- Email: [email protected]
Experience With Pericardiectomy for Constrictive Pericarditis Over Eight Decades
Experience With Pericardiectomy for Constrictive Pericarditis Over Eight Decades Takashi Murashita, MD, Hartzell V. Schaff, MD, Richard C. Daly, MD, Jae K. Oh, MD, Joseph A. Dearani, MD, John M. Stulak, MD, Katherine S. King, MS, and Kevin L. Greason, MD Departments of Cardiovascular Surgery, Cardiovascular Diseases, and Health Sciences Research, Mayo Clinic, Rochester, Minnesota
Background. The purpose of this study was to review the surgical outcomes of pericardiectomy for constrictive pericarditis and to examine risk factors for overall mortality in a contemporary period. Methods. We reviewed all patients who underwent pericardiectomy for constriction from 1936 through 2013. The investigation included constrictive pericarditis cases confirmed intraoperatively, all other types of pericarditis were excluded; 1,071 pericardiectomies were performed in 1,066 individual patients. Patients were divided into two intervals: a historical (pre-1990) group (n [ 259) and a contemporary (1990–2013) group (n [ 807). Results. Patients in the contemporary group were older (61 versus 49 years; p < 0.001), more symptomatic (NYHA class III or IV in 79.6% versus 71.2%; p < 0.001), and more frequently underwent concomitant procedures (21.4% versus 5.4%; p < 0.001) compared with those in the historical group. In contrast to the historical cases in which the etiologies of constriction were mostly
idiopathic (81.1%), nearly half of contemporary cases had a nonidiopathic etiology (postoperative 32.3%, radiation 11.4%). Although 30-day mortality decreased from 13.5% in the historical era to 5.2% in the contemporary era (p < 0.001), overall survival was similar after adjusting for patient characteristics. Risk factors of overall mortality in the contemporary group included NYHA class III or IV (HR 2.17, p < 0.001), etiology of radiation (HR 3.93, p < 0.001) or postcardiac surgery (HR 1.47, p < 0.001), and need for cardiopulmonary bypass (HR 1.35, p [ 0.014). Conclusions. There was a significant change in disease etiology over the study period. Long-term survival after pericardiectomy is affected by patient characteristics including etiology of constriction and severity of symptoms.
C
The first operation to relieve constrictive pericarditis at our institution was performed in 1936 [6]. Since then, more than 1,000 patients have undergone pericardiectomy for constriction. In this study we sought to document the changes in patient presentation during this 77-year interval and to analyze risk factors for early and late outcomes in a contemporary cohort.
onstrictive pericarditis results from inflammation and fibrosis of the pericardium, ultimately leading to impairment of diastolic filling and right heart failure [1]. Generally, constrictive pericarditis is irreversible, and once the diagnosis is made, pericardiectomy is indicated in patients with symptoms of heart failure, especially those requiring increasing doses of diuretics [2]. In Western countries, the underlying cause of constrictive pericarditis among many patients having pericardiectomy is unknown and termed idiopathic; presumably, these patients had unrecognized prior viral pericarditis [3–5]. In our practice, previous cardiac surgery has become increasingly common among patients with constrictive pericarditis who present for operation, and there appears to be increasing numbers of patients with prior mediastinal radiation therapy who present for surgical treatment of constrictive pericarditis. Accepted for publication May 19, 2017. Presented at the Poster Session of the Fifty-third Annual Meeting of The Society of Thoracic Surgeons, Houston, TX, Jan 21–25, 2017. Address correspondence to Dr Schaff, Department of Cardiovascular Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905; email: [email protected].
Ó 2017 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2017;-:-–-) Ó 2017 by The Society of Thoracic Surgeons
Material and Methods Study Design The study was approved by the Mayo Clinic institutional review board. We reviewed medical charts and our surgical database for all patients who underwent pericardiectomy for constrictive pericarditis at Mayo Clinic, Rochester, Minnesota. We included only patients with constrictive pericarditis that was confirmed intraoperatively and excluded patients with other types of pericarditis such as effusive, relapsing, and purulent pericarditis. We also excluded patients who underwent pericardial window procedures for drainage purposes and/or biopsy. We included patients who had pericardiectomy and concomitant procedures, such as coronary artery bypass grafting, cardiac valve repair or 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2017.05.063
2
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
replacement, and aortic procedures. From October 1936 through December 2013, pericardiectomy for constriction was performed in 1,066 patients. Five patients had a second operation but only data from the first procedure were included in the analysis. The frequency of pericardiectomy increased during the 77 years of records we studied as seen in Figure 1, which presents the number of cases by year of surgery. The frequency of cases was relatively low during the first 50 years of our experience. Since 1990, the number of procedures has increased noticeably; 259 cases (24.3%) were performed prior to 1990, and 807 cases (75.7%) from 1990 onwards. To determine changes in patient characteristics and surgical outcomes, we divided the period into two intervals using 1990 as the breakpoint. The historical era included all pericardiectomies performed before 1990, and the contemporary era of pericardiectomy was defined as 1990 through the present.
Definitions Patients with a history of previous cardiac surgery and pericardiotomy were categorized as having postoperative constrictive pericarditis. Patients with a history of mediastinal radiation were defined as having postradiation constrictive pericarditis. Patients with miscellaneous etiologies of pericarditis including tuberculosis, asbestosrelated, and autoimmune disease were grouped as other types of constrictive pericarditis. Patients who did not have one of the previous designations (ie, no specific etiology) were defined as having idiopathic constrictive pericarditis.
Fig 1. Number of pericardiectomies performed by surgery year (n ¼ 1,066).
Ann Thorac Surg 2017;-:-–-
Surgical Techniques In the contemporary era, the standard surgical approach for constrictive pericarditis is complete pericardiectomy through median sternotomy. Our technique for pericardiectomy has been reported previously [7, 8]. Briefly, it includes removal of all the anterior pericardium (phrenic nerve to phrenic nerve), the diaphragmatic pericardium, and, when accessible, a portion of pericardium posterior to the left phrenic nerve. This extent of pericardiectomy is the standard procedure in the contemporary era unless it was not technically feasible. The use of cardiopulmonary bypass depends on the severity of constriction, primarily the need to remove calcification involving the myocardium, as well as the need for concomitant procedures. In the historical era, left anterolateral and bilateral thoracotomy were the common approaches [9, 10].
Statistical Analysis Descriptive statistics were used to summarize baseline characteristics, and the Kaplan-Meier method was used to estimate both long-term and short-term rates of survival for both era cohorts separately. Univariate models based on Cox regression were first used to analyze the association of factors with both long-term and short-term mortality in the contemporary era. Because of the large of number of factors influencing 30-day mortality univariately and the relatively small number of events (n ¼ 42), multivariate analysis was not appropriate. Long-term mortality, however, was further analyzed in multivariate models to find significant risk factors, from
Ann Thorac Surg 2017;-:-–-
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
3
Operative Details
which adjusted estimates of hazard ratios (HRs) are reported. Specifically to predict time to death, we screened each candidate risk factor in univariate analyses, and only those detected at an alpha level of 0.05 were carried forward in the multivariate analysis. Partial likelihood ratio tests were used to justify removal of variables from the multivariate model at a 0.05 level of significance.
Table 2 presents details of the surgical data. In the contemporary era, operations were performed through a median sternotomy in 655 (81.2%) patients, cardiopulmonary bypass was used in 506 (62.7%), and aortic cross clamp was applied in 172 (34% of patients undergoing cardiopulmonary bypass). The most frequent concomitant procedure was tricuspid valve operation, followed by coronary artery bypass grafting.
