Kevin S. Attenhofer, MD
Statin medications and cholesterol management remain topics of debate over ten years after the publishing of the SPARCL (Stroke Prevention by Aggressive Reduction in Cholesterol Levels) trial. With many questions remaining, the authors investigated the effects of statin intensity and adherence on the long-term prognoses after acute ischemic strokes in Korea.
In this paper, Kim et al. present a retrospective cohort study based on nationwide Korean population-based health insurance data. They used diagnosis codes to identify patients with ischemic stroke and then collected prescription records (type, dosage, duration, and date) of statins. In order to determine adherence, they calculated the proportion of days covered (PDC) by any statin prescription for a period of 1 year after acute ischemic stroke. Poor adherence was PDC < 40%, intermediate adherence was PDC 40-80%, and good adherence was PDC > 80%. Their primary outcome was composite of recurrent stroke, myocardial infarction (MI) and all-cause death. Again, diagnosis codes were used to obtain this information. Recurrent stroke was defined as re-hospitalization with a primary diagnosis of I60-63. MI was defined as primary diagnosis code of I21.
After exclusions, this study included 8001 patients with a mean follow-up of 4.69±2.72 years. Patients were stratified by adherence as above, as well as statin intensity. Populations were fairly well matched with regards to baseline characteristics with the exception of the “No Statin” group, which appeared to have fewer overall comorbidities. The authors noted that the proportion of patients treated with a statin after ischemic stroke consistently increased from ~20% in 2002 to > 60% in 2012 and that adherence improved with strokes occurring in the latter years of their inclusion (notably after 2006). Also, good statin adherence was associated with good antithrombotic adherence and higher household income, among other confounders.
In this study, patients with good adherence to statin had a significantly longer event-free survival (Figure A). A sub-group analysis on this population demonstrated that event-free survival was also increased with increasing intensity of statin (Figure B).
Interpretation of these results is somewhat limited. The source data (diagnostic codes and prescription records) is not as accurate as traditional methods of clinical data collection, despite validation. There were also multiple confounding factors, including improved risk factor management over time, which put more emphasis on high intensity statins in later years of the study, as well as variable associated antithrombotic adherence (better antithrombotic adherence was seen in the good statin adherence group).
The subgroup analysis of the good adherence group is perhaps the most interesting aspect of this paper. A dose-response relationship between statin intensity and event-free survival would be an interesting adjunct to available data, which could change clinical practice to focus on intensity rather than LDL level. This would be a fascinating topic for future randomized clinical trials.