If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Outcomes following the inability to control the cystic duct due to a hostile triangle of Calot during cholecystectomy remain unknown. The purpose of this study was to analyze the safety and efficacy of subtotal cholecystectomy, with attention to the necessity for secondary interventions.
Methods
Sixteen thousand five hundred ninety six cholecystectomies from January 2002 to August 2014 were reviewed, identifying patients managed with subtotal cholecystectomy, defined as the inability to isolate/transect the cystic duct. After propensity matching, we investigated surgical indications, perioperative outcomes, and the necessity for secondary ERCP, percutaneous drainage, and completion cholecystectomy.
Results
65 (0.39%) patients underwent subtotal cholecystectomy; 54 (83.1%) began laparoscopically, of which 30 (55.6%) required conversion to laparotomy. Subtotal cholecystectomy, performed more frequently for acute cholecystitis (70.8% vs 34.6%), was associated with extended hospitalizations (4 d vs 2 d) and frequent surgical site infections (20% vs 4.6%). 25 (38.5%) subtotal cholecystectomy patients required ≥1 secondary intervention, and compared to standard cholecystectomy, underwent higher rates postoperative ERCP (30.8% vs 5.4%), percutaneous drainage (9.2% vs 1.5%), and completion cholecystectomy (6.2% vs 0%) [all P < 0.05].
Discussion
Subtotal cholecystectomy fails to control the cystic duct, resulting in significant morbidity. Most do not require completion cholecystectomy; however, patients demand close observation and, frequently, secondary interventions.
Introduction
The humbling experience of operating on a sick gallbladder, with a hostile triangle of Calot that fails to show its anatomy, namely the cystic duct in relationship to the portal structures, is not unfamiliar to general surgeons. The most dreaded complication in this setting is that of a common bile duct injury. When the critical view of safety cannot be established,
one surgical option to avoid major biliary injury is to perform a subtotal cholecystectomy. However, outcomes following this approach remain largely unknown.
In an extensive review of the literature, three major descriptions of the surgical technique are reported. The first technique includes controlling the cystic duct, either with clips, or from within Hartmann's pouch with suture, leaving only a strip of posterior mucosa adherent to the liver (Fig. 1a).
The latter two techniques include a nuance that our group relies upon in the definition of subtotal cholecystectomy, which is the surgeon's inability to safely ligate the cystic duct.
Figure 1Surgical technique for subtotal cholecystectomy. (a). The cystic duct is ligated, either with clips, or from within Hartmann's pouch with suture, leaving only a strip of posterior mucosa adherent to the liver. (b). Without ligating the cystic duct, the gallbladder is transected above the infundibulum and Hartmann's pouch is left open and drained. (c). Without ligating the cystic duct, the gallbladder is transected above the infundibulum and Hartmann's pouch is closed, either with a stapler or suture. Techniques illustrated in (b) and (c) meet our criteria for subtotal cholecystectomy, whereas Technique (a) does not
The largest single center study in the literature regarding subtotal cholecystectomy includes 168 patients; however, 157 of these patients had anatomy that allowed for identification and division of the cystic duct and artery.
In addition to the majority of their cohort undergoing an operation that we do not consider to be a subtotal cholecystectomy, their postoperative outcomes are limited to morbidity and mortality without discussion of secondary procedures.
However, most of the included references report a variety of inconsistently measurable outcomes, and with the combination of at least three major surgical techniques, the current body of literature remains relatively fickle, making it difficult for the reader to extrapolate the true expected outcomes after this operation.
The purpose of this study was to (i) clearly define subtotal cholecystectomy as an operation that specifically does not control the cystic duct, (ii) analyze a contemporary data set of an entire institutional experience with this operation, including laparoscopic and open approaches, (iii) report perioperative morbidity and mortality after subtotal cholecystectomy, and (iv) determine the subsequent rate of secondary ERCP, percutaneous drainage procedures, and completion cholecystectomy. The present study contributes to the literature by providing a more sophisticated understanding for subtotal cholecystectomy, and informing surgeons of what they can expect after removing a portion of the gallbladder without controlling the cystic duct.
