Letters

Alteplase for stroke

Uncertainty remains about efficacy

Patients and doctors are being misled by promotional pressures

American Heart Association explains how guidelines were formulated

Financial information is needed to ensure objectivity

Why were these authors of the commentaries chosen?

Author's reply

Authors of commentary reply

Uncertainty remains about efficacy

EDITORThe article by Lenzer and the associated commentary by Saver et al raise many serious issues, among which is the residual state of uncertainty concerning the efficacy of alteplase (tPA) in acute ischaemic stroke.¹ Confronted by opposing interpretations of the aggregate data published to date and the now known baseline imbalance in the severity of stroke in the National Institute of Neurological Diseases and Stroke (NINDS) trial, doctors are presented with a conundrum: what action do the data support?

The reported unwillingness of the investigators and sponsor of the National Institute of Neurological Diseases and Stroke trial to provide data for additional analysis is disturbing. Emanuel et al have described seven requirements for the ethical conduct of clinical research, among which is social and scientific value.² Social value presupposes the public dissemination of research results. I have formulated a standard for the scientific and ethical review of trials that elaborates on this requirement.³

Implicit in this requirement is the necessity for the public dissemination of the complete dataset acquired during a clinical trial. This allows interested investigators to apply recognised analytic techniques in an attempt to resolve (or diminish) residual uncertainty concerning the clinical implications of the trial's results. This ethical requirement has not yet been met for the National Institute of Neurological Diseases and Stroke trial.

Howard Mann, program associate. 
Division of Medical Ethics, University of Utah School of Medicine, Salt Lake City, UT 84132, USA howardm{at}xmission.com

Competing interests: None declared.

Patients and doctors are being misled by promotional pressures

EDITORAlteplase (tPA) is not proved for the treatment of stroke, and payment from the drug's manufacturers to research doctors degrades the premise of unbiased skill.

In their commentary Saver et al disclose ties to 81 for profit companies linked to treatments for stroke.¹ Sponsorship of this magnitude does not "channel the self interest of profit making companies to improving stroke care": it purchases spokesmen for manufacturers seeking to add credibility to their wares. Although sponsored doctors can perform impartial research, it is impossible to determine whether their research is truly unbiased and is much simpler to believe consultants without such ties.

We are troubled by the claim that alteplase has been proved in pooled data from six trials. Of these, only the National Institute of Neurological Diseases and Stroke (NINDS) trial rigorously measured the efficacy of alteplase within a predetermined three hours of stroke onset. The remainder of the trials were post hoc analyses that routinely ignored negative results and showed increased mortality in the alteplase treatment groups. Such data manipulation is contrary to meta-analysis methodology and confounds the truth.²

It is important to compare the benefit and harm in the National Institute of Neurological Diseases and Stroke trial. For every nine patients treated, one benefited. For every 16 treated, one was harmed by cranial haemorrhage. For every 34 treated, one died as a result of such haemorrhage. Thus for every four patients who benefited, one died. These numbers do not support the statement that treatment with alteplase is highly efficacious.

Given the animosity of this debate, we wonder if the future of medicine will rely on investigators with such ties to for profit companies and if the results of the National Institute of Neurological Diseases and Stroke trial will ever be confirmed by independent reanalysis or randomised trial. (Its data remain unavailable despite claims that its results are so compelling that further randomised controlled trials are unethical.)

Until this trial's conclusions are verified, and given the drug's potential for catastrophic harm, we strongly disagree with the premise that the public should be targeted in campaigns to increase the number of stroke patients treated with alteplase. In a treatment window of only three hours, doctors unfamiliar with the specific risks of alteplase cannot properly do all the assessments needed. Patients and doctors are being misled by promotional pressures to give the drug as widely as possible.

James Li, assistant professor of medicine. 
Larry A Nathanson, instructor of medicine. 
Division of Emergency Medicine, Harvard Medical School, Cambridge, MA 02115, USA jamesli{at}harvard.edu

Trevor J Mills, assistant professor. 
Division of Emergency Medicine, Southwestern Medical Center, University of Texas, Dallas, TX 75390, USA

Karin E Netland, clinical faculty, emergency medicine. 
Charity Hospital, Louisiana State University School of Medicine, New Orleans, LA 70112, USA

Richard Paula, clinical faculty, emergency medicine. 
Tampa General Hospital, University of South Florida, Tampa, FL 33606, USA

Douglas Ragland, clinical faculty, emergency medicine. 
John C Lincoln Hospital, Arizona School of Health Sciences, Phoenix, AZ 85020, USA

Competing interests: None declared.

