Although aortic valve replacement (AVR) and mitral valve replacement (MVR) are the most commonly performed prosthetic valve replacement operations, it is unclear whether clinical outcomes of paravalvular leakage (PVL) after MVR or AVR are different. It was hypothesized that clinical outcomes of PVL after AVR would be more favorable than after MVR because the pressure gradient is much larger in PVL occurring at the mitral position, which happens at the systolic phase, than at the aortic valve. Over a 12-year period, 82 patients with PVL were identified. After excluding patients who required immediate surgical repair for severe symptoms, patients with Behçet disease or infective endocarditis, and those with PVL involving both valves, 54 remaining patients (21 women, mean age 56 ± 14 years, 23 AVRs) with mild to moderate leakage constituted the study population. The end points were cardiac death, all-cause mortality, repeat surgery, and urgent admission for heart failure. During a median follow-up period of 35 months, there were 27 events, including 23 repeated surgeries, 2 cardiac deaths, 1 noncardiac death, and 1 admission for heart failure. Cox regression analysis revealed that the valve location of PVL was the only independent clinical predictor of event-free survival. The estimated 8-year event-free survival rate was significantly higher in patients with PVL after AVR than those after MVR (70 ± 12% vs 16 ± 8%, p <0.0001). In conclusion, PVL after AVR demonstrated more favorable long-term clinical outcomes compared to that after MVR. In patients who develop PVL after AVR, repeat surgery may be deferred. However, in patients with PVL after MVR, more aggressive therapeutic approaches should be considered.
Despite advances in surgical techniques, valve design, and postoperative and long-term management, prosthetic valve dysfunction and valve-related complications remain potentially life-threatening sequelae of valve replacement. Among the complications, paravalvular leakage (PVL) is uncommon but 1 of the most frequent causes leading to reoperation after prosthetic valve replacement. The consensus to recommend reoperation of PVL is PVL related to endocarditis, causing hemolysis needing repeated blood transfusions, or leading to severe symptoms. However, only few data are available regarding therapeutic options for the management of PVL accompanied by less severe hemolysis or mild to moderate symptoms. In addition, although aortic valve replacement (AVR) and mitral valve replacement (MVR) are the 2 most commonly performed prosthetic valve replacement operations, it is unclear whether clinical outcomes of PVL after mitral or AVR are different. Because the pressure gradient is much larger in PVL occurring at the mitral position, which happens at the systolic phase, than at the aortic valve, which occurs at the diastolic phase, we hypothesized that clinical outcomes of PVL after AVR would be more favorable than those of PVL after MVR.
Methods
We retrospectively reviewed all records of 82 patients diagnosed with PVL at our institution over a 12-year period (January 1997 to December 2009). Seven patients with Behçet disease, 4 patients with infective endocarditis, and 1 patient with PVL involving the mitral and aortic positions were excluded. Fifteen patients (2 with AVR, 13 with MVR) who required immediate surgical repair at the time of the presentation because of severe symptoms or requiring repeated blood transfusion for persisting hemolysis were also excluded. Ultimately, 55 patients with mild to moderate PVL constituted the study population.
The echocardiographic images of the included patients were reanalyzed by 2 experienced echocardiographers who were unaware of patients’ clinical data. Doppler color flow mapping was used to assess the competence of the prosthetic valves. A high-velocity, eccentric turbulent jet with its origin beyond the edge of the sewing ring was considered PVL. A laminar, low-velocity regurgitant jet with its origin within the orifice of the sewing ring was considered transvalvular. The presence of a regurgitant flow was also checked using continuous-wave Doppler. If transthoracic echocardiographic data were insufficient to diagnose PVL, transesophageal echocardiographic data were additionally analyzed. Nine patients with AVR (39% of AVRs) and 19 patients with MVR (59% of MVRs) underwent additional transesophageal echocardiographic examinations.
The severity of PVL was assessed semiquantitatively using visual estimation. PVL in the aortic position was considered mild if the ratio of the regurgitant jet to left ventricular outflow tract diameter was <25%. Mild, moderate, and large PVL of the mitral prosthetic valve were defined as maximum widths of the vena contracta of 1 to 2, 3 to 6, and ≥6 mm, respectively.
