Bicuspid aortic valve (BAV) is associated with ascending aortopathy predisposing to aneurysmal dilatation and dissection, even after successful aortic valve replacement (AVR). There is, however, scant evidence on which to make recommendations for prophylactic replacement of the ascending aorta at the time of AVR. The medical records of patients who underwent AVR for BAV without aortic replacement or repair from 1960 to 1995 were reviewed. Follow-up was by review of the medical record and postal questionnaire. Among 1,286 patients, the mean age at operation was 58 ± 14 years. During the follow-up interval (median 12 years, range 0 to 38), there were 13 documented aortic dissections (1%), 11 ascending aortic replacements (0.9%), and 127 documented cases of progressive aortic enlargement (9.9%). Fifteen-year freedom from aortic dissection, enlargement, or replacement was 89% (95% confidence interval [CI] 87% to 91%) and was lower in patients with documented aortic enlargement at the time of AVR (85%, 95% CI 81% to 89%) compared to those whose aortic dimensions were normal (93%, 95% CI 90% to 96%) (p = 0.001). Multivariate predictors of aortic complications included interval (subsequent) AVR (hazard ratio [HR] 3.5, 95% CI 2.3 to 5.4, p <0.001), concomitant coronary artery bypass grafting (HR 2.6, 95% CI 1.7 to 4.0, p <0.001), enlarged aorta (HR 1.8, 95% CI 1.3 to 2.6, p = 0.001), and history of tobacco abuse (HR 1.8, 95% CI 1.2 to 2.6, p = 0.003). Aortic dilatation did not predict mortality. In conclusion, despite a true risk for aortic events after AVR for BAV, the occurrence of aortic dissection was low. Any incremental surgical risk imposed by prophylactic replacement of the ascending aorta must be equally low.
It is now widely accepted that an inherent aortopathy exists independent of functional valvular pathology in many patients with bicuspid aortic valves (BAVs). The mean ascending aortic dimensions in patients with BAVs exceed those of patients with tricuspid aortic valves, and progressive enlargement or dilatation of the aorta, even after aortic valve replacement (AVR), has been documented. Because aortic dilatation is associated with an increased risk for aortic dissection and rupture, liberal prophylactic replacement of the ascending aorta at the time of AVR has been advocated, by some even in the presence of “a seemingly normal aorta.” BAV is the underlying pathology in as many as 1/3 of patients who undergo AVR in the Western world, however, any incremental operative risk imposed by such intervention will have a significant clinical impact. Despite these implications, there have been few previous studies concerning the risk for aortic complications after AVR for BAV and, accordingly, the indications for surgical repair. What data exist suggest risk stratification on the basis of aortic dimension at the time of AVR. Unfortunately, the limited size of the study populations, duration of follow-up, and number of aortic events lead investigators to use potentially misleading composite end points that render true risk assessment problematic. We therefore examined our entire institutional experience with AVR for BAV without aortic intervention from 1960 to 1995, with the aim of determining the long-term risk for aortic complications, including aortic dissection, rupture, and death, in patients with and without aortic enlargement.
Methods
Institutional review board approval from the Mayo Clinic for Education and Research was granted with an exception for obtaining study-specific consent. Using our institutional cardiovascular surgery database, which dates to the first case performed on cardiopulmonary bypass in 1955, we identified all patients who underwent AVR from 1960 to 1995, truncating the study to allow sufficient follow-up to observe late events. Demographic and clinical information as well as late follow-up, including survival data, was obtained via review of the medical records, including the surgical database and the Mayo registry, general follow-up postal surveys distributed by the division, and the Social Security Death Index.
Only patients with congenital BAVs as reported by the surgeon or pathologist were included. Patients with known connective tissue disorders, including Marfan syndrome and Ehlers-Danlos syndrome, were excluded. Those who underwent concomitant aortic root enlargement or replacement were also excluded, as were those having intervention on the ascending aorta either as ascending aortoplasty or graft replacement.
We examined outcomes for the entire population of 1,286 patients, and after stratification for aortic enlargement at the time of index AVR, because this parameter has been identified as a risk factor for aortic complications. Unfortunately, determining the presence of aortic enlargement is complicated by the evolution of imaging modalities over the course of the study interval. For patients who underwent more recent surgery, quantitative data were available from axial imaging studies, echocardiography, aortography, or intraoperative measurement. A diameter >40 mm was taken as a clinically relevant and practical cutoff from a surgical perspective, and therefore patients with aortic diameter exceeding this threshold were classified as “enlarged.” In those without such quantitative information, we sought narrative notation of aortic enlargement in the operative record. If no mention was made of aortic enlargement, the aorta was assumed to be of normal caliber. We performed analyses of the entire group, comparing the outcomes of those with “enlarged” and “normal” ascending aortas, and of the recent subset with quantitative information alone, comparing the outcomes of patients with aortic diameters <40 mm to those of patients with aortic diameters >40 mm.
