Intra-Operative Abdominal Drain Placement for Gallbladder Cancer Surgery and Risk of Infectious Complications

Abstract

Background:

Routine intra-operative abdominal drain placement (IADP) is not beneficial for uncomplicated cholecystectomies though outcomes in gallbladder cancer surgery is unclear. This retrospective study hypothesized that patients with IADP (+IADP) for gallbladder cancer surgery have a higher risk of post-operative infectious complications (PIC) compared with patients without IADP (−IADP).

Patients and Methods:

The 2014–2017 American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database was queried for +IADP and −IADP patients who had gallbladder cancer surgery. Post-operative infectious complications were defined as septic shock, organ/space infection (OSI), or percutaneous drainage. Multivariable analyses were performed to analyze the associated risk of PIC.

Results:

Of 385 patients, 237 (61.6%) were +IADP. The +IADP patients had higher rates of post-operative bile leak, OSI, re-admission, and increased length of stay (p < 0.05). The +IADP patients were not associated with increased risk of PIC (p > 0.05). Bile leak (odds ratio [OR], 10.61; p < 0.001), peri-operative blood transfusion (OR, 3.77; p = 0.003), biliary reconstruction (OR, 2.88; p = 0.018), and pre-operative biliary stent placement (OR, 3.02; p = 0.018) were the strongest associated risk factors of PIC. Patients with drains in place at or longer than 30 days post-operatively had an increased associated risk compared with patients who did not (OR, 6.88; 95% confidence interval [CI], 2.16–21.86; p < 0.001).

Conclusions:

More than 60% of gallbladder cancer surgeries included IADP and was not associated with an increased risk of PIC. Intra-operative abdominal drain placement was not associated with an increased risk of PIC, unless drains were left in place for 30 days or longer. Increased risk of PIC was associated with bile leak, peri-operative blood transfusion, pre-operative biliary stent placement, and biliary reconstruction.

Routine intra-operative abdominal drain placement (IADP) at the subhepatic space is not beneficial for uncomplicated elective cholecystectomy. Previously thought to be a benign strategy in draining excess intra-abdominal fluid or as prophylactic precautions against post-operative bile leaks, studies have shown that patients with IADP (+IADP) were associated with increased post-operative complications such as wound infection, bile leak, and length of stay (LOS) [1,2]. A Cochrane Review of uncomplicated laparoscopic and open cholecystectomy have found that IADP was associated with a 15-fold increased risks for abdominal collection and wound infection, respectively [3,4]. Currently, it is uncommon practice to use IADP in an uncomplicated laparoscopic or open cholecystectomy unless there is concern for biliary duct injury [5].

In contrast to the literature for uncomplicated cholecystectomy, routine use of IADP for hepatobiliary cancer surgery, especially in gallbladder cancer, remains unclear. Gallbladder cancer is relatively uncommon in the United States with an incidence of approximately one to two cases per 100,000 population [6]. To obtain negative margins, hepatic resections involving liver segments IVB and V are required, but may be extended beyond if necessary to include lymphadenectomy with or without bile duct resection [7]. Because of this, the inherent associated risk for bile leak may be relatively high compared with hepatectomy or elective uncomplicated cholecystectomy. Capussotti et al. [8] reported that resection of liver segment IV had a three-fold risk for bile leak in itself, thus making IADP a reasonable option for anticipating complications. The benefits of IADP have been described in other hepatobiliary cancer surgeries. Patients who are jaundiced pre-operatively may benefit from IADP because of associated higher incidences of bile leak and subphrenic collections [9]. Aldameh et al. [10] found that IADP in elective hepatic resections was not associated with increased LOS, mortality, or intra-abdominal post-operative fluid collections. Conversely, other studies have aligned with the uncomplicated cholecystectomy data. It has been found that routine IADP in hepatobiliary surgery may increase risks of wound infection, bile leak, and may not be as useful as it has been reported to have similar incidences of post-operative percutaneous drainage for fluid collection between patients with (+IADP) and without IADP (−IADP) [11,12]. There is a paucity of data in the use of IADP in gallbladder cancer surgery specifically and its clinical outcomes have yet to be described.

This present study sought to describe the risk of post-operative infectious complications (PIC) in +IADP and −IADP patients undergoing gallbladder cancer surgery. It was hypothesized that +IADP patients undergoing gallbladder cancer surgery were associated with a higher risk of PIC compared with −IADP patients.

