| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2000;17:222-227
© 2000 Elsevier Science NL
a Department of Cardiothoracic Surgery, Academic Medical Center of the University of Amsterdam, P.O. Box 22700, 1100 DE Amsterdam, The Netherlands
b Department of Cardiology, Academic Medical Center of the University of Amsterdam, 1100 DE Amsterdam, The Netherlands
Corresponding author. Tel.: +31-20-566-4323; fax: +31-20-696-2289
e-mail: patrickayaz{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
1.9 cm2/m2. The 30-day mortality was higher in group 1 than in group 2 (18.2 vs. 4.1%, P=0.005). Multivariate logistic regression analysis of the determinants of the 30-day mortality rendered a small valve area index (<1.9 cm2/m2) as an independent risk indicator: relative risk 4.3 (95% CI 1.6–9.5; P=0.0043). The difference in overall survival between the two groups was entirely due to the high 30-day mortality in the patients with small valve area indices, congestive heart failure being the main cause of death. Conclusions: By concentrating on the extreme lower tail of the normal distribution of the valve area index, a strong and independent relation was found between relatively small valves (valve area index <1.9 cm2/m2) and 30-day mortality. We found no influence of valve size on late mortality beyond the first 30 days.
Key Words: Mitral valve • Prosthesis • Surgery • Survival • Follow-up study
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
However, a recent study, using geometric valve area index, did not identify mitral prosthetic valve area index (VAI) as an independent risk factor for early or late mortality, NYHA class, major thrombo-embolism, anti-coagulant related hemorrhage, or reoperation [4]. The aim of the present study was to determine the role of mitral valve prosthesis–patient mismatch on survival in a cohort of consecutive patients after mitral valve replacement with a mechanical prosthesis, focusing on the lower tail of the normal distribution curve of the prosthetic valve area index (VAI).
|
2. Materials and methods |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
The final study population consisted of 428 patients (143 men, 285 women), mean age 61.7±11.2 years, range 22–83 years. Mitral stenosis was present in 76 (18%) patients, mitral regurgitation in 188 (44%) patients, and 164 (38%) patients had a combined lesion. Severe dyspnea, functional class III or IV according to the NYHA criteria was present in 81% of the patients. Isolated mitral valve replacement was performed in 75% of the patients; concomitant coronary bypass operation was performed in 19% and concomitant tricuspid annuloplasty in 6%. The size of prosthesis implanted ranged from 25 to 31 mm. The most frequently used prosthesis size was 27 mm. The prosthetic valves implanted were Medtronic Hall (n=270, 63%) and St. Jude (n=158, 37%).
2.1. Data collection
All data were collected retrospectively and processed in a structured database. Clinical, cardiac catheterization, Doppler echocardiography, operation and mortality data were gathered from patient hospital records. Preoperative symptoms were classified according to the New York Heart Association functional criteria. Any previous cardiac operations were defined as previous surgery. Prior myocardial infarction was recorded in cases of increased enzyme (CK-MB) level or electrocardiographic (ECG) verification. Left heart failure was defined as pulmonary edema or pulmonary vascular congestion, both auscultatory and radiographic, during the year before operation. Right ventricular failure was defined as the presence of increased central venous pressure, and/or hepatomegaly ascites or edema, at physical examination during the year before operation. Data on the presence of atrial fibrillation was obtained from the records of the referring physician or from rereading of the ECG. Cardiomegaly (cor-thorax ratio 50%) was determined by standard chest radiography. Left heart catheterization was performed in 397 patients and Doppler echocardiography in 339 patients.
Mitral valve lesions were classified as stenotic (mean pressure gradient 5 mmHg, mitral valve area <1 cm2 and regurgitation grade 
2/4), isolated regurgitation (Mean pressure gradient 5 mmHg and regurgitation grade 3) or mixed (mean pressure gradient 5 mmHg, regurgitation grade 3). Etiology of the mitral regurgitation was divided in ischemic and non-ischemic. Ischemic mitral regurgitation was said to be present if there was segmental ventricular wall asynergy or annulus dilatation in patients with coronary artery disease and no primary leaflet pathology, or in the case of papillary muscle rupture after myocardial infarction. Coronary artery disease was defined as the presence of luminal narrowing of >50% in coronary angiography. Diabetes mellitus was coded as present if the patient had either the insulin-dependent or the insulin-independent variety. Impaired renal function was defined as a creatinine level above 120 µmol/l. Operations were characterized as elective, urgent (operation during the same admission as the diagnostic evaluation) or emergency (operation within 48 h after hospitalization).