American Heart Association explains how guidelines were formulated

EDITOROn behalf of the American Heart Association I wish to set the record straight on claims that "money and optimistic claims buttress the `brain attack' campaign" for using alteplase (tPA) in stroke.¹

In accord with its established policies, the American Heart Association implemented a thorough, multilevelled process to prevent any individual or special interest from unduly influencing its guidelines 2000 for cardiopulmonary resuscitation and emergency cardiovascular care.² In addition, a formal evidence based system was used to evaluate scientific data, formulate recommendations, and classify final recommendations.

The development of the guidelines took more than 18 months and involved more than 300 of the world's leading resuscitation experts. The final published recommendations resulted from countless hours spent by panel committees, international scientific councils, an international editorial board, members of the Emergency Cardiovascular Care Committee of the American Heart Association, and writing groups.

All those who developed the guidelines adhered to the American Heart Association's stringent conflict of interest policies and procedures. Participants submitted disclosure forms reporting all relevant relationships with external organisations, groups, and companies. The chairs of panels and subcommittees reminded participants to report these relationships and, if necessary, asked them to refrain from discussion or voting. The relationships reported by panelists are published in the proceedings of the guidelines 2000 conference in the Annals of Emergency Medicine.³

In its more than 75 years the American Heart Association has earned its reputation as a trusted authority by remaining an independent, objective leader focused on reducing disability and death from cardiovascular disease.

David Faxon, president, American Heart Association. 
Cardiology Section, University of Chicago, Chicago, USA tagni.osentowski{at}heart.org

Competing interests: None declared.

Financial information is needed to ensure objectivity

EDITORSaver et al dismiss Lenzer's recommendation of avoidance of all potential bias as "unworkable and undesirable" and "extreme financial correctness" that would leave development of clinical guidelines to "ill equipped" non-experts.¹

Her recommendation is neither unworkable nor undesirable. Financial correctness is exactly what is needed to ensure objectivity. One need not have participated in research to be an expert; one need only have a thorough understanding of the subject and a comprehensive knowledge of the literature. If participation in related research might be valuable, the financial conflicts that arise from it could be mitigated by more research being funded with public money: that is a direction in which we should be moving.

Saver et al claim that thrombolytic agents are highly efficacious in stroke. This assertion is not supported by a single published study. The only positive randomised trial had a number needed to treat of nine and a number needed to harm of 17; this hardly supports the description "highly efficacious." All other numbers that seem positive are the product of statistical manipulation and data snooping. Efforts to identify possible benefit by post hoc analysis of unplanned subgroups are useful only to generate hypotheses to test in further prospective trials.

According to Saver et al, "Concerns about the everyday effectiveness . . . constitute a call to action, not resignation." The call to action of the Cleveland area experience is the re-examination of available data and recommendations and an attempt to replicate the results of the only randomised controlled trial (the National Institute of Neurological Diseases and Stroke (NINDS) trial) that seemed to show benefit. The abysmal results obtained when reporting is non-selective reinforce doubts as to whether they might be replicated in another formal randomised controlled trial.

The authors declare that "Physicians caring for acute stroke patients can and should master the key elements of thrombolytic care or allow patients to be diverted to specialised . . . centres where thrombolytic therapy can be expertly administered." Physicians should master the key elements of care that are of proved benefit. Fibrinolytic treatment is not yet one of those elements. Patients with potentially unstable disease should not be diverted in order to receive a treatment with a possibly unfavourable risk:benefit ratio, especially as most of them will be found not to be candidates for that treatment.

It may be true, as Saver et al claim, that treatment within 90 or 120 minutes would do more good than harm, but such a hypothesis has not been prospectively validated. Until that time, it seems appropriate to remember "First, do no harm."

Robert C Solomon, professor of medicine. 
Ohio Valley Medical Center, Wheeling, WV 26002, USA rcsmdnet{at}nauticom.net

Competing interests: None declared.

Why were these authors of the commentaries chosen?

