Searching for low phospholipid associated cholelithiasis among patients with post‐cholecystectomy biliary pain

Low phospholipid associated cholelithiasis (LPAC) is associated with variants of the adenosine triphosphate‐binding cassette subfamily B, member 4 (ABCB4) gene and is characterized by reduced phosphatidylcholine secretion into bile, impairing the formation of micelles and thus exposing bile ducts to toxic bile acids and increasing cholesterol saturation. LPAC is present in 1% of patients with gallstones and post‐cholecystectomy pain is common in this group. LPAC is an under‐appreciated cause of post‐cholecystectomy pain. The aim of this study is to assess a cohort of patients with post‐cholecystectomy pain to identify those with clinical features suggesting that further investigations for LPAC would be beneficial.


Introduction
Post-cholecystectomy pain is a common clinical problem, occurring in up to 33% of patients. 1Liver function tests (LFT) if performed at the time of pain or shortly afterwards may be transiently abnormal, confirming a bile duct problem.Causes of post-cholecystectomy pain include complications of the surgery, retained common bile duct stones and biliary spasm. 2 Low phospholipid associated cholelithiasis (LPAC) is a recently described clinical syndrome that should be added to this differential diagnosis.LPAC a clinical syndrome characterized by reduced secretion of phosphatidylcholine into bile and associated with variants of the adenosine triphosphate binding cassette, subfamily B, member 4 (ABCB4) gene. 3ABCB4 is a flippase that transports phosphatidylcholine from the hepatocyte into the bile in the biliary canaliculus. 4Defects in this transporter protein reduce the concentration of phosphatidylcholine in the bile.Phospholipids are 25% of bile dry weight and their primary function is to protect cell membranes from toxic hydrophobic bile acids by forming mixed micelles. 5Low phospholipid bile levels also destabilize micelles increasing biliary lithogenicity and promoting cholesterol crystallization. 6Functional variants of the ABCB4 gene have been described; phenotypic expression is variable and often overlapping disease manifestations are seen in families and with disease progression in an individual. 7,8evere reductions in biliary phospholipid secretion are usually associated with homozygous variants of ABCB4 and present in infancy as Progressive Familial Intrahepatic Cholestasis type 3 (PFIC 3) with cholestatic liver disease progressing to cirrhosis and liver failure. 9Less severe phenotypes associated with heterozygous ABCB4 variants include adults with unexplained chronic liver disease, drug induced liver disease and intrahepatic cholestasis of pregnancy (ICP). 7LPAC phenotypes have mild to moderately severe clinical illness often associated with heterozygous ABCB4 variants.LPAC overlaps with other phenotypes; ICP occurs in 42% of women with a history of pregnancy, chronic cholestatic liver disease may be present at the initial diagnosis or cirrhosis may develop many years after the initial presentation. 3,10There is often a family history of cholelithiasis at an early age or a family history or of one or more of the other ABCB4 variant phenotypes.
The term LPAC was introduced by Rosmorduc et al. in 2003 to describe a group of 32 patients referred for ABCB4 gene analysis. 11hese patients all had symptomatic cholelithiasis with at least one other criterion suggesting an ABCB4 variant: age less than 40 years, recurrent symptoms after cholecystectomy, typical ultrasound findings, family history of cholelithiasis, and clinical history of ICP.The patients were compared with controls and three independent clinical features associated with an ABCB4 variant were identified: recurrence of symptoms after cholecystectomy, typical imaging features, and onset of symptoms under 40 years of age.The presence of at least two of the three clinical features that predict the presence of an ABCB4 variant has been accepted as the clinical definition of LPAC syndrome.This is a very broad definition and includes all patients with post-cholecystectomy pain under the age of 40 years.At the time, the criteria were recognized as sensitive but not specific.Diagnostic certainty increases if typical imaging findings are seen on ultrasound or if genotyping identifies a known variant.2][13][14] Typical ultrasound findings in LPAC include intrahepatic hyperechoic foci, sludge and microlithiasis. 11The intrahepatic hyperechoic foci have also been subsequently described in patients with ABCB4 variants who are asymptomatic, and those presenting with ICP or unexplained cirrhosis. 15,16s more cases of LPAC have been reported the diagnostic criteria have evolved and the clinical features of the syndrome have been expanded to include: family history of gallstones, common bile duct stones at initial presentation, onset of symptoms at the end of or following pregnancy, persistently raised gamma-glutamyl transferase, response to ursodeoxycholic acid treatment and a personal or family history of ICP or hepatobiliary cancer. 17LPAC has a spectrum of severity of clinical features from the most severe with cholestatic liver disease progressing to cirrhosis or dilated intrahepatic ducts with large stones requiring surgery for recurrent cholangitis to much milder disease presenting with postcholecystectomy pain but normal imaging and liver function.
LPAC is often under recognized and under diagnosed by clinicians and can be the cause of post-cholecystectomy pain in up to 23% of younger patients. 13Diagnosis is important as identification of LPAC and subsequent treatment could avoid recurrent symptoms and long-term complications.The aim of this study is to assess a cohort of patients with post-cholecystectomy pain to identify those with clinical features suggesting that further investigations for LPAC would be beneficial.

