Enhancing upper tract urothelial carcinoma diagnosis: Utility of cytokeratin 17 and CK20/CD44/p53 immunohistochemical panel.

Upper tract urothelial carcinoma (UTUC) presents diagnostic challenges due to small biopsy specimen size, poor orientation, and technical obstacles that can yield equivocal diagnoses. This uncertainty often mandates repeated biopsies to evaluate the necessity of nephroureterectomy. Prior studies have suggested cytokeratin 17 (CK17) immunostain as an adjunctive tool for diagnosing bladder urothelial neoplasia in both urine cytology and tissue biopsy specimens. We evaluated the utility of CK17 in differentiating UTUC from benign urothelium and its ability to stratify low-grade from high-grade neoplasia. Our study involved a cohort of previously diagnosed cytology (n = 29) and tissue specimens from biopsies and resections (n = 85). We evaluated CK17 staining percentage in cytology and tissue samples and localization patterns in biopsy/resection samples. Our findings showed a statistically significant distinction (p < 0.05) between UTUC and benign tissue specimens based on full thickness localization pattern (odds ratio 8.8 [95% CI 1.53-67.4]). The percentage of CK17 staining failed to significantly differentiate neoplastic from non-neoplastic cases in cytology or tissue samples. Additionally, based on prior research showing the efficacy of CK20/CD44/p53 triple panel in bladder urothelial neoplasia, we utilized tissue microarrays to evaluate if these markers could distinguish UTUC from benign urothelium. We found that CK20/CD44/p53, individually or in combination, could not distinguish urothelial neoplasia from non-neoplasia. Full thickness CK17 urothelial localization by immunohistochemistry was highly reproducible with excellent interobserver agreement and may play a supplementary role in distinguishing upper tract urothelial neoplasia from benign urothelium.

A systematic review and meta-analysis of CK20, CD44, Ki67 and p53 as immunohistochemical markers in bladder carcinoma in situ.

BACKGROUND: Urothelial dysplasia and carcinoma in situ (CIS) are related to recurrence and progression of urothelial carcinoma. Differentiating CIS and dysplasia from reactive atypia is often difficult based only on histological features. The integration of histological findings with immunohistochemistry is used in routine practice to make a diagnosis of CIS and, for this purpose, the immunohistochemical markers CK20, CD44, Ki67 and p53 are used to supplement histology. In this work, we aimed to assess CK20, CD44, Ki67 and p53 as immunohistochemical markers in patients with CIS through a systematic review and meta-analysis. MATERIALS AND METHODS: A systematic review was performed by searching electronic databases for English-language studies published from January 2010 to April 2021. Studies were considered eligible if they evaluated the CK20, CD44, Ki67 and p53 expression in CIS. RESULTS: In total, 15 references were suitable for quantitative review. The overall rate of CK20, CD44, Ki67 and p53 expression in CIS was 43%, 31%, 44%, 38%, respectively. CONCLUSIONS: Our study supports the 2014 International Society of Urologic Pathology consensus that histological assessment remains the gold standard to diagnose urothelial CIS and suggests that a very close correlation between morphological, immunohistochemical and clinical data is essential to provide the best management for patients with bladder carcinoma.

Could double stain for p53/CK20 be a useful diagnostic tool for the appropriate classification of flat urothelial lesions?

