Author + information
- Larry A. Weinrauch, MD∗ ()
- ↵∗Harvard Medical School, 521 Mount Auburn Street, Suite 204, Watertown, Massachusetts 02472
The recent publication and editorial by Packer et al. (1,2) on flosequinan are welcome additions to the published data. This now published trial raises important issues. Perhaps despite delayed publication we can still learn more. The primary end point was all-cause mortality. The secondary endpoints were death due to cardiovascular reasons, mortality due to sudden death or pump failure, occurrence of hospitalization or the use of intravenous medications for heart failure and symptoms, and functional status. Clinicians and patients excited by reductions of heart failure symptoms among our sickest patients were dismayed when evidence of excess all-cause mortality emerged. As described, removal of the drug was associated with rapid return to premorbid heart failure class. Observations that reduction of symptoms and heart failure class were associated with decreased longevity led to uncomfortable discussions with patients. Many Class 4 heart failure patients offered the choice preferred a life less long with reasonable comfort over longevity served swollen and breathless. Sudden death was perceived as a better ending. The United States Food and Drug Administration (FDA) made this decision for the patients by removing the availability of the drug. We must consider this an appropriate decision at the time, but perhaps reconsider in designing trials whether all-cause mortality trumps all morbidity while understanding that short-term benefit in symptoms or biomarkers does not predict survival benefit as has been demonstrated repeatedly.
Several issues bear further examination:
1. Did testing of flosequinan in the 1990s predict what testing of the same drug would demonstrate today? The prevalence of use of digoxin at the time was higher; beta-blockade was likely lower, eplerenone and spironolactone not likely used, and ivabradine and angiotensin II receptor blocker neprilysin inhibitors unavailable. Left ventricular assist devices (LVADs) were in their infancy, and routine use of implantable cardioverter-defibrillators (ICDs) had not commenced. If the excess of deaths were sudden, would use of ICDs have saved this drug?
2. Were comorbidities balanced? If the excess deaths were related to diabetes, pulmonary disease, or infections, was there a relationship to the medication? If a greater percentage of the flosequinan patients were diabetic, would the results of the trial be correctly interpreted?
3. In our efforts to reduce all-cause mortality, we often forget that we learn more about patient care from adjudication of events than from counting of death certificates. In this trial, how were the endpoints adjudicated? What were rates of death due to defined cardiovascular causes (and what were these causes), pump failure, or sudden death?
4. Similar to problems of short-term markers, publication has become a marker of research productivity. But the peer review process ends not with publication, rather with discussion of the findings over time. This publication as a featured article in the “dead letter office” highlights an important effort that should be continued. Forgetting, shelving, or not publishing a “negative” trial may have adverse future consequences.
We thank the authors for their publication and the difficult effort in resurrecting this data and fulfilling our promise to those patients who have permitted us to share their information and care.
Please note: Dr. Weinrauch has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Packer M.,
- Pitt B.,
- Rouleau J.L.,
- Swedberg K.,
- DeMets D.L.,
- Fisher L.
- Packer M.