Methods
A retrospective review was performed of the medical records of all patients, younger than 40 years of age, who had a cholecystectomy for symptomatic gallstones at The Royal Melbourne Hospital between January 2016 and December 2017.Patients with postcholecystectomy pain admitted to hospital or presenting to outpatients were identified, data was collected with a standard case report form.Recorded variables included date of birth, gender, height, weight, date of recurrent symptoms, liver function tests during or shortly after an episode of pain and interval liver function tests, results of imaging including abdominal ultrasound and magnetic resonance cholangiopancreatography (MRCP).Other data recorded were a personal or family history of ICP and a family history of gallstones or hepatobiliary malignancy.Qualitative and quantitative variables are expressed as numbers (percentage) and median (range), respectively.Continuous variables were compared with Wilcoxon rank-sum tests or Kruskal-Wallis tests when more than two groups were compared.Quantitative variables were compared using Fisher's exact tests.A difference was considered statistically significant when the P < 0.05.The R software was used for all analysis. 18Approval for the study was obtained from the Melbourne Health Human Research Committee (QA2018013).

Results
Between 1 January 2016 and 31 of December 2017, 726 patients underwent a cholecystectomy for symptomatic gallstones.258 patients (35.5%) were under the age of 40 and their medical records were reviewed.The median age of the 258 patients was 32 years (range 17-40).197 (76%) were female and the median body mass index (BMI) of the cohort was 28.5 (range 17.1-53.3).The median outpatient post-operative clinic follow-up time was 19 days (range 3-1482).
Post-cholecystectomy abdominal pain was reported by 50 patients (19.4%); 41 patients were managed as hospital inpatients and nine as outpatients.Clinical features of each diagnostic group are shown in Table 1.Post-operative complications occurred in seven patients, a median of 1 week after the surgery, and the causes included three patients with post-operative collections, two patients with retained stones, one patient with a bile leak and one patient who had a blocked T tube.Liver function tests were abnormal in six of seven patients that were tested and cross-sectional imaging with CT, CT intravenous cholangiography (CT IVC), and MRCP was performed.The three patients with post-operative collections were managed conservatively with IV antibiotics, the patient with the bile leak underwent a laparoscopic washout and drain tube insertion, the two patients with retained stone underwent ERCP and the blocked T tube was unblocked and repositioned.
Residual self-limiting post-operative abdominal pain occurred in 12 patients presenting a median of 2.4 weeks post-cholecystectomy. LFTs were abnormal in seven of the eight patients in whom the tests were performed.Investigations performed included US, CT, and CT IVC, and the pain resolved spontaneously in all patients and did not recur.Other non-biliary causes of pain, shown in Fig. 1, were diagnosed in 23 patients with symptoms beginning a median of 31.4 weeks post-operatively.
Recurrent right upper quadrant pain, deranged liver function tests or pancreatitis occurred in eight patients a median 35.6 weeks after cholecystectomy.Two patients presented with recurrent pancreatitis due to alcohol and cystic fibrosis.One patient had recurrent common bile duct stones secondary to chronic haemolysis.
Out of the 50 patients with post-cholecystectomy pain five remaining patients (10%) had undiagnosed episodes of right upper quadrant pain at the time of the review.The patients were all females aged 19, 23, 24, 31 and 37 years of age, presenting 5, 11, 48, 6 and 11 months post-surgery, respectively.Liver function tests were abnormal with the episodes of pain in four patients and not done in the fifth.Repeated imaging excluded a recurrent stone or post-operative complication and the median bile duct diameter was 5 mm (range 3-6 mm).Abdominal US performed on all four patients did not identify diagnostic features of LPAC although the requests did not specifically alert staff to the possibility of a diagnosis of LPAC.These 5 patients all have a possible diagnosis of LPAC.
Table 1 outlines the patient characteristics classified by type of pain: post-operative complications, residual post-operative pain, other causes and possible LPAC.LPAC patients tended to have a lower median age 19 than the other groups, however, this difference did not reach statistical significance.The LPAC group had a similar median BMI to the other subgroups.Figure 2 shows the Kaplan-Meier plot for the estimated probability of the cause of pain based on the time of onset of the pain for the four different groups.The median time to onset of recurrent pain was longer for LPAC (26.6 weeks) and other causes (31.4 weeks) compared to post-operative complications and residual pain (0.9 and 2.4 weeks, respectively), Log-Rank test P < 0.01.
Of the 50 patients who had abdominal pain post-cholecystectomy only one patient had LPAC considered as a diagnosis.This patient had recurrent pancreatitis for other reasonscystic fibrosis.None of the five patients identified in this study who met the criteria  suggestive of LPAC had the diagnosis of LPAC considered by their treating clinicians.From the 258 patients, there were no imaging findings suggestive of LPAC consisting of evidence of intrahepatic macro or micro-lithiasis or intrahepatic bile duct dilatation.However, in all 258 patients including the five possible LPAC patients, no targeted liver ultrasounds were performed looking specifically for these findings or to confirm the diagnosis of LPAC.In the absence of US findings, the diagnosis of LPAC would be supported by secondary findings such as family history of cholecystectomy before the age of 50, a history of intrahepatic cholestasis of pregnancy or response to ursodeoxycholic acid treatment.Family history of gallstones was only documented in 33 of 258 (12.8%) of cases.Of those patients who had family history documented, 8 of 33 (25%) had first-degree relatives with gallstones and with only three of those patients having documented first-degree relatives with gallstones under the age of 40.An obstetric history was obtained in 69 of 197 (35%) of female patients.None of those female patients had a history of ICP.In all five patients with possible LPAC, a family history of gallstones or a history of ICP was not elicited.