BACKGROUND: The differential diagnosis between flat urothelial lesions [reactive urothelial atypia (RUA), atypia of unknown significance (AUS), urothelial dysplasia (UD) and carcinoma in situ (CIS)] has relevant prognostic and therapeutic implications. This crucial distinction could be very challenging but it is currently performed on hematoxylin and eosin (H&E) slides, with a great amount of partially discordant and/or not conclusive findings of the potential adjunctive role of immunohistochemistry. Herein, we tested double staining (DS) for p53/CK20 to verify if p53(+) cells, CK20(+) cells and double-positive cells (DPCs) are differentially expressed among these lesions and if p53/CK20 could be a useful tool in this diagnostic setting. METHODS: We tested 50, 9, 36 and 29 consecutive and retrospectively enrolled cases of RUA, AUS, UD and CIS, respectively. p53(+) cells, CK20(+) cells and DPCs were evaluated and compared by adopting the appropriate statistic tests (Mann-Whitney U and Kruskal-Wallis tests). RESULTS: We found that p53(+) cells (p = 0.000), CK20(+) cells (p = 0.000) and DPCs (p = 0.000) showed statistically significant differences among the different flat urothelial lesions. Besides, when dichotomized, both CIS and RUA are easily differentiable from their histological mimickers adopting all these markers; by contrast, AUS and UD did not reach statistically significant differences able to differentiate them from each other [p53(+) cells, p = 0.123; CK20(+) cells, p = 0.567; DPCs, p = 0.409], except if compared to CIS [AUS VS CIS: p53(+) cells, p = 0.013; CK20(+) cells, p = 0.000; DPCs, p = 0.000; UD vs CIS: p53(+) cells, p = 0.000; CK20(+) cells, p = 0.000; DPCs, p = 0.000]. CONCLUSIONS: p53(+) cells, CK20(+) cells and DPCs are differently expressed by flat urothelial lesions and p53/CK20 could be a time- and money-saving tool for the appropriate management of these lesions if applied to a routine scenario.

Association of current molecular subtypes in urothelial carcinoma with patterns of muscularis propria invasion.

Urothelial carcinoma is subdivided into luminal (L), basal (B), and p53-wild-type (WT) molecular subtypes, with basal and p53-WT groups showing more aggressive course and poor treatment response, respectively. The literature on molecular subtypes of UC includes a mixture of different stages. We investigated the molecular profile and outcome of pure cohort of muscle invasive bladder carcinoma (MIBC) considering two distinct patterns of muscularis propria (MP) invasion. Forty-three cystectomies harboring stage pT2 were retrospectively identified in 18 years. MP invasion was subclassified into patterns 1 (tumor encasing intact detrusor muscle bundles) and 2 (tumor dissecting/replacing detrusor muscle). Using IHC, B/L phenotypes, p53, and Ki67 were assessed, and survival data was collected. Pattern 1 invasion was noted in 16 (37%) and pattern 2 in 27 (63%), with mean age of pattern 1 being 10 years younger. B/L phenotypes were successfully determined in 83.7%; 48.8% and 34.8% revealed L and B phenotypes, respectively (indeterminate phenotype in 16.4%). Pattern 1 was associated with L phenotype (GATA3 and HER-2 expressions: p = 0.02 & p = 0.04, respectively). Ki67 ≥ 5/10HPF was noted in pattern 2 and B phenotype (p = 0.03). B phenotype showed association with p53-WT (p = 0.007). In median follow-up of 60.7 months, 63.6% of pattern 1 cases were alive without disease compared to 32% of pattern 2 (not significant). A panel of CK20 and GATA3 for luminal and CK5/6 and CK14 for basal subtypes can provide reliable molecular classification in UC. Also, morphology of MIBC can predict the molecular phenotype and the behavior of the UC.

Intestinal-type mucinous adenocarcinoma of the Bartholin gland in a perimenopausal woman. A case report and review of the literature.

We report the case of a 46-year-old woman who presented with a tumor on the left labium majus in the region of the Bartholin gland. Surgical excision revealed a mucinous adenocarcinoma of intestinal-type (CK20+, CDX-2+). Magnetic resonance imaging, computed tomography of the chest and abdomen and colonoscopy ruled out the presence of other tumors. A second immunohistochemical study showed negative results for GATA-3, mammaglobin and GCDFP-15. Molecular analysis revealed a mutation in exon 2 of the KRAS gene. We discuss its differential diagnosis and the importance of being aware of this unusual variant of a mucinous adenocarcinoma the Bartholin gland.

