Acalculous cholecystitis with gallbladder necrosis in a patient presenting without abdominal pain

Abstract

An 83-year-old man with a history of chronic myelogenous leukaemia in remission maintained with bosutinib presented with new-onset fevers. He denied pain and had no other focal symptoms. Ultrasound imaging revealed mild gallbladder wall thickening. Non-contrasted CT revealed right upper quadrant inflammation of indeterminate source. The diagnosis of acalculous cholecystitis was made on the third day when a CT with oral contrast demonstrated a remarkably inflamed biliary tree. The gallbladder was surgically removed and found to be necrotic. The case highlights an unusual presentation for a well-known condition. Both ultrasound and CT have limited diagnostic sensitivity for acalculous cystitis. This case adds to existing literature to support development of acalculous cholecystitis in non-critically ill patients. Clinicians should maintain awareness of this condition among patients presenting to the hospital or clinic with abdominal pain. Careful discussion with radiology and surgery is indicated to guide diagnostic testing when initial imaging results are indeterminate.

Background

Acute acalculous cholecystitis (AAC) makes up 5%–10% of cases of cholecystitis.¹ The condition has an incidence of 0.12% for the entire population but an incidence of 0.2%–0.4% among critically ill patients.^{2 3} Approximately 80% of cases occur in male patients and a majority are age 50 or older.² The condition has a peak incidence in the fifth and eighth decade of life.²

AAC typically has a presentation that is difficult to distinguish from that of acute calculous cholecystitis (ACC).⁴ For patients who are able to provide history, acute onset right upper quadrant abdominal pain and a positive Murphy sign on exam are both quite common.⁴ Both AAC and ACC may be confirmed with CT, ultrasound (US) and hepatobiliary iminodiacetic acid (HIDA) scan. AAC has traditionally been thought to occur in critically ill patients and those who had undergone prolonged fasting, received total parental nutrition or those who were in shock.^{1 5} However, new evidence has increasingly demonstrated that AAC may occur in patients presenting from the community and patients may not be severely ill at the time of presentation.^{1 6} Here, we present a case of AAC in a patient with chronic myelogenous leukaemia (CML) who presented fever but did not have abdominal pain or abnormal imaging at the time of presentation.

Case presentation

An 83-year-old man with a medical history significant for CML in remission with oral bosutinib therapy, hypertension, benign prostatic hyperplasia and a distant history of prostate cancer that was in remission presented to the emergency department due to new onset fevers. He was admitted 6 months earlier due to abdominal pain with fevers. At that time, blood cultures grew Escherichia coli and an abdominal US revealed gallbladder wall thickening, but did not meet clinical or radiographic criteria for cholecystitis. Afterwards, he was seen as a clinic patient by the surgery team and was not felt to need a cholecystectomy.

He developed fevers the day before presenting to our emergency department. He had no other focal symptoms. He reported feeling quite well aside from recording fevers. He did acknowledge intermittent chills and body aches which seemed to correlate to his fevers. He denied abdominal pain, nausea and vomiting. He had a fever to 102°F at presentation and mild tachycardia but was normotensive. He had a leucocytosis of 25 000 cells/mm³ without bandaemia and a lipase of 16. He had no abdominal tenderness and a negative Murphy sign. Due to suspected sepsis with unclear source, he was admitted to the internal medicine service.

Over the following 2 days, he had persistent fevers but his fever curve trended down slightly from its peak. He continued to deny abdominal pain or other symptoms. Due to the vague nature of his presentation, advanced further workup was initiated.

Investigations

He underwent US imaging of the gallbladder on the evening of admission (figure 1). It revealed a small degree of gallbladder thickening, but this was actually better than it had looked 6 months prior. The US did not show stones or biliary dilation and the sonographic Murphy sign was also negative.

Figure 1.

(Investigations) Right upper quadrant ultrasound supine view showing gallbladder wall thickening without evidence of cholelithiasis.

Two days after admission, a non-contrasted CT scan was ordered to assess the chest and abdomen (figure 2). The scan revealed inflammation of the right upper quadrant but was indeterminate as to the source of inflammation. Radiology felt a microperforation of the duodenum was possible and that his may have been causing associated inflammation in a focal area of bowel loops and extending into the biliary system.

Figure 2.

(Investigations) CT with oral contrast demonstrating an inflamed gallbladder with pericholecystic fluid. Box containing inflamed gallbladder with thickened wall and pericholecystic inflammation.

