Abstract: Background. Cardiac tamponade constitutes the most severe clinical consequence of a coronary perforation (CP) during percutaneous coronary intervention (PCI). Methods. We analyzed 23,399 PCIs and identified 73 patients complicated by CP (0.31%), of which 26 developed tamponade (0.11%). Patients were divided into two groups: CP with tamponade (group A) and CP without tamponade (group B). Characteristics and outcomes were compared and a multivariate model was developed to evaluate the independent contribution of tamponade on mortality. Results. All patients with tamponade were treated with pericardiocentesis. Tamponade increased the risk of death by more than 3-fold (odds ratio [OR] = 3.3; 95% confidence interval [CI],1.01-10.6; P=.047) relative to patients with CP but no tamponade. CP with tamponade during non-elective PCI carried a significantly worse long-term prognosis (P=.001) than the same complications during elective PCI. The use of glycoprotein IIb/IIIa inhibitors and bivalirudin did not differ between groups. Polytetrafluoroethylene-covered stents were used similarly between the two groups, while coil embolization was used more often in group A (P=.003). Emergent cardiac surgery was needed in 3 patients (11.5%) in group A versus 1 patient (2.1%) in group B. Conclusions. Tamponade complicating CP during PCI has short- and long-term major adverse effects. In-hospital mortality after tamponade and referral for emergent surgical treatment have decreased. Nonetheless, tamponade in this setting increases risk of death by >3-fold. PCI of complex lesions with an expected increased risk of perforation should be undertaken cautiously among patients with non-elective PCI because of increased early and late risk of death.
J INVASIVE CARDIOL 2013;25(10):486-491
Key words: pericardiocentesis, percutaneous coronary intervention
Cardiac tamponade is a grave but uncommon complication of percutaneous coronary intervention (PCI) and is frequently associated with poor outcome.1 Cardiac tamponade constitutes the most severe clinical manifestation of a coronary perforation (CP) and usually occurs after grade III CP as defined by the Ellis criteria.2 Few studies have specifically addressed characteristics and outcomes associated with cardiac tamponade during PCI.3,4 Additionally, due to the limited sample size, most of the previously published studies have included in their reports all 3 grades of CP. Furthermore, there is a dearth of evidence concerning the long-term effect of tamponade after successful initial management.
The objective of our study was to report the clinical, procedural, and angiographic characteristics of patients who experienced tamponade after PCI compared with patients experiencing CP without development of tamponade and to evaluate the short-term outcomes and long-term survival of these patients.
We analyzed a prospectively collected database of 23,399 PCIs performed at our institution, an urban tertiary-care teaching hospital, during an 8-year period (1999-2006) and identified 73 patients undergoing PCI complicated by CP, of which 26 developed tamponade. One patient who developed intraprocedural tamponade was found to have ventricular rupture secondary to acute myocardial infarction and was excluded from the database.
Patients were divided into two groups: group A (CP with tamponade) and group B (CP without tamponade). Tamponade was defined as the presence of fluid in the pericardial space requiring intervention during the procedure or before hospital discharge.5
The cineangiograms of all 73 patients were reviewed retrospectively and angiographic data were obtained. All relevant clinical information was recorded from the medical records, including age, gender, history of diabetes mellitus, hypertension, hyperlipidemia, heart failure, ejection fraction, previous myocardial infarction, previous PCI, previous coronary artery bypass graft (CABG), history of stroke, carotid artery disease, peripheral arterial disease, abdominal aortic aneurysm, chronic kidney disease, hemodialysis status, chronic obstructive pulmonary disease (COPD), history of smoking, and number of affected vessels (coronary artery disease). Procedural data were recorded and included status of elective or emergent PCI, use of glycoprotein (GP) IIb/IIIa inhibitors, use of bivalirudin, number of stents, intravascular ultrasound (IVUS) use prior to perforation, use of intraaortic balloon pump (IABP), measures to treat the CP (ie, protamine use, polytetrafluoroethylene [PTFE]-covered stent use, coil use, perfusion balloon use, and need for emergent cardiac surgery). Angiographic data, which included the culprit perforated vessel, the class of coronary lesion according to the American College of Cardiology/American Heart Association (ACC/AHA) lesion classification, angiographic evidence of coronary calcification (identified as readily apparent radio-opacities within the vascular wall), and presence of proximal tortuosity, were recorded. Proximal coronary tortuosity was defined as ≥2 bends of at least 45°. Short-term outcomes recorded included PCI success, length of hospital stay, ventricular tachycardia (VT) arrest, in-hospital mortality post emergent cardiac surgery, and overall in-hospital mortality.