Salivary Mucinous Adenocarcinoma Is a Histologically Diverse Single Entity With Recurrent AKT1 E17K Mutations: Clinicopathologic and Molecular Characterization With Proposal for a Unified Classification.

Mucin-producing salivary adenocarcinomas were historically divided into separate colloid carcinoma, papillary cystadenocarcinoma, and signet ring cell carcinoma diagnoses based on histologic pattern, but have recently been grouped together in the adenocarcinoma not otherwise specified category. It is currently unclear if these tumors represent 1 or more distinct entities and how they are related to well-circumscribed papillary mucinous lesions with recurrent AKT1 E17K mutations that were recently described as salivary intraductal papillary mucinous neoplasm. Here, we sought to evaluate the clinicopathologic and molecular features of salivary mucinous adenocarcinomas to clarify their classification. We identified 17 invasive mucin-producing salivary adenocarcinomas, 10 with a single histologic pattern, and 7 with mixed patterns. While most tumors demonstrated papillary growth (n=15), it was frequently intermixed with colloid (n=6) and signet ring (n=3) architecture with obvious transitions between patterns. All were cytokeratin 7 positive (100%) and cytokeratin 20 negative (0%). Next-generation sequencing performed on a subset demonstrated recurrent AKT1 E17K mutations in 8 cases (100%) and TP53 alterations in 7 cases (88%). Of 12 cases with clinical follow-up (median: 17 mo), 4 developed cervical lymph node metastases, all of which had colloid or signet ring components. Overall, overlapping histologic and immunohistochemical features coupled with recurrent AKT1 E17K mutations across patterns suggests that mucin-producing salivary adenocarcinomas represent a histologically diverse single entity that is closely related to tumors described as salivary intraductal papillary mucinous neoplasm. We propose a unified mucinous adenocarcinoma category subdivided into papillary, colloid, signet ring, and mixed subtypes to facilitate better recognition and classification of these tumors.

Comprehensive Gene Expression Analyses of Immunohistochemically Defined Subgroups of Muscle-Invasive Urinary Bladder Urothelial Carcinoma.

A number of urinary bladder urothelial carcinoma (UB UC) mRNA-based classification systems have been reported. It also has been observed that treatment response and prognosis are different for each molecular subtype. In this study, cytokeratin (CK)5/6 and CK20 immunohistochemistry (IHC) were performed, and IHC-based subgroup classification was applied. UB UC was classified into CK5/6 single-positive (SP), CK20 SP, double-positive (DP) and double-negative (DN) subgroups, and transcriptional analysis was performed. The results of gene ontology (GO) terms and functional analysis using differentially expressed genes indicate that, CK5/6 SP and DP subgroups were enriched in cell migration, immune activation, interleukin 6-Janus kinase-signal transducer and activator of transcription 3 (IL6-JAK-STAT3) signaling pathway and tumor necrosis factor-α signaling via the nuclear factor-κB (NF-κB) signaling pathway signature gene. In addition, compared with the other subgroups, the DN subgroup showed inhibited cell movement, cell migration, and cell activation. Furthermore, in survival analysis, the CK5/6 SP subgroup was significantly associated with poor progression-free survival (p = 0.008). The results of our study indicate that the CK5/6 positive subgroup exhibited high gene expression signature related to aggressive behavior and exhibited worse clinical outcome.

Expression of CD117, CK17, CK20, MUC4, villin and mismatch repair deficiency in pancreatic intraductal papillary mucinous neoplasm.

