19 J Med Dent SciN. Abe et al.J Med Dent Sci 2016； 63： 19－27

Endoscopic retrograde cholangiopancreatography (ERCP) brushing cytology often cannot distinguish adenocarcinoma from reactive epithelial changes. We attempted to improve the diagnostic sensitivity of ERCP using the following methods: systematic cytological evaluation, immunocytochemical examination of minichromosome maintenance proteins (MCM) 2 and p53, and a combination of these methods. ERCP specimens from 53 patients (13 benign and 40 malignant cases) were studied. First, we reclassified the cases into three categories according to the systematic cytological evaluation: negative, suspicious, and positive. Secondly, immunocytochemistry was performed for MCM 2 and p53. The cut-off values were set at 25% labeling index (LI) for MCM 2 and 10% LI for p53, respectively. We evaluated the sensitivity, specificity, and diagnostic accuracy. The sensitivity of the systematic cytological evaluation alone did not improve significantly, compared with the original screening examination (77% vs. 68%). The sensitivity of immunocytochemistry for MCM 2 and

p53 was 90% (P < 0.05) and 68%, respectively. Applying only the suspicious or positive categories, the sensitivity improved significantly to 93% for the combination of systematic cytological evaluation and immunocytochemistry for MCM 2 and p53 (P < 0.01). In conclusion, the combination of morphology and immunocytochemistry for MCM 2 and p53 may help to overcome the diagnostic cytological difficulties of pancreaticobiliary adenocarcinoma.

Key Words: ERCP brush cytology, systematic cytological evaluation, immunocytochemistry, minichromosome maintenance protein 2, p53.

Introduction

Pancreaticobiliary cancer is characterized by rapid progression and a poor prognosis. Therefore, quick diagnosis and treatment are required to improve the prognosis. Pancreaticobiliary cancer is usually diagnosed by brushing cytology or biopsy obtained during endoscopic retrograde cholangiopancreatography (ERCP). Although brushing cytology is important for establishing a diagnosis, previous studies have repeatedly indicated a low sensitivity of ERCP brushing cytology ranging widely from 19% to 78%, despite nearly a 100% specificity.1–6 This low sensitivity can be ascribed to two reasons: (1) sampling errors during ERCP brushing, resulting in only a few observable cells

Corresponding Author: Naoya Abe, MS, CT.Department of Pathology, Japanese Red Cross Medical Center, Tokyo, Japan. 4122 Hiroo, Shibuyaku, Tokyo 1508935, Japan Tel: +81334001311　Fax: +81334091604Email: byori@med.jrc.or.jpReceived August 18；Accepted December 25, 2015

Original Article

Systematic cytological evaluation and immunocytochemistry of minichromosome maintenance protein 2 and p53 significantly improve cytological diagnosis of pancreaticobiliary adenocarcinoma

Naoya Abe1), 5), Kozue Matsuo1), Toshio Kumasaka1), Koichi Naka1), Syoichi Hashimoto1), Tamiko Takemura1), Mutsunori Fujiwara2), Yukiko Ito3), Ryo Nakata3), Takuya Hashimoto4), Masatoshi Makuuchi4), Yurie Soejima5) and Motoji Sawabe5)

1) Departments of Pathology, Japanese Red Cross Medical Center, Tokyo, Japan2) Departments of Laboratory Medicine, Japanese Red Cross Medical Center, Tokyo, Japan3) Departments of Gastroenterological Medicine, Japanese Red Cross Medical Center, Tokyo, Japan4) Departments of Hepatobiliarypancreatic Surgery and Transplantation Surgery, Japanese Red Cross Medical Center, Tokyo, Japan5) Section of Molecular Pathology, Graduate School of Health Care Sciences, Tokyo Medical and Dental University, Tokyo, Japan

20 J Med Dent SciN. Abe et al.

on the glass slide; and (2) underdiagnosis caused by mild atypia, which may be visible in both benign reactive lesions and welldifferentiated adenocarcinoma.7 The latter problem always exists, since more than 90% of pancreaticobiliary cancer is adenocarcinoma. Hirooka et al. proposed an effective solution using a systematic cytological evaluation of pooled bile samples.8 This systematic evaluation was based on diagnostic criteria deduced from cytological findings collected from a multiinstitutional study group for bile cytology in Japan. Using this systematic cytological evaluation, cytotechnologists can diagnose adenocarcinoma based on the presence of atypical arrangements of cell clusters and nuclear atypia in the tumor cells. Wakasa et al. reported an improved sensitivity and accuracy of ERCP brush cytology using this systematic evaluation.1

