Pericardial tamponade as a complication of invasive cardiac procedures: a review of the literature

Introduction

Nowadays, open heart surgery procedures are being replaced with minimally invasive attempts. This trend has led to a growing number of invasive procedures in cardiology. Pericardial tamponade (PT) is one of the most severe complications after such procedures. In this paper we review studies and meta-analyses concerning the rate of iatrogenic PT and concisely present brief characteristics for each procedure separately.

Pericardial tamponade as a result of catheter-based procedures

PT occurs when the pressure in the pericardial space exceeds the pressure in one or more cardiac chambers. The occurrence of hemodynamic abnormalities and clinical symptoms depends on the rate of fluid accumulation relative to pericardial stretch and the effectiveness of compensatory mechanisms. Thus, abrupt intrapericardial content accumulation (i.e. hemorrhage from cardiac rupture) occurs in the context of a relatively stiff, unyielding pericardium and quickly overwhelms the pericardial capacity to stretch before most compensatory mechanisms can be activated. In those situations, volumes such as 50–100 ml of fluid may result in hemodynamic decompensation [1]. In the cases of slow increase in pericardial fluid volume there is more time for pericardial capacity to stretch and for compensative mechanisms to be activated, so even 2 l or more may accumulate before critical, life-threatening PT occurs [2].

Iatrogenic acute PT is a life-threatening complication that can lead to death. It often involves hemodynamic instability and requires cardiopulmonary resuscitation in 20% and blood transfusion in more than 25% of patients [3].

Within the cardiology lab, PT develops most often rapidly, usually as a result of perforation of the heart structures. The perforation may be caused by a guidewire, balloon dilator, sheath, pacemaker lead, or excessive ablation energy. The presentation depends in part on at least 5 factors: the size of the device responsible for the perforation, the structure that is perforated, such as atrial versus ventricular myocardium, left versus right chamber, the hemodynamic state during perforation, mechanical properties of the pericardium itself, and the coagulation status. The thicker wall of the left ventricle (LV) (≤ 10 mm) may act to seal small perforations, balancing the higher intra-chamber pressures in contrast to the right ventricle (RV) (≤ 4 mm). Small perforation of the ventricles in a patient without anticoagulation may not be clinically apparent. Perforation of the left atrium (LA) is potentially much more serious, primarily because procedures involving the LA are always associated with anticoagulation, and because LA pressure is typically higher than RA pressure. The pressure within the structure that is perforated is a major determinant of the development and severity of PT. Thus, conditions such as pulmonary hypertension and aortic valve stenosis may significantly affect hemodynamic instability in PT [4].

The risk of iatrogenic PT or pericardial effusion increases with the need for transseptal puncture and intraprocedural anticoagulation [5]. With growing acceptance of retrograde catheterization of the left ventricle, the use of the transseptal technique for diagnostic purposes has declined. However, in recent years, substantial renewed application of the transseptal method has occurred for special diagnostic and therapeutic purposes [6]. The procedures requiring transseptal puncture are the following: patent foramen ovale and ventricular septal defect closure, percutaneous heart valve repair or replacement (for mitral regurgitation and aortic stenosis), LA appendage occlusion and electrophysiological procedures such as pacemakers or cardioverter-defibrillators implantation and ablation procedures within the left heart [6].

Pericardial tamponade as a complication of atrial fibrillation ablation

Mentioning atrial fibrillation (AF) ablation, PT was observed as the most frequent complication leading to death. The incidence of PT is higher than with other procedures that employ transseptal catheterization. PT may be characterized as acute when occurring during or immediately after the procedure, or delayed when detected later than 1 h after completion of the procedure [7].

There is evolution of catheter-based AF-ablation techniques. Radiofrequency catheter ablation (RFCA) and cryo-balloon ablation (CBA) are the two standard ablation systems used for pulmonary vein antrum isolation at present. Anticoagulation strategies posit that anticoagulation therapy should be administered prior to or immediately following transseptal puncture during AF catheter ablation [8]. PT may occur any time during trans-septal puncture, extensive catheter manipulation, application of RF energy or steam pops under an intense anticoagulation regimen [7].

