Localized Cardiac Tamponade after Open-Heart Surgery

Transcription

1 Ann Thorac Cardiovasc Surg 2012; 18: Original Article Localized Cardiac Tamponade after Open-Heart Surgery doi: /atcs.oa Anna Grumann, 1 Leonel Baretto, 1 Anthony Dugard, 1,2 Pierre Morera, 3,4 Elisabeth Cornu, 3,4 Jean-Bernard Amiel, 1,2 and Pr Philippe Vignon 1,2,3 Purpose: To describe the clinical presentation and echocardiographic findings associated with localized tamponade after open-heart surgery. Methods: Retrospective analysis of a case series with a surgically proven diagnosis. Results: Among 23 patients with surgically proven localized cardiac tamponade after elective open-heart surgery, 5 patients (22%) died in the ICU from multiorgan failure. At the time of diagnosis (median delay: 2 days; range: 0 8 days), shock was present in 19 patients, 8 of them being hypotensive. Transthoracic echocardiography (TTE) depicted the localized cardiac tamponade in 3 of 4 examined patients, whereas transesophageal echocardiography (TEE) was always conclusive. The right atrium was primarily involved, solely (n = 11) or with the right ventricle (n = 5), whereas the left cardiac cavities were less frequently compressed (left atrium: n = 6, left ventricle: n = 1). The free wall curvature of the involved cardiac chamber was consistently inverted, and blood flow turbulences were depicted in 12 patients. Surgical removal of the compressive hematoma improved the clinical status of 18 patients (78%) who were discharged from the hospital. Conclusion: Since localized tamponade complicating open-heart surgery has various, nonspecific clinical presentations and TTE is not diagnostic, indications of TEE must be liberal in this setting to prompt diagnosis and surgical reoperation. Keywords: tamponade, shock, echocardiography, transesophageal echocardiography, Doppler Introduction 1 Medical-Surgical ICU, Dupuytren Teaching Hospital, Limoges, France 2 CIC-P 0801, Dupuytren Teaching Hospital, Limoges, France 3 University of Limoges, Limoges, France 4 Thoracic and Cardiovascular Surgery, Dupuytren Teaching Hospital, Limoges, France Received: December 2, 2011; Accepted: February 7, 2012 Corresponding author: Pr Philippe Vignon. Réanimation Polyvalente, CHU Dupuytren, 2 Ave. Martin Luther King, Limoges Cedex, France philippe.vignon@unilim.fr 2012 The Editorial Committee of Annals of Thoracic and Cardiovascular Surgery. All rights reserved. Cardiac tamponade is a life-threatening complication of cardiac surgery which incidence may reach 0.5 to 5.8%. 1) After open-heart surgery, single-chamber tamponade may potentially occur since the pericardium is frequently not closed, allowing blood clotting to develop unevenly around the heart and to compress any cardiac chamber. The diagnosis of localized tamponade in the postoperative setting may be challenging due to various clinical presentations. Noticeably, the classic clinical signs of pericardial tamponade are rarely present and the usually described hemodynamic profile (i.e., equalization of elevated right atrial, pulmonary artery diastolic and pulmonary capillary wedge pressures) may be blunted or even absent. 1,2) In the past decades, transthoracic echocardiography (TTE) has progressively supplanted right-heart catheterization for the diagnosis of clinically suspected tamponade because it is highly portable, rapid to obtain, and non invasive. 2) Transesophageal echocardiography (TEE) is recommended in the assessment of patients with unstable 524 Ann Thorac Cardiovasc Surg Vol. 18, No. 6 (2012)

