APHT exercise - please let me know anyone if I missed posting your entry:
--------------
Re.: Aphton
I would be careful here. It is of course difficult to come to a conclusion based on just a press release. In addition Aphton is sometimes claimed to be rather creative when it comes to their releases.
The problem I have with the release is of course that there is no control group. Instead Aphton tries to define as a control group those patients within the treatment group, who were not able to generate antibodies to G17. This at first sight looks striking, but is necessarily not indicative of success. There is the possibility that those who were not able to generate antibodies were generally seen "sicker" although Aphton tries to argue that they were not.
Erik
----------------
APTH Release
I don?t understand this trial without a control. I presume the RECIST criteria are used in assessing the tumor radiographically. How are responses confirmed? 2% are complete responders using RECIST criteria, but only 1% is have confirmed tumor response, suggesting one apparent complete response was not confirmed. I thought a complete response would be unequivocal.
Clearly generation of anti-G17 antibodies boded well for the patients in this trial, 10 rather than 3 months of life is meaningful and not merely statistically significant. The release addresses the question as to whether the ability to generate the anti-G17 antibodies is a correlate rather than a cause of prolonged survival, but it doesn?t fully answer the question. Looking at sub-groups defined by KPS strata is ok if it is accepted that KPS is a good predictor of survival, but only if. If KPS scores are independent of life expectance, then the ability to generate might suggest longer survival without playing a direct role in that survival. Maybe those patients that demonstrated that ability were inherently stronger in some way. It wasn?t simply that all the non-responders were completely immune-compromised, as the release pointed out. Nevertheless there still could be some explanation along those lines that would explain the results of the experiment without concluding that the drug had helped.
Ashley
---------------
peter,
i'd say 2.5
1. there was no comparator arm - - for all we know, g17dt
had no effect and whatever effect there was, was because of the standard of care drugs
2. intent to treat analysis could skew results
3. how many people were in intent to treat population?
4. use of subgroup analysis is spurious
5. is recist the standard by which tumors are measured?
6." The addition of G17DT did not exacerbate the known systemic toxicity profile of cisplatin and 5-FU." - - but perhaps it caused other bad side effects not associated with the standard of care drugs.
7. what were the side effects?
feel free to use my name.
sales
------------------------
My two cents on the Aphton results . . .
With the usual caveat about inter-trial comparisons . . .
I don't think Aphton has much here. While they've managed ~20% improvement in time to progression and median survival, it only amounts to about a month and a half in absolute terms. Meanwhile, the FUP regimen seems to demonstrate a much more durable response, though in a somewhat smaller population. The 1 year survival rate for the Aphton combo was 19%, while FUP alone has demonstrated 29%. In fairness to Aphton, it is possible that the level of care varied a lot at all those sites; perhaps it was better for the patients of the cited non-Aphton trials, thus yielding better 1 year rates?
If I'm an oncologist, I'm scratching my head over the ethical conundrum given the mixed and divergent comparison: better chance for stronger, more widespread response (Aphton + FUP combo) versus more durable response (FUP combo). If I'm an insurer, I'm thinking "the choice is not clear, so why should we pay for the more expensive version?"
But . . . There were two complete responses for the Aphton combo. None for FUP (Other regimens with FU have shown a couple of CRs, too, so maybe another FUP study and the law of averages would catch up with it. Maybe not.) The Japanese study found that FUP was more toxic than FU, anyhow. Maybe they should try a PII with just FU?
Finally, it would appear that DCF or ECF would be better standards of care (per the first & last referenced links), and the Aphton results look even more marginal against that.
Might get approved, but won't sell in this indication, says I.
Reference links:
biz.yahoo.com
jco.org
jco.org
jco.org
medscape.com (must register)
Cheers, Tuck
------------------
Peter,
(I think your idea about these collaborations are quite interesting. I don't have much time to post write now so sorry if things are a bit fragmented.)
