Vivek Gupta, Abhijit Chandra , Vishal Gupta, Ravi Patel, Amit Dangi, Ajay Pai

Department of Surgical Gastroenterology, King George’s Medical University, Lucknow, Uttar Pradesh 226003, India

Keywords: Acute calculus cholecystitis Cholecystectomy Cholecysto-enteric fistula Niemeier classification

ABSTRACT Background: Spontaneous gallbladder perforation (GBP) is an uncommon diagnosis. This study presented the experience of managing spontaneous GBP over nine years at a large, tertiary care university hospital in north India and investigated the outcomes and treatment strategies. Methods: A retrospective review of prospectively maintained digital database of consecutive patients was performed. All patients received medical and/or surgical treatment for spontaneous GBP in our depart- ment between January 2010 and June 2018. Results: We identified 151 patients (81 females and 70 males) with mean age of 53 years. Most common presenting features were pain (96.7%), fever (54.3%) and jaundice (31.1%). Most common cause was gall- bladder stones (84.8%) followed by common bile duct stones (30.5%), xanthogranulomatous cholecystitis (17.9%) and malignancy (11.9%). As per Niemeier classification, 8.6% had type 1 GBP (free perforation in peritoneal cavity), 76.2% had type 2 GBP (localized perforation) and 13.2% had type 3 GBP (cholecysto- enteric fistula). About 60% of the perforations were diagnosed preoperatively. Type 1 was more com- mon in patients with diabetes and also had the worst prognosis. Surgery was performed in 109 patients (72.2%). Seven patients (4.6%) had a postoperative morbidity of Clavien-Dindo III or higher. There were three mortalities in patients who underwent surgery. Conclusions: High index of suspicion is required for preoperative diagnosis of GBP, especially in types 2 and 3. Laparoscopic cholecystectomy can be difficult in these patients and patients may require open or partial cholecystectomy. Early diagnosis and step-up approach for the treatment of GBP is critical.

Introduction

Gallbladder perforation (GBP) is an uncommon diagnosis, usu- ally associated with acute cholecystitis. The etiology of GBP re- mains poorly understood and it presents a diagnostic and surgi- cal challenge [ 1 , 2 ]. Symptoms and signs of GBP may be mimicked by various other pathologies, such as cholecystitis, liver abscess and gallbladder (GB) malignancy. An early and correct diagnosis of GBP is imperative for proper management because of high morbid- ity and mortality associated with untreated disease [ 1 , 2 ]. GBP has been classified into three categories by Niemeier (1934): free per- foration into the peritoneal cavity (type 1), contained localized per- foration (type 2) and cholecysto-enteric fistulas (type 3) [3] . Type 2 is the most common type in previously reported series [ 1 , 4 , 5 ]. Cases of intrahepatic perforation of the gallbladder with liver ab- scess and cholecysto-hepatic communication are very rare but have been reported [6] . In this study, we present our experience with managing GBP at a large, tertiary care university hospital in north India.

Patients and methods

Patients and managing

This study was performed at a tertiary care university hospital in north India. All consecutive cases of spontaneous GBP occurring between January 2010 and June 2018 were retrospectively identi- fied from a prospectively maintained computerized database. GBP due to trauma and iatrogenic causes were excluded.

We followed a step-up approach for managing GBP patients. All patients were initially managed conservatively with intravenous fluids and antibiotics. Work up included routine blood investi- gations, imaging, blood and bile cultures. Patients with intra- abdominal collections were managed with percutaneous drain placement (PCD). Endoscopic retrograde cholangiopancreatography (ERCP) or percutaneous transhepatic biliary drainage (PTBD) was performed, if necessary, for control of sepsis. Surgery was per- formed at a later stage after adequate resuscitation and control of sepsis. Demographic and clinical information including age, sex, febrile episodes, jaundice, white blood cell counts, type of GBP, site of perforation, intraoperative data including type of surgical proce- dure, timing of surgery (interval or upfront), duration of surgery as well as postoperative data, including complications, length of stay and mortality were collected and analyzed. Patient’s records and investigations were reviewed, including X-ray abdomen, ultrasound (US), and CT scan to identify specific predictors of GBP preopera- tively.