Patients and Methods

This was a retrospective analysis using the 2014–2017 American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) Procedure-Targeted Hepatectomy database. The ACS-NSQIP database contains prospectively gathered data that includes pre-operative, intra-operative, and 30-day post-operative data for patients undergoing a hepatectomy at participating centers across the United States [13]. Adults (≥18 years old) who underwent a hepatectomy with a post-operative pathology report indicating the patient to have gallbladder cancer were queried and included in the study. Status of IADP was recorded and defined by the ACS-NSQIP data dictionary as the presence of a closed suction drain placed at the surgical bed at the conclusion of surgery [14]. The ACS-NSQIP and all participating hospitals are the source of the data used herein; they have not verified and nor are they responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Patients who did not undergo a hepatectomy and patients without a post-operative diagnosis of gallbladder cancer were excluded from this analysis.

The primary outcome was the risk of PIC, which was defined as septic shock, deep organ/space infection, and requirement of percutaneous drainage. Secondary outcomes included post-operative bile leak, organ/space infection (OSI), and post-operative intervention. The criteria for post-operative bile leak were different depending on the status of IADP. In +IADP patients, definition for bile leak included continued bilious drainage on or after post-operative day three, spontaneous wound drainage, bile output after post-operative percutaneous drainage with concentrations three-fold those of serum, and re-operation performed for a biloma. For −IADP patients, bile leak diagnosis was defined through a clinical diagnosis by the attending surgeon, spontaneous wound drainage, percutaneous drainage performed with bilious output, or reoperation with findings of a biloma. Post-operative intervention did not include re-operation. Instead, post-operative intervention included post-operative biliary stent placement via endoscopic retrograde cholangiography, percutaneous transhepatic biliary stent for bile obstruction/leak, and percutaneous drainage for OSI and hematoma.

Demographic data such as age, gender, race/ethnicity, medical comorbidities, and functional health prior to surgery were collected. Pre-operative biliary stenting because of obstructive jaundice and intra-operative liver texture, which included NSQIP-defined characteristics such as cirrhotic, congested, and fatty were recorded. Tumor-related information including tumor, node, and metastasis (TNM) staging based on pathology report were collected. Treatment-related information including use of neoadjuvant chemotherapy, operative approach, intra-operative ablation, Pringle maneuver during resection, biliary reconstruction, operative time, and number of concurrent partial resections were noted.

Data regarding concurrent partial hepatic resections note the number of additional hepatic resections in the liver required at the time of the primary procedure. There was no specific information on the type of hepatic resections (e.g., extended right hepatectomy, trisgementectomy). Information regarding concurrent ablation to destroy tumor cells in the liver during the primary procedure included radiofrequency, cryotherapy, microwave, and alcohol ablation. Operative approaches included laparoscopic/robotic and open procedures. Complications included bile leak, OSI, surgical site infection (SSI), need for transfusion, septic shock, pneumonia, deep vein thrombosis (DVT), and urinary tract infection. Need for post-operative intervention, re-admission, and mortality were also evaluated. Organ/space infection was classified as infection, other than the incision, that involves any part of the anatomy that was opened or manipulated during an operation within 30 days of the operation.

All variables were analyzed with descriptive statistics and coded as present or absent. Continuous data were reported as medians with interquartile range or means with standard deviations. Student t-test and Mann-Whitney U test were used to compare continuous variables. Categorical data were reported as percentages and compared using χ² test for bivariate analysis.

A multivariable logistic regression model was conducted to determine the associated risk of PIC in patients undergoing gallbladder cancer surgery that includes cholecystectomy with or without partial hepatectomy and in patients with drains still present at or longer than post-operative day 30 (at or beyond post-operative day 30). Post-operative complication was defined as septic shock, OSI, or percutaneous OSI drainage. In addition to controlling for age, female gender, smoking status, diabetes mellitus, steroid use, weight loss, pre-operative biliary stent placement, biliary reconstruction, neoadjuvant therapy, and peri-operative blood transfusion, two separate multivariable analyses were conducted controlling with and without post-operative bile leak.

Another multivariable logistic regression analysis was conducted to determine the associated risk of prolonged LOS. Prolonged LOS was defined as hospital day six or more. This multivariable logistic regression analysis controlled for bile leak, need for blood transfusion, post-operative intervention, DVT, septic shock, and OSI. Covariates for each of these models were chosen after review of the literature and discussion among co-authors [8,12,15,16]. The associated risks were reported with an odds ratio (OR) and 95% confidence intervals (CI). All p values were two-sided, with a statistical significance level of <0.050. All analyses were performed with SPSS Statistics, version 24 (IBM Corp, Armonk, NY).