Furthermore, advanced diagnostic techniques such as an immunofluorometric assay,3 a DNA methylation assay,9 and fluorescence in situ hybridization (FISH)4, 5 could offset the low sensitivity of cytology. Recently, the Papanicolaou Society of Cytology published the guidelines about these ancillary tests on pancreaticobiliary cytology.10 However, not all laboratories can introduce these advanced methods, since they require special equipment and highcost reagents. On the other hand, immunocytochemistry is more applicable and easier than these methods. Kim et al. reported an improved sensitivity for ERCP brush cytology using immunocytochemistry and recommended the routine use of p53 immunostaining in the evaluation of ERCP cytology.6 However, another study stated that p53 was not useful for cytological diagnosis, as it failed to provide a definite cutoff value for immunocytochemistry.11 The p53 overexpression due to genetic mutation causes cancers in many organs including pancreaticobiliary cancer. The frequency of p53 mutation was around 50% in pancreaticobiliary cancer.12, 13 Here we validate the usefulness of p53 immunocytochemical analysis in pancreaticobiliary cancer.

Another interesting tumor marker, in addition to p53, is the family of minichromosome maintenance (MCM) proteins.14 MCM proteins are intranuclear proteins that play important roles in the initiation of DNA replication through the activity of DNA helicase and act in a hexamer complex comprised of the MCM 2 through MCM 7 subunits.15 The contribution of MCM proteins to carcinogenesis has not yet been completely clarified, but MCM proteins are expressed during cell proliferation. The intranuclear localization of MCM 2 protein suggests that MCM 2 might be associated with DNA replication foci16 or that it might be the fastest phosphorylation

substrate for upper kinases (i.e., Cdc7, CDK), compared with other subunits, during the activation of the MCM complex.15 Therefore, MCM 2 may act as the functional center of MCM proteins, reflecting cell proliferation leading to carcinogenesis. We selected this marker MCM 2 because its mechanism was already clarified, and no immunocytochemical analysis for MCM 2 had been performed in pancreaticobiliary cancer.

In this study, we attempted to improve the sensitivity of ERCP brush cytology. First, we took a morphological approach to review cytological smears through a systematic cytological evaluation. Second, we took an objective approach to confirm the efficacy of MCM 2 and p53 immunocytochemistry, calculating reasonable cutoff values for the diagnosis of pancreaticobiliary adenocarcinoma.

Materials and Methods

PatientsThe subjects consisted of 53 histologically confirmed

cases of pancreaticobiliary adenocarcinoma diagnosed among 173 patients who underwent ERCP cytology between 2010 and 2012 at the Japanese Red Cross Medical Center. These 53 cases were comprised of 19 biopsy cases, 15 surgical cases, and 19 cases that had undergone both biopsy and surgery. The specimens included 40 malignant cases (23 cholangiocarcinomas, 15 pancreatic adenocarcinomas, and 2 gallbladder adenocarcinomas) and 13 benign cases (4 benign epitheliums, 7 chronic inflammations of the pancreaticobiliary tract, and 2 others). The patients were comprised of 33 men and 20 women with a median age of 73 years (range, 45–86 years). Benign cases were clinically followed up for a median of 35 (range 5–60) months. Informed consent for the medical procedure and the use of specimens for research was obtained from each patient. This research was approved by the ethical committee of the Japanese Red Cross Medical Center (No. 377).

Cytological specimens were prepared according to a modified Wakasa’s method.1 Briefly, the specimens were obtained from brush and fluid sediments collected from the ERCP brush sheath. These sediments were obtained by flushing the sheath in physiological salt solution and fixing in 95% ethanol. Histological slides were prepared from the usual formalinfixed and paraffinembedded blocks.

