Statin Withdrawal May Lead to Adverse Outcomes in Acute Stroke

Abstract & Commentary

By Alan Z. Segal, MD, Associate Professor, Department of Neurology, Weill-Cornell Medical College, Attending Neurologist, NewYork- Presbyterian Hospital. Dr. Segal is on the speaker's bureau for Borhringer-Ingelheim.

Synopsis: Withdrawal of statin therapy in acute ischemic stroke may lead to increases in death and disability.

Sources: Colivicchi F, et al. Neurology. 2007;69:904-910.

Statin therapy has a well-recognized role in the primary and secondary prevention of stroke. Statins may also have a neuroprotective effect in the setting of acute stroke. This has been well documented in animal models and high dose acute statin therapy is currently under investigation in human subjects. Pretreatment with statins is also likely to have benefit. Discontinuation of statins in the acute setting may precipitate vascular dysfunction and exacerbate ischemic events including both stroke and myocardial infarction.

In a single center study in Spain, Blanco and colleagues studied 215 patients with acute ischemic stroke, 89 of whom were previously taking statin medications. These patients were randomly assigned to have statin therapy withheld for three days (statin withdrawal group) or to be treated with 20 mg atorvastatin either orally or via nasogastric tube (statin treated group). All patients were treated with statins starting on day 4, including the 126 remaining patients who had not previously been treated with statin therapy (reference group).

At 3 months, 60% of patients in the statin withdrawal group met the primary outcome variable of death or dependence compared with 39% in the statin treated group. The adjusted odds ratio favoring a poor outcome among statin withdrawal patients was 2.39, increasing to 4.66 (1.46 - 14.91) after adjustment for age and stroke severity. Early neurological deterioration, defined as an increase of ≥ 4 points on the NIHSS, was observed in 65% of statin withdrawal patients compared with 21% of statin treated patients. Infarct volume was also greater in the statin withdrawal group, with a mean increase of 37 mL compared with treated patients. In post-hoc analyses, statin withdrawal patients also fared more poorly than patients in the reference group with regard to all endpoints-death and dependency, early neurological deterioration and infarct volume.

In a related study, Colivicchi and colleagues studied 631 patients with an ischemic stroke and followed them for one year to assess their adherence to statin therapy. Among 409 patients who received atorvastatin therapy, 163 discontinued this medication and among 222 patients who received simvastatin, 83 had stopped taking this at one year. Among the 631 patients, 116 (18%) died during one-year follow up. After adjusting for confounding variables, including stroke severity, discontinuation of statin was an independent predictor of mortality with a hazard ratio of 2.78. This effect was more pronounced with early discontinuation, leveling off in the 9-12 month interval. Discontinuation of anti-platelet therapy was also an independent predictor of death, though with a less profound effect (hazard ratio of 1.81).

Statin therapy was discontinued due to side effects (most commonly dyspepsia) in a minority of patients (29%) and was unexplained in the remaining 71%. Patients who discontinued statins were older and more commonly female. Statins were more likely to be continued among patients who were diabetic or who had a history of previous stroke.


As Blanco indicates, animal as well as human data strongly suggest that withdrawal of statin therapy in an acute stroke patient may impair vascular function and trigger a dangerous inflammatory and prothrombotic state. This raises a major red flag in our treatment of acutely hospitalized stroke patients. Patients on previous statin therapy who have the medication discontinued face a 4.7 fold increase in their risk of death or dependency due to their stroke. This effect is even more pronounced than among patients not previously receiving statin medications.

These data raise important practical implications regarding the "nuts and bolts" of emergency room and immediate hospital care for acute stroke patients. Patients with severe strokes, who cannot take oral medications due to dysphagia, must receive these via nasogastric tube. Such patients are commonly made NPO, with feeding and oral medication administration delayed until their swallow status can be clarified. These issues are particularly germane to patients receiving thrombolysis. Among patients receiving intravenous tPA, our protocol mandates placement of a Foley catheter prior to thrombolysis since such an invasive procedure cannot be performed once tPA has been administered. The same would likely apply to a nasogastric tube.

The data of Colivicchi are more difficult to understand. Discontinuation of statin therapy was highly associated with post-stroke mortality, but it is not clear if this was a cause, or more likely merely an effect of practice patterns among patients with more devastating strokes. It is testament to the inconclusive nature of this study that the justification for cessation of statin therapy was unexplained in over 70% of patients. In addition, while over 80% of the deaths were attributed to cardiovascular causes, data such as this, gleaned from death certificates, is unlikely to reflect the true etiology of their demise. Notably, recurrent stroke was not documented as the cause of death among any of the patients from whom statins were withdrawn.