The Symptomatic Outcomes of Cholecystectomy for Gallstones

Cholecystectomy is the definite treatment for symptomatic gallstones, and rates are rapidly rising. Symptomatic complicated gallstones are generally treated with cholecystectomy, but there is no consensus on the clinical selection of patients with symptomatic uncomplicated gallstones for cholecystectomy. The aim of this review is to describe symptomatic outcomes before versus after cholecystectomy in patients with symptomatic gallstones as reported in prospective clinical studies and to discuss patient selection for cholecystectomy. Following cholecystectomy, resolution of biliary pain is high and reported for 66–100%. Dyspepsia has an intermediate resolution of 41–91% and may co-exist with biliary pain but may also develop following cholecystectomy with an increase of 150%. Diarrhea has a high increase and debuts in 14–17%. Persisting symptoms are mainly determined by preoperative dyspepsia, functional disorders, atypical pain locations, longer duration of symptoms, and poor psychological or physical health. Patient satisfaction following cholecystectomy is high and may reflect symptom alleviation or a change in symptoms. Comparison of symptomatic outcomes in available prospective clinical studies is limited by variations in preoperative symptoms, clinical presentations, and clinical management of post-cholecystectomy symptoms. When selecting patients with biliary pain only in a randomized controlled trial, 30–40% still have persisting pain. Strategies for the selection of patients with symptomatic uncomplicated gallstones based on symptoms alone are exhausted. For the development of a selection strategy, future studies should explore the impact of objective determinants for symptomatic gallstones on pain relief following cholecystectomy.


Introduction
Cholecystectomy is currently the only definite treatment for symptomatic gallstones. Cholecystectomy rates have been rapidly rising in European countries within the past decades [1,2] and possibly also in pediatric populations [3]. Treatment of gallstones and cholecystectomy generally cause high healthcare costs [4]. High proportions of persistent symptoms and pain following cholecystectomy have previously been reported when reviewing prospective clinical studies [5], indicating impaired clinical guidelines and practice. This review provides an update on the symptomatic outcomes of cholecystectomy for symptomatic gallstones and discusses the selection of patients for surgery.
Gallstones are highly prevalent in general populations of the US and Europe and depend largely on ethnicity, age, and female sex [6]. The natural course of gallstones has been explored in one cohort, which included people with ultrasound screen-detected gallstones and who remained uninformed of their gallstone status. Only 18% developed symptomatic gallstones requiring hospital admission during long-term follow-up, and about half were due to symptomatic uncomplicated gallstones. The conclusion was that the natural course of gallstones is less aggressive than previously anticipated and that the vast majority of prevalent gallstones remain clinically silent and require no further treatment [7]. The current clinical challenge is selecting the right patients with symptomatic uncomplicated gallstones who will benefit the most from cholecystectomy.
The symptomatic complicated gallstones include acute cholecystitis, common bile duct stones, cholangitis, and pancreatitis while the uncomplicated include symptomatic gallstones in the absence of complications [8]. Cholecystectomy is recommended for all complicated gallstones [9][10][11][12][13][14]. In presence of symptomatic uncomplicated gallstones, current clinical practice guidelines have no consensus on which symptoms or patient characteristics should guide selection for cholecystectomy [9][10][11][12]14,15]. A randomized controlled trial of observation versus cholecystectomy concluded that the non-operative strategy caused fewer symptomatic complicated gallstones during follow-up when compared to post-cholecystectomy complications. Although the study proved the feasibility of observation in the presence of symptomatic uncomplicated gallstones [16], long-term follow-up of the population also showed cholecystectomy to be the preferred treatment [17]. Most practice guidelines recommend cholecystectomy for symptomatic uncomplicated gallstones [9][10][11][12]14], while observation is either disregarded [9] or only suggested [13]. The pooled postoperative morbidity following cholecystectomy is reported to be 1.6-5.3%, including bile duct injuries in 0.32-0.52% [18].
