Long-term survival results of a randomized multinational phase 3 trial of dacarbazine (DTIC) with or without Bcl-2 antisense (oblimersen sodium) in patients with advanced malignant melanoma

Reviewer: S. Jack Wei, MD
University of Pennsylvania School of Medicine
Last Modified: May 14, 2005

Presenter: J.M. Kirkwood
Presenter's Affiliation: University of Pittsburgh Cancer Institute
Type of Session: Scientific


  • Oblimersen sodium (Genasense, G3139) is an antisense oligonucleotide that selectively targets and specifically binds bcl-2 mRNA resulting in decreased bcl-2 protein translation and increased apoptosis when added to chemotherapy.
  • Results of this trial examining the use of G3139 in conjunction with dacabazine (DTIC) in the treatment of advanced maligant melanoma (MM) have previously been reported and showed no significant difference in overall survival (OS) compared to DTIC alone.
  • However, the minimum follow-up in this report was 6 months, and subsequent analyses with minimum follow-up of 15 and 18 months did show a significant difference in survival with the addition of G3139.
  • This update reports the outcomes in this study with a minimum follow-up of 24 months.

Materials and Methods

  • This trial was a multicenter, open-label trial (139 sites, 9 countries) which randomized 771 paients to either DTIC (1000mg/m2) alone OR G3139 (7 mg/kg/d x 5 ) + DTIC (1000 mg/m2).
  • Patients were stratified by ECOG performance status (0 vs. 1-2), liver metastases (yes vs. no), and no visceral metastases AND normal LDH vs. any viseral metastases OR elevated LDH
  • Patients were treated in q21 day cycles (max of 8 cycles)
  • Restaging was performed every 2 cycles
  • No cross-over of arms was allowed
  • Primary endpoint of analysis was OS with secondary endpoints of progression-free survival (PFS), antitumor response (RECIST), durable response (>=6mo), and safety


  • The two groups were well balanced for age, gender, and performance status
  • A higher percentage of patients receiving G3139 has LDH levels >2X normal (24% vs. 16%)
  • There was no difference in the mean and median dose of DTIC received between the two treatment groups.
  • Overall response was improved in the G3139 group (12.4% vs. 6.8%, p=0.007).
  • Durable response was improved in the G3139 group (7% vs. 3%, p=0.02).
  • Median PFS was improved in the G3139 group (2.4 mo vs. 1.6 mo, p=0.0003).
  • Median OS was non-significantly improved in the G3139 group (9.0 mo vs. 7.8 mo, p=0.08).
  • When analyzed by baseline LDH, patients with normal LDH showed a significant improvement with the use of G3139 (HR 0.80, p=0.02).
  • There was no improvement with G3139 in patients with LDH elevated >2x normal (p=0.85).
  • An increased incidence of grade 3-4 thrombocytopenia was seen in the G3139 group (16% vs. 6%); however, there was no difference in the rate of serious bleeding of need for platelet transfusions.
  • There was also a higher rate of grade 3-4 neutropenia in the G3139 group (21% vs. 13%) but no difference in serious neutropenic infections.
  • A higher rate of adverse events (AE) leading to discontinuation of therapy was seen inthe G3139 group (19% vs. 11%), but there was no difference in the rate of AE leading to death.

Author's Conclusions

  • The addition of G3139 to DTIC results in improved PFS in all patients and improved OS in patients with normal LDH levels.
  • The follow-up of this trial should be extended to 5 years.
  • Although the G3139 improves the outcomes in patients with advanced MM, the overall survival remains low and new combinations of chemotherapy should be explored to improve these outcomes.
  • The impact of bcl-2 on tumor tissue should be analyzed.
  • A new formulation of G3139 that can be delivered subcutaneously should be explored in clinical trials.

Clinical/Scientific Implications
The update presented here shows improved PFS with the addition of the bcl-2 antisense compound, G3139, to DTIC in patients with advanced MM.  Importantly, a subset of patients, namely those with normal LDH, were found to have improved OS with the addition of G3139.  It is important to note that this subset analysis is valid, given the fact that the patients were stratified for LDH level prior to randomization.  The benefit of G3139 was seen at multiple timepoints, including 12 and 18 months.  Further follow-up is needed to see if this difference holds up.  It is reasonable to consider the addition of G3139 to DTIC in patients with MM with normal LDH levels.  Given these results, further studies will likely focus on the use of G3139 in melanoma patients with earlier stage disease.  Although these results are promising, the overall outcome of these patients remains poor, and research into alternative combinations including G3139 are needed.