- Email: [email protected]
Pregnancy does not accelerate biological valve degeneration
International Journal of Cardiology 145 (2010) 418–421
Contents lists available at ScienceDirect
International Journal of Cardiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j c a r d
Pregnancy does not accelerate biological valve degeneration Julie Cleuziou a,⁎, Jürgen Hörer a, Harald Kaemmerer b, Alexandra Teodorowicz a, Jelena Kasnar-Samprec a, Christian Schreiber a, Rüdiger Lange a a b
Department of Cardiovascular Surgery, German Heart Center Munich, Technische Universität München, Germany Department of Pediatric Cardiology and Congenital Heart Disease, German Heart Center Munich, Technische Universität München, Germany
a r t i c l e
i n f o
Article history: Received 1 February 2010 Received in revised form 12 April 2010 Accepted 28 April 2010 Available online 3 June 2010 Keywords: Valves Bioprosthesis Pregnancy
a b s t r a c t Background: Pregnancy in women with biological heart valve prosthesis might lead to a faster degeneration of the prosthesis. We analyzed the rate of prosthesis replacement and the outcome of pregnancies in women with biological valve prosthesis. Methods: Between 1976 and 2006, 100 female patients aged 18–40 years at time of the study, underwent a valve replacement with a bioprosthesis or a homograft. At a mean interval of 10.8 ± 8 years, 87 patients were evaluated by a questionnaire. Results: A biological prosthesis was implanted in 45 patients (52%), while 42 patients (48%) received a homograft. After valve replacement, 33 patients (38%) had a total of 56 pregnancies with a live-birth rate of 77% (n = 43). There were 9 (16%) miscarriages, 4 (7%) therapeutic abortions and no stillbirths. There was no maternal death, thromboembolic event or structural valve deterioration during pregnancy. Out of 87 patients, 31 (36%) required a valve re-replacement at a mean time of 9.5 ± 5 years. Neither age, valve type, valve position nor pregnancy were a risk factor for a valve re-replacement. The freedom from valve rereplacement at 5 and 10 years, was 96 ± 3% and 73 ± 9%, respectively for patients after a pregnancy compared to 93 ± 4% and 52 ± 10%, respectively for patients without a pregnancy (p = 0.2). Conclusions: Pregnancy does not accelerate the risk for replacement of a biological heart valve prosthesis. Pregnancy in women with biological heart valves can be carried out without increased morbidity related to the implanted valve. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Young women requiring a heart valve prosthesis during their childbearing age must decide which type of prosthesis to choose. Mechanical prostheses are known to offer excellent long-term durability [1], but especially during pregnancy and delivery, there is an increased risk of bleeding from permanent anticoagulation or of thromboembolism from inadequate anticoagulation, which is associated with a high mortality rate [2–4]. Furthermore, anticoagulant medication is known to carry a risk for fetal malformations [2]. Bioprostheses have the advantage of not necessitating any anticoagulation but their durability is limited. In addition, faster deterioration of bioprostheses has been reported in young patients and especially during pregnancy [4,5]. However, there is some controversy regarding this topic and these reports are not confirmed by others, who could not associate pregnancy with an increased valve degeneration [6–8]. Nevertheless, re-replacements of bioprostheses
⁎ Corresponding author. Department of Cardiovascular Surgery, German Heart Centre Munich, Lazarettstrasse 36, D-80636 Munich, Germany. Tel.: +49 89 12 18 4111; fax: +49 89 12 18 4123. E-mail address: [email protected] (J. Cleuziou). 0167-5273/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2010.04.095
are unavoidable and early mortality rates between 0% [8] and 10% [9] have been reported. The current policy of the German Heart Centre Munich is to advise our female patients during childbearing age to opt for a bioprosthesis. The current study was designed to scrutinize our practice, focusing on the outcome of pregnancies as well as the rate and the outcome of valve re-replacement to analyze the influence of pregnancies on valve re-replacement. 2. Materials and methods One hundred female patients aged 18–40 years at time of the study, who underwent a valve replacement with a bioprosthesis or a homograft between 1976 and 2006, were identified from the database of the German Heart Centre Munich. A questionnaire was developed to enquire about pregnancy and childbirth as well as cardiac issues and possible re-replacement of the valve prosthesis. Patients who lived abroad or did not consent to respond to the questionnaire were excluded from the study (n = 7). The questionnaire included queries about the number of successful and failed pregnancies, complications during a pregnancy, delivery mode and the health status of the newborn. If complications had occurred during pregnancy, the patient's consent was obtained to contact the attending gynaecologist and enquire for further details. The cardiac condition of the patients was enquired with the questionnaire, focusing on functional class, heart rhythm disturbances, medication and cardiac interventions. Results from the last cardiac examination were obtained from our out-patient department or, if necessary, from the referring cardiologist. If a valve re-replacement
J. Cleuziou et al. / International Journal of Cardiology 145 (2010) 418–421 was required, the indication for surgery and the intraoperative findings were analyzed so that the cause of the re-replacement was indisputable.
419
Table 2 Position of the valve prostheses in 87 patients. Xenograft
Homograft
2.1. Statistical analysis Descriptive statistics are described as frequencies and percentages for categorical variables and as medians with ranges or means with standard deviation for continuous variables. The Kaplan–Meier method was used to estimate the probability of valve rereplacement. Differences between groups were analyzed by the log-rank test. The univariate analysis calculating risk factors for valve re-replacement was done using a Cox regression model. Statistical analysis was done using SPSS statistical software (Version 16.0.1, SPSS Inc., Chicago, IL).