Etiologies for post-cholecystectomy symptoms have recently been systematically reviewed and pooled into (1) co-existent diseases such as gastroesophageal reflux, peptic ulcer, or functional disorders; (2) physiological changes including duodenogastric reflux causing gastritis or bile acid malabsorption causing diarrhea; (3) retained or newly formed gallstones; (4) sphincter of Oddi dysfunction; (5) psychological distress; or (6) surgical complications such as common bile duct injury or incisional hernia. Symptoms may debut, but most often they persist [19]. Common bile duct stones may present with symptoms within 6 days to 18 years following laparoscopic cholecystectomy [20]. Gallstones may also be retained or newly formed in remnants of the cystic duct following cholecystectomy or of the gallbladder following a subtotal cholecystectomy. Estimates for complications due to retained or newly formed gallstones following cholecystectomy are reported with wide ranges between studies [19]. Although there are many possible reasons for postcholecystectomy symptoms, a third of patients have no identified causes [21].
Several study limitations may bias estimates of symptomatic outcomes following cholecystectomy in published clinical studies. The risk of bias may be due to either uncontrolled estimates without preoperative symptom assessment [22][23][24][25][26][27] or to retrospective assessments following surgery with a risk of recall bias [28,29]. On the other hand, estimates from prospective clinical studies may be limited by the short duration of follow-up [30][31][32][33]. Both retrospective and prospective studies may have a risk of attrition bias with either high or unreported proportions of patients lost to follow-up [34][35][36] or otherwise incomplete postoperative assessments such as unanswered questionnaires [37].
This review describes symptomatic outcomes before versus after cholecystectomy in patients with symptomatic gallstones. The focus is on declines or inclines in the proportion of patients with symptoms, on the debut of symptoms, and on the determinants of postcholecystectomy symptoms as reported in prospective clinical studies.

Biliary Pain Definitions and Determinants
Symptoms of biliary pain ascribed to gallstones have traditionally been termed "biliary colic" by pioneers in medical practice such as Sir William Osler. The original symptom complex was defined by a sudden debut, of intense and agonizing pain, localized in the right hypochondrium or epigastrium, with projection to the shoulder, and a duration of hours to days or possibly even weeks [38]. Similar but more simple definitions of biliary pain were later accepted in a series of gallstone screening studies in Italy during the 1980-90'ies which included abdominal pain during the last five years, with a duration of more than 30 min, and localized in the right hypochondrium and/or epigastrium [39][40][41]. Most definitions of biliary pain share similarities in regard to pain localization and duration, but may also include nausea or vomiting [42,43], not being relieved by bowel movements [42,44], a possible association to food ingestion [42], and forcing one to stop all current activity, to lie down, or to take analgesics [45,46]. All these biliary pain definitions rely on expert opinions or consensus statements rather than empirical evidence (Table 1).
In studies exploring gallstone prevalence in general or clinical populations, it has generally been challenging to associate gallstones with a specific symptom or complex. Although more or less strict definitions of biliary colic have been associated with gallstone prevalence when compared to controls without gallstones, the diagnostic accuracy and predictive test values are very low [47].
Only a few determinants for the development of symptomatic gallstones have been identified when studying gallstone cohorts through long-term follow-up. In the cohort including persons uninformed and unaware of the presence of gallstones at baseline as described above, determinants for the development of symptomatic uncomplicated gallstones were pain localized in the epigastrium, with a duration of hours, of moderate to extreme intensity, and with the need for analgesics. Pain at night determined symptomatic complicated gallstones only [48]. These findings share similarities with the originally described "biliary colic", except for pain projection. Pain projection has only been associated with gallstones in patients with emergency admissions in a cross-sectional study and with incident gallstones identified through multiple ultrasound examinations in a populationbased cohort study [49,50].
Determinants for the development of symptomatic gallstones of a more objective character have also been identified in cohort studies. These include younger age, female sex, multiple gallstones [7,51], gallstone size of 1 cm or above [7,51,52], higher body mass index (BMI) [51,53], gallbladder polyps, tobacco smoking, and a number of comorbidities [51].
