Histologically undetermined early acral melanoma in situ (HUAMIS) is rare but a diagnostic challenge, being clinically and dermoscopically MIS (late onset, a large size (>7 mm), parallel ridges pattern) but microscopically without recognizable cytological atypia. Cyclin D1 (CCND1) gene amplification is a genetic aberration occurring in the early radial growth phase of AMs and could thus help determine malignancy for this disease. We determine the value of CCND1 amplification by FISH as a diagnostic marker for HUAMIS. CCND1 amplification was examined in paraffin-embedded skin biopsies and excisions using a dual-probes fluorescence in situ hybridization (FISH) (11q13 and CEP11). One FISH-negative case 6 was additionally examined by Mypath Melanoma (qRT-PCR). Seventeen cases (12 dysplastic nevi, 3 AMIS, and 2 invasive AM) were served as negative controls for FISH. All six patients (4 females and 2 males) were Hispanic. Pigment lesions were on the left plantar foot (4), right third finger palm (1), and right thumb subungual (1). All cases showed similar clinical and dermoscopical characteristics, including late onset (50 to 74 years old), long duration (from 2 to 15 years), large-sized pigments (from 16 to 40 mm), and a parallel ridge pattern. Junctional melanocytes with no or minimal atypia from five cases showed CCND1 amplifications. Four of 5 cases were received 1st or/and 2nd wide excisions, which demonstrated foci of histologically overt MIS. One FISH-negative case 6 demonstrated "likely malignancy" scores (>2) by Mypath Melanoma (qRT-PCR). None of negative controls showed the amplification. We propose here a simple CCND1 FISH is a practical diagnostic test to determine the malignancy of the very early progression phase of AM preceding histopathologically defined MIS. Cases presented here could be an indolent subtype of AMIS characterized by carrying a long latent radial growth phase without vertical growth, mimicking lentigo maligna.

When South Florida became a hot spot for COVID-19 disease in March 2020, we faced an urgent need to develop test capability to detect SARS-CoV-2 infection. We assembled a transdisciplinary team of knowledgeable and dedicated physicians, scientists, technologists, and administrators who rapidly built a multiplatform, polymerase chain reaction- and serology-based detection program, established drive-through facilities, and drafted and implemented guidelines that enabled efficient testing of our patients and employees. This process was extremely complex, due to the limited availability of needed reagents, but outreach to our research scientists and multiple diagnostic laboratory companies, and government officials enabled us to implement both Food and Drug Administration authorized and laboratory-developed testing-based testing protocols. We analyzed our workforce needs and created teams of appropriately skilled and certified workers to safely process patient samples and conduct SARS-CoV-2 testing and contact tracing. We initiated smart test ordering, interfaced all testing platforms with our electronic medical record, and went from zero testing capacity to testing hundreds of health care workers and patients daily, within 3 weeks. We believe our experience can inform the efforts of others when faced with a crisis situation.

Follicular lymphoma (FL) is one of the most frequently diagnosed lymphomas in the United States and Europe. The definition of and basic approach to diagnosis and grading of FL is essentially unchanged in the recently updated revision of the World Health Organization (WHO) classification. FL is a biologically and histopathologically heterogeneous disease. Although there is an improved understanding of some FL variants and specific subtypes, there are cases whose recognition is particularly challenging, either because they have unusual features or represent examples of new or rare variants. Herein, we share a series of unusual and difficult to recognize FLs with the goal of increasing awareness of the expanding histopathologic variability in FL. Unusual FL discussed here include: FL with Castleman-like changes, FL with plasmacytic differentiation, and immunoglobulin G4-positive plasma cells in the setting of immunoglobulin G4-related disease, FL with marginal zone differentiation and involving mucosa-associated lymphoid tissue sites, diffuse FL variant expressing CD23 with STAT6 mutation, large B-cell lymphoma with IRF4 rearrangement, CD10-negative and MUM1-positive aggressive FL, and Epstein-Barr virus-positive FL.

Background: Partial monosomy 21 is a rare finding with variable sizes and deletion breakpoints, presenting with a broad spectrum of phenotypes. Case presentation: We report a 10-month-old boy with short stature, minor anomalies and mild motor delay. The patient had a monosomy 21 and duplication of the 21q22.11q22.3 region on the remaining derivative chromosome 21 which represents a partial 21q uniparental disomy of paternal origin, upd(21q22.11q22.3) pat. The abnormalities were characterized by karyotyping, FISH, chromosomal microarray, and genotyping. Conclusions: This is the first case showing a monosomy 21 compensated by upd(21q22.11q22.3) as a mechanism of genomic rescue. Because there is no strong evidence showing imprinting on chromosome 21, the uniparental disomy itself is not associated with abnormal phenotype but has reduced phenotype severity of monosomy 21. We reviewed the previously published cases with isolated 21q deletions and identified a common deletion of 5.7 Mb associated with low birth weight, length and head circumference in the 21q21.2 region.