21MCM 2 and p53 immunocytochemistry in ERCP cytology

Retrospective reevaluation of cytology results using a systematic cytological evaluation

The Papstained smears were reviewed using a systematic cytological evaluation procedure proposed by Hirooka et al.8 The diagnostic criteria are shown in Table 1. We classified the specimens into three categories: negative, suspicious, and positive (adenocarcinoma). These categories corresponded to the numbers of positive findings of arrangements of atypical cell clusters or nuclear atypia: none or one (negative), two (suspicious), and three (positive), respectively. Each specimen was blindly and independently reviewed by two cytotechnologists, and cases with different diagnoses were discussed to reach a consensus diagnosis.

Immunocytochemical examinationImmunocytochemistry was performed using the cell

transfer method and the universal immunoperoxidase polymer method. The cell transfer was performed according to the standard protocol.17, 18 Briefly, we immersed the stained slides in xylene, removed the coverslip, mounted the specimens in 1 mL of Malinol,

dried them overnight, softened the Malinol by soaking it in a 40°C hot water, peeled the cells embedded in the Malinol sheet from the glass slide, cut the sheet into a few pieces, and put the pieces on coated glass slides (Immunocoat YS; Muto Pure Chemicals Co., Ltd., Tokyo, Japan). Immunocytochemistry for both MCM 2 and p53 was then performed. The specimens were treated with 3% hydrogen peroxide for 10 min to block endogenous peroxidase activity. Then, the slides were autoclaved in 10–mM citrate buffer (pH6.0) at 121°C for 10 min for antigen retrieval. The specimens were washed using PBS buffer with 0.1% Tween 20 and were incubated overnight at 4°C with antiMCM 2 (phospho Ser 27) rabbit monoclonal antibody (clone EPR4119, 1 : 250 dilution; GeneTex, Irvine, CA) or anti–p53 mouse monoclonal antibody (clone DO–7, 1 : 100 dilution; DAKO, Carpinteria, CA). Then, the sections were incubated using universal immunoperoxidase polymer solutions (Histofine Simple Stain MAXPO; Nichirei, Tokyo, Japan) for 30 min. Peroxidase activity was detected using DAB for 10 min. Finally, the nuclei were counterstained with Gill’s hematoxylin. For the immunohistochemistry, the

Table 1 Diagnostic criteria of bile cytology

A. Large clusters

1. Irregularly overlapped nuclei

2. Irregularly arranged nuclei

3. Irregularly cluster margin

B. Small clusters and isolated cell

1. Enlarged nuclei

2. Irregularly shaped nuclei

3. Abnormal chromatin

C. Other notable findings

1. Necrotic background

2. Varying cell cluster size

D. Points to note

1. Do not make a judgment from limited abnormalities.

2. Even if morphological changes occur when bile juice is left as it is, cytological diagnosis may be possible by observing the nuclear structure.

3. Benign tumors and normal tissue display cytological features of equal internuclear distance and regularly arranged cytoplasm at cluster margins.

Cells or clusters are diagnosed as positive for adenocarcinoma if three factors are met for A or B.According to Hirooka et al. Report of the clinical study for methods to improve the diagnostic accuracy of bile cytology. Jpn Soc Clin Cytol. 2010; 49: 7–14.8

22 J Med Dent SciN. Abe et al.

sections were deparaffinized in xylene and similarly immunostained. The immunocytochemical staining was evaluated by calculating the labeling index (LI) (%) of positive nuclei. At least 200 epithelial nuclei were counted per case for both immunostainings among the cases with suspicious or malignant results.

Finally, we evaluated the efficacy of immunohistochemical examination by the combined diagnosis using the results of systematic cytological evaluation and immunohistochemical study. Considering the nature of immunocytochemical study as the ancillary analysis of cytological examination, we applied the combined diagnosis only to the suspicious and positive cases.

Statistical analysisThe sensitivity, specificity, and accuracy of the

cytological evaluations were calculated and compared between each evaluation using the χ2 test. Suspicious cases were included with the negative cases. The MCM 2 and p53 labeling indexes were compared between the benign and malignant groups using the MannWhitney Utest. The correlation between the immunocytochemistry and immunohistochemistry results

was analyzed using the Spearman rank correlation test. To obtain the point with highest sensitivity and lowest false positivity, namely the optimal cutoff values for MCM 2 and p53 for distinguishing benign and malignant cells were calculated using the receiver operating characteristics (ROC) curves. In MCM 2 and p53 immunocytology, the results were considered positive when either MCM 2 or p53 was positive. All the statistical analyses were performed using StatMate III software (ATMS, Tokyo, Japan). A value of P < 0.05 was considered statistically significant.

