-blinded, placebo-controlled study; eligibility criteria included measurable disease by RECIST, no prior cytotoxic chemotherapy, and no active brain metastases. Advanced melanoma patients (pts) stratified by stage (unresectable III vs IVM1a/M1b vs M1c) and ECOG PS (0 vs 1) were randomized to receive DTIC 1,000 mg/m2 q 21 days + oral placebo (PL) or oral SOR 400 mg bid continuously until the occurrence of progressive disease or intolerable toxicity. The primary endpoint was progression-free survival (PFS) of DTIC+SOR vs DTIC+PL. Using a two-sided test with a = 0.05, 77 PFS events were needed to detect a hazard ratio (HR) of 0.5 (SOR/PL) with 86 % power. The secondary endpoint was overall survival and tertiary endpoints were objective response rate (ORR), time to progression, and duration of response. Results: 101 pts were enrolled over 12 months (51 DTIC+SOR, 50 DTIC+PL). Treatment arms were balanced for age (median 58 yrs), gender (male 70%), PS (ECOG 1 39%), stage (Stage IV M1c 52%) and baseline LDH (>ULN 29%). At the time of analysis by independent assessment, the median PFS of DTIC+PL vs DTIC+SOR was 11.7 wks (95% CI 6.1, 17.9) vs 21.1 wks (95% CI: 16, 28); HR 0.67 [p=0.07]; PFS rate at Day 180 was 18% vs 41%; and ORR was 12% vs 24%. Survival data are immature. Toxicities of Grade 3 or higher (DTIC+PL vs DTIC+SOR) included neutropenia (12% vs 33%), leukopenia (6% vs 14%), thrombocytopenia (18% vs 35%), thrombosis/embolism (0% vs 6%), hypertension (0 vs 8%), hand-foot skin reaction (0 vs 4%), and CNS hemorrhage (0% vs 8%). 3 of the 4 pts with CNS hemorrhage had new brain metastases. No treatment-related deaths occurred in either arm. Conclusions: DTIC+SOR was well tolerated and showed a strong efficacy trend compared with DTIC+PL in median PFS, PFS rate at 6 months and ORR in chemotherapy-naïve pts with advanced melanoma. This regimen warrants further evaluation in larger clinical trial settings.

No significant financial relationships to disclose.

First-line treatment with bevacizumab (B) and high dose (HD) bolus aldesleukin (IL-2) in metastatic renal cell carcinoma (mRCC) patients (Pts)

15524

Background: The rationale for combining HD IL-2 with B includes potential synergistic immune interactions, non-overlapping toxicities, and potential for added clinical benefit. We have initiated a multi-center phase II study designed to estimate the efficacy of combination therapy of standard HD IL-2 and B therapy in mRCC pts. Methods: Pts with histologically confirmed mRCC, predominantly clear cell histology, measurable or evaluable disease, KPS of =80%, adequate end organ function for HD IL-2, and no underlying coagulopathy or thrombotic event are eligible for this study. One cycle consists of 84 days. B (10 mg/kg) IV is given every 2 wks beginning 2 wks prior to the first dose of IL-2. B is dosed 1 hr prior to initiating IL-2 on days IL-2 is given. HD IL-2 (600,000 IU/kg) IV Q8 hours (maximum 28 doses) is given during two 5-day courses separated by 9 days (starting on day 15 and 29). Results: We report the results of the first 15 of a planned 60 pts. The median age is 54 (range 40–73) with 9 men and 6 women. 14 pts have a MSKCC intermediate prognostic score, one pt has a poor prognostic score. In the first cycle, the median number of B doses was 7 of a planned 7 (range 2–7) and the median number of IL-2 doses was 17 of a planned 28 (range 6–26). There has been one treatment related death from unresponsive hypotension which occurred during the second cycle. Typical IL-2 toxicities have been noted thus far. Among a variety of correlative studies, we evaluated the serum L- ornithine (L-O) level, a byproduct of arginase-mediated arginine metabolism that has been shown to inversely correlate with TCR? chain expression. L-O level are significantly elevated in RCC pts possibly due to VEGF stimulation of arginase production. In 4 patients tested to date, peripheral blood L-O levels have dramatically decreased over the course of therapy. Conclusions: HD IL-2 and B can be given safely and may impact on immune regulatory pathways.