Conflict resolution in genomic variant interpretation is a critical step toward improving patient care. Evaluating interpretation discrepancies in copy number variants (CNVs) typically involves assessing overlapping genomic content with focus on genes/regions that may be subject to dosage sensitivity (haploinsufficiency (HI) and/or triplosensitivity (TS)). CNVs containing dosage sensitive genes/regions are generally interpreted as "likely pathogenic" (LP) or "pathogenic" (P), and CNVs involving the same known dosage sensitive gene(s) should receive the same clinical interpretation. We compared the Clinical Genome Resource (ClinGen) Dosage Map, a publicly available resource documenting known HI and TS genes/regions, against germline, clinical CNV interpretations within the ClinVar database. We identified 251 CNVs overlapping known dosage sensitive genes/regions but not classified as LP or P; these were sent back to their original submitting laboratories for re-evaluation. Of 246 CNVs re-evaluated, an updated clinical classification was warranted in 157 cases (63.8%); no change was made to the current classification in 79 cases (32.1%); and 10 cases (4.1%) resulted in other types of updates to ClinVar records. This effort will add curated interpretation data into the public domain and allow laboratories to focus attention on more complex discrepancies.

Monocytosis can develop during disease course in primary myelofibrosis simulating that seen in chronic myelomonocytic leukemia, and should not lead to disease reclassification. In contrast, at presentation, rare cases have clinical, morphologic, and molecular genetic features truly intermediate between primary myelofibrosis and chronic myelomonocytic leukemia. The taxonomy and natural history of these diseases are unclear. We identified cases which either: (1) fulfilled the 2008 World Health Organization criteria for primary myelofibrosis but had absolute monocytosis and, when available, chronic myelomonocytic leukemia-related mutations (ASXL1, SRSF2, TET2) or (2) fulfilled criteria of chronic myelomonocytic leukemia but had megakaryocytic proliferation and atypia, marrow fibrosis, and myeloproliferative-type driver mutations (JAK2, MPL, CALR). Patients with established primary myelofibrosis who developed monocytosis and those with chronic myelomonocytic leukemia with marrow fibrosis were excluded. By combining the pathology databases of two large institutions, six eligible cases were identified. Patients were predominantly male and elderly with monocytosis at diagnosis (average 17.5%/2.3 × 103/μl), organomegaly, primary myelofibrosis-like atypical megakaryocytes admixed with a variable number of chronic myelomonocytic leukemia-like hypolobated forms, variable myelodysplasia, marrow fibrosis and osteosclerosis. All had a normal karyotype and no myelodysplasia-associated cytogenetic abnormalities. Five of the patients in whom a more extensive molecular characterization was performed showed co-mutations involving JAK2 or MPL and ASXL1, SRSF2, TET2, NRAS, and/or KRAS. Disease progression has occurred in all and two have died. Rare patients present with features that overlap between primary myelofibrosis and chronic myelomonocytic leukemia and are thus difficult to classify based on current World Health Organization criteria. Biologically, these cases likely represent primary myelofibrosis with monocytosis, dysplasia, and secondary (non-driver) mutations at presentation. Alternatively, they may represent a true gray zone of neoplasms. Their clinical behavior appears aggressive and innovative therapeutic approaches may be beneficial in this particular subset.

The unparalleled heterogeneity in genetic causes of hearing loss along with remarkable differences in prevalence of causative variants among ethnic groups makes single gene tests technically inefficient. Although hundreds of genes have been reported to be associated with nonsyndromic hearing loss (NSHL), GJB2, GJB6, SLC26A4, and mitochondrial (mt) MT-RNR1 and MTTS are the major contributors. In order to provide a faster, more comprehensive and cost effective assay, we constructed a DNA fluidic array, CapitalBioMiamiOtoArray, for the detection of sequence variants in five genes that are common in most populations of European descent. They consist of c.35delG, p.W44C, p.L90P, c.167delT (GJB2); 309kb deletion (GJB6); p.L236P, p.T416P (SLC26A4); and m.1555A>G, m.7444G>A (mtDNA). We have validated our hearing loss array by analyzing a total of 160 DNAs samples. Our results show 100% concordance between the fluidic array biochip-based approach and the established Sanger sequencing method, thus proving its robustness and reliability at a relatively low cost.