In-hospital death was defined as death occurring during the index admission. Successful PCI was defined as PCI that achieved angiographic success without the major clinical complications of death, procedural-related myocardial infarction (MI), or need for emergent cardiac surgery. Emergent cardiac surgery was defined as unplanned surgery that was required emergently within 24 hours of PCI to treat a complication. Long-term patient survival data were secured from the United States Social Security Death Index database.
Statistical analysis. Continuous normally distributed data are presented as mean ± standard deviation and were analyzed with the unpaired t-test. Median values (min, max) are presented for data that were not normally distributed. Categorical data are presented as counts or proportions (percentage). Baseline clinical, procedural, and angiographic characteristics, as well as procedural outcomes, were compared between the two groups using either the Fisher’s exact test or the Mann-Whitney U-test. A multivariate model was developed to evaluate the independent contribution of CP with tamponade, history of CABG, and elective PCI on mortality. Kaplan-Meier survival estimates were calculated for the tamponade versus no tamponade groups, and adjusted for history of CABG and elective PCI. Subsequently, unadjusted Kaplan-Meier curves were estimated for the elective and non-elective PCI subgroups, separately. Finally, we performed a landmark analysis for all patients who survived the 30-day benchmark in order to evaluate possible long-term adverse effect of tamponade. A P-value <.05 was considered a priori to indicate statistical significance. SPSS software version 16 (SPSS, Inc) was utilized for the statistical analysis.
All patients with tamponade were treated with pericardiocentesis. Patients were followed for an average of 1865.74 ± 1037.77 days (range, 0-3401 days).
The baseline clinical, procedural, and angiographic characteristics of the patients in the two groups are shown in Table 1. More patients who experienced tamponade (group A) had elective non-emergent PCI procedures (P=.041), and no one had a history of previous CABG (P=.001). The use of GP IIb/IIIa inhibitors and bivalirudin was similar between the two groups.
The treatment options employed to manage CP are shown in Table 2. Protamine use was common but not universal for heparin recipients, and did not differ between groups. PTFE-covered stents were used similarly between the two groups, while coil embolization was used more often in group A (P=.003). There was a trend toward more use of IABP after CP complicated by tamponade (P=.061). Need for emergent cardiac surgery occurred in 3 patients (11.5%) after CP complicated by tamponade versus 1 patient (2.1%) after CP without tamponade (P=NS). There were no in-hospital deaths after emergent cardiac surgery in any of the groups. One patient underwent repair of apical VSD 3 days after successful treatment of CP with PTFE-covered stent; this patient was not included in the emergent cardiac surgery cohort.
In-hospital mortality was similar for the two groups (7.7% for group A and 4.3% for group B; Table 3). Death during the follow-up period occurred in 8 patients of group A and 10 patients of group B (Table 3). Successful PCI rate was similar between the two groups, but patients in group A had significantly longer length of hospital stay (median, 5.5 days [range, 0-22 days] compared to group B (median, 3 days [range 1-53 days]); P=.01. Ventricular tachycardia (VT) with arrest occurred more often after CP complicated by tamponade (P=.02). Of the 26 patients who experienced tamponade after CP (group A), 5 patients experienced VT (Table 3). In-hospital death occurred in 1 of these 5 patients. While in-hospital death was more frequent in patients with tamponade that experienced VT (1/5; 20%), when compared with patients who had tamponade not complicated by VT (1/21; 4.8%), the rates were not statistically significant (P=.35).
The development of tamponade as a complication of CP conferred a greater than 3-fold increase in the long-term probability of death (OR = 3.3; 95% CI, 1.01-10.65; P=.047; Table 4) compared with those who sustained CP without tamponade (Figure 1). However, the circumstances of PCI greatly influenced this risk (Table 4). Figure 2 reflects long-term unadjusted survival curves for patients experiencing CP with and without tamponade as a complication of elective PCI. In this subgroup, no impact of tamponade on mortality was observed. Figure 3 displays long-term unadjusted survival for non-elective PCI patients with the same complication. In this cohort, CP with tamponade was associated with significantly worse survival (P=.001; Figure 3). Of the 5 non-elective PCI patients who experienced CP with tamponade, 4 expired during the study period (80%), compared with 5 out of 21 (24%) non-elective PCI patients who experienced perforation without tamponade.
In order to evaluate the possibility that the development of tamponade confers a late mortality risk after coronary perforation, we conducted a landmark analysis beginning at 30 days after CP. The landmark analysis showed that the difference in survival between the two groups approached but did not reach statistical significance (data not shown).