According to the 2017 HRS/EHRA/ECAS/APHRS/SOLAECE expert consensus statement on catheter and surgical AF ablation, the rate of PT after AF ablation ranges from 0.2% to 5% [8]. It was based mostly on a prospective multicenter observational study performed by Cappato et al. [9], analyzing 45 115 AF ablations, where the incidence of PT was 2.3%. In this study PT was found as the most frequent fatal complication leading to intraoperative pump failure or post-operative early cardiac arrest. From the reported 32 deaths, 7 occurred due to tamponade (5 as acute and 2 as late PT), in comparison to atrioesophageal fistulas, as the second most frequent fatal complication, counting 5 deaths. In another multicenter prospective study of 6065 Medicare patients, analyzed by Ellis et al. [10], the incidence of PT due to AF ablation increased from 1.3% in 2001 to 3.6% in 2006. Patients who died experienced higher rates of perforation/tamponade (12.0% vs. 3.1% in the remainder of the study sample, p = 0.01).

In the study by Mujović et al. [11], in 2 of 12 cases, tamponade resulted in electromechanical dissociation and cardiorespiratory arrest, while in the remaining ten patients tamponade was associated with significant hypotension. Two patients required surgical treatment. Nine patients received blood transfusions; in seven of them auto-transfusion was carried out. It seems that direct auto-transfusion is simple and requires no additional equipment; it may abolish the need for allogeneic blood transfusion and can “buy” time until surgery [12]. However, direct auto-transfusion may cause systemic inflammation, and therefore the processing of the drained blood via the cell salvage system is recommended prior to its return [13]. Auto-transfusion of a larger volume of blood, i.e., more than 1500 ml, may lead to consumptive coagulopathy [12, 13], which occurred in one patient in this study. The rate of PT, its related mortality and management of PT after AF ablation procedures are summarized in Table I [9–11, 14–21].

Deshmukh et al. [22] had identified 93 801 AF patients treated with catheter ablation obtained from the Nationwide Inpatient Sample (NIS) data set from 2000 to 2010. They observed that in patients older than 80 years, the catheter ablation of AF was associated with a significantly higher total complication rate (9.37%, p < 0.001). From this group, cardiac complications were the most frequent adverse outcomes (2.54%). Within cardiac complications, CT was included but not defined alone. Also women overall had higher complication rates than men (7.51% vs. 5.49%, p < 0.001). A similar conclusion about sex differences was reached by Michowitz et al. [15], and Elayi et al. [23], who analyzed a group of 85 977 patients undergoing catheter ablation of AF. In this group PT appeared in 0.7% of men vs. 1.3% of women (p < 0.001) and the rate of at least one major and overall complications was significantly higher among women than among men, but without a significant difference in mortality between groups.

There are studies [24–26] proving no significant differences in the frequency of pericardial effusions/pericardial tamponade between CBA and RFCA. Meanwhile the recent dedicated meta-analyses have shown higher or considerably higher risk of PT in an RF catheter ablation group (Table II) [27–29].

Pericardial tamponade as a result of percutaneous coronary intervention

Pericardial tamponade is a rare complication of percutaneous coronary intervention (PCI). It is mostly caused by coronary artery perforation (CP) that may occur as the consequence of guide wire advancement, balloon inflation or rupture, and utilization of atherectomy devices [30, 31]. It usually occurs after grade III CP as defined by the Ellis criteria [32]. However, RV perforation due to temporary pacing wires was also found to be an important cause of PT [33, 34].

Pericardial tamponade following a CP is frequently associated with poor outcomes and may increase the risk of death by more than 3 fold (OR = 3.3; 95% CI: 1.01–10.65; p = 0.047) compared with patients who sustained CP without PT [35]. The incidence and predictors of CP have been studied in several large PCI series [30, 31, 36–43]. Generally summarizing, the reported risk factors for perforation include elderly patients, female gender, previous coronary artery bypass grafting, and use of rotational and laser atherectomy. Also PCI of chronic total occlusion (CTO) may be associated with higher risk for procedural complications, including coronary perforation. In the summary analysis from 65 studies by Patel et al., 419 perforations of 18 061 patients undergoing CTO PCI were reported. In this series, the reported rate of CP was 2.9% with a tamponade frequency of 0.2% [44].

Most PCI studies are focused on coronary artery perforation as the major complication, mentioning and calculating PT within this group. In a meta-analysis by Shimony et al. [45] involving 197 061 PCIs, the pooled incidence of CP was 0.43%. The overall percentage of PT after CP was 19.3 (11.9–28.9). In our review CP leads to PT in 11.5–35% (Table III).