Methods
With approval from the institutional review board, the Duke Health System was queried using an internal search engine for data extraction, Duke Enterprise Data Unified Content Explorer (DEDUCE), relying on Current Procedural Terminology (CPT) codes to identify patients that had undergone a cholecystectomy between January 2002 and August 2014.
DEDUCE functions as an electronic research tool designed to provide health care investigators access to patient-related clinical information, obtained through routine patient-centered care.
These data were cross-referenced and supplemented with targeted manual chart review. To identify all adult (18 years or older) patients in the Duke Health System having undergone a cholecystectomy, CPT codes (47562, 47563, 47564, 47600, 47605, 47610, 47612, 47620) and ICD procedure codes (51.22, 51.23) were utilized. To identify subtotal cholecystectomy patients, the ICD procedure codes 51.21 (other partial cholecystectomy) and 51.24 (laparoscopic partial cholecystectomy) were utilized.
Operative notes were reviewed to ensure consistency and accuracy, which included only those patients in which cholecystectomy was the intended operation. Additionally, subtotal cholecystectomy was defined as the surgeon's inability to control (identify, ligate, and transect) the cystic duct. Briefly, standard four-trocar laparoscopic cholecystectomies were most often performed. After dissection of inflammatory adhesions to the gallbladder, a laparoscopic grasper was used to retract the fundus of the gallbladder in a cephalad manner over the liver. Invariably, in this subset of patients, the triangle of Calot was densely inflamed, prohibiting safe dissection of the cystic duct and artery, making it impossible to achieve the critical view of safety. Most frequently, a ‘dome-down’ approach was utilized at this time, until either the surgeon's comfort level was exceeded and the gallbladder was transected at the infundibulum, or the operation was converted to an open procedure. During this dissection, some surgeons left remnant posterior gallbladder mucosa on the cystic plate. Additionally, opening the gallbladder to evacuate the contents, and using the internal queues to guide dissection in a safe manner, so as to avoid injury to the main biliary structures, was a commonly utilized tactic. Attempts to close the cystic duct with suture from within the gallbladder were not made. Operative drains were placed at the surgeon's discretion, typically in patients whose Hartmann's pouch was left open. For inclusion in this study, the following criterion were required: (i) failure to identify, ligate, and transect the cystic duct, either with external clips in the standard fashion or from within Hartmann's pouch using suture, and (ii) Hartmann's pouch was either closed with a stapler or suture (Fig. 1c), or a surgical drain was placed at the location of the open Hartmann's pouch (Fig. 1b).
Retrospective analysis of patient demographics, comorbidities, diagnosis and indications for cholecystectomy, radiology and operative reports, intraoperative variables (operative time, blood loss, conversion to open procedure), length of hospitalization, postoperative morbidity, and mortality, was performed. Additionally, patients were assessed for secondary interventions, including ERCP, percutaneous drainage procedures, and completion cholecystectomy.
Patients were grouped according to extent of surgical resection: subtotal cholecystectomy versus standard complete resection. Primary predictor variables included patient characteristics and demographics, as well as presence of specific comorbidities (including diabetes mellitus, chronic obstructive pulmonary disease, coronary artery disease, congestive heart failure, atrial fibrillation, peripheral vascular disease, and preoperative sepsis). The primary outcomes of interest were length of stay (LOS), hospital readmission, and interval mortality (in-hospital, 30-day, and 1-, 3-, and 5-year). Secondary outcomes included postoperative re-interventions (ERCP, percutaneous drainage and reoperation for completion cholecystectomy), as well as specific perioperative complications.
