Abstract
Background:
Outcomes and risk factors for surgical aortic valve replacement (SAVR) for aortic insufficiency (AI) have not been evaluated in a national cohort. We analyze the incidence, outcomes, and risk factors for SAVR for AI in the Society of Thoracic Surgeons Adult Cardiac Surgery Database.
Methods:
The national database was queried for patients with moderate or greater AI undergoing isolated SAVR between July 2011 and December 2018. Patients with moderate or greater aortic stenosis, acute dissection, active endocarditis, concomitant procedures, or emergent surgery were excluded. AI was staged using guideline criteria based on symptoms and ventricular remodeling. Operative mortality and morbidity were compared between stages and risk factors for operative mortality were identified.
Results:
12,564 patients underwent isolated SAVR for AI from 2011-2018. Patients were most frequently AI Stage D (7,019; 57.5%), compared to B (1,405; 11.2%), C1 (1,128; 9.0%), or C2 (1,325; 10.5%). Operative mortality was 1.1% overall, and increased between Stage C1, C2, and D (0.4 vs 0.7 vs 1.6%, p<0.01), along with major morbidity (5.1 vs 7.5 vs 9.9%, p<0.01). Patients with severe ventricular dilation and ejection fraction <30% experienced higher mortality (2.7 vs 1.0%, p<0.01). Mortality risk factors were symptomatic AI, decreased ejection fraction, age, weight, body surface area, and dialysis.
Conclusions:
Operative mortality and morbidity for isolated SAVR for AI is very low in a national cohort, providing a benchmark for future transcatheter approaches. Operative risk increases with advanced ventricular remodeling. SAVR prior to development of ventricular remodeling may be appropriate in severe AI patients.
Keywords: Aortic valve replacement, myocardial remodeling, outcomes, heart failure
Graphical Abstract
As chronic aortic insufficiency (AI) is initially well compensated and asymptomatic, the optimal timing of intervention is unclear.[1–5] Progressive myocardial damage leads to adverse remodeling with left ventricle (LV) dilation, symptoms, and loss of contractile function.[4] Uncorrected chronic AI decreases long-term survival even in clinically compensated patients. In a recent longitudinal study, asymptomatic severe AI patients with preserved LV ejection fraction (LVEF≥50%) experienced a 2.2% annual mortality risk, much higher than previously believed.[3] Conversely, surgical aortic valve replacement (SAVR) confers a long-term survival benefit with recovery of LV geometry and function.[2,5–8] SAVR prior to LVEF decline may restore life expectancy to match the general population.[2] Because long-term benefits are weighed against short-term risks of surgery, management decisions critically rely on accurate, individualized operative risk assessment. However, outcomes and risk factors for isolated SAVR for chronic AI have not been evaluated in a national cohort.
Severe AI patients with advanced LV remodeling and decreased LVEF have historically been considered high risk for surgery.[9–12] However, steady outcome improvements have been achieved through advancements in operative and myocardial protection techniques and perioperative critical care.[5,7,13] In contrast to previously reported hospital mortality rates of 10–20%, recent series of SAVR for chronic AI have reported operative mortality rates of 0.6–2.0%.[5,7,12,13] Patients with decreased LVEF and/or significant LV dilation have experienced the most dramatic improvements. Patients with low LVEF (≤50%) or very low LVEF (≤30-35%) have reported operative mortality rates approaching that of patients with preserved LVEF.[2,7,13] While encouraging, results have been limited to single-center reports from high-volume, experienced centers. It remains unknown if results and identified risk factors are widely applicable.
In a large, national cohort from the Society of Thoracic Surgeons Adult Cardiac Surgery Database (STS-ACSD), we sought to define contemporary outcomes after isolated SAVR for AI and identify risk factors to guide surgical decision making. The primary risk factors examined are measures of adverse remodeling, including symptom status, LVEF, and LV dilation. The primary outcome is operative mortality, with secondary outcomes of major morbidity and resource utilization.
Material and Methods
Data Source and Selection of Patient Cohort
The STS-ACSD (v2.73, v2.81, and v2.90) was queried for patients with moderate or greater AI undergoing isolated SAVR between July 2011 and December 2018. Variable merging between database versions is reported in Supplementary Table 1. Patients with moderate or greater aortic stenosis, prior cardiac surgery, acute aortic dissection, active endocarditis, concomitant valve, coronary artery bypass grafting, or aortic procedures, or emergent surgery were excluded (Figure 1). The most frequent exclusion was for mixed aortic valve disease (concomitant aortic insufficiency and stenosis). Patients with unknown AI severity, LVEF, symptom status, or operative mortality were excluded. This analysis of retrospective, de-identified data was deemed exempt by the Baylor College of Medicine Institutional Review Board.
Figure 1:
Patient selection and study inclusion.
