Acute Acalculous Cholecystitis in Hospitalized Patients With Hematologic Malignancies and Prognostic Importance of Gallbladder Ultrasound Findings.

Patients with hematological malignancies are prone to develop acute acalculous cholecystitis (AAC), a necroinflammatory disease of the gallbladder in the absence of cholelithiasis, especially during advanced phases of illness, chemotherapy, and neutropenic episodes. 1 - 4 Although the acalculous variety is less common than calculous cholecystitis, AAC is more common in critically ill patients. 5 Patients with hematological malignancies are often critically ill and in an advanced stage of disease and have higher rates of emergency room visits, hospitalization, and intensive care unit use. 6 - 8 Clinical diagnosis of AAC in these patients is difficult because clinical features such as right upper quadrant pain and fever may be attributable to other abdominal complications such as enterocolitis, and abnormal leukocyte counts can be caused by the underlying hematological malignancy. Compared with calculous cholecystitis, AAC is associated with higher rate of complications, with gangrene found in up to 50% and gallbladder perforation in up to 10% of cases on histological examination. 5 Therefore, early diagnosis and intervention are imperative whenever AAC is clinically suspected in patients with advanced hematological malignancies. Imaging is used to support the clinical diagnosis of AAC. Ultrasonography is often the first modality used because it is easily accessible at the bedside, quick, relatively inexpensive, and repeatable. Sonographic criteria commonly used for the diagnosis of AAC are gallbladder wall thickening, gallbladder distension greater than 4 mm, pericholecystic fluid, sludge, sloughed mucosal membrane, and intramural gas. 9 Different combinations of these findings have been used as criteria to diagnose AAC, with sensitivities ranging from 18% to 93%. 10 , 11 Imhof et al. found that sonomorphological changes were detected within 24 hours of symptom onset, and a triad of gallbladder hydrops, wall thickening, and sludge were detected by 11 days on average. 11 Wang et al. used a scoring system and found a higher tendency for complications in patients with a higher score; however, this finding was not statistically significant. 12 Cornwell et al. were able to preoperatively diagnose AAC in 13 of 14 patients using the major criteria of wall thickening, pericholecystic fluid, intramural gas, and sloughed mucosal membrane and the minor criteria of distension and sludge. 10 However, a few studies of critically ill patients have shown that some of these findings may be seen in patients without AAC. 13 , 14 Thus, uncertainty exists about the diagnostic value of sonographic criteria and their prognostic significance. Although limited studies have documented AAC in patients with leukemia, 1 , 4 to our knowledge no study has specifically documented the prognostic significance of AAC sonographic criteria in patients with hematological malignancies. A better understanding of what percentage of patients with similar right upper quadrant symptoms meet AAC sonographic criteria on ultrasound examination, along with the prognostic significance of these criteria, would improve clinical management and treatment plans. Therefore, the purpose of our study was to evaluate the prognostic significance of AAC sonographic criteria in hospitalized patients with hematological malignancies. We hypothesized that the presence of positive sonographic criteria could predict a higher rate of complications and/or mortality. To test our hypothesis, we retrospectively studied AAC sonographic criteria in hospitalized patients with hematological malignancies who had clinical findings suggestive of AAC (fever, right upper quadrant pain, abnormal liver function tests, and tenderness).