Baseline characteristics between groups were compared using Pearson's chi-square test for discrete variables and Student's t-test or Fisher's exact test for continuous data. Given the inherent differences between patients treated with subtotal versus complete cholecystectomy, propensity scores were then developed using a logistic regression model, which was defined as the probability of being treated with subtotal cholecystectomy, conditioned on the following variables: age at time of surgery, sex, and presence of diabetes mellitus, chronic obstructive pulmonary disease, congestive heart failure, coronary artery disease, atrial fibrillation, peripheral vascular disease, and presences of preoperative sepsis/SIRS. Patients treated with subtotal cholecystectomy were then matched to standard cholecystectomy controls using a nearest neighbor algorithm in an iterative process, first 3:1 but finalized as a 2:1 ratio following exclusion of inappropriate matches based on concomitant procedures (MatchIt: Nonparametric Preprocessing for Parametric Casual Inference). Balance between propensity-matched groups was assessed with standard statistical tests as well as calculated standard differences. Outcomes were compared in a similar manner, both before and after propensity adjustment. To examine the effect of surgical technique among patients undergoing subtotal cholecystectomy, a subgroup analysis was performed comparing patients treated with a laparoscopic versus open approach.
Type I error was controlled at the level of the comparison, and P-values <0.05 were considered statistically significant. All analyses were performed using R (version 3.1.2, R Foundation for Statistical Computing, Vienna, Austria).
Results
Baseline characteristics of all patients are shown in Table 1. Of the 16,596 patients that underwent cholecystectomy between January 2002 and August 2014, 65 (0.39%) patients underwent subtotal cholecystectomy. Hartmann's pouch was either closed with a stapler or suture in 50 patients (76.9%; Fig. 1c), or a surgical drain was placed at the location of the open Hartmann's pouch in 15 patients (23.1%; Fig. 1b). To compensate for possible, nonrandom, fundamental differences between subtotal and standard cholecystectomy patients, propensity analysis was performed, for which the matched baseline characteristics are also shown in Table 1. The propensity analysis yielded a 2:1 matched cohort with similar age and comorbidity profiles, providing a means to compare similarly sick patients undergoing subtotal cholecystectomy to those undergoing standard cholecystectomy.
Table 1Baseline patient characteristics of patients undergoing standard and subtotal cholecystectomy, unadjusted and following propensity matching
The etiology driving surgical intervention for the 65 patients undergoing subtotal cholecystectomy was acute cholecystitis in 46 (70.8%), chronic cholecystitis in 9 (13.8%), symptomatic cholelithiasis in 14 (21.5%), and gallstone pancreatitis in 7 (10.8%). In the matched 130 standard cholecystectomy patients, the indication for surgery was acute cholecystitis in 45 (34.6%), chronic cholecystitis in 15 (11.5%), symptomatic cholelithiasis in 68 (52.3%), biliary dyskinesia in 12 (9.2%), and gallstone pancreatitis in 20 (15.4%). 10 (15.4%) patients in the subtotal group and 7 (5.4%) patients in the standard cholecystectomy group had a prior percutaneous cholecystostomy tube. While the majority of records indicated a single indication for surgery, 11 patients in the subtotal cholecystectomy group and 30 patients in the standard cholecystectomy group carried two diagnoses (i.e. acute cholecystitis in the setting of gallstone pancreatitis).
Perioperative outcomes are shown in Table 2. In the propensity matched analysis, subtotal cholecystectomy patients were hospitalized longer after their operation (4 days vs 2 days), and while they had similar in-hospital and 30-day mortality, subtotal cholecystectomy patients had up to a 4-fold increase in 1-year (8.1% vs 1.6%), 3-year (16.7% vs 4.7%), and 5-year (33.3% vs 12%) mortality. The only death in the subtotal cholecystectomy group within 30 days of surgery was secondary to an electrolyte-related arrhythmia on the first postoperative day. Subtotal cholecystectomy patients also experienced higher rates of surgical site infection (20% vs 4.6%).
Table 2Outcomes for patients undergoing standard and subtotal cholecystectomy, following propensity matching
In a subset analysis, laparoscopic subtotal cholecystectomy was compared to all open (laparoscopic converted to open and planned open) subtotal cholecystectomies (Table 3). Postoperative complications, including surgical site infection, were similar between techniques. However, the postoperative length of stay was nearly double for the open subgroup, having a median length of stay of 5.5 days, compared to 3 days for the laparoscopic approach.