Here we report the case of a 30-year-old woman with relapsed acute myeloid leukemia (AML) who was treated with all-trans retinoic acid (ATRA) as part of investigational therapy (NCT02273102). The patient died from rapid disease progression following eight days of continuous treatment with ATRA. Karyotype analysis and RNA-Seq revealed the presence of a novel t(4;15)(q31;q22) reciprocal translocation involving the TMEM154 and RASGRF1 genes. Analysis of primary cells from the patient revealed the expression of TMEM154-RASGRF1 mRNA and the resulting fusion protein, but no expression of the reciprocal RASGRF1-TMEM154 fusion. Consistent with the response of the patient to ATRA therapy, we observed a rapid proliferation of t(4;15) primary cells following ATRA treatment ex vivo. Preliminary characterization of the retinoid response of t(4;15) AML revealed that in stark contrast to non-t(4;15) AML, these cells proliferate in response to specific agonists of RARα and RARγ. Furthermore, we observed an increase in the levels of nuclear RARγ upon ATRA treatment. In summary, the identification of the novel t(4;15)(q31;q22) reciprocal translocation opens new avenues in the study of retinoid resistance and provides potential for a new biomarker for therapy of AML.

Objectives: Primary mediastinal large B-cell lymphomas (PMLBCLs) are aggressive lymphomas with characteristic clinical, morphologic, and immunophenotypic features. “Double-hit” (DH) lymphomas are B-cell neoplasms characterized by a translocation involving MYC and either BCL2 or BCL6 . In the indexed literature, there are no reported cases of PMLBCL associated with DH or triple-hit events. Methods: Herein, we present a case of a 15-year-old girl with PMLBCL who had typical clinical, morphologic, and immunophenotypic features. Results: Fluorescent in situ hybridization studies showed rearrangements involving MYC and BCL6 . We also excluded the possibility of a reciprocal t(3;8) (3q27;8q24) BCL6 / MYC translocation. Conclusions: This case expands the current spectrum of lymphomas subtypes in which DH can be found and supports the rationale for cytogenetic testing for DH abnormalities in all cases of aggressive large B-cell lymphomas regardless of subtype.

The status of human epidermal growth factor receptor 2 (HER2, ERBB2) determines the eligibility of breast cancer patients to receive HER2-targeted therapy. The majority of HER2 testing in the U.S. is performed using a combination of immunohistochemistry (IHC) screening followed by fluorescence in situ hybridization (FISH) for IHC equivocal cases. In 2013, the American Society of Clinical Oncology (ASCO)/College of American Pathologists (CAP) updated the guideline for HER2 testing. This study evaluates the impact of the 2013 ASCO/CAP updated guideline on final HER2 FISH classification of breast cancers with an equivocal IHC result. For each case, we reported a FISH result according to the 2013 updated guideline and recorded a separated result using the 2007 guideline for investigational purpose. McNemar’s test and Bowker’s symmetry test were used to compare the classifications by the two guidelines. Among 172 HER2 IHC 2+ equivocal cases, use of the 2103 guideline changed classifications in 36 cases (21 %) when compared with the results expected by use of the 2007 guideline, and yielded a higher proportion of positive (28.5 vs. 23.3 %) and equivocal (16.3 vs. 4.1 %), and a lower proportion of negative (55.2 vs. 72.7 %) cases (p < 0.001). The major classification change with use of the updated guideline is from the HER2 FISH negative to equivocal in 26 cases (15 %). Our study has shown that implementation of the 2013 ASCO/CAP updated guideline has significant impact on HER2 classification for breast cancers with an equivocal HER2 IHC result and therefore increased the use of HER2-targeted therapy. Our data have also shown that reflex FISH is effective for final classification of the IHC equivocal cases and that polysomy 17 (CEP17 copy number ≥3/cell) is present in a significantly higher proportion of cases with an equivocal HER2 FISH classification.