In this study, 368 patients with advanced hematological malignancies underwent an abdominal ultrasound examination at our institution for right upper quadrant symptoms. Of these, 94 patients (25.5%) had clinical suspicion of AAC; these included 36 patients with acute lymphocytic leukemia, 33 with acute myelogenous leukemia, 10 with chronic myelogenous leukemia, 8 with non-Hodgkin lymphoma, and 7 with multiple myeloma. Of these 94 patients, 43 (45.7%) were positive for AAC based on ultrasound criteria, and 51 (54.3%) were negative for AAC. Patient characteristics in the 2 groups are shown in Table 2 . Among the 94 patients, ultrasonography showed abnormal gallbladder wall thickening in 66 (70%), pericholecystic fluid in 45 (48%), sloughed mucosa in 15 (16%), echogenic bile/sludge in 60 (64%), and gallbladder hydrops in 38 (40%). None of the ultrasound examinations showed intramural gas. Nine patients, all of whom were in the AAC-positive group, developed complications; these included 2 patients with gallbladder perforation, 3 with pericholecystic abscess, and 4 with septicemia. None of the patients developed gallbladder hemorrhage ( Table 2 ). We found no significant difference in the presence of any individual complication between the 2 groups ( Table 3 ). However, a positive ultrasound diagnosis for AAC was associated with the occurrence of complications, which were seen in 9 (20.9%) of 43 patients in the positive group and none in the negative group ( P = 0.0005, Table 3 ). Among ultrasonographic features, the presence of pericholecystic fluid (8 of 45 patients, P = 0.01), echogenic bile/sludge (9 of 51 patients, P = 0.02), and gallbladder distention > 4 cm (9 of 38 patients, P = 0.0002) were associated with the occurrence of complications ( Table 4 ). Sloughed mucosa was seen only in the positive group (15 of 43 patients), but only 1 patient with sloughed mucosa developed a complication ( P ≥ 0.99). Although 2 deaths occurred in the positive group, they were unrelated to gallbladder pathology. In the negative group, 21 (41%) of the 51 patients were diagnosed with AAC on follow-up diagnostic tests: 14 (27.4%) patients were positive on repeat ultrasound exams performed within 1 week, and 7 (13.7%) were positive on follow-up hepatobiliary scans performed within 1 week. Another 14 patients (27.4%) in the negative group remained negative on subsequent follow-up tests despite persistent right upper quadrant symptoms. None of the patients diagnosed with AAC on follow-up tests developed complications. Thirty-eight patients in the positive group underwent follow-up tests; of these, 13 (34%) patients remained positive the follow-up tests. Of the 6 patients in the positive group who underwent follow-up ultrasound examinations within 1 week, 1 patient (2.6%) remained positive and 5 (13%) were negative for AAC on the subsequent examinations ( Table 5 ). Patients with positive AAC diagnoses based on ultrasonographic findings had a significantly longer duration of fever and right upper quadrant symptoms than those with negative diagnoses. Specifically, the mean duration of symptoms was 7.8 days in the positive group and 3.9 days in the negative group ( P < 0.0001, Table 6 ). Complete resolution of fever and right upper quadrant symptoms through medical treatment occurred in 32 of the 43 patients in the positive group and in 45 of the 51 patients in the negative group. Cholecystostomy was required for 10 patients in the positive group and 4 patients in the negative group ( Table 2 ).