Table 3Outcomes among the subgroup of patients treated with subtotal cholecystectomy, laparoscopic versus open approach
Patients were evaluated for secondary interventions in the form of ERCP, percutaneous drainage procedures, and completion cholecystectomy (Table 4A). Overall, 25 (38.5%) subtotal cholecystectomy patients underwent at least one secondary intervention. Compared to the standard cholecystectomy group, patients undergoing a subtotal cholecystectomy had a nearly 6-fold higher rate of ERCP (30.8% vs 5.4%), with 20 subtotal cholecystectomy patients requiring 49 ERCPs, compared to 7 patients in the standard cohort undergoing 11 ERCPs. The time interval between the index operation and first ERCP was widely variable in the subtotal cholecystectomy cohort, ranging from the day of surgery to 195 days postoperatively (median = 8 days, mean = 29.5 days). Indications for ERCP in the first 20 days after subtotal cholecystectomy included biliary leak (10 patients, 50%), choledocholithiasis (1 patient, 5%), and inability to adequately clear the biliary tree with intraoperative cholangiography (1 patient, 5%). ERCP was performed 21 days or longer after subtotal cholecystectomy for delayed biliary leak (3 patients, 15%), choledocholithiasis (2 patients, 10%), biliary stricture (2 patients, 10%), and pancreatitis (1 patient, 5%). Conversely, the interval between the index operation and first ERCP for the standard cholecystectomy cohort ranged from day of surgery to 8 days postoperatively (median = 1 day, mean 2.4 days). Indications for postoperative ERCP in the standard cholecystectomy cohort included choledocholithiasis (4 patients, 57.1%), biliary leak (2 patients, 28.6%), and failed intraoperative cholangiography (1 patient, 14.3%). None of the reported postoperative ERCPs in either cohort were performed for extraction of preoperatively placed biliary stents.
Table 4The incidence of secondary interventions after subtotal cholecystectomy
Variable
Overall
Standard chole (n = 130)
Subtotal chole (n = 65)
P-value
A. Subtotal versus standard cholecystectomy, following propensity matching
Post-op ERCP
27 (13.8%)
7 (5.4%)
20 (30.8%)
<0.001
Percutaneous drain
8 (4.1%)
2 (1.5%)
6 (9.2%)
0.01
Completion cholecystectomy
4 (2.1%)
0 (0%)
4 (6.2%)
0.007
Variable
Overall
Laparoscopic (n = 24)
Open (n = 41)
P-value
B. Among patients treated with subtotal cholecystectomy, by surgical approach (laparoscopic versus open approach)
Additionally, subtotal cholecystectomy patients experienced a greater than 6-fold increase in percutaneous drainage procedures for bilomas or to aid in controlling biliary leak (9.2% vs 1.5%). Of the 15 patients whose gallbladder remnant was left open and drained operatively (Fig. 1b), only 1 patient (6.7%) required percutaneous drainage for biloma due to operative drain malposition. The remaining 5 percutaneous drains were placed in patients whose gallbladder remnant was closed (10%). Furthermore, 4 (6.2%) patients in the subtotal cohort required a completion cholecystectomy a median of 11.5 months (range 5–27 months) after their index operation for recurrent symptoms (2 with recurrent cholecystitis, 1 with recurrent symptomatic cholelithiasis, and 1 with choledocholithiasis), compared to no patients in the standard group requiring such intervention. All patients requiring a completion cholecystectomy received a laparoscopic subtotal cholecystectomy (Table 4B).
Discussion
The results of this study demonstrate that subtotal cholecystectomy, an operation in which the surgeon is unable to safely identify and control the cystic duct, is most frequently performed at the time of acute inflammation, usually in older patients with multiple medical comorbidities. Subtotal cholecystectomy patients remain in the hospital after surgery twice as long, and have surgical site infections four times more frequently, than standard cholecystectomy patients. The ultimate risk for subtotal cholecystectomy patients is the necessity of secondary interventions in nearly 40% of patients, including ERCP, percutaneous drainage, or completion cholecystectomy. Subtotal cholecystectomy patients in this study had a 6-fold higher rate of postoperative ERCP and percutaneous drainage, usually for biliary leak. Interval completion cholecystectomy was required for the minority of subtotal cholecystectomy patients, but none of the patients that underwent standard cholecystectomy required reoperative surgery.