Data Variables
Baseline characteristics, operative characteristics, and postoperative outcomes were collected using STS database definitions.[14] AI Stage was determined using AHA/ACC guideline criteria based on AI severity, LV dysfunction, and symptom status (Figure 2).[1] ‘Symptomatic’ patients were defined as current or prior heart failure, angina, or syncope. Independent variables with <10% missingness were populated with multiple imputation. Frequency of missing variables is reported in Supplemental Table 3. NYHA Class was used only for descriptive analysis, as data was missing for 26.8% of symptomatic patients. STS Predicted Risk of Mortality (STS-PROM) and Major Morbidity (STS-PROMM) are derived from the 2008 STS isolated valve surgery model.[15] We utilized 2008 STS risk models because this risk score was in use at the time of surgery for this cohort.[15] While this model is recalibrated quarterly for performance benchmarking, we used the model in its original form to maintain consistency with published model coefficients and the available STS online risk calculator.
Figure 2:
Aortic Insufficiency staging by guideline criteria. Percentages represent proportion of preceding node, except AI Stages indicate percentage of overall cohort.
Outcome Definitions
The STS database definition for operative mortality is death during index hospitalization or within 30 days. Major morbidity is defined as stroke, re-operation, deep sternal wound infection, renal failure (new dialysis and/or satisfying RIFLE criteria), or prolonged ventilation (≥24 hours).[16] Resource utilization was assessed by intensive care and hospital length of stay and readmission rates.
Statistical Analysis
Data analysis was performed using R (v3.6; R Foundation, Vienna, Austria). Patient characteristics and outcomes were compared between AI Stages with chi-squared or Kruskal-Wallis tests (α=0.05). Operative years were compared with 2-sided Cochran-Armitage trend tests. Observed-to-expected mortality ratios were calculated by bootstrapping with bias-corrected and accelerated confidence intervals using R package “boot”.[17] Relationships between operative mortality and LVEF, LV end-systolic diameter index (LVESDi), or LV end-diastolic diameter index (LVEDDi) were characterized by univariate logistic regression with restricted cubic splines using R package “rms” (knots at 10th, 50th, and 90th percentiles). Risk factors for operative mortality within the overall cohort were assessed by univariate logistic regression, with factors demonstrating p<0.10 on Wald tests selected for multivariable modeling. Candidate risk factors were determined a priori and are reported in Table 5. Age, weight, and LVEF were analyzed as continuous variables, with odds ratios reported in 10-point increments. The multivariable model was constructed by backward selection, with removal p-value of 0.15. Significance in the final model was defined as p<0.05 on Wald test. Model fit was assessed by Hosmer-Lemeshow goodness-of-fit and receiver operating characteristic curves. Indications for surgery could not be definitively ascertained in stage B or incompletely staged patients. Sensitivity analysis after excluding these patients was performed to confirm robustness of identified risk factors (Supplemental Table 2).
Table 5:
Predictors of Operative Mortality
| Univariate Logistic Regression for Operative Mortality | |||
|---|---|---|---|
| Risk Factor | Odds Ratio | 95% CI | p-value a |
| Age (per 10-pt increase) | 1.44 | 1.28–1.64 | <0.01 |