The mechanisms involved in subjective symptom generation are not fully understood but are believed to include a gallstone passage to the duodenum or an obstruction of the gallbladder or bile ducts causing biliary distention [38,54]. This should then activate visceral sensory neurons and create a sensation of pain [55]. In support of these long-held traditional beliefs, experimental distention of the gallbladder has been shown to cause pain in the right hypochondrium or epigastrium in patients with acute cholecystitis [56]. Table 1. Definitions of biliary pain in different types of studies.

+ not relieved by bowel movements
Clinical studies Ros and Zambon (1987) [42] Steady pain in the right hypochondrium/epigastrium Nausea or vomiting Duration of at least one hour Associated or not with food ingestion Unrelated to bowel movements Unassociated with discomfort at urination Heaton et al. (1991) [45] Abdominal pain attacks during the last year Duration of 30 min or more Localized in the upper abdomen Forces one to stop activities, lie down, or take analgesics Martinez de Pancorbo et al. (1997) [46] Abdominal pain in the right hypochondrium and/or epigastrium Forces one to lie down or take analgesics

Limitations of Studies Exploring Symptomatic Outcome
Most prospective clinical studies exploring symptomatic outcomes of cholecystectomy report results from unselected patient populations including both symptomatic complicated and uncomplicated gallstones. The latter may include symptoms of unspecified pain or other abdominal complaints, such as dyspepsia. Most of these studies of unselected populations report the presence of either biliary or unspecified pain at baseline [42,43,, and only a few include patients with mostly dyspepsia [58]. Exploration of selected study populations with dyspepsia as the main complaint at baseline has been performed in only a few studies [80][81][82], and one study has performed a subgroup analysis of patients with dyspepsia [43].
The identification of underlying diseases for post-cholecystectomy symptoms requires diagnostic examinations [19]. Most available studies do not report the performance of examinations at clinical follow-up or the identification of causes for post-cholecystectomy symptoms. Few studies report the performance of post-cholecystectomy cholangiography, endoscopy, ultrasound, or blood samples [42,57,59,60,62,63,65,66,69,71,83,84]. Even fewer studies report treatments for identified underlying diseases causing post-cholecystectomy symptoms before the final clinical assessments or how they managed patients in the analysis of the symptomatic outcome for cholecystectomy [42,57,66,83,84].
Symptom resolution is determined by a preoperative history of cholecystitis [70] and a higher age [62]. Persisting symptoms are determined by preoperative dyspeptic symptoms [63,70], symptom duration over 6 months, gastritis [93], poor self-rated health [70], higher Gastrointestinal Symptom Rating Scale, higher Hopkins Symptom Checklist for assessment of anxiety and depression [81], higher trait anxiety [43], higher Psychological Symptom Score [93], and an American Society of Anesthesiologists (ASA) score of III-IV [70]. Unsatisfying or unsuccessful cholecystectomy outcome is determined by flatulence, duration of days or more for pain episodes [59], symptom duration over 6 months, and poor self-rated health [70]. Both younger age and an age above 55 years have been associated with persisting symptoms [59,62] (Table 2). Female sex has been associated with postoperative symptoms in a retrospective study only [24]. Table 2. Determinants of postoperative symptoms identified in prospective clinical studies.
Definitions of dyspepsia have been without consensus and changed much throughout time. Flatulent dyspepsia was once a multi-symptomatic syndrome thought to be caused by gallstones but was also found to be due to other disorders [99]. The more recently described functional dyspepsia is defined according to the Rome IV criteria as one or more of bothersome postprandial fullness, early satiation, epigastric pain, or burning in the absence of structural disease at endoscopy [100]. These varying and broad definitions of dyspepsia make it challenging to assess changing symptoms between studies.
In studies of populations where dyspepsia is the main complaint at baseline, a minor resolution of dyspepsia for 17-28% has been reported [80,82]; however, populations are small and highly selected. One study including a larger population undergoing elective laparoscopic cholecystectomy found a high incline of 150% in patients reporting dyspeptic symptoms without biliary symptoms following cholecystectomy (preoperative 14% versus postoperative 35%) ( Table 3). In comparison, symptom relief in patients with preoperative biliary symptoms without or with dyspeptic symptoms was 88% and 77%, respectively [43]. Dyspeptic symptoms may thereby develop in unselected patient populations. Mechanisms of dyspeptic symptom development following cholecystectomy have been suggested in experimental studies and include duodenogastric reflux, histological gastritis [71,[101][102][103], and impaired postprandial gastric emptying [104].