The large, multicenter, prospectively collected and retrospectively analyzed study from the British Cardiovascular Intervention Society Database counting 527 121 cases revealed 470 (0.89%) patients having PT as a complication after PCI, 222 patients in the group with CP (14.18%) and 248 in the group without CP (0.05%) [36]. In the previous studies the percentage of PT was higher, 0.12% and 1.21% [33, 42].

In most PCI studies, PT was diagnosed in the cardiac catheterization laboratory by echocardiography or by fluoroscopy that revealed immobile heart borders or the extravasation of the blood from CP. However, late presentation (up to 24 h; mean time 2–4) has tended to occur. In the study by Fejka et al. [42] 14 of 31 PT had late appearance of PT (mean time: 4.4 h) and the most frequent mode of presentation was progressive hypotension culminating in 5 patients in cardiac arrest. In their study PT was associated with very high overall mortality (42%), while 64% of patients presenting PT required intra-aortic balloon pump, ventilatory support and blood transfusion, 61% cardiopulmonary resuscitation and 35% transvenous pacemaker implantation.

The treatment of CP depends on the perforated structure. With coronary perforations, a variety of approaches are possible by prolonged balloon inflation in addition to reversing anticoagulation, covered stents if the artery is large enough to accommodate these devices, or embolization for small vessels [4].

Pericardial tamponade as a result of permanent heart rhythm device implantation

The use of implantation of permanent heart rhythm devices (PHRD), which include permanent pacemakers (PPM) and implantable cardioverter defibrillators (ICD), is increasing due to the expansion of indications and aging of the population. Despite the fact that post-implantation pericardial effusion can be a sign of lead perforation (LP), there are other mechanisms leading to pericardial effusion, such as traumatic inflammation of the myocardium and pericardium from the lead screw, or irritation of the visceral pericardium via immune mediated mechanisms [46].

Lead perforation is a rare complication after PHRD implantation and may involve large veins, atrial or ventricular walls or coronary sinus. LP develops most often acutely (i.e., less than 24 h after the procedure), which may potentially result in PT or death. Also it can occur 24 h after the device implantation in a subacute or chronic fashion. Another classification distinguishes between early (symptoms occur up to 1 month after implantation) and delayed perforations [47].

The clinical presentation of LP may be different in the late form, with most patients presenting pacemaker malfunction, stabbing chest pain and shortness of breath. A distinguishing feature of delayed in opposition to acute LP is the decrease or absence of PT or death [48]. Published event rates for LP range from 0.1% to 0.8% for PPM and 0.14–5.2% for ICD leads; in those publications perforations occurred mostly within 1 month after implantation [48–50] (Table IV).

Factors that are thought to contribute to acute LP are similar to those in late LP: patient characteristics, concomitant therapies such as steroids or anticoagulants, implant techniques and the design characteristics of the lead [49]. Patient-related factors in PPM implantations include old age, female sex, low body mass index and for ICD implantations it is additionally worsened heart failure class, left bundle branch block and non-single-chamber ICD implant [50]. Thin heart muscle itself, such as in a patient with myotonic muscular dystrophy and dilated cardiomyopathy, may favor perforation [51]. The use of atrial leads, helical screw ventricular leads, active fixation and temporary stimulation was reported to increase the incidence of perforation [47, 52]. The risk factors concerning defibrillator leads are as follows: double spirals, number of shocks delivered, excessive length or small diameter of the lead, high resistance (small tip surface) and apical position [47, 53].

Ohlow et al. [46] prospectively observed 968 consecutive patients undergoing PHRD implantations who had undergone echocardiographic evaluation before and 24 h after the operation. Fourteen of them (1.44%) had had PT requiring pericardiocentesis (n = 12; 86%) or surgical treatment (n = 2, 14%). In 10 of those patients a hemorrhagic effusion suggested cardiac perforation of an implanted lead; acute pericarditis was observed in the remaining four patients.

