The present study suggests that (1) nearly 90% of AdCCs may have gene alterations of either MYB, MYBL1 or NFIB, suggesting the diagnostic utility of the FISH assay, (2) MYB or MYBL1 gene splits may be associated with local aggressiveness of the tumours and overexpression of MYC, which is one of the oncogenic MYB/MYBL1 targets and (3) MYC overexpression may be a risk factor for disease-free survival in AdCC.

There has been some debate as to whether a subset of metaplastic Warthin tumors (mWTs) harbor the mucoepidermoid carcinoma (MEC)-associated CRTC1-MAML2 fusion. We analyzed 15 tumors originally diagnosed as mWT (mWT-like tumors), 2 of which had concurrent MECs. We looked for the CRTC1/3-MAML2 fusion transcripts and performed immunohistochemistry for p63 and fluorescence in situ hybridization (FISH) for the MAML2 split. To localize MAML2 split-positive cells at the cellular level, whole tumor tissue sections were digitalized (whole-slide imaging [WSI]). The CRTC1-MAML2, but not CRTC3-MAML2 was detected in 5/15 mWT-like tumors. FISH-WSI results showed that all epithelial cells harbored the MAML2 split in fusion-positive mWT-like tumors and were totally negative in fusion-negative mWT-like tumors. A review of the hematoxylin and eosin-stained slides showed that morphology of the "metaplastic" epithelium was virtually indistinguishable between fusion-positive and fusion-negative tumors. However, oncocytic bilayered tumor epithelium, characteristic to typical WT, was always found somewhere in the fusion-negative tumors but not in the fusion-positive tumors. This distinguishing histologic finding enabled 5 pathologists to easily differentiate the 2 tumor groups with 100% accuracy. The age and sex distribution of fusion-positive mWT-like tumor cases was similar to that of fusion-positive MEC cases and significantly different from those of fusion-negative mWT-like tumor and typical WT cases. In addition, only fusion-positive mWT-like tumors possessed concurrent low-grade MECs. In conclusion, a subset of mWT-like tumors were positive for the CRTC1-MAML2 fusion and had many features that are more in accord with MEC than with WT. The term Warthin-like MEC should be considered for fusion-positive mWT-like tumors.

Mammary analogue secretory carcinoma (MASC) is a recently described low-grade carcinoma with morphologic and genetic similarity, including ETV6-NTRK3 fusion, to secretory carcinoma of the breast. ETV6 is frequently involved in other epithelial and nonepithelial tumors, and many fusion partners of ETV6 have been reported. In the present study, 14 Japanese MASC cases were clinicopathologically and molecularly analyzed. The median age of the patients was 39 years, and the male:female ratio was 6:8. All cases showed histopathologic findings compatible with those previously described for MASC and harbored an ETV6 split as visualized by fluorescence in situ hybridization. Two cases showed thick fibrous septa and invasive features including vascular or perineural tumor involvement, findings that are rare in MASC. In addition, in these 2 cases, non-NTRK3 genes appeared to fuse with ETV6 (ETV6-X fusion). NTRK1 and NTRK2, both members of the NTRK family, were not involved. Of the 14 MASC cases, the ETV6-NTRK3 fusion transcript was positive in 6 cases, and the relative expression level of the ETV6-NTRK3 fusion transcript was variable, ranging from 1 to 5.8. Results of the present study of MASC suggest that (1) ETV6 occasionally fuses with unknown non-NTRK3 genes, (2) ETV6-X cases might have an invasive histology, (3) for molecular diagnosis of MASC, fluorescence in situ hybridization to detect ETV6 splits is the method of choice, and (4) the expression level of the ETV6-NTRK3 fusion transcript is considerably variable. These findings provide a novel insight into the oncogenesis, histopathology, diagnosis, treatment, and prognosis of this newly recognized carcinoma.

The BIOMED-2 PCR protocol for targeting the IGH gene is widely employed for detecting clonality in B-cell malignancies. Unfortunately, the detection of clonality with this method is not very sensitive when paraffin sections are used as a DNA source. To increase the sensitivity, we devised a semi-nested modification of a JH consensus primer. The clonality detection rates of three assays were compared: the standard BIOMED-2, BIOMED-2 assay followed by BIOMED-2 re-amplification, and BIOMED-2 assay followed by semi-nested BIOMED-2. We tested more than 100 cases using paraffin-embedded tissues of various B-cell lymphomas, and found that the clonality detection rates with the above three assays were 63.9%, 79.6%, and 88.0%, respectively. While BIOMED-2 re-amplification was significantly more sensitive than the standard BIOMED-2, the semi-nested BIOMED-2 was significantly more sensitive than both the standard BIOMED-2 and BIOMED-2 re-amplification. An increase in sensitivity was observed in all lymphoma subtypes examined. In conclusion, tumor clonality may be detected in nearly 90% of B-cell lymphoma cases with semi-nested BIOMED-2. This ancillary assay may be useful when the standard BIOMED-2 fails to detect clonality in histopathologically suspected B-cell lymphomas.