The following day, oral contrast was provided and a repeat CT scan performed. The scan did not show contrast extravasation from the bowel, so perforation was considered unlikely. However, the gallbladder and biliary tree appeared to be remarkably inflamed on the repeat scan, though no extrahepatic biliary tract dilation was observed. Because the pericholecystic fat was inflamed, the reader felt that there was not only gallbladder thickening but also notable pericholecystic fluid. Pancreatic parenchyma did not show evidence of inflammation. In this context, he felt the inflammation involved the entire right upper quadrant and was the cause of bowel inflammation in that area.

Differential diagnosis

Before the results of the second CT scan were available, the differential diagnosis would have included cholecystitis, cholangitis, appendicitis, bowel perforation or obstruction, and mesenteric ischemia. Urinary sources of sepsis were also possible and could have included pyelonephritis or prostatitis. Neoplastic infiltration of the gallbladder was also an important consideration on the differential. Pancreatitis was ruled out due to the lack of inflammation on CT, normal lipase and lack of epigastric abdominal pain.

Treatment

The patient was initially started on broad spectrum antibiotics with vancomycin, cefepime and metronidazole. Vancomycin was removed after blood cultures returned without growth.

Following the second CT scan, surgery evaluated the patient and the patient consented to surgical removal of the gallbladder. There was no evidence of hepatic duct obstruction suggesting Mirizzi syndrome. Neoplastic infiltration was not observed and metastatic involvement was excluded.

Outcome and follow-up

Following cholecystectomy, pathological analysis revealed a necrotic gallbladder. The wall was mottled with green-tan areas possibly representing abscess-like activity and mottling along the mucosal lining consistent with a necrotising process. One reactive lymph node was identified. The mucosal aspect was dark-green and no discrete lesions were identified. No stones or gallbladder contents were present. The patient was transferred to the surgical team and had a quick recovery. His fevers and leucocytosis were resolved within 2 days postoperatively.

Discussion

AAC is defined as gallbladder inflammation in the absence of gallstones or cystic duct obstruction while ACC is defined by the presence of an obstructing gallstone in the cystic duct.¹ Gallstones form due to decreased solubility of cholesterol and bile salts in bile; cholecystitis develops after a stone becomes stuck in the cystic duct. The pathophysiology of AAC, meanwhile, is thought to be related to bile stasis and ischemia, often—but not always—in the setting of prolonged fasting.¹ It is thought that bile stasis contributes to the development of AAC through direct toxicity of congealed bile to the gallbladder epithelium and that ischaemia does so by damaging to all layers of the gallbladder wall.¹

Despite these differing mechanisms leading to gallbladder inflammation in AAC and ACC, there is significant overlap in pathological findings among the two diseases.⁷ While, grossly, the presence of erosions and ulceration secondary to gallstones is only found in ACC, bile infiltration, leucocyte margination, lymphatic dilation, epithelial degeneration and muscle layer necrosis are found on pathology in both diseases.^{7 8} In a comparison of 34 gallbladders with AAC and 28 with ACC, more leucocyte margination, increased focal lymphatic dilation and deeper muscle wall bile infiltration were found in AAC.⁷ Muscle layer necrosis was more widespread in ACC. However, no single finding was specific for either disease.⁷ In a separate study including five gallbladders with AAC and five with ACC, microangiography with subsequent histopathological analysis was performed immediately after cholecystectomy.⁹ The authors reported regular capillary network filling in ACC but poor filling in AAC; they concluded that disturbances in microcirculation were likely important for the pathogenesis of AAC but not ACC.⁹

Our case of AAC was unusual both in its presentation and in its radiological findings. From presentation until the time of surgery, the patient denied abdominal pain and did not have a Murphy sign. For patients who are able to provide history, the presence of right upper quadrant pain and a pattern of pain similar to ACC would be expected for a patient with ACC.³ Of 47 patients with AAC in one study, 39 (83%) reported right upper quadrant abdominal pain.¹ In a separate study including 27 patients, 21 (78%) reported abdominal pain.³ The absence of abdominal pain reduced our clinical suspicion for cholecystitis and led to a delay in diagnosis. Thankfully, he had an excellent clinical outcome, but his gallbladder had progressed to necrosis by the time of removal and if delayed further, he may have had a morbid turn of events. In review, this case represents an unusual presentation of a relatively common condition.

