Laparoscopic Cholecystectomy in Cardiogenic Shock And Heart Failure

Introduction

In the United States, the incidence of heart failure has increased from 5.7 million (2009–2012) to 6.5 million (2011–2014) in Americans greater than 20 years of age. ¹ This increase may be explained by factors, including diet, smoking, lack of exercise, and obesity. The low-flow state caused by heart failure and cardiogenic shock (CS) can result in hypoperfusion and end-organ ischemia. End-organ ischemia can present as kidney injury, altered mental status, and liver dysfunction. If the hypoperfusion is substantial or persists for an extended period, the patient does not survive.

Prior to death, the body disproportionately vasoconstricts the mesenteric vessels to divert blood to the systemic circulation. ² The cystic artery has a mean diameter of 1.6 mm and is a terminal artery, predisposing the gallbladder to ischemia during low-flow states. ³ Patients in CS can develop acute acalculous cholecystitis (AAC) due to the body’s diversion of blood flow to the systemic circulation. Because it reflects the underlying pathophysiology, we prefer to use the term ischemic cholecystitis. However, AAC encompasses this definition and will also be used in this article as it is most prevalent in the literature. AAC represents 2%–15% of acute cholecystitis cases. ⁴ When ischemic cholecystitis is left untreated, the gallbladder can necrose and perforate, causing mortality by septic shock. As management strategies of CS improve, patients who survive the initial cardiac insult are still at risk for mortality from additional sequelae, such as septic shock secondary to ischemic cholecystitis.

Gallbladder ischemia is difficult to detect in this patient population. Exam findings and conventional imaging are confounded by CS itself. Physical exam is often unrevealing with negative Murphy’s sign in many. ⁶ Patients are often febrile with leukocytosis. Additional sources of infection should be ruled out with blood cultures, urine cultures, chest imaging, and inspection of wounds associated with catheters and cardiac support devices. In the setting of CS, there is often congestive hepatopathy and ischemic hepatic injury resulting in transaminitis and hyperbilirubinemia. An elevated indirect bilirubin would be expected due to liver dysfunction resulting in the inability to conjugate. This is in contrast with calculous cholecystitis producing a direct hyperbilirubinemia due to cystic duct obstruction. In ischemic cholecystitis, patients may have both direct and indirect hyperbilirubinemia. Options for objective testing include ultrasound with Doppler, computed tomography of the abdomen, magnetic resonance cholangiopancreatography (MRCP), and diisopropyl iminodiacetic acid (DISIDA) scan (functional stimulated hepatobiliary iminodiacetic acid (HIDA) scan).

Patients with heart failure who develop CS despite medical therapy can be offered treatments, including ventricular assist devices with eventual cardiac transplantation. ⁵ LC has been proposed as treatment in patients with ventricular assist devices who develop ischemic cholecystitis, allowing them to be treated for cholecystitis and proceed toward eventual heart transplantation. ⁶ Because these patients are high-risk surgical candidates, the management of ischemic cholecystitis has been controversial. In high-risk patients with calculous or acalculous cholecystitis, a percutaneous cholecystostomy tube (PCT) can be placed by interventional radiology to decompress the gallbladder and drain infected bile. However, if there is gallbladder necrosis, patients with ischemic cholecystitis are unlikely to improve with this treatment. Definitive treatment with LC is a strategy for reversing the cascade toward sepsis and bowel ischemia following CS. This retrospective review discusses outcomes after cholecystectomy in patients with CS or heart failure.

Methods

Results

Of the 24 patients with heart failure who underwent LC, 19 (79%) were men and 5 (21%) were women. Twenty (83%) patients were white, 3 (13%) were African American, and 1 (4%) was American Indian (Table 1). Most patients were non-Hispanic (18 patients, 75%). At the time of surgery, most patients (15 patients, 63%) were New York Heart Association (NYHA) Class 4 (unable to carry out any physical activity without discomfort). Seven patients (29%) were NYHA Class 3 (marked limitations of activity), and 2 patients (8%) were NYHA Class 2 (slight limitations of physical activity). No patients were NYHA class 1 (no limitation of physical activity) (Table 2).