Aortic Pulmonic Tricuspid Mitral Aortic + mitral
n = 22 n = 50 n=8 n=6 n=1
Porcine
Pericardial
12 8 5 4 1
8 2 3 2 0
2 40 0 0 0
3. Results
3.2. Valve re-replacement
Six patients died during follow-up at a mean time of 11 ± 9 years after valve implantation. Five of them had a congenital heart disease and died ultimately from their underlying cardiac defect. One patient with rheumatic fever died of an unknown cause. Eighty seven questionnaires were evaluated at a mean interval of 11 ± 8 years after valve replacement. Mean age at time of valve replacement was 22 ± 9 years (range 3 months–39 years). Sixteen patients (18%) were below 15 years of age and 16 were above 30 years of age. A congenital heart disease was present in 75 patients (86%), whereas 12 patients (14%) had an acquired heart disease (Table 1). Forty five xenografts (52%) and 42 homografts (48%) were implanted. Xenografts included 30 porcine prostheses (67%) and 15 pericardial valves (34%). Homografts were most frequently implanted in the pulmonary position (n = 40, 95%), two were implanted in the aortic position (5%). Xenografts were most frequently implanted in the aortic position (n = 20, 44%), followed by the pulmonary (n = 10, 22%), tricuspid (n = 8, 18%) and mitral position (n = 6, 13%). One patient had a combined procedure with an aortic and mitral valve replacement each with a xenograft. The position of the valve prostheses is shown in Table 2.
Thirty one patients (36%) required a valve re-replacement at a mean time of 9.5 ± 5 years (median 8 years, range 1–21 years). The patient requiring a valve rereplacement at 1 year after the initial surgery was a patient with a Tetralogy of Fallot who presented with a distal stenosis of a homograft in pulmonary position. Most of the valve re-replacements were in pulmonary position (n = 19, 61%), 7 valves were in aortic position (23%), 2 (6%) in tricuspid, 2 (6%) in mitral valve position and one patient (3%) required a combined valve re-replacement in the aortic and mitral position. The major reason for a valve re-replacement was a structural degeneration of the valve prosthesis leading to a stenosis (n = 21, 68%). Eight patients (26%) had a non-structural valve degeneration resulting in a dysfunction but not in a degeneration of the prosthesis. The dysfunction was an incompetence in 6 patients (19%) and a right ventricular outflow tract obstruction below the homograft in 2 patients. Further 2 patients (6%) required a valve re-replacement due to endocarditis. There was no thromboembolism or valve thrombosis. Out of 42 homografts, 11 (26%) required a rereplacement at a mean time of 11 ± 7 years. Sixteen porcine (53%) and 4 pericardial (27%) valves required a re-replacement at a mean time of 9 ± 3 years, and 8 ± 5 years respectively. Only a pericardial valve prosthesis was a significant risk factor for a valve re-replacement (HR 5, CI 1–22, p = 0.01). Neither age, valve type, valve position nor pregnancy was a risk factor for a valve re-replacement (Table 4). A mechanical valve prosthesis was implanted as a second valve prosthesis and definitive solution in 12 patients (39%), 16 patients (52%) received a homograft in pulmonary position and 3 (10%) patients had a re-replacement of their bioprosthesis with a second xenograft prosthesis. The reason for choosing a xenograft during the re-replacement was in one case the tricuspid valve position, one patient was incompliant to oral anticoagulant medication and one patient desired to become pregnant. No patient died early after valve re-replacement. Postoperative morbidity included 3 (9.7%) re-explorations for bleeding and one phrenic nerve injury resulting in a temporary diaphragm paresis. Two patients died 10 and 6 years after valve rereplacement, respectively. None of them had been pregnant. Twelve (39%) of the patients requiring a valve re-replacement had been pregnant. There was no difference in the rate of valve re-replacement for patients having been pregnant and those who had not been pregnant. The freedom from valve re-replacement at 5 and 10 years, was 96 ± 3% and 73 ± 9% for patients after a pregnancy compared to 93 ± 4% and 52 ± 10% for patients without a pregnancy (Fig. 2, p = 0.2).
3.1. Outcome of pregnancies After valve replacement, 33 patients (38%) had a total of 56 pregnancies with a livebirth rate of 77% (n = 43). There were 9 (16%) spontaneous abortions, 4 (7%) therapeutic abortions and no stillbirths (Fig. 1). The reasons for therapeutic abortions were an increased maternal risk from a cardiac prospect in 3 patients and a severe malformation of the fetus in 1 patient. Obstetric complications during pregnancy included severe edema in 2 patients, 1 cervical incompetence requiring a surgical intervention, 1 patient with pre-eclampsia necessitating a premature caesarean section and 2 preterm deliveries triggered by premature labour. No patient was taking anticoagulant treatment during pregnancy. Cardiac complications were limited to rhythm disturbances in 4 patients. There was no maternal death, thromboembolic event or structural valve deterioration during pregnancy. The mode of delivery was left to the discretion of the attending obstetrician, who had the opportunity to consult the attending cardiologist regarding the cardiac risk. A spontaneous delivery was possible in 22 pregnancies (51%), whereas 21 newborns (49%) were delivered by a caesarean section. There were 3 preterm newborns. Newborns were born at a mean of 38.2 gestational weeks (range 36–40 gestational weeks) with a mean birth weight of 2975 g (range 2220–3940 g). Two newborns have a congenital heart defect (one a VSD and the other a mild pulmonary stenosis) and one has a hearing and speaking disorder (Table 3). Table 1 Congenital and acquired diseases requiring a valve replacement in 87 patients. Diagnosis
n (%)
Tetralogy of Fallot Congenital aortic valve disease Pulmonary atresia with VSD Ebstein's anomaly Congenital pulmonary stenosis Common arterial trunc TGA with VSD and PS DORV with PS Tricuspid valve disease Mitral valve prolaps CAVSD CCTGA with VSD and PS Sinus valsalva aneurysm Total
23 (26) 12 (14) 9 (10) 5 (6) 5 (6) 5 (6) 5 (6) 3 (3) 2 (3) 2 (2) 2 (2) 1 (1) 1 (1) 75 (86)
Rheumatic fever Infective endocarditis Aortic dissection Total
8 (9) 3 (3) 1 (1) 12 (14)
4. Discussion To reduce valve-related thromboembolism or bleeding as well as the risk of teratogenic effects in a patient during childbearing age, a bioprosthetic valve is currently the best option for valve replacement, although the limited durability of valve bioprostheses and some reports of faster degeneration of bioprosthetic valves, especially in younger patients as well as after a pregnancy are of concern. In addition, the mortality rate after the unavoidable re-replacement of a bioprosthesis needs to be addressed. The major finding of this study is that the outcome of pregnancy for patients with bioprostheses is good and that pregnancy does not accelerate the degeneration of bioprostheses. Furthermore, morbidity and mortality after valve re-replacement were very low in this study. Valve re-replacement was neither dependent on valve position, valve type nor patient age. The only significant risk factor for a valve re-replacement was the implantation of a pericardial valve. Although this finding is statistically significant, it needs to be addressed with some reservation because of the low rate of events and the multiple co-variates. Only 4 out of 15 patients with a pericardial valve had a valve re-replacement. This result might not stay significant with a higher number of events [10]. In large studies, pericardial valves show better results in longterm durability compared to porcine valves [11].
