International Differences in the Treatment of Sepsis: Title and subTitle BreakAre They Justified?

Imagine a patient with acute coronary syndrome on a flight from New York to Tokyo. Given how small the world of medicine has become, it is reasonable to assume that the treatment he/she would receive would be very similar no matter if the plane returned to New York or continued to Tokyo. Now imagine that the patient had septic shock instead of heart disease—landing in New York or Tokyo would result in drastically different treatment. In New York, early goal-directed therapy (EGDT) and drotrecogin alfa are widely used while the former is less common and the latter is unavailable in Tokyo. Conversely, for more than 10 years, a patient with sepsis would likely receive therapy with an endotoxin adsorber (polymyxin B hemoperfusion) in Tokyo, a treatment unavailable in New York.

Why such radical differences? Aren't physicians treating sepsis in Japan and the United States practicing evidence-based medicine? It turns out that evidence is indeed at the heart of these international differences—but the interpretation of the same evidence seems to vary. The opponents of EGDT and drotrecogin alfa argue that there is insufficient evidence to support their use. Widespread adoption of EGDT in the United States was based on one small single-center randomized trial.¹ Although approval for drotrecogin alfa was based on a large multicenter pivotal trial,² subsequent negative results in low-risk patients³ and children,⁴ combined with higher risk of bleeding, have led to doubts in many parts of the world about the effectiveness of the drug. Similarly, polymyxin B hemoperfusion lacks rigorous clinical trial data and is therefore considered unproven by many regulatory agencies.

However, if these therapies lack sufficient evidence, why are they avoided in some places and virtually “standard of care” in others? Part of the answer may be historical. For instance, achieving supraphysiologic hemodynamics for patients with sepsis through fluid resuscitation and vasoactive drugs has always been more of the “ICU culture” in the United States and Europe whereas “blood filtering” has been more readily accepted for sepsis in Asia.⁵ A more cynical interpretation might be that drotrecogin alfa was developed by a US company, whereas polymyxin B hemoperfusion was developed in Japan.

Against this backdrop, in this issue of JAMA, Cruz and colleagues⁶ report findings from the EUPHAS trial, testing the effectiveness of polymyxin B hemoperfusion in several Italian centers. This preliminary study is valuable as an example of “New Yorkers” testing a Japanese intervention. This kind of cross-community validation is refreshing and necessary but unfortunately only too rare. The results, although preliminary, suggest a number of interesting hypotheses and should provoke further study. This is essential given the significant ongoing problem that sepsis represents.

However, just as interesting as the results is the way the investigators and their oversight board chose to respond to results. Instead of producing the sort of solid scientific evidence that would likely eliminate the differences between the 2 communities, the investigators have, in essence, left New York and landed in Tokyo. But how did this happen and what does it say about the treatment of sepsis around the world?

The EUPHAS trial was not designed as a definitive trial with a patient-centered end point. Instead it was designed to determine whether polymyxin B hemoperfusion would result in improved mean arterial pressure and less requirement for vasopressors in patients with septic shock due to presumed abdominal infections. Given the modest differences in these end points it was therefore surprising that 28-day mortality (a secondary end point) was so drastically different between groups: 11/34 (32%) with polymyxin B vs 16/30 (53%) with conventional therapy. Even more surprising was the authors' response: “Results were discussed with the president of the ethics committee . . . who declared it unethical to deprive a potentially beneficial therapy to a group of patients that carry high mortality.”⁶

Thus, an oversight body has, on the basis of a secondary analysis of an underpowered study, changed the standard of care for a series of hospitals participating in the study and presumably affected the standard of care for other hospitals in the region. Of note, the statistical difference in mortality would be abolished by a different outcome in a single patient.

Another consideration is the potential influence of this decision on the practice of medicine. First, to suggest that it is unethical to deprive patients of polymyxin B hemoperfusion infers that clear and conclusive evidence is available that the therapy is effective. The current study does not provide such evidence but other studies have also found that polymyxin B hemoperfusion is effective⁷ —did the current study tip the evidence so far in favor of a positive effect that further study is unwarranted? Such a position is incongruous with other decisions being made by the medical community at large and is inconsistent with the authors' own conclusions.

For example, the Surviving Sepsis Campaign has recommended EGDT, tight glucose control, and drotrecogin alfa for patients with severe sepsis.⁸ Yet despite these recommendations additional studies have been called for and in the case of tight glucose control, the treatment has been found to be harmful.⁹ Multicenter studies of EGDT and drotrecogin alfa are under way, although the perception among some practitioners that these therapies are proven has made further study challenging. Indeed, this impression has also been a considerable barrier to conducting randomized trials of polymyxin B hemoperfusion in Japan. Compared with the evidence available from EUPHAS, much stronger evidence existed to establish the efficacy of high-intensity hemofiltration (35 vs 20 mL/kg/h) but a subsequent definitive trial was called for and ultimately showed no benefit.¹⁰ Even the EUPHAS investigators have called for further investigation into the effectiveness of polymyxin B hemoperfusion when they concluded that “Larger multicenter studies are indicated to confirm these encouraging findings in other patient populations.”⁶ Such a conclusion would seem difficult to reconcile with the belief that continuing the current trial would have been unethical.

Second, by failing to provide a definitive answer to the question of whether polymyxin B hemoperfusion is life saving, the oversight board has deprived millions of patients around the world of this therapy because few, if any, regulatory agencies would permit these results to be used for product approval. The perils of early termination of clinical trials are well known; the EUPHAS study is yet another example.

Third, even setting aside the issue of mortality and the early termination of the trial and taking the study's primary results at face value, the EUPHAS study provides strong rationale for doing exactly what the authors advocate—conduct further studies. The position taken by the oversight board to suggest that further study would be unethical should be rescinded. Indeed, willful ignorance of the true effect of polymyxin B hemoperfusion, whatever it is, is unethical. To keep this promising therapy in the shadows of opinion and insufficient evidence would be to withhold potentially beneficial treatment from patients deprived of it, should it be effective, or to put patients at risk, should it be harmful.

Although harm seems unlikely, it is not inconceivable. Extracorporeal circuits are never 100% biocompatible and thrombocytopenia is a known complication of polymyxin B hemoperfusion. Even the removal of endotoxin, a signal to the immune system of bacterial invasion, is not completely without risk.

In the broader context, a much higher standard of evidence is necessary prior to declaring that further study of medical therapy is unethical. Broad consensus exists on the protection of human research participants,¹¹ and these principles should always be followed. However, premature declaration of certainty serves to curtail the discovery of truth, and only by discovering the truth can the best care be provided for patients from Tokyo to New York and everywhere else.