Our results indicate that patients with advanced hematological malignancies diagnosed with AAC based on sonographic criteria have a higher rate of complications and longer duration of symptoms compared with patients with negative ultrasonographic findings. In our study, the positive group had a complication rate of 20.9%. This higher rate is in concordance with complication rates in some previous studies. For example, studies have shown a higher incidence of complications such as gangrene and gallbladder perforation in AAC compared with calculous cholecystitis. 15 In a series of 445 patients with postoperative AAC, Inoue et al. found a complication rate of 55% and mortality rate of 23%. 16 More recent studies have shown mortality rates ranging from 11% to 40%; the lower rates in some recent studies are likely due to the higher index of suspicion, improved imaging, and better treatment for AAC. 17 - 19 Previous studies attempted to find clinical or imaging features of AAC that predict complications. Wang et al. found that high white blood cell count and older age correlated with a higher incidence of complications. 12 They used an ultrasound scoring system and found a trend toward a higher score in patients who later developed severe complications, although the trend was not statistically significant. 12 Although our study was not intended to find specific predictors of complications, a positive ultrasound study based on 2 major or 1 major and 2 minor criteria showed a significantly higher rate of complications. Interestingly, although gallbladder distension and sludge were considered minor criteria, all 9 patients in our study who developed complications had both these findings. Other studies have shown that gallbladder sludge leads to subsequent complications such as cholecystitis, cholangitis, and pancreatitis; these symptoms can be prevented by cholecystectomy. 20 , 21 Overdistention of the gallbladder (which was seen in all 9 patients who experienced complications in our study) and increased intraluminal pressure are believed to cause ischemia, potentially leading to complications such as gangrene and gallbladder perforation. 22 Many risk factors have been implicated in causing AAC, including recent surgery, critical illness requiring intensive care unit admission, burns, immunosuppression, and sepsis.. 9 In patients with hematological malignancies, abdominal infections are life threatening, especially during periods of neutropenia, with reported mortality rates up to 26%. 3 In patients with hematological malignancies, AAC is more common than calculous cholecystitis, and there is a high risk of sepsis. 3 Leukemic infiltration of the gallbladder can also cause cholecystitis and has been reported as a rare presenting manifestation of leukemia. 23 Surgery for acute abdominal pain in patients with hematological malignancies is risky, with mortality rates up to 55%. Although early laparoscopic cholecystectomy is considered safe in patients with AAC, 24 individualized management based on prognosis and risk is recommended in patients with advanced malignancies and cholecystitis. 25 Another important feature of AAC is its rapidly progressive nature. Johnson et al. found that 40% of patients who underwent surgery 48 hours after the onset of symptoms had gallbladder perforation, whereas only 8% of those who underwent surgery earlier had perforation. 26 In our study, 41% of the patients in the negative group were found to be positive for AAC on follow-up ultrasound or hepatobiliary scans within 1 week. This finding reflects the rapidly progressive nature of AAC. Still, these patients had a lower rate of complications and duration of symptoms. These patients might have had early or milder presentation of AAC, resulting in initial negative ultrasound studies. Moreover, early diagnosis and treatment could explain the lower rate of complications and duration of symptoms in this group. These findings agree with the study by Jeffrey et al., who reported that 4 of 8 patients whose initial ultrasound studies were negative for AAC were found to have increased gallbladder wall thickness on follow-up ultrasound examinations performed within 24 hours. 27 Because of the rapidly progressive nature of AAC, several authors have emphasized the importance of daily follow-up ultrasound studies to help early detection of progression and to avoid misdiagnosis. 11 , 28 Hence, follow-up ultrasonography in borderline cases will help reduce complications and mortality by facilitating early diagnosis and prompt treatment. Some authors have shown that ultrasound findings of AAC are nonspecific and may be found in patients without AAC. Boland et al. found no significant correlation among sonographic gallbladder abnormalities with clinical and laboratory parameters. 13 But others have shown that when correlated with clinical features, ultrasonography is the investigation of choice for AAC in critically ill patients. 9 Hence, diagnosis should not rely solely on ultrasound findings and should be correlated with clinical and laboratory parameters. Our study has several limitations. Most importantly, it was a retrospective study based on search terms with inherent selection bias. Although it was possible to select a homogeneous group of hospitalized patients with a diagnosis of advanced hematological malignancy, the patients received different treatments for their primary cancer. We only included patients with right upper quadrant pain who underwent right upper quadrant ultrasound examination; it is possible that some patients who were not imaged or were imaged with other cross-sectional modalities were excluded, resulting in underestimation of complications. Also, the study was performed at a tertiary care cancer center. Thus, the generalizability of our findings in other settings is unclear. AAC was not pathologically proven in the majority of the patients (78 of 94); however, such patients in the acute phase rarely undergo cholecystectomy in routine clinical practice due to their critical condition and advanced malignancy. A significant number of ultrasound-negative patients (41%) were positive on follow-up studies; it would have been helpful to identify which ultrasound findings should prompt a follow-up imaging study, but our study did not find any statistically significant predictors. Finally, we did not assess the gallbladder for the resolution of sonographic findings following treatment because our objective was to assess their prognostic significance. In conclusion, patients with positive sonographic criteria and clinical suspicion of AAC should be treated aggressively because they are prone to have a higher complication rate and longer duration of symptoms. In borderline cases, a repeat ultrasound study allows early detection of AAC progression and prompt treatment to signifiantly reduce complications.