2 Localized Cardiac Tamponade hemodynamics after cardiac surgery. 3) Pericardial effusion and cardiac tamponade have long been described after open-heart surgery. 4) Nevertheless, localized tamponade mostly related to extrapericardial compressive hematoma may lead to atypical clinical presentations. We sought to describe both the clinical presentation and echocardiographic findings associated with localized tamponade after open-heart surgery based on a case series with a surgically proven diagnosis. We also aimed to identify potential factors associated with the outcome in these patients. Materials and Methods Patients We retrospectively studied 23 patients who were diagnosed with a localized tamponade after open-heart surgery during their stay in our intensive care unit (ICU) between 2001 and All patients underwent a TEE study and a surgical confirmation of the diagnosis. Patients characteristics were recorded upon admission to the ICU (baseline) and at the time of tamponade diagnosis. Shock was defined as the presence of clinical and biological findings associated with tissue hypoperfusion, with or without hypotension. 5) Echocardiography All echocardiographic studies were performed by experimented intensivists using an upper-end platform and a multiplane TEE probe (Sonos 5500, Hewlett Packard, Andover). Two-dimensional echocardiographic findings of cardiac tamponade were studied, but not respiratory variations of mitral and tricuspid doppler velocities which are not valid in ventilated patients. 2) The anatomical location, echocardiographic aspect (echo-dense or echo-free space), maximal anteroposterior and lateral dimensions, and area of the compressive effusion or hematoma were recorded. The presence of an inverted free wall curvature of the compressed cardiac chamber and intracavitary blood flow turbulence (color Doppler mapping) was noted. Statistical analysis Data are expressed in median and interquartile range (IQR). Quantitative variables were compared between survivors and non survivors using the Mann-Whitney U test. Percentages were compared using the Chi² test or the Fisher exact test, when necessary. A p-value less than 5% Table 1 Patients characteristics on ICU admission and at the time of tamponade diagnosis was considered as statistically significant. Results Baseline Tamponade (n = 23) (n = 23) SAPSII** 38 (30 44) - SOFA*** 9 (7 10) 10 (8 12) Heart rate (bpm) 95 (76 96) 97 (77 117) Systolic blood pressure 120 ( ) 100 (84 134)* (mmhg) Mean arterial pressure 82 (74 96) 70 (64 84)* (mmhg) Central venous pressure 12 (10 18) 18 (11 24)* (mmhg) Hemoglobin (g/dl) 10 (9 12) 9 (8 10)* Platelet count (/mm 3 ) ( ) ( ) Prothrombin time (%) 47 (33 61) 53 (37 69) ph 7.42 ( ) 7.36 ( ) Lactate (mmol/l) 2.5 ( ) 6.1 ( )* Creatinin (µmol/l) 188 (92 285) 186 ( ) ASAT (UI/L) 69 (39 520) 166 ( ) ALAT (UI/L) 50 (14 528) 124 ( ) Bilirubin (µmol/l) 23 (13 39) 29 (12 44) *: p <0.05 **: SAPSII: Simplified Acute Physiology Score II ***: Sepsis-related Organ Failure Assessment Among 466 patients who were admitted to our ICU after open-heart surgery during the study period, 23 had a documented localized cardiac tamponade (age: 74 years [IQR: 68 76]; 10 men; SAPS II: 38 [IQR: 30 44]). Elective surgery consisted in valve replacements in 14 patients (mitral: n = 9; aortic: n = 12), a coronary artery bypass grafting in 4 patients, and a combined procedure in the remaining patients. Median duration of cardiopulmonary bypass was 115 min (IQR: min), with a median cross-clamp time of 89 min (IQR: min). Tamponade was diagnosed after a median delay of 2 days (range: 0 20 days) after cardiac surgery. Reason for admission to the ICU was shock (n = 11), planned post-operative transfer (n = 7), clinically suspected sepsis (n = 2), or failure to wean the patient from the ventilator (n = 3). At the time of tamponade diagnosis, all patients but one were under vasopressor or inotropic support. Shock was present in 19 patients (83%), but only 8 (35%) of them were hypotensive (Table 1). Five patients (22%) died during their ICU stay from multiorgan failure despite surgical removal of the compressive hematoma. In the remaining 18 patients (78%), Ann Thorac Cardiovasc Surg Vol. 18, No. 6 (2012) 525

