Acute Cholecystitis Occurs More Commonly Than Expected Following Allogeneic Hematopoietic Stem Cell Transplantation and Is Ineffectively Diagnosed Using Abdominal Ultrasound

Having observed an unexpectedly high rate of acute cholecystitis following allogeneic hematopoietic stem cell transplantation (allo-HCT), we sought to define the risk factors for development of acute cholecystitis following allo-HCT, as well as the incidence, radiographic presentation, and outcomes of this disease in the transplant population.

Between January 2001 and December 2011, 644 patients underwent allo-HCT at our institution. Using ICD-9 codes, we screened this population for a diagnosis of acute cholecystitis or for cholecystectomy in the first year following HCT. Cases were confirmed through manual chart review, and were defined as having one or more imaging modalities positive for acute cholecystitis or having surgical gallbladder pathology consistent with acute cholecystitis. We then conducted a nested case-control study with controls randomly selected through incidence density sampling of the transplant cohort at a rate of 3:1, matching for age and sex. Using logistic regression, we evaluated multiple potential risk factors for the development of acute cholecystitis, including underlying hematologic malignancy, graft-versus-host-disease (GVHD), total parenteral nutrition (TPN) use, graft source, ABO incompatibility, >10% weight loss, conditioning regimen, and cytomegalovirus reactivation. Finally, we conducted a separate case-control study to evaluate the effectiveness of multiple radiographic studies in the diagnosis of acute cholecystitis in HCT recipients (n=32) compared to randomly selected non-transplant control patients who were diagnosed with acute cholecystitis at our institution (n=96).

The incidence of acute cholecystitis in the first year of transplant was 5.0% (32/644 patients). Of the 32 patients, 21 (65.6%) were male and 11 (34.4%) were female with a median age of 52.2 years (range, 24–75 years). Median time from HCT to diagnosis of acute cholecystitis was 56.5 days (range, 6–342 days). Twenty of 32 patients (62.5%) were treated with cholecystectomy, 7 (21.9%) with percutaneous cholecystostomy drainage, and the remaining 5 (15.6%) with conservative medical management. Twenty of the 32 patients who developed acute cholecystitis died within one year of HCT (62.5%), compared with 19 of 96 control patients (19.8%), (p<0.001). Of the 20 cholecystitis patients who died, three (15%) died as a direct result of acute cholecystitis; cause of death was septic shock in all 3 cases. HCT patients who received TPN had a higher risk of developing acute cholecystitis (multivariate OR, 3.41; p=0.009), while development of GVHD was protective against developing acute cholecystitis (multivariate OR, 0.28; p=0.006). When accounting only for development of grade III-IV GVHD, a trend towards protection remained but was not statistically significant (OR, 0.24; p=0.061). Ultrasound findings were equivocal for acute cholecystitis (vs. definitively positive or negative) more frequently in allo-HCT patients than in non-HCT patients (50.0% vs. 30.6%), although this difference was not statistically significant (p=0.063). Of the 32 case patients, 30 (93.8%) had US performed, and of these only 13 (43.3%) had an US positive for acute cholecystitis. Computed tomography (CT) scan was equivocal (versus definitively positive or negative) for acute cholecystitis at similar rates in allo-HCT and non-HCT patients (54.7%, vs. 41.2%, p=0.331), and cholescintigraphy (HIDA) scan was positive for cholecystitis at similar rates in both populations (80.0% vs. 77.4%, p=0.82).

Acute cholecystitis is a common complication of allo-HCT, especially in patients who received TPN. It is associated with a high 1-year mortality rate, and abdominal US has suboptimal performance in diagnosing it. Our data suggest that a high index of suspicion and use of alternative imaging modalities are required to make a prompt diagnosis and may improve the outcomes of allo-HCT patients.

No relevant conflicts of interest to declare.