| Female gender | 1.60 | 1.14–2.25 | 0.01 |
| Weight (per 10-kg increase) | 1.06 | 0.98–1.15 | 0.14 |
| Body surface area (per 0.1 m2 decrease) | 1.08 | 1.01–1.14 | 0.02 |
| Hypertension | 1.97 | 1.25–3.10 | <0.01 |
| Diabetes | 1.90 | 1.30–2.78 | <0.01 |
| End-stage renal disease | 3.72 | 1.87–7.35 | <0.01 |
| Peripheral vascular disease | 2.24 | 1.38–3.64 | <0.01 |
| Cerebrovascular Disease |
1.38 | 0.84–2.27 | 0.20 |
| Mitral Insufficiency (moderate or greater) | 2.19 | 1.43–3.24 | <0.01 |
| Prior MI | 2.32 | 1.45–3.54 | <0.01 |
| Symptomatic AI | 3.23 | 2.20–4.74 | <0.01 |
| Ejection fraction (%) (per 10-pt decrease) | 1.27 | 1.12–1.44 | <0.01 |
| LVESDi (mm/m2) (per 1-pt increase) | 1.03 | 0.99–1.07 | 0.14 |
| LVEDDi (mm/m2) (per 1-pt increase) | 1.01 | 0.98–1.05 | 0.46 |
| Multivariable Model for Operative Mortality | |||
| Risk Factor | Odds Ratio | 95% CI | p-value a |
| Age (per 10-year increase) | 1.45 | 1.26–1.66 | <0.01 |
| Body surface area (per 0.1-m2 decrease) | 1.20 | 0.73–0.96 | <0.01 |
| Weight (per 10-kg increase) | 1.26 | 1.06–1.46 | <0.01 |
| End-stage renal disease | 3.26 | 1.44–6.41 | <0.01 |
| Prior MI | 1.60 | 1.00–2.47 | 0.04 |
| Symptomatic AI | 2.26 | 1.52–3.47 | <0.01 |
| Ejection fraction (%) (per 10-point decrease) | 1.17 | 1.03–1.33 | 0.02 |
Wald test
AI– Aortic Insufficiency; LVESDi/LVEDDi– Left ventricular end-systolic/diastolic diameter, indexed to body surface area; MI– Myocardial infarction
Results
Study Population
In the United States, 12,564 chronic AI patients underwent isolated SAVR between 2011 and 2018. Preoperatively, 3,996 patients (31.6%) had reduced LVEF (<50%) and 2,537 patients had severe LV dilation (34.5% of patients with known LV dimensions) (Table 1). LV dimensions were documented in 7,368 patients (58.6%). Bicuspid aortic valve was noted in 3,082 patients (25.0%), and prior myocardial infarction was reported in 1,011 patients (8.0%). Patients most frequently presented with AI Stage D (7,019; 57.5%) versus B (1,405; 11.2%), C1 (1,121; 8.9%), or C2 (1,325; 10.5%) (Figure 2). Staging was incomplete for 1,694 patients (13.5%) due to missing LV dimensions. The most frequent symptom was dyspnea (6,319; 90.0%), followed by angina (983; 14.0%) and syncope (492; 7.1%). Symptomatic patients were NYHA Class I in 428 (6.1%), II in 1,809 (25.8%), III in 2,113 (30.1%), IV in 786 (11.2%), and unknown in 1,883 (26.8%). Demographics and comorbidities differed by AI Stage (Table 2). The most frequent comorbidities were hypertension (9,483; 75.5%) and diabetes (1,758; 14.0%). Operative characteristics were similar between AI Stages, except more frequent IABP utilization in Stage D (p<0.01) (Table 3).
Table 1:
Preoperative Aortic Insufficiency Staging and Cardiac Status
| Patient Characteristics | Overall Cohort (n=12,564) | Incomplete Staging (n=1,694) | Aortic Insufficiency Stage | p-value a | |||
|---|---|---|---|---|---|---|---|
| Stage B (n=1,279) |
Stage C1 (n=1,121) |
Stage C2 (n=1,451) |
Stage D (n=7,019) |
||||
| Aortic Insufficiency (AI) | |||||||
| Moderate | 11.3% (1,415) | 0.0% (0) | 44.3% (566) | 0.0% (0) | 8.7% (126) | 10.3% (723) | <0.01 |
| Severe | 88.7% (11,149) | 100.0% (1,694) | 55.7% (713) | 100.0% (1,121) | 91.3% (1,325) | 89.7% (6,296) | |
| Symptomatic AI | 55.9% (7,019) | 0.0% (0) | 0.0% (0) | 0.0% (0) | 0.0% (0) | 100.0% (7,019) | <0.01 |
| Bicuspid Aortic Valve | 25.0% (3,082) | 29.4% (488) | 26.1% (361) | 35.6% (395) | 30.1% (393) | 21.0% (1,445) | <0.01 |
| Treated Infective Endocarditis | 3.3%% (413) | 3.8% (65) | 2.6% (33) | 3.4% (38) | 1.9% (28) | 3.5% (249) | 0.01 |