Results

Demographics and patient characteristics

Of 385 patients with gallbladder cancer, 148 (38.4%) and 237 (61.6%) patients were −IADP and +IADP, respectively. Comparing −IADP patients with +IADP patients, there were no differences between cohorts in age (mean, 65.1 years vs. 67.3 years; p = 0.364), female (66.9% vs. 67.1%; p = 0.968), and race/ethnicity (p = 0.908). There were no differences in comorbidities including metastatic cancer (12.2% vs. 15.6%; p = 0.347), long-term steroid use (2.0% vs. 2.5%; p = 0.750), smoking (10.8% vs. 15.2%; p = 0.220), and diabetes mellitus (20.3% vs. 20.7%; p = 0.924) between −IADP patients and +IADP patients. There was no difference in functional health prior to surgery (p = 0.192) between the cohorts. Patients with +IADP had an increased incidence for pre-operative biliary stent placement for obstructive jaundice (17.7% vs. 4.1%; p < 0.001; Table 1). Tumor (p = 0.230) and metastatic (p = 0.133) gallbladder cancer staging were similar. A higher incidence of gallbladder cancer with higher nodal staging was found in +IADP patients (p = 0.003; Table 2).

Table 1.

Demographics of Patients with Gallbladder Cancer Undergoing Surgery with and without Drains

Demographics	−IADP (n = 148)	+IADP (n = 237)	p
Age, y, mean (±SD)	65.1 (12.0)	67.3 (10.8)	0.364
Female, n (%)	99 (66.9%)	159 (67.1%)	0.968
Race/ethnicity, n (%)			0.908
Asian	8 (5.4%)	18 (7.6%)
Black or African American	17 (11.5%)	25 (10.5%)
White	93 (62.8%)	144 (60.8%)
Hispanic	8 (6.5%)	19 (9.5%)
Comorbidities, n (%)
Metastatic cancer	18 (12.2%)	37 (15.6%)	0.347
COPD	6 (4.1%)	13 (5.5%)	0.528
Hypertension	84 (56.8%)	142 (59.9%)	0.540
Smoking	16 (10.8%)	36 (15.2%)	0.220
Diabetes mellitus	30 (20.3%)	49 (20.7%)	0.924
Long-term steroid use	3 (2.0%)	6 (2.5%)	0.750
Pre-operative biliary stent placed, n (%)	6 (4.1%)	42 (17.7%)	<0.001
Endoscopic	5 (3.4%)	38 (16.0%)
Percutaneous	1 (0.7%)	4 (1.7%)
Observed liver texture			0.398
Normal	43 (11.2%)	66 (17.1%)
Cirrhotic	4 (1.0%)	9 (2.3%)
Congested	1 (0.3%)	5 (1.3%)
Fatty	20 (5.2%)	20 (5.2%)
Functional health prior to surgery, n (%)			0.192
Independent	147 (99.3%)	232 (97.9%)
Partially dependent	0 (0.0%)	4 (1.7%)
Totally dependent	1 (0.7%)	0 (0.0%)

−IADP = patients without intra-operative abdominal drain placement; +IADP = patients with abdominal intra-operative drain placement; SD = standard deviation; COPD = chronic obstructive pulmonary disease.

Table 2.

TNM Staging of Patients with Gallbladder Cancer Undergoing Surgery with and without Drains

Demographics	−IADP (n = 148)	+IADP (n = 237)	p
Tumor stage			0.230
T0	6 (1.6%)	5 (1.3%)
T1	12 (3.1%)	12 (3.1%)
T2	52 (13.5%)	75 (19.5%)
T3	51 (13.2%)	99 (25.7%)
T4	3 (0.8%)	10 (2.6%)
Node stage			0.003
N0	4 (1.0%)	11 (2.9%)
N1	42 (10.9%)	91 (23.6%)
N2	0 (0%)	14 (3.6%)
Metastases stage			0.133
M0	76 (19.7%)	142 (36.9%)
M1	8 (2.1%)	15 (3.9%)

−IADP = patients without intra-operative abdominal drain placement; +IADP = patients with intra-operative abdominal drain placement; TNM = tumor, node, metastases.

Management and operative characteristics of gallbladder cancer

Patients +IADP were more likely to undergo neoadjuvant therapy compared with −IADP patients (13.5% vs. 5.4%; p = 0.020). There were no differences between −IADP and +IADP patients in operative approach (laparoscopic 17.0% vs. 11.8%; p = 0.446), concurrent partial resection of the liver (60.3% vs. 47.3%; p = 0.349), intra-operative ablation (4.7% vs. 6.3%; p = 0.453), and Pringle maneuver during resection (26.4% vs. 23.6%; p = 0.547). Incidence of biliary reconstruction was higher in +IADP patients compared with −IADP patients (29.1% vs. 6.8%; p < 0.001). Surgeries for +IADP patients were longer compared with −IADP patients (279.2 minutes vs. 205.9 minutes; p = 0.001; Table 3).