3 Grumann A, et al. Fig. 1 Illustrative example of a localized cardiac tamponade depicted by transesophageal echocardiography in a patient presenting with shock after open-heart surgery. In the bicaval view, the right atrium was compressed by a 5 cm-width hematoma (upper left, double-headed arrow). Color flow mapping depicted a mosaic of colors (blood flow turbulence) and marked reduction of right cardiac filling (upper right, arrow). In the upper transesophageal transversal view, the right atrium is nearly obliterated by the compressive hematoma (lower left, asterisks). No pericardial effusion around the hypertrophied left ventricle was evidenced in the transgastric long axis view (lower right). RA: right atrium; LA: left atrium; RV: right ventricle; LV: left ventricle surgical reoperation dramatically improved their clinical status and all survivors were alive at hospital discharge. TTE depicted the localized tamponade in only 3 out of 4 examined patients, whereas TEE was always conclusive (Fig. 1). In 4 patients (17%), the TEE study, which disclosed a localized tamponade, was performed for an indication unrelated to a shock (weaning failure, n = 3; suspected endocarditis: n = 1). In no instance were a circumferential pericardial effusion and a swinging heart observed. The right atrium was primarily involved, either solely (n = 11) or with the right ventricle (n = 5), whereas, the left cardiac cavities were less frequently compressed by the localized tamponade (left atrium: n = 6, left ventricle: n = 1). The compressive hematoma appeared as an echo-dense mass (n = 19) or as an echo-free space (n = 1), 526 or both (n = 3), of variable size (anteroposterior dimension: 80 mm [IQR: mm]; lateral dimension: 54 mm [IQR: mm]; area: 26.6 cm2 [IQR: cm2]). The free wall curvature of the involved cardiac chamber was consistently inverted. The compression consistently persisted throughout the cardiac cycle. Blood flow turbulences were depicted by color Doppler mapping of the compressed cardiac cavity in 12 patients (52%). On ICU admission, high lactate, ASAT and ALAT levels, and low platelet count were predictive factors of death (Table 2). At the time of tamponade diagnosis, the Sepsis-related Organ Failure Assessment (SOFA) score (p = ), ASAT level (p = 0.02), and systolic blood pressure (p = 0.03) were significantly related to the outcome, whereas, bilirubin (p = 0.08), lactate (p = 0.06) and Ann Thorac Cardiovasc Surg Vol. 18, No. 6 (2012)

4 Table 2 Univariate analysis of factors potentially related to outcome on ICU admission Survival (n = 18) Death (n = 5) Age (year) 74 (68 75) 75 (74 76) SAPSII** 37 (29 42) 44 (36 53) SOFA*** 8 (7 9) 10 (9 11) Heart rate (bpm) 101 (82 104) 107 (79 125) Systolic blood pressure (mmhg) 120 ( ) 125 (70 167) Mean arterial pressure (mmhg) 83 (77 93) 81 (47 111) Central venous pressure (mmhg) 19 (13 22) 18 (8 28) LV ejection fraction (%) 65 (50 75) 57 (43 60) Hemoglobin (g/dl) 9.6 ( ) 10.7 ( ) Platelet count (/mm 3 ) ( ) ( )* Prothrombin time (%) 52 (34 61) 35 (30 41) ph 7.42 ( ) 7.38 ( ) Lactate (mmol/l) 1.80 ( ) 7.05 ( )* Creatinin (µmol/l) 188 (71 327) 190 ( ) ASAT (UI/L) 61 (38 212) 1836 ( )* ALAT (UI/L) 32 (12 177) 1504 ( )* Bilirubin (µmol/l) 23 (19 34) 25 (9 52) Delay of reoperation (hours) 21 (31 161) 48 (11 182) Transfusion (units) 3 (0 7) 5 (2 6) Fresh frozen plasma (n) 1 (0 5) 3 (0 10) *: p <0.05 **: SAPSII: Simplified Acute Physiology Score II ***: Sepsis-related Organ Failure Assessment Localized Cardiac Tamponade ALAT levels (p = 0.07) tended to be lower in survivors. Discussion The present case series confirms that the clinical presentation of localized tamponade after open-heart surgery is nonspecific and distinct from that associated with typical cardiac tamponade physiology. 1,6) Russo et al. 1) reported that only one of ten post-cardiac surgery patients who sustained a localized tamponade had a classical clinical presentation. In our study, only 35% of patients had hypotension at the time of tamponade diagnosis, and 7 patients (30%) had a CVP <15 mmhg. This presumably accounts for the substantial delay of diagnosis. 