The technical challenges of operating on a severely diseased gallbladder are not unfamiliar amongst surgeons, where a hostile triangle of Calot fails to show its anatomy, and the inflammation prevents safe dissection and identification of the cystic duct in relationship to the portal structures. First reported in 1955 by Madding et al.,
and is described simply as the removal of only a portion of the gallbladder when inflammation precludes safe dissection and transection of the cystic duct.
Unfortunately, the current body of literature reports at least three major surgical techniques for subtotal cholecystectomy (Table 5), making it challenging to interpret this data. In the first reported technique (Fig. 1a), the surgeon controls the cystic duct with clips, or from within Hartmann's pouch using suture, leaving only a strip of posterior gallbladder mucosa adherent to the liver. This is the most commonly described approach, ranging in incidence from 2 to 48% of all cholecystectomies, and is associated with very few complications.
However, this technique, after ligating the cystic duct, should result in bile leak rates similar to standard cholecystectomy, and likewise should not pose a risk for recurrent symptoms; therefore, our group does not include this technique in the definition of subtotal cholecystectomy. We find it challenging to extrapolate the conclusions of these reports to our own patient population where the cystic duct is not ligated.
In the remaining two techniques, which are included in our definition of subtotal cholecystectomy, the surgeon is unable to safely control the cystic duct, and either drains an open Hartmann's pouch (Fig. 1b), or closes it with a stapler or suture (Fig. 1c). As demonstrated in Table 5, our study is the largest cohort of subtotal cholecystectomies utilizing the latter two techniques,
which leave the patient at risk for recurrent biliary symptoms due to an intact cystic duct. Based on this principle alone, we advise that the term subtotal cholecystectomy be reserved for an operation that fits this criteria, similar to the illustrations in Fig. 1b and c, where an inflamed triangle of Calot prohibits the surgeon from isolating and transecting the cystic duct.
Table 5Summary of previously published literature and current study, including surgical technique, postoperative morbidity (patients with a complication) and mortality, as well as secondary interventions
Author, year
All choles
Subtotal (% of all choles)
Subtotal cholecystectomies (% of Subtotal cholecystectomies)
Outcomes after subtotal cholecystectomy
Postoperative complications
30-day mortality (%)
Secondary interventions
Technique 1
Technique 2
Technique 3
PostopERCP (%)
Percutaneous drainage (%)
Completion cholecystectomy (%)
•
Cystic duct controlled with clips/stapler or from within Hartmann's pouch with suture
•
Posterior mucosa remains on liver
•
Cystic duct not controlled
•
Hartmann's pouch opened and drained
•
Posterior mucosa remains on liver (+/−)
•
Cystic duct not controlled
•
Hartmann's pouch closed with staples/suture
•
Posterior mucosa remains on liver (+/−)
Bornman et al., 1985
257
18 (7%)
18 (100%)
–
–
3
0 (0%)
NR
NR
NR
Cottier et al., 1991
290
11 (3.8%)
11 (100%)
–
–
2
0 (0%)
2 (18%)
NR
NR
Bickel et al., 1993
152
6 (3.9%)
6 (100%)
–
–
0
0 (0%)
NR
NR
NR
Katsohis et al., 1996
1654
34 (2.1%)
30 (88.2%)
4 (11.8%)
–
9
1 (3%)
3 (8.8%)
NR
0 (0%)
Michalowski et al., 1998
340
29 (8.5%)
27 (93.1%)
2 (6.9%)
–
16
1 (3.4%)
0 (0%)
2 (6.9%)
NR
Ransom, 1998
125
8 (6.4%)
7 (87.5%)
1 (12.5%)
–
3
0 (0%)
1 (12.5%)
NR
0 (0%)
Chowbey et al., 2000
1680
56 (3.3%)
56 (100%)
–
–
4
0 (0%)
1 (1.8%)
NR
NR
