Here's my analysis:
First, I'd like to distinguish between analysis of the underlying trial and analysis of the way they present it.
Second, pre-existing history is always significant, and APHT has not had a distinguished history on this front - their releases in the past have often been overly enthusiastic. Further, I know that Rick has always been skeptical of this project and this company, and that is another significant warning flag for me. On the plus side, they do have some partnerships with big pharma.
So let's look at the trial itself and the science behind it. On the trial itself, the obvious big issue is that it's not a comparative trial. It's over 100 patients, which is a decent size for a Phase II for a small company. It's US and Europe, which is good (no 3rd world locations). Enrollment in a trial like this is always going to be fairly tough simply because there is a narrow window for eligibility - not pre-treated and decent Karnofsky index with a life expectancy of at least 3-months, which means you have to get them essentially immediately after diagnosis. (Life expectancy for this disease untreated is not much more than 3 months).
Next the science. Note first that this is not traditional cancer immunotherapy (which would have given it close to three strikes for me immediately). Instead this is more along the lines of anti-hormone therapy, analogous to prostate and breast cancer treatment. Based on looking at this previously I am satisfied that knocking down gastrin is a pretty sensible thing to do in gastric and pancreatic ca. But immunizing against one's own hormones has to be hard - you have to overcome self-tolerance.
OK, now let?s look at the release. First I note that they present ITT numbers throughout - that's good - one whole set of spin possibilities go away. Second I note that they use the RECIST criteria - that's also good, as they are conservative and you don't have to worry about "home-grown" definitions of response.
Next I note they present the "Best" outcome under RECIST. This sounds dubious on the surface, but it isn't - I think it's the most reliable measure. "Confirmed" sounds like it should be better (and it is more conservative), but it relies on a follow-up at just the right time and so (in my view) it's actually less useful in making cross-trial comparisons.
So then what about the 51% response number? Given no comparator arm, you have to look cross-study, which is never desirable. I note they cite a similar trial - the EORTC trial, with (they say) a 20% response. My first thought is whether that was also a "Best" response or not. I dig up the trial (with some difficulty because they don't give a cite) and see that it was indeed a Best response, based on 81 patients. I note also the EORTC dosing was the same. Other EORTC numbers - 21% progressive disease, which is similar to this trial, so about the same number of patients have no response at all. For EORTC, 7.2 months median survival, no complete responses at all. So median survival somewhat over 20% longer for this trial.
So it looks to me based on the improved response rate (more than double) that there is something here.
Next, what to make of the sub-group analysis based on those that generated ant-G17 antibodies? 3 months vs. 10 months is a significant difference. Now an obvious question is whether the ability to generate antibodies is simply a proxy for overall status. They give a pretty good response to this in my mind, looking at both KS status and ability to generate other antibodies. But the possibility remains that all they have done is find a proxy for people who are going to die quickly. Now what difference does this analysis make from a practical perspective? Maybe not very much, but it is possibly/likely indicative that the drug is working like they say it is (and also raises the issue that some other, non-immunological, method of blocking gastrin might be better and help the other 20% as well).
So far, so good. But then they spoil things by referring to the 5-year survival rate with current treatment - for that they lose a point or two, to go with the point they lost by not emphasizing the non-randomness of their trial.
So where does this leave me overall? A non-randomized trial has a built-in 5/10 limit in my mind, but they probably earn that 5. The release itself is actually pretty good - I'd say a 7.
What about from an investment perspective? I actually own a few shares here, bought around a year ago at less than half the current price. Well the results are not spectacular, but then the additional tox costs were low as well. If you can figure out quickly who isn't generating antibody, then that improves the cost-benefit ratio (sort of like testing for Herceptin responsiveness, but not as clean).
So my overall sense is that they have probably something here, and I'll hang onto my few shares. Not tempted to increase my stake at these levels, however.
Peter
Here's the EORTC abstract:
Journal of Clinical Oncology, Vol 18, Issue 14 (July), 2000: 2648-2657
© 2000 American Society for Clinical Oncology
Final Results of a Randomized Phase III Trial of Sequential High-Dose Methotrexate, Fluorouracil, and Doxorubicin Versus Etoposide, Leucovorin, and Fluorouracil Versus Infusional Fluorouracil and Cisplatin in Advanced Gastric Cancer: A Trial of the European Organization for Research and Treatment of Cancer Gastrointestinal Tract Cancer Cooperative Group
By Udo Vanhoefer etc.
PURPOSE: To compare the efficacy and tolerability of etoposide, leucovorin, and bolus fluorouracil (ELF) or infusional fluorouracil plus cisplatin (FUP) with that of the reference protocol of fluorouracil, doxorubicin, and methotrexate (FAMTX) in advanced gastric cancer.
PATIENTS AND METHODS: A total of 399 patients with advanced adenocarcinoma of the stomach were randomized and analyzed for toxicity, tumor response, and progression-free and overall survival. Only reviewed and confirmed responses were considered. The analysis of remission was based on assessable patients with documented measurable lesions. The intent-to-treat principle, log-rank test, and Cox regression model were used for the statistical analysis of time-to-event end points.
RESULTS: The overall response rate for 245 eligible patients with measurable disease was 9% with ELF, 20% with FUP, and 12% with FAMTX, with no significant differences. One hundred twelve patients were eligible for efficacy in assessable, nonmeasurable disease. No change was observed in 66% of patients treated with ELF, 56% with FUP, and 55% with FAMTX. Two patients achieved a complete tumor regression (one each for ELF and FAMTX). With a median follow-up time of 4.5 years, the median survival times were 7.2 months with ELF, 7.2 months with FUP, and 6.7 months with FAMTX, respectively, with no significant differences. Nonhematologic and hematologic toxicities of ELF, FUP, and FAMTX were acceptable, with neutropenia being the major toxicity for all three regimens. Seven treatment-related deaths occurred (two with FUP and five with FAMTX).
CONCLUSION: All three investigated regimens demonstrate modest clinical efficacy and should not be regarded as standard treatment for advanced gastric cancer. New strategies should be considered to achieve a better clinical efficacy in the treatment of advanced gastric cancer. |