Table 3.

Management and Operative Characteristics of Gallbladder Cancer Surgery with and without Drains

Management characteristics	−IADP (n = 148)	+IADP (n = 237)	p
Neoadjuvant therapy,^a n (%)	8 (5.4%)	32 (13.5%)	0.020
Operative approach, n (%)			0.446
Laparoscopic/robotic^b	26 (17.0%)	28 (11.8%)
Open	122 (82.4%)	209 (88.2%)
Concurrent intra-operative ablation, n (%)	7 (4.7%)	15 (6.3%)	0.453
Pringle maneuver during resection, n (%)	39 (26.4%)	56 (23.6%)	0.547
Biliary reconstruction, n (%)	10 (6.8%)	69 (29.1%)	< 0.001
Number of concurrent partial resections, n (%)			0.349
None	78 (52.7%)	94 (39.7%)
1	42 (28.4%)	76 (32.1%)
2	23 (15.5%)	51 (21.5%)
>3	3 (2.0%)	10 (4.2%)
Total operative time [min], mean (± SD)	205.9 (±95.4)	279.2 (±136.1)	0.001

−IADP = patients without intra-operative abdominal drain placement; +IADP = patients with intra-operative abdominal drain placement; SD = standard deviation.

Includes locoregional liver ablation, portal vein embolization, pre-operative systemic chemotherapy.

Laparoscopic/robotic procedures includes those with open assist and unplanned open conversion.

Patient outcomes after gallbladder cancer surgery

It was found that +IADP patients had higher rates of post-operative bile leak (16.5% vs. 2.7%; p = 0.001), OSI (13.9% vs. 5.4%; p = 0.008), need for transfusion (21.9% vs. 4.7%; p < 0.001), septic shock (4.6% vs. 0.7%; p = 0.029), and DVT (1.8% vs. 0%; p = 0.035). There was no difference between the cohorts in terms of SSI (2.0% vs. 5.9%; p = 0.071), pneumonia (2.0% vs. 3.4%; p = 0.440), and renal failure (0% vs. 1.8%; p = 0.115). It was found that +IADP patients had higher rates of post-operative intervention (14.8% vs. 7.4%; p = 0.031), increased mean LOS (8.8 vs. 5.9 days; p < 0.001), and re-admission rate (12.7% vs. 4.7%; p = 0.010). There was no difference in unplanned return to the operating room (0.7% vs. 3.4%; p = 0.088) nor mortality between −IADP patients and +IADP patients (1.4% vs. 3.4%; p = 0.224; Table 4).

Table 4.

Gallbladder Cancer Surgery Outcomes with and without Drains

Outcomes	−IADP (n = 148)	+IADP (n = 237)	p
Complication, n (%)
Bile leak	4 (2.7%)	39 (16.5%)	0.001
OSI	8 (5.4%)	33 (13.9%)	0.008
SSI	3 (2.0%)	14 (5.9%)	0.071
Need for transfusion	7 (4.7%)	52 (21.9%)	<0.001
Septic shock	1 (0.7%)	11 (4.6%)	0.029
Unplanned intubation	1 (0.3%)	9 (2.3%)	0.061
Pneumonia	3 (2.0%)	8 (3.4%)	0.440
UTI	2 (1.4%)	5 (2.1%)	0.588
DVT	0 (0%)	7 (1.8%)	0.035
Renal failure	2 (0.5%)	10 (2.6%)	0.115
Need for post-operative intervention,^a n (%)	11 (7.4%)	35 (14.8%)	0.031
Biliary stent for biliary obstruction/leak	2 (1.4%)	6 (2.5%)
Percutaneous drain for pus from drain/aspirate	4 (2.7%)	14 (5.9%)
Intervention for bilirubin-rich fluid from drain/aspirate	1 (0.7%)	8 (3.4%)
Other intervention	4 (2.7%)	7 (3.0%)
Unplanned return to the operating room/re-operation	1 (0.7%)	8 (3.4%)	0.088
LOS, mean (± SD)	5.9 (±4.5)	8.8 (±8.3)	<0.001
Re-admission, n (%)	7 (4.7%)	30 (12.7%)	0.010
Mortality, n (%)	2 (1.4%)	8 (3.4%)	0.224

−IADP = patients without intra-operative abdominal drain placement; +IADP = patients with intra-operative abdominal drain placement; LOS = length of stay; SD = standard deviation; OSI = organ/space infection; SSI = surgical site infection; UTI = urinary tract infection; DVT = deep vein thrombosis.

