Registry Charts Trends in Mechanical Circulatory Support During PCI

The use of intra-aortic balloon pumps (IABPs) to support PCI in patients with cardiogenic shock has been declining in recent years, according to data from a nationwide registry published online August 18, 2015, ahead of print in Circulation. Additionally, this decline has not been accompanied by increased use of other, newer mechanical circulatory support devices. 

“Our findings likely reflect continued uncertainty of clinical benefit in the use of IABP and [other mechanical circulatory support] for high risk PCI,” write researchers led by Amneet Sandhu, MD, of the University of Colorado School of Medicine (Aurora, CO).

They examined data from the NCDR CathPCI registry on 76,474 patients who underwent PCI in the setting of cardiogenic shock at 1,429 centers from 2009 to 2013.

Slightly more than half (54%) received no mechanical circulatory support, 39% received IABP only, 3.5% received other mechanical circulatory support, and 3.6% received both IABP and other mechanical circulatory support. The latter category included TandemHeart (CardiacAssist) and Impella (Abiomed) percutaneous left ventricular assist devices (LVADs) and extracorporeal membrane oxygenators.

Compared with patients who received no mechanical circulatory support, those who did were more likely to have heart failure within the prior 2 weeks, LV systolic dysfunction, or cardiac arrest within 24 hours. Additionally, those receiving support also were more likely than those who did not to be in cardiogenic shock at the start of PCI (88% vs 75%; P < .001).

Temporal Trends

In 2009, 45% of patients undergoing PCI in the setting of cardiogenic shock received an IABP, while 6.7% received other mechanical circulatory support. From that time until late 2012, when the IABP-SHOCK II trial was published, the proportion of patients receiving IABP declined at an average rate of 0.25% per quarter. The trial showed no benefit of IABP therapy vs medical therapy alone in the treatment of shock patients. In the period immediately following publication of the trial, IABP use declined by 3.2%.   

However, by the first quarter of 2013 until the end of that year, rate of IABP use resumed only a marginal decline. Throughout the study the rates of use of other mechanical circulatory support devices did not change.

The predicted probability of IABP use in patients with cardiogenic shock varied broadly by site (hospital median 42%, interquartile range 26-51%), while the probability of other mechanical circulatory devices being used was clustered at a small number of hospitals (< 5% in half of hospitals and > 20% in less than one-tenth of hospitals). Centers with greater use of IABP or other devices were larger (> 600 beds), primarily university hospitals or teaching hospitals, and less likely to be private or community hospitals.

In a secondary analysis of 579,079 patients undergoing high-risk PCI, which included patients with cardiogenic shock, IABP use declined an average of 0.05% from 2009 to late 2012. Following the publication of IABP-SHOCK II, the percentage of those patients receiving IABP decreased by 1% but soon returned to what it had been prior to publication. Other mechanical circulatory support increased in use by an average of 0.02% from 2009 until late 2012, with no change in the months immediately following IABP-SHOCK II through the end of the study period.

According to the study authors, the lack of a more pronounced decline in IABP use since IABP-SHOCK II’s publication may stem from the therapy being downgraded to a class IIa indication for cardiogenic shock in the 2013 American College of Cardiology/American Heart Association STEMI guidelines.

Furthermore, they say, the clustering of newer mechanical circulatory support technologies at a fraction of hospitals with little uptake in the broader cardiology community is a novel finding. Explanations may relate to issues of availability of the technologies, operator unfamiliarity, patient clinical status, or lack of data showing improvement in outcomes, they suggest.

Data Divergence

The NCDR data, however, stand in contrast to several other studies, including one that examined National Inpatient Sample data and found a dramatic increase in short-term use of mechanical circulatory support over the past decade. But Dr. Sandhu and colleagues say they believe the NCDR database “provides a broader picture of [its] use in the specific setting of PCI.”

In a telephone interview with TCTMD, Jeffrey W. Moses, MD, of Columbia University Medical Center (New York, NY), disagreed, saying the disparity between this study and others is “difficult to reconcile,” adding that it “does not conform with other data we have seen in terms of the use of mechanical support over the last 2 to 3 years.”

Given the ongoing research on LVADs, the NCDR data may be slightly outdated. Recent registry data have indicated reduced mortality with the early use of the newer LVADs in shock patients and high risk PCI, Dr. Moses noted. Additionally, the Impella 2.5 device (Abiomed) received FDA approval earlier this year for use in high risk PCI, which likely will impact rates of use going forward.

“I don’t know if the data in NCDR will remain true if they were to look at it again this year or next year,” Dr. Moses said. “I would be surprised if more contemporary data does not show a marked uptick in use of these devices.”

Missing Details Muddy Interpretation

In an accompanying editorial, Garrick C. Stewart, MD, of Brigham and Women’s Hospital (Boston, MA), cautions that the broad classification of reduced cardiac index and/or hypotension used by the researchers to define shock “captures a heterogeneous group of patients with varying degrees of impairment,” adding that patient-level information regarding hemodynamics or use of vasoactive medications was not included.

Additionally, the category of ‘other mechanical circulatory support’ does not clarify which platforms were used. There also is no information provided, he notes, on how frequently support devices were use in “separate, later encounters for salvage therapy after initial PCI for shock was performed without support,” and no outcomes data (in-hospital or 30-day mortality) for comparison of support strategies.

Taken as a whole, Dr. Stewart says the study reflects the challenges of treating patients with shock.

“Entire regional health systems have been redesigned so patients can realize the proven benefits of early PCI,” he writes. “Yet whether or not a patient receives percutaneous support during PCI appears to have as much to do with where the ambulance happens to take the patient as how sick the patient might be. Protocols for the recognition of shock and triage to [mechanical circulatory support] often remain peculiar to each institution.”

According to Dr. Stewart, additional efforts are needed to classify the stages of cardiogenic shock and define how best and in whom to use temporary support devices.

---

Sources: 
1. Sandhu A, McCoy LA, Negi SI, et al. Utilization of mechanical circulatory support in patients undergoing percutaneous coronary intervention: insights from the NCDR. Circulation. 2015:Epub ahead of print.
2. Stewart GC. Finding traction for mechanical circulatory support during coronary interventions [editorial]. Circulation. 2015:Epub ahead of print. 

Disclosures:

• Drs. Sandhu and Stewart report no relevant conflicts of interest. 
• Dr. Moses reports serving as a consultant to Abbott and Boston Scientific. 

Related Stories:

• Mortality, Hospital Costs Drop as Short-term Mechanical Circulatory Support Use Grows
• Usefulness of IABP Questioned in MI Patients with Cardiogenic Shock
• IABP Therapy May Benefit Patients With Large STEMI, Persistent Ischemia Post-PCI