暂无摘要。

To improve the efficiency of pathological diagnoses, the development of automatic pathological diagnostic systems using artificial intelligence (AI) is progressing; however, problems include the low interpretability of AI technology and the need for large amounts of data. We herein report the usefulness of a general-purpose method that combines a hyperspectral camera with machine learning. As a result of analyzing bile duct biopsy and bile cytology specimens, which are especially difficult to determine as benign or malignant, using multiple machine learning models, both were able to identify benign or malignant cells with an accuracy rate of more than 80% (93.3% for bile duct biopsy specimens and 83.2% for bile cytology specimens). This method has the potential to contribute to the diagnosis and treatment of bile duct cancer and is expected to be widely applied and utilized in general pathological diagnoses.

BACKGROUND: The scoring system for bile cytology (SSBC) aims to improve bile cytology diagnostic accuracy. Here, the practicality of SSBC was verified by multiple cytotechnologists.
METHODS: Bile cytological specimens were evaluated by 24 cytotechnologists using SSBC. The samples were assessed before using the SSBC (first-time assessment) according to three categories: benign, indeterminate, and malignant. A first scoring evaluation (FSE) was then performed using SSBC; each item in the scoring system was classified as present or absent. After distributing an instruction sheet with diagnostic criteria, a second scoring evaluation (SSE) was performed using SSBC. Each method was evaluated using diagnostic accuracy and interobserver and intraobserver agreement.
RESULTS: Several samples were assessed as indeterminate in the first-time assessment. Although the specificity of the SSE improved, the sensitivity and accuracy decreased compared with those of the FSE. The overall interobserver agreement was fair for all parameters, including abnormal chromatin, irregular internuclear distances, irregularly overlapped nuclei, irregular cluster margins, and final evaluation in the FSE and SSE. The final evaluation by histological type exhibited slight agreement for well-differentiated tubular adenocarcinoma and almost perfect agreement for poorly differentiated tubular adenocarcinoma in the FSE and SSE. For moderately differentiated tubular adenocarcinoma, agreement was moderate in the FSE and fair in the SSE. For cholangitis, a slight agreement was observed in the FSE, which improved to fair in the SSE.
CONCLUSIONS: Although the SSBC is expected to improve specificity, there exists ambiguity regarding SSBC criteria and interindividual assessment differences. Therefore, the objective assessment method should be revised.

BACKGROUND: In primary sclerosing cholangitis (PSC), it is important to understand the cholangiographic findings suggestive of malignancy, but it is difficult to determine whether cholangiocarcinoma is present due to modifications caused by inflammation. This study aimed to clarify the appropriate method of pathological specimen collection during endoscopic retrograde cholangiopancreatography for surveillance of PSC.
METHODS: A retrospective observational study was performed on 59 patients with PSC. The endpoints were diagnostic performance for benign or malignant on bile cytology and transpapillary bile duct biopsy, cholangiographic findings of biopsied bile ducts, diameters of the strictures and upstream bile ducts, and their differences.
RESULTS: The sensitivity (77.8% vs. 14.3%, P = 0.04), specificity (97.8% vs. 83.0%, P = 0.04), and accuracy (94.5% vs. 74.1%, P = 0.007) were all significantly greater for bile duct biopsy than for bile cytology. All patients with cholangiocarcinoma with bile duct stricture presented with dominant stricture (DS). The diameter of the upstream bile ducts (7.1 (4.2-7.2) mm vs. 2.1 (1.2-4.1) mm, P < 0.001) and the diameter differences (6.6 (3.1-7) mm vs. 1.5 (0.2-3.6) mm, P < 0.001) were significantly greater in the cholangiocarcinoma group than in the noncholangiocarcinoma group with DS. For diameter differences, the optimal cutoff value for the diagnosis of benign or malignant was 5.1 mm (area under the curve = 0.972).
CONCLUSIONS: Transpapillary bile duct biopsy should be performed via localized DS with upstream dilation for the detection of cholangiocarcinoma in patients with PSC. Especially when the diameter differences are greater than 5 mm, the development of cholangiocarcinoma should be strongly suspected.

The specimen collection and subsequent pathological diagnosis of malignant biliary stricture (MBS) are difficult. This study aimed to determine whether the cell block (CB) method using overnight-stored bile is useful in the diagnosis of MBS. This trial was a single-arm prospective study involving a total of 59 patients with suspected MBS. The primary endpoint was cancer detectability and accuracy using the CB method, and a comparison with the detectability and accuracy achieved with bile cytology was made. The immunohistochemical sensitivity for maspin and p53 was also investigated in the CB and surgical specimens. We were able to collect bile from all 59 patients, and 45 of these patients were clinically diagnosed with MBS. The cancer detectability using the CB method (62.2%) was significantly higher than that using cytology (37.8%) (p = 0.0344). When CB was combined with biopsy, the rates of cancer detectability (75.6%) and accuracy (81.4%) increased. In eight patients who received surgical therapy, maspin- and p53-immunohistochemistry was applied to the surgical and CB specimens, and cancer cells in both specimens showed positive cytoplasmic and nuclear staining for maspin and nuclear staining for p53. The CB method is, thus, useful for detecting malignancy (UMIN000034707).

