Competing risks: Competing questions

Paraconduit hernia following esophagectomy: Is it safe to watch and wait?

OBJECTIVES

We hypothesized that emergency complications related to asymptomatic paraconduit hernias may occur less often than generally believed. Therefore, we assessed the occurrence and timing of paraconduit hernia diagnosis after esophagectomy, as well as outcomes of these asymptomatic patients managed with a watch-and-wait approach.

METHODS

From 2006 to 2021, 1214 patients underwent esophagectomy with reconstruction at the Cleveland Clinic. Among these patients, computed tomography scans were reviewed to identify paraconduit hernias. Medical records were reviewed for timing of hernia diagnosis, hernia characteristics, and patient symptoms, complications, and management. During this period, patients with asymptomatic paraconduit hernias were typically managed nonoperatively.

RESULTS

Paraconduit hernias were identified in 37 patients. Of these, 31 (84%) had a pre-esophagectomy hiatal hernia. Twenty-one hernias (57%) contained colon, 7 hernias (19%) contained pancreas, and 9 hernias (24%) contained multiple organs. Estimated prevalence of paraconduit hernia was 3.3% at 3 years and 7.7% at 10 years. Seven patients (19%) had symptoms, 4 of whom were repaired electively, with 2 currently awaiting repairs. No patient with a paraconduit hernia experienced an acute complication that required emergency intervention.

CONCLUSIONS

The risk of paraconduit hernia increases with time, suggesting that long-term symptom surveillance is reasonable. Emergency complications as a result of asymptomatic paraconduit hernias are rare. A small number of patients will experience hernia-related symptoms, sometimes years after hernia diagnosis. Our findings suggest that observation of asymptomatic paraconduit hernias (watch and wait) may be considered, with repair considered electively in patients with persistent symptoms.

Infective Endocarditis in Patients Addicted to Injected Opioid Drugs

BACKGROUND

Persons who inject drugs and require surgery for infective endocarditis have 2 potentially lethal diseases. Current postoperative rehabilitation efforts seem ineffective in preventing loss to follow-up, injection drug use relapse (relapse), and death.

OBJECTIVES

The purpose of this study was to characterize drug use, psychosocial issues, surgical outcome, and postoperative addiction management, as well as loss to follow-up, relapse, and mortality and their risk factors.

METHODS

From January 2010 to June 2020, 227 persons who inject drugs, age 36 ± 9.9 years, underwent surgery for infective endocarditis at a quaternary hospital having special interest in developing addiction management programs. Postsurgery loss to follow-up, relapse, and death were assessed as competing risks and risk factors identified parametrically and by machine learning. CIs are 68% (±1 SE).

RESULTS

Heroin was the most self-reported drug injected (n = 183 [81%]). Psychosocial issues included homelessness (n = 56 [25%]), justice system involvement (n = 150 [66%]), depression (n = 118 [52%]), anxiety (n = 104 [46%]), and post-traumatic stress disorder (n = 33 [15%]). Four (1.8%) died in-hospital. Medication for opioid use disorder prescribed at discharge increased from 0% in 2010 to 100% in 2020. At 1 and 5 years, conditional probabilities of loss to follow-up were 16% (68% CI: 13%-22%) and 59% (68% CI: 44%-65%), relapse 32% (68% CI: 28%-34%) and 79% (68% CI: 74%-83%), and mortality 21% (68% CI: 18%-23%) and 68% (68% CI: 62%-72%). Younger age, heroin use, and lower education level were predictors of relapse.

CONCLUSIONS

Infective endocarditis surgery can be performed with low mortality in persons who inject drugs, but addiction is far more lethal. Risk of loss to follow-up and relapse require more effective addiction strategies without which this major loss to society will continue.

Well-functioning bicuspid aortic valves should be preserved during aortic replacement for the ascending aortopathy phenotype

OBJECTIVES

Consensus has not been reached on whether or not to replace or preserve a well-functioning bicuspid aortic valve (BAV) in patients undergoing aortic replacement for the ascending phenotype of BAV aortopathy. We characterize morphology, evaluate progression of aortic regurgitation or aortic stenosis, and investigate the need for aortic valve replacement in patients whose well-functioning BAV was preserved during ascending aortic replacement ≥10 years prior.

