Tremelimumab dose selected for further testing due to favorable safety, tumor response in melanoma

Issue: February 25, 2009

Tremelimumab given at two dosing regimens afforded durable tumor responses in patients with metastatic melanoma, according to data from a phase-1/2 trial.

Researchers conducted a phase-1/2 trial to assess the safety of tremelimumab in multiple doses and to examine the efficacy of the agent and the appropriate dosing regimen.

To determine the recommended phase-2 dose, phase-1 IV infusions of tremelimumab at 3 mg/kg, 6 mg/kg or 10 mg/kg were given to 28 patients with metastatic melanoma for up to one year. During phase 2, 89 patients received tremelimumab 10 mg/kg once each month or 15 mg/kg every three months.

The researchers reported no dose-limiting toxicity during phase 1. Eighty-four patients were assessable for response during phase 2 at which time 10% reached objective antitumor responses; one complete response and three partial responses in each dosing regimen comprised the best overall objective response.

Most responses ranged from three to 30 or more months and the most common adverse events were diarrhea, rash and pruritus. Grade-3/4 adverse events had a frequency of 13% in the 15 mg/kg arm and 27% in the 10 mg/kg arm. Frequency of serious adverse events was 9% in the 15 mg/kg arm vs. 23% in the 10 mg/kg arm.

“Both phase-2 regimens generated durable tumor responses,” the researchers wrote. “Based on its more favorable safety profile, 15 mg/kg every three months was selected for further clinical testing.” – by Stacey L. Adams

J Clin Oncol. 2009;doi:10.1200/JCO.2008.19.2435.

The phase-3 data from ASCO, in terms of survival, show that tremelimumab given at 15 mg/kg every three months is a bit better than for the control arm of dacarbazine (DTIC-Dome, Bayer Healthcare) or temozolomide (Temodar, Schering). But, the difference was not statistically significant; median survival was a month better, but this is less than we were looking for. So we must ask: What is the real difference between clinical responses to tremelimumab and dacarbazine? In the recently published study of Camacho et al reporting the phase-1/2 experience and the larger phase-2 multicenter study I presented to ESMO and ASCO in 2008, as well as in the phase-3 experience that Toni Ribas presented to ASCO, there are very profound differences between the type of responses seen with anti-CTLA4-blocking antibodies and with conventional chemotherapy. About 8% to 10% of patients with melanoma have objective response to this modality and the majority of these have a very durable response. For example, in the phase-2 trial results I reported to ASCO and ESMO, 15 of 16 patients who had objective response with tremelimumab response was durable past six months -a very different pattern than we see for dacarbazine and temozolomide. Tremelimumab and ipilimumab are active agents in a fraction that is roughly the same as interleukin-2, interferon-alfa, and dacarbazine or temozolomide. In the largest trial of temozolomide ever conducted - results recently presented at ESMO by Patel et al.- response rates were 10% for dacarbazine vs. 14.8% for temozolomide, but there was no difference in the overall survival or the progression-free interval between these regimens. Dacarbazine and temozolomide responses do not seem to confer survival benefit in part because these responses are rarely durable. The 10% of patients who responded to the anti-CTLA4-blocking antibodies exhibit characteristically and qualitatively different kinds of responses in metastatic stage-4 disease.

When looking at the difference between the phase-3 and phase-1/2 trials, the phase-1/2 results clearly show that for tremelimumab there are very durable, high-quality responses--and the same can be said for ipilimumab. But the problem is, when you conduct a conventional trial in advanced metastatic melanoma to look at median survival, the survival median provides a different readout than the readout of durable responses. The quality of responses to tremelimumab and ipilimumab are still worthy of pursuit. The fact that the phase-3 trial of tremelimumab shows a little better outcome than dacarbazine in terms of median survival does not reflect these high-quality responses that go out well past six months, and will take different follow-up to document. So that is what is going on now: the long-term follow-up of those patients to make a qualitative assessment of the kind of response that they exhibited is really what we need.

Looking at the data from the phase-3 presentation by Ribas, the one thing that may have weighed differentially in favor of tremelimumab in the forest plot of the data was higher lactate dehydrogenase (LDH). This is curious when you think about it. It may or may not be real, but if patients who have higher LDH (perhaps worse disease) benefit more from tremelimumab than from dacarbazine or temozolomide, the study was designed in a way that may have disfavored the anti-CTLA4-blocking antibody. The protocol excluded those people who had higher than twofold elevated LDH – so, by the design of the trial, curiously the outcome may have been slanted in favor of temozolomide or dacarbazine.

Finally, it remains to test the role of anti-CTLA4 blocking antibodies in the disease setting where we have found the greatest relative benefit for other immunotherapy, such as IFN alpha. The adjuvant exploration of anti-CTLA4 blocking antibodies may show even greater relative benefit and warrants phase-3 study in relationto IFN alpha.

– John Kirkwood, MD

HemOnc Today Editorial Board member