Discussion
This study has identified undiagnosed recurrent biliary pain in 5 of 50 (10%) patients with post-cholecystectomy pain and an overall rate of 5 of 258 (1.9%) in cholecystectomies performed on patients under 40 years old.This presentation and other clinical features raise the possibility of undiagnosed LPAC in this cohort.
The recognition and diagnosis of LPAC is important.Patients with LPAC suffer from repeated episodes of biliary pain leading to hospital presentations and admission.Up to a third of patients experience complications necessitating high-risk procedures like endoscopic retrograde cholangiopancreatography (ERCP) for choledocholithiasis and cholangitis and even major liver surgery, long-term complications including secondary biliary cirrhosis and hepatobiliary cancer or resulted in prematurity and fetal distress associated with intrahepatic cholestasis of pregnancy. 12Identification of LPAC and subsequent treatment could avoid recurrent symptoms and progression leading to long-term complications.Observational studies have suggested that ursodeoxycholic acid is effective in reducing the frequency and severity of abdominal pain and associated complications. 12,13In addition, confirming the diagnosis will also reduce the need for further investigations of episodes of biliary pain particularly invasive investigations and treatment.
A stone in the bile duct is a common cause of biliary pain presenting several months after a cholecystectomy and is readily treated at ERCP. 20 Initial investigations should include appropriate imaging of the bile duct including US or MRCP or CT IVC.If the bile duct is clear the differential diagnosis should include consideration of sphincter of Oddi dysfunction (SOD), LPAC, and biliary spasm secondary to narcotic medication.
Recurrent episodes of biliary pain with no cause found on standard investigations is a challenging clinical problem.There were five such patients in this series and possible diagnoses include SOD and LPAC.These two diagnoses have similar clinical characteristics.Patients are diagnosed at a similar age, 38 and 36 years, females predominate 93% and 77%, and the diagnosis is often made well after the cholecystectomy, 4 and 6 years respectively for SOD and LPAC. 17,21Furthermore, the diagnostic criteria for SOD and LPAC overlap.In Dong's series of 308 patients with LPAC, 95% presented with biliary pain, the pain recurred after cholecystectomy in 86% and LFTs were transiently abnormal in 47% of these patients. 17he four patients with episodes of biliary pain with transiently abnormal LFTs and non-dilated bile ducts satisfy the diagnostic criteria for functional biliary sphincter disorder, previously called SOD type 2. The fifth patient with biliary pain and no evidence of abnormal LFTs would have previously satisfied a diagnosis of SOD type 3, however, this diagnosis has been abandoned.22 In our series, the features that increase the probability of a diagnosis of LPAC are the young age at cholecystectomy, 19, 23 and 24 years in three patients, and choledocholithiasis at presentation in one.All five patients should be further investigated with a targeted liver ultrasound and a detailed family history.Typical ultrasound features of intrahepatic microlithiasis, intrahepatic hyperechogenic foci and intrahepatic biliary stricturing and dilatation help confirm the diagnosis of LPAC.23 Many of these findings are subtle and difficult to detect, hence a repeat targeted liver ultrasound with expert radiologists examining for these LPAC findings should be requested.Our group has previously demonstrated that less than 10% of ultrasounds performed in the community detected ultrasound features