Beldi et al., 2003
345
46 (13.3%)
–
46 (100%)
–
4
1 (2.2%)
7 (15.2%)
1 (2.2%)
0 (0%)
Ji et al., 2006
3485
168 (4.8%)
157 (93.5%)
–
11 (6.5%)
13
0 (0%)
NR
1 (0.6%)
NR
Rohatgi et al., 2006
323
5 (1.5%)
–
5 (100%)
–
2
0 (0%)
1 (20%)
1 (20%)
NR
Sinha et al., 2007
889
28 (3.1%)
–
28 (100%)
–
5
0 (0%)
3 (10.7%)
1 (3.6%)
NR
Horiuchi et al., 2008
60
29 (48.3%)
26 (89.7%)
–
3 (10.3%)
2
0 (0%)
NR
NR
NR
Philips et al., 2008
1917
26 (1.4%)
–
26 (100%)
–
8
1 (3.8%)
5 (19.2%)
NR
NR
Nakajima et al., 2009
1226
60 (4.9%)
–
–
60 (100%)
4
0 (0%)
NR
NR
0 (0%)
Singhal et al., 2009
1150
52 (4.5%)
39 (75%)
–
13 (25%)
3
0 (0%)
4 (7.7%)
NR
NR
Tian et al., 2009
1558
48 (3.1%)
–
–
48 (100%)
4
0 (0%)
NR
NR
NR
Hubert et al., 2010
552
39 (7.1%)
39 (100%)
–
–
6
0 (0%)
2 (5.1%)
0
0 (0%)
Kuwabara et al., 2014
246
26 (10.6%)
–
–
26 (100%)
0
0 (%)
1 (3.8%)
NR
NR
Harilingam et al., 2016
993
64 (6.4%)
64 (100%)
–
–
8
0 (0%)
6 (12%)
NR
NR
Lidsky et al., 2017
16,596
65 (0.39%)
–
15 (23.1%)
50 (76.9%)
33
1 (1.5%)
20 (30.8%)
6 (9.2%)
4 (6.2%)
Abbreviations: Choles = cholecystectomies; NR = not reported.
The above definition is distinct from the recent definition proposed in the descriptive publication by Strasberg et al., who refers to two types of subtotal cholecystectomy.
Subtotal cholecystectomy-“Fenestrating” vs “Reconstituting” subtypes and the prevention of bile duct injury: definition of the optimal procedure in difficult operative conditions.
The first type is the subtotal fenestrating cholecystectomy, where the cystic duct is closed from within the lumen of the gallbladder using a purse-string technique.
Subtotal cholecystectomy-“Fenestrating” vs “Reconstituting” subtypes and the prevention of bile duct injury: definition of the optimal procedure in difficult operative conditions.
The alternative is that of a subtotal reconstituting cholecystectomy, where the cystic duct is not closed and the gallbladder remnant is reapproximated with suture or staples. Although the reconstituting subtotal cholecystectomy described by Strasberg and colleagues is consistent with the technique described in this study, we are hesitant to include to the fenestrating subtotal cholecystectomy in our definition, as this technique closes the cystic duct and therefore, theoretically, reduces the risk of bile leak and recurrent symptoms to that of a standard cholecystectomy.
Subtotal cholecystectomy-“Fenestrating” vs “Reconstituting” subtypes and the prevention of bile duct injury: definition of the optimal procedure in difficult operative conditions.
While our review of the literature demonstrates many important single-center contributions pertaining to subtotal cholecystectomy, two meta-analyses provide the most meaningful contribution to our understanding of this operation to date, and at least partially compensate for the infrequently performed operation.
Unfortunately, the relatively small studies included in each of these meta-analyses are limited by definition variability, multiple techniques, and inconsistently reported outcomes. In thirty manuscripts reviewed by Elshaer et al.,
all but 100 patients in three studies had their cystic duct ligated, and therefore represent a different patient population with an alternate postoperative risk profile. Additionally, the largest single institution study analyzed by Elshaer et al. included only 46 patients,
Also within the literature, Ji et al. published the largest single center experience reporting on outcomes after subtotal cholecystectomy, but the cystic duct was successfully ligated in 93.5% (157 of 168) of their patients, and therefore does not align with our definition of subtotal cholecystectomy.