Excludes surgery/re-operation.

Associated risk for post-operative infectious complications

On multivariable analysis, +IADP patients were not found to be associated with increased risk of PIC (OR, 1.70; 95% CI, 0.94–1.01; p = 0.253). Pre-operative biliary stent placement was also not associated with an increased risk of PIC (OR, 2.06; 95% CI, 0.88–4.85; p = 0.098). Only peri-operative blood transfusion (OR, 4.43; 95% CI, 1.99–9.90; p < 0.001) and biliary reconstruction (OR, 3.13; 95% CI, 1.40–7.00; p = 0.005) were associated with increased risk of PIC (Table 5).

Table 5.

Multivariable Analysis for Risk of Infectious Complications^a in Patients Undergoing Cholecystectomy for Gallbladder Cancer

Risk factor	OR	95% CI	p
+IADP	1.70	0.94–1.01	0.253
Pre-operative biliary stent placement	2.06	0.88–4.85	0.098
Peri-operative blood transfusion	4.43	1.99–9.90	<0.001
Biliary reconstruction	3.13	1.40–7.00	0.005
Neoadjuvant therapy	0.71	0.23–2.15	0.542

Controlled for age, female gender, smoking status, diabetes mellitus, steroid use, weight loss, pre-operative biliary stent placement, neoadjuvant therapy, biliary reconstruction, and need for transfusion.

OR = odds ratio; CI = confidence interval; +IADP = patients with intra-operative abdominal drain placement.

Defined as septic shock, deep organ space infection, and requirement of percutaneous drainage

Associated risk for post-operative infectious complications: Controlling for bile leak

When controlling for bile leak, the associated risk of IADP remained insignificant (OR, 1.07; 95% CI, 0.40–2.90 CI; p = 0.893). Bile leak (OR, 10.61; 95% CI, 4.40–25.58; p < 0.001), need for peri-operative blood transfusion (CO, 3.77; 95% CI, 1.58–9.01; p = 0.003), biliary reconstruction (OR, 2.88; 95% CI, 1.20–6.92; p = 0.018), and pre-operative biliary stent placement (OR, 3.02; 95% CI, 1.16–7.85; p = 0.018) were the strongest associated risk factors of PIC (Table 6).

Table 6.

Multivariable Analysis for Risk of Infectious Complications^a in Patients Undergoing Cholecystectomy for Gallbladder Cancer

Risk factor	OR	95% CI	p
+IADP	1.07	0.40–2.90	0.893
Bile leak	10.61	4.40–25.58	<0.001
Pre-operative biliary stent placement	3.02	1.16–7.85	0.018
Peri-operative blood transfusion	3.77	1.58–9.01	0.003
Biliary reconstruction	2.88	1.20–6.92	0.018
Neoadjuvant therapy	0.61	0.18–2.07	0.430

OR = odds ratio; CI = confidence interval; +IADP = patients with intra-operative abdominal drain placement.

Defined as septic shock, deep organ/space infection, and requirement of percutaneous drainage.

Subgroup analysis of +IADP patients who had drains at or beyond post-operative day 30 or longer versus those who did not

Within +IADP patients, 15 of 139 patients still had their drains by post-operative day 30. There were higher rates of infectious complications for those who still had drains at or beyond post-operative day 30 compared with those who did not (46.7% vs. 11.3%; p < 0.001). Patients who had drains at or beyond post-operative day 30 had greater rates of re-admission compared with patients who did not (33.3% vs. 6.5%; p = 0.001). Through a univariable logistic regression analysis for risk of infection, patients with drains at or beyond post-operative day 30 had an increased associated risk compared with patients who did not (OR, 6.88; 95% CI, 2.16–21.86; p < 0.001).

Subgroup analysis of patients with prolonged LOS

Patients with post-operative bile leak (OR, 4.18; 95% CI, 1.45–12.05; p = 0.008) and need for blood transfusions (OR, 3.65; 95% CI, 1.71–7.80; p = 0.001) were risk factors for prolonged LOS. Other post-operative complications such as post-operative intervention, DVT, shock, and organ/space infection were not associated with prolonged LOS (Table 7).

Table 7.