The primary end point was defined as the composite of cardiac death, all-cause mortality, repeat valve surgery, and urgent admission for heart failure during the follow-up period. A hospitalization for heart failure was defined as an unplanned, urgent admission for the management of heart failure. The occurrence of a clinical event was checked by reviewing the hospital records and by a telephone interview if needed.
The distributions of all relevant variables are reported either as percentages or as mean ± SD. The groups were compared using chi-square statistics for categorical variables and Student’s t test for continuous variables. Multivariate Cox proportional-hazards analysis was used to determine independent variables for event-free survival after diagnosis of PVL, with the variables showing statistical significance in univariate analysis between patients with and without events as covariates. For Kaplan-Meier analysis, we analyzed all clinical events by the time to the first event. A p value <0.05 was considered statistically significant.
Results
Clinical characteristics of the patients are listed in Table 1 . Patients with PVL after MVR were older, and women were more common. Most patients was diagnosed as having PVL >1 year after valve replacement surgery. In patients with PVL after MVR, patients with symptoms and hemolytic anemia were more common. There were no significant differences in the severity of the leakage, left ventricular dimensions, the ejection fraction, and mass index between PVL after AVR and that after MVR. However, the left atrial volume index was significantly larger and right ventricular systolic pressure higher in patients with PVL after MVR compared to those with PVL after AVR.
| Variable | All Patients (n = 55) | Location of Replaced Valve | ||
|---|---|---|---|---|
| PVL After AVR (n = 23) | PVL After MVR (n = 32) | p Value | ||
| Age (years) | 56 ± 14 | 52 ± 14 | 59 ± 13 | 0.038 |
| Women | 21 (38%) | 4 (17%) | 17 (53%) | 0.007 |
| Mechanical/tissue | 49/6 | 20/3 | 29/3 | 0.667 |
| Reason for valve replacement | 0.206 | |||
| Rheumatic | 9 (16%) | 1 (4%) | 8 (25%) | |
| Nonrheumatic | 20 (36%) | 11 (48%) | 9 (28%) | |
| Prosthetic valve failure | 17 (31%) | 7 (30%) | 10 (31%) | |
| Indeterminate | 9 (16%) | 4 (17%) | 5 (16%) | |
| Interval between operation and diagnosis of PVL | 0.025 | |||
| 1–7 days | 4 (7%) | 4 (17%) | 0 (0%) | |
| 8–30 days | 3 (6%) | 2 (9%) | 1 (3%) | |
| 31–365 days | 6 (11%) | 4 (17%) | 2 (6%) | |
| >1 year | 42 (76%) | 13 (57%) | 29 (91%) | |
| Presenting symptoms | 0.001 | |||
| Symptomatic | 19 (35%) | 3 (13%) | 16 (50%) | |
| Asymptomatic | 36 (66%) | 20 (87%) | 16 (50%) | |
| NYHA functional class ≥II | 14 (28%) | 3 (13%) | 11 (34%) | 0.117 |
| Hypertension | 5 (9%) | 1 (4%) | 4 (13%) | 0.311 |
| Diabetes mellitus | 5 (9%) | 1 (4%) | 4 (13%) | 0.311 |
| Presence of hemolytic anemia | 18 (33%) | 4 (17%) | 14 (44%) | 0.040 |
| Hemoglobin level at the time of diagnosis of PVL (mg/dl) | 9.9 ± 2.7 | 11.1 ± 2.5 | 9.3 ± 2.6 | 0.060 |
| Echocardiographic parameters | ||||
| Grade of leakage | 0.563 | |||
| Mild | 37 (67%) | 16 (70%) | 21 (66%) | |
| Moderate | 18 (33%) | 7 (30%) | 11 (34%) | |
| LV end-diastolic diameter (mm) | 54 ± 9 | 54 ± 8 | 53 ± 8 | 0.655 |
| LV end-systolic diameter (mm) | 37 ± 7 | 37 ± 6 | 37 ± 8 | 0.881 |
| LV ejection fraction (%) | 61 ± 9 | 60 ± 7 | 61 ± 10 | 0.825 |
| LV mass index (g/m 2 ) | 133 ± 45 | 133 ± 56 | 133 ± 36 | 0.980 |
| LA volume index (ml/m 2 ) | 73 ± 52 | 34 ± 16 | 105 ± 49 | <0.0001 |
| RV systolic pressure (mm Hg) | 40 ± 15 | 30 ± 8 | 50 ± 14 | 0.003 |
During a median follow-up period of 35 months, there were 27 events (49%), including 23 repeat surgeries (42%), 2 cardiac deaths (4%), 1 noncardiac deaths (2%), and 1 admission for heart failure (2%). In patients with PVL after AVR, there were 3 repeat surgeries, 1 cardiac death, and 1 noncardiac death. In contrast, there were 20 repeat surgeries, 1 cardiac death, and 1 admission for heart failure in patients with PVL after MVR. The estimated 8-year event-free survival rate was significantly higher in patients with PVL after AVR than in patients with PVL after MVR (70 ± 12% vs 16 ± 8%, p <0.0001; Figure 1 ). During the follow-up period, hemolytic anemia, which was less common in patients with PVL after AVR, recovered spontaneously in 75% of patients with PVL after AVR. In contrast, it recovered only in 14% of patients with PVL after MVR, and the other patients finally underwent reoperation without recovery of hemolytic anemia ( Figure 2 ).