The treatment options, including PCD, laparoscopic or open cholecystectomy, common bile duct (CBD) exploration or other sur- gical procedures, were compared with patient outcomes to deter- mine the best management strategy depending on the type of GBP.

Statistical analysis

Data were analyzed using SPSS version 27 (IBM Corp., Armonk, NY, USA). Continuous variables were reported using mean ± stan- dard deviation (SD) and categorical variables were reported us- ing number (percentage). Student’st-test and Chi-square test were used to compare between groups. Multivariate analyses were per- formed to study the effect of each parameter. AP＜0.05 was con- sidered statistically significant.

Results

Clinical presentation

There were 151 patients (81 females and 70 males) with mean age of 53.08 ± 13.04 years. The baseline characteristics are de- scribed in Table 1 . Right upper abdominal pain was the most com- mon presenting complaint (96.7%) followed by fever (54.3%) and jaundice (31.1%). Most common etiology was gallbladder stones (84.8%) followed by CBD stones (30.5%), xanthogranulomatous cholecystitis (XGC) (17.9%) and malignancy (11.9%). Diabetes was present in 19.9% of cases, while hypertension was present in 15.9%. At time of admission, 26.5% had acute kidney injury and 15.9% had hypotension. The mean duration between onset of symptoms to presentation to our center was 79.0 ± 4.5 weeks in type 1 GBP, 87.0 ± 5.1 weeks in type 2 GBP and 104.0 ± 6.7 weeks in type 3 GBP.

Imaging characteristics

Abdominal US was performed on all patients and abdominal CT scan was performed on 92 patients (60.9%). A correct preoperative diagnosis of GBP was made on imaging (US/CT) in 91 (60.3%) pa- tients. Fifty-seven (37.7%) of the perforations were diagnosed dur- ing surgery. Particularly, 90% of type 3 GBP were diagnosed intra- operatively during exploration for cholecystectomy (16 patients) or gallbladder malignancy (2 patients). Gallbladder stones were seen in 84.8% of the patients, while concomitant CBD stones were seen in 46 patients (30.5%). Abdominal US was able to locate site of GBP (sonological hole sign) preoperatively in 51 patients. A thick-walled gallbladder (wall thickness＞4 mm) was seen in 51.6% of the cases. CT revealed a mass lesion in 26 (28.3%) of patients. Eighteen of these 26 patients with a suspected gallbladder mass lesion were found to have gallbladder malignancy, while the other 8 had focal gallbladder perforation masquerading as a mass lesion. Findings of the US and the CT scan are summarized in Table 2 . Figs. 1-3 show radiological and intraoperative images in three types of GBP.

Table 1 Patient characteristics and summary of patients with sponta- neous gallbladder perforation ( n = 151).

Initial management

As per our step-up management protocol, patients were ini- tially managed with fluid resuscitation and intravenous antibiotics followed by PCD for intra-abdominal collections, biliary drainage or surgery. Medical management with antibiotics was successful in 21 (18.26%) patients. US guided PCD was done in 39 patients (25.8%). Forty-two patients (27.8%) with biliary obstruction under- went ERCP, while two underwent PTBD. Emergency surgery at pre- sentation was performed in 3 of 151 (1.9%) patients, all of whom had type 1 GBP ( Table 3 ). All three patients who underwent emer- gency surgery had CBD stones and required cholecystectomy along with CBD exploration and repair over T-tube.

Table 2 US and CT characteristics of patients with gallbladder perforation.

Table 3 Overview of initial management of gallbladder perforation.

Table 4 Comparison of three types of gallbladder perforation.

Step-up approach for management

After initial stabilization with conservative management or PCD, patients were re-evaluated for further procedures. A total of 106 patients (70.2%) underwent definitive cholecystectomy at an aver- age interval of 57 days. Twenty-eight patients underwent concomi- tant CBD exploration and 14 underwent biliary drainage proce- dures. Partial cholecystectomy was performed in 22 patients. Seven patients underwent extended cholecystectomy for gallbladder ma- lignancy. Mean postoperative hospital stay was 8.93 ± 5.27 days with 7 patients reporting morbidity of Clavien-Dindo III and higher. There were three mortalities within 30 days of admission. One was due to sepsis and the other two were due to non-surgical causes. No differences in age, sex, symptoms or associated pathology with respect to types of GBP were noted. Forty-two (27.8%) patients didnot undergo cholecystectomy due to refusal of surgery (n= 21, 13.9%), co-morbidities precluding surgery (n= 5, 3.3%) or advanced malignancy (n= 16, 10.6%).