METHODS

From January 1991 to August 2011, 191 patients with a well-functioning BAV underwent supracoronary aortic replacement (113 valves were minimally repaired). Aortic morphology was evaluated, aortic regurgitation grade and transvalvular aortic gradient modeled parametrically, and survival assessed by the Kaplan-Meier method. Median follow-up was 10 years.

RESULTS

Mean aortic diameter was 2.9 ± 0.53 cm at the annulus and 4.2 ± 0.55 cm at the sinuses. Mean maximum ascending diameter was 5.1 ± 0.49 cm. All patients exhibited a cusp-fusion BAV phenotype. Fifteen-year progression to severe aortic regurgitation was 3.2%. Mean aortic valve gradient began to rise 5 years postoperatively to 27 mm Hg by 14 years. Freedom from aortic valve replacement at 1, 5, 10, and 15 years was 100%, 95%, 83%, and 63%, respectively. Minimal valve repair was not associated with late aortic valve replacement. Fifteen-year survival was 74%.

CONCLUSIONS

Preserving a well-functioning BAV should be considered in carefully selected patients undergoing aortic replacement for the ascending phenotype of BAV aortopathy. The valves remain durable in the long term, with slow progression of regurgitation or stenosis, and low probability of aortic valve replacement through 10 years.

Isolated mitral valve endocarditis: Patient, disease, and surgical factors that influence outcomes

OBJECTIVE

The objectives of this study were to investigate patient characteristics, valve pathology, bacteriology, and surgical techniques related to outcome of patients who underwent surgery for isolated native (NVE) or prosthetic (PVE) mitral valve endocarditis.

METHODS

From January 2002 to January 2020, 447 isolated mitral endocarditis operations were performed, 326 for NVE and 121 for PVE. Multivariable analysis of time-related outcomes used random forest machine learning.

RESULTS

Staphylococcus aureus was the most common causative organism. Of 326 patients with NVE, 88 (27%) underwent standard mitral valve repair, 43 (13%) extended repair, and 195 (60%) valve replacement. Compared with NVE with standard repair, patients who underwent all other operations were older, had more comorbidities, worse cardiac function, and more invasive disease. Hospital mortality was 3.8% (n = 17); 0 (0%) after standard valve repair, 3 (7.0%) after extended repair, 8 (4.1%) after NVE replacement, and 6 (5.0%) after PVE re-replacement. Survival at 1, 5, and 10 years was 91%, 75%, and 62% after any repair and 86%, 62%, and 44% after replacement, respectively. The most important risk factor for mortality was renal failure. Risk-adjusted outcomes, including survival, were similar in all groups. Unadjusted extended repair outcomes, particularly early, were similar or worse than replacement in terms of reinfection, reintervention, regurgitation, gradient, and survival.

CONCLUSIONS

A patient- and pathology-tailored approach to surgery for isolated mitral valve endocarditis has low mortality and excellent results. Apparent superiority of standard valve repair is related to patient characteristics and pathology. Renal failure is the most powerful risk factor. In case of extensive destruction, extended repair shows no benefit over replacement.

Aortic allograft infection risk

OBJECTIVE

Intrinsic risk of infection of cryopreserved allograft aortic root replacements remains poorly understood despite their long history of use. The objective of this study was to determine this intrinsic risk of allograft infection and its risk factors when allografts are implanted for both nonendocarditis indications and infective endocarditis.

METHODS

From January 1987 to January 2017, 2042 patients received 2110 allograft aortic valves at a quaternary medical center, 1124 (53%) for nonendocarditis indications and 986 (47%) for endocarditis indications (670 [68%] prosthetic valve endocarditis). Staphylococcus aureus caused 193 of 949 cases of endocarditis (20%), 71 (7.3%) in persons who injected drugs. Periodic surveillance and cross-sectional follow-up achieved 85% of possible follow-up time. The primary end point was allograft infection in patients with nonendocarditis and endocarditis indications. Risk factors were identified by hazard function decomposition and machine learning.

