Multidetector computed tomography evaluation of obstructive jaundice: a cross-sectional study from a tertiary hospital of Nepal

Background: The utilization of advanced multidetector computed tomography (MDCT) technology along with postprocessing reconstruction techniques has significantly enhanced the clarity of visualization of the hepato-biliary tree. Therefore, this study was conducted to evaluate the diagnostic statistics of MDCT and its associated features in the evaluation of obstructive jaundice, with respect to surgical or histopathological diagnoses. Methods and methodology: The authors conducted a cross-sectional study among 30 participants with obstructive jaundice using purposive sampling. The authors calculated the diagnostic statistics of non-neoplastic and neoplastic types, along with specific etiologies of obstructive jaundice identified through MDCT using a statistical package for social sciences (SPSS) v21 and MedCalc v12.3.0. The ethical clearance was obtained from the institutional review committee of BPKIHS, Nepal (Ref no: Acd/291/075/076-IRC). Results: The sensitivity and the negative predictive value of MDCT for non-neoplastic cause to detect obstructive jaundice were 100% (95% CI: 79.41–100.00) and 100% (95% CI: 75.29–100.00), while the specificity and the positive predictive value for neoplastic cause to detect obstructive jaundice were 100% (95% CI: 79.41–100.00) and 100% (95% CI: 75.29–100.00), respectively. Similarly, the accuracy for either non-neoplastic or neoplastic cause was 96.67% (95% CI: 82.78–99.92). The most common cause of obstructive jaundice was choledocholithiasis (33.34%) followed by cholangiocarcinoma (20%), ampullary carcinoma (13.33%), and choledochal cyst (13.33%). The diagnostic accuracy of the individual etiology of common causes of obstructive jaundice ranged from 82.78 to 100%. Biliary obstruction was most frequently observed in the periampullary region (83.33%), followed by the proximal common bile duct (6.67%), hilar region (6.67%), and intrahepatic region (3.33%). Conclusion: The MDCT could serve as the initial and time-efficient excellent imaging modality for diagnosing various causes of obstructive jaundice with greater accuracy. It can differentiate non-neoplastic from neoplastic causes of obstructive jaundice.


Introduction
The initial assessment of obstructive jaundice involves the differentiation between intrahepatic and extra-hepatic biliary obstruction.Various noninvasive and invasive imaging modalities have been used to assess obstructive jaundice following a physical examination and biochemical parameters.Radiological investigations play an important role in determining treatment approaches [1,2] .
Although ultrasound is considered the primary imaging modality, its sensitivity varies widely, ranging from 20 to 80% depending on the underlying cause [3,4] .The presence of artifacts stemming from bowel gas, gall bladder and bile duct calculi, breathing movements, and obesity bring it down in the list of imaging modalities of choice in obstructive jaundice [5] .
We have employed multidetector computed tomography (MDCT) for the diagnostic assessment of biliary obstruction.Recent advancements in MDCT, coupled with postprocessing reconstruction techniques like multiplanar reformations (MPR) and minimal intensity projections, have significantly enhanced the visualization of the hepato-biliary tree.The MPR technique allows multiplanar exploration of the biliary ductal structure, while minimal intensity projections enables improved depiction of small biliary or pancreatic ducts [6,7] .MDCT has the capability to distinguish between benign and malignant strictures, stage complex biliary malignancies, and determine their involvement and invasion of adjacent organs as well as regional lymphadenopathy, detect metastases, and identify fluid in the peritoneal cavity [5,8] .Additionally, it is rapid and can be accomplished within a single breath-hold period [9] .
The aim of our study was to assess the diagnostic accuracy of MDCT in the evaluation of obstructive jaundice, with reference to surgical or histopathological diagnoses, and to explore the MDCT characteristics associated with of different causes of obstructive jaundice.

Study design, sample size, and sampling technique
In the Department of Radiodiagnosis and Imaging, we conducted a cross-sectional study within a hospital setting spanning a year, from 8 October 2019 to 7 October 2020.The ethical clearance was secured from the institutional review committee of B.P. Koirala Institute of Health Sciences, Dharan, Nepal (Ref no: Acd/ 291/075/076-IRC).The determination of our sample size drew inspiration from Mathew et al. [10] investigation, where MDCT demonstrated a sensitivity of 90% in diagnosing and distinguishing cholangiocarcinoma.Considering the proportion of cholangiocaracinoma to be 90% with a significance level set at 5%, an absolute precision of 1.96 and a finite population correction (consisting of 31 cases of obstructive jaundice that had undergone imaging in the preceding year), the calculated sample size was 30.
The study utilized purposive sampling to collect samples.All patients exhibiting clinical and biochemical features of obstructive jaundice who were referred for an MDCT scan and subsequently received a conclusive diagnosis through histopathology, cytopathology, or surgical findings were included as participants in the study.Patients with contraindications for contrastenhanced CT, those with nonobstructive cases of jaundice, and patients experiencing recurrent malignancy causing obstructive jaundice were excluded from the study.The research adhered to the strengthening the reporting of cohort, cross-sectional and case-control studies in surgery (STROCSS) criteria during its execution [11] .

