RE: Webinar 29th October23 Oct 2020 09:40
I wouldn't rush to overstate the results presented by jcyte. Based on the data in it's most basic form, they found a mean letter improvement of +2.8 in the sham group, +2.96 in the 3 million cell group, and +7.4 in the 6 million cell group.
They showed good data for a small group recieving the highest dose of 6 million cells when they went looking for it:
'In a post hoc analysis of this target population (~50% of the per-protocol population), an early and significant improvement in vision was seen in the higher-dose group, with average gain of 16 letters at month 12 compared with 2 letters in the control group.'
In the control group and 3 million dose group , when they applied the same 'target group' criteria, the results were +1.85 letters and -0.15 letter respectively. Now why would 6 million cell dose be effective for the 'target group', but a 3 million dose result in a loss of letters? To me it doesn't sound very scientific, and I think for phase 3 they will be changing their primary endpoint, as they will likely know that what they've gone looking for in the data is unlikely to reliably repeated in a larger population.
For comparision, Rene has presented data out to th e quivalent 12 month period (and up to 18 months for a couple of patients), with a mean change of +12 letters presented at 12 months. I'm sure if Rene were to expand the study and the cherry pick their own 'target population' in post hoc analysis, then they would be able to present some incredible data. However, I would prefer to just see data which shows the whole cohort tested, and the mean average improvement across the board, without having to massage the data in a way that will probably cause you to have to change your primary endpoint in future trials. Making data fit the results you want to see isn't particularly scientific in my opinion.