To simplify the statistical analysis of risk factors for ascending aortic complications, a composite end point of “aortic events” was defined as documented aortic dissection, ascending aortic replacement, interval aortic dilatation to a diameter ≥50 mm, or documented growth of the aorta >10 mm after AVR. Late cardiovascular-related mortality was defined as death attributed to myocardial infarction, congestive heart failure, cardiac arrest, or aortic catastrophe. More recent surgery was defined as surgery after the median year of operation (1983), and interval AVR was defined as prosthetic aortic valve repeat replacement before, or on the same day as, an aortic event.
Descriptive statistics for categorical variables are reported as frequencies and percentages, while continuous variables are reported as mean ± SD or as medians and ranges, as appropriate. Categorical variables were compared between patients with and without enlarged aortas using chi-square tests, and continuous variables were compared using 2-sample Student’s t tests or Wilcoxon’s rank-sum test, as appropriate.
Aortic events, overall mortality, and cardiac mortality were analyzed respectively as the end points for this study. For each end point, Kaplan-Meier method was used to draw freedom-from-event curves and calculate 15-year freedom-from-event statistics. Risk factor analysis was carried out among a group of baseline variables, includes age, gender, surgery year, type of valve, ascending aorta status, endocarditis, rheumatic fever, hypertension, smoking, hypercholesterolemia, diabetes, previous intervention, indication for operation, and coronary artery bypass grafting (CABG) as a concomitant procedure. Cox regression models were used to find the univariate and multivariate predictors of events. The multivariate model considered variables significant (p <0.05) on univariate analysis, with model selection using the stepwise method (backward and forward methods resulted in the same model). For exploratory purposes, age, gender, and surgery year were forced into each multivariate model regardless of univariate analysis results to determine if they made a difference in the multivariate model. If any of the 3 variables became significant, the final multivariate model included the variable. The same statistical analysis strategy was used in the analysis for the patients in the whole group as well as in subgroups. Overall survival was also compared with age- and gender-matched Minnesota white population using a locally written SAS macro (SAS Institute Inc., Cary, North Carolina). From the expected survival curves, the observed survival and expected survival crossed at approximately 5 years; survival differences between the patient group and the age- and gender-matched general population were first compared at a follow-up of 5 years and then out to 15 years. All statistical tests were 2 sided, with the α level set at 0.05 for statistical significance. The software used for analysis was SAS version 9.13.
Results
We identified 1,286 patients with congenital BAVs who underwent AVR at our institution from 1960 to 1995 ( Figure 1 ). Of them, 507 (39%) were noted in the medical record to have aortic enlargement (n = 432) or had imaging studies demonstrating diameters >40 mm (n = 75). Of the remaining 779 patients, quantitative information regarding aortic dimensions was available in 248 (32%). Therefore, in total, quantitative measurements were available for 323 patients, representing only 25% of the total study population. Most of these patients underwent operations relatively recently (mean year of surgery 1990 vs 1983 for the entire group). In the interest of balancing the competing aims of maximizing the duration of follow-up at risk for events versus precision in aortic measurement, we conducted analyses of the entire cohort and of the subset with quantitative aortic data only.
As listed in Table 1 , the mean age of the entire study population of 1,286 patients was 58 years. As expected, most patients were male, and aortic stenosis was the predominant valve pathology. Clinical histories of hypertension, hypercholesterolemia, and tobacco use were common. The operative mortality rate for the index operation was 2.8%.
