Percutaneous cholecystostomy for acute cholecystitis in high‐risk elderly patients

Authors: Wei‐Chen Lin, Chen‐Wang Chang, Cheng‐Hsin Chu

Abstract

Emergency cholecystectomy for acute cholecystitis (AC) is associated with high morbidity and mortality rates in elderly patients with significant comorbidities. The aim of this study was to evaluate percutaneous cholecystostomy for AC in elderly patients with various coexisting diseases. We retrospectively reviewed the records of 4311 patients with AC treated in Mackay Memorial Hospital between the years 2000 and 2015. The clinical course of AC was compared between nonelderly (age ≤70 years) and elderly patients (age >70 years). In total, 67 elderly patients and 32 nonelderly patients received percutaneous cholecystostomy. The rate of percutaneous cholecystostomy increased between the years 2011 and 2015 (from 2.5% to 12.2%) and this procedure was more common in the elderly group (p = 0.009). In addition, the comorbidities of ischemic heart disease and chronic kidney disease were higher in elderly than in nonelderly patients (p = 0.014 and p = 0.015, respectively). The American Society of Anesthesiologists' classification was higher in the elderly patients (p = 0.001). The overall survival‐free rate of recurrent cholecystitis in patients who did not receive cholecystectomy was not significantly different in the two groups. When compared with emergent cholecystectomy in high‐risk elderly patients, percutaneous cholecystostomy resulted in reduced hospital stay and morbidity (p = 0.002 and p = 0.013, respectively). Our results demonstrate that percutaneous cholecystostomy has become a common and early treatment for AC in high‐risk elderly patients with ischemic heart disease or chronic kidney disease. Identifying such patients will possibly improve clinical outcomes, reduce hospital stay and morbidity, and facilitate delayed laparoscopic cholecystectomy.

Introduction

Acute cholecystitis (AC) is one of the most common reasons for surgery in older patients. A previous study reported that 50% of women and 16% of men in their 70s have gallbladder disease [1]. In older patients, the disease is more frequent and severe, necessitating emergency surgery or other interventions [2]. Early cholecystectomy is the standard treatment for AC. Within 72 hours after the onset of symptoms and before the development of fibrosis, laparoscopic cholecystectomy (LC) may be a safe procedure because the anatomy is usually clear [3]. A Cochrane review of randomized controlled trials (1998–2003) evaluating early and late LC during AC showed that LC may shorten total hospital stay with no difference in rates of conversion to open cholecystectomy, morbidity, or mortality [4].

However, the morbidity (up to 41%) and mortality (up to 4.5%) associated with acute LC are high [[5], [6]], mainly because of anesthetic complications and the presence of serious underlying comorbidities. Therefore, high‐risk patients with AC are generally treated via conservative methods including intravenous fluid, antibiotic therapy, and percutaneous cholecystostomy (PC) while surgical intervention is delayed for 6–8 weeks [7].

Delayed presentation and significant comorbid illness in the elderly are associated with increased morbidity in elective and emergent surgery [8]. Because emergency surgery is poorly tolerated by older patients, the medical team needs to assess the timing of surgery thoroughly. If possible, surgery should be rescheduled as elective. However, age >70 years is a predictor of failure of conservative treatment [8]. Therefore, PC is a low‐risk management option when the patient is too ill to undergo surgery [9].

The present study was undertaken to compare the clinical factors for PC in elderly and nonelderly patients who did not respond to conservative treatment. The second objective was to compare the clinical outcomes of high‐risk elderly patients who received PC with those who received emergent cholecystectomy. Identifying such patients will improve the overall outcome and facilitate delayed LC.

Methods

Using International Classification of Diseases, 9th revision, codes, we identified 4311 patients with AC and 1247 patients above the age of 70 years admitted to Mackay Memorial Hospital from January 1, 2000 to September 30, 2015. Diagnosis was made based on medical history, physical examination findings, laboratory studies, and imaging findings (abdominal ultrasonography, cholescintigraphy hepatobiliary iminodiacetic acid scan, computed tomography, and magnetic resonance cholangiopancreatography). Emergency LC was recommended for patients with symptoms at 0–72 hours before admission. If the patient had a preoperative diagnosis of gallbladder perforation or gangrene change, open cholecystectomy wound be performed initially.