420
J. Cleuziou et al. / International Journal of Cardiology 145 (2010) 418–421
Fig. 1. Patients with a biological valve replacement and outcome of their pregnancies.
Accelerated degeneration of bioprosthetic valves in younger patients is a phenomenon which has been described by various authors [4,5,8,9]. In patients below 40 years of age, freedom from structural valve deterioration and reoperation ranges from 50% [12], 38% [8] down to 23% [13] at 10 years after implantation of a bioprosthesis. Although we could not demonstrate an influence of age on the rate of valve re-replacement, it was as high as 52% at 10 years after valve implantation in this population of young women with a mean age of 22 years. In patients with a mean age of 65 years, 20 year durability of porcine bioprostheses is described to be as high as 79% after implantation in a mitral valve position [14]. There might be a statistical significant influence of age in our population with a higher number of patients. Bortolotti et al. found that besides age, pregnancy was a risk factor for structural deterioration of bioprostheses [5]. In a study including 49 patients with bioprostheses, 35% of the valves needed to be replaced shortly after a pregnancy and 2 patients required urgent rereplacement during the pregnancy. This study from the year 1994 relates to patients who became pregnant mostly after 1985, and it can be presumed, that the newer generation of bioprostheses is better and will last longer as seen in our study [4]. Badduke et al. found that patients with bioprostheses who had been pregnant required more often a valve re-replacement. However, adding the age into the statistical analysis resulted in no difference for structural valve deterioration between patients who had been pregnant and those who had not [13]. Hanania et al. found premature failures of bioprosthetic valves in 5% of patients at a mean time of 1.5 years after implantation and attributed these to pregnancies [2].
Table 3 Outcome of pregnancies, delivery mode, maternal complications and neonatal outcome.
Pregnancy, n (%) Spontaneous abortion, n (%) Therapeutic abortion, n (%) Spontaneous delivery, n (%) Caesarean section, n (%) Obstetric complications, n (%) Cardiac complications, n (%) Gestational week, mean [range] Neonatal weight, mean, g Neonatal disease, n (%)
Xenograft n = 45
Homograft n = 42
33 (73) 5 (15) 3 (9) 14 (58) 10 (44) 0 1 (4) 38.5 [35–40] 3100 ± 480 1 (4)
23 (55) 4 (17) 1 (4) 10 (43) 8 (35) 5 (22) 3 (13) 38.3 [36–41] 2840 ± 450 2 (11)
All in all, reports are not consistent and our results do not support these findings. In our study, no patient required urgent valve rereplacement during or shortly after pregnancy. In addition, we could not identify pregnancy as a risk factor for valve re-replacement. It can be criticized that our patient population is inhomogeneous, because we included patients with different types of bioprostheses including homografts. However, if, as suggested by some authors, the high calcium turnover during pregnancy is the cause for the accelerated degeneration of bioprostheses [4,6,13], this effect should be valid in any bioprosthesis. Furthermore, our results demonstrated no significant difference in valve degeneration between homografts and xenografts, demonstrating comparable results between both. In a homogeneous population of young women with acquired valve disease and pericardial valve prostheses, also Salazar could not find any difference in valve degeneration between patients who had been pregnant and those who had not [9]. In a previous study homografts showed excellent outcome of pregnancy with no valve-related complications [3]. There are other reports showing that pregnancy did not increase structural deterioration or reduce survival of bioprosthetic valves [7,8]. Looking for more details in these conflicting literature, a remarkable feature is the date of the reports. Compared to older publications, more recent ones do not find an accelerated degeneration of bioprosthesis. This might show the better outcome of newer generations of bioprostheses. The inevitability of a redo-operation in patients with bioprostheses is a subject of discussion. Sbaorouni and Oakley recommend to avoid a
Table 4 Risk factors for valve re-replacement in 87 female patients. Variables
Age b 15 years Congenital heart disease Aortic valve position Mitral valve position Tricuspid valve position Pulmonary valve position Porcine valve prosthesis Pericardial valve prosthesis Homograft Previous operation Pregnancy
Univariate analysis Hazard rate
95% CI
p
3 1 0.1 0.2 0.06 0.1 2 5 0.5 0.8 0.7
0.5–18 0.2–4 0.01–1.2 0.02–3 0.003–1.3 0.01–2 0.9–6 1–22 0.2–1 0.3–2 0.3–2
0.2 0.9 0.08 0.2 0.07 0.2 0.08 0.01 0.1 0.7 0.4
J. Cleuziou et al. / International Journal of Cardiology 145 (2010) 418–421
421
[8] quoted in the literature. A caesarean section is more often indicated due to an obstetric than a cardiologic indication [17]. 4.1. Conclusion Pregnancy has no adverse effects on the degeneration of bioprosthetic valves. The present study shows that the outcome of pregnancies in women with a bioprosthesis is good and usually uncomplicated. Young women who desire to have children and need a heart valve replacement should opt for a bioprosthesis. Nevertheless, the outcome of a pregnancy in patients with a heart valve prosthesis will depend on their underlying heart disease and their cardiac wellbeing. And therefore these patients need thorough and steady support through specialists from before conception until after delivery. 4.2. Study limitations Fig. 2. Freedom from valve replacement for 87 patients with a biological valve replacement depending on a pregnancy.