Results
Early and Overall Mortality Between Eras
Baseline Characteristics
Thirty-day mortality in the historical era was 13.5% (35 of 259), and this decreased to 5.2% (42 of 807) in the contemporary era (p < 0.001). In contrast, late survival appears to have worsened; 5- and 10-year survival rates in the historical era were 71.0% and 59.4%, compared with 63.4% and 46.0% for pericardiectomies performed in the contemporary era (Fig 2). The median survival time of patients in the historical era was almost 19 years whereas for the contemporary era it was reduced by half with a median survival time of 8.9 years.
Table 1 summarizes the clinical characteristics of the 1,066 patients. Patients ranged in age from 1 to 84 years. Compared with the patients in the historical era, those in the contemporary era were older, had larger body surface area, and had higher New York Heart Association (NYHA) functional class. The etiologies of postoperative and postradiation were more frequently found in the contemporary era than in the historical era. Table 1. Baseline Characteristics of the Study Cohort Characteristics Age (years) Median Q1, Q3 Creatinine (mg/dL) Median Q1, Q3 Body surface area (m2) Median Q1, Q3 LV ejection fraction (%) Median Q1, Q3 Sex Male Female NYHA class I II III IV unknown Moderate or severe TR Atrial fibrillation Etiology Idiopathic Postoperative Radiation Other Follow-up period (year) Median Q1, Q3 LV ¼ left ventricle;
Total (n ¼ 1,066)
Historical Era (n ¼ 259)
Contemporary Era (n ¼ 807)
58.0 46.0, 68.0
49.0 31.0, 60.0
61.0 50.0, 69.0
1.2 1.0, 1.4
1.1 1.0, 1.4
1.2 1.0, 1.4
2.0 1.8, 2.2
1.8 1.7, 2.0
2.0 1.9, 2.2
60.0 53.0, 64.0
60.0 52.0, 69.0
60.0 53.0, 64.0
788 (73.9%) 278 (26.1%)
194 (74.9%) 65 (25.1%)
594 (73.6%) 213 (26.4%)
46 188 597 212 23
7 65 152 26 9
(2.8%) (26.0%) (60.8%) (10.4%) (3.5%) N/A N/A
39 123 445 186 14 220 208
(4.9%) (15.5%) (56.1%) (23.5%) (1.7%) (27.3%) (25.8%)
(81.1%) (8.5%) (3.5%) (6.9%)
447 261 92 7
(55.4%) (32.3%) (11.4%) (0.9%)
p Value <0.001
0.598
<0.001
0.936
0.679
<0.001
657 283 101 25
(4.4%) (18.0%) (57.2%) (20.3%) (2.2%)
(61.6%) (26.5%) (9.5%) (2.3%)
NYHA ¼ New York Heart Association;
210 22 9 18
N/A N/A <0.001
<0.001 9.7 2.6, 24.6 Q ¼ quartile;
2.4 0.2, 7.1 TR ¼ tricuspid regurgitation.
4
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
Ann Thorac Surg 2017;-:-–-
Table 2. Operative Details of Patients Undergoing Pericardiectomy Operative Details Sternotomy Thoracotomy Cardiopulmonary bypass Cardiopulmonary bypass time (min) Median Q1, Q3 Concomitant procedures Aortic valve operations Mitral valve operations Tricuspid valve operations Coronary artery bypass grafting Aortic cross clamp time (min) Median Q1, Q3 Operative status Elective Urgent
Total (n ¼ 1066)
Historical Era (n ¼ 259)
Contemporary Era (n ¼ 807)
p Value
741 (69.5%) 325 (30.5%) 535 (50.2%)
86 (33.2%) 173 (66.8%) 29 (11.2%)
655 (81.2%) 152 (18.8%) 506 (62.7%)
<0.001
73 51, 107 186 (17.4%) 38 (3.6%) 44 (4.1%) 102 (9.6%) 103 (9.7%)
104 59, 126 14 (5.4%) 4 (1.5%) 5 (1.9%) 2 (0.8%) 8 (3.1%)
72 51, 103 172 (21.4%) 34 (4.2%) 39 (4.8%) 100 (12.4%) 95 (11.8%)
36 24.2, 59.7
44.5 28, 67.5
35 23, 59.2
961 105
256 (98.8%) 3 (1.2%)
705 (87.4%) 102 (12.6%)
<0.001 0.658
<0.001
0.601
<0.001
Q ¼ quartile.
Consistent with this observation, an unadjusted Cox model of long-term mortality suggests elevated risk in those who had operations after 1990 compared with those who had operations before 1990 (unadjusted HR 1.82; 95% CI, 1.46 to 2.26; p < 0.001). Using the modified risk score from our model including age, etiology, NYHA
Fig 2. Overall survival of patients after pericardiectomy stratified by era.
classification, status of surgery, and need for cardiopulmonary bypass, however, we found that survival of patients in the contemporary era was slightly better than the historical era (adjusted HR 0.82; 95% CI, 0.63 to 1.07; p ¼ 0.140) when controlling for patient characteristics, although this was not statistically significant (Fig 3).
Ann Thorac Surg 2017;-:-–-
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
5
Fig 3. Covariate-adjusted survival function at the median risk for era controlling for age, New York Heart Association classification, cardiopulmonary bypass, status of surgery, and etiology.
Patient Characteristics Associated With 30-Day Mortality in the Contemporary Era Univariate proportional hazard analysis was used to assess the association of 30-day mortality for pericardiectomy and candidate risk factors in the contemporary era. Table 3 displays the unadjusted hazard ratios with 95% confidence intervals of all candidate patient variables. Female sex, chronic renal failure requiring hemodialysis, need for cardiopulmonary bypass, moderate-to-severe tricuspid valve regurgitation, higher creatinine levels, etiology of constrictive pericarditis, and increased NYHA functional classification were the factors significantly associated with increased 30-day mortality. Smaller body surface area and lower left ventricular ejection fraction were also associated with 30-day mortality.
Risk Factors for Overall Mortality in the Contemporary Era A univariate survival analysis was performed to test significance of candidate risk factors in the sample of 807 cases of contemporary pericardiectomy. All candidate factors found to be associated with long-term mortality were further evaluated in multivariate analysis. The final multivariate model included NYHA functional classification, etiology of constriction, need for cardiopulmonary bypass, preoperative renal dialysis, and urgency status of the surgery, along with linear terms of body surface area and age. The adjusted hazard ratios for these models are reported in Table 4. Compared with idiopathic etiology, the “other” categorization of etiology had the highest
adjusted hazard ratio of 4.33 (95% CI, 1.73 to 10.84). For patients whose etiology was radiation, the risk of death was almost 4 times greater than risk of those who had idiopathic constrictive pericarditis (adjusted HR 3.93; 95% CI, 2.8 to 5.50). When prior cardiac surgery was the presumed cause of constriction, the hazard increased by 47% (95% CI, 1.14 to 1.88) when compared with idiopathic reasons. The unadjusted survival curves stratified by etiologies of constriction are shown in Figure 4.
Comment This study reviews our surgical experience with pericardiectomy for constrictive pericarditis through 8 decades. The first pericardiectomy at Mayo Clinic was performed on October 31, 1936, for a 12-year-old boy by Dr Stuart W. Harrington. Four years later Dr Harrington reported his series of 9 pericardiectomies performed between 1936 and 1939; he wrote that 3 patients were cured, 3 improved, and 3 died after surgery [6]. In that publication and in a subsequent paper [9], Harrington referred to his procedure as partial pericardiectomy, recognizing that removal of only the anterior pericardium might not completely release the left ventricle. He did emphasize in these early operations the importance of separating the right ventricle from the diaphragm and finding the proper plane in dissection so as to not leave residual epicardial constriction. Both the surgical approach and the extent of pericardial resection have evolved since those early procedures.