RESULTS

During follow-up, 30 allografts (26 explanted) became infected in patients in the nonendocarditis group and 49 (41 explanted) in patients with endocarditis. At 20 years, the probability of allograft infection was 5.6% in patients in the nonendocarditis group and 14% in patients with endocarditis. Risk factors for allograft infection in patients in the nonendocarditis group were younger patient age and older donor age. Risk factors for allograft infection in patients with endocarditis were earlier implant year, injection drug use, and younger age. In patients with endocarditis, 18% of allograft infections were caused by the original organism.

CONCLUSIONS

The low infection rates, both in patients without and with endocarditis, support continued use of allografts in the modern era, in particular for the treatment of invasive endocarditis of the aortic root.

Aortic Valve Reconstruction With Autologous Pericardium Versus a Bioprosthesis: The Ozaki Procedure in Perspective

Background We assessed the Ozaki procedure, aortic valve reconstruction using autologous pericardium, with respect to its learning curve, hemodynamic performance, and durability compared with a stented bioprosthesis. Methods and Results From January 2007 to January 2016, 776 patients underwent an Ozaki procedure at Toho University Ohashi Medical Center. Learning curves, aortic regurgitation (AR), and peak gradient, assessed by serial echocardiograms, valve rereplacement, and survival were investigated. Valve performance and durability were compared with 627 1:1 propensity-matched patients receiving stented bovine pericardial valves implanted from 1982 to 2011 at Cleveland Clinic. Learning curves were observed for aortic clamp and cardiopulmonary bypass times, AR prevalence, and early mortality. Decreased aortic clamp time was observed over the first 300 cases. New surgeons performing parts of the procedure after case 400 resulted in a slight increase in aortic clamp and cardiopulmonary bypass times. Among matched patients, the Ozaki cohort had more AR than the PERIMOUNT cohort (severe AR at 1 and 6 years, 0.58% and 3.6% versus 0.45% and 1.0%, respectively; P[trend]=0.006), although with a steep learning curve. Peak gradient showed the opposite trend: 14 and 17 mm Hg for Ozaki and 24 and 28 mm Hg for PERIMOUNT at these times (P[trend]<0.001). Freedom from rereplacement was similar (P=0.491). Survival of the Ozaki cohort was 85% at 6 years. Conclusions Patients undergoing the Ozaki procedure had lower gradients but more recurrent AR than those receiving PERIMOUNT bioprostheses. Although recurrent AR is concerning, results confirm low risk and good midterm performance of the Ozaki procedure, supporting its continued use.

Time-related risk of pulmonary conduit re-replacement: a Congenital Heart Surgeons' Society Study

BACKGROUND

Patients receiving a right ventricle-to-pulmonary artery conduit in infancy will require successive procedures or replacements, each with variable longevity. We sought to identify factors associated with time-related risk of a subsequent surgical replacement (PC3) or transcatheter pulmonary valve insertion (TPVI) after a second surgically-placed PC (PC2).

METHODS

From 2002 to 2016, 630 patients from 29 Congenital Heart Surgeons' Society member institutions survived to discharge after initial valved PC insertion (PC1) at age < 2 years. Of those, 355 had undergone surgical replacement (PC2) of that initial conduit. Competing risk methodology and multiphase parametric hazard analyses were used to identify factors associated with time-related risk of PC3 or TPVI.

RESULTS

Of 355 PC2 patients (median follow-up of 5.3 years), 65 underwent PC3 and 41 TPVI. Factors at PC2 associated with increased time-related risk of PC3 were smaller PC2 Z score (Hazard Ratio [HR] 1.6, p<0.001), concomitant aortic valve intervention (HR 7.6, p=0.009), aortic allograft (HR 2.2, p=0.008), younger age (HR 1.4, p<0.001), and larger Z score of PC1 (HR 1.2, p=0.04). Factors at PC2 associated with increased time-related risk of TPVI were aortic allograft (HR: 3.3, p=0.006), porcine unstented conduit (HR 4.7, p<0.001), and older age (HR 2.3, p=0.01).

