Sorafenib is a TKI similar to XL 184. Its brand name is Nexavar and it is a drug Onyx partnered with Bayer and it is currently approved in 2 indications. It is approved for frontline hepatocellular cancer (HCC) and second line renal cell cancer (RCC). In 2009 worldwide revenue was $800+ million.
Interestingly, in its RCC indication, sorafenib was supplanted by sunitinib (Sutent), another TKI, this one a Pfizer drug. In both HCC and RCC sorafenib was approved based on phase 3 trials vs placebo with PFS endpoints. In both trials, the drug made its PFS endpoint and subsequently showed a survival advantage. In both indications, sorafenib did a good job arresting/slowing tumor growth, but had relatively low objective response rates, 10% in RCC and only 2% in HCC.
Sunitinib demonstrated a very high objective response rate in RCC, approx 40%. It was approved for refractory RCC based on acheiving a minimum 15% ORR at a 80% confidence level. To simplify, it had to show about a 20% ORR in a trial with at least 100 patients. Susequently in a phase 3 trial in RCC, sunitinib showed a PFS advantage vs the approved standard cytokine therapy, interferon alpha. To summarize, sorafenib was approved for 2nd line RCC by a placebo controlled trial. Sunitinib was initially approved for refractory patients based on ORR, then was approved for frontline based on besting the approved frontline therapy, IFN-a. It essentially trumped sorafenib in RCC without a direct comparison trial.
HCC was a different story. Sorafenib is the only systemic approved drug. In uncontrolled trials, Sutent showed promise, but had a low ORR, too low to employ a single arm ORR endpoint strategy as was done in RCC. Pfizer instead launched a full phase 3 trial vs a sorafenib control. Unlike RCC, sunitinib was found to be inferior in this indication and the trial was discontinued early.
That brings us to XL 184. In the randomized discontinuation trial, 16 HCC patients were evaluable for response. 3/16 were reported to have a partial response, including 2 sorafenib refractory patients. If XL 184 can show an ORR of approx 20% in sorafenib refractory patients in a sample size of 100+ patients, an accelerated conditional approval is a likely outcome. So far the sample size is very small, but the results are very encouraging.
One also has to speculate about a possible XL184 vs sorafenib frontline HCC trial. Will the superior response rate of XL184 translate into a survival advantage? In its phase three trial in HCC, of 299 patients treated with sorafenib, 7 enjoyed a partial response. Sorafenib 7/299 vs XL184 3/16. We can only wait and anticipate.
<<While it was a negative at the time, the new data makes it possible EXEL could make a better deal for xl184 than they had before.>>
Hey Vin. It's an exciting time at EXEL. They own 100% of a red hot drug. As investors, I hope we get to realize at least a sizable fraction of what I believe this is worth. There could be a partnership or an outright sale of the company. I hope that the recent data will get C. Icahn's attention and he will seek to play a higher profile role in the strategic moves that will unfold in the next 1-2 years.
I'd agree that the HCC, particularly the observation of responses in post-sorafenib HCC patients is very encouraging. As I count it, in the post-sorafenib group, there are 2 PRs in 9 patients (one is very early) which gives a 20-25% RR estimate. While that is very encouraging, to my knowledge, response rate has not been demonstrated to be a surrogate in HCC and hence I would say the deck is stacked against a single agent RR trial leading to accelerated approval. For the RDT, the HCC population treated with HCC is pretty good, all the more reason to require a randomized trial.
Interestingly, BMS has bet pretty big on HCC with the following trials with brivanib, a VEGFR, FGFR inhibitor:
1) progressed or intolerant to sorafenib(randomized trial with survival as primary endpoint)
2)in first line HCC patients vs. sorafenib (randomized trial with survival as primary endpoint)
If you look at the trial record links, the number of sites and duration of the respective trials is substantial.
With that said, my recollection of the brivanib phase 2 data (ASCO in the 2007-2009 time frame) is that is was not overly impressive.
The question for XL184 will really be, as you noted, how the XL184 data look over time in terms of both safety and efficacy. In looking over the safety data, I was more concerned by the profile of this agent in this population than I had been in other settings. There is 100% incidence of fatigue, including 17% of grade 3. I noted a few grade 3 episodes of thrombocytopenia and also a 67% incidence of diarrhea including an 11% incidence of grade 3. For a chronically adminstered drug, these side effects can end up being intolerable.