Objective Mucinous cystadenocarcinoma of appendix is a rare entity. Differentiating mucinous cystadenocarcinoma from mucinous cystadenoma is very challenging and depends on establishing the presence of malignant cells in the appendix wall. The invasion may be very difficult to assess in some cases, especially in early stages of the disease, which could have devastating prognostic effects on patients. Therefore, it is necessary to develop an ancillary test that can differentiate the mucinous cystadenocarcinoma from mucinous cystadenoma. So far, there is no report available about the role of differentially expressed miRNAs in the diagnosis of appendiceal mucinous cystadenocarcinoma. Materials and Methods Six confirmed mucinous appendiceal cystadenocarcinoma and twelve mucinous appendiceal cystadenoma cases were selected. The total RNAs were extracted from the formalin-fixed paraffin-embedded specimen of these cases. The comprehensive miRNA microarray expression profiling from pooled aliquots of RNA samples from these two entities were analyzed to detect the differentially expressed miRNAs in mucinous cystadenocarcinoma. The best seven differentially expressed miRNAs were validated in individual cases by quantitative reverse transcriptase PCR (qRT-PCR). Results The microarray miRNA expression profiling analysis revealed 646 miRNAs that were differentially expressed in the mucinous cystadenocarcinoma. Among these differentially expressed miRNAs, the expression of 80 miRNAs showed statistical difference (p<0.01). The quantitative RT-PCR validated that the expression of miR-1, miR-4328 was significantly down regulated in mucinous cystadenocarcinoma compared to the mucinous cystadenoma (p<0.05). On the other hand, the expression of miR-200b, miR-200c, miR-451, miR-223 and miR-21 were significantly upregulated in mucinous cystadenocarcinoma (p<0.05). Conclusion The expression levels of miRNAs tested were significantly altered in the appendiceal mucinous cystadenocarcinoma samples compared to the mucinous cystadenoma. These data suggest that the miRNA expression in mucinous appendiceal neoplasm may help to supplement the morphological evaluation in distinguishing benign from malignant tumors.

Introduction: We describe for the first time, a novel chromosomal translocation in AML. The cells bearing this translocation proliferated rapidly following stimulation with ATRA in vitro. A 30-year old woman presented with monocytic AML. Karyotype revealed: 46,XX,t(9;11)(p22;q23)[13]/46,XX[7]. Mutational analysis showed an activating mutation in NRAS (c.34G>A; p.G12S). She achieved remission with standard induction therapy followed by cytarabine consolidation in the absence of an optimal donor. Following a disease free interval of 8 months, she relapsed. Karyotype at relapse: 46,XX,t(4;15)(q31;q22),t(9;11)(p22;q23)[20] (Figure 1). Mutational analysis revealed no new mutations. In the setting of chemotherapy refractory disease, she was enrolled on a phase I clinical trial combining escalating doses of TCP (tranylcypromine, Parnate¨) with fixed doses of ATRA (NCT02273102). The patient died from rapid disease progression shortly after her first cycle of therapy. Unexpectedly, we observed a rapid proliferation in the patient’s blasts following treatment with ATRA in vitro. We hypothesized that the novel t(4;15) translocation was involved in the regulation of retinoic acid (RA) signaling, and may have contributed to the rapid disease progression observed in this patient when she was treated with ATRA. To this end, we fully characterized the fusion gene and created a cell line bearing the translocation. Using these cells, we will further elucidate the mechanisms accounting for a rare and potentially clinically relevant effect of ATRA, in vitro.

333 Background: Identification of pts with exquisite sensitivity and/or durable responses to targeted therapies may lead to improved patient selection and allow for more rational treatment designs. Exceptional responders to everolimus in NET including pancreatic (PNET) were observed in our cohort of pts. PTEN is a key negative regulator of the phosphatidylinositol 3-kinase(PI3K)/Akt and mammalian target of rapamycin (mTOR) pathway. Loss of PTEN tumor suppressor gene function, usually due to deletion, leads to PI3K/Akt/mTOR pathway activation. Inthis study, we explored the role of PTEN as a potential predictive marker of everolimus in pts with NET including PNET. Methods: Between 2010 and 2014, pts with well-differentiated unresectable and metastatic gastrointestinal NET treated at our institution with everolimus were identified. 17 patients had pathology specimens available for testing. PTEN loss detection by FISH was carried out using a commercially available probe for cytoband 10q23, and by IHC using a commercially available antibody. Patients’ response to everolimus was evaluated through August 2014. The primary outcome was PFS and PTEN status was correlated with PFS for any potential association. Results: The median age was 59 years (range 45-78); 7 were male; 8 had PNET (2 gastrinomas, 1 insulinoma and 5 non-functional); 7 had small bowel NET and 2 unknown primary. All pts received everolimus starting at 10mg daily and octreotide LAR. Of the pts with PNETs, 3 had PTEN loss by FISH. Of those, one did not tolerate everolimus. The PFS for the other two pts was 8 and 24 months respectively. Among the 4 pts with intact PTEN; PFS was 14, 26, 3 and 12 months. 1 patient had insufficient tumor for testing. PTEN FISH is ongoing in the 9 non-pancreatic NET pts. PTEN expression by IHC is also ongoing and will be reported at the meeting. Conclusions: Testing for PTEN loss by FISH is feasible. Due to the small sample size, the role of PTEN loss could not be defined as a predictive marker in PNET. Testing on additional cases is ongoing and will be presented at the meeting.