Numerous studies have addressed the incidence, risk factors, diagnosis, and management of CP after PCI, but the literature contains very few studies specifically evaluating the incidence of tamponade and its relationship to short- and long-term survival. Additionally, most previously published studies have included tamponade that arose from perforation of the right ventricle by a temporary pacing wire rather than the coronary artery. Several of these investigations were conducted more than a decade ago, and the equipment used for PCI, as well as our understanding and management of coronary perforation during PCI, have evolved since that time. Furthermore, although we now use ablative devices less often, the complexity of the clinical and anatomic patient spectrum has broadened (eg, chronic total occlusion interventions). The purpose of this study was to evaluate clinical, procedural, and angiographic characteristics of patients who experienced CP complicating PCI, and to compare outcomes of those who develop tamponade as a result of CP to those who experience perforation without tamponade.
Over an 8-year period, we observed CP in 73 PCI patients (0.31%) and cardiac tamponade in 26 patients (36%) with identified CP (0.11% of all PCI patients). The rate of tamponade in our study was similar to previously reported series (Table 5). The observed in-hospital mortality after tamponade (7.7%) was somewhat lower than previously reported by most other investigators, although two studies have reported zero mortality.3,4,6-16 In-hospital mortality was numerically higher for patients with versus without tamponade (7.7% vs 4.3%, respectively), but this difference was not statistically significant. In contrast to some previous studies,3,4,6-19 we found that emergent cardiac surgery was needed infrequently, being required in just 11.5% of patients with tamponade. This finding is consistent with reports from more recent studies11,13,20 and may represent improvement in complication identification, equipment, and interventional techniques over time.
As reported in previous studies,11,13 we observed that patients who experienced tamponade had complex anatomy with the majority of treated lesions classified as ACC/AHA class C (73.1%). Evidence of angiographic calcium was noted in 85.6% of cases. The left anterior descending coronary artery was the intervention site in 57.7% of the tamponade cases, which is somewhat different than observed by other researchers.4,17 There was no patient in group A that had a history of previous CABG (P=.001), which was consistent with other reports that suggest a “protective” effect of previous thoracotomy through scar formation.
Although the in-hospital mortality was not statistically different between the two groups, we found that tamponade conferred a more than 3-fold risk of death among patients with CP during the study period. Increased mortality risk may be related to factors that are directly connected to the impact of tamponade and/or its treatment, such as a greater risk of cardiac arrest and subsequently higher incidence of complications (eg, complications related to rib fractures from cardiac compressions, or greater use of occluding coils to manage perforation, with consequent myocardial infarction). Mortality may also be increased by factors related to patients’ underlying disease severity rather than tamponade per se: patients with CP and tamponade were more likely to have a history of previous stroke and, as noted, were more likely to have prior CABG compared with patients not experiencing tamponade. We adjusted for these differences in our multivariate model of survival, but we recognize that the small sample size limited our ability to develop a robust adjusted model.
The observation that CP with tamponade in the setting of elective PCI decreased the likelihood of death by 73% compared with the same complication in the setting of non-elective PCI is consistent with data that show higher risk of adverse cardiovascular events in patients presenting with acute coronary syndrome. The observation that CP with tamponade complicating non-elective PCI is linked to a significantly worse prognosis (P=.001) during a mean follow-up period of greater than 5 years underscores the importance of prevention in this high-risk subgroup. Although a 30-day landmark analysis did not demonstrate an important difference in survival after 30 days between the two groups, the numerical differences were remarkable, and this analysis is acknowledged to be limited by the small sample size. Thus, an important association between CP with tamponade and heightened late mortality cannot be excluded.
Study limitations. The study was retrospective and descriptive and has all the inherent limitations of such studies. Due to the small sample size, given the rarity of this complication and accompanied events, the probability of type 1 error is increased and the results should be interpreted cautiously, especially where the number of events is limited. Additionally, the presented data may not represent practice in other institutions.
The rate of tamponade after PCI remains grossly unchanged from earlier reports, but related in-hospital mortality and referral for emergent surgical treatment have decreased. As in the past, early diagnosis and immediate implementation of therapy to restore hemodynamic stability and seal the perforation are keys to a successful outcome. We observed that cardiac tamponade may impact both short- and long-term prognosis: tamponade complicating non-elective interventions carries a significantly greater risk of death than tamponade complicating elective interventions, including the possibility of increased late mortality. With this in mind, complex PCI procedures (especially attempts to reopen chronic total occlusions) that have greater perforation risk should be undertaken among higher-risk patients only after consideration of the short- and long-term risk from tamponade.
- Holmes DR Jr, Nishimura R, Fountain R, Turi ZG. Iatrogenic pericardial effusion and tamponade in the percutaneous intracardiac intervention era. JACC Cardiovasc Interv. 2009;2(8):705-717. Review.