In the latest study Moazzami et al. [54] reported their findings after analyzing the United States National Inpatient Sample (USNIS) database from 922 549 patients implanted with PPM. PT occurred in 2695 (0.28%) patients. The authors found that female sex, implantation of dual-chamber pacemakers, and chronic liver disease predicted greater odds of PT, whereas hypertension and atrial fibrillation were associated with lower odds of tamponade. The association of chronic liver disease with PT may be related to the potential for bleeding due to coagulopathy, systemic tissue characteristics from liver disease, or anatomic consideration from hepatomegaly. The protective association of atrial fibrillation may be related to implantation of fewer atrial leads, enlarged LA or atrial fibrosis, whereas hypertension may be related to hypertrophy of the cardiac chambers [55].

Temporary transvenous pacing (TTP) with electrodes guided to the RV is burdened with the risk of PT as well. Performing analysis about complications and outcomes of over 360 000 TTPs using the USNIS database, Metkus et al. defined the rate of PT as 0.6% [56].

A review of studies concerning the rate of PT and LP after PHRD implantation, management and PT-related death is presented in Table IV.

With the growing number of PHRD implantations in recent years, in parallel, a rise in lead malfunction and recalls has resulted in increased transvenous lead extractions (TLE) [57]. In 2016 the biggest multicentre prospective overview of TLE safety and efficacy conducted by the EHRA, entitled The European Lead Extraction ConTRolled Registry (ELECTRa), was published [57]. The primary endpoint was TLE safety defined by in-hospital procedure-related major complications including death. In 3510 patients 6493 leads including 4917 (75.7%) pacing and 1576 (24.3%) ICD leads were targeted for extraction. The mean dwell time of extracted leads was 6.4 ±5.4 years (median: 5 years, IQR: 2–9). Indications for TLE were infective in 52.8%. Among 58 deaths, 17 were procedure-related. The most common procedure-related complications were cardiovascular complications requiring pericardiocentesis or surgical repair occurring in 49 (1.4%) patients. Apart from thoracic and peripheral vascular lesions, 28 patients had cardiac avulsion and 2 cardiac avulsion with thoracic vascular tears. The clinical manifestations of these complications were PT, haemothorax and hemorrhagic shock. The authors did not report the exact number of PTs, but we may assume that this number may correspond to the number of cases concerning cardiac avulsion [30]. However, the authors provided accurate information about 17 causes of procedure-related deaths. According to their results, PT led to death in 6 cases, undergoing surgical treatment in all of them. This emphasizes what a hazardous complication pericardial tamponade is. Moreover, procedure-related major complications and death were more common in female patients (OR = 2.11, 95% CI: 1.23–3.62, p = 0.0067), leads with a dwell time > 10 years (OR = 3.54, RR: 1.6–7.83, p = 0.0018), with the use of powered sheaths (OR = 2.4, 95% CI: 1.41–4.09, p = 0.0013) and a femoral approach (OR = 3.60, 95% CI: 1.64–7.87).

Pericardial tamponade as a result of left atrium appendage occlusion

Occluding the left atrial appendage (LAA) is an alternative treatment for stroke prevention in high-risk patients with contraindications to oral anticoagulants. The risk of causing PT is due to the fact that the LA appendage itself can be extremely thin-walled.

In addition to surgical technique, percutaneous methods of LAA closure were developed. For clinical use both Watchman and the Amplatzer Cardiac Plug (ACP) devices have been approved. Moreover, in 2013 a second generation of the ACP, the Amplatzer Amulet left atrial appendage occluder, was released [58]. The Watchman device is basically a plug that should be precisely implanted to avoid both its protrusion into the LA as well as the creation of a cul-de-sac where thrombus may form. The ACP consists of two parts joined by a central pin. Being short, the ACP can be implanted in a shallow position in the LAA, as only the proximal 2 cm are needed for its occlusion. The occlusive disc permits the complete closure of the LAA orifice [59]. In a prospective randomized controlled trial by Holmes et al. [60], with the intervention group consisting of 463 LAAC implantations, the most frequent primary safety event was severe pericardial effusion (defined as the need for percutaneous or surgical drainage). It occurred in 22 (4.8%) patients; 15 of them were treated with pericardiocentesis and 7 underwent surgical intervention. None of those patients died, although length of hospital stay in these patients was longer than in the control group (244 patients) without severe pericardial effusion (median 4 days longer). Effusion rates declined with investigator experience. In a meta-analysis by Wei et al. [61] assessing the efficacy and safety of transcatheter LAA closure in patients with nonvalvular AF, the incidence of pericardial effusion/tamponade was estimated as 0.02 (95% CI: 0.02–0.03). The studies assessing the rate of PT, its management and PT-related mortality after LAA closure are summarized in Table V.