| Mitral Insufficiency (Moderate or greater) | 10.8 (1,363) | 6.0% (101) | 6.8% (87) | 6.4% (72) | 9.2% (133) | 13.8% (970) | <0.01 |
| Prior Myocardial Infarction | 8.0% (1,011) | 4.0% (67) | 4.7% (60) | 3.7% (42) | 6.0% (87) | 10.8% (755) | <0.01 |
| Ejection Fraction (%) | 55.0 [45.0-60.0] | 57.0 [55.0-60.0] | 60.0 [55.0-63.0] | 58.0 [54.0-60.0] | 45.0 [40.0-48.0] | 52.0 [40.0-60.0] | <0.01 |
| ≥ 50% | 68.2% (8,568) | 100.0% (1,694) | 100.0% (1,279) | 100.0% (1,121) | 18.4% (267) | 59.9% (4,207) | <0.01 |
| < 50% | 26.5% (3,327) | 0.0% (0) | 0.0% (0) | 0.0% (0) | 76.8% (1,115) | 31.5% (2,212) | <0.01 |
| < 30% | 5.3% (669) | 0.0% (0) | 0.0% (0) | 0.0% (0) | 4.8% (69) | 8.5% (600) | <0.01 |
| LV Dimensions Documented | 58.6% (7,368) | 0.0% (0) | 82.7% (1,058) | 100.0% (1,121) | 63.5% (922) | 60.8% (4,267) | <0.01d |
| LVESD (mm) | 41.0 [35.0-48.0] | - | 33.0 [29.0-37.0] | 41.5 [38.0-45.0] | 48.0 [43.0-53.0] | 42.0 [35.0-49.5] | <0.01d |
| LVESDi (mm/m2)b | 20.5 [17.4-24.1] | - | 16.7 [14.7-18.7] | 20.4 [18.6-22.0] | 24.6 [21.5-27.1] | 20.9 [17.5-24.9] | <0.01d |
| LVEDD (mm) | 59.0 [52.0-65.0] | 59.0 [52.6-63.0] | 51.0 [46.0-55.9] | 61.0 [58.0-65.0] | 63.0 [58.0-68.8] | 59.0 [52.0-65.0] | <0.01d |
| LVEDDi (mm/m2)b | 29.3 [26.0-32.7] | 29.4 [26.3-32.9] | 25.7 [23.1-27.8] | 30.2 [28.1-32.4] | 32.1 [28.6-35.5] | 29.5 [26.1-33.0] | <0.01d |
| Severe LV Dilationc | 34.5% (2,537) | - | 3.9% (40) | 23.5% (263) | 65.2% (606) | 37.8% (1,617) | <0.01d |
Between-stages comparison
Indexed to body surface area
LVESD>50mm, LVESDi>25mm, or LVEDD>65mm (% patients with LV dimensions documented)
comparison excludes incompletely staged patients
LV– left ventricle; LVESD/LVEDD– Left ventricular end-systolic/diastolic diameter
Table 2:
Patient Demographics and Comorbidities
| Patient Characteristics | Overall Cohort (n=12,564) | Incomplete Staging (n=1,694) | Aortic Insufficiency Stage | p-value a | |||
|---|---|---|---|---|---|---|---|
| B (n=1,279) |
C1 (n=1,121) |
C2 (n=1,451) |
D (n=7,019) |
||||
| Age (yrs) (Median, [IQR]) | 59.0 [49.0-68.0] | 58.0 [47.0-67.0] | 59.0 [50.0-69.0] | 56.0 [45.0-66.0] | 57.0 [46.0-68.0] | 60.0 [50.0-69.0] | <0.01 |
| Female gender (n, % of group) | 25.0% (3,141) | 24.6% (416) | 27.6% (353) | 19.4% (218) | 17.6% (256) | 27.0% (1,898) | <0.01 |
| Weight (kg) | 85.0 [73.0-99.3] | 84.6 [73.0-98.7] | 85.0 [73.5-98.4] | 87.5 [76.0-99.0] | 83.1 [72.5-97.0] | 85.0 [72.0-100.0] | <0.01 |
| Body surface area (m2) | 2.0 [1.8-2.2] | 2.01 [1.8-2.2] | 2.00 [1.8-2.2] | 2.04 [1.9-2.2] | 2.00 [1.8-2.2] | 2.00 [1.8-2.2] | <0.01 |
| Hypertension | 75.5% (9,483) | 70.7% (1,195) | 72.5% (927) | 68.7% (770) | 70.0% (1,015) | 79.5% (5,576) | <0.01 |
| Diabetes | 14.0% (1,758) | 8.8% (149) | 10.9% (139) | 8.1% (9) | 10.2% (148) | 17.6% (1,231) | <0.01 |
| End-stage renal disease | 1.8% (220) | 0.8% (14) | 0.5% (7) | 0.8% (9) | 1.6% (23) | 2.4% (167) | <0.01 |
| Peripheral vascular Disease | 6.1% (769) | 4.4% (75) | 6.1% (78) | 3.6% (40) | 4.7% (68) | 7.2% (508) | <0.01 |
| Cerebrovascular disease | 9.1% (1,143) | 7.0% (119) | 7.8% (99) | 6.9% (77) | 6.4% (93) | 10.8% (755) | <0.01 |
| Operative Risk | |||||||
| STS-PROM | 1.0% [0.6-1.7] | 0.7% [0.5-1.1] | 0.8% [0.5-1.3] | 0.7% [0.5-1.0] | 0.8% [0.5-1.3] | 1.3% [0.8-2.2] | <0.01 |
| STS-PROMM | 11.2% [8.2-16.2] | 8.9% [7.0-11.7] | 9.0% [7.0-12.2] | 8.4% [6.7-10.8] | 9.9% [7.6-13.2] | 13.5% [9.8-19.4] | 0.03 |
Between-stages comparison
STS-PROM(M)–predicted risk of operative mortality (or major morbidity)