Results

Retrospective review of cytology results using a systematic cytological evaluation

Representative cytological images are shown in Figure 1. Cluster of cells diagnosed as “suspicious of malignant” had the enlarged irregularlyshaped nuclei lacking crowded chromatin, being present in the normally arranged cluster of cells (Figure 1B). Large cluster of cells diagnosed as “malignant” showed the atypical arrangement of large clusters containing over 30 cells:

Figure 1 Cytological features of ERCP brushing cytology(A) Cluster of benign epithelium cells. (B) Cluster of cells diagnosed as Suspicious of malignant. (C) Atypical arrangement of large clusters containing over 30 cells diagnosed as Malignant. (D) Small cluster with nuclear atypia of each cells diagnosed as Malignant. (Pap-stain, ×400)

23MCM 2 and p53 immunocytochemistry in ERCP cytology

They had irregularly overlapped nuclei with disarrayed polarity or disproportionate intracellular distances (irregularly arranged nuclei) and nuclear protrusion from the cell cluster or a complicated divergence of clusters (irregular cluster margins) (Figure 1C). Small cluster of cells diagnosed as “malignant” showed the nuclear atypia of each cell with doublesized (enlarged nuclei) or distorted nuclei (irregularly shaped nuclei) and overcrowded and unevenly distributed chromatin patterns (abnormal chromatin) (Figure 1D). Table 2 shows a summary of the systematic cytological evaluation results. The systematic cytological evaluation improved the sensitivity from 68% to 77% and the accuracy from 75% to 83%. The recognition of two cytological important factors, irregularly arranged nuclei (A–2, Table 1) and enlarged nuclei (B–1, Table 1), seemed to lead to a more accurate diagnosis (data not shown).

Immunocytochemical examinationImmunocytochemical staining for MCM 2 and p53 was

positive in the nuclei of epithelial cells (Figure 2). The benign cells lacked p53 expression (Figure 2A, E), but the malignant cells showed high expression of p53 (Figure 2B, F). MCM 2 immunostaining showed low expression in benign cells (Figure 2C, G) and high expression in malignant cells (Figure 2D, H). Table 3 shows the labelling indexes of MCM 2 and p53 for the immunocytochemistry and immunohistochemistry evaluations. The median MCM 2 and p53 labelling indexes of the malignant cases were significantly higher than those of the benign cells using both immunocytochemistry and immunohistochemistry, but MCM 2 positivity in the benign cases was not very low (16%) for both the immunocytochemistry and immunohistochemistry findings. One case with chronic pancreatitis showed high expression of p53 probably due to marked inflammation or nonspecific reaction by drying the specimen before fixation. Positive correlations

Table 2 Comparison of cytological evaluations between the original screening and the systematic cytological review

Histological diagnosis

Cytological diagnosis (n) Sensitivity Specificity

Accuracy (95% CI)Positive Suspicious Negative

Original screening examinationMalignant 27 12 1 68% (54%–82%)

100% 75% (63%–87%)Benign 0 6 7

Systematic cytological evaluationMalignant 31 6 3 77% (64%–90%)

100% 83% (73%–93%)Benign 0 4 9

CI, confidence interval

Figure 2 Results of immunostaining for p53 and MCM 2Immunostaining for p53 (A, B, E, F) and MCM 2 (C, D, G, H) using immunocytochemical (ICC; A–D) and immunohistochemical (IHC; E–H) stainings. Paired photos (E–G and F–H) were taken at the same site of the serial paraffin sections. (All images, ×400)

24 J Med Dent SciN. Abe et al.

were obtained between the immunocytochemistry and immunohistochemistry results for MCM 2 (ρ = 0.34, P < 0.05) and for p53 (ρ = 0.51, P < 0.001).