MDCT imaging of participants with obstructive jaundice
The MDCT was performed using a multislice CT scanner (ECLOS 16; HITACHI), excluding contraindications to CECT.The process involved a noncontrast CT scan followed by a contrast CT scan while in the supine position.Positive oral contrast and intravenous contrast images were obtained.Three-dimensional reconstructions employing thin planar slicing (0.625 mm) and MPR were performed in coronal and sagittal planes to enhance the visualization of the intraluminal and wall lesions of the biliary tract [12,13] .
During the analysis of noncontrast images, observations were made for the presence of calculi, calcifications, masses, the baseline Hounsfield Units (HU) of the mass, and the condition of the common bile duct (CBD) wall.Subsequently, on postcontrast images, attention was given to confirming the presence of a mass, assessing its size, shape and margins, adjacent infiltration status, and enhancement pattern.Enhancement was labeled significant if there was an HU difference of more than 15 on postcontrast images compared to the baseline HU [14][15][16] .Features such as an ill-defined mass with pronounced postcontrast enhancement, an intraluminal polypoidal mass with adjacent thickening and/or enhanced biliary wall, abnormal invasion of surrounding structures or loss of fat plane, and evidence of metastasis were indicative of malignant mass [14,[17][18][19][20][21] .
In cases presenting with obstructive jaundice features, CBD dilatation was identified when the maximum transverse diameter exceeded 7 mm [22][23][24] and postcholecystectomy, when exceeding 10 mm [25] , regardless of age.For the main pancreatic duct (MPD), a diameter greater than 2.5 mm was considered dilated [26][27][28] .A diameter surpassing 2 mm and/or more than 40% compared to an adjacent portal vein branch diameter indicated the presence of intrahepatic biliary radical dilation [29] .In scenarios involving biliary obstruction at the periampullary region, dilatation of both CBD and the MPD was categorized as the double duct sign [14,28,30,31] .
Concerning biliary strictures, a neoplastic/malignant cause was considered if there was a long segment (greater than 1.5 cm) [32] , abrupt narrowing, wall thickness exceeding 1.5 mm, and enhancement.Conversely, a smooth gradual tapering with no abnormal enhancement indicated a non-neoplastic/benign cause of stricture [15,33] .A lack of visualization of the confluence of the right and left hepatic ducts accompanied by abrupt tapering and the presence of intrahepatic biliary radical dilation pointed toward a likelihood of hilar cholangiocarcinoma [21,[34][35][36] .Distension of the gall bladder was denoted when the long axis dimension exceeded 10 cm and/or transverse diameter surpassed 4 cm [37][38][39][40] .
Calculi and non-neoplastic strictures within the biliary tract were confirmed through surgical findings.Other observations

HIGHLIGHTS
• The multidetector computed tomography could serve as the initial and time-efficient imaging modality for diagnosing various causes of obstructive jaundice.• Multidetector computed tomography can differentiate nonneoplastic from neoplastic causes of obstructive jaundice.• The most common cause of obstructive jaundice was choledocholithiasis. • Biliary obstruction was most frequently observed in the periampullary region.were corroborated by both surgical and cyto/histopathology findings.The MDCT diagnosis was formulated based on these findings and then compared with the ultimate diagnosis.The final diagnosis was determined from cyto/histopathological reports in 20 cases and surgical findings in 10 cases.The accuracy of the MDCT was assessed, taking surgery and/or cyto/histopathology as the definitive diagnosis, as applicable.The recorded findings were documented on pro forma for subsequent data analysis.

