More Trouble For tPA at the BMJ

The BMJ recently stirred the pot again on a subject near and dear to our hearts – no, not exit block – but intravenous thrombolysis for acute stroke. For a therapy that’s been approved within 3 hours for nearly two decades, and was extended to 4.5 hours by guidelines in Europe and the United States since 2008, it ought be surprising there is still serious discussion regarding the validity of the results.

Just two years ago, the BMJ allowed members of the Australasian college to debate the risks and benefits for tPA in acute stroke. The accompanying unscientific poll of bmj.com readers came out in favor of skepticism, with 54% of the 612 respondents stating the risks outweighed the benefits. Now, in the context of the UK regulators re-examining the evidence in support of alteplase, the BMJ has published an examination specifically regarding the 3 to 4.5 hour time window.

In contrast to the individual-patient meta-analysis, the Cochrane Review, and many other interpretations of the applicable data, this work is unique in that is led by an outsider – neither a professionally entrenched stroke neurologist or emergency medicine physician. The author, Brian Alper, is a family physician whose professional affiliation is with a knowledge translation medical reference product. Rather than carrying the baggage of twenty years of contentious debate, colored by pharmaceutical sponsorship, his take is of one, colloquially, without a horse in the race.

Dr. Alper et al critques three different interpretations of the evidence regarding the extended time window: The American Heart Association guidelines, the 2014 Cochrane review, and the 2014 individual patient meta-analysis. The crux of each analysis focuses on ECASS III, IST-3, and the integration of such data sources. ECASS III, he points out, enrolled twice as many patients with prior stroke into the control arm – and excluding those with prior stroke from the final analysis diminishes the treatment effect to a 95% CI of 0.89 to 1.59 for mRS 0-1. The 3 to 4.5 hour cohort from IST-3, then, found no benefit to alteplase with a 95% CI of 0.50 to 1.07 for mRS 0-2. These two data sources are then combined to make up the bulk of the individual patient meta-analysis, and despite their seeming contradiction, their synthesis remains strongly positive in favor of thrombolysis.

The critique of the Cochrane review is couched in quite revealing language, noting “Cochrane reviews usually provide the most valid syntheses of evidence of effects of interventions”. As previously critiqued, these authors again take Wardlaw et al to task for discrepancies between their ultimate conclusions and the data presented beyond 3 hours. Dr. Wardlaw has already responded at bmj.com to begin what will hopefully be a provocative and enlightening discussion.

Dr. Alper et al make a call for data sharing such that all can examine the individual patient results, similar to the fascinating third-party re-analyses undertaken of the osletamivir data. They also call for additional research to further independently evaluate efficacy. Unfortunately, a paucity of new data is forthcoming. A cursory search of stroke trials indicates a 200 patient trial from China enrolling in the 3 to 4.5 hour time window, with little additional information. Other potentially forthcoming data includes TESPI, an open-label trial from Italy whose recruitment ended last year. The remaining trials, rather than further enlightening us regarding safety under the existing license, aim to expand eligibility by time window or elimination of other exclusions.

Will this most recent dissent produce any meaningful result? It is too early to tell – but, certainly, for our patients’ sake, I hope so.

(Visited 229 times, 1 visits today)