6
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
Ann Thorac Surg 2017;-:-–-
Table 3. Univariate Proportional Hazard Regression Analysis of Each Independent Variable Predicting 30-Day Mortality in the Contemporary Era With Log-Rank p Value Variable Sex (n ¼ 807) Male Female Diabetes (n ¼ 775) No Yes Dialysis (n ¼ 806) No Yes Grouped NYHA (n ¼ 793) I and II III IV Tricuspid valve regurgitation (n ¼ 770) None to mild Moderate to severe Rhythm (n ¼ 804) Atrial fibrillation Sinus Other Redo sternotomy (n ¼ 806) No Yes Etiology (n ¼ 807) Idiopathic Postoperative Radiation Other Cardiopulmonary bypass (n ¼ 807) No Yes Concomitant procedures (n ¼ 803) No Yes Operative status (n ¼ 807) Elective Urgent Age (years) Creatinine Body surface area LV ejection fraction CI ¼ confidence interval;
LV ¼ left ventricle;
n
Events
Hazard Ratio (95% CI)
594 213
25 (4%) 17 (8%)
– 1.93 (1.04, 3.57)
616 159
36 (6%) 5 (3%)
– 0.52 (0.20, 1.32)
789 17
39 (5%) 3 (18%)
– 3.56 (1.10, 11.54)
162 445 186
1 (1%) 16 (4%) 25 (13%)
– 5.79 (0.77, 43.64) 22.29 (3.02, 164.50)
550 220
20 (4%) 19 (9%)
– 2.42 (1.29, 4.54)
208 528 68
15 (7%) 24 (5%) 3 (4%)
– 0.64 (0.34, 1.22) 0.58 (0.17, 2.02)
567 239
23 (4%) 19 (8%)
– 1.98 (1.08, 3.63)
447 261 92 7
15 14 13 0
– 1.59 (0.77, 3.30) 4.35 (2.07, 9.15) 0.00 (0.00, )
301 506
9 (3%) 33 (7%)
– 2.21 (1.06, 4.63)
631 172
30 (5%) 12 (7%)
– 1.46 (0.75, 2.84)
705 102 807 686 802 779
27 15 42 32 41 40
p Value 0.033
0.160
0.023
<0.001
0.004
0.358
0.025
<0.001 (3%) (5%) (14%) (0%)
0.030
0.269
<0.001 (4%) (15%) (5%) (5%) (5%) (5%)
3.91 1.01 1.35 0.23 0.92
– (2.08, (0.98, (1.14, (0.08, (0.89,
7.35) 1.03) 1.60) 0.63) 0.95)
0.594 <0.001 0.005 <0.001
NYHA ¼ New York Heart Association classification.
We favor midline sternotomy for most patients, as it gives wide access to the pericardium and facilitates cannulation if cardiopulmonary bypass is necessary for dissection, control of bleeding, or performing other cardiac procedures. This may be especially important with regard to management of tricuspid regurgitation [11]. In contrast to other operations where cardiac repair often results in lessening of tricuspid valve regurgitation, pericardiectomy and release of constriction may lead to transient right ventricular and tricuspid annular dilatation with worsening of functional valve leakage. Further, it
had been our experience that need for tricuspid valve annuloplasty is difficult to predict, and because persistent valve leakage is associated with reduced survival after pericardiectomy, we have a liberal policy for tricuspid valve annuloplasty when regurgitation worsens with pericardiectomy [11, 12]. We would advise concomitant tricuspid valve repair for any patient who has grade 3 or 4 of 4 regurgitation when pericardiectomy is completed. The ventricles can be decorticated through an anterolateral thoracotomy [10, 13] but we generally reserve that approach for cosmetic considerations.
Ann Thorac Surg 2017;-:-–-
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
7
Table 4. Estimated Hazard Ratios With 95% Confidence Intervals for the Multivariate Proportional Hazard Analysis Predicting Overall Mortality in the Contemporary Era (n ¼ 787) Variable Age Body surface area NYHA Etiology
Cardiopulmonary bypass Operative Status Dialysis
Reference
Hazard Ratio
I and II I and II Idiopathic Idiopathic Idiopathic No Elective No
1.465 0.485 1.710 3.415 4.326 1.467 3.930 1.348 1.401 1.941
10 years III IV Other Postoperative Radiation Yes Urgent Yes
95% Confidence Limits 1.329 0.303 1.199 2.328 1.727 1.143 2.811 1.064 1.028 1.003
1.614 0.776 2.441 5.009 10.836 1.884 5.494 1.708 1.910 3.756
p Value <.001 0.003 <.001 <.001
0.014 0.033 0.049
NYHA ¼ New York Heart Association classification.
Adequate extent of pericardiectomy has been debated [14] and some surgeons continue to use anterior pericardiectomy (from right to left phrenic nerves), believing this adequately releases the ventricles. We favor a more extensive resection [4, 13], in particular freeing the inferior walls of the right and left ventricles and removing the diaphragmatic pericardium as well as pericardium posterior to the left phrenic nerve when possible. Failure to
release the inferior aspect of the ventricles and remove the diaphragmatic pericardium may lead to residual constriction and recurrent symptoms [15]. As seen in Figure 1, there has been an increase in the number of pericardiectomies performed for constrictive pericarditis at our institution since 1990. To explore possible explanations for this change, we divided the surgical experience into the two eras. There was a
Fig 4. Overall survival of patients after pericardiectomy stratified by etiology of constriction in the contemporary era. (postop ¼ postoperative.)
8
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
significant change in disease etiology comparing these two time periods, and this likely accounts for some of the increase in number of procedures. In the historical era, idiopathic constriction was considered to be the cause of constriction in more than 80% of all patients, and tuberculosis was the cause in many of the remaining cases. In the contemporary era, there are increasing numbers of patients with constriction presenting with a history of previous cardiac surgery or remote mediastinal radiation therapy. Thus, some of the increase in the number of procedures may be explained by these other etiologies, which reflect advancement in modern medicine. But other factors also account for this change in practice, including a multidisciplinary specialty clinic and cardiologists with expertise in pericardial disease [1] and the advances in diagnosis of constrictive pericarditis by Doppler echocardiography [16]. The latter is especially important in patients with constriction who have normal thickness pericardium in whom the diagnosis would not be identified by other imaging techniques [17]. Previous studies, including those from our institution, have reported a wide range in perioperative mortality after pericardiectomy for constrictive pericarditis, varying from from 0% to 18.6% [3, 4, 11, 13, 18–27]. In our series, 30-day mortality in the historical era was 13.5%, whereas it decreased to 5.2% in the contemporary era. Isolated pericardiectomy for idiopathic constrictive pericarditis has a very low risk of early death, and early mortality of pericardiectomy after coronary artery bypass surgery is acceptably low, at 3% [28]. Thus, although continued advances in perioperative support may improve this relatively low perioperative risk [29], survival of patients after operation is poorer than expected. Indeed, overall unadjusted survival of patients in the contemporary era was poorer than survival of patients having the operation in the historical era despite the finding that perioperative mortality was lower by more than 50%. Even after adjustment of survival curves for known risk factors, overall outcome of patients in the contemporary era was not significantly different from that of the earlier cohort. Many of the risk factors for late death after pericardiectomy—such as patient age, body surface area, etiology of constriction, and comorbid conditions—are immutable. Lower body surface area was associated with mortality, likely corresponding to patients with cachexia. The most important variables that clinicians might control with medications are severity of right heart failure and chronicity of disease, which are reflected in our analysis by preoperative NYHA class. Many other studies have also documented the increased risk of death after pericardiectomy in patients with advanced heart failure and associated hepatic and renal dysfunction [3, 18, 19, 23, 26]. It seems clear then, that further improvement in outcome depends on prompt referral for pericardiectomy when signs and symptoms of heart failure are present in patients with chronic constrictive pericarditis. Once the diagnosis of constriction is made, we recommend consideration of pericardiectomy for most patients with suspected idiopathic etiology who are
Ann Thorac Surg 2017;-:-–-
otherwise fit and have no serious comorbid conditions that would increase operative risk. Often, a treadmill exercise test with measurement of maximal oxygen consumption is useful to assess impaired cardiac output in patients who deny symptoms. Patients who have constriction due to prior mediastinal radiation can be observed if symptoms are well controlled with low doses of diuretics, but operation would be discussed with the patients if symptoms progress or there is need to escalate the frequency or dose of diuretics. Management is individualized for elderly patients and those with previous cardiac surgery, in particular patients with previous coronary artery bypass who are graft-dependent.