CONCLUSIONS

Aortic allograft as PC2 was associated with increased time-related risk of both PC3 and TPVI. Surgeons may reduce risk of these subsequent procedures by not selecting an aortic homograft at PC2, and by oversizing the conduit when anatomically feasible.

Factors associated with mortality or transplantation versus Fontan completion after cavopulmonary shunt for patients with tricuspid atresia

OBJECTIVE

Tricuspid atresia with normally related great vessels (TA) is considered the optimal substrate for the Fontan pathway. The factors associated with death or transplantation after cavopulmonary shunt (CPS) are underappreciated. We aimed to determine factors associated with CPS-Fontan interstage death/transplantation versus transition to Fontan in TA.

METHODS

A total of 417 infants younger than 3 months of age with TA were enrolled (January 1999 to February 2020) from 40 institutions into the Congenital Heart Surgeons' Society TA cohort. Parametric competing risk methodology was used to determine factors associated with the competing end points of death/transplantation without Fontan completion, and transition to Fontan.

RESULTS

CPS was performed in 382 patients with TA; of those, 5% died or underwent transplantation without transition to Fontan and 91% transitioned to Fontan by 5 years after CPS. Prenatal diagnosis (hazard ratio [HR], 0.74; P < .001) and pulmonary artery band (PAB) at CPS (HR, 0.50; P < .001) were negatively associated with Fontan completion. Preoperative moderate or greater mitral valve regurgitation (HR, 3.0; P < .001), concomitant mitral valve repair (HR, 11.0; P < .001), PAB at CPS (HR, 3.0; P < .001), postoperative superior vena cava interventions (HR, 9.0; P < .001), and CPS takedown (HR, 40.0; P < .001) were associated with death/transplantation.

CONCLUSIONS

The mortality rate after CPS in patients with TA is notable. Those with preoperative mitral valve regurgitation remain a high-risk group. PAB at the time of CPS being associated with both increased risk of death and decreased Fontan completion may represent a deleterious effect of antegrade pulmonary blood flow in the CPS circulation.

Valve Academic Research Consortium 3: updated endpoint definitions for aortic valve clinical research

Aims 

The Valve Academic Research Consortium (VARC), founded in 2010, was intended to (i) identify appropriate clinical endpoints and (ii) standardize definitions of these endpoints for transcatheter and surgical aortic valve clinical trials. Rapid evolution of the field, including the emergence of new complications, expanding clinical indications, and novel therapy strategies have mandated further refinement and expansion of these definitions to ensure clinical relevance. This document provides an update of the most appropriate clinical endpoint definitions to be used in the conduct of transcatheter and surgical aortic valve clinical research.

Methods and results 

Several years after the publication of the VARC-2 manuscript, an in-person meeting was held involving over 50 independent clinical experts representing several professional societies, academic research organizations, the US Food and Drug Administration (FDA), and industry representatives to (i) evaluate utilization of VARC endpoint definitions in clinical research, (ii) discuss the scope of this focused update, and (iii) review and revise specific clinical endpoint definitions. A writing committee of independent experts was convened and subsequently met to further address outstanding issues. There were ongoing discussions with FDA and many experts to develop a new classification schema for bioprosthetic valve dysfunction and failure. Overall, this multi-disciplinary process has resulted in important recommendations for data reporting, clinical research methods, and updated endpoint definitions. New definitions or modifications of existing definitions are being proposed for repeat hospitalizations, access site-related complications, bleeding events, conduction disturbances, cardiac structural complications, and bioprosthetic valve dysfunction and failure (including valve leaflet thickening and thrombosis). A more granular 5-class grading scheme for paravalvular regurgitation (PVR) is being proposed to help refine the assessment of PVR. Finally, more specific recommendations on quality-of-life assessments have been included, which have been targeted to specific clinical study designs.