Six patients (25%) had a pacemaker–defibrillator or defibrillator as the only cardiac device (Table 2). Four patients (17%) had no device. Fourteen patients (58%) had one or more of the following devices: left ventricular assist device (LVAD), extracorporeal membrane oxygenation (ECMO), Impella, tandem heart, or total artificial heart. Specifically, there were 3 patients (13%) with LVAD and defibrillator, 3 patients (13%) with Impella only, and 2 patients (8%) with ECMO only. There was 1 patient (4%) with each of the following: LVAD only; ECMO, Impella, and defibrillator; ECMO, Impella, defibrillator, and tandem heart; Impella and defibrillator; or total artificial heart.

Comorbidities, in descending order of frequency included 21 patients on anticoagulation (88%), 14 with smoking history (58%), 12 with hypertension (50%), 9 dependent on mechanical ventilation (38%), 10 dependent on continuous renal replacement therapy (CRRT) (42%), 10 with prior myocardial infarction (42%), 7 with obstructive sleep apnea (29%), 7 with diabetes (29%), 6 with chronic kidney disease (25%), 6 with pulmonary hypertension (25%), 6 with prior percutaneous coronary intervention (25%), 5 receiving steroids (21%), 5 with prior stroke (21%), 3 with chronic obstructive pulmonary disease (13%), 1 with asthma (4%), and 1 with cirrhosis (4%) (Table 1). There were no patients with nonalcoholic steatohepatitis. Four patients (17%) had a PCT preoperatively.

Average interval between diagnosis and surgery was 15 days (Table 1). Pneumoperitoneum of 10 mmHg was tolerated by all. The intended operation was completed in all patients with the exception of one, who instead underwent diagnostic laparoscopy with cholecystostomy tube placement due to a complex ventral hernia limiting exposure. No patients were converted to open. Average operative time was 83 minutes. Average length of stay was 24.7 days. The most frequently observed complications included 17 patients requiring reintervention within 30 days (71%), 13 patients with bleeding requiring transfusion (54%), and 9 patients requiring reoperation within 30 days (38%) (Table 3). Nine patients (37.5%) underwent a total of 21 reoperations (Table 4). Of these, only 1 reoperation (4%) was related to the cholecystectomy. Reoperation was more likely (odds ratio; 95% CI) in patients with Impella (17.5; 1.22–884.45), vasopressors (6; .66–75.61), inotropes (32; 2.25–1532.93), and dialysis (5.5; .68–48.77). Less common complications included reintubation in 4 patients (17%), readmission within 30 days in 3 patients (13%), myocardial infarction in 1 patient (4%), stroke in 2 patients (8%), deep vein thrombosis in 1 patient (4%), prolonged ventilation >48 hours in 1 patient (4%), deep space infection in 1 patient (4%), and acute kidney injury in 2 patients (8%) (Table 3). Mortality within 30 days occurred in 5 patients (20.8%), with the interval between cholecystectomy and mortality ranging from 6 to 30 days. Complications that were not observed in any patients included pulmonary embolism, surgical site infection, and urinary tract infection.

At 30 days, discharge destinations included home (12 patients, 50%), home with hospice (2 patients, 8%), and rehab (1 patient, 4%) (Table 3). Six patients (25%) remained hospitalized after 30 days. Thirty-day mortality occurred in 5 patients (20.8%), including the 2 patients discharged with home hospice. Mortality prior to discharge occurred in one additional patient at 44 days postop.