Limitations This study has usual limitations of a retrospective analysis, including selection and reporting biases. Additional limitations include incomplete data from patients in the historical era, lack of information on functional status, and the fact that many changes have occurred in surgical methods and intraoperative and postoperative care over this 8-decade experience. Finally, another limitation in analyzing trends of etiologies over time is that these data are from a single, referral institution and not populationbased.
Conclusions We have experienced an increased number of patients requiring pericardiectomy for constrictive pericarditis, and this may be explained in part by important changes in etiology of constriction. In current practice, postoperative constriction and radiation-induced pericardial disease are common. The poor outcome of patients with constrictive pericarditis and preoperative NYHA class III and IV disability supports early referral for pericardiectomy.
References 1. Miranda WR, Oh JK. Constrictive pericarditis: a practical clinical approach. Prog Cardiovasc Dis 2017;59:369–79. 2. Adler Y, Charron P, Imazio M, et al. 2015 ESC guidelines for the diagnosis and management of pericardial diseases: the Task Force for the Diagnosis and Management of Pericardial Diseases of the European Society of Cardiology (ESC) Endorsed by: The European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2015;36: 2921–64. 3. Ling LH, Oh JK, Schaff HV, et al. Constrictive pericarditis in the modern era: evolving clinical spectrum and impact on outcome after pericardiectomy. Circulation 1999;100: 1380–6. 4. George TJ, Arnaoutakis GJ, Beaty CA, Kilic A, Baumgartner WA, Conte JV. Contemporary etiologies, risk factors, and outcomes after pericardiectomy. Ann Thorac Surg 2012;94:445–51. 5. Imazio M, Gaita F, LeWinter M. Evaluation and treatment of pericarditis: a systematic review. JAMA 2015;314:1498–506. 6. Harrington SW, Barnes AR. Diagnosis and surgical treatment of chronic constrictive pericarditis. South Surg 1940;IX: 459–84. 7. Cho YH, Schaff HV. Surgery for pericardial disease. Heart Fail Rev 2013;18:375–87. 8. Murashita T, Schaff HV. Pericardiectomy, constrictive and effusive. In: Grover FL, Mack MJ (eds). Master Techniques in
Ann Thorac Surg 2017;-:-–-
9. 10. 11.
12.
13.
14. 15.
16. 17. 18. 19.
Surgery: Cardiac Surgery. Alphen aan den Rijn: Wolters Kluwer; 2016:387–96. Harrington SW. Chronic constrictive pericarditis: partial pericardiectomy and epicardiolysis in twenty-four cases. Ann Surg 1944;120:468–85. McCaughan BC, Schaff HV, Piehler JM, et al. Early and late results of pericardiectomy for constrictive pericarditis. J Thorac Cardiovasc Surg 1985;89:340–50. Busch C, Penov K, Amorim PA, et al. Risk factors for mortality after pericardiectomy for chronic constrictive pericarditis in a large single-centre cohort. Eur J Cardiothorac Surg 2015;48:e110–6. G ongora E, Dearani JA, Orszulak TA, Schaff HV, Li Z, Sundt TM. Tricuspid regurgitation in patients undergoing pericardiectomy for constrictive pericarditis. Ann Thorac Surg 2008;85:163–70. Chowdhury UK, Subramaniam GK, Kumar AS, et al. Pericardiectomy for constrictive pericarditis: a clinical, echocardiographic, and hemodynamic evaluation of two surgical techniques. Ann Thorac Surg 2006;81:522–9. Glower D. Invited commentary. Ann Thorac Surg 2012;93: 1240–1. Cho YH, Schaff HV, Dearani JA, et al. Completion pericardiectomy for recurrent constrictive pericarditis: importance of timing of recurrence on late clinical outcome of operation. Ann Thorac Surg 2012;93:1236–40. Welch TD, Ling LH, Espinosa RE, et al. Echocardiographic diagnosis of constrictive pericarditis: Mayo Clinic criteria. Circ Cardiovasc Imaging 2014;7:526–34. Talreja DR, Edwards WD, Danielson GK, et al. Constrictive pericarditis in 26 patients with histologically normal pericardial thickness. Circulation 2003;108:1852–7. Zhu P, Mai M, Wu R, Lu C, Fan R, Zheng S. Pericardiectomy for constrictive pericarditis: single-center experience in China. J Cardiothorac Surg 2015;10:34. Avgerinos D, Rabitnokov Y, Worku B, Neragi-Miandoab S, Girardi LN. Fifteen-year experience and outcomes of
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
20.
21. 22. 23. 24. 25.
26. 27. 28.
29.
9
pericardiectomy for constrictive pericarditis. J Card Surg 2014;29:434–8. Ha J-W, Oh JK, Schaff HV, et al. Impact of left ventricular function on immediate and long-term outcomes after pericardiectomy in constrictive pericarditis. J Thorac Cardiovasc Surg 2008;136:1136–41. Kang SH, Song JM, Kim M, et al. Prognostic predictors in pericardiectomy for chronic constrictive pericarditis. J Thorac Cardiovasc Surg 2014;147:598–605. Lin Y, Zhou M, Xiao J, Wang B, Wang Z. Treating constrictive pericarditis in a chinese single-center study: a five-year experience. Ann Thorac Surg 2012;94:1235–40. Szab o G, Schmack B, Bulut C, et al. Constrictive pericarditis: risks, aetiologies and outcomes after total pericardiectomy: 24 years of experience. Eur J Cardiothorac Surg 2013;44:1023–8. Thompson JL, Burkhart HM, Dearani JA, Cetta F, Oh JK, Schaff HV. Pericardiectomy for pericarditis in the pediatric population. Ann Thorac Surg 2009;88:1546–50. Tiruvoipati R, Naik RD, Loubani M, Billa GN. Surgical approach for pericardiectomy: a comparative study between median sternotomy and left anterolateral thoracotomy. Interact Cardiovasc Thorac Surg 2003;2:322–6. Tokuda Y, Miyata H, Motomura N, et al. Outcome of pericardiectomy for constrictive pericarditis in Japan: a nationwide outcome study. Ann Thorac Surg 2013;96:571–6. Gillaspie EA, Stulak JM, Daly RC, et al. A 20-year experience with isolated pericardiectomy: Analysis of indications and outcomes. J Thorac Cardiovasc Surg 2016;152:448–58. Gillaspie EA, Dearani JA, Daly RC, et al. Pericardiectomy after previous bypass grafting: analyzing risk and effectiveness in this rare clinical entity. Ann Thorac Surg 2017;103: 1429–33. Beckmann E, Ismail I, Cebotari S, et al. Right-sided heart failure and extracorporeal life support in patients undergoing pericardiectomy for constrictive pericarditis: a risk factor analysis for adverse outcome. Thorac Cardiovasc Surg 2016 Nov 17 [Epub ahead of print].