Pericardial tamponade as a result of transcatheter aortic valve implantation (TAVI)

The European Society of Cardiology guidelines advocate the use of TAVI in patients with severe aortic stenosis and high risk for SAVR (surgical aortic valve replacement), favoring TAVI mostly in older patients [62]. In a retrospective cohort study evaluating 16 755 patients diagnosed with AS in the Japanese healthcare setting, in-hospital outcomes between TAVI and SAVR were evaluated. The incidence of pericardial tamponade was significantly higher in the SAVR patients (1.5% in SAVR vs. 0.5% in TAVI; p = 0.03) [63]. There are three major pathophysiological situations that may lead to PT during TAVI: first, annular or aortic root rupture during balloon valvuloplasty and valve implantation with subsequent arterial bleeding into the pericardium; second, perforation of the right RV caused by the temporary pacing lead; and third, perforation of the LV by an extra-stiff guidewire during its placement or at later stages of the procedure [64, 65]. In existing literature, PT has been described as occurring in 0.2–4.3% of cases, with a higher probability in retrograde trans-vascular techniques than with trans-apical access [66, 67] (Table V).

Pericardial tamponade as a result of percutaneous mitral valve repair with the Mitra-Clip system

Percutaneous edge-to-edge mitral valve repair using the Mitra-Clip device represents a less invasive treatment option for patients with symptomatic severe mitral regurgitation [68]. With the need for cardiologic techniques such as trans-septal puncture as well as navigation of catheter devices within the LA and LV, the risk of tamponade may grow. There are two large prospective multicenter studies that have analyzed the risk and outcomes of complications during and after Mitra-Clip based on the German Transcatheter Mitral Valve Interventions register [69] and the European ACCESS-EU registry [70] (Table V [60, 66, 67, 69–73]). The risk of PT was 1.9% and 1.1% respectively, suggesting that transseptal puncture followed by advancement of the 24 Fr guiding sheath is safe [69].

Discussion

Pericardial tamponade is a rare complication of invasive cardiac procedures, mostly associated with poor outcomes. It generally develops acutely, but late forms with atypical presentation may occur. The incidence depends on patients’ comorbidities, concomitant pharmacotherapy, sex, age, operators’ skills, the size of the device responsible for the perforation, the structure of the heart that is perforated, the hemodynamic state during perforation and mechanical properties of the pericardium itself. Some procedures, especially those requiring transseptal puncture or perioperative anticoagulation, favor CT.

Summarizing our review, the overall rate of iatrogenic PT varies among published studies from 0.089 to 4.8%, with the highest rate after left atrial appendage closure, transcatheter aortic valve implantation, and atrial fibrillation ablation. Concerning electrophysiological procedures, the rate of PT was quite divergent between multicenter and single center studies (Tables I, IV). Generally, the PT-related post-procedural mortality ranged from 0% to 50% (medium 7.3%) and was highest in LAAC and TAVI groups, then PCI, PHRD and the AF ablation procedures respectively. The need for surgical treatment after iatrogenic PT ranged from 0% to 60% (medium 15%). The most frequent group of PTs requiring surgical treatment was the PCI group, the second group consisted of LAAC and TAVI procedures and the last group consisted of AF ablation and PHRD procedures. These conclusions may not be precise. Many of the mentioned studies only determine predictors of iatrogenic PT or the endpoint of those studies was only the rate of PT, and a follow-up was not performed.

Additionally, we have observed that older age and female sex were common risk factors predisposing to PT. Some techniques as well as echocardiographic or fluoroscopic guidance during such procedures may decrease the overall complication rate.

Conclusions

Pericardial tamponade is a rare complication of invasive cardiac procedures with the overall rate of iatrogenic PT varying among published studies from 0.089% to 4.8%, with the highest rate after left atrial appendage closure, transcatheter aortic valve implantation, and atrial fibrillation ablation. Iatrogenic CT was associated with poor outcomes as high incidence of in-hospital death and need for surgical intervention.

Conflict of interest

The authors declare no conflict of interest.