of LPAC identified by expert radiologists.19 Positive ABCB4 genotyping would also confirm the diagnosis, however, it is expensive, difficult to obtain and only positive in up to 57% of patients.12,24 If imaging is normal and genotyping is unavailable, then other features in the clinical presentation may support the diagnosis of LPAC.Our study suggests that patients with LPAC experience recurrent pain after 6 months post-cholecystectomy. Abnormal LFTs with right upper quadrant pain or pancreatitis are suggestive of a biliary cause and LFTs are often abnormal with episodes of biliary colic in LPAC.17 Other supporting clinical features of LPAC are a normal BMI, an absence of cholecystitis and a family history of hepato-biliary cancer.A family history of gallstones especially in first-degree relatives as well as a history of intrahepatic cholestasis of pregnancy are also associated with LPAC.24 These important facets of clinical history are not usually elicited in patients who present with symptomatic gallstones and help differentiate from alternative diagnoses like SOD type II.In our cohort, a family history of gallstones was only documented in 33 of 258 (12.8%) of cases while an obstetric history was only obtained in 69 of 197 (35%) of female patients.Hence, the diagnosis of LPAC can be missed by clinicians.
None of the five possible LPAC patients had the diagnosis of LPAC considered by their treating clinicians and hence family history and obstetric history was not elicited.All five patients had normal ultrasounds for investigation of recurrent biliary pain.
There were limitations to this study.This was a retrospective chart review and hence there was incomplete data and the accuracy was limited by the documentation from the clinicians.The diagnosis of LPAC in each of the five patients could not be confirmed through this retrospective review as LPAC was not considered by their treating clinicians.Hence, a targeted ultrasound looking for typical LPAC findings of microlithiasis or hyperechogenic foci was not requested nor did the clinicians elicit further history such as family history or a history of intrahepatic cholestasis of pregnancy.Furthermore, with limited follow-up to identify further episodes of recurrent pain and no trial of ursodeoxycholic acid, these five patients have only presumed LPAC diagnoses.
As this is a retrospective study, ongoing follow-up was not performed.We plan to invite the LPAC patients identified in our audit for further workup for LPAC and to participate in our randomized control trial assessing the efficacy of ursodeoxycholic acid to improve pain and quality of life among patients who suffer from post-cholecystectomy pain with features consistent with LPAC (Clinical trial registration number: ACTRN12621000450819).

Conclusion
LPAC is an under recognized cause of post-cholecystectomy painup to 2% of patients undergoing a cholecystectomy under the age of 40 could have LPAC.Clinicians should take a family history and obstetric history to alert them to the diagnosis of LPAC.Identification of potential LPAC patients is vital as they may benefit from genetic testing and ursodeoxycholic acid treatment to prevent recurrent symptoms and long-term complications.

Table 1
Clinical features † †Table 1 excludes three patients with explained causes of recurrent right upper painalcohol related pancreatitis, cystic fibrosis-related pancreatitis and recurrent stones from chronic haemolysis.BMI, body mass index; CT IVC, computed tomography intravenous cholangiogram; CT, computed tomography; LFT, liver function test; LPAC, low phospholipid associated cholelithiasis; MRCP, magnetic resonance cholangiopancreatography; US, ultrasound.