Multivariable Analysis for Risk of Increased Length of Stay (>6 Hospital Days) in Patients Undergoing Cholecystectomy for Gallbladder Cancer

Risk factor	OR	95% CI	p
Bile leak	4.18	1.45–12.05	0.008
Need for blood transfusion	3.65	1.71–7.80	0.001
Post-operative intervention	2.17	0.83–5.67	0.113
DVT	1.46	0.13–16.25	0.757
Septic shock	3.66	0.42–32.30	0.242
OSI	2.09	0.19–22.86	0.547

Controlled for bile leak, need for blood transfusion, post-operative intervention, DVT, shock, OSI. OR = odds ratio; CI = confidence interval; +IADP = patients with intra-operative abdominal drain placement; DVT = deep vein thrombosis; OSI = organ/space infection.

Discussion

This study is the first to examine the clinical outcomes of IADP in patients undergoing gallbladder cancer surgery using a large national database. Most cholecystectomies for gallbladder cancer involved IADP. Although IADP was associated with higher rates of infectious complications, bile leak, re-admission, and longer lengths of stay, IADP was not associated with an increased risk of PIC contrary to our hypothesis. Rather, use of pre-operative biliary stent, biliary reconstruction, bile leak, and peri-operative blood transfusion were associated with an increased risk of PIC.

It is well accepted that the routine use of IADP during an uncomplicated, laparoscopic, or open cholecystectomy does not confer any benefit. Contrary to this belief, the present study's large national analysis found that more than 60% of gallbladder cancer surgery cases still involve IADP. This is similar to a prior single-institution studies that had reported the use of IADP to occur in 45%–76% of cases [12]. The routine use of IADP in gallbladder cancer surgery may be the result in part of its more radical operative strategy, especially if TNM staging of gallbladder cancer calls for additional hepatic resections, lymphadenectomy, or removal of bile ducts. In support of this, +IADP patients had nearly a three- and four-fold higher rate of biliary reconstruction and pre-operative biliary stent placement, respectively. Furthermore, +IADP patients had longer operative times by as much as an additional 80 minutes compared with −IADP patients. These variables could serve as a surrogate for perceived complexity of the case by the surgeon. Thus, IADP may be a sign of good surgical judgment for patients at significant risk of bile leak and its subsequent complications.

There was no difference in mortality rate between cohorts despite +IADP patients having higher rates of PIC, and more than a six-fold higher rate of post-operative bile leak compared with −IADP patients. Nevertheless, these complications were associated with longer LOS and a three-fold rate of re-admission in this cohort. In other areas in which IADP is considered, well-powered randomized controlled trials are needed to truly determine the benefit of this practice and identify a subset of high-risk patients who will.

The infectious risk of routine IADP in uncomplicated cholecystectomies has largely driven its use away from regular practice. In a single-center retrospective study of 104 patients with gallbladder cancer, Dixon et al. [12] found that IADP may also confer an infectious risk after gallbladder cancer surgery. They hypothesized that IADP may cause wound infections and bile leaks because of the close proximity to the skin at which the drain exits and from irritative injury to the biliary stump, respectively. However, the present study found no difference in associated risk for PIC in +IADP patients when controlling for other risk factors that may contribute to tissue healing and PIC such as smoking, diabetes mellitus, steroid use, and peri-operative blood transfusion, factors not previously included in the study by Dixon et al. [12]. There is, however, a relation between PIC and those who have their drains left in place for long-term post-operative management. Among +IADP patients who had the drain in place at or beyond post-operative day 30 there was a nearly seven-fold increased associated risk of PIC compared with those who had their drains removed earlier.

The present study comprised a larger population, encompassing more than 400 hospitals, which contributes more to the strong generalizability of the data. Overall, it appears that the associated risk of PIC may rely on the judgment of the surgeon to stratify each case for the risk for a bile leak. The need for pre-operative biliary stent placement, resection of segment IV, hepatectomy of segment I, left hepatectomy extending to segment I, and transection out of the main porta fissure may help surgeons anticipate a bile leak [8,16]. If an IADP was necessary, Kawai et al. [17] found in their prospective study on 104 patients that its removal on post-operative day four was an independent factor in reducing the incidence of intra-abdominal infections compared with removal later in the hospital stay. Hence, until high-quality randomized controlled trials are available, judicious use of an IADP at the surgeon's discretion with early removal appears warranted.

There are other factors that may contribute to PIC other than bile leak. The need for peri-operative blood transfusion has been demonstrated to increase the risk of infectious complications in many cancer-related surgeries [18,19]. In the present study on patients with gallbladder cancer, peri-operative blood transfusion increased the associated risk of PIC and prolonged LOS by nearly four-fold and 3.5-fold, respectively. This is similar to the data on pancreaticoduodenectomy and hepatectomy as peri-operative transfusions increased the risk of post-operative infection including pneumonia, sepsis, and septic shock by 40%–80% [15,20]. This risk may be compounded in gallbladder cancer surgery as the nature of its operative intervention has an inherently high risk for PIC to begin with because of bile leak. Transfusion-related immunomodulation (TRIM) is a phenomena that involves suppression of neutrophil chemotaxis, natural killer cell activity, production of T-cells, donor HLA molecules, or immunoglobulins after blood transfusion [21 –24]. Moreover, blood transfusions have been described to impair clearance of bacterial pathogens, increasing the risk of infectious complications [18,19]. Future efforts to minimize blood loss through hemorrhage control adjuncts, development of cancer-related surgery transfusion guidelines, and advancement of minimally invasive surgeries may ameliorate this issue [25 –27].

