Can Amgen’s Talimogene become an optimal therapy for melanoma?
| Jun 1, 2013
Co-authored by Mara Jeffress, PhD, Associate Consultant
The melanoma market has been transformed in the past two years, following the discovery of BRAF mutations that are observed in approximately 50% of melanoma patients. BRAF-mutant patients are primarily treated with the BRAF inhibitor Zelboraf® (vemurafenib, Genentech/Roche) and this area has been a hot bed of research with recent approvals of Tafinlar® (dabrafenib, GlaxoSmithKline, another BRAF inhibitor) and Mekinist® (trametinib, GlaxoSmithKline, MEK inhibitor). BRAF wild-type patients primarily receive Yervoy® (ipilimumab, Bristol Myers-Squibb) but there is a clear opportunity for other targeted agents to be developed for this patient population. One melanoma contender is talimogene laherparepvec (formerly known as OncoVex GM-CSF, BioVex/Amgen) – a novel oncolytic immunotherapy injected intra-tumorally. This modified herpes simplex virus causes tumor lysis while also stimulating a systemic anti-tumor immune response. This new approach to melanoma therapy was evaluated in a global, randomized, open-label, Phase III trial (OPTiM; NCT00769704) in 436 patients with unresected stage IIIB, IIIC or IV disease. Patients were randomized 2:1 to receive either talimogene laherparepvec intralesionally every two weeks or GM-CSF subcutaneously for the first 14 days of each 28-day cycle.
Early results of the OPTiM trial, released in March, confirmed that the primary endpoint of durable response rate (DRR), defined as the rate of complete or partial response lasting continuously for at least six months, was met (16% versus 2%, p<0.0001). Additional results were presented on June 1 in an oral abstract presentation (#LBA9008) at the 2013 American Society of Clinical Oncology (ASCO) meeting in Chicago. As expected, treatment-naïve patients had higher DRR (24% versus 0%) than previously treated patients (DRR 10% versus 4%). Interestingly, the DRR varied by stage, with Stage III B/C patients (30% of patients) achieving an impressive 33% (versus 0%) with talimogene. Stage IV M1a/M1b/M1c patients (27/21/22% of patients) did not do as well and achieved 16% (versus 2%), 3% (versus 4%) and 8% (versus 4%), respectively. Was this variation in response by stage due to the availability of injectable sites in non-metastatic patients? Did stage III patients receive more injections or a higher dose than stage IV patients? This analysis is planned, according to the presenter, Dr. Andtbacka, however final overall survival (OS) data will be telling and will ultimately hold more weight with the FDA. A trend toward OS (HR = 0.79, 95% CI, 0.61-1.02, not yet significant) was also observed at a predefined interim analysis. The median OS favored talimogene (23.3 versus 19.0 mos) and the curves are diverging. Final OS results are expected in late 2013, potentially in time for the European Society of Medical Oncology (ESMO) meeting in October.
The time to treatment failure (TTF) also favored talimogene with mTTF of 8.2 versus 2.9 mos (HR=0.42, p<0.0001). In addition, the overall response rate (ORR) of 26% (versus 6%) compares favorably to competitors nivolumab (anti-PD-1, Bristol Myers-Squibb; 28%) and Yervoy (anti-CTLA-4; 11%) and includes 11% of patients who achieved a complete response (CR). Of these contenders, Yervoy is the only approved immunotherapy, and it currently enjoys a large share (41%) of the first-line BRAF wild-type population, a solid share (17%) of the BRAF-mutant patients, and is the top regimen in second-line (22-29%) regardless of BRAF status, according to Kantar Health’s CancerMPact® Treatment Architecture US, 2013 report.
The most frequently observed adverse events (AEs) for talimogene versus GM-CSF were fatigue (50% versus 36%), chills (49% versus 9%), pyrexia (43% versus 9%), nausea (36% versus 20%), flu-like symptoms (31% versus 15%), injection site pain (28% versus 6%) and vomiting (21% versus 9%). The only serious adverse event (grade 3-4) greater than 2% was cellulitis (2% versus <1%). Median treatment duration was much lower in the control arm (10 weeks) than the talimogene arm (23 weeks), a fact that Dr. Andtbacka attributed to a high rate of patients in the control arm coming off the trial early. Treatment discontinuations due to AEs were very low (3.8% versus 2.4%). This adverse event profile looks favorable, but does not align with Amgen’s morning press release which stated “serious adverse events occurred in 26 percent of talimogene laherparepvec patients and 13 percent of GM-CSF patients,” which leaves us wondering if the press release is in error or if incomplete AE data was presented. For comparison: nivolumab grade 3-4 AEs occurred in 20% of patients and included lymphopenia (3%), fatigue (2%), diarrhea (2%), abdominal pain (2%) and increased serum lipase (2%). And notably, the toxicity profile for nivolumab appears to be much better than that of Yervoy.
While unusual, the use of durable response rate as a primary endpoint was agreed to as part of a Special Protocol Assessment (SPA) agreed upon by the FDA. The fact that the pivotal Phase III trial of talimogene met its primary endpoint of improving durable response rate bodes well. However, it is unclear whether or not results for this endpoint alone will be sufficient for approval. Recent approvals of other products in this setting have been primarily based on PFS or OS. Thus, the hopes for talimogene may hinge upon the mature OS data expected later this year. In addition, even if approved, physicians may not be entirely comfortable with the non-standard comparator arm, which was a point of criticism by the discussant, Dr. Margolin. She called for more trials of this promising new therapy including new immunotherapy combination trials. Amgen is looking ahead with a newly initiated Phase I/II trial examining the combination of talimogene with Yervoy (NCT01740297).