Table 3:
Procedural Characteristics
| Patient Characteristics | Overall Cohort | Incomplete Staging | Aortic Insufficiency Stage | p-value a | |||
|---|---|---|---|---|---|---|---|
| B | C1 | C2 | D | ||||
| Cardiopulmonary Bypass Time (min) (Median,[IQR]) | 91 [73-115] | 91 [73-114] | 94 [74-125] | 90 [70-113] | 92 [73-116] | 91 [73-114] | <0.001 |
| IABP Usage/Timing of Insertion | |||||||
| None | 98.6% (12,376) | 100.0% (1,691) | 99.7% (1,275) | 99.7% (1,117) | 98.6% (1,429) | 97.8% (6,864) | <0.001 |
| Preoperative | 0.1% (11) | 0.0% (0) | 0.0% (0) | 0.0% (0) | 0.0% (0) | 0.2% (11) | |
| Intraoperative | 1.1% (143) | 0.0% (0) | 0.2% (2) | 0.1% (1) | 1.2% (18) | 1.7% (122) | |
| Postoperative | 0.2% (27) | 0.0% (0) | 0.2% (2) | 0.2% (2) | 0.2% (3) | 0.3% (20) | |
Between-stages comparison
IABP– Intra-aortic balloon pump
Clinical Outcomes by Aortic Insufficiency Stage
Operative mortality was 1.1% overall, with significant differences between AI Stages (Table 4). Operative mortality for Stage B patients was 0.5%. Operative mortality increased between Stage C1, C2, and D (0.4 vs 0.7 vs 1.6%, p<0.01). Patients with both LVEF <30% and severe LV dilation had higher operative mortality (2.7 vs. 1.0%, p<0.01). There was no difference in operative mortality (1.1 vs 1.2%, p=0.62) between patients with known (n=7,368) and unknown (n=5,196) LV dimensions, suggesting LV dimensions were not missing due to underlying differences in operative risk. The overall observed-to-expected ratio for operative mortality was 0.75 (95% CI 0.65-0.80). The observed-to-expected ratio was <1 for each AI stage. (Figure 3).
Table 4:
Operative Outcomes and Resource Utilization
| Outcome Measure | Overall Cohort | Incomplete Staging | Aortic Insufficiency Stage | p-value a | |||
|---|---|---|---|---|---|---|---|
| B | C1 | C2 | D | ||||
| Operative Mortality (n,%) | 1.1% (144) | 0.6% (10) | 0.5% (7) | 0.4% (4) | 0.7% (10) | 1.6% (113) | <0.01 |
| Major Morbidity | 8.2% (1,024) | 6.0% (101) | 4.9% (63) | 5.1% (57) | 7.6% (110) | 9.9% (693) | <0.01 |
| Stroke | 0.2% (21) | 0.2% (3) | 0.2% (2) | 0.2% (2) | 0.2% (3) | 0.2% (11) | 0.99 |
| Renal failure | 1.3% (158) | 0.9% (15) | 0.4% (5) | 0.5% (6) | 1.1% (16) | 1.7% (116) | <0.01 |
| Reoperation for bleeding | 3.0% (375) | 2.6% (44) | 1.6% (20) | 2.9% (33) | 3.4% (49) | 3.3% (229) | 0.01 |
| Permanent pacemaker | 3.0% (377) | 2.6% (44) | 2.7% (34) | 2.2% (25) | 3.3% (48) | 3.2% (226) | 0.25 |
| Prolonged ventilation | 5.3% (667) | 3.3% (56) | 3.4% (43) | 2.4% (27) | 4.1% (60) | 6.9% (481) | <0.01 |
| Resource UtiliZation | |||||||
| ICU Length of stay (hours) | 118 [73-211] | 109 [77-155] | 108 [65-155] | 77 [52-99] | 132 [76-197] | 129 [77-228] | 0.024 |
| Length of stay (days) | 6.0 [5.0-8.0] | 5.0 [4.0-7.0] | 5.0 [4.0-7.0] | 5.0 [4.0-6.0] | 5.0 [4.0-7.0] | 6.0 [5.0-10.0] | <0.01 |
| 30-Day Readmission | 10.1% (721) | 9.1% (95) | 9.8% (72) | 9.7% (63) | 7.1% (62) | 11.2% (429) | 0.01 |
Between-stages comparison
Figure 3:
Observed-to-expected operative mortality ratio for overall cohort and aortic insufficiency stages. Error bars represent 95%CI.
Major morbidity increased between Stage C1, C2, and D (5.1 vs 7.5 vs 9.9%, p<0.01) (Table 4). The most frequent complications were prolonged ventilation (5.3%), permanent pacemaker (3.0%) and re-operation (3.0%). Resource utilization increased between Stage C1, C2, and D, with increased length of stay in intensive care (77 vs 111 vs 129 hours, p=0.02) and overall (5 vs 5 vs 6 days, p<0.01).