EDITORLenzer's article on conflicts of interest surrounding recommendations for the use of alteplase (tPA) in ischaemic stroke is important.¹ Despite Saver et al's startling suggestion in their commentary that it would be undesirable for influential organisations to rely on experts without any financial conflict of interest, most readers will be disturbed by the facts that Lenzer gives. We can decide for ourselves whether the changes she suggests are cogent.

I am less sanguine, however, about the choice of authors of the accompanying commentaries. All four authors have financial relations with makers of alteplase. Couldn't the BMJ have found even one author without such a tie? Is it tacitly supporting Saver et al's extraordinary claim that it is impossible to be an expert on a subject about which one does not have a conflict of interesteven when the subject is conflict of interest itself?

Equally distressing is the BMJ's decision to publish an unopposed dismissal of criticism of alteplase for stroke. Lenzer is not a scientist, so it was appropriate that she merely referenced concerns of critics without trying to argue their points in detail. Would the BMJ have solicited two contrary commentaries had her article praised the use of alteplase? (The BMJ could hardly have predicted Warlow's cautionary stance, since he is an investigator in the third international stroke trial (IST-3), "designed to generate the data needed to persuade everyone involved in stroke medicine . . . that rt-PA [recombinant tPA] thrombolysis should be more widely available.")

It is ironic that Saver et al's commentary was published immediately before an article from the series on evidence based medicine, as it is replete with distorted and selective reasoning that violates fundamental principles of clinical epidemiology. Allowing these authors to dismiss Lenzer's concernswith the implication that they are experts, while she is merely an intemperate journalistwithout dissent from opposing experts encourages readers to pay less attention not only to the debate about stroke but also to the influence of money on policy recommendations. "So what if there are competing interests," one might conclude, "if the evidence is so clearly favourable," as Saver et al proclaim.

If the BMJ felt compelled to solicit this attack on critics of alteplase why did it not also solicit a balancing article so that readers could understand the debate and reach their own conclusions? Any number of us who believe that the evidence is far from conclusive, and who worry that widespread community use of alteplase for stroke might do more harm than good, would surely have been willing to write that article.

Jerome R Hoffman, professor of medicine. 
University of California at Los Angeles School of Medicine, Los Angeles, CA 90077, USA jrh{at}ucla.edu

Competing interests: JRH has provided expert consultation in lawsuits against physicians involving the issue of use of alteplase (tPA) in acute stroke. He has taken no personal compensation for this work, having donated all fees to the UCLA (University of California at Los Angeles) emergency medicine residency programme.

Author's reply

EDITORAs noted by Mann and Li et al, ethical questions arise when the full datasets of research trials are not made public and when the guidelines issued in response to those trials are made by those with competing financial interests.

Faxon writes: "All those who developed the guidelines . . . submitted disclosure forms reporting all relevant relationships with external organisations, groups, and companies." He omits to mention, however, that the American Heart Association kept the "disclosures" of its stroke panellists secret, refusing to allow the public to know what was "disclosed." Moreover, public disclosures came only after I started an investigation into why the association did not mention financial conflicts with its original guidelines. This is a good first step. However, it would be far better if the association enacted a policy change that would require routine public disclosure of all competing interests for all of its guideline authors.

The claim by Saver et al in their commentary to my article that experts cannot be found without financial conflicts coincides with the recent and profoundly disturbing assertion by the editor of the New England Journal of Medicine that it cannot find experts without financial conflicts (a point contested by a former editor of JAMA (ABC News, 12 June 2002),¹ prompting it to accept review articles by experts who will be limited to $10 000 annually from each pharmaceutical company. Saver et al list ties to 81 corporations; if an expert consults for five to 10 companies $50 000-100 000 is added annually to his or her income.

One need only recall the thyroid storm debacle to realise the risks of mixing promotional interests with scientific inquiry.² The manufacturer of Synthroid wanted to prove that it was superior to far cheaper generic preparations, so Flint Labs (later acquired by Knoll) hired Dr Betty Dong to prove their case. When her research failed to yield the desired results, Knoll launched a seven year campaign of harassment against Dr Dong when it learnt that she planned to publish her data despite its negative implications for sales. Knoll successfully blocked publication of her results until 1997, when they were published in JAMA, along with a damning editorial about bias and delay.³ Clearly, private financial interests did not serve the public health interestsor that of the public budget.