It seems that biliary pain and dyspeptic symptoms may co-exist, and that cholecystectomy resolves dyspeptic symptoms in unselected populations where most patients report pain at baseline. However, dyspepsia may also incline or debut following cholecystectomy. Resolution of dyspeptic symptoms in populations where they are the main preoperative complaint seems inconsistent based on the currently available studies. Larger studies exploring populations with functional dyspepsia or, alternatively, analysis of subgroups should be performed in the future. Until further, cholecystectomy is not indicated for the presence of dyspepsia only.
Diarrhea is believed to be due to continuous bile excretion following cholecystectomy, which may also cause bile acid malabsorption [107]. Another proposed mechanism includes a shortened gut transit due to accelerated colonic passage [108]. More recently, a decreased microbial diversity in patients with post-cholecystectomy diarrhea has been demonstrated when compared to patients without diarrhea following cholecystectomy or to healthy controls [109,110].
Estimating the prevalence of post-cholecystectomy diarrhea is challenging. Bile acid malabsorption is diagnosed with a SeHCAT (selenium homocholic acid taurine) test that shows low absorption of radiolabelled bile acids. Tests are usually performed several years following cholecystectomy [111,112]. A retrospective study with a follow-up of four years has shown diarrhea to debut in 12% [108]. Prospective studies with multiple postoperative assessments report inclining proportions of patients with diarrhea within the first month following cholecystectomy, only to decline again during the following months [93,96,106,113]. Available prospective studies have a maximum follow-up duration of 12 months, which, therefore, may exaggerate the incidence of diarrhea [66,92,93,96,105,106]. A more exact prevalence of post-cholecystectomy diarrhea has yet to be explored in future clinical studies. Such studies should have longer and more systematic follow-ups of an entire cholecystectomy cohort including both symptom assessments and functional tests.

Other Functional Disorders Outcome
Constipation is reported with minor declines of 0-47% following cholecystectomy [66,74,93,96,105]. Mechanisms of changing bowel function following cholecystectomy, including increased colonic transit, have also been suggested to correct constipation [105,108]. However, changes in constipation following cholecystectomy are of a minor magnitude and may, until further notice, also be ascribed to random changes during follow-up.
Abdominal bloating or distention is reported to resolve for 21-73% of patients following cholecystectomy in unselected populations [59,61,65,66,74,76,97]. Bloating or distention is part of irritable bowel syndrome according to the Manning criteria [114]. It is also associated with irritable bowel syndrome and other functional bowel disorders according to the Rome IV criteria [115]. Currently, no studies have explored symptom outcomes following cholecystectomy in populations with mainly functional gastrointestinal disorders.
A recent large prospective observational study found that patients with functional dyspepsia and/or irritable bowel syndrome at baseline were significantly less likely to be pain-free six months after cholecystectomy when compared to those without (41% vs. 64%). Biliary colic was effectively resolved following cholecystectomy, independently of concomitant functional dyspepsia and/or irritable bowel syndrome [79]. Other studies have also found that patients with functional gastrointestinal disorders are less likely to resolve symptoms [85] and that patients with irritable bowel syndrome have fewer improvements in gastrointestinal quality of life scores following cholecystectomy [116]. Patients with irritable bowel syndrome have an increased risk of cholecystectomy that is not due to an increased risk of gallstones [117].
Currently, it seems that patients with mainly functional gastrointestinal disorders should be carefully selected for cholecystectomy based on the presence of biliary pain if offered surgery at all.

