Management of Penetrating Cardiac Injuries With Pericardial Window and Drainage in Select Patients

Key Takeaways

• Penetrating cardiac injury can be managed with pericardial window, lavage, and drainage in select patients.

Introduction

Penetrating cardiac trauma is highly lethal and its management poses a challenge to trauma surgeons. Penetrating cardiac injuries are rare, but prehospital mortality is high.^1,2 Patients that survive to reach the hospital require rapid assessment, diagnosis, and intervention. Pericardial tamponade, hemorrhagic shock, and cardiopulmonary arrest can occur, even in patients that present hemodynamically stable. ³

Focused assessment with sonography for trauma (FAST) is the initial imaging study, which can be performed by the bedside surgeon. Traditionally, FAST that demonstrates fluid within the pericardium mandates sternotomy. Pericardial window can be performed when FAST is equivocal or if FAST is negative, and there is high suspicion for cardiac injury based on injury trajectory or subsequent computed tomography (CT) imaging. Blood found during pericardial window also traditionally mandates sternotomy. However, sternotomies can result in significant morbidity for patients and increased cost to hospitals.^4,5 Recent literature suggests select hemodynamically stable patients may undergo pericardial window, lavage, and drainage instead of sternotomy.^6,7 Bleeding can cease in low velocity cardiac injuries and pericardial injuries, and simple drainage of the pericardium can alleviate the potential for tamponade.

This study aims to describe our institutional experience with penetrating cardiac injuries, management strategies, and outcomes.

Methods

The study was approved by the institutional review board at the University of Tennessee Health Science Center. All patients who sustained penetrating chest trauma and underwent pericardial window and/or sternotomy from January 2017 to December 2021 were identified from the trauma registry at the Elvis Presley Trauma Center in Memphis, Tennessee. Charts were reviewed for data regarding patient demographics, injury severity, time to operation, imaging findings, indication for operation, and operative interventions. Collected data from the medical records was merged with additional information from the trauma registry to create the database.

Focused assessment with sonography in trauma (FAST) was considered positive, negative, or equivocal based on the attending trauma surgeon’s interpretation documented in the medical record. Computed tomography (CT) scans were reviewed by an attending radiologist, and findings considered significant were pneumopericardium, hemopericardium, or missile fragments in myocardium.

Patients were stratified by operative intervention [PW + sternotomy vs PWLD] and compared. Pericardial window, lavage, and drainage was performed by opening the pericardium, lavaging saline into the pericardium until the drainage was clear, and placement of a red rubber catheter or chest tube in the pericardium. Outcome measures include mortality, length of stay (LOS), ICU LOS, ventilator days, complications, and additional interventions.

All data were analyzed using SAS version 9.4 (SAS Institute, Cary, NC). Comparisons between patients were performed using Student’s t-test, chi-square analysis, Wilcoxon rank sum, or Fisher’s exact test where appropriate. A logistic regression model was developed to determine predictors of therapeutic sternotomy. Variables with a value of <.2 on univariable analysis were selected as prospective covariates. Stepwise multivariable logistic regression (MLR) analysis was performed to determine independent predictors of therapeutic sternotomy. Differences were considered significant at P < .05.

Results

Of the 146 patients included in the study, 20 patients underwent immediate sternotomy, 19 underwent PW followed by sternotomy, 97 patients underwent diagnostic PW, and 10 patients underwent PWLD. Overall, patients were most commonly male (91%), 28 years old (IQR 22, 37), with an ISS of 19 (IQR 23, 29) and GCS 15 (IQR 15, 15). FAST was performed in 90% of patients, of which 35% were interpreted as positive or equivocal. Preoperative CT imaging was performed in 23% of patients, of which 88% were considered significant. Of all sternotomies, 87% were therapeutic. Overall mortality was 10% (Table 1).

OPEN IN VIEWER

Table 1.

Overall Patient Population Characteristics.

	All
N	146
Male (%)	91
Age (years)	28 (22, 37)
Injury severity score	19 (23, 29)
Glasgow Coma Scale	15 (15, 15)
Base excess	−5 (−8, −1)
FAST performed (%)	90
Preoperative CT imaging performed (%)	23
Mortality (%)	10

Of the 126 patients who underwent pericardial window, 74% were performed using subxiphoid technique, 17% using transdiaphragmatic technique, and 9% using intercostal technique. FAST was found to have 93% sensitivity, 21% specificity, and 79% positive predictive value. There were 7 patients with a left hemothorax and negative FAST, found to have a positive PW and cardiac injury mandating sternotomy and repair.