[Table: see text]

Phase II trial of sunitinib in bevacizumab-refractory metastatic renal cell carcinoma (mRCC): Updated results and analysis of circulating biomarkers

5035

Background: Sunitinib malate is an oral, multitargeted tyrosine kinase inhibitor with antiangiogenic and antitumor activity. This study evaluated the safety and activity of sunitinib in mRCC patients (pts) previously treated with the VEGF-neutralizing antibody, bevacizumab. Levels of angiogenic biomarkers, including plasma VEGF and soluble VEGFR-3 (sVEGFR-3), were assessed for predictive significance with clinical response. Methods: Pts were required to have mRCC with disease progression following bevacizumab- based therapy, measurable disease, ECOG performance status 0 or 1, and adequate organ function. Pts were treated with sunitinib 50 mg daily in 6-week cycles (4 weeks on, followed by 2 weeks off). The primary endpoint was objective response according to RECIST. Plasma VEGF and sVEGFR-3 levels were measured in pre-treatment samples and at multiple timepoints on study. Results: A total of 61 pts were enrolled. The objective partial response rate was 23% (95% CI: 13%, 36%); 35 pts (57%) demonstrated stable disease. The median duration of response was 36 weeks (95% CI: 26, NA) and progression-free survival was 30 weeks (95% CI: 18, 34). Plasma VEGF levels increased from baseline (3-fold mean elevation), while plasma sVEGFR-3 levels decreased from baseline (40% mean reduction). Pre-treatment VEGF levels were significantly higher in pts (n=34) with <10 weeks between cessation of bevacizumab and start of sunitinib (p<0.001); ELISA specificity suggests that detected VEGF is not bevacizumab-bound. Pre-treatment sVEGFR-3 levels were significantly lower at baseline in responding pts vs. non-responding pts (p<0.0318). A greater reduction in sVEGFR-3 levels was seen in responding pts vs. non-responding pts (p<0.10). Pretreatment VEGF and VEGF fold-changes did not differ according to clinical response. Conclusions: Sunitinib has significant antitumor activity in bevacizumab-refractory mRCC pts, suggesting absence of cross-resistance between bevacizumab and sunitinib. Biomarkers including plasma VEGF and sVEGFR-3 may have predictive potential in sunitinib-treated patients.

No significant financial relationships to disclose.

VEGF-Trap (V-T) overcomes defects in dendritic cell (DC) differentiation without improvement in antigen-specific immune responses

21087

Background: Induction of antitumor immune responses requires adequate DC function. Well-documented DC defects in cancer mediated through tumor-derived vascular endothelial growth factor (VEGF) may lead to tumor escape and limit efficacy of cancer vaccines. V-T with extracellular domains of VEGFR 1/ 2 coupled to Fc IgG1 binds all VEGF-A isoforms. Methods: To determine if inhibition of VEGF signaling improves immune responses, we evaluated 15 cancer pts treated with V-T (2.0–7.0 mg/kg) IV over 1 hr Q 2 wks on a phase I trial. To evaluate immune parameters, PBMC were collected prior to, days 15, 29, and 57 of V-T treatment. Ligand blockade was achieved in all patients. Mature DCs (myeloid and plasmacytoid subsets), and regulatory T cells (T regs) were characterized by expression pattern of CD11c, CD86, CD40, CCR7, CD83, CD123, CD4, CD25, and GITR. T cell function was assessed by allogeneic mixed leukocyte reaction (MLR) and proliferation to tetanus toxoid, influenza, or PHA. Results: V-T treatment did not affect the presence of the total population of DCs, myeloid or plasmacytoid subsets, or myeloid-derived suppressor cells (MDSC). It significantly increased the proportion of mature DCs expressing CD86, CD40, CD83, and CCR7R (p<0.05) without overall improvement in T cell immunity. At 8 wks, 11 pts had either stable disease (SD) or partial response (PR) to the therapy and 4 pts had progressive disease (PD), both with no significant differences in the proportion of MDSC or total DCs. The proportion of mature DCs was significantly increased in pts with SD and PR (p<0.025), but remained unchanged in PD pts. However, functional tests were similar in both groups. Another subset analysis revealed significant improvement in immune responses in 6 pts who had stable or decrease in the MDSC proportion, but not in 9 pts with increases in MDSC. During V-T treatment, Tregs further increased above baseline in pts (p=0.048). Conclusions: These data demonstrate that inhibition of VEGF signaling improves differentiation of DCs in cancer patients, while alone it was not sufficient to improve immune responses. These data illustrates the multifaceted nature of immune deficiency in cancer pts and points to a need for complex approaches to modulation of immune reactivity.

