Insulin glargine and malignancy: an unwarranted alarm.

www.thelancet.com Vol 374 August 15, 2009 511 and for the rest by perceived increased bleeding risk. The last was based on inability to comply with monitoring by international normalised ratio, predisposition to falling or head trauma, persistent blood pressure above 160/100 mm Hg, previous serious bleeding on warfarin, severe alcohol misuse for more than 2 years, peptic ulcer disease, thrombocytopenia, or the need for chronic use of a non-steroidal anti-infl ammatory drug. Clearly, these criteria were rather loose, being put forward by either the physician or the patient. Therefore double antiplatelet therapy cannot been seen as an alternative to warfarin for patients with atrial fi brillation in general. Are the patients in ACTIVE-A very diff erent from the patients in ACTIVE-W? The strong risk factors for stroke, such as age and CHADS2 score, a clinical predictor for stroke in atrial fi brillation, were almost identical (table). As expected, the stroke rate in patients on double antiplatelet therapy was also similar in the double antiplatelet therapy groups in both ACTIVE-A and ACTIVE-W, which strongly suggests that the patients also had the same baseline bleeding risk. So it seems that the populations of patients in both trials were similar. The lowest stroke rate per year was seen in the warfarin group in ACTIVE-W, with a similar major bleeding rate as double antiplatelet therapy in both ACTIVE-A and ACTIVE-W. Although ACTIVE-A underscores the role of platelets in stroke in patients with atrial fi brillation, double antiplatelet therapy for stroke prevention should be given only to patients who are defi nitely ineligible for warfarin. This group could include patients who refuse to undergo monitoring or those mentally not able to take the various doses of warfarin mandated by the monitoring. Perceived unacceptably high risk of bleeding itself cannot make patients ineligible for warfarin, as clearly shown in the published ACTIVE trials, because the bleeding rate with double antiplatelet therapy in both studies were very similar to the bleeding rate with warfarin. Therefore warfarin should remain the cornerstone of stroke prevention in atrial fi brillation.