Specific characteristics of subtypes of GBP ( Table 4 )

Type 1 GBP patients (n= 13), presented with acute biliary peri- tonitis and all were diagnosed preoperatively by imaging. These patients had higher incidence of hypotension (46.2%) and acute kidney injury (84.6%). Prevalence of diabetes (8 patients, 61.5%) in type 1 GBP was higher than that of other two types (P＜0.05). They also had higher postoperative mortality (P＜0.05). For initial management, 7 patients required PCD and 3 underwent emergency surgery. For definitive treatment, 3 patients underwent interval cholecystectomy, including 2 laparoscopic conversions to open procedure. Partial cholecystectomy was performed in 2 pa- tients and bilio-enteric anastomosis was performed in 2 (chole- dochojejunostomy: 1, Roux-en-Y hepaticojejunostomy: 1). Table 5 summarizes the surgical procedure performed in three types of GBP.

Table 5 Surgical management of gallbladder perforation.

Seventy-six patients (66.1%) of type 2 GBP were diagnosed pre- operatively by imaging. US guided PCD placement was done in 31 patients for peri-cholecystic collection. Eighty-five patients (73.9%) underwent interval cholecystectomy. CBD exploration with T-tube drainage was done in 20 (17.4%) patients and another 9 (7.8%) re- quired bilio-enteric drainage (choledochoduodenostomy: 4, Roux- en-Y hepaticojejunostomy: 5). Partial cholecystectomy was per- formed in 14 (12.2%) patients, whereas extended cholecystectomy was performed in 5 (gallbladder malignancy: 3, XGC: 1, tubercu- lar cholecystitis: 1). Laparoscopic cholecystectomy was attempted in 27 patients but could be completed in only 10 (8.7%) patients and the rest required conversion to open surgery.

Patients with type 3 GBP had much longer duration of symp- toms (104.0 ± 6.7 weeks). Eighteen (90.0%) patients with type 3 perforations were diagnosed during surgical exploration. Fif- teen patients (75.0%) underwent open cholecystectomy, 3 patients underwent laparoscopic converted to open cholecystectomy with dismantling of fistula. Two patients (10.0%) underwent extended cholecystectomy, including segmental colectomy in a patient with XGC and segmental colectomy with sleeve duodenal resection in a patient with locally advanced gallbladder malignancy. Five patients underwent CBD exploration with T-tube drainage and 3 underwent choledochoduodenostomy. One patient underwent cholecystectomy with tube duodenostomy. One patient had advanced malignancy and one refused operative intervention and had to be discharged on personal request.

Choledocholithiasis

Forty-six (30.5%) patients with GBP had concomitant choledo- cholithiasis. Eighteen of these patients were managed with ERCP clearance. Twenty-eight patients required intraoperative CBD ex- ploration due to failed ERCP clearance. Majority of patients (20/28) requiring CBD exploration had type 2 GBP.

Malignancy

Among 18 patients with perforation in cases of gallbladder ma- lignancy, 9 were explored of which only 3 underwent definitive surgery with extended cholecystectomy while 6 were found inop- erable on open exploration and trial dissection. The remaining 9 patients had metastasis at time of presentation. Most of these pa- tients had type 2 GBP (16/18) and one had type 1 and one type 3 GBP and fundus was the common site (14/18).

XGC

Twenty-seven (17.9%) patients with GBP had XGC reported on final biopsy with the majority of such patients (26/27) presenting with type 2 GBP and 1 patient presenting with type 3 GBP.

Discussion

GBP is recognized as an uncommon diagnosis, mostly associ- ated with acute cholecystitis, biliary obstruction and pancreatico- biliary malignancy. Incidence vary between 2%-11% in acute chole- cystitis [ 4 , 7 ]. In a national German quality control database of over 45 0 0 0 patients with acute cholecystitis, the incidence of acute perforated cholecystitis was 9.7% [8] .

