Treatment of Obstruction in Hypertrophic Cardiomyopathy

Executive Summary

The WG first discussed whether it would be feasible to compare these modalities with a randomized clinical trial (RCT).The WG participants felt that an RCT would not be feasible due to the rarity of the condition, the much smaller subset of patients who experience intractable symptoms due to outflow obstruction, and the need for such patients to be eligible and willing to be randomized to either procedure. In addition, all felt that such a trial would need to focus on clinical outcomes and not physiologic surrogates, but this would be difficult because morbid events are rare. All of these factors were felt to result in a prohibitively large sample size. Other very important factors arguing against the feasibility of an RCT included a perception that physician advocates of either procedure might be unwilling to randomly assign treatment and the small number of U.S. centers (about five) equally adept at SSM and ASA to allow individual patient randomization. One large center estimated the volume of isolated SSM at approximately 100 procedures per year, but the participants were unaware of accurate estimates of the volume of these procedures in the US. (A CPT code specific for ASA will not be implemented until 10/2014.)

The discussion turned to observational registry data and existing resources that might potentially be leveraged for a comparative effectiveness study of these treatments. Most centers with expertise in this area have existing local clinical registries. A "registry of registries" was discussed but most felt that the research questions would best be answered with a prospective registry with standard data elements. In addition to local U.S. registries, registries exist in the United Kingdom, Canada, and Europe, and a prospective French research registry is currently enrolling patients. The group heard of the efforts of the NHLBI-funded Cardiovascular Research Network (CVRN), which provides a platform for electronic medical record-based research, but it was recognized that, while the catchment area for the funded centers includes 11 million patients, it does not contain centers with large volumes of either procedure. Representatives of professional society registries discussed their existing data in this arena. The Society of Thoracic Surgeons database currently has information limited to diagnosis, surgical procedure performed, and short term outcomes, but does have 90% participation nationally. Specific modules can be created for specific procedures, such as has been done for maze surgery in those centers with a large volume of surgery for atrial fibrillation. The STS has had research experience comparing the surgical data with the catheter-based procedure data of the American College of Cardiology's National Cardiovascular Data Registry (NCDR). The NCDR currently does not capture alcohol septal ablation procedures but collaborates with STS on the Transcatheter Valve Therapy (TVT) Registry. It was thought that many centers with expertise in HCM therapies are likely centers that already participate in TVT and that this might provide an opportunity for leveraging existing resources.

The WG participants next discussed the key elements needed to perform a comparison of these procedures in a prospective registry. While the group agreed on the need to include both short-term and long-term clinical outcomes and quality of life, the group was divided on whether the questions were best answered by the broadest base of centers possible, as opposed to including only those that were considered expert in one or both procedures. This consideration is complicated by the WG's perception that surgical septal myectomy is a difficult and low-volume procedure that is performed primarily in surgical centers of excellence, whereas alcohol septal ablation might be performed by cardiac interventionalists with less procedure-specific training, so that variability in technique and expertise may well be greater. Because of this technical variability, it was considered essential that a registry include specific procedural details to inform best practices and accurate co-morbidities to allow a fair comparison of outcomes since the patients referred for each procedure may be very different patient populations.

Finally, the statistical considerations of research based on a prospective registry were discussed. It was noted that high quality results may be obtained from prospective registries that are designed and audited with the same rigor as an RCT. It might also be possible to embed an RCT of interest into a well-designed registry. The low clinical event rate remains a statistical problem in a registry as well as in an RCT; it is possible that even with long periods of follow-up there may not be enough difference in outcomes to influence any change in practice. Because of the nature of the disease process, arrhythmic events in an HCM registry would be hard to attribute to the therapy chosen rather than to the myopathy itself unless the differences between treatment groups were very large. On the other hand, some participants noted that advances in knowledge over the past decade in the form of small clinical studies and a growing clinical experience have allayed some of the earliest concerns about the arrhythmic potential of ASA.

Points of consensus and differences of opinions

The WG agreed on the importance of a comparative effectiveness study, especially in light of the costs of these procedures, and a need first to obtain an accurate estimate of the number of procedures which may be best accomplished through professional societies' databases. There was consensus that an RCT would be extremely difficult to complete. The highest quality observational data would likely come from a prospectively designed registry, perhaps added on to existing registries, and an endeavor similar to, or added to, the TVT might provide a good opportunity to leverage existing infrastructure. The registry should include long-term outcomes and granular procedural details. Although the rarity of clinical events might make it difficult to find a large enough difference in outcomes to change practice, valuable information about best technical practices could be obtained, leading to more standardized performance of procedures. The group was divided on whether a registry is best limited to expert centers vs. broad-based centers; however, the latter would be more in concert with the principles of comparative effectiveness studies. The WG felt that a longer, in-person discussion or workshop to better define the details of such a registry, or obstacles to participation, would be welcomed.

Working Group Members

• Eugene Braunwald, MD (WG Chair), Harvard Medical School, Boston, MA
• Ralph Brindis, MD, NCDR and University of California at San Francisco, San Francisco, CA
• Bernard Gersh, MD, PhD, Mayo Clinic, Rochester, MN
• David Holmes, Jr., MD, Mayo Clinic, Rochester, MN
• Kerry Lee, PhD, Duke Clinical Research Institute, Durham, NC
• Barry Maron, MD, Minneapolis Heart Institute Foundation, Minneapolis, MN
• Martin Maron, MD, Tufts University Medical School, Boston, MA
• Christine Seidman, MD, Harvard Medical School, Boston, MA
• David M. Shahian, MD, STS and Harvard Medical School, Boston, MA
• Stephen Sidney, MD, MPH, Kaiser Permanente Northern California, Oakland, CA
• Nicholas Smedira, MD, Cleveland Clinic, Cleveland, OH

DCVS Participants

• Alice M. Mascette, MD
• Diane Bild, MD
• David A. Lathrop, PhD, (staff contact)
• Marissa Miller, DVM, MPH
• Gail Pearson, MD, ScD

Last Updated: June 2013