These events should serve as a warning shot: the line between science and marketing is being rapidly erased. Experts without competing interests may be hard (though not impossible) to find, but it begs the question: what is to be done? So long as public funding is curtailed, many scientists and not for profit organisations will be driven into the arms of corporations. Unless we are willing to support a degree of public funding necessary to attain truly disinterested and objective science, potentially critical conflicts of interest of this sort are inevitable.

Jeanne Lenzer, medical investigative journalist. 
Oak Ridge Journalism, Ellenville, NY 12428, USA Jlenzer1{at}csi.com

Competing interests: None declared.

Authors of commentary reply

EDITORThe statements of Solomon and Li et al about the number needed to harm with alteplase (tPA) are misleading. tPA causes more patients to bleedthe number needed to treat to cause symptomatic intracerebral haemorrhage does approximate 17. However, tPA also prevents an approximately equal number of patients from experiencing symptomatic worsening from stroke extension, cerebral herniation, and other complications of large infarcts. Baldly stated, if you receive tPA the risk is increased that you may bleed and die. If you don't receive tPA the risk is increased that you may herniate and die. The salient number needed to harm is the net sum of these two factors, and across all under three hour trials there is no net harm. Analysing all seven trials with available under three hour data (NINDS 1 and 2, ECASS 1 and 2, ATLANTIS A and B, Haley 1993), death occurred in 83/479 (17.3%) of tPA treated patients and 83/478 (17.4%) of placebo treated patients (P=0.9). Thus, the number needed to treat to produce benefit from tPA is as low as two, the number needed to treat to cause net harm approaches infinity. These numbers amply support the statement that tPA is highly efficacious.

Hoffman wishes the BMJ had solicited commentaries more accurately reflecting the balance of opinion on tPA in acute stroke. So do we. It is important, however, to realise what such balanced commentaries would look like. The typical opinion page dichotomy of one pro opinion and one con opinion that the BMJ arranged provides an entirely misleading view of the state of informed opinion. Among American experts on stroke there is overwhelming consensus that tPA is efficacious. A balanced set of expert commentaries would include an order of magnitude greater number of positive opinions than negative opinions.

Hoffman also erects a straw man. We never suggested that "it is impossible to be an expert on a subject about which one does not have a conflict of interest." Rather, we merely suggested that many experts will have minor competing interests. In this regard, it is noteworthy that in his letter Hoffman has declared his own financial competing interest with regard to the use of tPA in stroke. If he were to be true to the absolutist position on financial conflicts that he has advanced, he will now absent himself from advising influential and independent organisations on the tPA in stroke issue.

Li et al repeat the claim that the National Institute of Neurological Diseases and Stroke (NINDS) tPA study was a single trial. This myth was long ago demonstrated to be false¹ but persists among tPA contrarians as an article of faith impervious to actual evidence. However devoutly the tPA contrarians wish they only had a single NINDS-tPA trial with which to contend, the fact is there were two trials, as the Food and Drug Administration recognised when ascertaining that the evidence for the benefit of tPA in acute stroke was quite adequate to approve the indication.²

We will close by again seeking common ground. We concur, as before, with calls to bar experts with major financial competing interests from service on guideline committees and to require experts with minor financial competing interests to disclose them publicly. And we reiterate our concurrence with policy statements of the Brain Attack Coalition and the American College of Emergency Physicians that urge emergency physicians and other acute care providers to become expert in acute stroke care, including the use of tPA for acute stroke, and to place their hospitals on standby and divert patients to designated stroke centres where treatment can be expertly delivered. ^{3 4}

Jeffrey L Saver, associate professor, neurology. 
Chelsea S Kidwell, assistant professor, neurology. 
Sidney Starkman, professor, emergency medicine and neurology. 
UCLA Stroke Center, Department of Neurology, and Department of Emergency Medicine, University of California, Los Angeles, USA

JLS, CSK, and SS have served as site investigators in acute stroke clinical trials sponsored by several (15, 11, and 17 respectively) pharmaceutical and biotechnology companies, including Genentech and Boehringer-Ingelheim; have received speaking honorariums from several (12, 5, 8) pharmaceutical companies, including Genentech and Boehringer-Ingelheim; and have served as consultants on scientific advisory boards for several (7, 1, 5) pharmaceutical and biotechnology companies developing acute stroke treaments, including Boehringer-Ingelheim and Genentech. JLS and SS have provided expert consultation in lawsuits against physicians involving the issue of use of tPa in acute stroke.