- first I think this study should never have been preformed as a single arm study. The lack of a control group makes interpretation of results questionable. If they had enrolled another 100 pts in a control group and gotten reasonable results they may have been able to file for approval. To me this trial was a waste of money.
- the release states the primary endpoint was tumor response but then goes on to report "Best overall" and less prominently "confirmed" response rates with out saying which is the primary. I would guess that confirmed response is primary as usual requirement to have response confirmed by at least 2 consecutive assessments. Also they don't tell us if there is any statistical test associated with the primary endpoint. eg ( response rate > 20%).
- The results with the immune responsiveness and survival are of no value. One can not conclude if showing immune activity is an effect of ,or a cause of patients living longer and thus getting more doses. These kinds of analyses seem popular with many, however FDA will ignore these as will good cancer researchers.
- there are no confidence limits on the various efficacy estimates( response rates. median survival, median time to progression).
- The statement about safety could be a bit shady. It claims to not exacerbate the known toxicity but really doesn't say that it didn't cause any new toxicity. Perhaps it is just a poor choice of words but I prefer a more clear statement accompanied by a list of most frequent toxicites and their frequencies. If you really have nothing t hide than show the results.
- Bottom line to me this trial shows 28% response (19%-37% 95% CI) not different from vs. historical 20%. but may be some signs of efficacy.
my overall rating of the quality and completeness of the release is a 5.
bio_kruncher
----------------------
Hi Peter.
APHT -
sorry, didn't have enough time to really think about it, as other stuff on.
Gut reaction on a quick readthrough is give it a 5. Didn't pick up on any blatant dishonesty, but couldn't see that the results were terrifically exciting given lack of control arm. Would have liked more detail on figures, and find the feelgood quotes at the end unnecessary in this context.
Like the idea you came up with here a lot, & would hope to give it a bit more effort next time...
Nigel
-----------------
I'm not in a position to evaluate any of the science based assertions, as you know, Peter. So when Aphton states that Gastrin 17 is believed to be a central growth factor or initiating signal for cell growth.... etc, I have to take it at face value.
The trial was small (it was a Phase II) and spread over 42 sites, which might lead one to question whether there was any effective overriding control to ensure the comparability of the findings from site to site. At the very least, it's an unwieldy study (made necessary by the relative uncommonness of the disease?). OTOH, it was a phase II study, without a control arm, tho there is a reference to a European study. APHT asserts comparability. We have to take it at face value, since there is no description of the methodology of the European study.
APHT hangs a good part of its hat on the finding that those who generated antiG17 antibody response had a substantially greater benefit than those who did not generate such a response. There apparently is a possible objection that the non-responders were immunocompromised. This is answered by the statement there is evidence that ?many? of the non responders were immunocompetent. How many?
The conclusions are modest ? APTH has not found the Holy Grail, and it doesn't claim to. It claims only that it would be a useful adjuvant therapy to the standard cisplatin 5-FU therapy.
Your question was not whether the study was convincing, but whether the release intended to deceive. You did not ask whether it would lead anyone to buy the stock (in my case it wouldn't). My conclusion is that the release was fuzzy but not deceptive.
Give it a 7.
No problem with making this public. (There must be a fishhook in there that I've missed, otherwise you wouldn't have set it as assignment.)
Tom
------------
I thought it might be interesting for those of you who have a medical or science background (or have developed it) to see how someone lacking in the appropriate tools might look at this case. As background, I've been following biotech stocks for many years, and have developed a "feel" for evaluating company material. But I'm not in the same league with the board's usual contributers.