One explanation for the subtle clinical picture could be that the patient had an undiagnosed chronic cholecystitis that had worsened. Six months prior, the patient had been admitted with E. coli-sepsis and, although no source was identified, right upper quadrant US did identify a thickened gallbladder wall. Chronic acalculous cholecystitis is known to present more insidiously and with less severe symptoms.¹⁰ If the patient indeed had undiagnosed chronic acalculous cholecystitis, then his presentation and course during our report here may have been due to development of a necrotising infection, which is consistent with the final pathology report.

While the patient’s physical examination was unusual for AAC, his illness severity was not unprecedented. AAC has traditionally been thought of as a disease of the critically ill.⁴ It is often suspected in patients with an unknown source of sepsis and may be provoked by prolonged fasting, total parenteral nutrition and other critical illness states.¹⁰ In contrast to this common teaching, we identified studies where ACC is described in non-critically ill patients and identified in the outpatient setting in some cases. One study found that among 47 patients diagnosed with AAC, 36 of them (77%) developed the illness at home and without evidence of acute illness or trauma.¹¹ A second study identified 22 patients diagnosed with AAC over a 43-month period. Twenty of the 22 (91%) presented as outpatients.⁶ Our case adds another example of the condition ocurring in a relatively well-appearing patient without shock or need for intensive care. It is a reminder that clinicians should keep ACC in their differential diagnosis for a patient presenting with abdominal pain, regardless of illness severity.

The other unusual aspect of this case is the presentation by imaging. His US only revealed thickening of the gallbladder wall. It was repeated 2 days later just before going to his operation and continued to not demonstrate other features of cholecystitis. These findings highlight the limited sensitivity of US for confirming acute cholecystitis, which ranges from 28% to 92%.^{1 11} However, if findings are seen on US, specificities are consistently reported at more than 90%.¹¹ The entire right upper quadrant appeared inflamed on contrasted CT imaging. Without contrast, the read favoured perforation of the small bowel. However, after excluding this with oral contrast, the pattern of inflammation was more clearly identified as originating from the gallbladder. The contrasted CT also identified pericholecystic fluid accumulation and inflammation of the surrounding mesenteric fat. These findings were not appreciated on an US later that night. To diagnosis ACC with imaging, there are major and minor diagnostic findings. Major findings include gallbladder wall thickening >3 mm, a positive Murphy sign, a striated gallbladder, the presence pericholecystic fluid, mucosal sloughing and intramural gas, while minor findings include gallbladder distension >5 cm and echogenic bile.¹² Our patient had one major finding (gallbladder wall thickening >3 mm) after initial imaging, but only met diagnostic criteria after the second CT. When considering a diagnosis of AAC or ACC, careful discussion with radiology and the surgical team to determine the most appropriate diagnostic imaging may be useful when a preliminary US is not conclusive.

In our review of the literature, we did not find any association between his oral tyrosine kinase inhibitor or his diagnosis of CML with development of cholecystitis.¹³ Bosutinib, which our patient had been taking for several months, is associated with hepatotoxicity and gastrointestinal side effects but not cholecystitis.¹⁴ His CML was also in remission and after surgery and his leucocyte count was under 10 000 with no blasts present. The CML was being managed with maintenance bosutinib, a second-generation tyrosine kinase inhibitor. This drug is associated with liver enzyme elevation in up to 25% of patients.¹⁵ His liver enzymes had been elevated for many months and this was attributed to the drug. His liver function tests were not significantly different than they had been in the preceding months during his admission. However, following removal of the gallbladder, his bilirubin returned to a normal level from over 2 mg/dL to 0.5 mg/dL at the time of discharge. It is plausible that the diffuse inflammation seen by radiology in the right upper quadrant had led to his elevated bilirubin. Interestingly, his alkaline phosphatase was normal throughout the admission.

Learning points.

• Acalculous cholecystitis may present in non-critically ill patients and most patients will have symptoms that resemble typical cholecystitis.

• Diagnostic imaging with ultrasound and CT scan both have a sensitivity of under 50% in confirming acalculous cholecystitis.

• When the diagnosis is suspected but initial imaging is indeterminate, careful discussion with radiology and the surgical team may be helpful in guiding the diagnostic approach or deciding to intervene surgically based on clinical symptoms.

• This case represents a rare presentation for acalculous cholecystitis. The absence of abdominal pain, despite fever, leucocytosis and ultimately necrosis of the gallbladder, is remarkable.