Ours is the largest tertiary care and referral center of one of the most populous state of north India in the Gangetic plains. This region has one of the highest incidences of gallbladder stone dis- ease and gallbladder malignancy [ 5 , 9-11 ]. In spite of high inci- dence, a number of patients present very late in the course of dis- ease. Kapoor and McMichael [9] found that patients with gallblad- der stone disease in India presented very late for surgical treat- ment, the median duration of symptoms being 34 months. A num- ber of reasons are possible for this. Many of our patients come from small towns and villages. Rural residence is associated with lower literacy rates, poorer socio-economic status and poorer ac- cess to medical care [10] . Initial symptoms of abdominal discom- fort or pain may be managed conservatively without proper diag- nosis. Many patients prefer to use herbal/indigenous medical ther- apy to dissolve gallbladder stones in order to avoid surgery. Pa- tients also delay surgery due to economic reasons. A similar pat- tern is also noticed for gallbladder malignancy, which is endemic in our region [11] . Late presentation and persistent cholecystitis might lead to gallbladder wall inflammation and weakening of gallbladder wall, leading to perforation and increased complica- tions. Large scale education of general public and regional physi- cians is required to reduce morbidity and mortality of gallstone related diseases in our region.

In our series, prevalence of CBD stones in GBP was 30.5%, which is much higher than the prevalence of CBD stones in gallbladder stone disease (8.6%) [12] . Distal obstruction by CBD stone can lead to increased luminal pressure in gallbladder. This can cause pro- gressive distension and compromised blood supply which even- tually leads to necrosis, gangrene and perforation of gallbladder wall [13] . Prevalence of diabetes in our study was 19.9% which is higher than overall prevalence of diabetes in gallbladder stone disease in general population (11.6%) [14] . Incidence of diabetes was highest in type 1 GBP. Earlier studies have also demonstrated higher co-morbidities in patients with type 1 GBP in the form of coronary artery disease, diabetes, malignancy and cirrhosis [ 4 , 15 ]. About 12% of the patients with GBP had associated malignancy in the current series. Prevalence of XGC among GBP in our study was 17.9% which is higher than usual. In a study by Hale et al. [16] , the prevalence of XGC in excised gallbladder specimen in general population in India was 8.8%.

The most common type of perforation was type 2 (76.2%) which is consistent with previous studies [ 17 , 18 ]. Fundus was the most common site of perforation. Derici et al. [1] also found fundus to be the most common site (60%) as it is the least vascular and most distal part of gallbladder.

In our study, there were similar numbers of male and female patients with GBP. Other studies had higher number of male pa- tients associated with GBP [ 1 , 7 , 17 ]. No cause for this high preva- lence in males has been identified. Roslyn et al. [17] found that types 1 and 2 GBP tend to occur in younger age groups whereas type 3 GBP is found more commonly in the elderly patients. Gall- bladder stone related diseases as well as gallbladder malignancies have a very high incidence in our region and are seen at a much younger age compared to that of the Western population [9] .

Clinical signs and symptoms are usually similar to acute un- complicated cholecystitis except for type 1 GBP which may show signs of generalized peritonitis and sepsis. GBP in acute cholecys- titis should be suspected in patients who become toxic or develop sudden clinical deterioration for unexplained reasons [19] .

Imaging plays a very important role in diagnosing and manag- ing GBP. Three sonographic signs are described for GBP. The first is the hole sign, which is the direct visualization of the perfora- tion in the gallbladder wall. Contrast enhanced CT scan is supe- rior to US in depicting the hole sign [20] . The second sign is di- rect demonstration using color Doppler imaging of flow between the lumen of the gallbladder and peri-cholecystic abscess across a defective gallbladder wall [21] . The third sign is the detection of calculi within the peri-hepatic collection, indicating that gall- bladder has ruptured [22] . Sood et al. [2] demonstrated sonological hole sign in 70% of patients by the use of high-resolution scanner. CT is more accurate in showing free intra-peritoneal fluid, peri- cholecystic fluid and abscess along with defect on the wall due to perforation [22] . Magnetic resonance imaging has also been used in the diagnosis of GBP [23] .