Among pancreatic intraductal papillary neoplasms, gastric, intestinal, and pancreatobiliary intraductal papillary mucinous neoplasm (IPMN), intraductal oncocytic papillary neoplasm (IOPN), and intraductal tubulopapillary neoplasm (ITPN) have been defined, differing regarding association with invasive carcinoma and prognosis. Immunohistochemistry (IHC) can help in the distinction of these neoplasms, but a proportion is unclassifiable using recommended markers. Hence, additional markers useful for the typing of pancreatic intraductal papillary neoplasms are needed. The reported frequencies of the different types of IPMNs in surgical series vary to some extent, and such data based on Danish patients are currently lacking. Besides, the role of mismatch repair (MMR) deficiency in these neoplasms has not been fully elucidated. We aimed to evaluate the frequency of different types of pancreatic intraductal papillary neoplasms in a Danish cohort. Furthermore, we aimed to examine the utility of CD117, CK17, CK20, MUC4, and villin as markers for their distinction, in addition to the recommended markers MUC1, MUC2, MUC5AC, MUC6 and CDX2, and to evaluate the frequency of MMR deficiency. We typed 40 consecutively resected pancreatic intraductal papillary neoplasms according to the WHO criteria from 2019. IHC for CD117, CDX2, CK17, CK20, MLH1, MSH2, MSH6, MUC1 (H23), MUC1 (Ma695), MUC2, MUC4, MUC5AC, MUC6, PMS2, and villin was performed and evaluated using a five-tiered semiquantitative scale. A subset of the tumours was examined with PCR for microsatellite instability (MSI). Most tumours were intestinal (40 %) and gastric (40 %) IPMNs, followed by pancreatobiliary (17 %) IPMNs and IOPN (3 %). All cases were MMR proficient. We found a higher expression of MUC4, CK20 and villin in intestinal compared to gastric IPMNs (p < 0.01, p < 0.001 and p < 0.001). MUC4 was more strongly expressed in intestinal compared to pancreatobiliary IPMNs, while the opposite was found for CK17 (p < 0.05 and p < 0.05). IOPN showed strong CD117 expression (score 4), while all gastric IPMNs were negative and 50 % and 29 % of intestinal and pancreatobiliary IPMNs only showed weak expression (score 1). Our data suggest that CK20, MUC4 and villin may aid in the identification of intestinal IPMNs, while CK17 and CD117 may aid in the identification of pancreatobiliary IPMNs and IOPN, in some cases. However, additional studies evaluating these markers in pancreatic intraductal papillary neoplasms are needed.

Cutaneous Adenosquamous Carcinoma of the Scalp With Intestinal Phenotype.

Cutaneous adenosquamous carcinoma is a mixed, squamous and glandular, rare malignant tumor of the skin characterized by a mixed, squamous, and glandular differentiation. Few cases of this tumor have been so far reported, and even fewer have been thoroughly studied by immunohistochemistry. We report here an exceptional case of cutaneous adenosquamous carcinoma which showed immunohistochemically features of intestinal differentiation, namely because of the expression of keratin 20 and CDX2, a marker of gastrointestinal tumors.

CK20 versus AMACR and p53 immunostains in evaluation of Urothelial Carcinoma in Situ and Reactive Atypia.

Ancillary testing with immunohistochemistry has shown recent promise in the workup of equivocal bladder lesions. We read with interest the recent findings of Alston et al., who assessed the diagnostic utility of alpha-methylacyl-CoA racemase (AMACR) in comparison to cytokeratin 20 (CK20) in evaluation of atypia in challenging flat urothelial lesions in the differential between carcinoma in situ (CIS) and reactive atypia. AMACR was reported to be a somewhat more specific but less sensitive marker for CIS than CK20, though showing weaker intensity. Spurred by their report, with the knowledge that we had consistently and consecutively performed AMACR, CK20, and p53 on flat urothelial lesions challenging enough to reach intradepartmental consensus, we performed a retrospective review. Similarly, we found that AMACR was less sensitive (80%) and more specific (100%) than CK20, with the same caveat of less staining intensity. Additionally, our p53 review identified a significant rate (~ 27%) of equivocal/non-informative findings. Taken together, our experience in this consecutive cohort confirms the impression of Alston et al. regarding the utility and challenges of AMACR use, while highlighting challenges with p53, which we plan to use more sparingly prospectively.