First, there is no control arm for comparison. A major "watchout" in my mind, altho I know there are situations where this is not practical. In this case, I don't see an obvious reason for not having a control. Next, there are only 102 subjects even tho it's a multi-country study involving 42 sites. Makes me wonder how difficult it was to recruit patients and whether these patients are representative. Finally, results are not convincing and are not clearly stated. There is the implication of success, but no explicit commitment to followup with P3. The success is implied by bunches of complicated (to me) data that makes it look like the company is trying to sell something instead of report results. Good results have a way of being understandable and simple. Finally, the press release was from Miami <G>. My overall impression is that I would never make a bet on this company based on this study. I would give it a "3".
fredlhayes
----------------------
The release states that:
"Importantly, patients who generated anti-G17 antibodies (N = 79) lived significantly longer than patients who did not generate anti-G17 antibodies (N=24)."
I can take their samples and design an assay where all of their samples are positive for antibody. I can take their samples and design an assay where they are all negative, and yet a positive control would indicate that the assay was just peachy.
The company has existed for 22 years, but a Medline search turns up one irrelevant study from the company. What basis, therefore, do we have to judge the quality of their most fundamental observations? Can you find the trust in the biotech background of the founder and CEO?
Subset analysis of an uncontrolled phase II trial.
You can put my name on it.
(Rick)
--------------
I was intrigued by the Aphton challenge, but didn't have time to respond this morning.
If you want a number (0 = scum, 10 = very reputable), my score is 3. This means, to me, uninvestable, but above the level of penny-stock hypesters.
Here's the major flaw: "It is estimated that there are approximately 570,000 patients with gastric cancer in the US, Europe and Japan, alone. The prognosis for the overwhelming majority of these patients is very poor."
So it would remain, if Aphton's drug lives up to modest promise.
The sleaziest comment they made was the citing the five year survival rate. For a drug with (at best) 10 months median survival, and that in a combination, comparison to five year survival is not apt-:).
If I took Aphton's past into the rating, I'd lower them to a 2. They have been around forever, have not produced anything, and have been thoroughly sleazy. However, I am fastidious in reserving the 1's and 0's for the overtly fraudulent. Which abound in rising markets....
(John)
-----------
Score = 2.5
With the caveat that I know NOTHING about cancer trials, and have therefore avoided investing in oncology Co's:
This trial doesn't advance their cause any. I am puzzled by the design.
The PR really does seem to be an attempt to draw attention away from the relevant comparison (with the EORTC cohort) towards the responder/non-responder comparison, which is discussed in excruciating detail with meaningless P values.
What's the status of the NCI trial, which is a dose-finding trial?
Given the eagerness with which they pump out PR's on interim analyses, I would expect that if their Ph III trial supported approval for monotherapy in pancreatic cancer, we would have heard about it.
And perhaps that's my point - I would perhaps rate the PR in question a 3 or charitable 4, but looking at the track record of spin, their credibility is way down.
(DNA-Jock)
-----------------
Aphton analysis revised...Peter, if you don't mind, please replace this one with the previous one I sent....
I will attempt an analysis without rating since I have always found the midpoint #'s in the 1-10 scale quite subjective as in the 1-10 scale for pain evaluation...
Overall, I believe the study did show that it does contribute to the median survival time as an adjuvant in this population...I've never seen data reported with ITT consistently at every analysis as this PR did...
Red flags: No randomized cohort with which to compare the data...big one for me...
Secondly the disparity between best response and confirmed response was 51%/28% respectively...what does this mean? the PR did not address this.
<<* Median survival for patients treated with the combination was 8.9 months (ITT).>>>
It does show improved survival since historically this population survival is about 3 months.
<<<A sub-group analysis based on KPS, ranging from 70-100, showed that the survival benefit of G17 responders compared to G17 non-responders remained statistically significant (p less than 0.0001, log rank). At each KPS stratum, G17 responders lived significantly longer than G17 non-responders. This indicates that the survival benefit of G17 responders compared to non-responders was not solely a function of health status.>>>>
Not enough subgroup #'s available...
<<< At each KPS level among the G17 non-responders there were anti-DT antibody responders, indicating that many of the G17 non-responders were immunocompetent, providing a separate indication that survival benefit was not a function of health status.>>>>
This tells us that although one may be an antibody DT responder, it's not a direct, early predictor of who's tumor will stabilize, regress, or which patient will live longer....