The performance of the MCM 2 and p53 immunocytoche

mistry as a diagnostic test for pancreaticobiliary malignancy is shown as an ROC curve in Figure 3. The tests were capable of discriminating patients with adenocarcinoma from those without; the areas under the ROC curves were 92% (95% CI, 85%–98%) for MCM 2 and 79% (95% CI, 69%–89%) for p53. Appropriate cutoff values were then determined using these ROC curves. To distinguish adenocarcinoma cells and benign cells, the optimal cutoff value for MCM 2 was LI = 25%, while that for p53 was LI = 10%. The cytological smears were judged as being positive or negative according to these cutoff values.

Table 4 shows the results for the sensitivity, specificity, and accuracy of MCM 2 immunocytochemistry, p53 immunocytochemistry, and a combined diagnosis based on a systematic cytological examination and MCM 2 and p53 immunocytochemistry. The 10 suspicious cases diagnosed by the systematic cytological evaluation comprised of 2 positive cases of MCM 2, 1 positive case of p53, 4 positive cases of both MCM 2 and p53, and 3 negative cases. The sensitivity of MCM 2 immunocytochemistry was relatively high (90%), as was the specificity of p53 (92%). We obtained a high

Figure 3 Receiver operating characteristic curve of an immunocytochemical analysis based on MCM 2 and p53 immunocytochemistry

The area under the curve for MCM 2 (%) was 92% (95% CI, 85%–98%), while that for p53 was 79% (95% CI, 69%–89%).

Table 3 Labelling indexes of p53 and MCM 2

Median labelling index, % (range)P values

Malignant cases Benign cases

ICCp53 40% (0%–100%) 2% (0%–11%) 0.004

MCM 2 65% (2%–96%) 16% (0%–50%)

25MCM 2 and p53 immunocytochemistry in ERCP cytology

sensitivity (90%), but a low specificity (77%), for MCM 2 immunocytochemistry. This lower specificity might be attributable to 3 falsepositive cases of MCM 2 immunocytochemistry. No common pathological features were seen among these three cases. For the combined diagnosis based on a systematic cytological examination and MCM 2 and p53 immunocytochemistry positivity, we obtained a 100% sensitivity and a 93% accuracy, but the specificity decreased to 69%. This reduction in specificity resulted from the presence of 4 falsepositive diagnoses among the benign cases, consisting of 1 suspicious and 3 negative cytological diagnosis (Table 5). When the immunocytochemistry evaluations were applied to only the suspicious or positive cases in the systematic cytological evaluation, we obtained the best results: 93% (P < 0.01) for sensitivity, 92% (P = 0.31) for specificity, and 93% for accuracy.

Discussion

This study showed that a systematic cytological evaluation alone could not significantly improve the sensitivity and accuracy of ERCP brush cytology. However, MCM 2 and p53 immunocytochemistry showed higher expression levels in adenocarcinoma than in benign lesions, and we successfully obtained a suitable cutoff value for MCM 2 and p53. The combined evaluation of systematic cytology and MCM 2 and p53 immunocytochemistry resulted in a high sensitivity (93%) and specificity (92%) for distinguishing benign and malignant lesions.

Low sensitivity of ERCP cytologyERCP brush cytology plays a very important role for

pancreaticobiliary lesions, since physicians can decide on the therapeutic strategy depending on the results of the cytological diagnosis. However, the diagnosis is often difficult, and there are actually no standard cytological

methods with a high sensitivity. The sensitivity of our original screening was 68%.

This result was relatively good, compared with those of previous reports. Sampling errors have been suggested to be a major cause of the poor sensitivity of ERCP cytology.7 In most previous studies, a onetime direct smear of bile was used for the cytological sample.1–6 Wakasa et al. reported an improved sensitivity by collecting the cells from the ERCP brush sheath.1 They also indicated that smears should not be dried and should contain low amounts of mucin and blood. We used a modification of Wakasa’s method, which seemed to yield a high sensitivity.