Discussion

This retrospective study examines the outcomes of patients with heart failure or CS undergoing LC for presumed AAC. The average patient age was 53. Nineteen patients (79%) were male. Twenty patients (83%) were white. The most common chronic comorbidities included anticoagulation (10 patients, 88%), history of smoking (14 patients, 58%), hypertension (12 patients, 50%), and history of myocardial infarction (10 patients, 42%). Four patients (17%) had prior treatment with PCT. At the time of surgery, many patients had additional interventions for their underlying cardiac disease, including ventricular assist devices (14 patients, 58%), vasopressors (11 patients, 46%), CRRT (10 patients, 42%), inotropes (10 patients, 42%), and mechanical ventilation (9 patients, 38%). Average operative time was 83 minutes. The most common complications were reintervention within 30 days (17 patients, 71%), bleeding requiring transfusion (13 patients, 53%), and reoperation within 30 days (9 patients, 37.5%). Mortality occurred in 5 patients (20.8%). All mortalities were attributed (directly or indirectly) to complications of underlying cardiac disease.

Included patients were mostly male, which is consistent with reported literature that CS and AAC both disproportionately affect men.^7–9 The most common comorbidities were anticoagulation, hypertension, prior myocardial infarction, and smoking history. There are multiple indications for anticoagulation in patients with heart failure, including stents, atrial fibrillation, left ventricular thrombus, deep venous thrombosis, or ventricular assist device. ¹⁰ The comorbidities represented are consistent with those commonly seen in patients with heart failure. ¹¹ Advancements in medical management and technology for heart failure, including ventricular assist devices, enable patients with heart failure to live longer. However, this presents new issues, such as the development of AAC from a low-flow state. ¹² Once acute cholecystitis (calculous and acalculous alike) is diagnosed, the preferred treatment is cholecystectomy within 3 days; this is associated with fewer postoperative complications than patients who undergo delayed surgery. ⁹ When patients with AAC are of unacceptable surgical risk, they can be treated with PCT. In a retrospective study of patients 65 years or older with Grade III calculous or acalculous cholecystitis (defined by the Tokyo guidelines as the presence of organ dysfunction in cardiovascular, neurological, respiratory, renal, hepatic, or hematological systems), heart failure was among the factors with the greatest odds ratio of being treated with PCT rather than cholecystectomy,^13,14 Despite being a frequently used treatment option in this patient population, PCT is not a perfect substitute for surgery. A 2018 study randomized high-risk surgical patients with AAC to treatment with either PCT or LC; after an interim analysis, the trial was concluded due to higher rates of major postprocedural complications (65%) with PCT as compared with LC (12%). ¹⁵ The complications defined in this analysis included infectious and cardiopulmonary complications within 1 month, need for reintervention within 1 year, or recurrent biliary disease within 1 year. PCT has been shown to be an effective treatment in some cases of AAC. However, a retrospective study of patients with AAC showed that those with sepsis and shock had no difference in survival with PCT versus no intervention, but improved survival with cholecystectomy. ¹⁶ The authors concluded that PCT might not provide benefit to the sickest patients and that cholecystectomy should be considered. In patients 65 or older with Grade III cholecystitis of all etiologies, treatment with PCT has been associated with greater odds of mortality, composite morbidity (defined as neurological, cardiovascular, thromboembolic, respiratory, and infectious complications), and readmission compared with LC and open cholecystectomy. ¹³ Early LC avoids delayed complications and additional expenses of recurrent cholecystitis or interval cholecystectomy. This is important in this patient population, as congestive heart failure has been identified as a risk factor for relapse in AAC following initial treatment with PCT, likely attributable to ongoing hypoperfusion and ischemia. ¹⁷ The presentation of AAC can be atypical and progress rapidly. ¹⁸ Diagnosis in critically ill patients is complicated by limitations in patient participation. Gangrenous cholecystitis and gallbladder perforation are both indications for urgent LC. ¹⁹ The incidence of gangrenous cholecystitis or gallbladder perforation in AAC has been estimated between 37% and 81%, which is more common than in acute calculous cholecystitis.^8,18,20,21 In a study of 22 patients undergoing cholecystectomy for AAC, gangrene was observed significantly more frequently in AAC (59%) than in calculous cholecystitis (27%) (p < .05), highlighting the role of ischemia in its pathogenesis. ⁸ It follows that this patient population may differ from those achieving definitive therapy with PCT.