Cochran-Armitage analysis showed no trend in operative mortality by year (Z=0.11, p=0.91). Major morbidity trended downward over the study period (Z= −2.51, p=0.01) (Figure 4).
Figure 4:
Overall operative mortality and major morbidity by operative year.
Measures of Left Ventricular Function and Relationship to Operative Mortality
Patients with preserved LVEF (≥50%) had lower mortality than patients with moderately (30-50%) reduced LVEF (0.9 vs 1.7%, p<0.01) or severely (≤30%) reduced LVEF (0.9 vs 2.1%, p<0.01). Operative mortality did not differ between patients with moderately and severely reduced LVEF (1.7 vs 2.1%, p=0.53). Patients with severely reduced LVEF exhibited greater LV dilation than patients with moderately reduced or preserved LVEF (median LVESDi 29.0 vs 23.7 vs 19.1 mm; p<0.01).
Relationships between operative mortality and LV function (LVEF, LVESDi, LVEDDi) are demonstrated in Figure 5. Declining LVEF appeared to steadily increase operative mortality risk throughout the observed range of LVEF values. In contrast, LVESDi and LVEDDi did not demonstrate a positive association with mortality until the upper end of their range (i.e., severe dilation).
Figure 5:
Relationship between operative mortality and left ventricular dysfunction. LVEF–Left ventricular ejection fraction, LVEDD–Left ventricular end-diastolic diameter; LVESD–Left ventricular end-systolic diameter.
Predictors of Operative Mortality
Logistic regression modeling of risk factors for operative mortality is presented in Table 5. The multivariable model (Table 5) demonstrated acceptable calibration (Hosmer-Lemeshow test, p=0.37) and good discrimination (c=0.72). Symptomatic AI significantly increased odds of operative mortality (OR 2.51, 95%CI 1.68-3.73). Additional risk factors were decreased LVEF, prior MI, age, female sex, and preoperative dialysis. (Table 5).
Comment
In 12,564 patients with chronic AI undergoing SAVR in the STS-ACSD between 2011 and 2018, operative mortality (1.1%) and major morbidity (8.2%) were very low in a contemporary, national cohort. Patients with advanced AI Stage experienced higher operative mortality and major morbidity. Symptomatic AI had developed in 56% of patients prior to operative intervention and was an independent predictor of operative mortality. Contrary to single-center reports, reduced LVEF increased risk for operative mortality, particularly with concomitant LV dilation. This suggests surgery should be considered prior to development of severe LV remodeling. However, operative outcomes remain acceptable in these patients, suggesting it can still be safe and effective to offer surgery after advanced remodeling and/or symptoms develop.
Prior studies detailing operative outcomes in chronic AI have been single-center reports from valve centers of excellence. Mentias and colleagues reported the Cleveland Clinic experience in patients with chronic AI, LVEF ≥50%, and mild or no symptoms undergoing aortic valve (AV) surgery, with or without concomitant procedures.[2] Operative mortality was 0.6% for 336 patients undergoing isolated AV surgery. Murashita and colleagues at Mayo Clinic reported operative mortality of 0.75% in 530 patients undergoing AV surgery +/− concomitant procedures.[5] Kaneko and colleagues from Brigham and Women’s Hospital reported outcomes from 485 chronic AI patients undergoing SAVR +/− revascularization from 2002-2013, with higher operative mortality (1.9%) than the present study.[13] However, this cohort may have been at greater operative risk due to prior cardiac surgery (22.5%) and concurrent revascularization (18.7%). Overall, the present study corroborates prior reports using a national cohort and suggests excellent postoperative outcomes are not restricted to a few highly specialized centers.
The impact of LVEF has been explored in prior studies of SAVR for chronic AI .[5,7,8,13] While severely reduced LVEF (≤30-35%) was historically associated with high in-hospital mortality, contemporary outcomes have improved dramatically and several authors have suggested outcomes are now equivalent to preserved LVEF patients. Bhudia and colleagues detailed outcome trends after SAVR for chronic AI at Cleveland Clinic, and found LVEF<30% did not increase mortality risk in operations after 1985.[7] This was reinforced by Murashita and colleagues, who reported their historical ~14% operative mortality rate after SAVR for chronic AI patients with LVEF<35% improved to 0% (0/32) over a similar period.[5] Kaneko and colleagues also found no operative mortality difference between patients with low (≤35%), moderately reduced (36-50%), and preserved (>50%) LVEF.[13] However, these reports were limited by small numbers of low LVEF patients. The present study includes 669 chronic AI patients with LVEF<30%, by far the largest cohort to date. In contrast to single-center studies, we found decreased LVEF remains an independent risk factor for operative mortality (OR 1.16 per 10-point decrease). However, outcomes remained acceptable with severely reduced LVEF, and this should not be considered a contraindication for surgery.