Data entry and analysis
The collected data were tabulated in the Microsoft Excel 2019 v16.0 (Microsoft) and analyzed using the statistical package for social sciences (SPSS) version 11.5, IBM SPSS v21 (IBM) and MedCalc for Window version 12.3.0(MedCalc-Software).We expressed the categorical data as frequency and percentages, and the continuous data as mean SD.Similarly, the sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of non-neoplastic and neoplastic type of obstructive jaundice as well as the individual etiology of obstructive jaundice detected on MDCT were determined in reference to histopathological/cytopathological and/or surgical diagnosis.

Results
Out of the 30 participants meeting the inclusion criteria, threefifths of the participants (60%, 18) were female.The mean age of the participants were 54.90 19.88 years with the minimum and maximum being 10 and 87 years.About two-fifths of the participants belonged to the age group 31-60 years (40%, 12) and 61-90 years (43.30%,13) (Table 1).Out of 30 participants, nine cases were diagnosed as choledocholithiasis on MDCT, which were subsequently confirmed on surgery and one falsely diagnosed case of CBD sludge on MDCT was confirmed as choledocholithiasis on surgery.Similarly, seven cases were diagnosed as cholangiocarcinoma on MDCT of which six cases were confirmed as cholangiocarcinoma on cyto/histopathological examination.One case was falsely diagnosed as a stricture type of cholangiocarcinoma on MDCT, which was diagnosed as carcinoma head of the pancreas on histopathology.Likewise, three cases were diagnosed as benign biliary stricture on MDCT, of which two cases were confirmed as benign biliary stricture postoperatively.One case of benign biliary stricture on MDCT came out to be a periampullary metastatic deposit from gall bladder carcinoma on histopathology examination (Supplementary File, Supplemental Digital Content 1, http:// links.lww.com/IJSGH/A31)(Table 2).
In our study for specific etiologies, the individual sensitivity, specificity, PPV, and NPV of MDCT were calculated (Table 3).
Out of the 30 cases of obstructive jaundice, more than fourfifths (83.3%, 25) had periampullary obstruction, more than onethirds (36.7%, 11) had mass, and more than four-fifths (80.0%, 24) had both CBD and IHBR type biliary dilatation.Similarly, out of the 16 cases of obstructive jaundice due to malignant obstruction, more than two-thirds, that is 78.6 and 68.8% had abrupt distal tapering of the dilated biliary tract due to malignant and overall cases, respectively.Excluding the hilar and intrahepatic level of obstruction, gall bladder was distended in 22.2% of the participants.Likewise, out of periampullary obstruction in 25 cases, the MPD dilatation was absent in three-fourths (76%, 19) (Table 4).
About the characteristics of the mass causing the obstructive jaundice, the arterial phase enhancement, ill-defined margins, and surrounding invasions were present in 63.6, 81.8, and 36.4%,respectively (Table 5).

Discussion
In our current study, our objective was to investigate both the diagnostic statistics and the MDCT features associated with obstructive jaundice.The average age of participants in our study was 54.9 19.8 years, with over 60 years (43.3%)being the maximum age group.This finding aligns with the studies conducted by Darwish [41] and Mathew et al., where the majority of the participants fell within the fifth to sixth decades of life.Female predominance was observed in our study, with a male-to-female ratio of 1:1.5.Similarly, Anderson et al. [42] , also reported female predominance with a male-to-female ratio of 1:1.6.
Concerning the MDCT's capacity to differentiate between benign and malignant lesions, our study correctly identified 13 out of 14 cases as malignant, with one false negative case.The agreement with the final diagnosis was nearly perfect (Kappa value 0.933).This is consistent with findings from multiple other studies.For instance, Mathew et al. [10] study correctly identified 22 out of 23 malignant cases, achieving a sensitivity of 100%, specificity of 95.65%, and an accuracy of 98%.Similarly, Reiman et al. [44] research accurately predicted malignancy in 25 of 27 patients (92%) and benign disease in 13 of 17 participants (77%).Likewise, Ahmetoglue et al. [45] study on MDCT for the assessment of patients with biliary obstruction showed both sensitivity and specificity of 94% for the diagnosis of malignant obstruction.
In our study, a mass lesion causing biliary obstruction was detected in 78.5% (11 out of 14) of malignant cases through MDCT.This outcome corresponds to findings from Reiman et al. and Mathew et al. [10,44] .Excluding choledocholithiasis and choledochal cyst, 68.8% (11/16) of participants had abrupt distal tapering.This correlates with a study by Agrawal et al. [33] , which showed that abrupt tapering was indicative of malignancy.In this study, among 25 cases with periampullary pathologies, six (24%) had dilatation of MPD and all of these were malignant.None of the benign periampullary pathology showed dilatation of MPD.This finding is supported by Tanaka et al. [27] study, which associated MPD dilatation of over 2.5 mm with periampullary malignant pathology.A similar observation was reported by Kim et al. [49] , showing the double duct sign in ampullary carcinoma (52%) and carcinoma head of pancreas (62%).Moreover, Krishna et al. [30] concluded that dilation of both MPD and CBD on imaging scans suggests pancreatic malignancy.
In our study, heterogeneous enhancement in delayed phase and peripheral rim enhancement in arterial phase were seen in 80 cases and 20%, respectively (Figs. 1 and 2).Darwish et al. [48] study also showed heterogeneous enhancement in the delayed phase in cases of cholangiocarcinoma.Similar imaging characteristics were reported in studies by Olthof et al. [50] and Chung et al. [36] .
Despite its contributions, our study does have limitations.The sample size is relatively small, and the data was derived from a single center, which might introduce potential confounding factors.

