Lesion Water Content on CT Predicts 12-Month Outcome and Modifies Thrombectomy Effect in Large-Core Stroke-A Secondary Analysis of the TENSION Trial.

In large-core ischemic stroke, outcomes after thrombectomy remain heterogeneous, and imaging markers that capture treatment-modifiable injury beyond lesion extent are needed. We investigated whether lesion water content on noncontrast CT, quantified as net water uptake (NWU), predicts 12-month functional outcome and modifies the effect of endovascular thrombectomy (EVT). This secondary analysis of the randomized TENSION trial included 198 patients with admission CT suitable for NWU quantification and available 12-month modified Rankin Scale (mRS) scores; 168 had paired 24-h CT for follow-up NWU (NWUFU) assessment. The primary outcome was ordinal 12-month mRS. Multivariable ordinal and logistic regression tested associations and the NWU × EVT interaction, with discrimination assessed by area under the receiver operating characteristic curve (AUC). Median baseline NWU was 16.4% and NWUFU 28.9%. At 12 months, 45 patients (23%) had mRS 0-3 and 111 (56%) had died. Higher baseline NWU independently predicted worse 12-month mRS (adjusted common odds ratio [acOR], 1.10 per percentage-point; 95% CI, 1.05-1.16; P < 0.001) and modified the EVT effect (interaction acOR, 1.17; 95% CI, 1.06-1.30; P = 0.002), with greater discrimination than ASPECTS. NWUFU was the strongest single imaging predictor (acOR, 1.10; 95% CI, 1.06-1.14; P < 0.001; AUC, 0.78 for mRS 0-3) and superseded baseline NWU when co-modeled. Among 97 90-day survivors, higher NWUFU independently predicted late mortality (adjusted odds ratio, 1.09; 95% CI, 1.02-1.18; P = 0.019). Lesion water content may therefore provide a clinically accessible CT biomarker linking treatment-modifiable injury at baseline with realized post-treatment injury and long-term outcome.