1) As previously reported, 6 12) TTE failed to identify the compressed cardiac cavity in 1/4 of our patients, whereas TEE was always conclusive. During the first week following open-heart surgery, Flynn et al. 13) showed that TTE provides adequate visualization of cardiac ventricles in only 58% of patients, since numerous factors are known to interfere with image quality in ventilated ICU patients. 14) In addition, localized tamponade predominantly involve the right or left atrium, 15,16) which are best imaged by TEE. 6,8,17) Although TTE has allowed the depiction of localized cardiac tamponade in ventilated patients after open-heart surgery, 1,15,16) TEE must be obtained since a false positive or false negative result may have major consequences in this clinical setting. Although chest CT scan is an alternative imaging modality to depict localized tamponade of various origins, 18) TEE has the unparalleled advantage of allowing immediate diagnosis at the bedside of unstable patients who cannot be safely transferred to the Radiology suite. Localized tamponade results from an elective compression of one or several cardiac cavities, whereas cardiac tamponade physiology typically occurs secondary to the rapid development of a circumferential pericardial effusion which compresses all chambers equally. 8) Accordingly, none of our patient exhibited a circumferential pericardial effusion and typical echocardiographic findings associated with cardiac tamponade were not observed, as previously reported. 1,7,9) It has long been known that the hemodynamic effects of tamponade mainly result from atrial rather than ventricle compression. 17) Interestingly, an atrial compression was observed in 90% of our patients. TEE consistently disclosed the inversion of free wall curvature of the involved cardiac chamber and turbulent blood flow reflected the severe reduction of Ann Thorac Cardiovasc Surg Vol. 18, No. 6 (2012) 527

5 Grumann A, et al. cardiac filling in half of our patients. 12) As in 6 of our patients, the localized tamponade may less frequently involve a ventricle. 1,16) In this case, the inversion of free wall curvature may be more subtle and localized in the vicinity of the compressive hematoma. In all our patients, the surgeon confirmed the presence of clots or of bloody fluid at the surface of the heart which accounted for the aspect of an echo-dense mass 8,12) or of an echo-free space 9,15) compressing cardiac cavities. Since the pericardium is frequently left open after cardiac surgery, the compressive hematoma usually develops in the mediastinal space rather than within the pericardium. 12) Surgical removal of the compressive clot or fluid collection resulted in a rapid improvement in survivors, whereas the remaining patients died in the ICU from multiorgan failure, which resulted from prolonged tissue hypoperfusion prior to the reoperation. The present mortality rate is in keeping with that reported by Russo et al., which reached 20%. 1) High lactate and ASAT / ALAT levels, and low platelet count on admission were associated with death. Elevation of circulating liver enzymes presumably reflected the hepatic congestion secondary to cardiac tamponade. This retrospective series has several limitations. Invasive hemodynamic monitoring was not performed in our patients with surgically proven localized tamponade since echocardiography is routinely used for the assessment of patients with cardiorespiratory compromise in our ICU. 14) Nevertheless, authors previously reported that right-heart catheterization was not commonly contributive in these patients. 