No significant financial relationships to disclose.

A phase I/II trial of sorafenib (S) with bevacizumab (B) in metastatic renal cell cancer (mRCC) patients (Pts)

3031

Background: In an effort to enhance the efficacy of vascular endothelial growth factor (VEGF) pathway blockade in mRCC we initiated a Phase I-II trial of combination S and B to block VEGFR signaling and VEGF binding, as well as, platelet derived growth factor receptor β (PDGFR) signaling. Methods: Pts with measurable (RECIST) mRCC, adequate organ function, and PS 0–1 were eligible for this trial. Cohorts of 6 pts were enrolled at 3 sites to define the MTD and DLT of the combination of S and B. The schedule was B IV q 14 days and S daily with cycles of 28 days. Response and toxicity were assessed at the end of 2 cycles. Dose levels began with S at 200mg BID and B at 5 mg/kg (level 1) with the hope to reach phase II doses of both agents (S at 400mg BID and B at 10mg/kg). The MTD or the phase II dose of each agent would then be administered to up to 45 pts with mRCC, clear cell histology, and prior nephrectomy. Results: A total of 18 patients have been enrolled to date, with 15 completing their first response evaluation. Pts were median age 61 years (46–74 range); M/F: 15/3; PS: 0/1= 12/6; 17 clear cell, 1 chromophobe, 11 prior nephrectomy; 4 with prior cytokine therapy. Two pts in level 1 experienced DLT with recurrent and intolerable (grade 3) hand-foot syndrome (HFS). An additional 6 patients were treated at dose level -1 (S at 200mmg QD and B at 3mg/kg) with no DLTs. Six more pts have been treated at dose level 1 with Vit B6 at 300mg/d in an attempt to minimize HFS. Toxicity data in this cohort is incomplete. Additional toxicities among the 18 pts included grade 3 hypertension (4), grade 3 proteinuria (2), and grade 2 stomatitis (3). Responses including 4 objective PRs, and 4 pts with 20–30% regression have been seen in the 14 evaluated pts. Only 2 patients have had disease progression. Conclusions: The phase II doses for this combination have not yet been established, but will likely be lower than the full phase II doses of the individual agents. The toxicities from HFS (primarily) and hypertension and stomatitis appear to be limiting. Even at the initial low doses of S and B, significant anti-tumor activity has been observed. The completion of this phase I trial will be reported. The phase II studies in mRCC are highly anticipated. Support by phase I contract UO-1 CA099177

[Table: see text]

Surgical resection for stage IV melanoma: A Southwest Oncology Group trial (S9430)

8019

Background: Surgery is the treatment of choice for isolated distant metastases of melanoma. Based on a number of retrospective reports, patients (pts) who have undergone complete resection of metastatic disease can have an excellent survival. Methods: In SWOG, we prospectively evaluated surgical resection for stage IV melanoma. Pts enrolled prior to surgery; treatment following resection was at the discretion of the treating physician. Resected specimens were collected, whenever possible, for future molecular analyses. Results: Over a 9 year period from 1996 to 2005, 77 pts were enrolled from 18 centers, with 7 centers accruing ≥4 pts. Of 77 pts, 10 were incompletely resected and 5 had no evidence of stage IV disease. Therefore, 62 pts (81%) felt to have resectable stage IV actually did, and were included in the analysis. Pts characteristics were: median age 54 yrs (range 23–81); M:F 69%:31%; PS 0–1 100%; prior adjuvant IFN 45%. Resected sites included skin/soft tissue 40%; distant LN 21%; lung 13%; liver 8%; CNS 5%; bone 2%; other visceral sites 27%. Post-surgical complications included 1 grade IV pulmonary embolus and 1 grade III liver toxicity. Following surgery but prior to any further recurrence, 18 pts received adjuvant treatment including 8 with IFN and 7 with radiotherapy. After surgery, median progression-free survival (PFS) was 6 mos (95% CI 3–7 mos) with 9 patients (15%) remaining progression-free. Median overall survival (OS) was 21 mos (95% CI 15–28 mos); 3- and 4-yr OS was 33% and 29%, respectively. Conclusions: These results provide an estimate of prognosis for resected stage IV disease from a diverse patient base. They illustrate the potential for prolonged OS even with a short PFS. In some cases, pts can be repeatedly resected for long-term control of their disease. Pts with resectable stage IV melanoma are appropriate candidates for inclusion in adjuvant therapy trials.