Background. The purpose of this study was to review the surgical outcomes of pericardiectomy for constrictive pericarditis and to examine risk factors for overall mortality in a contemporary period. Methods. We reviewed all patients who underwent pericardiectomy for constriction from 1936 through 2013. The investigation included constrictive pericarditis cases confirmed intraoperatively, all other types of pericarditis were excluded; 1,071 pericardiectomies were performed in 1,066 individual patients. Patients were divided into two intervals: a historical (pre-1990) group (n [ 259) and a contemporary (1990–2013) group (n [ 807). Results. Patients in the contemporary group were older (61 versus 49 years; p < 0.001), more symptomatic (NYHA class III or IV in 79.6% versus 71.2%; p < 0.001), and more frequently underwent concomitant procedures (21.4% versus 5.4%; p < 0.001) compared with those in the historical group. In contrast to the historical cases in which the etiologies of constriction were mostly
idiopathic (81.1%), nearly half of contemporary cases had a nonidiopathic etiology (postoperative 32.3%, radiation 11.4%). Although 30-day mortality decreased from 13.5% in the historical era to 5.2% in the contemporary era (p < 0.001), overall survival was similar after adjusting for patient characteristics. Risk factors of overall mortality in the contemporary group included NYHA class III or IV (HR 2.17, p < 0.001), etiology of radiation (HR 3.93, p < 0.001) or postcardiac surgery (HR 1.47, p < 0.001), and need for cardiopulmonary bypass (HR 1.35, p [ 0.014). Conclusions. There was a significant change in disease etiology over the study period. Long-term survival after pericardiectomy is affected by patient characteristics including etiology of constriction and severity of symptoms.
C
The first operation to relieve constrictive pericarditis at our institution was performed in 1936 [6]. Since then, more than 1,000 patients have undergone pericardiectomy for constriction. In this study we sought to document the changes in patient presentation during this 77-year interval and to analyze risk factors for early and late outcomes in a contemporary cohort.
onstrictive pericarditis results from inflammation and fibrosis of the pericardium, ultimately leading to impairment of diastolic filling and right heart failure [1]. Generally, constrictive pericarditis is irreversible, and once the diagnosis is made, pericardiectomy is indicated in patients with symptoms of heart failure, especially those requiring increasing doses of diuretics [2]. In Western countries, the underlying cause of constrictive pericarditis among many patients having pericardiectomy is unknown and termed idiopathic; presumably, these patients had unrecognized prior viral pericarditis [3–5]. In our practice, previous cardiac surgery has become increasingly common among patients with constrictive pericarditis who present for operation, and there appears to be increasing numbers of patients with prior mediastinal radiation therapy who present for surgical treatment of constrictive pericarditis. Accepted for publication May 19, 2017. Presented at the Poster Session of the Fifty-third Annual Meeting of The Society of Thoracic Surgeons, Houston, TX, Jan 21–25, 2017. Address correspondence to Dr Schaff, Department of Cardiovascular Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905; email: [email protected].
Ó 2017 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2017;-:-–-) Ó 2017 by The Society of Thoracic Surgeons
Material and Methods Study Design The study was approved by the Mayo Clinic institutional review board. We reviewed medical charts and our surgical database for all patients who underwent pericardiectomy for constrictive pericarditis at Mayo Clinic, Rochester, Minnesota. We included only patients with constrictive pericarditis that was confirmed intraoperatively and excluded patients with other types of pericarditis such as effusive, relapsing, and purulent pericarditis. We also excluded patients who underwent pericardial window procedures for drainage purposes and/or biopsy. We included patients who had pericardiectomy and concomitant procedures, such as coronary artery bypass grafting, cardiac valve repair or 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2017.05.063
2
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
replacement, and aortic procedures. From October 1936 through December 2013, pericardiectomy for constriction was performed in 1,066 patients. Five patients had a second operation but only data from the first procedure were included in the analysis. The frequency of pericardiectomy increased during the 77 years of records we studied as seen in Figure 1, which presents the number of cases by year of surgery. The frequency of cases was relatively low during the first 50 years of our experience. Since 1990, the number of procedures has increased noticeably; 259 cases (24.3%) were performed prior to 1990, and 807 cases (75.7%) from 1990 onwards. To determine changes in patient characteristics and surgical outcomes, we divided the period into two intervals using 1990 as the breakpoint. The historical era included all pericardiectomies performed before 1990, and the contemporary era of pericardiectomy was defined as 1990 through the present.
Definitions Patients with a history of previous cardiac surgery and pericardiotomy were categorized as having postoperative constrictive pericarditis. Patients with a history of mediastinal radiation were defined as having postradiation constrictive pericarditis. Patients with miscellaneous etiologies of pericarditis including tuberculosis, asbestosrelated, and autoimmune disease were grouped as other types of constrictive pericarditis. Patients who did not have one of the previous designations (ie, no specific etiology) were defined as having idiopathic constrictive pericarditis.
Fig 1. Number of pericardiectomies performed by surgery year (n ¼ 1,066).
Ann Thorac Surg 2017;-:-–-
Surgical Techniques In the contemporary era, the standard surgical approach for constrictive pericarditis is complete pericardiectomy through median sternotomy. Our technique for pericardiectomy has been reported previously [7, 8]. Briefly, it includes removal of all the anterior pericardium (phrenic nerve to phrenic nerve), the diaphragmatic pericardium, and, when accessible, a portion of pericardium posterior to the left phrenic nerve. This extent of pericardiectomy is the standard procedure in the contemporary era unless it was not technically feasible. The use of cardiopulmonary bypass depends on the severity of constriction, primarily the need to remove calcification involving the myocardium, as well as the need for concomitant procedures. In the historical era, left anterolateral and bilateral thoracotomy were the common approaches [9, 10].
Statistical Analysis Descriptive statistics were used to summarize baseline characteristics, and the Kaplan-Meier method was used to estimate both long-term and short-term rates of survival for both era cohorts separately. Univariate models based on Cox regression were first used to analyze the association of factors with both long-term and short-term mortality in the contemporary era. Because of the large of number of factors influencing 30-day mortality univariately and the relatively small number of events (n ¼ 42), multivariate analysis was not appropriate. Long-term mortality, however, was further analyzed in multivariate models to find significant risk factors, from
Ann Thorac Surg 2017;-:-–-
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
3
Operative Details
which adjusted estimates of hazard ratios (HRs) are reported. Specifically to predict time to death, we screened each candidate risk factor in univariate analyses, and only those detected at an alpha level of 0.05 were carried forward in the multivariate analysis. Partial likelihood ratio tests were used to justify removal of variables from the multivariate model at a 0.05 level of significance.
Table 2 presents details of the surgical data. In the contemporary era, operations were performed through a median sternotomy in 655 (81.2%) patients, cardiopulmonary bypass was used in 506 (62.7%), and aortic cross clamp was applied in 172 (34% of patients undergoing cardiopulmonary bypass). The most frequent concomitant procedure was tricuspid valve operation, followed by coronary artery bypass grafting.
Results
Early and Overall Mortality Between Eras
Baseline Characteristics
Thirty-day mortality in the historical era was 13.5% (35 of 259), and this decreased to 5.2% (42 of 807) in the contemporary era (p < 0.001). In contrast, late survival appears to have worsened; 5- and 10-year survival rates in the historical era were 71.0% and 59.4%, compared with 63.4% and 46.0% for pericardiectomies performed in the contemporary era (Fig 2). The median survival time of patients in the historical era was almost 19 years whereas for the contemporary era it was reduced by half with a median survival time of 8.9 years.
