Influence of Coronary Artery Bypass and Age on Clinical Performance After Aortic and Mitral Valve Replacement With Biological and Mechanical Prostheses
Abstract The influence of prosthetic type, age, and coronary artery bypass grafting (CABG) on valve-related complications by valve position was evaluated in a population of 2353 bioprosthesis patients (mean age, 66.5 years; range, 13 to 89 years) and in a population of 1112 mechanical prosthesis patients (mean age, 59.1 years; range, 13 to 91 years). The follow-up was complete to 96% and 98%, respectively, for the bioprosthesis and mechanical prosthesis groups. The patient groups were evaluated by actuarial assessment of survival and valve complications and composites. Preoperative factors were evaluated for determination of significant independent predictors by multivariate proportional-hazard regression analysis. CABG was an influential factor in the actuarial analysis. Survival was superior for aortic mechanical replacements without CABG and for mitral replacements, both biological and mechanical, without CABG (P<.05). The freedom from thromboembolism (TE) and antithromboembolic hemorrhage (ATH) was greater for biological prostheses with and without CABG for aortic replacements (P<.05) but not for mitral replacements (P=NS). The freedom from valve-related mortality was not influenced by CABG for either position (P=NS). The freedom from valve-related reoperation was greater for biological prostheses with CABG than without CABG for both aortic and mitral replacements (P<.05). The evaluation of covariates as independent predictors revealed CABG to be a nonpredictor for aortic valve replacement (AVR) (P=NS) but a predictor of survival and valve-related reoperation for mitral valve replacement (MVR) (P<.05). With AVR, biological prostheses predicted greater freedom from valve-related mortality and TE and ATH (P<.05) and for MVR for TE and ATH (P<.05). Valve position was not a predictor otherwise (P=NS). Advancing age was a significant covariate for AVR for all areas of assessment. Advancing age for MVR provided a lower rate of survival and greater freedom from valve-related mortality and reoperation (P<.05). The clinical performance of valvular prostheses is influenced to a greater degree by valve type and age than by CABG.
There has been a choice between BPs and MPs as valvular substitutes for cardiac valve replacement surgery for more than 20 years.1 The past 2 decades have seen major advancements in valvular designs and materials for MPs and tissue preservation and stent designs for BPs. Major TE, thrombosis, and ATH remain the predominant complications of MPs, while structural valve deterioration necessitating reoperation is the major complication of BPs. Cardiac surgeons, cardiologists, and patients must make a suitable choice between prosthesis types that will optimize quality of life and minimize the risk of permanent impairment (residual morbidity) and the potential for reoperation. BPs provide the greatest opportunity to optimize quality of life and avert long-term care for anticoagulant management. BPs have been identified as the prostheses of choice for the elderly, especially for AVR for those more than 65 years of age. Several investigators, including us,2 3 4 5 6 have identified the influence of age on structural failure of BPs and necessity for reoperation. There has been no documentation on other patient-related variables or confounders, either singly or in combination, that affect life expectancy and provide the opportunity for use of BPs.
The purpose of this article is to evaluate the clinical performance of BPs and MPs when age and CABG are considered confounders.
From 1982 to 1993, 2353 patients received the Carpentier-Edwards supra-annular porcine BP, and these patients constitute the BP population. The population implanted with MPs during the same time frame comprises 1112 patients who received St Jude Medical, Carbomedics, Duromedics, and Björk-Shiley Monostrut MPs. The mean age of the BP group is 66.5 years (range, 13 to 89 years) and of the MP group, 59.1 years (range, 13 to 91 years). The total follow-up for the BP group was 11 810 patient-years and for the MP group, 3275 patient-years. The follow-up was complete to 96.3% and 98.4%, respectively, for the BP and MP groups.
The patient population is detailed in Table 1⇓. The BP group consisted of 1378 AVRs with 502 (36.4%) CABG procedures and 975 MVRs with 298 (30.6%) CABGs. The MP group comprised 551 AVRs with 114 (20.7%) CABGs and 561 MVRs with 107 (19.1%) CABGs. Each patient population by valve replacement and with or without CABG was subdivided, as documented, into age groups of <60 years, 60 to 69 years, and ≥70 years.
The standard guidelines and definitions of terms of the Society of Thoracic Surgeons have been used for formulation of the valve-related complications and composites.7 The survival and freedom curves for complications and composites of complications were calculated by the actuarial method of Cutler-Ederer and illustrated in yearly intervals with the number of operations entering at chosen intervals, the cumulative proportions surviving or remaining, and standard errors. The Lee-Desu statistic was used to compare the survival and freedom curves. Multivariate proportional-hazard regression analysis (Cox model) was applied to assess which preoperative factors were significant independent predictors of survival and other valve-related complications and composites of valve-related complications (BMDP programs 2L, University of California, Los Angeles). The examined covariates were the following: age groupings (<60, 60 to 69, ≥70 years), CABG (yes/no), and valve type (mechanical/biological). A two-tailed value of P<.05 was considered statistically significant for the study.