| Variable | All 1,286 Patients | 1,286 Patients | 323 Patients With Quantitative Measurements | ||||
|---|---|---|---|---|---|---|---|
| Normal (n = 779) | Enlarged (n = 507) | p Value | <40 mm (n = 248) | ≥40 mm (n = 75) | p Value | ||
| Age (years) | 58 ± 14 | 59 ± 14 | 56 ± 15 | <0.01 | 60 ± 15 | 62 ± 12 | 1.00 |
| Gender (male) | 932 (72%) | 550 (71%) | 382 (75%) | 0.06 | 166 (67%) | 61 (81%) | 0.02 |
| History of rheumatic fever | 162 (13%) | 91 (12%) | 71 (14%) | 0.22 | 32 (13%) | 14 (19%) | 0.21 |
| Indication for operation | |||||||
| Aortic stenosis | 989 (77%) | 625 (80%) | 364 (72%) | <0.01 | 214 (86%) | 62 (83%) | 0.07 |
| Aortic regurgitation | 85 (7%) | 38 (5%) | 47 (9%) | 6 (2%) | 4 (5%) | ||
| Both | 191 (15%) | 96 (12%) | 95 (19%) | 26 (10%) | 9 (12%) | ||
| Infective endocarditis | 60 (5%) | 31 (4%) | 29 (6%) | 0.15 | 5 (2%) | 4 (5%) | 0.22 |
| Previous cardiac surgery | 73 (6%) | 41 (5%) | 32 (6%) | 0.43 | 16 (6%) | 8 (11%) | 0.22 |
| History of hypertension | 444 (35%) | 270 (35%) | 174 (34%) | 0.90 | 90 (36%) | 30 (40%) | 0.56 |
| History of tobacco use | 696 (54%) | 414 (53%) | 282 (56%) | 0.38 | 141 (57%) | 46 (61%) | 0.49 |
| History of hypercholesterolemia | 486 (38%) | 282 (36%) | 203 (40%) | 0.24 | 79 (32%) | 37 (49%) | 0.02 |
| Coronary artery bypass grafting | 222 (17%) | 154 (20%) | 68 (13%) | <0.01 | 63 (25%) | 26 (35%) | 0.12 |
| Diabetes mellitus | 78 (6%) | 57 (7%) | 21 (4%) | 0.02 | 29 (12%) | 8 (11%) | 0.81 |
| Tissue bioprosthesis | 458 (36%) | 307 (39%) | 151 (30%) | <0.01 | 129 (52%) | 44 (59%) | 0.31 |
| Operative mortality | 2.80% | ||||||
Among the 1286 patients, 160 aortic events occurred in 130 patients during the median follow-up interval of 12 years (range 0 to 38), with 2% of patients lost to follow-up within 3 years after surgery. Specifically, there were 13 aortic dissections (1%), 11 aortic replacements (0.9%), and 127 instances of documented progressive aortic enlargement (9.9%). Freedom from aortic events was 89% (95% confidence interval [CI] 87% to 91%) at 15 years ( Figure 2 ). Overall survival was 52% (95% CI 49% to 55%) at 15 years; survival was no different from an age- and gender-matched Minnesota population at 5 years but poorer at 15 years (p <0.001; Figure 2 ). Freedom from cardiovascular mortality was 78% (95% CI 75% to 81%; Figure 2 ).
When the study group was stratified by aortic dimensions, those with enlargement more often had aortic regurgitation as the predominant functional pathology (9% vs 5%) or as a component of mixed disease (19% vs 12%) than those with normal aortas (p <0.001). Hypertension, tobacco abuse, and hypercholesterolemia were similar, but diabetes mellitus and coronary artery disease requiring CABG were more common in those with normal aortas. Tissue prostheses were also more commonly used in those patients with normal aortas, perhaps in part because of a slightly older mean age.
Two patients with normal aortas at the time of AVR underwent interval aortic replacement (0.3%), compared to 9 (1.8%) with enlarged aortas (p = 0.01). Subsequent aortic enlargement occurred in 51 patients (7%) with normal aortas, compared to 75 patients (15%) with enlarged aortas (p <0.001). Fifteen-year freedom from aortic events was 93% (95% CI 90% to 95%) in those with normal aortas compared to 85% (95% CI 81% to 89%) in those with enlarged aortas (p <0.001), with 15-year freedom from documented dissection of 99% (95% CI 99% to 100%) and 98% (95% CI 97% to 100%) for patients with normal aortas and those with enlarged aortas, respectively ( Figure 3 ). There was a slight difference in all-cause mortality between the 2 groups, as shown in Figure 3 , but it favored those with enlarged, not normal, aortas. Fifteen-year survival from all-cause mortality was 50% (95% CI 46% to 54%) in patients with normal aortas and 56% (95% CI 51% to 60%) in those with enlarged aortas (p = 0.02). This difference no longer existed in the multivariate model. There was no difference in cardiovascular mortality between the subgroups ( Figure 3 ).