The operations were performed with the standard techniques of cardiopulmonary bypass, including hemodilution and moderate systemic hypothermia with the myocardium protected by cold crystalloid cardioplegia. The mechanical prostheses were implanted in the mitral annulus using interrupted everting mattress sutures. The prosthetic valve area index (VAI) was calculated from the geometric valve area (cm2) divided by the body surface area (m2). We used specifications for geometric valve area supplied by the manufacturer according to prosthesis size. Body surface area was calculated by the formula of Dubois [5].
Data on the clinical course and functional status of hospital survivors were by collected one of us (A.P.Y.) from the clinical records and outpatient records, at the end of follow-up (January to April 1998). If this approach was not feasible, the attending cardiologist or family practitioner answered a questionnaire. The occurrence of death was established by information on all patients from the Census Bureau. Information on causes of death was obtained from clinical records, postmortem examination or a questionnaire to the attending physician. If no information on causes or circumstances of death was available (e.g. patients who emigrated); patients were classified as circumstances unknown. Follow-up was 98% complete, mean follow-up was 68 months (range 0–177 months) and comprised 2434 patient-years.
2.2. Statistical analysis
Continuous variables are presented as mean±SD and as median and range depending on the distribution. The occurrence of categorical variables was expressed as percentages. Student's t-test for unpaired samples and mid-point exact tests analysis [6] were carried out for comparison of continuous and categorical variables, respectively. Continuous variables for within sub-group analysis were compared by the Mann–Whitney test. Cumulative survival rates were estimated using the Kaplan–Meier procedure and compared with the log-rank test. Background mortality was calculated by the method of Verheul et al. [7,8]. The following potential confounders, showing a relation with the outcome with a P-value of less than 0.2 were entered into the model, either as a continuous variable or dichotomized (1=present, 0=other): age (entered as continuous variable), gender, preoperative functional class (III or IV vs. other), prior myocardial infarction, left-sided heart failure, right-sided heart failure, atrial fibrillation, coronary artery disease, mitral regurgitation (mitral regurgitation vs. mitral stenosis and/or combined lesion), ischemic mitral regurgitation (ischemic vs. non-ischemic etiology), creatinine (<120 vs. 120 µmol/l), diabetes mellitus, operation year (<1988 vs. 1988) and emergency operation. Variables with a P-value of less than 0.2 in the univariate analysis, with possible clinical relevance for the mortality, were entered as covariates in the stepwise multivariate regression analysis. Multivariate analysis of 30-day mortality was performed with logistic regression analysis and multivariate analysis of survival was performed using Cox proportional hazards regression method. Dichotomous odds ratios, calculated by multivariate logistic regression, were converted into relative risk by the method of Zhang et al. [9]. A P-value of less than 0.05 was designated as statistically significant. Statistical analysis was performed with SPSS for Windows version 7.5 (SPSS, Chicago, IL).
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
The study population was divided into two groups according to their VAI. The cut-off point was the 10th percentile of the VAI which was 1.918 cm2/m2. Group 1 consisted of 33 patients with a VAI<1.9 cm2/m2 (1.43–1.89 cm2/m2, median 1.82 cm2/m2) and Group 2 consisted of 395 patients who had a VAI 1.9 cm2/m2 (1.9–2.98 cm2/m2, median 2.26 cm2/m2). Patient characteristics according to VAI (cut-off point 1.9 cm2/m2) are listed in Table 1.
|
3.1. Early mortality
The 30-day mortality was higher in patients in group 1 (VAI <1.9 cm2/m2) than in group 2 (18.2 vs. 4.1%, P=0.005). In group 1, significantly more patients died of congestive heart failure in the first 30 days (9.1% in group 1 vs. 1.8% in group 2, P=0.035). Congestive heart failure was also the main cause of death in group 1. Other causes of 30-day mortality (intractable bleeding, major infection/sepsis, LV-wall rupture, acute myocardial infarction, sudden death and anti-coagulant related) were not significantly different between the two groups.