bioprosthesis in young women who wish to become pregnant, as the risk for a re-operation is increased. But they don't mention their morbidity and mortality after re-replacement of a bioprosthesis [4]. In our patient population, there were no early deaths after valve rereplacement. This is consistent with a report from the year 2000, where no patient died after valve re-replacement [8]. On the other hand, after re-replacement of a bioprosthesis, other authors report a mortality rate between 2.7% [6] and 8.7% [13]. Although this rate is high, it reflects another era of cardiac surgery, as these patients received their valve replacement between 1975 and 1987. There is a growing experience in redo-operations, especially in centres treating young adults with congenital heart disease. Mortality in adults with congenital heart disease after a redo-operation is as high as 7.6% [15] and is associated with cyanosis and the number of previous operations [16]. Another limiting factor in these patients is the nature of their underlying heart disease and the presence of chronic heart failure, which will be a strong factor on deciding to conceive and on the outcome of a pregnancy. Live-birth rate in our study was 77% with 2 newborns having a congenital heart disease. This result is comparable to other reports with live-birth rates in women with bioprostheses ranging from 74.5% [6] to 83% [4]. The rate of abortions is excluded in some studies, as spontaneous abortions could be missed and their causes can be multifactorial. We detected a rate of 16% of spontaneous abortions, while other reports describe rates from 5% [17], 11% [6] up to 22% [9]. However, the rate of therapeutic abortions can be an evidence of maternal distress or fetal malformation and needs to be addressed. Out of 4 therapeutic abortions in our population, 3 were due to maternal risk. Jamieson et al. reported a therapeutic abortion rate of nearly 15% [6]. Stillbirths are rare [17] and were not experienced in our patients. Premature newborns though, can occur in up to 13% [17] and were present in 3 newborns (7%) in the present study, although none of them was born before a gestational age of 36 weeks and their birth weight was not below 2000 g. Sadler et al. found that cardiac complications were associated with preterm delivery [3]. Maternal outcome in our study was excellent, with a low rate of obstetric and cardiac complications. Some large studies show that the outcome of pregnancy in women with cardiac disease is dependent on the underlying cardiac condition and carries a mortality as high as 2.7% [17]. The most common cardiac complication described in patients after valve replacement is cardiac failure [3]. The mode of delivery is strongly dependent on the decision of the attending obstetrician as well as established regional and era-depending differences. In our patients, just more than half of the newborns were delivered spontaneously, which is below the rates of 60% [9] up to 81%
This is a retrospective study over a long period of time. Changes in surgical and postoperative management might have influenced the results and cannot be ruled out. Furthermore, the patient population is heterogeneous and includes patients with acquired and congenital heart valve diseases. Nevertheless, this should not impact the results as the goal of the study was to analyze the influence of pregnancy on biological valves, independent of the underlying cardiac disease. Acknowledgment The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [18]. References [1] Remadi JP, Baron O, Roussel C, et al. Isolated mitral valve replacement with St Jude Medical prosthesis. Circulation 2001;103:1542–5. [2] Hanania G, Thomas D, Michel PL, et al. Pregnancy and prosthetic heart valves: a French cooperative retrospective study of 155 cases. Eur Heart J 1994;15:1651–8. [3] Sadler L, McCowan L, White H, Stewart A, Bracken M, North R. Pregnancy outcomes and cardiac complications in women with mechanical, bioprosthetic and homograft valves. Br J Obstet Gyn 2000;107(2):245–53. [4] Sbarouni E, Oakley CM. Outcome of pregnancy in women with valve prosthesis. Br Heart J 1994;71:196–201. [5] Bortolotti U, Milano A, Mazzuco A, et al. Pregnancy in patients with a porcine valve bioprosthesis. Am J Cardiol 1982;50:1051–4. [6] Jamieson WRE, Miller DC, Akins CW, et al. Pregnancy and bioprostheses: influence on structural valve deterioration. Ann Thorac Surg 1995;60:S282–287. [7] North RA, Sadler L, Stewart AW, McCowan LME, Kerr AR, White HD. Long-term survival and valve-related complications in young women with cardiac valve replacements. Circulation 1999;99:2669–76. [8] Mihaljevic T, Paul S, Leacche M, Rawn JD, Cohn LH, Byrne JG. Valve replacement in women of childbearing age: influences on mother, fetus and neonate. J Heart Valve Dis 2005;14:151–7. [9] Salazar E, Espinola N, Roman L, Casanova JM. Effect of pregnancy on the duration of bovine pericardial bioprostheses. Am Heart J 1999;137:714–20. [10] Concato J, Peduzzi P, Holford TR, Feinstein AR. Importance of events per independent variable in proportional hazards analysis. Background, goals and general strategy. J Clin Epidemiol 1995;48:1495–501. [11] Gao G, Wu YX, Grunkemeier GL, Furnary AP, Starr A. Durability of pericardial versus porcine aortic valves. JACC 2004;44(2):384–8. [12] Yun KL, Miller DC, Moore KA, et al. Durability of the Hancock MO bioprosthesis compared with standard aortic valve bioprostheses. Ann Thorac Surg 1995;60: S221–228. [13] Badduke ER, Jamieson WR, Miyagishima RT. Pregnancy and childbearing in a population with biologic valvular prostheses. J Thorac Cardiovasc Surg 1991;102: 179–86. [14] Myken PSU, Bech-Hanssen O. A 20-year experience of 1712 patients with the Biocor porcine bioprosthesis. J Thorac Cardiovasc Surg 2009;137:76–81. [15] Berdat PA, Immer F, Pfammatter JP, Carrel T. Reoperations in adults with congenital heart disease: analysis of early outcome. Int J Cardiol 2004;93:239–45. [16] Dore A, Glancy DL, Stone S, Menashe VD, Somerville J. Cardiac surgery for grown-up with congenital heart patients: survey of 307 consecutive operations from 1991 10 1994. Am J Cardiol 1997;80:906–13. [17] Avila WS, Rossi EG, Ramires JAF, et al. Pregnancy in patients with heart disease: experience with 1000 cases. Clin Cardiol 2003;26:135–42. [18] Coats AJ. Ethical authorship and publishing. Int J Cardiol 2009;131:149–50.
