Discussion
In this register-based study based on a primary health care population, we found that initiation of simvastatin was associated with a minor, but statistically significant increase in INR of 0.32 (95%CI: 0.25-0.38, p<0.001), peaking approximately 4 weeks after initiation. A similar increase in INR was observed with atorvastatin and rosuvastatin initiation, although failing to reach statistical significance.
The main strength of this study is the use of data obtained from daily routine work in primary health care with limited risk of selection bias. One limitation of this study is the assessment of drug use by prescription data alone. The level of adherence to statin therapy cannot be determined and neither can the precise date of initiation of statin treatment. Early discontinuation of statins due to side effects could minimize the effect on the change in INR.23Furthermore, we had no available information about other factors that might impact the INR level e.g. lifestyle factors, herbal medications as well as relevant diagnoses diagnosed solely in primary health care, however, due to the within-person comparison, most of such factors can reasonably be assumed constant. Finally, we do not report clinical outcomes. However, increases in INR are well known to increase the risk of severe bleeding.1,26,27 This is supported by a study that found initiation of statins to increase the risk of gastrointestinal bleeding in chronic warfarin users.28In our study the proportion of patients with INR > 5 increased from 1.3 to 3.2% after statin treatment was initiated. Importantly, however, when scrutinizing the INR changes at the level of the individual, this increase reflected a small overall increase and not a pronounced increase in a subset of patients.
Our findings from primary health care in the Copenhagen area support the recent findings from specialized anticoagulation clinics in Sweden presented by Andersson et al, which found an increase in INR from 2.43 to 2.58 in 5637 patients on warfarin treatment initiating simvastatin, also peaking about 4 weeks after simvastatin initiation.10 A more pronounced increase in INR was found in a small study, where INR increased from 2.5 at baseline to 3.2 after simvastatin initiation in 29 patients in stable warfarin treatment.7
The latency in the INR increase, peaking after four weeks of concomitant treatment is surprising. Of note, a similar course was seen in the study by Andersson et al.10 This does not correspond to statins directly inhibiting the CYP enzymes responsible for the metabolism of warfarin, as this would lead to a faster onset of the INR increase, as e.g. seen for azole antinfungals.29 To our knowledge, no alternative mechanisms have been proposed. As such, additional work identifying the mechanism through which statin use potentiates the effect of warfarin is warranted.
In conclusion, initiation of simvastatin, atorvastatin, and rosuvastatin led to a minor increase in INR in patients treated with warfarin, peaking about 4 weeks after statin initiation. The magnitude of the change in INR is for most patients likely to be of limited clinical relevance. Individual risk stratification including age, use of medication and other diseases should be applied, when deciding if increased INR monitoring should be performed during the initiation of statin treatment.