Our treatment strategy was based on step-up approach. Patients were initially managed conservatively and US guided PCD place- ment was done selectively for control of sepsis. Patients with bil- iary obstruction underwent ERCP or PTBD. Only 3 patients needed emergency laparotomy (all type 1 GBP patients), while the rest had elective procedures after control of sepsis. Type 3 GBP re- quired repair of fistula along with cholecystectomy [ 24 , 25 ]. All type 3 GBP in our series had cholecysto-enteric fistula. Ibrarullah et al. [26] have described more complex fistula, such as cholecysto- biliary fistula as a type 4 GBP. Kochar et al. [27] suggested con- solidating the various fistulas (i.e. cholecysto-biliary, cholecysto- cutaneous, cholecysto-enteric and cholecysto-hepatic) into the cat- egory of type 3 perforation. Some novel alternatives to surgery include US guided trans-duodenal (or trans-gastric) gallbladder drainage with stenting or trans-papillary endoscopic gallbladder stenting [28] .

Cholecystectomy in GBP is more difficult than a regular chole- cystectomy. Laparoscopic cholecystectomy in GBP has higher con- version rates and longer operative time. In literature, rate of con- version was almost 3 times higher in patients with acute GBP com- pared to that of patients without GBP and the rate of bile duct injury was also 3 times higher in presence of GBP [8] . In our se- ries laparoscopy was attempted in 32 (21.2%) patients but could be completed in only 10 (6.6%) patients, while rest required con- version to open surgery. Due to severe adhesions, partial cholecys- tectomy was done in 22 cases in our series. An algorithm for the management of GBP is proposed ( Fig. 4 ).

Fig. 1. Free (type 1) gallbladder perforation. A: CT scan showing breach in gallbladder wall (white arrow) & peritoneal collection. B: Operative photographs showing breach in gallbladder wall (black arrow).

Fig. 2. CT scan showing contained (type 2) gallbladder perforation. A: Massively enlarged gallbladder with common bile duct stone. B: Fundal perforation with peri-hepatic collection. C: Intra-hepatic, white arrow breach in gallbladder wall. D: Operative photograph showing perforation in gallbladder wall.

Fig. 3. Cholecysto-gastric fistula (type 3) gallbladder perforation. Intraoperative photographs showing fistula before ( A ) and after ( B ) dismantling.

Fig. 4. Algorithm of management of gallbladder perforation (GBP) by step-up approach. US: ultrasound; ERCP: endoscopic retrograde cholangiopancreatography; PTBD: percutaneous transhepatic biliary drainage; CBD: common bile duct.

In our study, a total of 3 patients (2.0%) died due to compli- cations of GBP. A previous study reported mortality as high as 42% [16] . However, advances in critical care, perioperative manage- ment and image guided interventions have reduced the mortality to 7%-16% [ 13 , 29 , 30 ].

The major limitation of the case series includes its retrospec- tive nature. Our long-term follow-up data were not included in this study and only inpatient morbidity and mortality data were analyzed. Another limitation was the lack of comparative group (cholecystitis without perforation).

In conclusion, early diagnosis of GBP using multimodal diag- nostic workup is of paramount importance and requires high in- dex of suspicion. Step-up approach based on patient condition and type of GBP can reduce the morbidity and mortality. Laparoscopic cholecystectomy can be difficult in these patients and patients may require open or partial cholecystectomy.

Acknowledgments

We thank Dr. Animesh Chandra, Science Writer (Institute for Translational Sciences, University of Texas Medical Branch, Galve- ston, USA), for editing assistance.

CRediTauthorshipcontributionstatement

VivekGupta:Conceptualization, Data curation, Formal analy- sis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing.AbhijitChandra:Conceptual- ization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Valida- tion, Visualization, Writing - original draft, Writing - review & edit- ing.VishalGupta:Conceptualization, Data curation, Formal analy- sis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing.RaviPatel:Data curation, Formal analysis, Investigation, Methodology, Software, Writing - original draft, Writing - review & editing.AmitDangi:Data curation, For- mal analysis, Investigation, Methodology, Software, Writing - origi- nal draft, Writing - review & editing.AjayPai:Data curation, For- mal analysis, Investigation, Methodology, Software, Writing - re- view & editing.

Funding

None.

Ethicalapproval

This study was approved by the Ethics Committee of local hos- pital.

Competinginterest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the sub- ject of this article.