The decision to perform PC was made by a senior surgeon based on a high risk‐benefit ratio for AC. The indications for PC in patients with calculous cholecystitis were classified into four (1) high risk for surgery due to associated severe comorbidity; (2) severe cholecystitis (not responding to conservative management); (3) patients who refuse cholecystectomy; and (4) suspected empyema of the gallbladder. Patients with gallstone pancreatitis, choledocholithiasis, hepatobiliary or intestinal malignancy, or autoimmune biliary disease were excluded.

PC was performed by a specialized interventional radiology team under ultrasound or computed tomography guidance in the interventional radiology unit. The distended gall bladder was visualized and local anesthetic infiltrated into the skin and subcutaneous tissue. The gall bladder was then cannulated with a plastic pig‐tail catheter using the Seldinger technique. A small volume of contrast agent was injected and fluoroscopy was used to confirm the position of the catheter and determine the patency of the biliary ductal system. Initial aspirated bile was cultured and antibiotic sensitivity of isolated organisms was established. The catheter was then anchored to the skin.

The clinical course of AC was compared between nonelderly onset (age ≤70 years) and elderly onset patients (age >70 years). We compared clinical characteristics such as development of PC symptoms, total length of PC, American Society of Anesthesiologists (ASA) class, and severity on admission. We also compared clinical outcomes such as morbidity, mortality, time to surgery, conversion rate to open surgery, total length of PC, recurrent cholecystitis rate, and postoperative length of stay. Morbidity was classified into PC complications (i.e., bleeding, catheter blockage and dislodgement, and failure to resolve the AC) and surgical complications (i.e., biliary tree injury, bleeding, abdominal abscess, and wound infection).

AC severity was assessed based on the Tokyo Guidelines, 2013 [10]. Grade I (mild AC) is defined as AC in a patient with no organ dysfunction and limited disease in the gallbladder, making cholecystectomy a low‐risk procedure. Grade II (moderate AC) is defined as AC in a patient with no organ dysfunction but extensive gallbladder disease, which makes safely performing cholecystectomy difficult. Grade III (severe AC) is defined as AC in a patient with organ dysfunction.

Statistical analysis

Descriptive statistics for continuous variables were calculated and reported as mean ± standard deviation. The categorical variables were described using frequency distributions and reported as n (%). The p values were based on a t test for continuous variables, and the chi‐square or Fisher exact test was used for categorical variables. Statistical analysis was performed using SPSS (12.0, Windows version; SPSS Inc., Chicago, IL, USA). Tests were two‐tailed and p = 0.05 was considered statistically significant.

The event of recurrent cholecystitis in patients who did not receive cholecystectomy was evaluated using survival analysis. The time to PC tube removal was considered to begin and end at the date of recurrent cholecystitis or last known follow‐up. The cumulative probabilities of event‐free survival were estimated using the proportional hazards regression model.

Results

We examined the records of 4311 patients treated in Mackay Memorial Hospital between 2000 and 2015 for AC (Figure 1). LC rate increased from 10.7% in 2000 to 60.3% in 2015. However, the rate did not noticeably increase after 2011. By contrast, the rate of open cholecystectomy decreased from 34.2% in 2000 to 7.1% in 2015. Conservative treatment rate decreased from 55.1% in 2000 to 17.6% in 2015. PC rate markedly increased from 0.5% in 2005 to 12.2% in 2015.

When divided into nonelderly and elderly onset groups (Figure 2), the rate of open cholecystectomy decreased and LC increased during 2000–2015, but this was not statistically significant (p = 0.242 and p = 0.235, respectively). Since the introduction of PC to our hospital in 2004, this procedure has become more common in the elderly group than in the nonelderly (p = 0.009). PC rate in the elderly increased from 7.2% in 2010 to 31.8% in 2015.

Ninety‐three patients (52 men and 41 women) received PC for AC during the 15‐year period (Table 1). Sixty‐six percent (n = 61) of the patients were older than 70 years (mean age 80.3 years). The mean duration from symptom onset to PC was 2.8 ± 2.0 days (range, 1–11 days) in the nonelderly group and 2.7 ± 2.2 days (range, 1–15 days) in the elderly group (p = 0.904). The mean duration of insertion (removal after surgery or symptom relief) was longer in the elderly group 17.6 ± 14.7 days (range, 1–49 days) than in the nonelderly group 15.1 ± 12.8 days (range, 3–46 days), but this difference was not statistically significant (p = 0.071).

