A few days ago my friend, who goes by the handle Persevering on the TDI Forum, posted an evaluation of the PALS who have been self-reporting their experience with the Neuraltus Phase 2 trial of NP001 (which I have discussed here in the past). While the data looks exciting, I must caution that this isn't official data and makes some assumptions which, if wrong, could totally invalidate the evaluation. However, I have enough faith in Persevering to bring this to your attention. For those who don't have Patients Like Me accounts, I reproduce the graphs here.
I have analyzed the NP001 phase 2 data from PLM to date.
I am aware of 37 PLM members who have participated in the trial and 34 who have entered sufficient data to track progress. This is roughly 1/3 of the listed full trial enrollment, and offers a great sample to predict actual trial results. The trial consists of 6 infusion cycles over 21 weeks (147 days), followed by 4 follow-up visits adding 16 more weeks. 13 have completed dosing. 11 will complete within 4 weeks, and all within 8 weeks.
In my thinking, since each infusion cycle is 4 weeks total, the first non-dosing follow up at week 25 (175 days) ends the dosing phase, and I have chosen to review data to day 175 in terms of FRS change. The variable of interest is FRS slope, and is very commonly used to evalauate efficacy for ALS clinical trials. For example, it was used in the recent report of Dexpramipexole phase 2 data, and was used for the official Lithium clinical trials in the US.
In my opinion it is possible to review reports of side effects for commonality to approximate those getting drug (either dose) versus placebo. I do not believe it is 100% accurate, and I also expect those getting a higher dose to have more or greater side effects, but it is not entirely possible to segregate the drug groups (high vs. low). Some may not be prone to side effects on drug and others could experience "nocebo" effects, so again this is not a true substitute for unblinding, which should occur later and allow a timely re-analysis of PLM data.
With that assumption/disclaimer, the summary statistics are:
(Green is improved FRS score, yellow is stable FRS down to -0.49/mo, and red is loss of FRS)
With side effects:
- n = 20
- Mean ALSFRS-r slope: 0.00
- Standard Deviation ALSFRS-r slope: 1.24
Without side effects:
- n = 14
- Mean ALSFRS-r slope: -1.01
- Standard Deviation ALSFRS-r slope: 0.68
Comparison: 100.4% mean improvement
The most exciting outcome to me is that the difference in progression rate is very statistically significant (p<0.05), even with only 34 data points! This would be unprecedented for a phase 2 ALS clinical trial. For example, the recent exciting report regarding Dexpramipexole, with a 31% mean improvement versus placebo in the first 12 weeks had a p-value ~ 0.20, a value 4 times higher than acceptable to conclude drug is better than placebo, by convention (based on a sample size of 53).
There is rumor of an official NP001 trial efficacy review soon, based on all data available on January 24, 2012. Let's hope for statistical significance there as well, and potentially FDA accelerated approval, as occurs for cancer and HIV drugs.
Thanks to Persevering for the work and for the guest-blog!