Wound haematoma following defibrillator implantation: incidence and predictors in the Shockless Implant Evaluation (SIMPLE) trial Free

Abstract

Introduction

Implantable cardioverter defibrillators (ICDs) have been shown to prevent sudden cardiac death and improve survival in large randomized trials,¹ leading to a widespread use of these devices.² However, the benefits of ICD therapy are in part offset by device-related morbidity, including inappropriate shocks and implant-related complications.² Pocket haematoma is a common complication after device implantation; with an overall incidence reaching up to 15% in some patient sub-groups.^3–7 A clinically significant pocket haematoma may have serious implications, including the need for prolonged interruption of oral anticoagulation (OAC),^3,8 increased risk of infection,⁹ and the possible need for surgery to evacuate the haematoma.¹⁰ These patients also experience more peri-operative pain and worse quality of life.³

Several authors have identified potential predictors of pocket haematoma, including patient characteristics, surgical techniques, and the use and management of anti-thrombotic medications. However, most of these studies are small and single-centre in nature, and their findings have not been consistent.^4,6,7 Understanding all of the predictors of wound haematoma is an important step to further prevent this complication. Peri-operative heparin bridging has been recognized as an important risk for haematoma, and recent randomized trials demonstrated that a change in the peri-operative management of OAC substantially reduces the rate of the wound haematoma.³

The purpose of the present analysis is to further explore frequency and predictors of pocket haematoma in patients having ICD implantation in the large, multi-centre Shockless Implant Evaluation (SIMPLE) trial.

Methods

The design and main results of SIMPLE have been published previously.¹¹ Briefly, SIMPLE was a 2500-patient, multi-centre, randomized study evaluating whether ICD implantation without defibrillation testing (DT) is non-inferior to implantation with DT. The primary outcome was the composite of ineffective appropriate shocks or arrhythmic death. All patients provided written informed consent. For the current analysis, a clinically significant pocket haematoma was prospectively defined as device-pocket haematoma that necessitated further surgery (i.e. haematoma evacuation) or interruption of anticoagulation or anti-platelet therapy. This definition was similar with that used in the seminal publication in this field.⁴ Data regarding the development of pocket haematoma were collected at patient discharge and at the 30-day post-surgical follow-up visit.

Statistical analysis

The baseline characteristics were summarized as mean and standard deviation for normally distributed continuous variables, median and interquartile range for non-normally distributed variables, and frequency and proportion for categorical variables. Comparison between patients with and without pocket haematoma was performed using independent t-test or Wilcoxon rank-sum test for continuous variables and χ² test or Fisher's exact test for categorical variables. Univariate logistic regression analysis was applied to investigate the effect of individual predictors, and subsequently multivariable logistic regression model was applied to potential predictors selected from univariate analysis based on statistical significance (P < 0.1). Potential non-linear effects of continuous predictors were inspected using restricted cubic splines. Results were presented as odds ratio (OR) with corresponding 95% confidence interval (CI). Statistical analyses were conducted using SAS 9.2 software (SAS Institute, Inc., Cary, NC, USA). Unless otherwise specified, a two-sided P-value of ≤0.05 was considered statistically significant.

Results

A clinically significant pocket haematoma developed in 56 of 2500 patients (2.2%). Among these 56 patients, 15 required only evacuation, 34 required only interruption of pre-/post-operative OAC or anti-platelet therapy, and 7 required both. Patients who developed pocket haematoma were older and more likely to have permanent atrial fibrillation (AF), history of stroke, and impaired renal function (Table 1). Furthermore, they were more likely to have undergone an upgrade from a permanent pacemaker to ICD and to have a sub-pectoral ICD placement (Table 1). There were no significant differences in terms of body mass index (BMI) between the two groups.

Among the 56 patients with haematoma, 28 received peri-operative OAC, which was more frequent than among patients who did not develop pocket haematoma (50.0 vs. 28.4%, P < 0.001). Similarly, patients developing haematoma were more likely to receive bridging therapy with low-molecular-weight heparin (LMWE) (37.5 vs. 17.5%, P < 0.001) or unfractionated heparin (UFH) (16.1 vs. 5.2%, P = 0.003); however, the median pre-operative international normalized ratio (INR) was not significantly different between the two groups (Table 1). The 721 patients who received peri-operative OAC therapy had a higher stroke risk profile compared with those without OAC. They were more likely to be older (64.7 ± 10.4 and 62 ± 12 years, respectively, P < 0.001); to have a history of stroke (10.4 vs. 5.6%, P < 0.001 and 7.2 vs. 3.7%, P < 0.001), AF (50.9 vs. 12.2%, P < 0.001), heart failure (80.6 vs. 67.2%, P < 0.001), and impaired renal function (26.6 vs. 16.1%, P < 0.001); and to receive biventricular devices implantation (35.5 vs. 25.7%, P < 0.001). Device type (DDD vs. VVI vs. CRT) and the use of anti-platelet therapy were not significantly correlated with risk of haematoma (Table 1).

On univariate analysis, significant predictors of pocket haematoma were bridging with UFH or LMWE before the procedure (OR 3.47, 95% CI 1.66–7.23, P < 0.001; and OR 2.80, 95% CI 1.61–4.86, P < 0.001, respectively), the use of any pre- or post-operative anticoagulation (OR 2.50, 95% CI 1.47–4.26, P < 0.001), upgrade from pacemaker to ICD (OR 3.21, 95% CI 1.42–7.26, P= 0.005), sub-pectoral ICD placement (OR 2.54, 95% CI 1.47–4.39, P < 0.001), impaired renal function (OR 2.21, 95% CI 1.26–3.88, P = 0.006), prior stroke (OR 3.02, 95% CI 1.50–6.09, P = 0.002), and older age (OR 1.05, 95% CI 1.02–1.07, P = 0.002).