Table 1 summarizes the clinical characteristics of the 1,066 patients. Patients ranged in age from 1 to 84 years. Compared with the patients in the historical era, those in the contemporary era were older, had larger body surface area, and had higher New York Heart Association (NYHA) functional class. The etiologies of postoperative and postradiation were more frequently found in the contemporary era than in the historical era. Table 1. Baseline Characteristics of the Study Cohort Characteristics Age (years) Median Q1, Q3 Creatinine (mg/dL) Median Q1, Q3 Body surface area (m2) Median Q1, Q3 LV ejection fraction (%) Median Q1, Q3 Sex Male Female NYHA class I II III IV unknown Moderate or severe TR Atrial fibrillation Etiology Idiopathic Postoperative Radiation Other Follow-up period (year) Median Q1, Q3 LV ¼ left ventricle;
Total (n ¼ 1,066)
Historical Era (n ¼ 259)
Contemporary Era (n ¼ 807)
58.0 46.0, 68.0
49.0 31.0, 60.0
61.0 50.0, 69.0
1.2 1.0, 1.4
1.1 1.0, 1.4
1.2 1.0, 1.4
2.0 1.8, 2.2
1.8 1.7, 2.0
2.0 1.9, 2.2
60.0 53.0, 64.0
60.0 52.0, 69.0
60.0 53.0, 64.0
788 (73.9%) 278 (26.1%)
194 (74.9%) 65 (25.1%)
594 (73.6%) 213 (26.4%)
46 188 597 212 23
7 65 152 26 9
(2.8%) (26.0%) (60.8%) (10.4%) (3.5%) N/A N/A
39 123 445 186 14 220 208
(4.9%) (15.5%) (56.1%) (23.5%) (1.7%) (27.3%) (25.8%)
(81.1%) (8.5%) (3.5%) (6.9%)
447 261 92 7
(55.4%) (32.3%) (11.4%) (0.9%)
p Value <0.001
0.598
<0.001
0.936
0.679
<0.001
657 283 101 25
(4.4%) (18.0%) (57.2%) (20.3%) (2.2%)
(61.6%) (26.5%) (9.5%) (2.3%)
NYHA ¼ New York Heart Association;
210 22 9 18
N/A N/A <0.001
<0.001 9.7 2.6, 24.6 Q ¼ quartile;
2.4 0.2, 7.1 TR ¼ tricuspid regurgitation.
4
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
Ann Thorac Surg 2017;-:-–-
Table 2. Operative Details of Patients Undergoing Pericardiectomy Operative Details Sternotomy Thoracotomy Cardiopulmonary bypass Cardiopulmonary bypass time (min) Median Q1, Q3 Concomitant procedures Aortic valve operations Mitral valve operations Tricuspid valve operations Coronary artery bypass grafting Aortic cross clamp time (min) Median Q1, Q3 Operative status Elective Urgent
Total (n ¼ 1066)
Historical Era (n ¼ 259)
Contemporary Era (n ¼ 807)
p Value
741 (69.5%) 325 (30.5%) 535 (50.2%)
86 (33.2%) 173 (66.8%) 29 (11.2%)
655 (81.2%) 152 (18.8%) 506 (62.7%)
<0.001
73 51, 107 186 (17.4%) 38 (3.6%) 44 (4.1%) 102 (9.6%) 103 (9.7%)
104 59, 126 14 (5.4%) 4 (1.5%) 5 (1.9%) 2 (0.8%) 8 (3.1%)
72 51, 103 172 (21.4%) 34 (4.2%) 39 (4.8%) 100 (12.4%) 95 (11.8%)
36 24.2, 59.7
44.5 28, 67.5
35 23, 59.2
961 105
256 (98.8%) 3 (1.2%)
705 (87.4%) 102 (12.6%)
<0.001 0.658
<0.001
0.601
<0.001
Q ¼ quartile.
Consistent with this observation, an unadjusted Cox model of long-term mortality suggests elevated risk in those who had operations after 1990 compared with those who had operations before 1990 (unadjusted HR 1.82; 95% CI, 1.46 to 2.26; p < 0.001). Using the modified risk score from our model including age, etiology, NYHA
Fig 2. Overall survival of patients after pericardiectomy stratified by era.
classification, status of surgery, and need for cardiopulmonary bypass, however, we found that survival of patients in the contemporary era was slightly better than the historical era (adjusted HR 0.82; 95% CI, 0.63 to 1.07; p ¼ 0.140) when controlling for patient characteristics, although this was not statistically significant (Fig 3).
Ann Thorac Surg 2017;-:-–-
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
5
Fig 3. Covariate-adjusted survival function at the median risk for era controlling for age, New York Heart Association classification, cardiopulmonary bypass, status of surgery, and etiology.
Patient Characteristics Associated With 30-Day Mortality in the Contemporary Era Univariate proportional hazard analysis was used to assess the association of 30-day mortality for pericardiectomy and candidate risk factors in the contemporary era. Table 3 displays the unadjusted hazard ratios with 95% confidence intervals of all candidate patient variables. Female sex, chronic renal failure requiring hemodialysis, need for cardiopulmonary bypass, moderate-to-severe tricuspid valve regurgitation, higher creatinine levels, etiology of constrictive pericarditis, and increased NYHA functional classification were the factors significantly associated with increased 30-day mortality. Smaller body surface area and lower left ventricular ejection fraction were also associated with 30-day mortality.
Risk Factors for Overall Mortality in the Contemporary Era A univariate survival analysis was performed to test significance of candidate risk factors in the sample of 807 cases of contemporary pericardiectomy. All candidate factors found to be associated with long-term mortality were further evaluated in multivariate analysis. The final multivariate model included NYHA functional classification, etiology of constriction, need for cardiopulmonary bypass, preoperative renal dialysis, and urgency status of the surgery, along with linear terms of body surface area and age. The adjusted hazard ratios for these models are reported in Table 4. Compared with idiopathic etiology, the “other” categorization of etiology had the highest
adjusted hazard ratio of 4.33 (95% CI, 1.73 to 10.84). For patients whose etiology was radiation, the risk of death was almost 4 times greater than risk of those who had idiopathic constrictive pericarditis (adjusted HR 3.93; 95% CI, 2.8 to 5.50). When prior cardiac surgery was the presumed cause of constriction, the hazard increased by 47% (95% CI, 1.14 to 1.88) when compared with idiopathic reasons. The unadjusted survival curves stratified by etiologies of constriction are shown in Figure 4.
Comment This study reviews our surgical experience with pericardiectomy for constrictive pericarditis through 8 decades. The first pericardiectomy at Mayo Clinic was performed on October 31, 1936, for a 12-year-old boy by Dr Stuart W. Harrington. Four years later Dr Harrington reported his series of 9 pericardiectomies performed between 1936 and 1939; he wrote that 3 patients were cured, 3 improved, and 3 died after surgery [6]. In that publication and in a subsequent paper [9], Harrington referred to his procedure as partial pericardiectomy, recognizing that removal of only the anterior pericardium might not completely release the left ventricle. He did emphasize in these early operations the importance of separating the right ventricle from the diaphragm and finding the proper plane in dissection so as to not leave residual epicardial constriction. Both the surgical approach and the extent of pericardial resection have evolved since those early procedures.