Contents lists available at ScienceDirect
International Journal of Cardiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j c a r d
Pregnancy does not accelerate biological valve degeneration Julie Cleuziou a,⁎, Jürgen Hörer a, Harald Kaemmerer b, Alexandra Teodorowicz a, Jelena Kasnar-Samprec a, Christian Schreiber a, Rüdiger Lange a a b
Department of Cardiovascular Surgery, German Heart Center Munich, Technische Universität München, Germany Department of Pediatric Cardiology and Congenital Heart Disease, German Heart Center Munich, Technische Universität München, Germany
a r t i c l e
i n f o
Article history: Received 1 February 2010 Received in revised form 12 April 2010 Accepted 28 April 2010 Available online 3 June 2010 Keywords: Valves Bioprosthesis Pregnancy
a b s t r a c t Background: Pregnancy in women with biological heart valve prosthesis might lead to a faster degeneration of the prosthesis. We analyzed the rate of prosthesis replacement and the outcome of pregnancies in women with biological valve prosthesis. Methods: Between 1976 and 2006, 100 female patients aged 18–40 years at time of the study, underwent a valve replacement with a bioprosthesis or a homograft. At a mean interval of 10.8 ± 8 years, 87 patients were evaluated by a questionnaire. Results: A biological prosthesis was implanted in 45 patients (52%), while 42 patients (48%) received a homograft. After valve replacement, 33 patients (38%) had a total of 56 pregnancies with a live-birth rate of 77% (n = 43). There were 9 (16%) miscarriages, 4 (7%) therapeutic abortions and no stillbirths. There was no maternal death, thromboembolic event or structural valve deterioration during pregnancy. Out of 87 patients, 31 (36%) required a valve re-replacement at a mean time of 9.5 ± 5 years. Neither age, valve type, valve position nor pregnancy were a risk factor for a valve re-replacement. The freedom from valve rereplacement at 5 and 10 years, was 96 ± 3% and 73 ± 9%, respectively for patients after a pregnancy compared to 93 ± 4% and 52 ± 10%, respectively for patients without a pregnancy (p = 0.2). Conclusions: Pregnancy does not accelerate the risk for replacement of a biological heart valve prosthesis. Pregnancy in women with biological heart valves can be carried out without increased morbidity related to the implanted valve. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Young women requiring a heart valve prosthesis during their childbearing age must decide which type of prosthesis to choose. Mechanical prostheses are known to offer excellent long-term durability [1], but especially during pregnancy and delivery, there is an increased risk of bleeding from permanent anticoagulation or of thromboembolism from inadequate anticoagulation, which is associated with a high mortality rate [2–4]. Furthermore, anticoagulant medication is known to carry a risk for fetal malformations [2]. Bioprostheses have the advantage of not necessitating any anticoagulation but their durability is limited. In addition, faster deterioration of bioprostheses has been reported in young patients and especially during pregnancy [4,5]. However, there is some controversy regarding this topic and these reports are not confirmed by others, who could not associate pregnancy with an increased valve degeneration [6–8]. Nevertheless, re-replacements of bioprostheses
⁎ Corresponding author. Department of Cardiovascular Surgery, German Heart Centre Munich, Lazarettstrasse 36, D-80636 Munich, Germany. Tel.: +49 89 12 18 4111; fax: +49 89 12 18 4123. E-mail address: [email protected] (J. Cleuziou). 0167-5273/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2010.04.095
are unavoidable and early mortality rates between 0% [8] and 10% [9] have been reported. The current policy of the German Heart Centre Munich is to advise our female patients during childbearing age to opt for a bioprosthesis. The current study was designed to scrutinize our practice, focusing on the outcome of pregnancies as well as the rate and the outcome of valve re-replacement to analyze the influence of pregnancies on valve re-replacement. 2. Materials and methods One hundred female patients aged 18–40 years at time of the study, who underwent a valve replacement with a bioprosthesis or a homograft between 1976 and 2006, were identified from the database of the German Heart Centre Munich. A questionnaire was developed to enquire about pregnancy and childbirth as well as cardiac issues and possible re-replacement of the valve prosthesis. Patients who lived abroad or did not consent to respond to the questionnaire were excluded from the study (n = 7). The questionnaire included queries about the number of successful and failed pregnancies, complications during a pregnancy, delivery mode and the health status of the newborn. If complications had occurred during pregnancy, the patient's consent was obtained to contact the attending gynaecologist and enquire for further details. The cardiac condition of the patients was enquired with the questionnaire, focusing on functional class, heart rhythm disturbances, medication and cardiac interventions. Results from the last cardiac examination were obtained from our out-patient department or, if necessary, from the referring cardiologist. If a valve re-replacement
J. Cleuziou et al. / International Journal of Cardiology 145 (2010) 418–421 was required, the indication for surgery and the intraoperative findings were analyzed so that the cause of the re-replacement was indisputable.
419
Table 2 Position of the valve prostheses in 87 patients. Xenograft
Homograft
2.1. Statistical analysis Descriptive statistics are described as frequencies and percentages for categorical variables and as medians with ranges or means with standard deviation for continuous variables. The Kaplan–Meier method was used to estimate the probability of valve rereplacement. Differences between groups were analyzed by the log-rank test. The univariate analysis calculating risk factors for valve re-replacement was done using a Cox regression model. Statistical analysis was done using SPSS statistical software (Version 16.0.1, SPSS Inc., Chicago, IL).