Conclusions 

Acknowledging the dynamic and evolving nature of less-invasive aortic valve therapies, further refinements of clinical research processes are required. The adoption of these updated and newly proposed VARC-3 endpoints and definitions will ensure homogenous event reporting, accurate adjudication, and appropriate comparisons of clinical research studies involving devices and new therapeutic strategies.

Valve Academic Research Consortium 3: Updated Endpoint Definitions for Aortic Valve Clinical Research

AIMS

The Valve Academic Research Consortium (VARC), founded in 2010, was intended to (i) identify appropriate clinical endpoints and (ii) standardize definitions of these endpoints for transcatheter and surgical aortic valve clinical trials. Rapid evolution of the field, including the emergence of new complications, expanding clinical indications, and novel therapy strategies have mandated further refinement and expansion of these definitions to ensure clinical relevance. This document provides an update of the most appropriate clinical endpoint definitions to be used in the conduct of transcatheter and surgical aortic valve clinical research.

METHODS AND RESULTS

Several years after the publication of the VARC-2 manuscript, an in-person meeting was held involving over 50 independent clinical experts representing several professional societies, academic research organizations, the US Food and Drug Administration (FDA), and industry representatives to (i) evaluate utilization of VARC endpoint definitions in clinical research, (ii) discuss the scope of this focused update, and (iii) review and revise specific clinical endpoint definitions. A writing committee of independent experts was convened and subsequently met to further address outstanding issues. There were ongoing discussions with FDA and many experts to develop a new classification schema for bioprosthetic valve dysfunction and failure. Overall, this multi-disciplinary process has resulted in important recommendations for data reporting, clinical research methods, and updated endpoint definitions. New definitions or modifications of existing definitions are being proposed for repeat hospitalizations, access site-related complications, bleeding events, conduction disturbances, cardiac structural complications, and bioprosthetic valve dysfunction and failure (including valve leaflet thickening and thrombosis). A more granular 5-class grading scheme for paravalvular regurgitation (PVR) is being proposed to help refine the assessment of PVR. Finally, more specific recommendations on quality-of-life assessments have been included, which have been targeted to specific clinical study designs.

CONCLUSIONS

Acknowledging the dynamic and evolving nature of less-invasive aortic valve therapies, further refinements of clinical research processes are required. The adoption of these updated and newly proposed VARC-3 endpoints and definitions will ensure homogenous event reporting, accurate adjudication, and appropriate comparisons of clinical research studies involving devices and new therapeutic strategies.

Competing risks analysis of time-to-event data for cardiovascular surgeons

OBJECTIVES

The purpose of this article is to provide thoracic and cardiovascular surgeons with a practical step-by-step strategy to use in collaboration with a biostatistician for implementation of competing risks analysis when analyzing time-to-event data. Patients may have an outside event that precludes the event of interest. Traditional time-to-event analysis incorrectly assumes noninformative censoring in this scenario, which will lead to invalid results and conclusions.

METHODS

The steps are (1) to determine whether competing risks analysis is needed, (2) to perform a nonparametric analysis, (3) to perform a model-based analysis, (4) to interpret the results, and (5) to compare to traditional survival analysis methods. We apply our approach to a hypothetical cardiovascular surgery example in determining the hazard of mortality after the stage 3 Fontan operation associated with prematurity among patients with hypoplastic left heart syndrome who had successful completion of Norwood stage 1 while incorporating mortality during the stage 2 bidirectional Glenn procedure as a competing risk. We apply nonparametric, semiparametric, and parametric methods.

RESULTS

Although Cox regression establishes prematurity as a significant risk factor of mortality after stage 3 (hazard ratio, 1.26; 95% confidence interval, 1.06-1.50; P = .009), the competing risks analysis with the Fine-Gray model accounting for mortality after stage 2 determines that prematurity is not a significant predictor (hazard ratio, 1.07; 95% confidence interval, 0.90-1.27; P = .467).

CONCLUSIONS

This article provides a practical step-by-step approach for making competing risks more accessible for cardiac surgeons collaborating with a biostatistician in analyzing and interpreting time-to-event data.