The occurrence of ischemic heart disease and chronic kidney disease was higher among elderly patients than among nonelderly patients (p = 0.014 and p = 0.015, respectively). However, there were no statistical differences between the nonelderly and elderly onset groups with respect to the following diabetes mellitus, hypertension, liver cirrhosis, chronic obstructive pulmonary disease, and cerebrovascular accident (p = 0.09, p = 0.07, p = 0.502, p = 0.086, and p = 0.317, respectively). Ninety‐five percent of the elderly patient group were high risk (ASA III/IV) as opposed to 68% of the nonelderly group (p = 0.001). With respect to AC severity, 21 (34.4%) of 61 elderly patients belonged to Grade III, which was greater than 9 (25.8%) out of 32 nonelderly patients (p = 0.531).

Regarding clinical outcomes, bacteremia and mortality rates were higher among the elderly patients than among nonelderly patients (14.8% vs 9.7% p = 0.462; and 13.1% vs 6.5% p = 0.310), respectively (Table 2). All‐cause mortality was associated with comorbidity, but not caused by PC. PC morbidity rate was 26.2% in the elderly patients [catheter blockage and dislodgement (n = 11), bleeding (n = 3), wound infection (n = 1), and bile leaks (n = 1)], and 15.5% in the nonelderly patients [catheter blockage and dislodgement (n = 4), and bleeding (n = 1)]. There was no statistically significant difference in morbidity (p = 0.245).

In high‐risk patients that received PC for AC, only 34 (55.7%) of 61 elderly patients underwent surgical cholecystectomy, compared with 23 (71.1%) of 32 nonelderly patients (p = 0.170). Conversion from laparoscopic to open surgery (usually because of the difficulty with dissection at the gall bladder and Calot's triangle) was more frequent in elderly patients than in nonelderly patients (21.4% vs. 8.7%, p = 0.340). There was no statistically significant difference in gall bladder gangrene change or perforation during surgery between these two groups (p = 0.325 and p = 0.799, respectively). The postoperative hospital stay in the nonelderly group was longer than that in the elderly group (6.2 ± 4.6 days vs. 5.4 ± 3.1 days, p = 0.325). There were no cases of bile duct injury or death after surgery.

Eight (29.6%) out of 27 elderly patients and two (22.2%) out of nine nonelderly patients experienced recurrent cholecystitis after PC removal. One elderly patient received repeated PC for recurrent cholecystitis while others recovered after conservative treatment. The overall survival‐free rate of recurrent cholecystitis in patients who did not receive cholecystectomy was not significantly different between the two groups (p = 0.511, log‐rank; Figure 3).

When comparing high‐risk elderly patients with AC who received emergent cholecystectomy (including LC and open cholecystectomy) or PC, 275 patients received emergent cholecystectomy and 61 patients received PC (Table 3). These two groups showed homogeneity with respect to age (p = 0.274), ASA classification (p = 0.218), and each comorbidity. The hospital stay was significantly longer in elderly patients who received emergent cholecystectomy than PC (21.9 ± 13.1 vs. 17.7 ± 12.9; p = 0.002). Morbidity was more common in the emergent cholecystectomy group than in the PC group (44% vs. 26.2%, p = 0.013). Mortality was slightly higher in the PC group but the difference was not statistically significant (p = 0.098).

Discussion

The efficacy of delayed cholecystectomy after PC management is unclear. A large population study of 155,322 cases of severe gallstone disease in Taiwan showed that the emergent cholecystectomy rate was higher than elective cholecystectomy rate in the elderly group (age >80 years) from 1997 to 2005, while the elective cholecystectomy rate was higher in the nonelderly group [11]. However, elderly patients are often poor candidates for surgery because of concomitant medical problems. PC provides an alternative treatment for AC in elderly patients with high surgical risk and has lower morbidity and mortality than emergency cholecystectomy [7].

In the present study, we evaluated the indication and outcome of PC for AC in elderly patients. The PC rate in the elderly group increased significantly between 2010 and 2015. The ASA classification was also higher in elderly patients. Ischemic heart disease and chronic kidney disease were common in the elderly group. Although the two groups had no difference in outcomes, we believe that a lower proportion of elderly patients underwent cholecystectomy, which prevented recurrent cholecystitis, when compared with the nonelderly patients.

Complicated cholecystitis occurs more commonly in elderly patients who have comorbidities [2]. Since the first percutaneous drainage procedure for gallbladder empyema was performed by Radder in 1980, it has become a useful alternative treatment to surgery, allowing high‐risk patients with comorbidities to recover from acute illness before proceeding to cholecystectomy [[7], [12], [13]].