6
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
Ann Thorac Surg 2017;-:-–-
Table 3. Univariate Proportional Hazard Regression Analysis of Each Independent Variable Predicting 30-Day Mortality in the Contemporary Era With Log-Rank p Value Variable Sex (n ¼ 807) Male Female Diabetes (n ¼ 775) No Yes Dialysis (n ¼ 806) No Yes Grouped NYHA (n ¼ 793) I and II III IV Tricuspid valve regurgitation (n ¼ 770) None to mild Moderate to severe Rhythm (n ¼ 804) Atrial fibrillation Sinus Other Redo sternotomy (n ¼ 806) No Yes Etiology (n ¼ 807) Idiopathic Postoperative Radiation Other Cardiopulmonary bypass (n ¼ 807) No Yes Concomitant procedures (n ¼ 803) No Yes Operative status (n ¼ 807) Elective Urgent Age (years) Creatinine Body surface area LV ejection fraction CI ¼ confidence interval;
LV ¼ left ventricle;
n
Events
Hazard Ratio (95% CI)
594 213
25 (4%) 17 (8%)
– 1.93 (1.04, 3.57)
616 159
36 (6%) 5 (3%)
– 0.52 (0.20, 1.32)
789 17
39 (5%) 3 (18%)
– 3.56 (1.10, 11.54)
162 445 186
1 (1%) 16 (4%) 25 (13%)
– 5.79 (0.77, 43.64) 22.29 (3.02, 164.50)
550 220
20 (4%) 19 (9%)
– 2.42 (1.29, 4.54)
208 528 68
15 (7%) 24 (5%) 3 (4%)
– 0.64 (0.34, 1.22) 0.58 (0.17, 2.02)
567 239
23 (4%) 19 (8%)
– 1.98 (1.08, 3.63)
447 261 92 7
15 14 13 0
– 1.59 (0.77, 3.30) 4.35 (2.07, 9.15) 0.00 (0.00, )
301 506
9 (3%) 33 (7%)
– 2.21 (1.06, 4.63)
631 172
30 (5%) 12 (7%)
– 1.46 (0.75, 2.84)
705 102 807 686 802 779
27 15 42 32 41 40
p Value 0.033
0.160
0.023
<0.001
0.004
0.358
0.025
<0.001 (3%) (5%) (14%) (0%)
0.030
0.269
<0.001 (4%) (15%) (5%) (5%) (5%) (5%)
3.91 1.01 1.35 0.23 0.92
– (2.08, (0.98, (1.14, (0.08, (0.89,
7.35) 1.03) 1.60) 0.63) 0.95)
0.594 <0.001 0.005 <0.001
NYHA ¼ New York Heart Association classification.
We favor midline sternotomy for most patients, as it gives wide access to the pericardium and facilitates cannulation if cardiopulmonary bypass is necessary for dissection, control of bleeding, or performing other cardiac procedures. This may be especially important with regard to management of tricuspid regurgitation [11]. In contrast to other operations where cardiac repair often results in lessening of tricuspid valve regurgitation, pericardiectomy and release of constriction may lead to transient right ventricular and tricuspid annular dilatation with worsening of functional valve leakage. Further, it
had been our experience that need for tricuspid valve annuloplasty is difficult to predict, and because persistent valve leakage is associated with reduced survival after pericardiectomy, we have a liberal policy for tricuspid valve annuloplasty when regurgitation worsens with pericardiectomy [11, 12]. We would advise concomitant tricuspid valve repair for any patient who has grade 3 or 4 of 4 regurgitation when pericardiectomy is completed. The ventricles can be decorticated through an anterolateral thoracotomy [10, 13] but we generally reserve that approach for cosmetic considerations.
Ann Thorac Surg 2017;-:-–-
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
7
Table 4. Estimated Hazard Ratios With 95% Confidence Intervals for the Multivariate Proportional Hazard Analysis Predicting Overall Mortality in the Contemporary Era (n ¼ 787) Variable Age Body surface area NYHA Etiology
Cardiopulmonary bypass Operative Status Dialysis
Reference
Hazard Ratio
I and II I and II Idiopathic Idiopathic Idiopathic No Elective No
1.465 0.485 1.710 3.415 4.326 1.467 3.930 1.348 1.401 1.941
10 years III IV Other Postoperative Radiation Yes Urgent Yes
95% Confidence Limits 1.329 0.303 1.199 2.328 1.727 1.143 2.811 1.064 1.028 1.003
1.614 0.776 2.441 5.009 10.836 1.884 5.494 1.708 1.910 3.756
p Value <.001 0.003 <.001 <.001
0.014 0.033 0.049
NYHA ¼ New York Heart Association classification.
Adequate extent of pericardiectomy has been debated [14] and some surgeons continue to use anterior pericardiectomy (from right to left phrenic nerves), believing this adequately releases the ventricles. We favor a more extensive resection [4, 13], in particular freeing the inferior walls of the right and left ventricles and removing the diaphragmatic pericardium as well as pericardium posterior to the left phrenic nerve when possible. Failure to
release the inferior aspect of the ventricles and remove the diaphragmatic pericardium may lead to residual constriction and recurrent symptoms [15]. As seen in Figure 1, there has been an increase in the number of pericardiectomies performed for constrictive pericarditis at our institution since 1990. To explore possible explanations for this change, we divided the surgical experience into the two eras. There was a
Fig 4. Overall survival of patients after pericardiectomy stratified by etiology of constriction in the contemporary era. (postop ¼ postoperative.)
8
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
significant change in disease etiology comparing these two time periods, and this likely accounts for some of the increase in number of procedures. In the historical era, idiopathic constriction was considered to be the cause of constriction in more than 80% of all patients, and tuberculosis was the cause in many of the remaining cases. In the contemporary era, there are increasing numbers of patients with constriction presenting with a history of previous cardiac surgery or remote mediastinal radiation therapy. Thus, some of the increase in the number of procedures may be explained by these other etiologies, which reflect advancement in modern medicine. But other factors also account for this change in practice, including a multidisciplinary specialty clinic and cardiologists with expertise in pericardial disease [1] and the advances in diagnosis of constrictive pericarditis by Doppler echocardiography [16]. The latter is especially important in patients with constriction who have normal thickness pericardium in whom the diagnosis would not be identified by other imaging techniques [17]. Previous studies, including those from our institution, have reported a wide range in perioperative mortality after pericardiectomy for constrictive pericarditis, varying from from 0% to 18.6% [3, 4, 11, 13, 18–27]. In our series, 30-day mortality in the historical era was 13.5%, whereas it decreased to 5.2% in the contemporary era. Isolated pericardiectomy for idiopathic constrictive pericarditis has a very low risk of early death, and early mortality of pericardiectomy after coronary artery bypass surgery is acceptably low, at 3% [28]. Thus, although continued advances in perioperative support may improve this relatively low perioperative risk [29], survival of patients after operation is poorer than expected. Indeed, overall unadjusted survival of patients in the contemporary era was poorer than survival of patients having the operation in the historical era despite the finding that perioperative mortality was lower by more than 50%. Even after adjustment of survival curves for known risk factors, overall outcome of patients in the contemporary era was not significantly different from that of the earlier cohort. Many of the risk factors for late death after pericardiectomy—such as patient age, body surface area, etiology of constriction, and comorbid conditions—are immutable. Lower body surface area was associated with mortality, likely corresponding to patients with cachexia. The most important variables that clinicians might control with medications are severity of right heart failure and chronicity of disease, which are reflected in our analysis by preoperative NYHA class. Many other studies have also documented the increased risk of death after pericardiectomy in patients with advanced heart failure and associated hepatic and renal dysfunction [3, 18, 19, 23, 26]. It seems clear then, that further improvement in outcome depends on prompt referral for pericardiectomy when signs and symptoms of heart failure are present in patients with chronic constrictive pericarditis. Once the diagnosis of constriction is made, we recommend consideration of pericardiectomy for most patients with suspected idiopathic etiology who are
Ann Thorac Surg 2017;-:-–-
otherwise fit and have no serious comorbid conditions that would increase operative risk. Often, a treadmill exercise test with measurement of maximal oxygen consumption is useful to assess impaired cardiac output in patients who deny symptoms. Patients who have constriction due to prior mediastinal radiation can be observed if symptoms are well controlled with low doses of diuretics, but operation would be discussed with the patients if symptoms progress or there is need to escalate the frequency or dose of diuretics. Management is individualized for elderly patients and those with previous cardiac surgery, in particular patients with previous coronary artery bypass who are graft-dependent.