Influence of CABG
The influence of CABG and age on patient survival by valve type and valve position is shown in Figs 1⇓ and 2⇓. For AVR the overall survival for MPs without CABG was 78.8±4.2% at 8 years after replacement, greater than for BPs without CABG (69.8±2.0% at 8 years), and MPs with CABG (69.3±6.8% at 5 years) (P<.05). For MVR, the overall survival was for MPs without CABG, 74.2±4.1% at 8 years and for BP without CABG, 63.8±2.1% at 8 years (P<.05). The survival for BP without CABG was greater than for BP with CABG (45.6±4.1% at 5 years) (P<.05).
The freedom from TE and ATH for AVR and MVR is illustrated in Figs 3⇓ and 4⇓. For AVR, the freedom from TE and ATH for BP without CABG was 84.6±1.7% and for MP without CABG, 71.7±5.3%, at 8 years after replacement (P<.05). The freedom for AVR was 83.5±2.5% for BP with CABG and 70.1±8.7% for MP with CABG, at 8 years (P<.05). CABG and valve type did not influence freedom from TE and ATH for MVR (P=NS).
Valve-related mortality was not influenced by CABG for AVR and MVR, for the overall population (P=NS) (Figs 5⇓ and 6⇓). Figs 7⇓ and 8⇓ illustrate the freedom from valve-related reoperation. CABG influenced the freedom for AVR and MVR. For AVR, the freedom from valve-related reoperation was 96.8±1.4% for BP with CABG and 91.0±1.4% for BP without CABG at 8 years (P<.05). For MVR, the freedom was 88.0±3.4% for BP with CABG and 83.4±2.1% for BP without CABG (P<.05) at 8 years.
Influence of Prosthesis Type
The influence of prosthesis type on patient survival by valve position is shown in Figs 1⇑ and 2⇑. For AVR, the survival for MP without CABG was greater than BP without CABG (P<.05), and for MVR the same relationship applied for replacement without CABG (P<.05). The relationship did not exist for replacement with CABG (P=NS).
The freedom from TE and ATH for AVR and MVR is shown in Figs 3⇑ and 4⇑. In MVR, the prosthesis type had no influence on TE and ATH. For AVR, freedom from BP without CABG was greater than MP without CABG (P<.05) and greater for BP with CABG over BP without CABG (P<.05).
The prosthesis type had no influence on freedom from valve-related mortality or reoperation for either AVR or MVR (P=NS).
Influence of Age (Groups)
Patient survival is influenced significantly for all patient groups except for the mechanical AVR with CABG, mechanical MVR without CABG, and mechanical MVR with CABG groups.
The BP AVR without CABG had the only significant findings, by age groups, for total and major TE and ATH (Table 2⇑). The age groups had no influence in the MVRs (P=NS).
The freedom from valve-related reoperation is also presented in Tables 2⇑ and 3⇑. The freedom for AVR with BPs without CABG was more significant (P<.05) for patients 60 to 69 years of age than those <60 years of age, and for those ≥70 years of age than those 60 to 69 years of age. For BP with CABG for AVR, significance for ≥70 years of age was greater than for <60 years of age, and for MP with CABG, significance for 60 to 69 years of age was greater than <60 years of age (P<.05). The freedom for MVR was affected by age groups only for the BP without CABG ≥70 years than <60 years (P<.05).
Covariate Assessment (Proportional-Hazard Regression)
The covariates, namely, CABG, valve type, and age groups (<60, 60 to 69, and ≥70 years) were evaluated for the aortic position (Table 4⇓) and for the mitral position (Table 5⇓) by the multivariate proportional-hazard regression analysis.
In the aortic position, CABG had no influence on survival or valve-related complications or composites of valve-related complications (P=NS). The prosthesis type had influence on freedom from valve-related mortality and TE and ATH. The freedom from valve-related mortality was greater for the BP group (P<<.05) and from TE and ATH was greater for the BP group (P<<.05). The freedom from major TE and ATH was lower with BPs but was not significant (P=NS).
Age was a significant covariate for AVR for all areas of assessment. Survival was lower with advancing age (P<<.05). The freedom from TE and ATH was less with older age as was freedom from valve-related mortality (P<<.05). The freedom from reoperation was greater with advancing years (P<<.05).
The assessment by the multivariate proportional-hazard regression analysis was different for the mitral position. CABG reduced survival (P<.05) and created greater freedom from reoperation (P<.05). The BPs had greater freedom from TE and ATH than MPs (P<.05).
