Comparison of MRCP and ERCP in the evaluation of common bile duct and pancreatic duct pathologies

Background Magnetic resonance cholangiopancreatography (MRCP) is a non-invasive imaging modality that has high diagnostic accuracy for a wide range of bile duct and pancreatic duct pathologies. Endoscopic retrograde cholangiopancreatography (ERCP) is still the gold standard for the exploration of the biliopancreatic region. Aim The aim of the study was to compare the diagnostic accuracy of MRCP with that of ERCP in the diagnosis of bile duct and pancreatic duct pathologies. Material and methods A total of 60 patients with common bile duct (CBD) and pancreatic duct pathologies detected on MRCP were subsequently evaluated by ERCP in this observational study. A comparison of MRCP findings with ERCP was made. Results MRCP had a sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of 88.1%, 94.4%, 97.3%, 72.7%, and 90%, respectively, in diagnosing choledocholithiasis in comparison to ERCP. For CBD dilation, the sensitivity was 90.91%, specificity was 93.75% and the PPV, NPV, and accuracy were 97.56%, 78.95%, and 91.67%, respectively, for MRCP. In CBD stricture, MRCP showed a sensitivity, specificity, PPV, NPV, and accuracy of 83.33%, 97.92%, 90.91%, 95.92%, and 95%, respectively. In pancreatic duct dilatation, the sensitivity, specificity, PPV, NPV, and accuracy were all 100%. Pancreatic duct stricture showed a sensitivity, specificity, PPV, NPV, and accuracy of 80%, 98%, 88.89%, 96.08%, and 95%, respectively. For the diagnosis of periampullary carcinoma, the sensitivity, specificity, PPV, NPV, and accuracy rate of MRCP were 80%, 98%, 88.89%, 96.08%, and 95%, respectively. Conclusion No significant difference was found between MRCP and ERCP in diagnosing those six pathologies.


Introduction
Accurate methods of detecting common bile duct (CBD) and pancreatic disease in patients are important to both surgeons and endoscopists for planning an effective interventional strategy; therefore, there is a need for less invasive, safe and highly sensitive diagnostic procedures.
Various invasive and non-invasive diagnostic techniques have been employed to achieve this aim. Non-invasive techniques, such as ultrasound and CT scanning (abdomen and pelvis) are widely used in the preliminary investigations of pancreaticobiliary disease. These techniques, though easily available and less expensive, have limitations in terms of sensitivity, such as the low sensitivity of ultrasonography for detecting common duct calculi, which means that the diagnosis of several common conditions, such as tumours, calculi, sclerosing cholangitis and chronic pancreatitis, may require invasive procedures (1).
Endoscopic retrograde cholangiopancreatography (ERCP) is often regarded as a definitive diagnostic test (2,3). ERCP combines the use of endoscopy and fluoroscopy and has additional therapeutic advantages (3). Although currently ERCP is almost exclusively used for therapeutic purposes, it is still regarded as an important tool in the evaluation of biliary diseases, especially malignancy (4). As ERCP is invasive, the associated disadvantages are as follows: (1) it requires direct cannulation of the common bile duct or pancreatic duct; (2) it requires sedation; (3) it is more operator-dependent, requiring more experienced personnel; and (4) it uses ionising radiation. ERCP is also associated with a complication rate of 1%-7%, including haemorrhage, sepsis, pancreatitis, and bile leak. About one in four complications is severe. The overall complication rate appears relatively consistent across time (2)(3)(4).
In the evaluation of biliary and pancreatic duct anatomy and obstruction, magnetic resonance cholangiopancreatography (MRCP) is a non-invasive, radiation-free, non-operatordependent, multiplanar, and safe alternative to diagnostic ERCP (5). MRCP was developed in 1991, and techniques have improved since then. MRCP makes use of heavily T2-weighted sequences, which significantly increase the signal of static or slow-moving fluid-filled structures, such as bile and pancreatic ducts, and that leads to increased duct to background contrast (6). The most recent software available includes fast, high spatial resolution MRCP sequences, such as a heavily T2-weighted turbo-spin-echo (TSE), single-shot rapid acquisition with relaxation enhancement (RARE), and half-Fourier single-shot TSE (HATSE), which provide clear projectional images that are similar to those provided by ERCP procedures (7).
According to recent studies, the diagnostic accuracy of MRCP is comparable with that of ERCP for the evaluation of extrahepatic bile duct and pancreatic duct abnormalities, such as choledocholithiasis, malignant obstruction of the bile and pancreatic ducts, congenital anomalies, chronic pancreatitis, benign strictures due to sclerosing cholangitis, demonstrating pancreatic pseudocyst, and in cases of biliary cystadenoma and cystadenocarcinoma (7,8). The advantages of MRCP include the following: 3D imaging and image reformatting; it is non-invasive; patients tolerate it well; and it is an excellent diagnostic tool in situations where ERCP is difficult, hazardous, or impossible (7).
In our study, we put forward the hypothesis that MRCP can provide relevant information regarding CBD and pancreatic duct pathology. The aim of the study was to compare the diagnostic accuracy of MRCP with that of ERCP in common bile duct and pancreatic duct pathologies and to correlate both techniques.

