Post–Cardiac Surgery Mechanical Support
A Tool for Decision Making
The report by Hernandez et al1 in the present issue of Circulation is timely and important. Compiled from the Society of Thoracic Surgeons voluntary database, it comprises an analysis of the largest series of patients treated with mechanical support to date. The authors have reviewed outcomes of 5735 patients who had a ventricular assist device (VAD) placed for low cardiac output after having undergone a cardiac surgical procedure. Remarkably, this represents <0.5% of all cardiac operations. The authors observed significant improvement in outcome for these patients over a recent 10-year study period. Survival improved from 38.5% to 59.2%. Similarly, during the same time period, there was a reduction in major assist-device morbidity of bleeding, stroke, and renal failure. These results are in keeping with those from smaller, single-institution reports.2–4
Article p 606
The authors attribute the improved results to advances in technology and more standardized postoperative care protocols. This has encouraged surgeons to implant assist devices earlier, before occurrence of irreversible end-organ damage.5 The most important determinant of poor outcome was whether the index procedure was being performed for salvage or reoperation. Preoperative need for dialysis was also an important predictor of mortality. These observations support those of others.6 Table 6 in the article by Hernandez et al1 shows that the majority of patients who had VAD placement as a “bridge” to transplantation or as permanent lifetime therapy (destination) had left-sided support only, whereas those who were expected to recover had mostly right-sided support. Also, the authors noted that only 7 centers performed >50 VAD implantations over the 10-year study period.
Typically, there are 3 scenarios that can lead to the need for mechanical support early after cardiac surgery. First, VAD/transplantation can be used to provide cardiac support postoperatively in high-risk cases. In such cases, patients may be asked to provide their consent to the procedure ahead of time for this potential need. Second, there can be unexpected poor postoperative cardiac function. This use of VAD is unanticipated. Third, in the setting of percutaneous coronary intervention/CABG for acute myocardial infarction and cardiogenic shock, VAD support may be required for clinical salvage.
The data presented probably reflect these 3 scenarios. One can expect a trend toward certain referral centers performing higher-risk procedures with VAD/transplantation backup. Patients undergoing such procedures include those with low left ventricular ejection fraction due to a variety of causes. They also usually have a dilated left ventricle, possibly with mitral valve regurgitation. These patients can experience low cardiac output postoperatively due to isolated preexisting left ventricular dysfunction. In the present report, most of these patients received isolated left-sided support. In such patients, the right ventricles are typically adapted to chronically elevated pulmonary venous hypertension and will not require support postoperatively. In such scenarios, the assist device provides cardiac support in case there is low cardiac output after cardiopulmonary bypass. Most of these patients will go on to cardiac transplantation after a period of time to allow end-organ recovery and physical therapy.
Next is the case of a bridge to recovery, when right ventricular support was required more frequently. In such instances, the hope is to achieve myocardial recovery. Most often, these are patients with unanticipated cardiac failure. Biventricular support is occasionally required, but the right ventricle is primarily failing. In these patients, an assist device is used temporarily as a bridge to recovery. Transplantation will be required rarely but remains a possibility. These patients may be transferred to referral hubs in the postoperative period to take advantage of the expertise in high-volume VAD programs.7–9
Finally, the authors observed that preoperative cardiogenic shock was ranked tenth among preoperative risk factors for mortality. The better-than-expected outcomes in these patients likely reflect early institution of assist-device support, before the occurrence of end-organ damage such as renal or hepatic failure.
In summary, Hernandez and colleagues are to be commended for bringing these data to light using the extensive Society of Thoracic Surgeons database. Rapidly deployable devices are currently available in many if not most cardiac surgery centers. Once on support, patients have the option of being transferred to referral centers if transplantation or other advanced care is required. Such hub-and-spoke arrangements represent opportunities for partnerships. Encouraging results such as those reported in the present study should stimulate surgeons to implant devices early in the course of low postoperative cardiac output, before end-organ damage becomes irreversible.
Dr Marelli is an occasional consultant for Abiomed Inc in Danvers, Mass. Dr Shemin reports no conflicts.
The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.
Hernandez AF, Grab JD, Gammie JS, O’Brien SM, Hammill BG, Rogers JG, Camacho MT, Dullum MK, Ferguson TB, Peterson ED. Decade of short-term outcomes in post–cardiac surgery ventricular assist device implantation: data from the Society of Thoracic Surgeons’ National Cardiac Database. Circulation. 2007; 116; 606–612.
Dekkers RJ, FitzGerald DJ, Couper GS. Five-year clinical experience with Abiomed BVS 5000 as a ventricular assist device for cardiac failure. Perfusion. 2001; 16: 13–18.