I'd have to believe that BMS talked to the FDA about the possibility of doing a RR trial in the post-sorafenib population and were told no though there is no way to know for sure.
Nonetheless, these data will need to mature on both the safety and efficacy front before being able to get the FDA to buy into a single arm study. I won't say never, but I would imagine it would take a pretty clean population of approximately 100-150 patients with a response rate of >25% (confidence intervals in the range that would exclude less than 20%RR). The duration of response would have to be pretty significant as well. The patients shown on the poster have done pretty well so on this fron I'm encouraged so far. The other issue is pronostic factors within the patient population. There would have to be a pretty robust argument that the population is not predisposed to have a better prognosis by looking at a variety of prognostic factors in the population treated vs. the generic HCC population (a very difficult argument to make). Looking at where the presented trial is enrolling (USA and Taiwan) may present some challenges with having a well defined population.
My guess is that they use this trial to address multiple issues to optimize either a randomized phase 2 trial (which if very positive could be used for registration based on PFS and survival) or a phase 3.
I guess the concern here might be enrollment rate. If it's taken about 9 months to get almost 10 treated post-sorafenib patients, getting 100 such patients could take a while. I think for this, we'll need to wait until after ASCO next year (unless there is something at ASCO GI).
Good luck all and despite the potential conerns I raise, note that I'm encouraged by what looks like a data set that clearly demonstrates activity in a difficult indication.
<<Good luck all and despite the potential conerns I raise, note that I'm encouraged by what looks like a data set that clearly demonstrates activity in a difficult indication.>>
I agree Doc. I see 3 very encouraging signals. The activity on HRPC bony mets, the ovarian cancer response rate and the HCC response rate. I wouldn't be surprised to see good activity on Brca bony mets also.
The side effect profile is a valid concern. One of the things that stuck out in the EORTC presentation was the mention that 40% of patients do get a dose reduction. The initial experience in GBM seems to have driven home the message that they need to be aggressive in managing the side effects as they occur so as to prevent a forced discontinuation.
I don't see much trouble recruiting for pivotal phase 2 trial if there is one. With an uncontrolled trial there is no risk of randomization to placebo. It's a daily pill or capsule. Once word is out that the drug has shown preliminary efficacy, it shouldn't be a hard sell . There are other competing trials, but XL184 on paper has to look like a pretty good alternative for patients who have run out of approved options.
Now that the EORTC abstracts are out, I suspect recruitment in all the tumor types will become easier. We will hear some discussion on the clinical plans at the R&D day next week, plus possible updates to the RDT data set.
<my recollection of the brivanib phase 2 data (ASCO in the 2007-2009 time frame) is that is was not overly impressive. >
You are right. The data weren't impressive as reported at ASCO.
Results: Data are available for 22 pts. The median age of brivanib-treated pts was 55 years, 73% were male, and 73% were Asian. Of the 22 pts, 20 had failed sorafenib and 2 thalidomide. Most AEs were low grade (CTC grade 1 or 2). The most frequently reported AEs (>25%) were fatigue (63.6%), diarrhea and anorexia (40.9% each), vomiting and hypertension (36.4% each), and constipation and nausea (27.3% each). Nineteen pts had an investigator assessment of efficacy, and 11 (58%) had stable disease. Twenty-one pts had measurements of alpha fetoprotein (AFP) levels (82% had elevated AFP at baseline) and 43% had a >50% reduction on study. Conclusions: These preliminary data suggest that brivanib alaninate is generally well tolerated in pts who have failed sorafenib or thalidomide treatment.
<I'd have to believe that BMS talked to the FDA about the possibility of doing a RR trial in the post-sorafenib population and were told no though there is no way to know for sure.>
I don't think they have. Brivanib data didn't suggest RR trial was a possibility.
Going against Nexavar directly in HCC is too risky. By ASCO next year, we might have better idea if XL184 should go for Nexavar refractory indication.
On a different note, Nexavar is one of the successes from RDT. RCC cohort from phase 2 RDT was used for Nexavar approval in RCC.
Ernie, in reply to speculating of XL184 extending life. My comment would be that if XL184 shrinks tumors why would it not extend life more than DNDN. My beief after viewing some of its images affecting tumors that it will not only extend life, but maybe a cure as well if detected early. I have viewed some of XL'S tumor shrinkage of graphs and it is very exciting seeing what it does to help patients. I suspect that we may hear some good news coming soon on that front.