Immunohistochemical staining patterns of Ki-67 and p53 in florid reactive urothelial atypia and urothelial carcinoma in situ demonstrate significant overlap.

Flat urothelial lesions with atypia may pose significant diagnostic challenges. Given frequent increased proliferation rates in florid reactive urothelial atypia and limited studies on the interpretation of p53 stains in the urothelium (following current standard guidelines for correlation with P53 mutation status), we sought to further study the discriminatory value of Ki-67 and p53 for florid reactive urothelial atypia versus urothelial carcinoma in situ (CIS). Bladder specimens diagnosed as reactive urothelial atypia (n = 40) and CIS (n = 40) were assessed by immunohistochemical staining with antibodies for Ki-67, p53, CD44, and CK20. Immunoreactivity was scored based on percent cells positive for Ki-67 and pattern of reactivity with p53 (aberrant: diffuse strong positive or negative; normal: patchy/wild type). CD44 and CK20 reactivity patterns served as adjunctive internal validation controls for reactive urothelial atypia and CIS, as previously described. In reactive urothelial atypia, Ki-67 ranged from 0% to 90% (mean, 34% ± 26) with 30 cases (75%) having >10%. In CIS, Ki-67 ranged from 5% to 95% (mean, 50% ± 25) with 17 cases (43%) having >50%. In all 40 cases (100%) of reactive urothelial atypia, p53 expression had a wild-type pattern. In CIS, aberrant p53 expression was identified in 15 cases (37%): 3 cases (7%) were p53 negative (i.e. null phenotype) and 12 cases (30%) showed strong and diffuse nuclear reactivity (in >85% of cells). The remaining 25 cases (63%) of CIS had a p53 wild-type pattern of expression. Cytoplasmic CK20 immunoreactivity in umbrella cells was seen in 34 cases (85%) of reactive urothelial atypia, and 6 cases (15%) were negative. In addition, 35 cases (88%) of reactive urothelial atypia demonstrated full-thickness CD44 expression, while 5 cases (12%) had expression confined to the basal/parabasal layers of the urothelium. Strong and diffuse CK20 positivity was present in 39 cases (98%) of CIS, and patchy positivity was detected in 1 case (2%). None of the CIS cases overexpressed CD44: 16 cases (40%) showed focal expression in the nonneoplastic basal cell layer; 24 cases (60%) demonstrated no staining. In summary, Ki-67 has poor discriminatory value for reactive urothelial atypia versus CIS and adds little to the classic CK20/CD44 immunophenotype. While p53 sensitivity for CIS is relatively low (30%) and interpretation as either wild type or negative may be challenging in a small subset of cases, strong and diffuse nuclear reactivity was 100% specific in the distinction from florid reactive urothelial atypia in this cohort.

Does the addition of AMACR to CK20 help to diagnose challenging cases of urothelial carcinoma in situ?

BACKGROUND: Urothelial carcinoma in situ (CIS) in the bladder can be difficult to diagnose due to factors including procedural artifact, minimal tissue sampled, therapy-related changes, and various CIS growth patterns. Prior data has demonstrated an increase in alpha-methylacyl-CoA-racemase (AMACR) in urothelial CIS, but there is no information on its utility for diagnosing difficult cases. The aim of this investigation was to assess the expression of AMACR that was ordered on equivocal bladder cases during clinical practice. METHODS: Transurethral resections of the bladder in which AMACR and CK20 were performed during diagnostic workup were identified and cases with a final diagnosis of CIS (n = 22) or non-neoplastic urothelium (n = 30) were selected. Additionally, cases in which a diagnosis of CIS was rendered without IHC (n = 20) were selected and tested for AMACR expression. RESULTS: Sensitivity of AMACR for CIS diagnosed with IHC during clinical practice was 73% and specificity was 97%, while CK20 was 95% sensitive and 80% specific. Sensitivity of AMACR in CIS diagnosed without IHC was 100%. In all groups, AMACR had inconsistent intensity, compared to CK20 which had consistent, strong intensity. CONCLUSIONS: AMACR was usually positive in urothelial CIS and negative in non-neoplastic urothelium. However, it is important to note that AMACR was less sensitive in difficult cases, while CK20 was more sensitive with more consistent, strong staining compared to AMACR.