This HIPAA-compliant retrospective study was approved by The University of Texas MD Anderson Cancer Center’s institutional review board with a waiver for informed consent. The study subjects were identified through a search of the institutional radiology database for abdominal ultrasound studies performed for right upper quadrant symptoms in hospitalized patients with advanced hematological malignancies (leukemia, lymphoma, or myeloma) from August 2011 to August 2016. We defined advanced hematological malignancies as incurable disease at presentation or recurrent disease not eligible for potentially curative treatments per oncology notes. The search yielded 368 patients hospitalized with advanced hematological malignancies who had abdominal ultrasound examinations for right upper quadrant symptoms. Among these 368 patients, we excluded 124 patients who had gallbladder calculi, 58 patients who had calculus cholecystitis, 50 patients who had a history of cholecystectomy, 24 patients who had liver disease that could have contributed to right upper quadrant symptoms, and 14 patients who had more than 1 malignancy; another 4 patients were excluded because of limited ultrasound examination of the gallbladder inconclusive findings. In total, 94 consecutive patients who had advanced hematological malignancies and were suspected to have AAC (median age, 42.5 y; range, 6-64 y) were included in the analysis. All the patients had clinical features suspicious for AAC at the time of ultrasound examination. Specifically, all 94 patients had fever, abnormal liver function test results, right upper quadrant pain, and tenderness. All ultrasound studies were performed in a single institution on Philips IU 22 US machines (Philips Medical Systems, Bothell, WA, USA) using a curvilinear transducer (3-8 MHz). The studies were performed using standardized ultrasound examination protocols by 6 certified ultrasound technologists with experience ranging from 1 to 30 years. Static images and cine loops of the gallbladder were recorded in longitudinal and transverse planes through intercostal, subcostal, and decubitus subcostal approaches. All the images were reviewed by fellowship-trained abdominal radiologists while the patient was still in the department. The images were stored and reviewed on a standard picture archiving and communication system (PACS). Two abdominal radiologists with 5 years and 1 year of experience, who were blinded to the original ultrasound report, reviewed the gallbladder images in consensus on a PACS monitor and documented their findings. The gallbladders were evaluated for the presence of sonographic criteria 9 as shown in Table 1 . An ultrasound study was considered positive for AAC if at least 2 major criteria or 1 major and 2 minor criteria were present. Sonographic findings in a representative case are shown in Figure 1 . We also reviewed these patients’ electronic medical records and collected clinical data, which included patients’ age, sex, primary hematological malignancy, and absolute neutrophil count. The clinical course over 3 months following the ultrasound examination from the initial time of clinical suspicion was reviewed for the duration of symptoms, treatment, complications, follow-up imaging tests, and outcome (mortality). Our analysis included only complications related to cholecystitis that were diagnosed on subsequent imaging tests or pathology; these included gallbladder perforation, pericholecystic abscess, gallbladder hemorrhage, and septicemia due to cholecystitis. The patients were divided into 2 groups: patients with ultrasound-positive AAC and those with ultrasound findings negative for AAC. In the 2 groups, the duration of symptoms and the rate of complications or gallbladder-specific mortality rate were compared. The Wilcoxon rank sum test was used to compare the duration of symptoms, and the Fisher exact test was used to compare complication rates in the 2 groups. All statistical calculations were performed by using IBM SPSS statistics for Macintosh, Version 24.0 and SAS software for windows, version 9.4 (SAS Institute, Cary, NC).