Materials and methods
The study was conducted in the Department of Radiodiagnosis, Apollo Hospital Bhubaneswar, the Department of Radiodiagnosis, Kalinga Institute of Medical Sciences, Bhubaneswar, and the Department of Gastroenterology and Hepatology, Apollo Hospital Bhubaneswar, over a period of 1 year, between December 2018 and November 2019. It was a prospective observational study in which patients with common bile duct or pancreatic duct pathologies detected in MRCP were evaluated by ERCP.
The inclusion criteria for our study included patients with CBD pathologies (choledocholithiasis, CBD stricture, CBD dilatation), pancreatic duct pathologies (pancreatic duct stricture, pancreatic duct dilatation), and periampullary carcinomas (those that arise from the head of the pancreas, the distal common bile duct, and within 2 cm of the major papilla in the duodenum) detected on MRCP who were subsequently evaluated by ERCP.
The exclusion criteria for our study were patients with absolute contraindications to the MRCP technique, such as patients with incompatible implants, patients with claustrophobia, patients with CBD, and pancreatic duct pathologies, in whom therapeutic interventions were not indicated, any case that developed complications during the procedure and all cases of failed cannulation.
CBD dilatation in adults is defined as a common bile duct measuring ≥8 mm at its widest part. A dilatation of 8 mm-12 mm is mild, 12 mm-16 mm moderate, 16 mm-20 mm severe, and >20 mm extremely severe.
CBD strictures are narrowing segments of the intrahepatic or extrahepatic biliary ductal system. The narrowing impedes the normal antegrade flow of bile, causing proximal dilatation.
A dilated main pancreatic duct was diagnosed when it measured >3 mm in the head and <2 mm in the body and tail regions of the pancreas.
This study protocol was approved by Institutional Ethics Committee Apollo Hospitals, Bhubaneswar (registration no. ECR/246/Inst/OR/2013/RR 2016 on 12 April 2019). The studies were conducted in MRI and ERCP units.

MRCP and MRI techniques
In this study, an eight-channel 1.5-T MRI scanner (Signa HDxt; GE Healthcare) using torso phased-array coils was used for the MRCP techniques. Pre-procedural preparation included 6 h of fasting to promote gallbladder filling.
(1) Three phase gradient-echo localising images were obtained and used to plan the MRCP sequences. All the sequences were required during a single breath-hold. The entire examination was usually completed within 20 min. MRCP was performed before ERCP and the results were evaluated by senior consultant radiologists.

ERCP technique
ERCP was performed using an Olympus CV150 duodenoscope, and fluoroscopic images were obtained using a Philips BV Libra system. Pre-procedural preparation included at least 12 h of fasting. Omnipaque TM (iohexol) contrast was used and the procedure was performed under anaesthesia using propofol. ERCP was performed by a well-trained and experienced endoscopist. Cholangiograms were obtained.

Statistical analysis
Data were collected on different pancreato-biliary pathologies. The diagnostic test evaluation calculator MEDCALC software was used to calculate the outcome measures of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of MRCP with reference to ERCP as the gold standard in respect to different pancreato-biliary pathologies. A chi-square test was used to compare both groups (i.e., cholangiography and pancreatography) using IBM SPSS Statistics 24.0 software (SPSS South Asia Pvt. Ltd).

Results
A total of 60 patients (46.7% men, 53.3% women; age range 23-79 years; mean age 54.35 ± 29.62 years) were included in the study. The majority of cases (46.7%) belonged to the 50-60-year age group, with nearly one-quarter belonging to the 65 years and older age group.

Choledocholithiasis
On MRCP, the location, size, and number of stones were in accordance with the ERCP examination. The stone size was <5 mm in 59% of cases, while stones >5 mm were observed in 41%. MRCP correctly diagnosed 37 of 42 patients with ERCPproven CBD calculi and 17 of 18 patients without calculi. Of the 60 cases, 37 were true positives, one was a false positive, five were false negatives, and 17 were true negatives.
In our study, the bile duct diameter (mean = 8.6 mm) was measured in the setting of choledocholithiasis. Of the five falsenegative MRCP cases, all had dilated bile ducts >10 mm on ERCP.

Stricture of the pancreatic duct
Of the 10 pancreatic duct stricture cases diagnosed by ERCP, MRCP correctly diagnosed eight cases, and of 50 unaffected patients, MRCP excluded 49 cases. The numbers of true-positive, true-negative, false-positive, and false-negative cases reported by Kumar

Periampullary carcinoma
MRCP correctly diagnosed 9 out of 10 cases diagnosed by ERCP. MRCP correctly excluded 50 cases out of 50 unaffected cases. The numbers of true-positive, false-positive, true-negative, and falsenegative cases were 9, 0, 50, and 1, respectively. With reference to ERCP, the sensitivity was high at 90% (95% CI: 55.50-99.75) and the specificity was high at 100% (95% CI: 92.89-100.00). The PPV, NPV, and accuracy were 100%, 98.04%, and 98.33%, respectively. The correlation of MRCP and ERCP in the evaluation of the six different pathologies is shown in Table 1.
Choledocholithiasis was found in 63% of cases of MRCP and 70% of cases of ERCP. MRCP and ERCP detected CBD strictures in 18% and 20% of cases, respectively. Both techniques detected  Frontiers in Medical Technology pancreatic duct dilatation in 18% of cases each and pancreatic duct stricture in 15% and 16% of cases, respectively. Periampullary carcinoma was diagnosed in 15% and 16% of MRCP cases, respectively. There was no significant difference between MRCP and ERCP in the evaluation of these pathologies (Figures 1 and 2).