Patients of advancing years with prostheses placed in the mitral position had a lower survival rate (P<<.05), greater freedom from valve-related mortality (P<<.05), and greater freedom from valve-related reoperation (P<<.05). The freedom from TE and ATH was less with advancing years but not significant (P=NS).
Cardiovascular surgeons, cardiologists, and patients continue to have the choice of BPs or MPs as valvular substitutes. The choice of prosthesis for the individual patient has depended on the clinical status of the patient, the confidence of the surgeon in the prosthesis, and the risk of valve-related complications. The literature has provided extensive documentation on individual prostheses and combination of prostheses and nonrandomized assessments of BPs and MPs, but few randomized clinical trials have been reported. The randomized clinical trials have included the Veterans Administration Study on Valvular Heart Disease and the Edinburgh Heart Valve Trial.8 9 10
The randomized trials have provided several conclusions, the majority known from nonrandomized studies. Bleeding complications from anticoagulation are predominant in the mechanical valve populations. The prevalence rates of TE, thrombosis, and prosthetic valve endocarditis are the same for MPs and BPs. Reoperations are necessary for structural failure of BPs and paravalvular leak of MPs. Porcine BPs fail more frequently in the mitral position than in the aortic position 5 or more years after implantation. The increased risk of reoperation with BPs appeared to be a high price to pay for the reduced risk of bleeding afforded by the avoidance of anticoagulation. The freedom from death, reoperation, major embolism, and endocarditis was less with porcine BPs, especially in the mitral position.
The influence of patient-related variables on valve-related complications to facilitate the choice of BPs versus MPs has received limited attention. Age as a patient-related variable has been evaluated by a limited number of investigators.2 3 4 5 6 The results of these reports have delineated the indications that numerous centers follow for the implantation of BPs and MPs. BPs are indicated for AVR for patients ≥65 years of age and for MVR for patients ≥70 years of age who do not require anticoagulation, as well as women during child-bearing years or patients with contraindications to anticoagulation. MPs are considered for patients categorized as younger than those designated for BPs. The purpose of this article is to consider the influence of both age and concomitant myocardial revascularization on patient survival and significant valve-related complications and composites of these complications, namely, valve-related mortality, residual morbidity, and reoperation.
The documented influence of age will initially be reviewed. In 1988, Jamieson and colleagues3 demonstrated the advancing freedom from structural valve deterioration for both AVR and MVR with each decade of life at implantation. The study also demonstrated the greater freedom from structural failure with AVR than MVR. Jamieson and coauthors2 in 1991 again reported that age was a predictor of structural valve deterioration, but that structural failure had little influence on valve-related mortality and residual morbidity. In the same year, the authors4 identified mitral position as well as age as predictors of structural valve deterioration.
The results of BP use in the elderly have been reported since 1988.11 12 13 The detailed report on structural valve deterioration by Burr and colleagues13 summarizes the overall consideration. The long-term freedom from structural failure for AVR with the Carpentier-Edwards porcine BP was 98% for age group 65 to 69 years at 15 years after replacement, 95% for 70 to 79 years at 13 years, and 100% for ≥80 years at 13 years. For MVR, the freedom was 85% at 7 years for the age group 65 to 69 years, 95% at 7 years for 70 to 79 years, and 100% at 6 years for patients ≥80 years. We generally concluded that if structural failure was in progress it did not advance so as to necessitate reoperation or contribute to mortality. Pelletier and coinvestigators5 reported that for the age group 65 to 80 years, the freedom from structural valve deterioration was 93% and the freedom from reoperation was 92% at 12 years with the Carpentier-Edwards porcine BP. These authors concluded that age was a determinant of durability of BPs but not of the overall success of the management.
The present study evaluates the clinical performance of BPs and MPs to 8 years after AVR or MVR with consideration of age groups (<60, 60 to 69, ≥70 years) and concomitant CABG. The study was performed to evaluate the hypothesis that since coronary artery disease reduced life expectancy, the indications for BPs could be extended to younger age groups with the valve-related mortality and residual morbidity and reoperation less than for MPs.
CABG compromised survival with mechanical AVRs from 79% at 8 years after replacement to 69% at 5 years. The same relationship existed for BP MVR; the survival without CABG was 64% and with CABG was 46%. Concomitant CABG extended greater freedom from reoperation for AVR and MVR BPs at 8 years. The freedom from reoperation for AVR, at 8 years, was 97% for BPs with CABG and 91% for BPs without CABG. The freedom from reoperation for MVR, at 8 years, was 88% for BPs with CABG and 83% for BPs without CABG. For AVR with CABG, the freedom from reoperation, at 8 years, was 97% for BPs and 86% for MPs. These observations generally concur that concomitant CABG compromises survival and translates into greater freedom from reoperation over MPs especially in the aortic position. Reoperations with MPs are generally early and due to periprosthetic leak and prosthetic valve endocarditis, while reoperations for BPs occur generally after 8 years, for structural valve deterioration.