No significant financial relationships to disclose.

Bevacizumab, erlotinib, and imatinib in the treatment of patients with advanced renal cell carcinoma: Update of a multicenter phase II trial

4594

Background: Most patients with clear cell RCC have loss of von-Hippel-Lindau (VHL) gene function, ultimately leading to overexpression of VEGF, transforming growth factor (TGF-α), epidermal growth factor (EGF), and platelet-derived growth factor (PDGF). In this phase II trial, we evaluate the combination blockade of VEGF, and the EGF and PDGF receptors as treatment for metastatic RCC. Methods: Eligibility: metastatic clear cell RCC; 0–2 previous systemic regimens; ECOG PS 0–1; no previous anti-angiogenesis or EGF receptor inhibitor therapy; no active CNS metastases; adequate organ function; no history of thromboembolic disease; informed consent. All patients received bevacizumab 10 mg/kg IV q 2 wks, erlotinib 150 mg PO daily, and imatinib 400 mg PO daily. Patients were evaluated for response after 8 weeks; treatment continued until tumor progression. Results: 92 patients entered this trial between 6/04 and 4/05. Pertinent clinical characteristics: no previous systemic treatment, 71%; Motzer prognostic category low/intermediate/high 47%/50%/3%. 84 patients (91%) received at least 8 weeks of therapy. Objective responses were seen in 11 patients (12%); an additional 52 patients (57%) had stable disease/minor response at first reevaluation. After a median follow-up of 11 months, projected median and 1-year progression-free survivals are 9 months and 40%, respectively. One-year overall survival is 62%. The addition of imatinib to bevacizumab/erlotinib added substantial toxicity. Grade 3/4 toxicity included: diarrhea, 48%; skin rash, 27%; fatigue, 23%; nausea/vomiting, 20%. 2 patients discontinued treatment as a result of toxicity; all patients who received treatment for > 8 weeks required dose reductions of imatinib and/or erlotinib. Conclusions: The addition of imatinib, a PDGF receptor inhibitor, to bevacizumab/erlotinib does not appear to substantially improve the efficacy of the 2-drug combination. This 3-drug combination produced moderate to severe toxicity in the majority of patients, primarily exacerbations of diarrhea, fatigue, and skin rash. Further development of this 3-drug regimen is not recommended.

[Table: see text]

Phase II study of erlotinib given daily for patients with metastatic melanoma (MM)

18002

Background: Erlotinib is a small molecule tyrosine kinase inhibitor (TKI) that targets epidermal growth factor receptor (EGFr). The EGFr is a potential therapeutic target because it is expressed by a number of malignancies, including melanocytic lesions, and in some plays an important role in the biology of the cancer. Our aim was to conduct a phase II study evaluating erlotinib in patients (pts) with measurable metastatic melanoma. Methods: Eligibility criteria included measurable disease, ECOG PS = 0–1, and adequate organ function. Pts were eligible if they received up to one prior therapy for metastatic disease. Pts received a daily dose of erlotinib 150 mg. The primary outcomes were overall response rate and response duration. The study had a two-stage design with closure at 14 pts if there were no objective RECIST responses. Secondary outcomes included overall safety and tolerability of erlotinib. Results: Between August 2003 and August 2004, a total of 14 pts with MM were accrued. The majority of pts were male 12:2 = M, median age = 57.5 yrs (range 38 to 80 yrs). Stage of disease included M1a (n = 7), M1b (n = 5), M1c (n = 2) and performance status was equally divided between 0 and 1 (7 pts each). Seven pts (50%) had prior adjuvant therapy and six pts (43%) had at least one prior therapy for metastatic disease. Four pts (29%) had no prior therapy. No objective responses were observed. Four pts (29%) had stable disease at their initial 8-week evaluation of which only 2 had SD>6 months (228 and 365 days). One pt with SD withdrew from the study on day 68 for grade II toxicities. There were no Grade III/IV hematological or biochemical toxicities. Grade III toxicities were diarrhea and anorexia, each in a single patient. Most pts (n = 12) experienced at least a grade I dermatological toxicity manifested as an acneiform rash and/or pruritis. The median progression free survival (PFS) of all 14 pts was approximately 60 days with a range of 35 to 365 days. The median PFS of those with SD was 192.5 days. Conclusions: Erlotinib given daily at 150 mg is well tolerated in pts with MM. However, it has minimal to no single agent activity. Any further investigation of this drug should be pursued in combination with other agents only if a strong scientific/clinical rationale exists.