Background The National Comprehensive Cancer Network (NCCN) guidelines recommend considering postoperative radiotherapy (PORT) for completely resected T1/2N0M0 salivary mucoepidermoid carcinomas when they show tumor spillage, perineural invasion, or intermediate/high‐grade histology. CRTC1/3‐MAML2 fusions have been associated with a favorable clinical outcome. Methods Forty‐seven T1/2N0M0 mucoepidermoid carcinoma cases positive for CRTC1/3‐MAML2 fusions were completely resected and were not treated with PORT. Results Pathologically, none of the cases showed tumor spillage or perineural invasion. Cases with intermediate/high‐grade histology numbered 9 (19%) to 26 (55%) with the currently used 3 different grading systems. During the follow‐up (median 60 months), locoregional tumor recurrence occurred in 4 cases, which were treated with surgery alone. At the last follow‐up (median 60 months; 7‐160), all patients were alive with no evidence of disease. Conclusion An excellent prognosis may be achieved without PORT in T1/2N0M0 mucoepidermoid carcinoma patients positive for CRTC1/3‐MAML2 fusions when the tumors are completely resected without tumor spillage.

Adenoid cystic carcinoma (AdCC), one of the most common salivary gland carcinomas, usually has a fatal outcome. Epidermal growth factor receptor (EGFR) pathway gene mutations are important in predicting a patient's prognosis and estimating the efficacy of molecular therapy targeting the EGFR pathway. In this study of salivary gland AdCC (SAdCC), we looked for gene mutations in EGFR, RAS family (KRAS, HRAS, and NRAS), PIK3CA, BRAF, and AKT1, using a highly sensitive single-base extension multiplex assay, SNaPshot. Out of 70 cases, EGFR pathway missense mutations were found in 13 (18.6%): RAS mutations in 10 (14.3%), EGFR in one (1.4%), and PIK3CA in 5 (7.1%). None of the cases showed an EGFR deletion by direct sequencing. Concurrent gene mutations were found in three cases (4.3%). EGFR pathway mutations were significantly associated with a shorter disease-free (p = 0.011) and overall survival (p = 0.049) and RAS mutations were as well; (p = 0.010) and (p = 0.024), respectively. The gene fusion status as determined by a FISH assay had no significant association with mutations of the genes involved in the EGFR pathway. In conclusion, EGFR pathway mutations, especially RAS mutations, may be frequent in SAdCC, and associated with a poor prognosis for the patient.

These findings suggest that CTAs may be expressed in a variety of salivary gland carcinomas, especially in those with higher histological grades. In addition, MAGE-A, which is frequently expressed in AdCC cases, may be a useful prognostic factor for poorer locoregional recurrence-free survival.

Mucosa-associated lymphoid tissue (MALT) lymphoma mainly consists of three types of tumor B cells, small (centrocyte-like), scattered large transformed, and intraepithelial. However, it is difficult to differentiate tumor B cells from reactive B cells at the cellular level. We examined five cases of API2-MALT1 fusion-positive MALT lymphoma of the lung. A single paraffin section for each case was subjected to sequential retrieval of whole-slide imaging (WSI) data of hematoxylin and eosin (HE) staining, immunofluorescence staining for CD79a, and fluorescence in situ hybridization (FISH) for the MALT1 split. We counted the number of MALT1 split-positive or MALT1 split-negative cells among CD79a-positive cells. The MALT1 split was detected in 59, 46, and 76 % of small, large, and intraepithelial B cells, respectively. A review of the HE-WSI data showed that cytomorphological distinction between the MALT1 split-positive and MALT1 split-negative B cells was virtually impossible. None of CD79a-negative lymphoid cells, epithelial cells, and microvascular endothelial cells was positive for MALT1 splits. As API2-MALT1 fusion is an early and critical event in the lymphomatogenesis, our findings are best interpreted as that a considerable number of B cells, either small, large, or intraepithelial, are reactive cells and that it is difficult to distinguish cytomorphologically between tumor B cells and reactive B cells. These findings suggest that the tumor architecture may be the central factor for making a correct histopathological diagnosis of MALT lymphoma. The sequential WSI of HE staining, immunofluorescence staining, and FISH as described here is a useful tool for pathological analysis at the cellular level.