Conclusion
In our study on obstructive jaundice, the MDCT demonstrated a sensitivity of 92.86%, specificity of 100%, PPV of 100%, NPV of 94.12%, and an accuracy of 96.67% in detecting neoplastic causes.It also had a sensitivity of 100%, specificity of 92.86%, PPV of 94.12%, NPV of 100%, and accuracy of 96.67% in detecting non-neoplastic causes.The most common causes of obstructive jaundice were choledocholithiasis followed by cholangiocarcinoma.Biliary obstruction was most frequently observed in the periampullary region (83.33%),followed by the proximal CBD (6.67%), hilar region (6.67%), and intrahepatic region (3.33%).On MDCT, the majority of neoplastic causes of obstructive jaundice (78.5%) exhibited a mass lesion as well as abrupt distal tapering of the biliary tract at the site of obstruction.Neoplastic causes were further characterized by the presence of enhancement and invasion of adjacent structures, irregular wall with abrupt distal tapering of biliary tract, distended gall bladder, nonvisualization of the confluence of the right and left hepatic ducts, and a double duct sign.These features suggest that MDCT can serve as initial and time-efficient excellent imaging modality for diagnosing various causes of obstructive jaundice, with an accuracy ranging from 82.78 to 99.92%.

Figure 1 .
Figure 1.Postcontrast axial (A) and reformatted coronal (B) sections show ill-defined heterogeneously enhancing lesion in hilar region with poor interface with adjacent liver parenchyma.Confluence of right and left hepatic ducts is not visualized and there is presence of intrahepatic biliary radical dilation (IHBRD).It was diagnosed as hilar cholangiocarcinoma on histopathology.Photomicrograph (C) of hilar cholangiocarcinoma shows well defined tumor glands interspersed with poorly differentiated small tumor groups and single tumor cells (Hematoxylin and eosin stain 200 × ).

Figure 2 .
Figure 2. Postcontrast coronal reformatted image (A) showing heterogeneously enhancing mass lesion in head of pancreas with ill-defined margin, infiltration into surrounding structures, upstream dilatation of biliary tree, and distended gall bladder.Postoperative Whipple's specimen (B) showing mass in head of pancreas infiltrating into the duodenum.On histopathological examination, it was confirmed as carcinoma head of pancreas.Photomicrograph (C) of ductal adenocarcinoma of head of pancreas shows nuclear atypia and desmoplasia along with atypical cells predominantly forming glands (Hematoxylin and eosin stain 400 × ).

Table 1
Background characteristics of the study (n = 30).

Table 2
Comparisons of MDCT diagnosis with the final histopathological diagnosis of the cause of obstructive jaundice and diagnostic statistics of MDCT to detect obstructive jaundice.
The CI for sensitivity, specificity, and accuracy are 'exact' Clopper-Pearson CIs.The CI for the likelihood ratios is calculated using the 'Log method'.The CI for the predictive values is the standard logit CI. a Final diagnosis.bThese values are in ratio (not %).cThese values are dependent on disease prevalence.FN, false negative; FP, false positive; NLR, negative likelihood ratio; PLR, positive likelihood ratio; TN, true negative; TP, true positive.study,the reported prevalence from this study was 53.33 (95% CI: 34.33-71.66)and 46.67% (95% CI: 28.34-65.67)(Table

Table 3
Diagnostic statistics of individual diagnosed cases of obstructive jaundice detected in MDCT in reference to surgical or cyto-/ histopathological diagnosis as final diagnosis.

Table 5
The characteristics of mass causing obstructive jaundice detected in MDCT.