No significant financial relationships to disclose.

Efficacy and safety of sunitinib malate (SU11248) in bevacizumab-refractory metastatic renal cell carcinoma (mRCC)

4522

Background: Sunitinib malate (SU11248) is an oral, multitargeted tyrosine kinase inhibitor of the vascular endothelial growth factor receptor (VEGFR) family, platelet-derived growth factor receptor (PDGFR) and other related receptors. It has demonstrated anti-tumor activity in cytokine-refractory mRCC patients (pts). The activity of sunitinib in pts refractory to VEGF binding agents such as bevacizumab, however, has not been evaluated. It was hypothesized that tumor resistance to bevacizumab may be driven, in part, through pathways sensitive to inhibition by sunitinib. A phase II study evaluating the activity of sunitinib in bevacizumab-refractory mRCC was thus conducted. Methods: Pts with mRCC who demonstrated RECIST-defined disease progression within 3 months after bevacizumab-based therapy were treated with sunitinib (50 mg daily, 4 weeks of a 6-week cycle). Additional eligibility included measurable disease, clear cell histology, ≤ 2 prior systemic regimens, prior nephrectomy, performance status 0 or 1 and adequate organ function. The primary endpoint was objective response by RECIST criteria. A single-stage design was employed to test the null hypothesis that the true response rate is ≤ 5% versus the alternative hypothesis that the true response rate is ≥ 15%. Results: Accrual of 60 patients has been completed. Baseline characteristics include a median age of 59 years; 92% of pts had ≥ 2 metastatic sites and 23% had prior radiotherapy. Thirty-two of 60 pts enrolled are evaluable for response; 28 pts are too early for assessment. Twenty-six pts (81%) demonstrated some degree of tumor shrinkage, including, 4 pts (13%; 95% CI 4%, 29%) demonstrating an objective partial response. The most common treatment-related adverse events (AEs) included fatigue, diarrhea, dysgeusia, and nausea. Serious treatment-related AEs included fatigue, diarrhea, nausea and one fatal cerebral hemorrhage; 3 pts withdrew due to an AE. Conclusions: Sunitinib has substantial antitumor activity in bevacizumab-refractory mRCC pts, suggesting that sunitinib may inhibit signaling pathways involved in bevacizumab resistance. The precise mechanisms of response to sunitinib in bevacizumab-refractory tumors will require additional studies.

[Table: see text]

A multi-institutional phase II trial of a novel inpatient schedule of continuous interleukin-2 with interferon alpha-2b in advanced renal cell carcinoma: major durable responses in a less highly selected patient population

BACKGROUND

A prospective multi-institutional phase II trial was undertaken to define the activity and toxicity of a unique decrescendo infusion of interleukin-2 (IL-2) in combination with interferon (IFN) in patients with metastatic renal cell carcinoma. The identical regimen has shown promise in advanced melanoma.

PATIENTS AND METHODS

Between February 1997 and March 1999, 47 patients with metastatic renal cell carcinoma, from five institutions, were treated with outpatient s.c. IFN (10 mU/m2/day) on days 1-5, followed by inpatient IL-2 via continuous i.v. decrescendo infusion [18 million International Units (MIU) (I mg)/m2/6 h, followed by 18 MIU/m2/12 h, then 18 MIU/m2/24 h and 4.5 MIU/m2/24 h for the following 3 days] on days 8-12, in a hospital ward without intensive care unit (ICU)-type monitoring. Treatment was repeated every 4 weeks. In contrast to high dose IL-2 protocols, patient eligibility did not require pulmonary function tests and allowed serum creatinine up to 2 mg/dl.