Predicting outcomes in non-muscle invasive (Ta/T1) bladder cancer: the role of molecular grade based on luminal/basal phenotype.

Bladder cancer tumors can be divided into two molecular subtypes referred to as luminal or basal. Each subtype may react differently to current chemotherapy or immunotherapy. Likewise, the technology required for comprehensive molecular analysis is expensive and not yet applicable for routine clinical diagnostics. Therefore, it has been suggested that the immunohistochemical expressions of only two markers, luminal (CK20+, CK5/6-) and basal (CK5/6+, CK20-), is sufficient to identify the molecular subtypes of bladder cancer. This would represent a molecular grade that could be used in daily practice. Molecular classification is done using immunohistochemistry to assess luminal-basal phenotype based on tissular expression of CK20 and CK5/6 as surrogate for luminal or basal subtypes, respectively. A series of 147 non-muscle-invasive bladder carcinoma cases was selected, and the tumors were divided into four subgroups based on the presence of CK20 and/or CK5/6, that is, null (CK20-, CK5/6-), mixed (CK20+, CK5/6+), basal (CK20-, CK5/6+), and luminal (CK20+, CK5/6-) categories. Survival analysis was estimated using the Kaplan-Meier method and the log-rank test. Hazard ratios were calculated by Cox multivariate analysis. The molecular grade included cases with null (n = 89), mixed (n = 6), basal (n = 20), and luminal (n = 32) phenotypes with differences in recurrence-free, progression-free and cancer-specific survival associated with molecular-grade categories in patients with low- or high-grade Ta, or high-grade T1 tumors. The multivariate analysis identified the luminal phenotype as a predictor of more aggressive neoplasms. Our findings provide a rationale to investigate luminal and basal subtypes of bladder cancer using two gene expression signatures as surrogate markers and show that non-muscle-invasive bladder carcinoma can be stratified into biologically and clinically different subgroups by using an immunohistochemical classifier.

Prognostic Significance of "Nonsolid" Microscopic Metastasis in Merkel Cell Carcinoma Sentinel Lymph Nodes.

Our recent work regarding Merkel cell carcinoma sentinel lymph node (SLN) metastasis found that "solid" pattern microscopic metastasis conferred worse prognosis than the "nonsolid" ones. The goals of the present study were to (1) compare the prognostic significance/outcomes of 2 diagnostic groups-patients with a nonsolid pattern of SLN metastasis and those with diagnostically negative SLN biopsies (SLNB), and (2) evaluate the durability of SLN metastasis after extensive sectioning. Five-level, step-wise sectioning at 250-µm intervals was performed in all SLN blocks with an immunohistochemical stain for CK20 on all levels. The presence and pattern of metastases were recorded and analyzed as were corresponding patient and tumor parameters. Median follow-up durations for all patients (n=38), positive SLNB (n=16) and negative SLNB (n=22) groups were 56.3, 50.4, and 66.8 months, respectively. Overall survival (OS) and disease-specific survival (DSS) did not differ between the 2 diagnostic groups (OS P=0.65, DSS P=0.37) but did differ by immune status (immunocompetent vs. immunosuppressed, OS P=0.03, DSS P=0.005) and primary tumor category (OS P<0.0001, DSS P=0.001). On deeper sectioning, all 16 diagnostically positive SLNB continued to show nonsolid microscopic metastasis, and 32% (7/22) diagnostically negative SLNB revealed nonsolid metastasis. DSS was worse for sinusoidal-pattern metastasis versus all others (P=0.02). Five of 38 patients (13%) died of disease; the only immunocompetent patient had sinusoidal-pattern metastasis discovered in a diagnostically negative SLNB. Our data suggest that outcome for nonsolid metastasis is similar to that of negative SLNB with the exception of the sinusoidal pattern, which was associated with worse outcome. Larger studies are warranted to quantify and compare microscopic metastatic tumor burden by pattern and confirm whether the sinusoidal pattern confers an intermediate prognostic risk between solid and other nonsolid microscopic metastases.