No transcatheter therapy is currently approved for pure AI, and prior studies have reported increased complications and decreased procedural success with transcatheter valve replacement (TAVR) in these patients.[21] Without a viable transcatheter option, the results of the current study suggest patients with chronic AI should be referred earlier for consideration of SAVR. Importantly, patients presenting after development of symptoms and/or LV dysfunction should still be considered for SAVR as morbidity and mortality remained low in these patients. This study may serve as a comparator for evaluating success of TAVR strategies currently in development for pure AI.
Limitations
This study has limitations inherent to retrospective analysis. Though the STS-ACSD captures >90-95% of cardiac operations in the United States, patients managed medically would not be captured.[20] Surgery may be performed in comparatively healthy patients, and prior reports suggest patients with older age or depressed LVEF are frequently not referred for surgery.[8] Thus, our results cannot be taken as representing expected results in every patient presenting with severe chronic AI.[2,8] Given low operative mortality in the study cohort, we were not able to identify characteristics defining prohibitive risk. As we excluded concomitant aortic operations to isolate effects of AI Stage and remodeling, additive risk in such patients will need to be addressed in future studies. In addition to more extensive surgery, patients with requiring aortic replacement may differ in underlying cardiovascular disease burden. While observed-to-expected mortality ratios were below one for all stages of chronic AI, model calibration could improve with the newer 2018 STS risk model and/or incorporation of recalibration factors.[18–19] A subset of our cohort did not have a readily identifiable indication for surgery under current guidelines. This may reflect limitations in echocardiographic documentation in the STS-ACSD, as only the most recent examination prior to surgery is recorded. However, results remained consistent in sensitivity analyses excluding these patients. Finally, we used multivariable regression to minimize confounding but cannot exclude potential unmeasured confounders such as frailty, control of medical comorbidities, or surgeon/institution case volume.
Outcomes in pure AI patients with symptoms and LV dysfunction would ideally be compared to non-surgical management, as operative risk may be balanced by the annual mortality risk of uncorrected valve dysfunction.[3] While surgical databases do not allow for comparison of surgery versus medical treatment, prior mid-term studies have demonstrated a mortality benefit to SAVR in these patients.[2,5,8]
Conclusions
This study found operative mortality and morbidity is very low after isolated SAVR for chronic AI in the United States. Over half of patients had progressed to symptomatic AI prior to operative intervention, and these patients were at higher operative risk. More advanced AI stage and LV remodeling also increased operative risk. Together with longitudinal studies demonstrating a significant annual mortality risk with uncorrected severe chronic AI, these results suggest SAVR prior to the development of LV remodeling may be appropriate in patients with severe AI.
Supplementary Material
Acknowledgment:
Data provided by The Society of Thoracic Surgeons’ National Database Participant User File Research Program with analysis performed at the investigators’ institution(s). Views or opinions are solely those of the author(s) and do not represent those of The Society of Thoracic Surgeons.
Abbreviations
- AHA/ACC
American Heart Association/American College of Cardiology
- AI
Aortic insufficiency
- LV
Left ventricle
- LVEF
Left ventricular ejection fraction
- LVESDi
Left ventricular end-systolic diameter index
- LVEDDi
Left ventricular end-diastolic diameter index
- NYHA
New York Heart Association
- RIFLE
Risk, injury, failure, loss of kidney function, end-stage kidney disease
- SAVR
Surgical aortic valve replacement
- STS-ACSD
Society of Thoracic Surgeons Adult Cardiac Surgery Database
- STS-PROM
Society of Thoracic Surgeons predicted risk of mortality
- STS-PROMM
Society of Thoracic Surgeons predicted risk of mortality or morbidity
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- [1].Nishimura RA, Otto CM, Bonow RO, et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: A report of the american college of cardiology/american heart association task force on practice guidelines. Circulation. 2014;129:2440–2492.. [DOI] [PubMed] [Google Scholar]