RESULTS

Among 44 eligible patients, 57% (25) had their primary in place, 57% (25) had bone or visceral involvement, and only 4% (2) had lung as their only site of disease. The overall response rate in 43 response-evaluable patients was 16.3% [95% confidence interval (CI) 5.3 to 27.3], with three complete responses and four partial responses observed. The median survival was 13 months; nine patients remain alive at >23 months. The median duration of response is 36 months (range 11.5 to 48+ months). Toxicity was modest, consisting of typical cytokine-induced systemic symptoms and rare organ dysfunction. Severe grade 4 toxicity occurred in only 13% of the 130 cycles.

CONCLUSIONS

This unique, reasonably well tolerated IL-2/IFN combination induced a modest response rate with a number of durable remissions. While the optimal IL-2-based regimen for the treatment of advanced renal cell carcinoma remains elusive, the present regimen should attract considerable interest. This is based on tumor activity very similar to high dose IL-2 in a patient population not as carefully selected for optimal organ function.

High-dose interferon alfa-2b significantly prolongs relapse-free and overall survival compared with the GM2-KLH/QS-21 vaccine in patients with resected stage IIB-III melanoma: results of intergroup trial E1694/S9512/C509801

PURPOSE

Vaccine alternatives to high-dose interferon alfa-2b therapy (HDI), the current standard adjuvant therapy for high-risk melanoma, are of interest because of toxicity associated with HDI. The GM2 ganglioside is a well-defined melanoma antigen, and anti-GM2 antibodies have been associated with improved prognosis. We conducted a prospective, randomized, intergroup trial to evaluate the efficacy of HDI for 1 year versus vaccination with GM2 conjugated to keyhole limpet hemocyanin and administered with QS-21 (GMK) for 96 weeks (weekly x 4 then every 12 weeks x 8).

PATIENTS AND METHODS

Eligible patients had resected stage IIB/III melanoma. Patients were stratified by sex and number of positive nodes. Primary end points were relapse-free survival (RFS) and overall survival (OS).

RESULTS

Eight hundred eighty patients were randomized (440 per treatment group); 774 patients were eligible for efficacy analysis. The trial was closed after interim analysis indicated inferiority of GMK compared with HDI. For eligible patients, HDI provided a statistically significant RFS benefit (hazard ratio [HR] = 1.47, P = .0015) and OS benefit (HR = 1.52, P = .009) for GMK versus HDI. Similar benefit was observed in the intent-to-treat analysis (RFS HR = 1.49; OS HR = 1.38). HDI was associated with a treatment benefit in all subsets of patients with zero to > or = four positive nodes, but the greatest benefit was observed in the node-negative subset (RFS HR = 2.07; OS HR = 2.71 [eligible population]). Antibody responses to GM2 (ie, titers > or = 1:80) at days 29, 85, 365, and 720 were associated with a trend toward improved RFS and OS (P2 = .068 at day 29).

CONCLUSION

This trial demonstrated a significant treatment benefit of HDI versus GMK in terms of RFS and OS in melanoma patients at high risk of recurrence.

Pilot trial of interleukin-2 with granulocyte colony-stimulating factor for the mobilization of progenitor cells in advanced breast cancer patients undergoing high-dose chemotherapy: expansion of immune effectors within the stem-cell graft and post-stem-cell infusion

PURPOSE

To evaluate whether administration of interleukin-2 (IL-2) with granulocyte colony-stimulating factor (G-CSF) improves mobilization of immune effector cells into the stem-cell graft of patients undergoing high-dose chemotherapy and autografting.

PATIENTS AND METHODS

We performed a trial of stem-cell mobilization with IL-2 and G-CSF in advanced breast cancer patients receiving high-dose chemotherapy with cyclophosphamide, thiotepa, and carboplatin and stem cells followed by IL-2. The trial defined immune, hematologic, and clinical effects of IL-2 in this setting.

