Interleukin-6, C-Reactive Protein, and Recurrence After Stroke: A Time-Course Analysis of Individual-Participant Data

BACKGROUND: Inflammation promotes atherogenesis. Randomized controlled trials of anti-inflammatory therapies for prevention after stroke have not yet demonstrated clear benefit. IL-6 (interleukin-6) and hsCRP (high-sensitivity C-reactive protein) are independently associated with major adverse cardiovascular events poststroke and may guide patient selection in future randomized controlled trials. Optimal timing of hsCRP/IL-6 measurement poststroke is unknown, as early blood levels may be confounded by the inflammatory response to brain infarction. METHODS: Using individual-participant data from a systematic review, we performed a time-course analysis to investigate the association between hsCRP/IL-6 and recurrent events stratified by timing of sampling. The prespecified coprimary end points after sample measurement were: (1) recurrent major adverse cardiovascular events (first major coronary event, recurrent stroke, or vascular death) and (2) recurrent stroke (ischemic, hemorrhagic, or unspecified). The poststroke dynamics of IL-6/hsCRP were analyzed by plotting their median (interquartile interval) concentrations within each tenth of the sampling timeframe. Acute/postacute phases were defined for each biomarker according to the shape of this relationship. RESULTS: There were data for 9798 patients from 11 studies (19 891 person-years follow-up, 10 observational cohorts, and 1 randomized trial). Each marker was measured once. IL-6 was markedly elevated <24 hours poststroke compared with postacute levels (≥24 hours; 11.6 versus 3.02 pg/mL; P <0.001). HsCRP was elevated for 10 days. IL-6 was associated with recurrent major adverse cardiovascular events in the postacute phase (≥24 hours; risk ratio, 1.30 [CI, 1.19–1.41], per unit log e IL-6), but not in the acute phase (<24 hours; risk ratio, 1.10 [CI, 0.98–1.25]; P interaction =0.03). After adjustment for risk factors/medication, the association remained for postacute IL-6 when analyzed per log e unit (risk ratio, 1.16 [CI, 1.05–1.66]) and per quarter increase (risk ratio, 1.55 [CI, 1.19–2.02]; Q4 versus Q1), but not if measured acutely. Similar findings were observed for recurrent stroke. There was no evidence of time-dependent interaction with hsCRP. CONCLUSIONS: Timing of sample measurement after stroke modifies the association with recurrent major adverse cardiovascular events for IL-6 but not hsCRP. These data inform future randomized controlled trial designs incorporating biomarker-based selection of patients for anti-inflammatory therapies.