Abstract Introduction: Identification of patients with exquisite sensitivity and/or durable responses to targeted therapies may lead to improved patient selection and allow for more rational treatment designs. Loss of PTEN tumor suppressor gene function, usually due to deletion, leads to activation of phosphatidylinositol 3-kinase/Akt and mammalian target of rapamycin (mTOR) signaling. The RADIANT-3 trial of everolimus in advanced PNETs demonstrated a significant prolongation of progression-free survival (PFS) from 4.6 months with placebo versus 11 months with everolimus. Despite the improvement in PFS, the response rate was only 5% among patients receiving everolimus. The index case for our study was an exceptional responder who had a significant radiological response as well as a PFS of 24 months, which were both better than expected from the literature. This led to the hypothesis that there may be mutational changes in genes affecting the mTOR pathway that could predict sensitivity to mTOR inhibitors. In this study, we specifically explored the role of PTEN loss as a potential predictive marker. Methods and Materials: Between 2010 and 2014, patients with well-differentiated unresectable and metastatic PNETs treated at the University of Miami/Sylvester Comprehensive Cancer Center and Jackson Memorial Hospital with everolimus were identified. Eight patients had pathology specimens available for testing. PTEN loss detected by Fluorescence In Situ Hybridization (FISH) was carried out using a commercially available probe for cytoband 10q23. Patients' response to everolimus was evaluated through June 2014. The primary outcome was PFS. PTEN expression by immunohistochemistry (IHC) will also be performed and the results will be compared to those obtained by FISH. Results: The median age was 60 years (range 45-78). 50% of the patients were females and 50% were males. Two patients had gastrinomas, 1 patient had an insulinoma, and five patients had non-functional PNETs. All patients had unresectable metastases to the liver. In addition to sandostatin LAR, the patients received everolimus starting at a dose of 10mg daily. Three patients were found to have deletion of PTEN. Of those, one patient did not tolerate everolimus and the PFS for the other two was 8 and 24 months respectively. Detection of PTEN loss by FISH yielded no results in 2 patients due to insufficient tumor left in the specimen. PFS in these 2 patients was 24 and 4 months respectively. Testing is ongoing in the last 3 patients and the PFS for these patients is 3, 10 and 13 months. PTEN expression by IHC is also ongoing and will be reported at the meeting. Conclusion: The index case for this study had a PTEN deletion and had a partial response to treatment and prolonged disease control for 2 years with everolimus. Testing is ongoing in additional cases to determine if there is a consistent correlation between PTEN loss by FISH and PFS. Citation Format: Moh'd Khushman, Joyce Slingerland, Yao-Shan Fan, Monica Garcia-Buitrago, Ernesto Bustinza, Maria Restrepo, Daniel Sussman, Caio Rocha-Lima, Peter Hosein. Exploring phosphatase and tensin homolog (PTEN) loss as a potential predictive marker for response to everolimus in patients with pancreatic neuroendocrine tumors (PNETs). [abstract]. In: Proceedings of the AACR Special Conference: Targeting the PI3K-mTOR Network in Cancer; Sep 14-17, 2014; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(7 Suppl):Abstract nr A01.