- [2].Mentias A, Feng K, Alashi A, et al. Long-Term Outcomes in Patients With Aortic Regurgitation and Preserved Left Ventricular Ejection Fraction. J Am Coll Cardiol. 2016;68:2144–53. [DOI] [PubMed] [Google Scholar]
- [3].Detaint D, Messika-Zeitoun D, Maalouf J, et al. Quantitative Echocardiographic Determinants of Clinical Outcome in Asymptomatic Patients With Aortic Regurgitation. JACC Cardiovasc Imaging. 2008; 1:1–11. [DOI] [PubMed] [Google Scholar]
- [4].Lee JKT, Franzone A, Lanz J, et al. Early detection of subclinical myocardial damage in chronic aortic regurgitation and strategies for timely treatment of asymptomatic patients. Circulation. 2018;137:184–96. [DOI] [PubMed] [Google Scholar]
- [5].Murashita T, Schaff HV, Suri RM, et al. Impact of Left Ventricular Systolic Function on Outcome of Correction of Chronic Severe Aortic Valve Regurgitation: Implications for Timing of Surgical Intervention. Ann Thorac Surg. 2017;103:1222–8. [DOI] [PubMed] [Google Scholar]
- [6].Fiedler AG, Bhambhani V, Laikhter E, et al. Aortic valve replacement associated with survival in severe regurgitation and low ejection fraction. Heart. 2018;104:835–40. [DOI] [PubMed] [Google Scholar]
- [7].Bhudia SK, McCarthy PM, Kumpati GS, et al. Improved Outcomes After Aortic Valve Surgery for Chronic Aortic Regurgitation With Severe Left Ventricular Dysfunction. J Am Coll Cardiol. 2007;49:1465–71. [DOI] [PubMed] [Google Scholar]
- [8].Kamath AR, Varadarajan P, Turk R, et al. Survival in patients with severe aortic regurgitation and severe left ventricular dysfunction is improved by aortic valve replacement results from a cohort of 166 patients with an ejection fraction ≤35%. Circulation. 2009;120:134–8. [DOI] [PubMed] [Google Scholar]
- [9].lung B, Baron G, Butchart EG, et al. A prospective survey of patients with valvular heart disease in Europe: The Euro Heart Survey on valvular heart disease. Eur Heart J. 2003;24:1231–43. [DOI] [PubMed] [Google Scholar]
- [10].Klodas E, Enriquez-Sarano M, Tajik AJ, Mullany CJ, Bailey KR, Seward JB. Optimizing Timing of Surgical Correction in Patients With Severe Aortic Regurgitation: Role of Symptoms. J Am Coll Cardiol. 1997;30:746–52. [DOI] [PubMed] [Google Scholar]
- [11].Bonow RO, Nikas D, Elefteriades JA. Valve Replacement for Regurgitant Lesions of the Aortic or Mitral Valve in Advanced Left Ventricular Dysfunction. Cardiol Clin. 1995;13:73–83. [PubMed] [Google Scholar]
- [12].Stone PH, Clark RD, Goldschlager N, Selzer A, Cohn K. Determinants of prognosis of patients with aortic regurgitation who undergo aortic valve replacement. J Am Coll Cardiol. 1984;3:1118–26. [DOI] [PubMed] [Google Scholar]
- [13].Kaneko T, Ejiofor JI, Neely RC, et al. Aortic Regurgitation with Markedly Reduced Left Ventricular Function Is Not a Contraindication for Aortic Valve Replacement. Ann Thorac Surg. 2016;102:41–7. [DOI] [PubMed] [Google Scholar]
- [14].Society of Thoracic Surgeons. Adult Cardiac Database: Data Collection Sheet 2019. https://www.sts.org/registries-research-center/sts-national-database/adult-cardiac-surgery-database.
- [15].O’Brien SM, Shahian DM, Filardo G, et al. The Society of Thoracic Surgeons 2008 Cardiac Surgery Risk Models: Part 2-lsolated Valve Surgery. Ann Thorac Surg. 2009;88:S23–42. [DOI] [PubMed] [Google Scholar]
- [16].Badhwar V, Rankin JS, Thourani VH, et al. The Society of Thoracic Surgeons Adult Cardiac Surgery Database: 2018 Update on Research: Outcomes Analysis, Guality Improvement, and Patient Safety. Ann Thorac Surg. 2018;106:8–13. [DOI] [PubMed] [Google Scholar]
- [17].Carpenter J, Bithell J. Bootstrap confidence intervals: when, which, what? A practical guide for medical statisticians. Stat Med. 2000;19:1141–64. [DOI] [PubMed] [Google Scholar]
- [18].Jin R, Furnary AP, Fine SC, Blackstone EH, Grunkemeier GL. Using Society of Thoracic Surgeons Risk Models for Risk-Adjusting Cardiac Surgery Results. Ann Thorac Surg. 2010;89:677–82. [DOI] [PubMed] [Google Scholar]
- [19].O’Brien SM, Feng L, He X, et al. The Society of Thoracic Surgeons 2018 Adult Cardiac Surgery Risk Models: Part 2—Statistical Methods and Results. Ann Thorac Surg. 2018;105:1419–28. [DOI] [PubMed] [Google Scholar]
- [20].Fernandez FG, Shahian DM, Kormos R, et al. The Society of Thoracic Surgeons National Database 2019 Annual Report. Ann Thorac Surg. 2019;108:1625–32. [DOI] [PubMed] [Google Scholar]
- [21].Jiang J, Liu X, He Y, et al. Transcatheter Aortic Valve Replacement for Pure Native Aortic Valve Regurgitation: A Systematic Review. Cardiology 2018;141(3):132–40. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.