Although unfavorable patient characteristics (like a significantly higher prevalence of ischemic mitral regurgitation, prior myocardial infarction, coronary artery disease and emergency surgery) in patients with a small VAI might have been responsible for the higher early mortality, multivariate logistic regression analysis showed that a small VAI (<1.9 cm2/m2) was an independent risk factor for 30-day mortality (relative risk 4.3, 95% CI 1.6–9.5, P=0.0043). Other independent risk factors were age and emergency surgery (see Table 2).
|
3.2. Late mortality and survival
Survival according to VAI (Group 1, VAI <1.9 cm2/m2 vs. Group 2, VAI 1.9 cm2/m2) and the background mortality is shown in Fig. 1. The estimated 10-year survival was 51% in group 1 and 56% in group 2. The observed difference between both groups is not due to the estimated background survival derived from the healthy general population. On the contrary, the estimated background survival of the patients with small VAI (group 1) is somewhat better than that of group 2. The difference between the background survival and the observed survival which could be designated as the estimated excess or valve-related mortality is greater in group 1. All long-term differences in survival can be explained by the early 30-day mortality. In Kaplan–Meier curves of the 30-day survivors (Fig. 2), there is no difference in survival between the groups over a period of 10 years (log-rank test P=0.98).
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
The present study concentrated on the lower tail of the distribution curve of the geometric VAI (<1.9 cm2/m2). This discrepancy of our results with those of Fernandez et al. may be due to the different distribution of body surface area of the patients in the latter study with an age range from 1 to 84 years, or to the fact that the lower tail of the distribution of VAI was not specifically addressed, to unknown confounding factors, or it may be due to chance. The causes of early death in our series give some support to the theory that extreme valve prosthesis–patient mismatch may cause heart failure in the early postoperative phase. Theoretically, this might be attributed to the combination of a persisting relatively higher afterload of the right ventricle after implantation of a narrow prosthetic valve and stunning of the right ventricle in the early postoperative phase.
From our results it was conspicuous that, among patients with a small VAI (the lower tail of the normal distribution curve of the VAI, group 1), those who died had the smallest VAI. The patients in our series with small VAIs had a significantly larger body surface area and smaller cor-thorax ratio (on the standard chest X-ray), which may point out to the relatively smaller sizes of their heart and subsequent small annulus. There was no difference between the survival of patients with a Medtronic Hall or a St. Jude prosthesis.
The limitations of our study are obvious. It is not a randomized controlled study and it rests critically on only 33 patients with six deaths. It is conceivable that the correlation between narrow valves and early death, even if real, is not entirely causal. There may be hidden causes why surgeons prefer small valves in patients who have a bad prognosis. However, the effect remained after adjusting for possible confounders and the high significance of this relation means that any such hypothetical unknown confounder would have to be of similar force and highly correlated with small VAI to explain it.
With respect to the late survival, our results confirm those of Fernandez et al. In the range studied we found no long-term deleterious effect of a small mitral VAI. All patients (7) in group 1 (VAI <1.9 cm2/m2) who died in the first year did so in the first 61 days. It might be that a small VAI eliminates those with the greatest co-morbidity from our cohort, so the survival curve flattens thereafter. Logically there must be valve sizes that severely impair cardiac function, but general surgical practice may largely avoid this. Where exactly the permissible lower boundary of the VAI lies will be hard to determine in adults. Observations in children with implanted valves might perhaps help to delineate it. After all, in children re-replacement of mitral valve prosthesis because of somatic growth leading to valve prosthesis–patient mismatch is a well-known problem, especially if the child has been operated upon at a young age because of a parachute mitral valve complex. Unfortunately, in studies addressing this subject it is not documented at which level of VAI signs of heart failure develop [10–12]. However, from two reported studies it may be inferred that severe heart failure leading to re-replacement of the prosthetic valve in patients who survive the acute phase (first 60 days) of MVR may occur at a geometric mitral valve area index of 1.2 cm2/m2 [13,14]. The VAI of patients surviving the first 30 days in our population ranged from 1.67 cm2/m2 to 2.97 cm2/m2 (mean 2.25 cm2/m2), which was well above the aforementioned value of 1.2 cm2/m2. This might explain why the late mortality between the 30-day survivors of group 1 and group 2 in our study population was not significantly different.