Mammary analog secretory carcinoma of salivary gland is a recently described entity with unique morphologic, clinical, and genetic characteristics, including the characteristic t(12;15)(p13;q25) with ETV6-NTRK3 translocation found in secretory carcinomas of the breast. Before their initial description, these salivary gland tumors were generally diagnosed as acinic cell carcinoma or adenocarcinoma. For the purpose of this study, all cases of salivary gland acinic cell carcinoma, cribriform cystadenocarcinoma, and adenocarcinoma, not otherwise specified (NOS), diagnosed over a 10-year period were retrieved from our surgical pathology files. There were a total of 11 cases diagnosed as acinic cell carcinoma, 10 cases of adenocarcinoma, NOS, and 6 cases of cribriform cystadenocarcinoma. All slides were reviewed by two pathologists (AP, CGF) and tumors that show morphologic features of mammary analog secretory carcinoma according to the recent literature were selected. This process narrowed down the initial number to six cases originally diagnosed as acinic cell carcinoma, three cases originally diagnosed as adenocarcinoma, NOS, and one case originally diagnosed as cribriform cystadenocarcinoma. The 10 cases were subjected to immunohistochemistry for S-100, mammaglobin, and ANO1, as well as fluorescence in situ hybridization analysis for t(12;15)(p13;q25) with ETV6-NTRK3 fusion rearrangement. The ETV6-NTRK3 gene rearrangement was detected in three tumors. These three tumors, initially diagnosed as acinic cell carcinomas, stained positive for S-100 and mammaglobin, and negative for ANO1 by immunohistochemistry. Two of the three patients were male (2/3). In summary, mammary analog secretory carcinoma is a newly described diagnostic entity that should be in the differential diagnosis of salivary gland tumors that morphologically mimic other neoplasms, mainly acinic cell carcinomas. They differ from conventional acinic cell tumors immunohistochemically and molecularly. Positivity for mammaglobin and S-100, and negativity for ANO1 are useful screening tools before confirmatory molecular studies.

Lymphoma is a common malignancy observed in companion animals. This type of naturally occurring neoplasia has been uncommonly reported in great apes. Diffuse large B-cell lymphoma was diagnosed in an 8-yr-old captive orangutan (Pongo pygmaeus) with gastrointestinal disease by histologic and immunohistochemical methodologies. The orangutan was treated with three cycles of combination chemotherapy (intravenous Rituxan, cyclophosphamide, doxorubicin, and vincristine). The primate has been in good health and exhibiting normal behaviors for more than 15 mo following treatment.

Introduction: Acute T-cell leukemia/lymphoma (ATLL) is a highly aggressive malignancy caused by HTLV-I, which is endemic in Japan, the Caribbean, and South America. ATLL carries a dismal prognosis and is generally incurable with conventional chemotherapy. ATLL is challenging to study at the molecular level, in part due to its complex genetic alterations likely resulting from years of HTLV-I driven T-cell proliferation and accumulation of genetic damage prior to malignant transformation. While no specific chromosome or genetic abnormalities have been proven to contribute to the pathogenesis of ATLL, older comparative genomic hybridization (CGH) studies performed in Japanese patients have demonstrated frequent genetic lesions (gains and losses) involving specific chromosomal regions, thus limited information exists about the chromosomal abnormalities occurring in the African ATLL variant commonly seen in the Western World.

KEY CLINICAL MESSAGE:Acute myeloid leukemia (AML) with t(8:16) is an infrequent acute leukemia subtype. It can occur de novo or more frequently therapy-related. The presence of blasts with monocytoid morphology and erythrophagocytosis suggest the presence of the t(8;16).

BACKGROUND AND AIM:Diagnosis of pancreatic malignancy is often based on cytological specimens collected by endoscopic ultrasound guided fine needle aspiration (EUS FNA). Several factors can decrease sensitivity of EUS FNA for pancreatic cancer: well-differentiated tumors, pancreatitis, blood, necrosis and slides with low cellularity. The objective of this study is to report on the use of fluorescence in situ hybridization (FISH) analysis combined with cytology in pancreatic masses.METHODS:EUS database and medical records of patients referred for EUS between January 2009 through august 2013 were reviewed. Data on cytology, FISH and surgical pathology were reviewed. Surgical pathology, death or extended clinical follow-up were used to verify correct diagnosis of malignancy. FISH performed using a four-set DNA probe for chromosomes 3, 7, 17, and band 9p21 in patients with inconclusive immediate cytology reading. Sensitivity of cytology and FISH were compared.RESULTS:Study cohort comprised of 104 patients with FISH analysis on EUS FNA specimens of pancreatic masses (74 adenocarcinoma, 7 neuroendocrine tumor and 23 benign. Sensitivity of cytology and FISH for carcinoma was respectively: 62% and 81%. Sensitivity of FISH + cytology was 89%. The specificity of FISH and cytology was 100%. The most common abnormality on FISH was a 9p21 deletion seen in 43 patients (58%) followed by polysomy of 7 (46%). FISH detected malignancy in 23 patients with negative cytology.CONCLUSIONS:In patients with inconclusive immediate cytology reading, FISH is superior to cytology and improves overall sensitivity. The 9p21 deletion is the most common abnormality seen in this cohort of patients with pancreatic cancer.