BLEEDING RISK WITH PRADAXA AND STATIN USE

A new analysis suggests older atrial-fibrillation patients on dabigatran (Pradaxa, Boehringer Ingelheim) who take lovastatin or simvastatin have a more than 40% higher relative risk of major hemorrhage compared with those given other commonly prescribed statins^[1].

Use of lovastatin and simvastatin was not associated with a higher risk of stroke or transient ischemic attack (TIA) relative to atorvastatin, pravastatin, fluvastatin, or rosuvastatin (adjusted odds ratio [OR] 1.33; 95% CI 0.88–2.01), but risk was increased for major hemorrhage (adjusted OR 1.46; 95% CI 1.17–1.82).

Principal investigator and pharmacist Dr Tony Antoniou (St Michael's Hospital, Toronto, ON) told heartwire from Medscape that the absolute risk of these events is quite low, but the finding is still important because a lot of patients taking dabigatran are also on statins.

"It highlights a previously unrecognized drug interaction and indicates you could use other statins so patients could still derive the benefits of statins without getting the excess risk that might be imparted by lovastatin and simvastatin," he said.

The article, published November 21, 2016 in CMAJ, suggests that, unlike the comparator statins, lovastatin and simvastatin are potent inhibitors of intestinal P-glycoprotein, which increases systemic dabigatran exposure.

Commenting to heartwire on the controversial study, Dr Thomas Deering (Piedmont Heart Institute, Atlanta, GA) said there are mechanistic reasons to think the finding might have validity, but "most important, we should not overreact and should figure out how to approach patients in a situation when we have one publication and not great, definitive data."

He suggested one possibility may be to have patients extend the time between administration of dabigatran and lovastatin or simvastatin. Indeed, the investigators write that "the phenomenon of competitive P-glycoprotein inhibition and spacing of medication should attenuate any observed effect in our analysis."

Deering added, "We've had both these statins and dabigatran for a long time, so it's a little bit interesting that pretty far down the road we're identifying a mechanism, and if true—and I don't think we can say that yet—that mechanism could significantly influence how we wind up treating patients."

Dr Isabelle Van Gelder (University Medical Center Groningen, the Netherlands) told heartwire , "The mechanism makes sense because it is a well-known interaction, but it is a little strange that these data became available 7 years after the first large clinical trial and that RE-LY investigators did not find such an interaction."

She noted that in vitro data support the mechanistic explanation and a similar interaction with risk of major hemorrhage has been observed with other drugs that inhibit P-glycoprotein.

Van Gelder said the findings clearly warrant well-defined in vivo and human studies and cited several limitations of the current study, including a lack of information on statin adherence, very small numbers of patients on lovastatin and simvastatin, and importantly, the definition of hemorrhage as any hemorrhagic event resulting in hospital admission or visit to an emergency department.

"An emergency-department visit doesn't mean it was a major bleed—not our typical definition of a major bleed at least," she said

The findings were drawn from two studies that used administrative health databases. The first study comprised 397 patients diagnosed with an ischemic stroke or TIA and 1588 matched controls. The second study comprised 1117 patients diagnosed with major hemorrhage and 4465 matched controls. All were Ontario residents at least 66 years of age who started dabigatran between May 1, 2012 and March 31, 2014 and had received a single statin within 60 days preceding their index date.

Among the 1985 participants diagnosed with stroke or TIA, 986 were on atorvastatin, six on fluvastatin, and 12 on lovastatin. Among the 5582 participants diagnosed with major hemorrhage, the corresponding numbers were 2778, 26, and 27, respectively.

Dr Paul D Thompson (Hartford Hospital, CT) told heartwire that the study and its hypothesis are interesting, but "a major flaw is that [neither] the researchers nor anyone else has actually measured dabigatran levels while on different statins and compared levels."

Thompson also observed that lovastatin and simvastatin are older drugs approved before atorvastatin and the other comparators; consequently, patients on these older statins may have been on them longer because they were sicker. Duration of statin use was controlled for only in the stroke analysis, and there were a fair number of other differences between cases and controls, including higher digoxin and nitrate use among cases.

"Both statins and anticoagulation for strokes save lives and people should not jump to conclusions and stop their drugs" based on a single observational analysis, he said.

Antoniou said residual confounding is possible despite controlling for many variables and that researchers didn't adjust for duration of statin use in the hemorrhage analysis because it was so similar between cases and controls (median 4.6 vs 4.4 years; P=0.05).

He said the biggest strength of the study is that it is population based and limitations are lack of lab data, data on markers of kidney function, and reliable data on nonprescription drugs that can influence bleeding. Finally, the findings may not apply to younger patients, who may have fewer risk factors for stroke or hemorrhage.