In conclusion, by concentrating on the extreme lower tail of the normal distribution curve of the valve area index (VAI <1.9 cm2/m2), a strong and independent relation was found between relatively narrow valves and 30-day mortality, congestive heart failure being the main cause of death. We found no influence of a small VAI on late mortality beyond the first 30 days. The effects of mitral valve prosthesis–patient mismatch in adults may be difficult to observe because surgeons in general are well aware of the problem and manage to avoid it.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
Appendix A. Conference discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
Dr Yazdanbakhsh: We studied New York Heart Association class after the operation, and there was no difference between these two groups, we found no difference between these two groups. Exercise capacity was not studied systematically.
Dr Deac: Because a treadmill test can make the difference maybe in the quality of life if the mortality results didn't show the difference.
Dr O. Oto (Izmir, Turkey): Did you find any difference in the two groups from the atrial fibrillation rate point of view and left atrial diameters?
Dr Yazdanbakhsh: Postoperatively, we didn't look at the incidence of atrial fibrillation.
Dr G. Rizzoli (Padova, Italy): What does a small mitral valve mean for you if the median body surface area of your patients is 1.7? What is a small valve for a 1.7 patient?
Dr Yazdanbakhsh: At a body surface area of 1.65, you could best put in a Medtronic Hall prosthesis of 25 mm, and you would be well away from the danger zone.
Dr J. Revuelta (Santander, Spain): Do you think that the preservation of the mitral valve tissue could be a risk factor in producing a valve area index lower than 1.9 square centimeters in a patient with a small body surface area? I mean preservation of the mitral valve subvalvular apparatus during valve replacement. We all try to maintain the mitral valve, the continuity between the papillary muscle and the annulus to preserve the left ventricular function.
Dr Yazdanbakhsh: Yes.
Dr Revuelta: Do you think this could be a risk factor in patients with a small surface area for producing a valve area index lower than 1.9?
Dr Yazdanbakhsh: I am not sure about that, but I think it could be, because you would have a smaller area for fitting the valve.
Dr Revuelta: So, you are not recommending to maintain mitral valve tissue there?
Dr Yazdanbakhsh: In a patient with a relatively small orifice you should put the largest possible valve in. If the valvular tissue is interfering with putting in a sufficiently large valve, then I would say take it away, yes.
Dr E. Baudet (Bordeaux, France): I have just one question. You mentioned that in patients with a valve area index under 1.9 cm2/m2, early mortality was higher, but you did not explain the reasons or the causes of this early mortality.
Dr Yazdanbakhsh: The cause of the mortality?
Dr Baudet: Of early mortality.
Dr Yazdanbakhsh: In group 1, the patients with smaller valve area indices, the major and significant cause of mortality was congestive heart failure.
|
References |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Totaro and V. Argano Patient prosthesis mismatch after mitral valve replacement: Myth or reality? J. Thorac. Cardiovasc. Surg., September 1, 2007; 134(3): 697 - 701. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B.-K. Lam, V. Chan, P. Hendry, M. Ruel, R. Masters, P. Bedard, B. Goldstein, F. Rubens, and T. Mesana The impact of patient-prosthesis mismatch on late outcomes after mitral valve replacement J. Thorac. Cardiovasc. Surg., June 1, 2007; 133(6): 1464 - 1473. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Magne, P. Mathieu, J. G. Dumesnil, D. Tanne, F. Dagenais, D. Doyle, and P. Pibarot Impact of Prosthesis-Patient Mismatch on Survival After Mitral Valve Replacement Circulation, March 20, 2007; 115(11): 1417 - 1425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ruel, F. D. Rubens, R. G. Masters, A. L. Pipe, P. Bedard, and T. G. Mesana Late incidence and predictors of persistent or recurrent heart failure in patients with mitral prosthetic valves J. Thorac. Cardiovasc. Surg., August 1, 2004; 128(2): 278 - 283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Masuda, H. Kado, H. Tatewaki, Y. Shiokawa, and H. Yasui Late results after mitral valve replacement with bileaflet mechanical prosthesis in children: evaluation of prosthesis-patient mismatch Ann. Thorac. Surg., March 1, 2004; 77(3): 913 - 917. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