Patients who underwent PW + sternotomy were compared to patients who underwent PWLD. Patients were similar in age, gender, initial hemodynamics, and GCS. Those who underwent PW + sternotomy had a higher ISS (26 vs 13, P = .034) and worse base excess (−8 vs 1, P = .003) compared to patients who underwent PWLD. There was no difference in LOS, ICU LOS, vent days, or mortality. The time to operative intervention was shorter in the PW + sternotomy group (30 vs 75 minutes, P = .046) (Table 2). There were 7 patients in the PW + sternotomy group and 8 patients in the PWLD group who underwent preoperative CT imaging, all of which were interpreted as positive. In the PW + sternotomy group, the right ventricle was most commonly injured (42%), followed by left ventricle (16%), pericardium (16%), right atrium (11%), left atrium (5%), and internal mammary artery (5%) (Table 3). There was one patient who underwent negative sternotomy. One patient had a sternal wound infection and subsequent dehiscence. In the PWLD group, one patient returned to the OR for recurrent pericardial effusion and no patients required sternotomy. Pericardial drains were removed after 2.6 days.

OPEN IN VIEWER

Table 2.

Comparison of Patients Who Underwent PW + Sternotomy vs PWLD.

	PW + Sternotomy	PWLD	P
N	19	10	.863
Male (%)	94%	89%	.548
Systolic blood pressure	116 (88, 137)	123 (115, 129)	.337
Pulse	108 (23)	99 (28)	.369
Injury severity score	26 (16, 33)	13 (10, 22)	.034
Glasgow Coma Scale	15 (15, 15)	15 (15, 15)	.580
Base excess	−8 (−11, −4)	1 (−4, 2)	.003
Time to operative intervention (minutes)	30 (21, 51)	75 (45, 83)	.046
Intensive care unit LOS (days)	4 (4, 9)	6 (2, 13)	.901
LOS (days)	10 (9, 14)	7 (5, 22)	.587
Ventilator days	3 (0, 4)	0 (0, 0)	.113
Mortality (%)	11%	0%	.452

OPEN IN VIEWER

Table 3.

Cardiac Injuries Identified After PW + Sternotomy.

N	19
Right ventricle	8 (42%)
Left ventricle	3 (16%)
Pericardium	3 (16%)
Right atrium	2 (11%)
Left atrium	1 (5%)
Internal mammary artery	1 (5%)
No injury	1 (5%)

When MLR was performed, ISS was an independent predictor of therapeutic sternotomy (OR 1.160; 95% CI 1.006-1.338, P = .0616). A positive FAST and missile/wound trajectory as an indication for operation were not predictors of therapeutic sternotomy.

Discussion

Penetrating cardiac injuries are relatively rare, representing .16% of all trauma admissions, but are highly lethal. ⁸ Up to 90% of victims don’t survive at the scene or die enroute to the hospital. ⁹ Campbell et al found rapid operative intervention is the best predictor for survival in patients who arrive at the hospital alive. ¹⁰ Limited experience and knowledge of these injuries has resulted in sparse literature with small sample sizes and few randomized trials.^8,11

Initial assessment is performed with FAST; however, this is operator dependent. A recent study found FAST to be 79% sensitive and 92% specific, whereas PW was 96% to 100% sensitive and 96% to 100% specific for determining cardiac injuries. ¹² Pneumomediastinum and hemothorax are the most likely culprits of false negative results. ⁷ The current study was sensitive but lacked specificity, likely a result of the number of patients with a negative FAST and left hemothorax who had a true cardiac injury.

A majority of the PWs in this study were performed using a subxiphoid approach. In a review by Selvakumar, subxiphoid, transdiaphragmatic, and laparoscopic approaches resulted in similar efficacy and safety. ¹³

Median sternotomy is the traditional management for penetrating cardiac injuries, which can result in significant morbidity. While the diagnostic yield of PW has been known for some time, therapeutic management with PW, lavage, and drainage has only recently been reported in the literature.^6,11,13,14 Nicol et al performed a randomized study in South Africa comparing patients who underwent sternotomy and PWLD in patients with documented cessation of bleeding on PW. Pericardial window, lavage, and drainage resulted in shorter LOS and no difference in mortality. ¹⁵ In the United States, Chestovich et al identified 5 patients who underwent PW, lavage with saline until cessation of bleeding, and drainage with no difference in LOS or mortality. ⁶ The current study found similar results: in hemodynamically stable patients, PWLD can be safely performed with no difference in mortality.

Penetrating mechanism and hemodynamic instability have been demonstrated to be predictors of therapeutic sternotomy. ¹⁴ The current study only included patients with penetrating mechanism and identified higher ISS to be a predictor of therapeutic sternotomy. Interestingly, positive FAST was not a predictor of therapeutic sternotomy. This finding may be due to the significant number of false negative results. Patients with CT findings demonstrating hemopericardium, pneumopericardium, or missile in the myocardium should prompt further investigation with a PW but does not necessarily mandate sternotomy.

The study has several limitations. This a retrospective study and the accuracy of the data is subject to the documentation in the chart. The FAST imaging studies were not re-examined by a surgeon or radiologist to determine their accuracy. There is no institutional guideline for the management of penetrating cardiac injuries in the hemodynamically stable patient, and operative management was performed at the discretion of the attending surgeon.

Conclusion

Penetrating cardiac injury can be managed with pericardial window, lavage, and drainage in select patients. Positive FAST, significant findings on CT imaging, and trajectory of wounds do not mandate sternotomy. A negative FAST in the setting of a left hemothorax does not rule out a cardiac injury and should be investigated further.