Limitations This study has usual limitations of a retrospective analysis, including selection and reporting biases. Additional limitations include incomplete data from patients in the historical era, lack of information on functional status, and the fact that many changes have occurred in surgical methods and intraoperative and postoperative care over this 8-decade experience. Finally, another limitation in analyzing trends of etiologies over time is that these data are from a single, referral institution and not populationbased.
Conclusions We have experienced an increased number of patients requiring pericardiectomy for constrictive pericarditis, and this may be explained in part by important changes in etiology of constriction. In current practice, postoperative constriction and radiation-induced pericardial disease are common. The poor outcome of patients with constrictive pericarditis and preoperative NYHA class III and IV disability supports early referral for pericardiectomy.
References 1. Miranda WR, Oh JK. Constrictive pericarditis: a practical clinical approach. Prog Cardiovasc Dis 2017;59:369–79. 2. Adler Y, Charron P, Imazio M, et al. 2015 ESC guidelines for the diagnosis and management of pericardial diseases: the Task Force for the Diagnosis and Management of Pericardial Diseases of the European Society of Cardiology (ESC) Endorsed by: The European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2015;36: 2921–64. 3. Ling LH, Oh JK, Schaff HV, et al. Constrictive pericarditis in the modern era: evolving clinical spectrum and impact on outcome after pericardiectomy. Circulation 1999;100: 1380–6. 4. George TJ, Arnaoutakis GJ, Beaty CA, Kilic A, Baumgartner WA, Conte JV. Contemporary etiologies, risk factors, and outcomes after pericardiectomy. Ann Thorac Surg 2012;94:445–51. 5. Imazio M, Gaita F, LeWinter M. Evaluation and treatment of pericarditis: a systematic review. JAMA 2015;314:1498–506. 6. Harrington SW, Barnes AR. Diagnosis and surgical treatment of chronic constrictive pericarditis. South Surg 1940;IX: 459–84. 7. Cho YH, Schaff HV. Surgery for pericardial disease. Heart Fail Rev 2013;18:375–87. 8. Murashita T, Schaff HV. Pericardiectomy, constrictive and effusive. In: Grover FL, Mack MJ (eds). Master Techniques in
Ann Thorac Surg 2017;-:-–-
9. 10. 11.
12.
13.
14. 15.
16. 17. 18. 19.
Surgery: Cardiac Surgery. Alphen aan den Rijn: Wolters Kluwer; 2016:387–96. Harrington SW. Chronic constrictive pericarditis: partial pericardiectomy and epicardiolysis in twenty-four cases. Ann Surg 1944;120:468–85. McCaughan BC, Schaff HV, Piehler JM, et al. Early and late results of pericardiectomy for constrictive pericarditis. J Thorac Cardiovasc Surg 1985;89:340–50. Busch C, Penov K, Amorim PA, et al. Risk factors for mortality after pericardiectomy for chronic constrictive pericarditis in a large single-centre cohort. Eur J Cardiothorac Surg 2015;48:e110–6. G ongora E, Dearani JA, Orszulak TA, Schaff HV, Li Z, Sundt TM. Tricuspid regurgitation in patients undergoing pericardiectomy for constrictive pericarditis. Ann Thorac Surg 2008;85:163–70. Chowdhury UK, Subramaniam GK, Kumar AS, et al. Pericardiectomy for constrictive pericarditis: a clinical, echocardiographic, and hemodynamic evaluation of two surgical techniques. Ann Thorac Surg 2006;81:522–9. Glower D. Invited commentary. Ann Thorac Surg 2012;93: 1240–1. Cho YH, Schaff HV, Dearani JA, et al. Completion pericardiectomy for recurrent constrictive pericarditis: importance of timing of recurrence on late clinical outcome of operation. Ann Thorac Surg 2012;93:1236–40. Welch TD, Ling LH, Espinosa RE, et al. Echocardiographic diagnosis of constrictive pericarditis: Mayo Clinic criteria. Circ Cardiovasc Imaging 2014;7:526–34. Talreja DR, Edwards WD, Danielson GK, et al. Constrictive pericarditis in 26 patients with histologically normal pericardial thickness. Circulation 2003;108:1852–7. Zhu P, Mai M, Wu R, Lu C, Fan R, Zheng S. Pericardiectomy for constrictive pericarditis: single-center experience in China. J Cardiothorac Surg 2015;10:34. Avgerinos D, Rabitnokov Y, Worku B, Neragi-Miandoab S, Girardi LN. Fifteen-year experience and outcomes of
MURASHITA ET AL PERICARDIECTOMY FOR CONSTRICTION
20.
21. 22. 23. 24. 25.
26. 27. 28.
29.
9
pericardiectomy for constrictive pericarditis. J Card Surg 2014;29:434–8. Ha J-W, Oh JK, Schaff HV, et al. Impact of left ventricular function on immediate and long-term outcomes after pericardiectomy in constrictive pericarditis. J Thorac Cardiovasc Surg 2008;136:1136–41. Kang SH, Song JM, Kim M, et al. Prognostic predictors in pericardiectomy for chronic constrictive pericarditis. J Thorac Cardiovasc Surg 2014;147:598–605. Lin Y, Zhou M, Xiao J, Wang B, Wang Z. Treating constrictive pericarditis in a chinese single-center study: a five-year experience. Ann Thorac Surg 2012;94:1235–40. Szab o G, Schmack B, Bulut C, et al. Constrictive pericarditis: risks, aetiologies and outcomes after total pericardiectomy: 24 years of experience. Eur J Cardiothorac Surg 2013;44:1023–8. Thompson JL, Burkhart HM, Dearani JA, Cetta F, Oh JK, Schaff HV. Pericardiectomy for pericarditis in the pediatric population. Ann Thorac Surg 2009;88:1546–50. Tiruvoipati R, Naik RD, Loubani M, Billa GN. Surgical approach for pericardiectomy: a comparative study between median sternotomy and left anterolateral thoracotomy. Interact Cardiovasc Thorac Surg 2003;2:322–6. Tokuda Y, Miyata H, Motomura N, et al. Outcome of pericardiectomy for constrictive pericarditis in Japan: a nationwide outcome study. Ann Thorac Surg 2013;96:571–6. Gillaspie EA, Stulak JM, Daly RC, et al. A 20-year experience with isolated pericardiectomy: Analysis of indications and outcomes. J Thorac Cardiovasc Surg 2016;152:448–58. Gillaspie EA, Dearani JA, Daly RC, et al. Pericardiectomy after previous bypass grafting: analyzing risk and effectiveness in this rare clinical entity. Ann Thorac Surg 2017;103: 1429–33. Beckmann E, Ismail I, Cebotari S, et al. Right-sided heart failure and extracorporeal life support in patients undergoing pericardiectomy for constrictive pericarditis: a risk factor analysis for adverse outcome. Thorac Cardiovasc Surg 2016 Nov 17 [Epub ahead of print].