Two additional studies published by Nakajima et al. and Tian et al. in 2009 include similar sized cohorts of patients that do meet our criterion for subtotal cholecystectomy, however, they do not report on the rates of secondary interventions.
To our knowledge, the present study includes the largest series of patients having undergone a subtotal cholecystectomy that fulfills the criterion described herein, reporting postoperative morbidity and mortality, as well as the necessity for secondary interventions in the form of ERCP, percutaneous drainage, and completion cholecystectomy; therefore, we provide the most thorough assessment of this patient population to date (Table 5).
Furthermore, the available meta-analyses conclude that outcomes after subtotal cholecystectomy are similar to those of standard cholecystectomy.
However, these studies do not account for selection bias and the potential for two different patient populations undergoing either subtotal or standard cholecystectomy. Hypothesizing that subtotal cholecystectomy patients represent a different patient cohort, one in which patients are older and comorbidities are more common, we performed a propensity analysis to optimize the likelihood of contrasting similarly ill patients under our care, therefore yielding the ability to focus on outcomes as a reflection of technique: subtotal versus standard cholecystectomy. Our goal was to provide surgeons with a more sophisticated appreciation for this operation, and the projected postoperative course for patients that require it.
Importantly, 84% of subtotal cholecystectomy patients underwent surgery for acute or chronic inflammation, compared to 46% in the propensity matched standard cholecystectomy cohort. This finding is likely the primary factor that impeded safe dissection, ligation, and subsequent transection of the cystic duct.
Postoperatively, subtotal cholecystectomy patients experience relatively similar complications, with the exception of surgical site infections. Compared to prior studies that report combined wound infection rates of 2.6%,
our cohort experienced a nearly 10-fold higher incidence in such infectious complications. We attribute this discrepancy to variation in technical definitions; our patients did not have their cystic duct ligated, and therefore had the potential for ongoing bile drainage and ensuing infectious complications. Additionally, this patient cohort requires a length of hospitalization that is double that of standard cholecystectomies, likely due to the greater than 50% conversion rate from a laparoscopic to open approach in our series.
Failure to control the cystic duct in subtotal cholecystectomy patients resulted in a secondary intervention rate of 38.5%, with 25 of 65 patients requiring at least one intervention after the index operation. Compared to the current body of literature, we report the highest rates of postoperative ERCP, nearly the highest rate of percutaneous biloma drainage, and the highest necessity for completion cholecystectomy (Table 5). While we could easily attribute this observation again to the surgical technique, only having included patients whose surgeon was unable to control the cystic duct, we report a more meaningful comparison of subtotal to standard cholecystectomy patients. Propensity matching demonstrated an increase in the need for immediate (first 20 days after index operation) and delayed (greater than 20 days after index operation) postoperative ERCP (Table 4A) in the subtotal cholecystectomy patients, which, at our institution and others, is a commonly utilized therapy for the management of biliary leak.
We attribute this observation to the more prevalent inflammatory state associated with the subtotal cholecystectomy cohort, which leads to friable remnant tissue that is surely prone to breakdown. Combined with the surgeon's inability to ligate the cystic duct, there exists an increased likelihood for ongoing bile leak. Additionally, closing an edematous Hartmann's pouch in the setting of such inflammation may result in bile leak when the edema subsides and the suture or staple line loosens.
If the bile leak continues after ERCP and/or a sizable biloma develops, it is our practice to perform percutaneous drainage as well. Our utilization of such percutaneous interventions is more frequent than many previous reports, likely because prior studies report on a surgical technique that ligates the cystic duct and therefore reduces the potential for bile leak.
Additionally, although reported, this study is not powered to detect a difference in postoperative drainage procedures between patients whose gallbladder remnant was closed, versus those that were left open and drained. Lastly, in the propensity-matched group, 4 patients in the subtotal cholecystectomy cohort required completion cholecystectomy for recurrent symptoms, a median of 11.5 months after the index operation. This outcome may be attributed to an intact cystic duct with potential to drain into the gallbladder remnant, which likely serves as a reservoir for bile stasis, stone formation, and recurrent symptoms.