Discussion
In our study, the elderly age group (>50 years) is commonly affected by these pathologies (p = 0.001), which is consistent with the study conducted by Kimura et al. (8). The number of affected women is slightly higher than that of affected men in the present study, which correlates with the study conducted by Ko et al. (9). According to O'Connor et al., choledocholithiasis and CBD dilatation are some of the most common biliary tract pathologies (1), which matches the findings of our study.
In this study, MRCP correctly diagnosed 37 of 42 patients with ERCP-proven CBD calculi and 17 of 18 patients without calculi. The location, size, and number of stones were consistent with the ERCP examination. According to Griffin et al. (10), compared with ERCP, MRCP has a sensitivity, specificity, PPV, and NPV of 86%, 93%, 87%, and 82%, respectively, in diagnosing choledocholithiasis. According to Vitellas et al. (6), MRCP is comparable to ERCP in diagnosing choledocholithiasis, with sensitivities and specificities in the range of 81%-100% and 85%-100%, respectively. In another study, the sensitivity of MRCP and ERCP for identifying choledocholithiasis was 80% and 90%, respectively. The overall agreement between MRCP and ERCP was 90.6% (11).
The results of our study show that the diagnostic accuracy of MRCP in detecting choledocholithiasis was 90% and comparable to the results of previous studies (6, 10-12). A false-negative diagnosis had occurred, as multiple small (2 mm) intrahepatic duct stones were missed on the MRCP. Stones were probably missed because of the lack of contrast between the stones and surrounding liver, with no high signal bile outlining the stones.
It is known that the sensitivity of MRCP for detecting choledocholithiasis decreases with bile duct dilatation (72% for bile duct diameter >10 mm vs. 88.9% for diameter ≤10 mm) (11,13). In our study, five false-negative cases on MRCP had dilated bile ducts >10 mm on ERCP examination.
According to Chan et al., in diagnosing CBD dilatation, MRCP showed a sensitivity of 95%, specificity of 85%, PPV of 82%, and NPV of 96% compared to ERCP (14). Hintze et al. found the sensitivity and PPV of MRCP in detecting bile duct   (18). Takehara et al. also found an agreement of 83%-92% for diagnosing pancreatic duct dilatation by MRCP and ERCP (19). In our study, the sensitivity, specificity, PPV, NPV, and accuracy of MRCP in detecting pancreatic duct dilatation were in accordance with the results of the abovementioned studies.
According to Sugita et al. (21), the sensitivity, specificity, accuracy, PPV, and NPV of high-resolution MRI for the evaluation of periampullary carcinoma were 88%, 100%, 96%, 100%, and 94%, respectively. They observed that MRCP can accurately detect the location, extension, and origin of periampullary carcinoma and is beneficial in the preoperative staging of tumours.
According to Pamos et al. (22), the sensitivity and specificity of MRCP compared to ERCP in diagnosing periampullary carcinoma were 100% and 83%, respectively. In our study, the results for MRCP in diagnosing periampullary carcinoma were high (sensitivity = 90%, specificity = 100%) and comparable to the results of the above-mentioned studies. MRCP exhibited a high level of diagnostic precision for obstructive jaundice.
According to the most recent studies, MRCP has the potential to replace diagnostic ERCP in a wide variety of bile duct anomalies (bile stones, benign and/or malignant strictures, CBD), which would be a significant advancement in the field as this could reduce the frequency of invasive procedures undergone by the patient. This would curtail the occurrence rate of potential complications associated with ERCP (4, 12).
The present study has some limitations. A small sample size is an important limitation. This is due to the limited duration of the study and the difficulty in finding cases that underwent both MRCP and ERCP, as only those patients with therapeutic interventions indicated were subjected to ERCP examinations. There are various CBD and pancreatic duct pathologies. However, in our study, because of the above-mentioned limitations, we could evaluate the diagnostic accuracy of only six pathologies.

Conclusion
MRCP, with reference to ERCP, has high sensitivity, specificity, PPV, NPV, and accuracy, with no significant difference in diagnosing those six pathologies. In the case of stricture of the CBD and pancreatic duct, MRCP can accurately reveal the location and extent of stricture. MRCP offers the additional advantage of cross-sectional imaging in cases of periampullary neoplasm and can accurately evaluate tumour invasion into surrounding tissue.

Data availability statement
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author.

Ethics statement
The studies involving human participants were reviewed and approved by ECR/246/Inst/OR/2013/RR-2016 on 12 April 2019 at Apollo Hospital, Bhubaneswar. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.