BPs provide greater freedom from TE and ATH for AVR with and without CABG. For AVR, the freedom was 85% for BPs without CABG and 72% for MPs without CABG at 8 years, and with CABG, 84% and 70%, respectively. The freedom from major TE and ATH and residual morbidity for BPs was 96% and for MPs was 94%, at 8 years, for major TE plus hemorrhagic complications of anticoagulants, and 99% and 96% for residual morbidity or permanent impairment.
The evaluation of the covariates by multivariate proportional-hazard regression analysis provided further refinement from the actuarial analysis of the influence on clinical performance. The influence of CABG, valve type, and age groupings was assessed individually on survival, valve-related complications, and composites of valve-related complications.
For AVR, survival was influenced only by advancing age and not by coronary artery disease and valve type as was illustrated by the actuarial analysis. Valve-related mortality was influenced by valve type and age, with greater freedom for BPs and lesser freedom for advancing years. The freedom from reoperation was influenced only by advancing age. The freedom from TE and ATH was greater for BPs and less for advancing age. The freedom from major TE and ATH was influenced negatively by advancing age. The freedom from residual morbidity (permanent impairment) followed freedom from major TE and ATH.
For MVR, survival was influenced negatively by CABG and advancing age. The freedom from valve-related mortality was influenced only by age with less freedom from advancing age. The freedom from reoperation was greater for MVR in the presence of CABG. Advancing age also increased the freedom from reoperation. The freedom from TE and ATH was greater for BPs. The freedom from major TE and ATH and residual morbidity was not influenced by coronary artery disease, valve type, or age groups.
In summary, this study reveals that, at 8 years after valve replacement by actuarial analysis, CABG compromises survival with AVR MPs and MVR BPs while providing greater freedom from reoperation for AVR and MVR BPs. Also by actuarial analysis, AVR with CABG provides greater freedom from reoperation with BPs than MPs. On the contrary, by regressive analysis, CABG did not influence AVR, survival, or freedom from reoperation but did so for MVR. Age was the influencing factor on survival and freedom from reoperation for both positions.
This study of patient-related variables on valve-related complications for comparison of utilization of BPs and MPs must be extended from 8 years to 12 and 15 years. The University of British Columbia has BP experience to 15 and 18 years but not equivalent MP experience.
Selected Abbreviations and Acronyms
|AVR||=||aortic valve replacement|
|CABG||=||coronary artery bypass grafting|
|MVR||=||mitral valve replacement|
Presented in part at the 67th Scientific Sessions of the American Heart Association, Dallas, Tex, November 14-17, 1994, and published in abstract form (Circulation. 1994;90[pt 2]).
- Copyright © 1995 by American Heart Association
Jamieson WRE, Tyers GFO, Miyagishima RT, Janusz MT, Ling H. Carpentier-Edwards porcine bioprostheses: comparison of standard and supra-annular prostheses at seven years. Circulation. 1991;84(suppl III):III-145-III-152.
Pelletier LC, Carrier M, Leclerc Y, Dyrda I, Gosselin G. Influence of age on late results of valve replacement with porcine bioprostheses. J Cardiovasc Surg (Torino). 1992;33:526-533.
Hammermeister KE, Henderson WG, Burchfiel CM, Sethi GK, Souchek J, Oprian C, Canton AB, Folland E, Khuri S, Rahimtoola S. Comparison of outcome after valve replacement with a bioprosthesis versus a mechanical prosthesis: initial 5 year results of a randomized trial. J Am Coll Cardiol. 1987;10:719-732.
Hammermeister KE, Sethi GK. Comparison of occurrence in bleeding systemic embolism, endocarditis valve thrombosis, and re-operation between patients randomized between a mechanical prosthesis and a bioprosthesis: results from a VA randomized trial. J Am Coll Cardiol. 1991;17:362A. Abstract.
Pupello DF, Bessone LN, Hero SP, Lopez-Cuenca E, Glatterer MS Jr, Ebra G. The Carpentier-Edwards bioprosthesis: a comparative study analyzing failure rates by age. J Cardiac Surg. 1988;3(suppl 3):369-374.
Burr LH, Jamieson WRE, Munro AI, Miyagishima RT, Janusz MT, Ling H, Hayden RI, Tutassaura H, Fradet G, Gudas VM, Tyers GFO. Structural valve deterioration in elderly patient populations with the Carpentier-Edwards standard and supra-annular porcine bioprostheses: a comparative study. J Heart Valve Dis. 1992;1:87-91.