Aortic Pulmonic Tricuspid Mitral Aortic + mitral
n = 22 n = 50 n=8 n=6 n=1
Porcine
Pericardial
12 8 5 4 1
8 2 3 2 0
2 40 0 0 0
3. Results
3.2. Valve re-replacement
Six patients died during follow-up at a mean time of 11 ± 9 years after valve implantation. Five of them had a congenital heart disease and died ultimately from their underlying cardiac defect. One patient with rheumatic fever died of an unknown cause. Eighty seven questionnaires were evaluated at a mean interval of 11 ± 8 years after valve replacement. Mean age at time of valve replacement was 22 ± 9 years (range 3 months–39 years). Sixteen patients (18%) were below 15 years of age and 16 were above 30 years of age. A congenital heart disease was present in 75 patients (86%), whereas 12 patients (14%) had an acquired heart disease (Table 1). Forty five xenografts (52%) and 42 homografts (48%) were implanted. Xenografts included 30 porcine prostheses (67%) and 15 pericardial valves (34%). Homografts were most frequently implanted in the pulmonary position (n = 40, 95%), two were implanted in the aortic position (5%). Xenografts were most frequently implanted in the aortic position (n = 20, 44%), followed by the pulmonary (n = 10, 22%), tricuspid (n = 8, 18%) and mitral position (n = 6, 13%). One patient had a combined procedure with an aortic and mitral valve replacement each with a xenograft. The position of the valve prostheses is shown in Table 2.
Thirty one patients (36%) required a valve re-replacement at a mean time of 9.5 ± 5 years (median 8 years, range 1–21 years). The patient requiring a valve rereplacement at 1 year after the initial surgery was a patient with a Tetralogy of Fallot who presented with a distal stenosis of a homograft in pulmonary position. Most of the valve re-replacements were in pulmonary position (n = 19, 61%), 7 valves were in aortic position (23%), 2 (6%) in tricuspid, 2 (6%) in mitral valve position and one patient (3%) required a combined valve re-replacement in the aortic and mitral position. The major reason for a valve re-replacement was a structural degeneration of the valve prosthesis leading to a stenosis (n = 21, 68%). Eight patients (26%) had a non-structural valve degeneration resulting in a dysfunction but not in a degeneration of the prosthesis. The dysfunction was an incompetence in 6 patients (19%) and a right ventricular outflow tract obstruction below the homograft in 2 patients. Further 2 patients (6%) required a valve re-replacement due to endocarditis. There was no thromboembolism or valve thrombosis. Out of 42 homografts, 11 (26%) required a rereplacement at a mean time of 11 ± 7 years. Sixteen porcine (53%) and 4 pericardial (27%) valves required a re-replacement at a mean time of 9 ± 3 years, and 8 ± 5 years respectively. Only a pericardial valve prosthesis was a significant risk factor for a valve re-replacement (HR 5, CI 1–22, p = 0.01). Neither age, valve type, valve position nor pregnancy was a risk factor for a valve re-replacement (Table 4). A mechanical valve prosthesis was implanted as a second valve prosthesis and definitive solution in 12 patients (39%), 16 patients (52%) received a homograft in pulmonary position and 3 (10%) patients had a re-replacement of their bioprosthesis with a second xenograft prosthesis. The reason for choosing a xenograft during the re-replacement was in one case the tricuspid valve position, one patient was incompliant to oral anticoagulant medication and one patient desired to become pregnant. No patient died early after valve re-replacement. Postoperative morbidity included 3 (9.7%) re-explorations for bleeding and one phrenic nerve injury resulting in a temporary diaphragm paresis. Two patients died 10 and 6 years after valve rereplacement, respectively. None of them had been pregnant. Twelve (39%) of the patients requiring a valve re-replacement had been pregnant. There was no difference in the rate of valve re-replacement for patients having been pregnant and those who had not been pregnant. The freedom from valve re-replacement at 5 and 10 years, was 96 ± 3% and 73 ± 9% for patients after a pregnancy compared to 93 ± 4% and 52 ± 10% for patients without a pregnancy (Fig. 2, p = 0.2).
3.1. Outcome of pregnancies After valve replacement, 33 patients (38%) had a total of 56 pregnancies with a livebirth rate of 77% (n = 43). There were 9 (16%) spontaneous abortions, 4 (7%) therapeutic abortions and no stillbirths (Fig. 1). The reasons for therapeutic abortions were an increased maternal risk from a cardiac prospect in 3 patients and a severe malformation of the fetus in 1 patient. Obstetric complications during pregnancy included severe edema in 2 patients, 1 cervical incompetence requiring a surgical intervention, 1 patient with pre-eclampsia necessitating a premature caesarean section and 2 preterm deliveries triggered by premature labour. No patient was taking anticoagulant treatment during pregnancy. Cardiac complications were limited to rhythm disturbances in 4 patients. There was no maternal death, thromboembolic event or structural valve deterioration during pregnancy. The mode of delivery was left to the discretion of the attending obstetrician, who had the opportunity to consult the attending cardiologist regarding the cardiac risk. A spontaneous delivery was possible in 22 pregnancies (51%), whereas 21 newborns (49%) were delivered by a caesarean section. There were 3 preterm newborns. Newborns were born at a mean of 38.2 gestational weeks (range 36–40 gestational weeks) with a mean birth weight of 2975 g (range 2220–3940 g). Two newborns have a congenital heart defect (one a VSD and the other a mild pulmonary stenosis) and one has a hearing and speaking disorder (Table 3). Table 1 Congenital and acquired diseases requiring a valve replacement in 87 patients. Diagnosis
n (%)
Tetralogy of Fallot Congenital aortic valve disease Pulmonary atresia with VSD Ebstein's anomaly Congenital pulmonary stenosis Common arterial trunc TGA with VSD and PS DORV with PS Tricuspid valve disease Mitral valve prolaps CAVSD CCTGA with VSD and PS Sinus valsalva aneurysm Total
23 (26) 12 (14) 9 (10) 5 (6) 5 (6) 5 (6) 5 (6) 3 (3) 2 (3) 2 (2) 2 (2) 1 (1) 1 (1) 75 (86)
Rheumatic fever Infective endocarditis Aortic dissection Total
8 (9) 3 (3) 1 (1) 12 (14)
4. Discussion To reduce valve-related thromboembolism or bleeding as well as the risk of teratogenic effects in a patient during childbearing age, a bioprosthetic valve is currently the best option for valve replacement, although the limited durability of valve bioprostheses and some reports of faster degeneration of bioprosthetic valves, especially in younger patients as well as after a pregnancy are of concern. In addition, the mortality rate after the unavoidable re-replacement of a bioprosthesis needs to be addressed. The major finding of this study is that the outcome of pregnancy for patients with bioprostheses is good and that pregnancy does not accelerate the degeneration of bioprostheses. Furthermore, morbidity and mortality after valve re-replacement were very low in this study. Valve re-replacement was neither dependent on valve position, valve type nor patient age. The only significant risk factor for a valve re-replacement was the implantation of a pericardial valve. Although this finding is statistically significant, it needs to be addressed with some reservation because of the low rate of events and the multiple co-variates. Only 4 out of 15 patients with a pericardial valve had a valve re-replacement. This result might not stay significant with a higher number of events [10]. In large studies, pericardial valves show better results in longterm durability compared to porcine valves [11].
