Keller, Daniel M. PhD
PHILADELPHIA—In a randomized, controlled, Phase III trial, intralesional injection of a herpes simplex virus (HSV) engineered to express human granulocyte macrophage colony-stimulating factor (GM-CSF) was superior to use of subcutaneously injected GM-CSF alone in providing durable responses in Stage IIIB/IV melanoma. Interim analysis also suggests an advantage in overall survival in Stage IIIB disease.
The viral vector, talimogene laherparepvec (T-VEC) [OncoVEX(GM-CSF); BioVex], is an oncolytic immunotherapy consisting of an HSV Type 1 engineered to selectively replicate in tumors and express human GM-CSF to exert both local and systemic antitumor effects.
Reporting results of the OPTim (OncoVEX Pivotal Trial in Melanoma) here at the International Meeting of the Society for Melanoma Research, Howard L. Kaufman, MD (until recently Director of Rush University Cancer Center and now Associate Director for Clinical Sciences at Rutgers Cancer Institute of New Jersey), said that although ipilim-umab and other new therapies show efficacy in metastatic melanoma, not all tumors respond and additional new therapies are still greatly needed.
Explaining the mechanism, he said that locally, T-VEC virus replicates in tumor cells and lyses them. GM-CSF expression activates dendritic cells, which present tumor-specific antigen to CD4+ helper T cells, activating CD8+ cytotoxic T cells. These cytotoxic T cells can then circulate and kill distant tumor cells.
The trial included patients with unresectable Stage IIIB/IV melanoma with at least one injectable lesion, randomized 2:1 to intralesional T-VEC once (295 patients), and then after three weeks, T-VEC every two weeks. A control arm of 141 patients received subcutaneous GM-CSF for 14 days of every 28-day cycle. The primary endpoint was a durable response rate (complete or partial) beginning any time within 12 months of initiation of therapy and lasting six months or longer. Stage IV patients could have up to three visceral metastases other than lung lesions, none larger than 3 cm. Participants could not have any open herpetic skin lesions or be on chronic anti-herpetic agents.
The T-VEC and GM-CSF groups were well balanced, and 70 percent of patients had Stage IV disease; half were older than 65, 57 percent were men; 47 percent were being treated in this study as first-line therapy; and 58 percent were HSV-positive.
Kaufman reviewed previously reported results, in which the response rate was about 16 percent with patients treated with T-VEC compared with two percent for those receiving GM-CSF alone, giving an unadjusted odds ratio of 8.9, and this was significant, he said, meeting the primary endpoint, as determined by an independent assessment committee.
Objective responses were determined by the investigators, giving an objective response rate (sum of the complete and partial responses) of 26.4 percent for patients receiving T-VEC compared with 5.7 percent for those treated with GM-CSF alone. “We had an almost 11 percent complete response rate in patients with T-VEC, and this included regression of not only injected lesions but also had to include regression of all non-injected lesions as well,” he said. The complete response rate with GM-CSF alone was 0.7 percent.
New data include prespecified subgroup analyses and estimates of overall survival in the intent-to-treat population.
For all of the subsets, the patients treated with T-VEC generally had better response in terms of both durable and objective response rates: “There was an especially significant response in patients who had unresected Stage IIIB or C or IV/M1a, and the durable response rate was 33 percent compared with zero percent [for GM-CSF] in the Stage IIIB/C population, and in terms of objective response rate, was 52 percent for patients receiving T-VEC compared with two percent for GM-CSF for patients with Stage IIIB/C,” he said.
For Stage IV/M1a, the rate of durable responses was 16 percent for T-VEC versus two percent with GM-CSF, and objective responses were 27 versus two percent, respectively.
Patients who received T-VEC as first-line therapy had a durable response rate of 24 percent and an objective response rate of 38 percent, versus zero and five percent, respectively, with GM-CSF. About half of responding patients did so after initially progressing. HSV-1 positivity or negativity did not appear to affect the response to T-VEC.
Most Responders Still Responding
Kaufman noted that at the last assessment most responders were still in remission, adding that several of the T-VEC responders have made it past 12 months, regardless of whether they are complete or partial responders. Although far fewer patients responded to GM-CSF (just eight patients) than to T-VEC (78 percent, with twice as many patients in that arm), similar proportions of responders were estimated to still be in response past 12 months—53 percent for T-VEC and 46.9 percent for GM-CSF.
Interim overall survival estimates, representing 85 percent of the required 290 events, are a median of 23.3 months for patients receiving T-VEC and 19.0 months for those on GM-CSF, he said—a difference of 4.3 months, and giving a hazard ratio of 0.79. Key covariates independently favoring T-VEC in this interim analysis are first-line therapy, Stage IIIB/C disease, no visceral disease, and an ECOG performance status of 1.
Patients tolerated T-VEC fairly well, he said. The major adverse events were fatigue and chills, affecting about half of the patients receiving T-VEC, with roughly one third or more experiencing fever, nausea, influenza-like illness, or injection site pain. Fatigue was generally low grade. Grade 3/4 instances of cellulitis, fatigue, vomiting, dehydration, deep vein thrombosis, or tumor pain each affected about two percent of patients. Vitiligo was reported in five percent of patients on T-VEC and in percent on GM-CSF alone.
Final analysis is expected to be available sometime later this year, he said, adding that the subset analysis suggested that T-VEC may be most useful in treating patients with Stage IIIB/C or IV/M1a disease and in those patients who receive T-VEC as first-line therapy. He and his colleagues also plan to test T-VEC in combination with other immunotherapies to try to boost the response.
Caution from Vernon Sondak
Following the presentation, Vernon Sondak, MD, Chief of the Division of Cutaneous Oncology at H. Lee Moffitt Cancer Center, strongly advised during the question-and-answer period that the study results must be interpreted in light of the design of the study and patient inclusion criteria, including the endpoint, the control arm, patient selection, and the choice of patients with or without uninjected lesions: “All of these factors are so critical in, first of all, getting a successful trial and then subsequently figuring out where this kind of therapy fits into the management algorithm for our patients,” he said.
Responding, Kaufman thanked him for the comment and added that designing such studies has been a challenge. Particularly in light of what researchers have been learning about immunotherapy in the past two years and delayed kinetics of the response, “it makes it very challenging to pick an endpoint that we could really use as a surrogate of overall survival before we enter into these large Phase III studies,” he said. He mentioned that he has done work looking at durable responses, which appear to be a reasonable surrogate endpoint for overall survival given that many of the patients who responded in this trial had initially progressed before going on to have durable responses.
Answering another question from the audience, Kaufman said that prior therapy with T-VEC did not appear to change the response to subsequent ipilimumab.
In a video interview at the meeting with OT reporter Dan Keller, Patrick Hwu, MD, Chair of the Department of Melanoma Medical Oncology at the University of Texas MD Anderson Cancer Center, discusses topics from a session on immunotherapy of melanoma that he led, specifically: drugs that unblock the immune response, combined immunotherapies, and immunotherapies combined with other targeted drugs. He also summarizes presentations on the need for biomarkers for patient selection and as potential therapeutic targets themselves, the utility of immunotherapy plus radiation, and stimulation of antigen-presenting cells.
© 2014 by Lippincott Williams & Wilkins, Inc.