- Ellis SG, Ajluni S, Arnold AZ, et al. Increased coronary perforation in the new device era. Incidence, classification, management, and outcome. Circulation. 1994;90(6):2725-2730.
- Von Sohsten R, Kopistansky C, Cohen M, Kussmaul WG III. Cardiac tamponade in the “new device” era: evaluation of 6999 consecutive percutaneous coronary interventions. Am Heart J. 2000;140(2):279-283.
- Fejka M, Dixon SR, Safian RD, et al. Diagnosis, management, and clinical outcome of cardiac tamponade complicating percutaneous coronary intervention. Am J Cardiol. 2002;90(11):1183-1186.
- Brindis R, Fitzgerald S, Anderson H, Shaw R, Weintraub W, Williams J. The American College of Cardiology-National Cardiovascular Data Registry (ACC-NCDR): building a national clinical data repository. J Am Coll Cardiol. 2001;37(8):2240-2245.
- Ajluni SC, Glazier S, Blankenship L, O’Neill WW, Safian RD. Perforations after percutaneous coronary interventions: clinical, angiographic, and therapeutic observations. Cathet Cardiovasc Diagn. 1994;32(3):206-212.
- Gunning MG, Williams IL, Jewitt DE, Shah AM, Wainwright RJ, Thomas MR. Coronary artery perforation during percutaneous intervention: incidence and outcome. Heart. 2002;88(5):495-498.
- Witzke CF, Martin-Herrero F, Clarke SC, Pomerantzev E, Palacios IF. The changing pattern of coronary perforation during percutaneous coronary intervention in the new device era. J Invasive Cardiol. 2004;16(6):257-301.
- Javaid A, Buch AN, Satler LF, et al. Management and outcomes of coronary artery perforation during percutaneous coronary intervention. Am J Cardiol. 2006;98(7):911-914 (Epub 2006 Aug 7).
- Shirakabe A, Takano H, Nakamura S, et al. Coronary perforation during percutaneous coronary intervention. Int Heart J. 2007;48(1):1-9.
- Kiernan TJ, Yan BP, Ruggiero N, et al. Coronary artery perforations in the contemporary interventional era. J Interv Cardiol. 2009;22(4):350-353.
- Doll JA, Nikolsky E, Stone GW, et al. Outcomes of patients with coronary artery perforation complicating percutaneous coronary intervention and correlations with the type of adjunctive antithrombotic therapy: pooled analysis from REPLACE-2, ACUITY, and HORIZONS-AMI trials. J Interv Cardiol. 2009;22(5):453-459 (Epub 2009 Aug 20).
- Al-Lamee R, Ielasi A, Latib A, et al. Incidence, predictors, management, immediate and long-term outcomes following grade III coronary perforation. JACC Cardiovasc Interv. 2011;4(1):87-95.
- Romaguera R, Sardi G, Laynez-Carnicero A, et al. Outcomes of coronary arterial perforations during percutaneous coronary intervention with bivalirudin anticoagulation. Am J Cardiol. 2011;108(7):932-935 (Epub 2011 Jul 23).
- Hendry C, Fraser D, Eichhofer J, et al. Coronary perforation in the drug-eluting stent era: incidence, risk factors, management and outcome: the UK experience. EuroIntervention. 2012;8(1):79-86.
- Fukutomi T, Suzuki T, Popma JJ, et al. Early and late clinical outcomes following coronary perforation in patients undergoing percutaneous coronary intervention. Circ J. 2002;66(4):349-356.
- Shimony A, Zahger D, Van Straten M, et al. Incidence, risk factors, management and outcomes of coronary artery perforation during percutaneous coronary intervention. Am J Cardiol. 2009;104(12):1674-1677.
- Shimony A, Joseph L, Mottillo S, Eisenberg MJ. Coronary artery perforation during percutaneous coronary intervention: a systematic review and meta-analysis. Can J Cardiol. 2011;27(6):843-850.
- Gruberg L, Pinnow E, Flood R, et al. Incidence, management, and outcome of coronary artery perforation during percutaneous coronary intervention. Am J Cardiol. 2000;86(6):680-682.
- Kini AS, Rafael OC, Sarkar K, et al. Changing outcomes and treatment strategies for wire induced coronary perforations in the era of bivalirudin use. Catheter Cardiovasc Interv. 2009;74(5):700-707.
From the Lenox Hill Hospital, Heart and Vascular Institute of New York, Department of Cardiac and Vascular Interventional Services, New York, New York.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.
Manuscript submitted May 28, 2013, provisional acceptance given June 20, 2013, final version accepted July 2, 2013.
Address for correspondence: Ioannis Stathopoulos, 30-10 38th Street, Astoria, NY 11103. Email: firstname.lastname@example.org