Incomplete evacuation of gallbladder contents at the time of subtotal cholecystectomy could also contribute to the observed recurrence of symptoms necessitating completion cholecystectomy. Interestingly, we observed that only patients in the laparoscopic subtotal cholecystectomy cohort underwent reoperation to remove the gallbladder remnant, whereas those that underwent a converted or planned open operation required no such intervention. While this study is not powered to statistically compare completion cholecystectomy rates between laparoscopic and all open subtotal cholecystectomies, it is our suspicion that surgeons persisting laparoscopically should be more aggressive to remove more of the anterior wall of the gallbladder, so as to more adequately evacuate the gallbladder contents and leave behind a smaller remnant less prone to disease recurrence.
Although the results of this study provide surgeons with the most comprehensive understanding of subtotal cholecystectomy to date, there remain several limitations. First, this non-randomized, retrospective study has its inherent limitations, including the observation that some patients had multiple documented indications for surgery. Additionally, the decision to perform a subtotal cholecystectomy is not made preoperatively, and therefore randomization is not feasible. Furthermore, this series is underpowered to determine the benefit or detriment to laparoscopic versus open subtotal cholecystectomy. Given subtotal cholecystectomy is a relatively uncommon operation, a multi-center analysis in the future may include a larger cohort to draw such conclusions. However, given the variability and impact of surgical technique and critical definitions on patient inclusion criteria, identifying surgeons performing what we define as subtotal cholecystectomy may be challenging. Furthermore, we determined that a critical chart review of 16,596 patients would likely be prone to inconsistencies in data procurement. Therefore, we performed a propensity analysis to compensate for and minimize selection bias, rather than examine the entire Duke Health System cohort.
Conclusions
This study provides a concrete definition for subtotal cholecystectomy, which includes the surgeon's inability to safely identify, isolate, and ligate the cystic duct. In doing so, surgeons can begin to appreciate what it means to perform a subtotal cholecystectomy, and develop realistic expectations of how these patients will perform postoperatively. The decision to perform a subtotal cholecystectomy should be based on natural stopping points for reasons of safety while attempting to perform a standard cholecystectomy.
Based on the results of this study, subtotal cholecystectomy, while perhaps cautious, is associated with a cost. Although subtotal cholecystectomy in the setting of a hostile Triangle of Calot avoids common bile duct injury, postoperative morbidity is not insignificant. Compared to standard cholecystectomy, surgeons performing subtotal cholecystectomy can anticipate a postoperative length of stay that is twice as long and more frequent surgical site infections. Patients also require close monitoring for bile leak-related events and recurrent symptoms, as secondary interventions are required in more than one-third of subtotal cholecystectomy patients whose cystic duct is not safely identified and ligated.
As such, aggressive biliary duct clearance, attempts to internally control the cystic duct from within the gallbladder lumen, and more liberal use of operative drainage, even when Hartmann's pouch is closed, may help to minimize some of the morbidity and subsequent secondary interventions observed in this study.
Funding sources
None.
Conflict of interest
None declared.
Acknowledgements
The authors of this manuscript would like to acknowledge Stan Coffman (Medmedia Solutions, Durham, NC) for providing the medical illustrations.
Legend
A-fib
atrial fibrillation
COPD
chronic obstructive pulmonary disease
CHF
congestive heart failure
CAD
coronary artery disease
ERCP
endoscopic retrograde cholangio-pancreatography
LOS
length of stay
NS
non-significant
PVD
peripheral vascular disease
SSI
surgical site infection
SIRS
systemic inflammatory response syndrome
References
Strasberg S.M.
Avoidance of biliary injury during laparoscopic cholecystectomy.
Subtotal cholecystectomy-“Fenestrating” vs “Reconstituting” subtypes and the prevention of bile duct injury: definition of the optimal procedure in difficult operative conditions.
30516-6&id=gr1.jpg){kind=link}