There were several limitations to this study including those inherent to a database-driven retrospective analysis. The present study is limited by the number of participating ACS-NSQIP centers, reporter bias, misclassification, and missing data. The database does not have the granularity to provide specific nuances of a procedure. For example, there was no information on the exact type of hepatic resection (e.g., extended right hepatectomy, trisegementectomy). Patient status of IADP is also affected by selection bias as surgeons may use an IADP because of factors that are unable to be controlled for in this study, such as surgeon preference and intra-operative findings. It is also possible IADP use may lead to the overdiagnosis of bile leaks. For those −IADP patients, it is much more difficult to diagnose a bile leak because patients did not undergo routine post-operative imaging and most of the diagnosis was dependent on the clinical suspicion of the attending surgeon.

Data concerning other blood products transfused such as platelets, fresh frozen plasma, and cryoprecipitate were not included in the ACS-NSQIP database [28]. Allogenic and autologous blood transfusion were also not distinguished [29,30]. Last, the ACS-NSQIP database is limited to 30-day outcomes; long-term functional and cancer-related clinical outcomes are not available. Because the study is a retrospective analysis, it is unable to make claims regarding causality. In all, our study could still play a role in clarifying management of IADP in gallbladder cancer because the low incidence of gallbladder cancer may make it extremely difficult to perform an adequately powered, randomized controlled trial that would be required to prove causality. Future studies through a multicenter or multinational approach may be necessary.

Conclusions

In conclusion, more than 60% of gallbladder cancer surgeries included IADP and had higher rates of infectious complications, re-admission, and longer LOS. The use of IADP was not associated with an increased risk of PIC, unless drains were left in place at or beyond post-operative day 30. Instead, peri-operative blood transfusion was associated with a four-fold increased risk of PIC in addition to pre-operative biliary stent placement, biliary reconstruction, and bile leak. Future prospective randomized studies are needed to determine if IADP is truly a source of PIC or rather indicative of good surgical judgment for concern of post-operative bile leak. Although it appears that IADP in gallbladder cancer surgery does not have an associated risk with PIC, surgeons must be aware of the risk factors of PIC including those that may potentially be modifiable such as pre-operative biliary stent placement, reconstruction, and peri-operative blood transfusions.

Footnotes

Authors' Contributions

Conceptualization, methodology, investigation, writing, and review of the manuscript: Hasjim, Grigorian, Nahmias. Original draft of the manuscript: Hasjim. Formal analysis: Grigorian. Reviewing, editing, and validation of the study. Jutric, Wolf, Yamamoto, Imagawa. Direct supervision of the investigation: Nahmias.

Funding Information

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author Disclosure Statement

The authors have no relevant financial or non-financial interests to disclose. The study was deemed exempt from the institutional review board and consent due to the use of a national deidentified database.

References

Farha

, Chang

, Matthews

. Drainage in elective cholecystectomy. Am J Surg, 1981; 142:678–680.

Monson

JRT

, Guillou

, Keane

FBV

, et al. Cholecystectomy is safer without drainage: The results of a prospective, randomized clinical trial. Surgery, 1991; 109:740–746.

Gurusamy

, Samraj

, Mullerat

, Davidson

. Routine abdominal drainage for uncomplicated laparoscopic cholecystectomy. Cochrane Database Syst Rev, 2007;(4):CD006004.

Gurusamy

, Samraj

. Routine abdominal drainage for uncomplicated open cholecystectomy. Cochrane Database Syst Rev, 2007;(2):CD006003

Pekolj

, Alvarez

, Palavecino

, et al. Intraoperative management and repair of bile duct injuries sustained during 10,123 laparoscopic cholecystectomies in a high-volume referral center. J Am Coll Surg, 2013; 216:894–901.

Carriaga

, Henson

. Liver, gallbladder, extrahepatic bile ducts, and pancreas. Cancer, 1995; 75:171–190.

NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines^®) Hepatobiliary Cancers. 2019. https://www.nccn.org/professionals/physician_gls/pdf/hepatobiliary.pdf (Last accessed March 1, 2021).