Table 2 lists the intergroup comparison of clinical characteristics and initial echocardiographic parameters between the patients with and without events. Clinical characteristics including age, gender, reason for valve replacement, interval between operation and PVL diagnosis, prevalence of hypertension, diabetes mellitus and hemolytic anemia, hemoglobin level at diagnosis of leakage, and New York Heart Association functional class were largely comparable between the groups. There were also no statistically significant differences in severity of PVL, left ventricular diameters, the ejection fraction, mass index, and left atrial volume index between the groups. However, in patients who had events, PVL after MVR was more frequent than that after AVR. Cox regression analysis demonstrated that valve location, that is, whether PVL occurred after MVR or AVR, was the only independent clinical predictor of event-free survival ( Table 3 ).
| Variable | Without Events | With Events | p Value |
|---|---|---|---|
| (n = 28) | (n = 27) | ||
| Age (years) | 55 ± 14 | 57 ± 14 | 0.414 |
| Women | 11 (39%) | 10 (37%) | 0.864 |
| Aortic valve/mitral valve | 18/10 | 5/22 | 0.001 |
| Mechanical/tissue | 25/3 | 25/2 | 0.999 |
| Reason for valve replacement | 0.249 | ||
| Rheumatic | 4 (14%) | 5 (19%) | |
| Nonrheumatic | 14 (50%) | 10 (37%) | |
| Prosthetic valve failure | 5 (18%) | 10 (37%) | |
| Indeterminate | 5 (18%) | 2 (7%) | |
| Interval between operation and diagnosis of PVL | 0.362 | ||
| ≤1 year | 9 (33%) | 6 (22%) | |
| >1 year | 19 (67%) | 21 (78%) | |
| Presenting symptoms | 0.011 | ||
| Symptomatic | 5 (18%) | 14 (52%) | |
| Asymptomatic | 23 (82%) | 13 (48%) | |
| NYHA functional class ≥II | 5 (19%) | 9 (33%) | 0.214 |
| Hypertension | 2 (7%) | 3 (11%) | 0.999 |
| Diabetes mellitus | 3 (11%) | 2 (7%) | 0.999 |
| Presence of hemolytic anemia | 7 (25%) | 11 (41%) | 0.214 |
| Hemoglobin level at diagnosis of PVL (mg/dl) | 10.1 ± 2.9 | 9.8 ± 2.4 | 0.958 |
| Echocardiographic parameters | |||
| Grade of leakage | |||
| Mild | 19 (68%) | 18 (67%) | 0.500 |
| Moderate | 9 (32%) | 9 (33%) | |
| LV end-diastolic diameter (mm) | 54 ± 9 | 54 ± 7 | 0.846 |
| LV end-systolic diameter (mm) | 38 ± 8 | 37 ± 6 | 0.791 |
| LV ejection fraction (%) | 60 ± 9 | 61 ± 9 | 0.519 |
| LV mass index (g/m 2 ) | 123 ± 33 | 148 ± 58 | 0.190 |
| LA volume index (ml/m 2 ) | 62 ± 44 | 89 ± 60 | 0.194 |
| RV systolic pressure (mm Hg) | 38 ± 17 | 45 ± 10 | 0.401 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