420
J. Cleuziou et al. / International Journal of Cardiology 145 (2010) 418–421
Fig. 1. Patients with a biological valve replacement and outcome of their pregnancies.
Accelerated degeneration of bioprosthetic valves in younger patients is a phenomenon which has been described by various authors [4,5,8,9]. In patients below 40 years of age, freedom from structural valve deterioration and reoperation ranges from 50% [12], 38% [8] down to 23% [13] at 10 years after implantation of a bioprosthesis. Although we could not demonstrate an influence of age on the rate of valve re-replacement, it was as high as 52% at 10 years after valve implantation in this population of young women with a mean age of 22 years. In patients with a mean age of 65 years, 20 year durability of porcine bioprostheses is described to be as high as 79% after implantation in a mitral valve position [14]. There might be a statistical significant influence of age in our population with a higher number of patients. Bortolotti et al. found that besides age, pregnancy was a risk factor for structural deterioration of bioprostheses [5]. In a study including 49 patients with bioprostheses, 35% of the valves needed to be replaced shortly after a pregnancy and 2 patients required urgent rereplacement during the pregnancy. This study from the year 1994 relates to patients who became pregnant mostly after 1985, and it can be presumed, that the newer generation of bioprostheses is better and will last longer as seen in our study [4]. Badduke et al. found that patients with bioprostheses who had been pregnant required more often a valve re-replacement. However, adding the age into the statistical analysis resulted in no difference for structural valve deterioration between patients who had been pregnant and those who had not [13]. Hanania et al. found premature failures of bioprosthetic valves in 5% of patients at a mean time of 1.5 years after implantation and attributed these to pregnancies [2].
Table 3 Outcome of pregnancies, delivery mode, maternal complications and neonatal outcome.
Pregnancy, n (%) Spontaneous abortion, n (%) Therapeutic abortion, n (%) Spontaneous delivery, n (%) Caesarean section, n (%) Obstetric complications, n (%) Cardiac complications, n (%) Gestational week, mean [range] Neonatal weight, mean, g Neonatal disease, n (%)
Xenograft n = 45
Homograft n = 42
33 (73) 5 (15) 3 (9) 14 (58) 10 (44) 0 1 (4) 38.5 [35–40] 3100 ± 480 1 (4)
23 (55) 4 (17) 1 (4) 10 (43) 8 (35) 5 (22) 3 (13) 38.3 [36–41] 2840 ± 450 2 (11)
All in all, reports are not consistent and our results do not support these findings. In our study, no patient required urgent valve rereplacement during or shortly after pregnancy. In addition, we could not identify pregnancy as a risk factor for valve re-replacement. It can be criticized that our patient population is inhomogeneous, because we included patients with different types of bioprostheses including homografts. However, if, as suggested by some authors, the high calcium turnover during pregnancy is the cause for the accelerated degeneration of bioprostheses [4,6,13], this effect should be valid in any bioprosthesis. Furthermore, our results demonstrated no significant difference in valve degeneration between homografts and xenografts, demonstrating comparable results between both. In a homogeneous population of young women with acquired valve disease and pericardial valve prostheses, also Salazar could not find any difference in valve degeneration between patients who had been pregnant and those who had not [9]. In a previous study homografts showed excellent outcome of pregnancy with no valve-related complications [3]. There are other reports showing that pregnancy did not increase structural deterioration or reduce survival of bioprosthetic valves [7,8]. Looking for more details in these conflicting literature, a remarkable feature is the date of the reports. Compared to older publications, more recent ones do not find an accelerated degeneration of bioprosthesis. This might show the better outcome of newer generations of bioprostheses. The inevitability of a redo-operation in patients with bioprostheses is a subject of discussion. Sbaorouni and Oakley recommend to avoid a
Table 4 Risk factors for valve re-replacement in 87 female patients. Variables
Age b 15 years Congenital heart disease Aortic valve position Mitral valve position Tricuspid valve position Pulmonary valve position Porcine valve prosthesis Pericardial valve prosthesis Homograft Previous operation Pregnancy
Univariate analysis Hazard rate
95% CI
p
3 1 0.1 0.2 0.06 0.1 2 5 0.5 0.8 0.7
0.5–18 0.2–4 0.01–1.2 0.02–3 0.003–1.3 0.01–2 0.9–6 1–22 0.2–1 0.3–2 0.3–2
0.2 0.9 0.08 0.2 0.07 0.2 0.08 0.01 0.1 0.7 0.4
J. Cleuziou et al. / International Journal of Cardiology 145 (2010) 418–421
421
[8] quoted in the literature. A caesarean section is more often indicated due to an obstetric than a cardiologic indication [17]. 4.1. Conclusion Pregnancy has no adverse effects on the degeneration of bioprosthetic valves. The present study shows that the outcome of pregnancies in women with a bioprosthesis is good and usually uncomplicated. Young women who desire to have children and need a heart valve replacement should opt for a bioprosthesis. Nevertheless, the outcome of a pregnancy in patients with a heart valve prosthesis will depend on their underlying heart disease and their cardiac wellbeing. And therefore these patients need thorough and steady support through specialists from before conception until after delivery. 4.2. Study limitations Fig. 2. Freedom from valve replacement for 87 patients with a biological valve replacement depending on a pregnancy.