Capussotti

, Ferrero

, Viganò

, et al. Bile leakage and liver resection: Where is the risk?. Arch Surg, 2006; 141:690–694.

Cherqui

, Benoist

, Malassagne

, et al. Major liver resection for carcinoma in jaundiced patients without preoperative biliary drainage. Arch Surg, 2000; 135:302–308.

10.

Aldameh

, McCall

, Koea

. Is routine placement of surgical drains necessary after elective hepatectomy? Results from a single institution. J Gastrointest Surg, 2005; 9:667–671.

11.

Fong

, Brennan

, Brown

, et al. Drainage is unnecessary after elective liver resection. Am J Surg, 1996; 171:158–162.

12.

Dixon

, Vollmer

, Sahajpal

, et al. An aggressive surgical approach leads to improved survival in patients with gallbladder cancer: A 12-year study at a North American Center. Ann Surg, 2005; 241:385–394.

13.

ACS NSQIP Meritorious Hospitals. www.facs.org/quality-programs/acs-nsqip/meritorious (Last accessed December 7, 2019).

14.

ACS-NSQIP Operations Manual. 2019;1–39.

15.

Dosch

, Grigorian

, Delaplain

, et al. Perioperative blood transfusion is associated with an increased risk for post-surgical infection following pancreaticoduodenectomy. HPB, 2019; 21:1577–1584.

16.

Povoski

, Karpeh

, Conlon

, et al. Preoperative biliary drainage: Impact on intraoperative bile cultures and infectious morbidity and mortality after pancreaticoduodenectomy. J Gastrointest Surg, 1999; 3:496–505.

17.

Kawai

, Tani

, Terasawa

, et al. Early removal of prophylactic drains reduces the risk of intra-abdominal infections in patients with pancreatic head resection: Prospective study for 104 consecutive patients. Ann Surg, 2006; 244:1–7.

18.

Houbiers

, Velde

, Watering

, et al. Transfusion of red cells is associated with increased incidence of bacterial infection after colorectal surgery: A prospective study. Transfusion, 1997; 37:126–134.

19.

Gianotti

, Pyles

, Alexander

, et al. Identification of the blood component responsible for increased susceptibility to gut-derived infection. Transfusion, 1993; 33:458–465.

20.

Hallet

, Mahar

, Nathens

, et al. The impact of perioperative blood transfusions on short-term outcomes following hepatectomy. Hepatobiliary Surg Nutr, 2018; 7:1–10.

21.

Ghio

, Contini

, Mazzei

, et al. Soluble HLA class I, HLA class II, and Fas ligand in blood components: A possible key to explain the immunomodulatory effects of allogeneic blood transfusions. Blood, 1999; 93:1770–1777.

22.

Ottonello

, Ghio

, Contini

, et al. Nonleukoreduced red blood cell transfusion induces a sustained inhibition of neutrophil chemotaxis by stimulating in vivo production of transforming growth factor-1 by neutrophils: role of the immunoglobulinlike transcript 1, sFasL, and sHLA-I. Transfusion, 2007; 47:1395–1404.

23.

Vamvakas

, Blajchman

. Transfusion-related immunomodulation (TRIM): An update. Blood Rev, 2007; 21:327–348.

24.

Peterson

CGB

, Skoog

, Venge

. Human eosinophil cationic proteins (ECP and EPX) and their suppressive effects on lymphocyte proliferation. Immunobiology, 1986; 171:1–13.

25.

Yoon

, Han

, Cho

, et al. Is laparoscopy contraindicated for gallbladder cancer? A 10-year prospective cohort study. J Am Coll Surg, 2015; 22:847–853.

26.

Haghighi

, Wang

, King

, et al. In-line radiofrequency ablation to minimize blood loss in hepatic parenchymal transection. Am J Surg, 2005; 190:43–47.

27.

Gumbs

, Jarufe

, Gayet

. Minimally invasive approaches to extrapancreatic cholangiocarcinoma. Surg Endosc, 2013; 27:406–414.

28.

Sarani

, Dunkman

, Dean

, et al. Transfusion of fresh frozen plasma in critically ill surgical patients is associated with an increased risk of infection. Crit Care Med, 2008; 36:1114–1118.

29.

Mezrow

, Bergstein

, Tartter

. Postoperative infections following autologous and homologous blood transfusions. Transfusion, 1992; 32:27–30.

30.

Heiss

, Jauch

, Delanoff

, et al. Beneficial effect of autologous blood transfusion on infectious complications after colorectal cancer surgery. Lancet, 1993; 342:1328–1333.