Selection of Patients for Cholecystectomy
The high rates of biliary pain resolution reported in prospective clinical studies indicate a true benefit of cholecystectomy, and it appears beneficial to select patients with biliary pain for cholecystectomy and, vice versa, refrain from surgery in its absence. A recent large randomized controlled trial, the SECURE trial, has challenged this assumption. Patients with uncomplicated gallstones and abdominal pain were selected for cholecystectomy through allocation to either a restrictive or a usual care strategy. In the usual care strategy, the selection was at the discretion of the physician. In the restrictive strategy, cholecystectomy was advised only to those patients who fulfilled a five-criteria symptom complex including (1) severe pain attacks, (2) pain duration of 15-30 min or longer, (3) pain localized in the epigastrium or right upper quadrant, (4) pain radiating to the back, and (5) a positive response to simple analgesics. If cholecystectomy was deferred accordingly at the first clinical assessment, the need was reconsidered during follow-up assessments. At one-year follow-up, only 72% versus 98% of patients allocated to the restrictive strategy were treated per protocol when compared to the usual care strategy. Thereby, a large cross-over of patients to cholecystectomy was seen in the restrictive strategy arm. The primary outcome was the proportion of patients being pain-free at follow-up, which was no different for the restrictive and usual care strategies (64% versus 63%, respectively). The only identified differences were fewer cholecystectomies in the restrictive strategy. The study concluded suboptimal pain reduction for both strategies [118]. Parallel with the SECURE trial, a prospective cohort study, the Success trial, developed a prediction model for clinically significant pain reduction following cholecystectomy, which was validated in the study population of the SECURE trial [78]. The identified determinants for pain reduction have been referred to in this present review (see Section 5 and Table 2).
The SECURE trial demonstrated the clinical challenges of withholding patients from a protocolled treatment and, even more, it challenged the definitions and predictive value of biliary pain. Despite protocolled attempts to select patients for cholecystectomy based on strict biliary pain definitions, the proportions of patients with persisting pain were unacceptably high.
Patient selection for cholecystectomy relying only on patient-reported symptoms has been clinical practice for a long time but is suboptimal if the aim is a pain-free state following surgery. Although prediction scores for symptomatic outcomes of cholecystectomy are of limited availability and even fewer are validated [78], available studies seem to have exhausted the exploration of preoperative patient-reported symptoms for post-cholecystectomy outcomes. Future clinical studies should focus on exploring more objective variables for symptom resolution following cholecystectomy. Determinants for the development of symptomatic gallstones as identified in previous larger cohort studies (see Section 2) are candidate explorative variables for future studies [7,[51][52][53]. Due to a consensus on the need for cholecystectomy in the treatment of symptomatic complicated gallstones in clinical practice guidelines, future studies should confine to populations of symptomatic uncomplicated gallstones. If determinants for symptom resolution following cholecystectomy are identified, future randomized controlled trials may explore selection strategies yet further to improve management strategies for the treatment of symptomatic uncomplicated gallstones.

Conclusions
Studies exploring symptomatic outcomes following cholecystectomy mostly include heterogeneous populations with a variety of clinical gallstone presentations. Further, they fail to report a diagnostic workup or treatment for post-cholecystectomy symptoms. This challenges the comparison of symptomatic outcomes between studies. Although biliary pain seems to resolve following cholecystectomy in most patients, dyspeptic symptoms may co-exist with pain and may both resolve or develop following cholecystectomy. Patient satisfaction following cholecystectomy is high, which may reflect a symptom alleviation or, alternatively, a change in symptoms. Persisting complaints following surgery are mostly determined by preoperative functional disorders, atypical pain locations, long duration of symptoms, frequent and long pain episodes, and poor psychological or physical health. Even when the selection of patients for cholecystectomy is restricted to the presence of biliary pain, unacceptably high proportions of patients have persisting pain. Biliary pain is ill-defined and mostly based on consensus statements, but it currently remains the only criterion for cholecystectomy in the presence of symptomatic uncomplicated gallstones. The exploration of patient-reported symptoms only for the prediction of cholecystectomy symptomatic outcome has been exhausted. Future prospective clinical studies should explore the impact of objective determinants for symptomatic gallstones on pain resolution following cholecystectomy in patients with presumed symptomatic uncomplicated gallstones.