Immunocytochemistry of MCM 2 and p53The overexpression of MCM protein is observed in

some malignant tumor tissues, such as cancers of the endocervix, breast, and anus.19–21 Various expressions of MCM immunocytochemistry proteins have been reported in nonmalignant lesions as well, including reactive mesothelial cells (9.8% ± 6.6%),22 lowgrade CIN of the endocervix (26.9%, range 0.5%–68.2%),19 and benign breast tissues (1%–35%).20 A previous study reported that MCM 2 was expressed in pancreaticobiliary benign tissue (5%, range 0%–33%).3 In line with these study results, we observed the mild expression of MCM 2 in benign cells (16%). Some types of nonneoplastic stimulation to the biliary epithelium might cause the expression of MCM proteins. Therefore, a higher cutoff point of 25% LI for MCM 2 immunocytochemistry may be valid for distinguishing benign cells from malignant cells. The correlation of MCM 2 and p53 positivity between immunocytochemistry and immunohistochemistry showed low to intermediate positive correlations. Therefore we consider that ERCP cytological samples could reflect the whole sample. p53 was also expressed more strongly in malignant cases than in benign cases but resulted in a low sensitivity (68%). Stewart et al.

Table 5 Immunocytochemical results with different histological and cytological diagnoses

Histological diagnosis

Immunocyto- chemical analyses

Cytological diagnosis [n / n (%)]

Positive Suspicious Negative

Malignant n = 40

p53 positive 21/31 (68%) 5/6 (83%) 1/3 (33%)

MCM 2 positive 28/31 (90%) 5/6 (83%) 3/3 (100%)

Benign n = 13

p53 positive 0/0 (0%) 0/4 (0%) 1/9 (11%)

MCM 2 positive 0/0 (0%) 1/4 (25%) 2/9 (22%)

MCM 2, minichromosome maintenance protein 2

26 J Med Dent SciN. Abe et al.

indicated that p53 immunocytochemistry alone is not useful for ERCP cytology.11 However, the combined use of MCM 2 and p53 immunocytochemistry and a systematic cytological evaluation dramatically increased the sensitivity. We recommend that MCM 2 and p53 immunocytochemistry should be combined with a systematic cytological examination for ERCP cytology.

In this study, we adopted the cell transfer method using previously stained slides because of the retrospective nature of this study. Other preparation methods, such as cellblock and liquidbased cytology (LBC), do exist. LBC can enable effective specimens to be produced even from the relatively small number of cells that are typically obtained using ERCP brushing or endoscopic ultrasoundguided fine needle aspiration (EUSFNA). LBC is also an apparently better method for immunocytochemistry and has now become a standard cytological preparation method. However, some laboratories cannot introduce this system because it is more costly than conventional preparations. Superior cytological results could potentially be obtained if LBC were to be utilized at our laboratory.

Study limitationThis study had an apparent selection bias. Benign

cases usually received only followup care, while most of the advanced malignant cases were treated with chemotherapy alone without surgery. Thus, only 31% (53 / 173 cases) of all the ERCP cytological specimens could be used in this study, and the number of benign cases (13 cases) was much lower than the number of malignant cases (40 cases).

ConclusionsWe showed that the combined use of systematic

cytological evaluation and MCM 2 and p53 immunocytochemistry was capable of improving the sensitivity of ERCP cytology. FISH and DNA assays might be attractive and objective examinations, but in terms of specificity, cytological evaluation seems to be a superior examination. The results of immunochemical analyses should be applied only to the suspicious or malignant cases because not a few benign cases show positive results. To obtain higherquality examinations, a prospective immunocytochemistry study examining large numbers of samples and ancillary marker antigens should be performed and the introduction of an LBC system should be considered in the future.

Acknowledgements

We are deeply thankful to all the staff members of the Department of Pathology, Japanese Red Cross Medical Center, for the preparation of the slides. We also wish to thank Ms. Mio Nakagawa for editing the English in this paper.

References

1. Wakasa A, Inayama K, Honda T, et al. Brushing cytology of the biliary tract: bile juice from the ERCP sheath tube provides cellrich smear samples. Acta Cytopthol. 2014; 58(4):398–405.

2. Renshaw AA, Madge R, Jiroutek M, et al. Bile duct brushing cytology: statistical analysis of proposed diagnostic criteria. Am J Clin Pathol. 1998; 110:635–640.

3. Ayaru L, Stoeber K, Webster GJ, et al. Diagnosis of pancreaticobiliary malignancy by detection of minichromosome maintenance protein 5 in bile aspirates. Br J Cancer. 2008; 98:1548–1554.

4. Smoczynski M, Jablonska A, Matyskiel A, et al. Routine brush cytology and fluorescence in situ hybridization for assessment of pancreatobiliary strictures. Gastrointest Endosc. 2012; 75:65–73.