RESULTS

Of 32 patients enrolled, nine received G-CSF alone for mobilization. Twenty-one of 23 patients mobilized with IL-2 plus G-CSF had stem cells collected with more mononuclear cells than those receiving G-CSF (19.3 v 10.4 x 10(8)/kg; P =.006), but fewer CD34(+) progenitor cells (6.9 v 22.0 x 10(6)/kg; P =.049). The IL-2 plus G-CSF-mobilized patients had greater numbers of activated T (CD3(+)/CD25(+)) cells (P =.009), natural killer (NK; CD56(+)) cells (P =.007), and activated NK (CD56 bright(+)) cells (P: =.039) than those patients mobilized with G-CSF. NK (P =.042) and lymphokine-activated killer (LAK) (P =.016) activity was increased in those mobilized with IL-2 + G-CSF, whereas G-CSF-mobilized patients had a decline in cytolytic activity. In the third week posttransplantation, immune reconstitution was superior in those mobilized with IL-2 plus G-CSF based on greater numbers of activated T cells (P =.003), activated NK cells (P =.04), and greater LAK activity (P =.003). The 16 of 21 IL-2 + G-CSF-mobilized patients with adequate numbers of stem cells (> 1.5 x 10(6) CD34(+) cells/kg) collected engrafted rapidly posttransplantation.

CONCLUSION

The results demonstrate that G-CSF + IL-2 can enhance the number and function of antitumor effector cells in a mobilized autograft without impairing the hematologic engraftment, provided that CD34 cell counts are more than 1.5 x 10(6) cells/kg. Mobilization of CD34(+) stem cells does seem to be adversely affected. In those mobilized with IL-2 and G-CSF, post-stem-cell immune reconstitution of antitumor immune effector cells was enhanced.

Interleukin 10-induced thrombocytopenia in normal healthy adult volunteers: evidence for decreased platelet production

Recombinant human interleukin 10 (rhuIL-10) inhibits the production of proinflammatory cytokines and has shown promise in the treatment of inflammatory bowel disease. Clinical trials have been accompanied by a reversible decline in platelet counts. We conducted a randomized, double-blinded, placebo-controlled, parallel group trial in 12 healthy volunteers to investigate the aetiology of rhuIL-10-induced thrombocytopenia. Eight volunteers received 8 microg/kg/d of rhuIL-10 subcutaneously, while four subjects received a placebo alone for 10 d. A reversible decline in the platelet counts from a mean of 275 x 10(9)/l to 164 x 10(9)/l was observed in the IL-10-treated cohort (P = 0.012). A fall in the haemoglobin mean levels was also observed in the IL-10-treated cohort from 13.7 to 11.7 g/dl (P = 0.011). No significant change was observed in the bone marrow cellularity or myeloid/erythroid ratio or in the number of megakaryocytes per high-powered field (HPF). A fall was observed in the number of megakaryocyte colony-forming units (CFU-MKs) after the administration of IL-10 compared with those receiving the placebo (P = 0.068). No difference in the change in granulocyte-macrophage CFUs (CFU-GMs), mixed lineage CFUs (CFU-GEMMs) or erythroid burst-forming units (BFU-Es) was observed when comparing the IL-10- vs. placebo-treated groups (P > 0.465). Serum cytokine levels of thrombopoietin (TPO). IL-6 and granulocyte-macrophage colony stimulating factor (GM-CSF) were not decreased following IL-10 administration. In fact, both TPO and GM-CSF appeared to be slightly increased in the serum. All subjects underwent In111-labelled platelet survival studies with liver/spleen scans to assess splenic sequestration prior to and then on day 7 of treatment. A significant reduction in splenic sequestration of platelets (P =0.012) was observed in the IL-10-treated group, but not in the placebo-treated subjects.

Amplification and overexpression of HER-2/neu are uncommon in advanced stage melanoma

The HER-2/neu proto-oncogene is a useful prognostic and predictive biomarker in breast cancer. In addition, use of a humanized monoclonal antibody against HER-2/neu has recently been shown to have efficacy in the treatment of metastatic breast cancer. In order to examine the potential of HER-2/neu as a biomarker and as a target for HER-2/neu monoclonal antibody treatment in melanoma, we examined the HER-2/neu status in 40 advanced stage melanomas. Using fluorescence in situ hybridization for determining the gene amplification status and immunohistochemistry for detecting protein overexpression, we found that only one out of 40 cases of melanoma had an altered HER-2/neu status. These results demonstrated that HER-2/neu amplification and overexpression are not common in advanced stage melanoma and thus, HER-2/neu would have limited value as a biomarker or as a target for immunotherapy in melanoma.