As listed in Table 2 , risk factors for aortic events by univariate analysis included interval AVR (hazard ratio [HR] 3.2, 95% CI 2.1 to 4.9, p <0.001), concomitant CABG at initial AVR (HR 2.4, 95% CI 1.6 to 3.6, p <0.001), use of a bioprosthesis (HR 2.1, 95% CI 1.5 to 3.0, p <0.001), smoking (HR 2.0, 95% CI 1.4 to 2.8, p <0.001), an enlarged aorta at AVR (HR 1.8, 95% CI 1.3 to 2.6, p = 0.001), and more recent surgery (HR 1.6, 95% CI 1.1 to 2.3, p = 0.02). By multivariate analysis, interval AVR (HR 3.5, 95% CI 2.3 to 5.4, p <0.001), CABG at initial AVR (HR 2.6, 95% CI 1.7 to 4.0, p <0.001), aortic enlargement (HR 1.8, 95% CI 1.3 to 2.6) p = 0.001), and smoking (HR 1.8, 95% CI 1.2 to 2.6, p = 0.003) remained predictors of aortic events.
| Variable | HR | p Value | HR | p Value |
|---|---|---|---|---|
| Univariate analysis (n = 1,286) | Multivariate analysis (n = 1,286) | |||
| Interval AVR | 3.2 | <0.001 | 3.5 | <0.001 |
| Concomitant CABG | 2.4 | <0.001 | 2.6 | <0.001 |
| Tissue valve prosthesis | 2.1 | <0.001 | — | NS |
| Smoking history | 2.0 | <0.001 | 1.8 | 0.003 |
| Enlarged ascending aorta (≥40 mm) | 1.8 | 0.001 | 1.8 | 0.001 |
| Surgery after 1983 | 1.6 | 0.02 | — | NS |
| Female gender | 0.7 | 0.07 | — | NS |
| Univariate analysis (n = 323) | Multivariate analysis (n = 323) | |||
| Smoking history | 2.3 | 0.02 | 2.3 | 0.02 |
| Enlarged ascending aorta (≥40 mm) | 1.7 | NS | — | — |
| Concomitant CABG | 1.6 | NS | — | — |
| Subsequent AVR | 1.5 | NS | — | — |
| Tissue valve prosthesis | 1.1 | NS | — | — |
| Female gender | 0.7 | NS | — | — |
| Surgery after 1983 | — | — | — | — |
Predictors of all-cause mortality ( Table 3 ) by multivariate analysis included age >60 years at AVR (HR 2.5, 95% CI 2.1 to 2.9, p <0.001), diabetes mellitus (HR 2.1, 95% CI 1.6 to 2.7, p <0.001), CABG at AVR (HR 1.3, 95% CI 1.1 to 1.6, p = 0.003), and use of a tissue prosthesis (HR 1.3, 95% CI 1.1 to 1.5, p = 0.006). Two protective variables identified by multivariate analysis that were not present on univariate analysis were female gender (HR 0.8, 95% CI 0.7 to 1.0, p = 0.01) and surgery after 1983 (HR 0.6, 95% CI 0.5 to 0.8, p <0.001). Aortic size was not predictive of overall mortality.
| Variable | HR | p Value | HR | p Value |
|---|---|---|---|---|
| Univariate analysis (n = 1,286) | Multivariate analysis (n = 1,286) | |||
| Age ≥ 60 years | 2.4 | <0.001 | 2.5 | <0.001 |
| Diabetes mellitus | 2.5 | <0.001 | 2.1 | <0.001 |
| Concomitant CABG | 1.7 | <0.001 | 1.3 | 0.003 |
| Tissue valve prosthesis | 1.6 | <0.001 | 1.3 | 0.006 |
| Hypertension | 1.2 | 0.01 | — | NS |
| Enlarged ascending aorta (≥40 mm) | 0.9 | 0.02 | — | NS |
| Female gender | 0.9 | NS | 0.8 | 0.01 |
| Mixed aortic insufficiency/regurgitation | 0.8 | 0.01 | — | NS |
| Surgery after 1983 | 1.0 | NS | 0.6 | <0.001 |
| Aortic insufficiency | 0.6 | <0.001 | — | NS |
| Previous cardiac surgery | 0.7 | 0.014 | — | NS |
| Infective endocarditis | 0.6 | 0.00 | — | NS |
| Univariate analysis (n = 323) | Multivariate analysis (n = 323) | |||
| Age ≥ 60 years | 2.7 | <0.001 | 1.9 | 0.002 |
| Tissue valve prosthesis | 2.1 | <0.001 | 2.0 | <0.001 |
| Tobacco use | 1.2 | NS | 1.5 | 0.03 |
| Diabetes mellitus | 1.4 | NS | 1.9 | 0.004 |
| Concomitant CABG | 1.1 | NS | 1.7 | 0.002 |
| Subsequent AVR | 2.4 | <0.001 | 0.5 | 0.05 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
