[Skip to Navigation]
Comment & Response
May 2016

Mechanical Circulatory Support and Rationale for Future Research—Reply

Author Affiliations
  • 1Division of Cardiology, University of California-San Francisco, San Francisco
  • 2Department of Internal Medicine, University of California-San Francisco, San Francisco
JAMA Intern Med. 2016;176(5):715-716. doi:10.1001/jamainternmed.2016.0813

In Reply We agree with Romeo et al that the trials with the intra-aortic balloon pump (IABP) have largely been negative, and perhaps more trials with this device are unlikely to yield significantly different results. However, randomized clinical trials (RCTs) with newer circulatory support devices would be useful. In the dearth of effective medical interventions to rescue patients with pump failure and circulatory shock, mechanical support could be helpful if it is proven to improve outcome. Newer circulatory support devices like Impella (Abiomed) potentially have superior hemodynamic profile; however, they have not been shown to improve hard clinical outcomes. As shown by Khera et al,1 the use of Impella continues to increase over years. Increasing comfort with its use and dwindling confidence in the IABP are undoubtedly responsible. Yet there is no RCT confirming significant improvement in clinical outcomes, especially in patients with cardiogenic shock. It is surprising that based on small underpowered studies devices like Impella are quickly becoming the standard of care against which even newer devices like Percutaneous Heart Pump (Thoratec Inc) are being compared (NCT02468778). If we believe that Impella or other newer devices indeed improve outcomes in cardiogenic shock, do we not need to prove this with a compelling RCT against standard medical therapy? If we fail to do this, are we not repeating the same mistakes that led to the use the IABP for 40 years without outcome data? We contest that new RCTs are desperately needed to prove the efficacy of newer mechanical circulatory devices in improving hard outcomes like mortality. Ethical dilemmas in randomizing sick patients are legitimate but should not be prohibitory. The IABP SHOCK-II trial2 showed that patients in cardiogenic shock could be successfully randomized and compared with standard medical therapy. Moreover, because sick patients are most likely to derive the potential benefit from mechanical circulatory support, a relatively small sample size should sufficiently yield statistically significant results. Careful selection of the study population with concerted efforts to avoid enrollment of low risk patients would also be important to preserve the power of the study.3

Add or change institution