On January 3, 2013, EMPOWER, the Phase III clinical trial of Dexpramipexole was determined by Biogen Idec to have failed because no ascertainable differences could be identified between the medication and placebo subjects administered in the trial. Until this time, Dexpramipexole was considered the most promising drug to ever emerge for the ALS patient community.
Yet as stated by Biogen Idec,
“The trial did not meet its primary endpoint, a joint rank analysis of function and survival, and no efficacy was seen in the individual components of function or survival. The trial also failed to show efficacy in its key secondary endpoints. Additional analyses of multiple subpopulations failed to demonstrate any efficacy among these groups. Based on these results, Biogen Idec will discontinue development of Dexpramipexole in ALS.” (3)
The failure of the Dexpramipexole Phase III trial, which showed notable promise and benefit in both the Phase I and Phase II clinical trials is the most recent in a long line of failed Phase III trials for the ALS patient community. (4)
But is it the failure of the drug or the failure of the trial construct that is at issue? ALS WORLDWIDE believes it is the trial construct that needs to be changed for the following reasons:
1. ALS is, in all likelihood, not a single disease but rather a generic title given to a heterogeneous group of phenotypes or similar conditions. Cancer is a title given to many different diseases. In fact, there is not one form of breast cancer, but rather more than six different types of breast cancer, each requiring a different form of treatment based on the molecular signature of each tumor. (5)
2. ALS progresses through several, even many, disparate paths. One patient may have drop foot, another dysarthria and dysphagia, and yet another loss of upper peripheral limb strength and mass. Eventually, while all may end up in the same place, an ongoing quality of life can be best brought about through proper identification and treatment of each genetic or other molecular subgroup, as has been demonstrated in the treatment of cancer.
3. The selection of clinical trial subjects who are symptomatic for less than 24 months, as is now standard in ALS clinical trials, produces a group of patients, both those in medication and placebo groups, who will live longer than the duration of the trial. Therefore, survivability as a criterion is without merit unless trials are carried out for years, which is currently untenable.
4. Subgroups within this and other ALS clinical trials may have experienced benefit, and in some cases, significant benefit. Because the percentage of patients within the trials that experienced such benefit may be only 20 to 40 percent instead of a more robust 60 to 80 percent, entire trials may have been deemed as failures.
5. Until now, little or no attention could be given toward the identification of common genetic and other molecular identifiers for subgroups that may experience trial benefit.
6. It should be acknowledged that clinical trial benefit for ALS patients should include three evaluation subsets: a) bona fide improvement of symptoms; b) provable maintenance of symptoms; and c) reduced rate of decline of symptoms. Each of these subsets has independent value in determining efficacy.
7. With more than 100 genetic permutations now known to affect both familial and sporadic ALS, pre-human testing of the SOD1 mouse continues to be an inelegant model for subsequent human treatment, frequently leading to false positives. These included minocycline, creatine, Celebrex, Neurontin, Lithium, arimoclomol, Myotrophin, pioglitazone, Copaxone, Olesoxime, xaliproden, Indinavir and ceftriaxone. With the current awareness of a less clear distinction between sporadic and familial ALS and the frequency with which positive SOD1 mouse trials have led to human clinical trial failures, it is also possible that false negatives may have existed.
According to James Bennett, MD, PhD, and Bemiss Professor of Neurology at Virginia Commonwealth University, clinical trials for ALS/MND could be improved significantly in the following ways:
- Because sporadically occurring ALS is likely a heterogeneous disorder, grouping patients according to genetics and/or molecular attributes would conceivably or likely improve trial results and knowledge about patient subsets. The question as to how that might be done successfully, as is currently the situation in many cancer studies, is yet to be determined. We are working on characterizing this molecular heterogeneity, as are others.
- We must use historical controls from the large Phase III clinical trials, instead of requiring a placebo-controlled trial. This would parallel the now-successful use of historical controls in cancer trials.
- Open-label small studies provide early signals of potential efficacy that can justify carrying out larger studies.
- FDA consideration and approval is needed to allow two experimental drugs in a single subject, as long as there are trial limbs for the experimental drugs given individually.
- The FDA should provide preliminary approval for promising therapies, allowing early access, while requiring testing in larger studies and with genetic "footprint" characterization.