For patients with AC who did not respond to nonoperative treatment, the results of treatment with emergency LC or PC followed by delayed LC are similar in low‐risk (ASA I) patients, but the conversion rate to open surgery is significantly higher in high‐risk (ASA II or III) patients who are treated with emergency LC [14]. Furthermore, advanced age (p < 0.001, odds ratio 2.33) and patients with higher ASA classes (p < 0.001, odds ratio 2.31) had a statistically significant increase in morbidity [15]. Open cholecystectomy was associated with a significantly higher incidence of morbidity than LC (47% vs 7%, p < 0.001) [15]. Our study showed that elderly patients who received PC had a higher risk (ASA III and IV) than nonelderly patients. Therefore, PC followed by delayed LC should be considered in elderly patients and those in higher ASA classes to reduce morbidity.

When considering the indication for PC, one study showed that the common comorbidities were hypertension (76%), ischemic heart disease (54%), diabetes (37%), and renal failure (22%) [16]. We obtained similar results in our study. Ischemic heart disease or chronic kidney disease was a common comorbidity in the elderly patients.

The cause of conversion was a repetitively progressive inflammation, which was accompanied by a distended and edematous walled gallbladder. Gall bladder distension is a predictor for failure of conservative treatment [8]. One previous study suggested that initial conservative treatment followed by delayed operation could not diminish the morbidity and conversion rate for AC [17]. However, in this study, delayed cholecystectomy was performed without any bridging treatment such as PC [17]. Another prospective study showed that patients underwent PC and delayed LC with a low conversion rate (8%) [18]. In our study, subsequent to PC, 34 of the 61 elderly patients underwent a cholecystectomy. Twenty‐eight patients had an LC, and six patients had LC converted to an open procedure. The conversion rate was higher in the elderly group than the nonelderly group but without statistical significance (21.4% vs. 8.7% p = 0.340).

In a 10‐year retrospective analysis, Abi‐Haidar et al [19] reported that PC was associated with longer hospital stay and higher morbidity compared with cholecystectomy in patients with AC. However, older age and higher comorbidities were found in the PC group. Another study showed that PC had a shorter operation time and shorter postoperative hospital stay than emergency cholecystectomy [20]. In our study, the mortality rate was higher in the elderly group but was not statistically significant. The postoperative hospital stay was shorter in the elderly group but was also not statistically significant which may be attributed to a selection bias as it was to be expected that elderly patients treated with cholecystectomy after PC were in a better clinical condition than those not treated with LC. Our study showed that PC could reduce hospital stay and morbidity in high‐risk elderly patients more than emergent cholecystectomy, while mortality was slightly higher in the PC group but this was not statistically significant. In all, our findings suggest that PC should be performed in elderly patients with severe comorbidities rather than improving surgical outcomes of LC for severe AC.

There is some controversy concerning whether cholecystectomy should be performed after PC in elderly high‐risk patients. Our study found there was no statistically significant difference in the overall survival‐free rate of recurrent cholecystitis between the nonelderly and elderly group. One study showed a recurrence rate of 4.1% for cholecystitis and a one‐year survival rate of 82.2% [21]. They proposed that PC is a definitive treatment for AC and cholecystectomy is not necessary due to limited survival and the low recurrence rate of cholecystitis in high‐risk elderly patients with AC [21]. However, another study showed that cholecystectomy for AC in elderly patients should be performed during initial hospitalization to prevent recurrent cholecystitis, multiple readmissions, and increased costs [22].

Our study has certain important limitations. First, this was a retrospective study that used administrative data. Therefore, analysis was limited to the information available about the exact time from onset of symptoms to presentation. It is possible that some patients who received PC had symptoms for more than 48–72 hours. Furthermore, the prolonged hospitalization might be associated with higher emergent surgery rates in the decade 2000–2010 and operations were performed by less experienced surgeons in the night. In addition, this analysis was also limited to short‐term mortality and did not consider long‐term mortality or complications.

In summary, PC has become a common early treatment for AC in the elderly and has resulted in improved clinical outcomes such as reduced hospital stays and morbidity. PC may be followed by elective cholecystectomy or conservative management in high‐risk elderly patients with systemic disease such as ischemic heart disease or chronic kidney disease.

Acknowledgments

The authors would like to thank the team of radiologist at Mackay Memorial Hospital.