The risk of conversion to open cholecystectomy (and the associated increase in morbidity and mortality) has been stated as a reason to avoid LC in critically ill patients. ²² However, all patients in this study tolerated the procedure (0% conversion to open). Nine patients (37.5%) underwent a total of 21 reoperations, most of which were related to cardiac devices (Table 4). Out of 21 total reoperations, 1 (4%) was related to LC. Patient 16 underwent diagnostic laparoscopy due to concern for bowel ischemia in the setting of worsening lactic acidosis on postop days 4 and 5 from LC and total artificial heart placement, respectively. Bowel ischemia was ruled out and the patient was taken for chest exploration, where a 400 cc clot was evacuated. The patient’s acidosis was attributed to impaired venous return to the right atrium.

Thirty-day mortality occurred in 5 (20.8%) patients. Causes of death included CS (POD 6, 14, 27), complication of LVAD and multisystem organ failure (POD 30), and multisystem organ failure (POD 22). The nature of these mortalities reflects the compromised cardiac state of these patients and the risks inherent to their necessary cardiac interventions. Ventricular assist devices are associated with complications, including thromboembolism, hemorrhage, infection, right ventricular failure, and multisystem organ failure.^5,10,23 The in-hospital mortality rate for CS has been reported between 27% and 51%, ²⁴ which is higher than the mortality rate for patients undergoing LC in this study (20.8%). Reported in-hospital mortality following PCT for AAC is 8.3–19%,^25,26 which is near the mortality rate of 20.8% in this study. As per the 2018 Tokyo Guidelines, LC can be performed for patients with Grade III cholecystitis (etiology unspecified) if the following are true: it is decided the patient can withstand surgery, the surgeon has extensive experience, and intensive care management is available. ¹⁴ The guidelines further consider cardiovascular dysfunction and renal dysfunction as “favorable organ system failure,” stating that these dysfunctions may improve with the treatment of acute cholecystitis. ¹⁴ There are interesting implications here. First, if mortality following LC in CS is truly less than the mortality rate of CS, it is possible that treating cholecystitis eliminates one cause of death in this patient population. Second, if mortality following PCT and LC are truly similar, the increased potential for benefit with cholecystectomy would favor surgical management.

Limitations include small sample size with limited demographic variety. Because this is an observational study of critically ill patients with various interventions, attributing a change in condition following LC to the operation is not possible. Additionally, while the pathophysiology of AAC is of interest, it is possible that elements of calculous, acalculous, acute, and chronic cholecystitis contributed to the included cases. This challenge extends to literature review, where studies often include patients with shock, heart failure, or AAC, but not in combination. Studies do not always differentiate between calculus and acalculous etiologies; some even exclude acalculous cholecystitis. Despite this, the observations remain clinically relevant given the challenges of diagnosing AAC in the critically ill.

Future efforts should generate larger datasets to determine the incidence of and rates of complications in patients with heart failure or CS who develop AAC. Pathological findings should be used to account for circumstances in which an ischemic etiology occurs in the setting of incidental cholelithiasis. Randomized controlled trials comparing outcomes of PCT versus LC in CS will be important for developing management guidelines. Future studies should include long-term follow-up of patients receiving PCT to identify delayed morbidity, mortality, and expense. Finally, using a term specific to the pathogenesis rather than the presence or absence of stones, such as ischemic cholecystitis, can be helpful in future discussions.

It follows from our current pathophysiologic understanding of AAC that LC would be beneficial to patients who can tolerate it. Patients with decompensated heart failure or CS have reason to be at increased risk for developing AAC. This can be difficult to diagnose and progress quickly if treatment is delayed. PCT may be inferior to cholecystectomy in this patient population. LC for AAC in patients with CS remains high risk but should be considered when in a facility with an experienced surgeon, cardiothoracic anesthesia, and an ICU. In patients at risk for sepsis, surgery is an option for definitive treatment after thoughtful discussions with the patient and family.