Diagnostic and prognostic roles of CK20 in the pathology of urothelial lesions. A systematic review.

Cytokeratin 20 (CK20) is one of the most common immunohistochemical markers in the routine practice of a pathology lab, as biopsies from the urinary tract encompass a wide spectrum of lesions which may pose issues in their detection and classification. In this review, we aim to outline the diagnostic accuracy and prognostic value of CK20 in flat urothelial lesions, papillary non-invasive and invasive urothelial carcinoma, molecular subgroups and variant histology, and we briefly discuss its limitations and potential pitfalls.

[Combined application of immunohistochemical markers to identify pathologic subtypes of ampullary carcinoma and its clinical significance].

Objective: To investigate the expression of immunomarkers CK7, CK20, CK17, CDX2, MUC1 and MUC2 in primary adenocarcinoma of the ampulla of Vater, to explore the role of these markers in the histopathologic subclassification of ampullary carcinoma; and to provide biologic basis for precision treatment of patients with different types of ampullary carcinoma. Methods: Forty-two cases of primary ampullary carcinoma were collected at Peking University People's Hospital, from 2012 to 2018 year. There were 22 males and 20 females. Aged range 42 to 88 years old, with mean aged (62±11) years. Among the patients, 6 was high differentiation, 19 median differentiation, and 17 low differentiation. Immunohistochemical studies on the expression of CK7, CK20, CK17, CDX2, MUC1 and MUC2 were performed in 42 cases of primary ampullary carcinoma. The relationship between different ampullary carcinoma subtypes and clinicopathologic survival data was analyzed using SPSS 16.0 statistical software. Results: Three histopathologic subtypes were observed. Among 42 cases, 8(19.0%)were classified as intestinal subtype, which showed a positive expression rate of 8/8 for both CK20 and CDX2, and 5/8 for MUC2. Both CK7 and CK17 were weakly expressed in one case (1/8). No expression was observed for MUC1 in this subtype. Twenty-two (52.4%,22/42) cases were classified as pancreaticobiliary subtype, which showed a positive expression rate of 100.0%(22/22) for both CK7 and MUC1, and 90.9% (20/22) for CK17. No expression was observed for CK20, CDX2 and MUC2.The remaining 12 (28.6%) cases were classified as mixed subtype, which showed variable expression patterns. The expression frequencies of these 6 immunomarkers in different subtypes of ampullary carcinoma did not correlate with various clinicopathologic factors such as patient gender and age, tumor size, histologic differentiation, pancreatic and bile duct invasion, or the depth of duodenal invasion. However, stage Ⅲ+Ⅳ diseases were more commonly seen in pancreaticobiliary type (63.6%,14/22) than intestinal type (2/8) and mixed type (3/9; χ(2)=6.508, P=0.039). Follow-up data showed a trend of better survival rate for patients with intestinal subtype than those with mixed and pancreaticobiliary subtypes. Conclusions: Ampullary carcinoma can be subclassified into three different subtypes using a panel of six immunomarkers, especially for the identification of subtypes of poorly differentiated carcinoma. CK7, CK17 and MUC1 are major markers of pancreaticobiliary subtype, whereas CK20, CDX2 and MUC2 are useful markers for intestinal subtype. The mixed subtype variably expresses these markers. The prognosis of patients with intestinal subtype appears better than that of pancreaticobiliary and mixed subtypes. Accurate subtyping of ampullary carcinoma is clinically important to patient management and prognosis assessment.

Neurofilament is superior to cytokeratin 20 in supporting cutaneous origin for neuroendocrine carcinoma.