CI-980 in advanced melanoma and hormone refractory prostate cancer

INTRODUCTION

CI-980 is a novel chemotherapeutic agent that inhibits polymerization of tubulin. Preclinical studies have indicated a high level activity of this agent against various tumor cell lines.

METHODS

13 malignant melanoma patients who had failed prior chemotherapy and/or immunotherapy and 13 hormone refractory prostate cancer patients, including 4 who had received prior chemotherapy, were treated in 2 separate NCI-supported clinical trials. Subjects received a recommended phase II dose of CI-980 of 4.5 mg/m2/day by continuous infusion for 72 hours every 3 weeks.

RESULTS

No activity was seen in either study. Toxicity was tolerable with neutropenia being the most common, significant toxicity. Among the melanoma patients, 15% and 31% developed grade 3 and grade 4 neutropenia, while 7% and 38% of the prostate patients developed grade 3 and grade 4 neutropenia, respectively.

CONCLUSIONS

CI-980 at this dose and schedule is ineffective against malignant melanoma and hormone refractory prostate cancer.

Daily subcutaneous ultra-low-dose interleukin 2 with daily low-dose interferon-alpha in patients with advanced renal cell carcinoma

A limited institution Phase II pilot study was performed using a very low-dose combination of daily s.c. interleukin (IL)-2 with IFN-alpha-2b in patients with advanced renal cancer in an attempt to duplicate or increase the response documented with higher dose schedules without the attendant toxicity profile. We selected a dose of IL-2 with documented immunological activity and combined it with clinically active low-dose IFN. Between August 1994 and September 1996, 19 patients with metastatic renal cell carcinoma, who had been judged incapable of tolerating high-dose i.v. IL-2, were treated with IL-2 (1 million units/m2/day) and IFN (1 million units/day), administered s.c. daily. All treatments were administered on an outpatient basis. Virtually all patients had bulky tumor burden with multiple sites of involvement, including five patients with bone metastases. No major objective responses were observed; however, one patient experienced a minor response lasting 13 months, with an associated improvement in performance status. Median survival was 6 months, and 1-year survival was 16%. Toxicity was generally mild and consisted almost entirely of constitutional symptoms. No serious grade 3 or 4 toxicity was observed, although two patients withdrew from treatment due to treatment-related fatigue. On therapy, mild eosinophilia but no lymphocytosis was noted; in fact, peripheral lymphocyte counts decreased, only to rebound after treatment was discontinued. No toxic deaths occurred. Despite the reasonable tolerability of this daily low-dose s.c. regimen, we conclude that this regimen is an ineffective treatment in metastatic renal cell carcinoma patients who are incapable of tolerating high-dose i.v. IL-2.

Phase II trial of biochemotherapy with interferon alpha, dacarbazine, cisplatin and tamoxifen in metastatic melanoma: a Southwest Oncology Group trial

The therapeutic benefit of adding interferon alpha (IFNalpha) to established single-agent and combination chemotherapy regimens for the treatment of metastatic melanoma has not been proven. We designed the present study to estimate the response rate of IFNalpha, dacarbazine, cisplatin and tamoxifen in patients who had not been treated with systemic therapy for advanced disease. Using a schedule similar to that which had previously been shown to favor IFNalpha plus dacarbazine over dacarbazine alone, we treated patients with an "induction" regimen of IFNalpha, 15 mU m(-2) day(-1) intravenously 5 days/week for 3 weeks. Following induction, schedules of IFNalpha, 5 mU m(-2) day(-1) subcutaneously three times a week, and tamoxifen, 10 mg orally twice a day, were begun. Dacarbazine, 250 mg m(-2) day(-1) and cisplatin 33 mg m(-2) day(-1) for 3 consecutive days were repeated every 4 weeks, and subcutaneous IFNalpha and oral tamoxifen were continued until the discontinuation of chemotherapy. We treated 25 patients (18 men and 7 women, median age 52 years) and observed only 1 objective response (response rate 4%, 95% confidence interval 0.1%-20%). The toxicities of the regimen consisted of moderate myelosuppression and constitutional side-effects. On the basis of the low antitumor activity of this regimen, we do not recommend it for further study or for use as standard therapy of metastatic melanoma.