BACKGROUND: Bile cytology is useful in diagnosing biliary tract lesions, albeit often challenging due to equivocal findings. To achieve better diagnoses for clinical decisions, we conducted cytomorphological and immunocytochemical studies of bile cytology cases.
METHODS: We re-evaluated 40 bile cytology cases with initial equivocal diagnoses, taken from the cytology records of Jichi Medical University Hospital, including 1778 bile cytology specimens. First, we assessed the cases by the diagnostic bile cytology criteria of the Japanese Society of Clinical Cytology. Second, we searched for useful immunocytochemical markers by extensive immunohistochemical analyses using tissue microarray for 10 antibodies: S100P, IMP3, GLUT1, p53, S100A4, Mapsin, MUC17, CD10, MDM2, and SMAD4. Microarrays were from 257 extrahepatic bile duct carcinoma cases. To elucidate the utility of immunocytochemistry, we applied selected markers to immunocytochemical evaluation of the equivocal cases after cell transfer.
RESULTS: The criteria indicated a sensitivity 60%, specificity 87%, and accuracy 70%. Irregularly overlapping (88%), arranged (96%), and shaped (76%) nuclei were more common in malignant cases, while enlarged nuclei were more frequent in benign cases (67% vs. 28%). We applied S100P and IMP3, which showed higher accuracy (88% and 77%) in tissue microarray, to immunocytochemistry. The sensitivity of S100P and IMP3 were 69% and 70%, respectively. The specificity of S100P and IMP3 were 50% and 100%, respectively.
CONCLUSIONS: The criteria showed a certain effectiveness even in challenging cases, and some pitfalls associated with reactive changes of benign cells. Although comprehensive diagnosis including cytomorphology seems preferable, S100P and IMP3 are promising immunocytochemical markers.

A new device with metallic wires for scrape cytology was developed.
To compare the diagnostic performance of scrape cytology and conventional cytology during endoscopic retrograde cholangiopancreatography for biliary strictures.
A total of 420 cases with biliary stricture underwent transpapillary bile cytology. Among them, there are 79 cases with scrape cytology using the new device (scrape group) and 341 cases with conventional cytology (control group). Seventy-two and 174 cases underwent biliary biopsy at the same time as bile cytology in the scrape and control group, respectively.
The sensitivity for malignancy of bile cytology in the scrape and control group was 41.2% [pancreatic cancer (PC): 23.1%, biliary cancer (BC): 52.5%] and 27.1% (PC: 16.3%, BC: 38.0%), respectively (P = 0.023). When analyzed PC and BC, respectively, there was no significant difference between the two groups. In the both groups, the sensitivity was significantly higher for BC than PC. In the scrape group, there was no difference in the sensitivity between cytology and biopsy [39.7% (PC: 17.4%, BC: 55.3%)], but in the control group, a significantly lower sensitivity was observed with cytology than biopsy (36.4% (PC: 19.7%, BC: 50.0%)) (P = 0.046). When analyzed PC and BC, respectively, there was no significant difference between cytology and biopsy. The sensitivity of combined cytology and biopsy was 55.6% (PC: 30.4%, BC: 71.1%) in the scrape group and 47.0% (PC: 24.6%, BC: 64.3%) in the control group.
Scrape bile cytology for biliary strictures may be superior to conventional cytology.

The sensitivity of bile cytology for biliary tract cancer varies from 6-64%, and hence remains unsatisfactory. Sialylated carbohydrate antigen KL-6 mucin is positive in biliary tract cancer tissues and serum KL-6 levels are significantly increased in intrahepatic ductal adenocarcinoma patients compared with healthy individuals. The aim of the present study was to evaluate the usefulness of the KL-6 concentration of bile for the diagnosis of biliary tract cancer. Bile cytology and measurements of bile KL-6 concentration were conducted for 43 patients (25 biliary tract cancers and 18 benign biliary disease). The concentration of KL-6 in the bile of the biliary tract cancer group was compared with the benign biliary disease group. The diagnostic ability was assessed by using receiver operating characteristic curves (ROC). The mean KL-6 concentration of bile for biliary tract cancer (34.6±51.6 U/ml) was increased compared with benign biliary disease (5.2±3.9 U/ml, P<0.001). The area under the ROC for diagnosis of biliary tract cancer was 0.84 for benign biliary disease. When the cut-off level of the KL-6 concentration of bile was 8.6 U/ml, the sensitivity, specificity, and accuracy of the KL-6 concentration of bile alone for the diagnosis of biliary tract cancer were 72, 89, and 79%, respectively. Adding the bile KL-6 concentration to bile cytology measurements, the sensitivity for the diagnosis of biliary tract cancer was increased significantly (100%, P=0.0184). The KL-6 concentration of bile may strengthen the sensitivity of bile cytology for biliary tract cancer.

BACKGROUND: Obstructive jaundice is frequently caused by bile duct strictures. Determination of malignant strictures is crucial for the initiation of appropriate treatment. Cytologic examination of bile drainage fluid is an easy and reproducible method of detecting malignant cells. This method, however, frequently yields indeterminate results, such as atypia or suspicious of malignancy, due to difficulties in differentiating malignancy from benign atypia. Immunocytochemical assessment of p53 expression by cells in bile drainage fluid may enhance the ability to detect malignancy.
METHODS: A total of 139 samples of bile drainage fluid were obtained from 80 patients. Following cytologic examination, the samples were incubated with antibody to p53. The performance of cytology with and without p53 immunocytochemistry was evaluated, with reference to surgical or clinical findings of benign and malignant biliary strictures.
RESULTS: Bile drainage cytology alone had a sensitivity of 31.6% and a specificity of 98.4% in the identification of malignant strictures, whereas the combination of p53 immunocytochemistry and bile drainage cytology had a sensitivity of 80.3% and a specificity of 92.1%. P53 immunocytochemistry alone had a sensitivity of 64.5% and a specificity of 92.7% for the identification of malignant strictures in bile drainage samples with atypical cytology, and a sensitivity of 85.0% and a specificity of 100.0% in samples with suspicious of malignancy.
CONCLUSIONS: The addition of p53 immunocytochemistry to bile drainage cytology can be useful in identifying malignant strictures in samples showing indeterminate results on bile drainage cytology. Diagn. Cytopathol. 2017;45:592-597. © 2017 Wiley Periodicals, Inc.