5. Barr Fritcher EG, Caudill JL, Blue JE, et al. Identification of malignant cytologic criteria in pancreatobiliary brushings with corresponding positive fluorescence in situ hybridization results. Am J Clin Pathol. 2011; 136:442–449.

6. Kim YS, Kim HG, Han J, et al. The Significance of p53 and Kras Immunocytochemical Staining in the Diagnosis of Malignant Biliary Obstruction by Brush Cytology during ERCP. Gut Liver. 2010; 4:219–225.

7. Selvaggi SM. Biliary brushing cytology, Cytopathology. 2004; 15:74–79.

8. Hirooka Y, Nakaizumi A, Oka T, et al. Report of the clinical study for methods to improve the diagnostic accuracy of bile cytology. Jpn Soc Clin Cytol. 2010; 49:7–14. (in Japanese)

9. Shin SH, Lee K, Kim BH, et al. Bilebased detection of extrahepatic cholangiocarcinoma with quantitative DNA methylation markers and its high sensitivity. J Mol Diagn. 2012; 14:256–263.

10. Lester JL, Hormoz E, Armando CF, et al. Utilization of ancillary studies in the cytologic diagnosis of biliary and pancreatic lesions: the Papanicolaou Society of Cytopathology guidelines for pancreatobiliary cytology. Diagn Cytopathol. 2014; 42(4):351–362

11. Stewart CJ, Burke GM. Value of p53 immunostaining in pancreaticobiliary brush cytology specimens. Diagn Cytopathol. 2000; 23:308–313.

12. Wang Y, Yamaguchi Y, Watanabe H, et al. Usefulness of p53 gene mutations in the supernatant of bile for diagnosis of biliary tract carcinoma: comparison with K ras mutation. J Gastroenterol. 2002; 37(10): 831–9.

27MCM 2 and p53 immunocytochemistry in ERCP cytology

13. Wang Y, Yamaguchi Y, Watanabe H, et al. Detection of p53 gene mutations in the supernatant of pancreatic juice and plasma from patients with pancreatic carcinomas. Pancreas. 2004; 28(1): 13–9.

14. Giaginis C, Giagini A, Tsourouflis G, et al. MCM–2 and MCM–5 expression in gastric adenocarcinoma: clinical significance and comparison with Ki67 proliferative marker. Dig Dis Sci. 2011; 56:777–785.

15. Zhiying Y, Ishimi Y, Masai H. Roles of Mcm 7 and Mcm 4 Subunits in the DNA Helicase Activity of the Mouse Mcm 4 / 6 / 7 Complex. J Biol Chem. 2002; 277(45):42471–42479

16. Masata M, Juda P, Raska O, et al. A fraction of MCM 2 proteins remain associated with replication foci during a major part of S phase. Folia Biol (Praha). 2011; 57:3–11.

17. Gong Y, Joseph T, Sneige N. Validation of commonly used immunostains on celltransferred cytologic specimens. Cancer. 2005; 105:158–164.

18. Wu HH, Jones KJ, Cramer HM. Immunocytochemistry performed on the celltransferred direct smears of the

fineneedle aspirates: a comparison study with the corresponding formalinfixed paraffinembedded tissue. Am J Clin Pathol. 2013; 139:754–758.

19. Nicol AF, Lapa e Silva JR, et al. Evaluation of MCM–2 expression in TMA cervical specimens. PLoS One. 2012; 7: e32936.

20. Nasir A, Chen DT, Gruidl M, et al. Novel molecular markers of malignancy in histologically normal and benign breast. Patholog Res Int. 2011; 2011: 489064.

21. Scarpini C, White V, Muralidhar B, et al. Improved screening for anal neoplasia by immunocytochemical detection of minichromosome maintenance proteins. Cancer Epidemiol Biomarkers Prev. 2008; 17:2855–2864.

22. Kimura F, Kawamura J, Watanabe J, et al. Significance of cell proliferation markers (Minichromosome maintenance protein 7, topoisomerase IIalpha and Ki–67) in cavital fluid cytology: can we differentiate reactive mesothelial cells from malignant cells? Diagn Cytopathol. 2009; 38:161–167.