Phenylpropanolamine (norephedrine)

Stroke

Anecdotal reports of stroke in individuals who have taken phenylpropanolamine have appeared from time to time.

• A 37-year-old woman took an over-the-counter formulation containing phenylpropanolamine 100 mg [ ]. About 90 minutes later she developed very severe bilateral headache resistant to analgesics. Her blood pressure was 180/100 mmHg but fell rapidly to 110/70 mmHg. A CT scan showed multiple small frontal and parietal hemorrhages, and angiography showed extensive segmental vasospasm. She was treated with nimodipine and prednisolone, followed by verapamil. She made an uneventful recovery, and there was angiographic resolution of the vascular lesions.

• Stroke occurred in an 8-year-old boy on chronic peritoneal dialysis after he took phenylpropanolamine [ ]. He developed occipital infarcts and was found to have extremely high concentrations of phenylpropanolamine in his blood and dialysis fluid. Although the voluntary recall was in effect, the family already had a bottle of phenylpropanolamine at home.

• A 57-year-old woman had a sudden thunderclap headache after exposure to phenylpropanolamine and subsequently developed posterior reversible encephalopathy syndrome complicated by occipital intracranial hemorrhage with cerebral vasoconstriction [ ].

The authors of the first report reviewed a number of other case reports of phenylpropanolamine-induced cerebral vasospasm associated with hemorrhages and drew a parallel with similar effects of amfetamines.

In November 2000 the FDA asked manufacturers to stop marketing products containing phenylpropanolamine and subsequently banned such products, whether as nasal decongestants or appetite suppressants. This was prompted by the publication by the Hemorrhagic Stroke Project of a 5-year case-control study of 702 patients who had had hemorrhagic strokes and 1376 control subjects aged 18–49 years [ ]. The odds ratio for hemorrhagic stroke in those who had taken phenylpropanolamine was 1.49 (95% CI = 0.84, 2.64). In subgroup analyses by type of formulation the OR was 1.23 (0.68, 2.24) for cough/cold remedies and 16 (1.38, 184) for appetite suppressants. There was a trend toward an increased risk of stroke with cough and cold remedies in women (OR = 3.1; CI = 0.9, 11). The risk of stroke associated with phenylpropanolamine was significant only in women (adjusted OR = 17; 95% CI = 1.5, 182) and for men and women combined, the adjusted odds ratio was 16 (CI = 1.4, 184). However, only women reported using appetite suppressants containing phenylpropanolamine, and the ABBA study [ ], described in detail below, suggested that the more evident risk in women cannot be attributed to a difference in the exposure rate between the sexes.

In a reappraisal of these results it was suggested that these results could have occurred by chance, selection and recall bias, and confounding factors, such as hypertension, cigarette smoking, and heavy alcohol consumption [ ]. Among the potential conflicts of interest declared by the authors of this review, one was a defence lawyer for Delaco in the litigation over phenylpropanolamine and the other a member of various data and safety monitoring boards for several drugs companies and was an expert witness in the litigation over phenylpropanolamine. Subsequently, others criticized this conclusion [ ], claiming that the article “was intended only for litigation advantage to phenylpropanolamine manufacturers in pending court cases”, and pointing to various suggested defects in the reappraisal and to other evidence, at that time published in abstract form, but later published in full, showing that phenylpropanolamine contained in cold remedies increases the risk of hemorrhagic stroke, particularly in women [ ]. Both of the authors of this letter had acted on behalf of the plaintiffs in litigation, one as an expert witness, the other as a lawyer. Another letter raised the question of “mercenary epidemiology” [ ]. These letters elicited a robust response [ ].

In a consecutive stroke registry since 1988, 22 patients (10 men and 12 women) had strokes associated with over-the-counter sympathomimetic drugs [ ]. There was intracerebral hemorrhage in 17, subarachnoid hemorrhage in four, and ischemic stroke in one. Stroke was associated with the use of phenylpropanolamine (75–675 mg) in 16 patients.

Korean authors analysed the risk of hemorrhagic stroke in individuals taking phenylpropanolamine as a cold remedy [ ]. They included 940 patients with hemorrhagic stroke that occurred between October 2002 and March 2004 and 1880 matched controls. Of those who had had a stroke, 16 (1.7%) had been exposed to phenylpropanolamine in the preceding 2 weeks, compared with 14 (0.74%) of the controls. There were no significant differences in the men (50% of each group), but in the women there was an increased risk of stroke associated with exposure to phenylpropanolamine, and this was graded according to dose, more recent date of exposure, and longer duration. The overall odds ratios were about 2 for exposure versus no exposure; over 5 for an exposure within 3 days; over 3 for an exposure of greater than 3 days; and over 2 for a dose of greater than 75 mg/day. The risk was much greater in women.