bioprosthesis in young women who wish to become pregnant, as the risk for a re-operation is increased. But they don't mention their morbidity and mortality after re-replacement of a bioprosthesis [4]. In our patient population, there were no early deaths after valve rereplacement. This is consistent with a report from the year 2000, where no patient died after valve re-replacement [8]. On the other hand, after re-replacement of a bioprosthesis, other authors report a mortality rate between 2.7% [6] and 8.7% [13]. Although this rate is high, it reflects another era of cardiac surgery, as these patients received their valve replacement between 1975 and 1987. There is a growing experience in redo-operations, especially in centres treating young adults with congenital heart disease. Mortality in adults with congenital heart disease after a redo-operation is as high as 7.6% [15] and is associated with cyanosis and the number of previous operations [16]. Another limiting factor in these patients is the nature of their underlying heart disease and the presence of chronic heart failure, which will be a strong factor on deciding to conceive and on the outcome of a pregnancy. Live-birth rate in our study was 77% with 2 newborns having a congenital heart disease. This result is comparable to other reports with live-birth rates in women with bioprostheses ranging from 74.5% [6] to 83% [4]. The rate of abortions is excluded in some studies, as spontaneous abortions could be missed and their causes can be multifactorial. We detected a rate of 16% of spontaneous abortions, while other reports describe rates from 5% [17], 11% [6] up to 22% [9]. However, the rate of therapeutic abortions can be an evidence of maternal distress or fetal malformation and needs to be addressed. Out of 4 therapeutic abortions in our population, 3 were due to maternal risk. Jamieson et al. reported a therapeutic abortion rate of nearly 15% [6]. Stillbirths are rare [17] and were not experienced in our patients. Premature newborns though, can occur in up to 13% [17] and were present in 3 newborns (7%) in the present study, although none of them was born before a gestational age of 36 weeks and their birth weight was not below 2000 g. Sadler et al. found that cardiac complications were associated with preterm delivery [3]. Maternal outcome in our study was excellent, with a low rate of obstetric and cardiac complications. Some large studies show that the outcome of pregnancy in women with cardiac disease is dependent on the underlying cardiac condition and carries a mortality as high as 2.7% [17]. The most common cardiac complication described in patients after valve replacement is cardiac failure [3]. The mode of delivery is strongly dependent on the decision of the attending obstetrician as well as established regional and era-depending differences. In our patients, just more than half of the newborns were delivered spontaneously, which is below the rates of 60% [9] up to 81%
This is a retrospective study over a long period of time. Changes in surgical and postoperative management might have influenced the results and cannot be ruled out. Furthermore, the patient population is heterogeneous and includes patients with acquired and congenital heart valve diseases. Nevertheless, this should not impact the results as the goal of the study was to analyze the influence of pregnancy on biological valves, independent of the underlying cardiac disease. Acknowledgment The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [18]. References [1] Remadi JP, Baron O, Roussel C, et al. Isolated mitral valve replacement with St Jude Medical prosthesis. Circulation 2001;103:1542–5. [2] Hanania G, Thomas D, Michel PL, et al. Pregnancy and prosthetic heart valves: a French cooperative retrospective study of 155 cases. Eur Heart J 1994;15:1651–8. [3] Sadler L, McCowan L, White H, Stewart A, Bracken M, North R. Pregnancy outcomes and cardiac complications in women with mechanical, bioprosthetic and homograft valves. Br J Obstet Gyn 2000;107(2):245–53. [4] Sbarouni E, Oakley CM. Outcome of pregnancy in women with valve prosthesis. Br Heart J 1994;71:196–201. [5] Bortolotti U, Milano A, Mazzuco A, et al. Pregnancy in patients with a porcine valve bioprosthesis. Am J Cardiol 1982;50:1051–4. [6] Jamieson WRE, Miller DC, Akins CW, et al. Pregnancy and bioprostheses: influence on structural valve deterioration. Ann Thorac Surg 1995;60:S282–287. [7] North RA, Sadler L, Stewart AW, McCowan LME, Kerr AR, White HD. Long-term survival and valve-related complications in young women with cardiac valve replacements. Circulation 1999;99:2669–76. [8] Mihaljevic T, Paul S, Leacche M, Rawn JD, Cohn LH, Byrne JG. Valve replacement in women of childbearing age: influences on mother, fetus and neonate. J Heart Valve Dis 2005;14:151–7. [9] Salazar E, Espinola N, Roman L, Casanova JM. Effect of pregnancy on the duration of bovine pericardial bioprostheses. Am Heart J 1999;137:714–20. [10] Concato J, Peduzzi P, Holford TR, Feinstein AR. Importance of events per independent variable in proportional hazards analysis. Background, goals and general strategy. J Clin Epidemiol 1995;48:1495–501. [11] Gao G, Wu YX, Grunkemeier GL, Furnary AP, Starr A. Durability of pericardial versus porcine aortic valves. JACC 2004;44(2):384–8. [12] Yun KL, Miller DC, Moore KA, et al. Durability of the Hancock MO bioprosthesis compared with standard aortic valve bioprostheses. Ann Thorac Surg 1995;60: S221–228. [13] Badduke ER, Jamieson WR, Miyagishima RT. Pregnancy and childbearing in a population with biologic valvular prostheses. J Thorac Cardiovasc Surg 1991;102: 179–86. [14] Myken PSU, Bech-Hanssen O. A 20-year experience of 1712 patients with the Biocor porcine bioprosthesis. J Thorac Cardiovasc Surg 2009;137:76–81. [15] Berdat PA, Immer F, Pfammatter JP, Carrel T. Reoperations in adults with congenital heart disease: analysis of early outcome. Int J Cardiol 2004;93:239–45. [16] Dore A, Glancy DL, Stone S, Menashe VD, Somerville J. Cardiac surgery for grown-up with congenital heart patients: survey of 307 consecutive operations from 1991 10 1994. Am J Cardiol 1997;80:906–13. [17] Avila WS, Rossi EG, Ramires JAF, et al. Pregnancy in patients with heart disease: experience with 1000 cases. Clin Cardiol 2003;26:135–42. [18] Coats AJ. Ethical authorship and publishing. Int J Cardiol 2009;131:149–50.