"The data from this trial are quite encouraging for patients who suffer from gastric cancer," stated Jaffer Ajani, M.D., from the MD Anderson Cancer Center.>>>
I find it interesting that he's commenting on this since he is the PI of the largest gastric cancer P3 ongoing trial, he presented interims at ASCO 2003....see Ajani, et al, Abstract 999, ASCO 2003 and medscape.com
Ajani goes on to say "There is no approved therapy for gastric cancer and current standards of care add only a small clinical benefit at a cost of significant side-effects.>>>
Technically true re: 5-FU and cisplat, but doesn't mention his own trial 5FU+cisplatin+docetaxel which showed 48% year end survival...but in actual practice, clinicians at ASCO were so impressed with his data that they are now using it in practice although not officially approved...a bit deceptive on his part...
Ajani also says <<<The data shown in this trial marks a critical step in adding a safe, targeted biological approach to existing therapies to attack this typically fatal life-threatening disease.">>>>
Yes, true, promising...wonder if this is a prelude to Aphton progressing this trial to Phase 11b's or the old fashioned P3's...with G17DT added to Ajani's cocktail of 5FU, cisplatin +docetaxel...
<<<<Gastrin 17 and gly-gastrin 17 are believed to be central growth factors, or the initiating signals, for cell growth, cell proliferation and metastasis, or spread, in pancreatic, gastric (i.e. stomach), esophageal, colorectal and other gastrointestinal (GI) system cancers. This signaling cascade is triggered by gastrin binding to the large numbers of gastrin receptors which appear, de novo, in the great majority of cases, on tumor cell surfaces of the gastrointestinal system.>>>>>>>>
The above deceptive...gastrin has essential activities such as regulating gastric juices essential for proper digestion, and yes it is a necessary growth factor for endothelial cell growth for stomach lining which is where it "gets into trouble" by overprolfiterating...but clearly all the mechanisms causing gastric cancer have not been uncovered...in this PR, Gastrin is the culprit, leaving the reader with the sense that they've found the CAUSE...
<<<<Interrupting this process by immunizing the patient with Aphton's anti-gastrin immunogen is specifically targeted immunotherapy. This specificity of targeting only cancer cells occurs because gastrin is not normally secreted and gastrin receptors are not normally found on cells in the GI system, unless they are malignant, or on the path to malignancy (except for cells involved with normal acid secretion).>>>>>
Again deceptive and not completely true....G17DT is specific to GI system but not specific to malignant cells...it goes to non cancerous gastrin producing cells(those producing acids as noted in PR and those noncancerous producing growth factors for healthy endothelial stomach lining...it has not shown to be an autoimmune issue yet....
<<Gastrin also stimulates the secretion and expression of other important growth factors and receptors within and on the surfaces of the cancer cells involved in tumor growth. Hence, inhibiting gastrin inhibits all of the foregoing factors that contribute to tumor growth and spread, resulting in tumor cell death.>>>
Wonderful news...but rather lofty. Imo, if the above was true, this would be a cure and not just an life extension adjuvant...
<<There is only one large, randomized, phase III clinical trial that has been reported in the medical literature with cisplatin plus 5 FU for patients with advanced gastric cancer. In that trial the tumor response rate was 20%, as reported by the European Organization for Research and Treatment of Cancer (EORTC) which conducted the trial.>>>
While technically true...they don't mention Ajani's largest gastric cancer Phase III for 5FU,cisplatin+doxetaxel...link above...
Overall...could have been a much shorter PR with facts and less deceptive fluff interwoven with nuggets of truth...
Their history speaks for it's need to ingratiate itself to Wall Street and it's investors....
Thanks again Peter, look forward to other's opinions....
Zeta
-------------
My entry follows in next post.
Peter |