1,9) Although the small sample size did not allow us to use a multivariate analysis in order to determine independent risk factors of death, the present case series of localized tamponade after open-heart surgery is currently the largest published to the best of our knowledge. Since TTE was not systematically performed, its diagnostic capacity could not be compared with that of TEE. Conclusion Localized tamponade complicating open-heart surgery has various, non specific clinical presentations. Surgical removal of the compressive hematoma is key to improve the clinical status of the patient. TTE is not adequately suited to ascertain the diagnosis while TEE is conclusive. Accordingly, indications for performing a TEE examination in the setting of circulatory failure developing after open-heart surgery must be liberal to prompt surgical removal of the compressive hematoma. Disclosure Statement All authors declare having no personal or financial conflict of interest. References 1) Russo AM, O Connor WH, Waxman HL. Atypical presentations and echocardiographic findings in patients with cardiac tamponade occurring early and late after cardiac surgery. Chest 1993; 104: ) Spodick DH. Acute cardiac tamponade. N Engl J Med 2003; 349: ) Cheitlin MD, Alpert JS, Armstrong WF, et al. ACC/ AHA Guidelines for the Clinical Application of Echocardiography. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Clinical Application of Echocardiography). Developed in collaboration with the American Society of Echocardiography. Circulation 1997; 95: ) Kuvin JT, Harati NA, Pandian NG, et al. Postoperative cardiac tamponade in the modern surgical era. Ann Thorac Surg 2002; 74: ) Antonelli M, Levy M, Andrews PJ, et al. Hemodynamic monitoring in shock and implications for management. International Consensus Conference, Paris, France, April Intensive Care Med 2007; 33: ) D Cruz IA, Kensey K, Campbell C, et al. Two-dimensional echocardiography in cardiac tamponade occurring after cardiac surgery. J Am Coll Cardiol 1985; 5: ) Simpson IA, Munsch C, Smith EE, et al. Pericardial haemorrhage causing right atrial compression after cardiac surgery: role of transoesophageal echocardiography. Br Heart J 1991; 65: ) Kochar GS, Jacobs LE, Kotler MN. Right atrial compression in postoperative cardiac patients: detection by transoesophageal echocardiography. J Am Coll Cardiol 1990; 16: ) Sangalli F, Colagrande L, Manetti B, et al. Hemodynamic instability after cardiac surgery: transoesophageal echocardiographic diagnosis of a localized pericardial tamponade. J Cardiothorac Vasc Anesth 2005; 19: ) Ionescu A, Wilde P, Karsch KR. Localized pericardial tamponade: difficult echocardiographic diagnosis of a rare complication after cardiac surgery. J Am Soc Echocardiogr 2001; 14: ) Berge KH, Lanier WL, Reeder GS. Occult cardiac tamponade detected by transoesophageal echocardiography. Mayo Clin Proc 1992; 67: ) Ananthasubramaniam K, Jaffery Z. Postoperative right atrial compression by extracardiac hematoma: transesophageal echocardiographic diagnosis in the 528 Ann Thorac Cardiovasc Surg Vol. 18, No. 6 (2012)

6 Localized Cardiac Tamponade critically ill patient. Echocardiography 2007; 24: ) Flynn BC, Spellman J, Bodian C, et al. Inadequate visualization and reporting of ventricular function from transthoracic echocardiography after cardiac surgery. J Cardiothorac Vasc Anesth 2010; 24: ) Vignon P, Mentec H, Terré S, et al. Diagnostic accuracy and therapeutic impact of transthoracic and transesophageal echocardiography in mechanically ventilated patients in the ICU. Chest 1994; 106: ) Chamoun G, Farah MG. Echocardiographic diagnosis of right atrial tamponade. Chest 1990; 97: ) Tardif JC, Taylor K, Pandian NG, et al. Right ventricular outflow tract and pulmonary artery obstruction by postoperative mediastinal hematoma: delination by multiplane transoesophageal echocardiography. J Am Soc Echocardiogr 1994; 7: ) Fowler NO, Gabel M. The hemodynamic effects of cardiac tamponade: mainly the result of atrial, not ventricular, compression. Circulation 1985; 71: ) Gulati GS, Sharma S. Pericardial absess occurring after tuberculous pericarditis: image morphology on computed tomography and magnetic resonance imaging. Clin Radiol 2004; 59: Ann Thorac Cardiovasc Surg Vol. 18, No. 6 (2012) 529