By Jamie L.W. Kennedy, MD, FACC
Associate Professor, Division of Cardiology, Advanced Heart Failure & Transplant Cardiology, University of California, San Francisco
Dr. Kennedy reports no financial relationships relevant to this field of study.
SYNOPSIS: A retrospective single-center analysis of patients with isolated severe tricuspid regurgitation revealed no survival benefit from tricuspid surgery.
SOURCE: Axtell AL, Bhambhani V, Moonsamy P, et al. Surgery does not improve survival in patients with isolated severe tricuspid regurgitation. J Am Coll Cardiol 2019;74:715-725.
The presence of significant tricuspid regurgitation (TR) carries a poor prognosis; however, the benefit of intervention has not been studied rigorously. Most TR is secondary to left-sided heart disease, either valvular or myopathic, or pulmonary vascular disease. Primary TR from leaflet pathology is less common. Current expert consensus guidelines recommend tricuspid intervention for patients undergoing left-sided valve surgery and for patients with severe primary TR with either symptoms or progressive right heart dilation/dysfunction. Intervention for secondary TR in the setting of left heart failure, either systolic or diastolic, or pulmonary vascular disease is not addressed in the guidelines. Isolated tricuspid surgery, either repair or replacement, is a surprisingly high-risk procedure, with the in-hospital mortality rate ranging from 8% to 10%. Volumes are low, and there is not a Society of Thoracic Surgeons risk model for either repair or replacement.
Axtell et al used a single-center echocardiography database to identify patients with severe TR between November 2001 and March 2016. Patients with moderate or worse aortic, mitral, and pulmonic valve disease were excluded, leaving 3,276 patients with “isolated” TR. These patients were followed through April 2018; median follow-up was 2.6 years. Patient care decisions, such as frequency of echocardiograms, medications, and surgery, were made by clinicians independent of the study. Over the course of the study, 171 patients underwent surgery (143 repair and 28 replacement) and 29 underwent concomitant coronary artery bypass grafting. Not surprisingly, compared to those medically treated, patients undergoing tricuspid surgery were younger, with higher average ejection fraction (52% vs. 57%) and lower rates of coronary disease (31% vs. 16%), heart failure (83% vs. 72%), diabetes (4% vs. 1%), and chronic kidney disease (39% vs. 20%). Interestingly, 15 patients in the medical management group underwent coronary artery bypass grafting without tricuspid intervention. From the isolated TR patient panel, a propensity-matched sample was selected. The sample consisted of pairs of similar patients managed medically and surgically. It is unclear from the paper which variables were used for matching. Interestingly, there was substantially more secondary TR in the medical group and primary TR in the surgical group. The medical and surgical groups were similar in terms of right ventricular dilation and estimated right ventricular systolic pressure (RVSP). There was an apparent survival benefit with surgery at this point of the analysis; however, the authors adjusted for the time between diagnosis of severe TR and surgery (median, 3.7 months), the immortal time bias. Following this adjustment, there was no difference in survival between the two groups. Among patients who did undergo surgery, there was no difference in survival between repair vs. replacement.
As expected, surgical outcomes were worse for older patients and those with heart failure. There may have been significant clinical changes in the time between diagnosis and surgery. The authors performed a second analysis using clinical parameters at the time of surgery and found no difference in the outcome. The authors repeated the propensity matching analysis, including echo-derived RVSP. They found no difference in survival between medically and surgically managed patients. Finally, Axtell et al attempted to correct for those lost to follow-up by studying a “loyalty cohort” of patients who received most of their care at the study institution. Again, there was no difference between medical and surgical groups. The authors concluded that surgery for isolated TR does not improve survival. They mentioned the possibility that delayed intervention allows progressive right heart dysfunction and end-organ dysfunction, thereby limiting the benefit of intervention. However, the median time between echo diagnosis of severe TR and surgery was only 2.9 months in the overall cohort and 3.7 months in the propensity-matched sample.
A significant limitation in many studies of TR is the combined analysis of primary and secondary (or organic and functional) etiologies. The study by Axtell et al was no exception. Patients with moderate or severe aortic, mitral, or pulmonic valve disease were excluded, but patients with secondary TR due to left ventricular systolic or diastolic dysfunction were included, as were patients with pulmonary vascular disease. The clinicians caring for these patients clearly took etiology into account in their management decisions. Only 39% of patients in the surgical arm of the propensity-matched group had secondary TR vs. 79% in the medically managed arm. Unfortunately, the rate of secondary TR for the entire cohort was not reported. Despite the preponderance of primary TR in the surgical patients, there was no survival benefit to surgery, an intriguing finding as current guidelines advocate for intervention in primary TR.
I am puzzled by the lack of information on hepatic function in these patients. Congestive hepatopathy and cardiac cirrhosis are well-recognized complications of long-standing TR. Liver disease is a powerful predictor of outcomes for patients undergoing cardiac surgery. Perioperative mortality for Child-Pugh class A is reasonable at 5%, marginal at 35% for class B, and likely prohibitive at 70% for class C. For this study, I imagine patients with significant hepatic impairment were medically managed. In light of this, it is again intriguing that surgery was not beneficial.
Based on the results of this study, tricuspid surgery may not improve survival and should be undertaken cautiously, after thorough preoperative evaluation (including a complete hemodynamic study and end organ assessment). I hope the authors of future studies differentiate primary and secondary TR and capture information on hepatic impairment. Similar to our experience with functional mitral regurgitation, we may need a well-conducted percutaneous intervention trial to truly assess the role of tricuspid intervention.