Expediting MRI-based proof of concept stroke trials using an earlier primary endpoint
Purpose: Prior to pivotal phase III trials of acute stroke therapies, proof of concept MRI trials are used to assess likelihood of success. Surrogate MRI stroke trials could be expedited if subacute DWI lesion volume replaced late T2-weighted lesion volume as the primary endpoint.
Methods: EPITHET patients - acute ischemic stroke presenting within 3-6 hours of onset, were randomised to t-PA or placebo. Prior to unblinding, correlation between subacute DWI lesion volume (3 to 5 days after acute stroke) and outcome T2 lesion volume (90 days after acute stroke) was assessed. The correlation between outcome T2 lesion volume and NIHSS score on day 90 was compared to the correlation between subacute DWI lesion volume and NIHSS score on day 90.
Results: 100 patients received baseline scans and blinded drug. Of these, 29 patients did not have either outcome or the subacute scans, including 18 deaths, 6 patients who had either poor quality images or who were not rescanned for other reasons, 4 patients have not yet completed the study, one patient withdrew consent. Outcome median NIHSS score of the remaining 71 patients was 4 (0 to 21). Median subacute DWI lesion volume was 48.9 ml (0.0 to 384.4), median 90 day T2-weighted lesion volume was 28.6 ml (0.5 to 252.8). Subacute DWI lesion volume and 90 day T2-weighted lesion volume were strongly correlated (spearman’s rho:0.92,p<0.001). The correlation coefficient of subacute DWI lesion volume and NIHSS score was 0.79 (95%CI [0.686,0.866]) and that between late T2 lesion volume and NIHSS score was 0.74 (95%CI [0.615,0.831]).
Conclusions: The close correlation of subacute DWI and late T2 lesion volumes as well as the similar correlations of subacute and late lesion volume with clinical outcome suggest that subacute imaging at day 3-5 is an appropriate outcome time for proof of principle MRI-based stroke trials.