AIM: Primary cutaneous neuroendocrine carcinoma, or Merkel cell carcinoma (MCC), cannot be distinguished morphologically from small-cell neuroendocrine carcinomas (SmCC) from other sites. Immunohistochemistry is required to confirm cutaneous origin, and is also used for detection of sentinel lymph node (SLN) metastases of MCC. Cytokeratin 20 (CK20) expression is commonly used for these purposes, but is negative in some MCC cases, and has unclear specificity. We evaluated immunohistochemistry for neurofilament and CK20 in MCC compared with SmCC from other sites. METHODS AND RESULTS: We evaluated neurofilament expression in 55 MCC specimens from 39 unique patients, including nine CK20-negative MCC tumours. Neurofilament expression was observed in 42 of 55 (76.4%) MCC cases, including seven of nine (77.8%) CK20-negative MCC cases. Neurofilament was expressed in nine of 12 (75%) Merkel cell polyomavirus-positive tumours and five of 10 (50%) virus-negative tumours. Compared to a standard immunohistochemical panel (cytokeratin cocktail and CK20), neurofilament was 87.5% sensitive for detecting SLN metastases. Neurofilament and CK20 expression was also assessed in 61 extracutaneous SmCC from 60 unique patients, with primary sites including lung (27), bladder (18), cervix (3), gastrointestinal tract (3), sinonasal tract (2) and other sites (7). The specificity of neurofilament and CK20 for MCC versus non-cutaneous SmCC was 96.7% and 59.0%, respectively. CONCLUSIONS: Neurofilament has superior specificity to CK20 in distinguishing MCC from non-cutaneous SmCC. Neurofilament is frequently expressed in CK20- and virus-negative MCC tumours. Limitations of neurofilament immunohistochemistry include lower sensitivity than CK20 and subtle staining in some tumours. However, our findings indicate that neurofilament is useful for excluding non-cutaneous SmCC.

Immunohistochemistry of cytokeratin (CK) 5/6, CD44 and CK20 as prognostic biomarkers of non-muscle-invasive papillary upper tract urothelial carcinoma.

AIMS: Immunohistochemical (IHC) staining for cytokeratin (CK) 5/6, CD44 and CK20 has been significantly associated with the prognosis of urinary bladder urothelial carcinoma, and probably reflects its molecular characteristics. We aimed to investigate the IHC-based subgroups and their prognostic effects on non-muscle-invasive papillary upper tract urothelial carcinoma (UTUC). METHODS AND RESULTS: IHC staining for CK5/6, CK20 and CD44 was analysed in 211 patients with non-muscle-invasive papillary UTUC. Staining was classified as showing a negative, positive or normal pattern. We found that CK5/6-negative, CD44-negative and CK20-positive tumours were distinctly high-risk subgroups that were associated with high grade (CK5/6-negative, P < 0.001; CD44-negative, P < 0.001; CK20-positive, P = 0.017) and frequent intravesical recurrence (CK5/6-negative, P = 0.002). Using survival analysis with Kaplan-Meier and log-rank tests, we found that these IHC subgroups were correlated with poor progression-free (CK5/6-negative, P = 0.001; CD44-negative, P = 0.009; CK20-positive, P = 0.031) and cancer-specific (CK5/6-negative, P = 0.009) survival. Furthermore, CK5/6 negativity was an independent prognostic factor for shorter progression-free (P = 0.009) and cancer-specific (P = 0.045) survival. CK5/6 improved Harrell's C-indices for progression-free (0.68-0.77, P = 0.029) and cancer-specific (0.59-0.77, P < 0.001) survival. When markers were combined, luminal-like subtypes showed poor prognoses. CONCLUSIONS: We demonstrated that IHC staining for CK5/6, CD44 and CK20 was significantly associated with the clinicopathological characteristics and prognoses of patients with non-muscle-invasive papillary UTUC. The IHC subgroups may be correlated with the molecular characteristics of non-muscle-invasive papillary UTUC.