On multivariable analysis (Table 2), the use of heparin bridging, upgrade from permanent pacemaker, prior stroke, sub-pectoral ICD location, and age remained significant independent predictors of pocket haematoma, while impaired renal function and the use of any pre- or post-operative anticoagulation were not confirmed to be significant predictors.

Among 56 patients with pocket haematoma, 6 (10.7%) developed device-related infection, 5 in the totality of patients who received a re-operation (22.7%), and 1 (2.9%) in the subgroup who only required interruption of anti-thrombotic therapy (P = 0.03).

Discussion

The present analysis from SIMPLE trial shows that pocket haematoma remains a clinically relevant complication following ICD implantation. It requires re-operation or interruption of anti-thrombotic therapy in 2.2% of all patients undergoing ICD implantation and among 3.9% of those receiving peri-operative anticoagulation. Multivariate analysis identified three patient sub-groups at increased risk of pocket haematoma: patients with prior stroke, older age, and those having an upgrade from a permanent pacemaker; as well as two potentially modifiable technical risk factors: the use of heparin bridging and sub-pectoral ICD placement. Furthermore, our study shows that among patients with pocket haematoma those who underwent haematoma evacuation had a higher risk of infection than those who were managed with interruption of anti-thrombotic therapy alone.

The overall 2.2% rate of pocket haematoma in this analysis is similar to prior studies, which found rate of up to 5%.^6,7 Some variation between reports is expected, given differences in patient populations and in the definition of pocket haematoma.

As in our report, prior studies identified heparin bridging as one of the strongest predictors of pocket haematoma.^3,8,12 In some studies, the rate of haematoma among patients receiving heparin bridging has been as high as 17–31%.^12,13 This was the rationale for the BRUISE-CONTROL study, which enrolled individuals receiving warfarin who were scheduled for pacemaker or ICD surgery to either continue with warfarin throughout the peri-operative period or interrupt warfarin and have bridging with heparin. In this study, patients randomized to heparin bridging had more than a four-fold increase in haematoma rate (16 vs. 3.5%). Guidelines now endorse a strategy of continued OAC for warfarin-treated patients with a moderate or high risk of thrombotic events.¹⁴ At present, there is little guidance for the growing number of patients receiving chronic therapy with direct oral anticoagulants; however, a randomized trial is underway in this population (BRUISE-CONTROL-II; NCT01675076).

It is less clear how to manage anti-platelet therapy around the time of surgery. Prior reports suggest that aspirin (ASA) monotherapy does not statistically increase the incidence of pocket haematoma and thus does not require pre-operative discontinuation.^4,6,7 In contrast, some authors reported that combined use of ASA and thienopyridine is associated with an increased incidence of pocket haematoma,^6,15 so many advocate interruption of thienopyridine therapy for 3–5 days in patients without a recent coronary stent placement or acute coronary syndrome.¹⁶ In our large study, pre-procedure anti-platelet therapy did not independently affect pocket haematoma rate; however, thienopyridine use was low, and precise details of peri-operative management were not recorded.

The sub-muscular approach to ICD implantation involves transecting the pectoralis major muscle, with some inherent bleeding risk. Existing data are contradictory regarding the association between sub-muscular placement of ICD devices and wound haematoma. Early reports, using larger, older generation ICDs suggested that the sub-muscular approach was associated with a reduction of pocket complications.^17,18 However, Wiegand et al.⁶ showed a slight higher pocket haematoma rate with sub-muscular placement. In our series, despite the use of smaller, modern ICDs, sub-muscular placement was associated with a two-fold increased risk of haematoma development. These results suggest that sub-muscular placement may no longer be the safer routine surgical technique for device placement; although it may be useful in selected populations, such as patients with low body mass.

Several studies did not find increased incidence of pocket haematomas for upgrade procedures compared with first device implantation.^3,19 Only one study has reported increased complications, including haematoma, after generator replacement accompanied by upgrade to CRT devices.²⁰ In contrast, we found that upgrade procedures from permanent pacemakers were associated with a higher risk of pocket haematoma in our large multi-national trial. One possible explanation for the discrepancy may be differences in the technique of device upgrade, including the degree of capsule removal, and in the longer surgical procedure time, which has been reported as an independent predictor of pocket haematoma.⁷

In contrast to prior reports,^3,6 gender, diabetes, and other cardiovascular diseases did not independently influence pocket haematoma rate, nor did the type of ICD implanted. Contradictory results have been previously reported regarding ICD type: Kutinsky et al.⁴ failed to show an association between the number of implanted leads and pocket haematoma, while Tischenko et al.,¹² in a high-risk population receiving anticoagulant treatment, showed that the number of implanted leads was an independent predictor of haemorrhagic complications.

Limitations of the methods

The present study must be interpreted in light of certain methodological limitations. The choice of implantation technique and of peri-operative anticoagulation strategy were not randomized and pre-specified in the SIMPLE trial and thus left at discretion of the operator. Although this is a large prospective series of ICD implantations, the number of events is low, limiting the power of the present analysis to detect all associations. On the other hand, the objectivity of the definition of pocket haematoma compared with other prospective reports makes our analysis quite reflective of the real clinical practice.

Conclusions

Pocket haematoma remains a clinically relevant complication of ICD implantation. Heparin bridging, sub-muscular ICD placement, and upgrading from a pacemaker are significant predictors of pocket haematoma. Bridging with heparin or LMWE should be avoided whenever possible, and the decision to undertake sub-pectoral device placement should include consideration of the increased risk of pocket haematoma. Furthermore, physicians should recognize that surgical evacuation of pocket haematoma is associated with a higher risk of infection.

Funding

The SIMPLE trial was funded by Boston Scientific.

Conflict of interest: none declared.

References