Background: Multiple studies have compared the performance of artificial intelligence (AI)ebased models for automated skin cancer classification to human experts, thus setting the cornerstone for a successful translation of AI-based tools into clinicopathological practice.

When a unilocular pancreatic cyst is found, without history of pancreatitis and gallstones, having low serum tumor markers levels and negativity of CA 72-4 and MCA in the cyst fluid, the diagnosis of the macrocystic variant of the serous cystadenoma may be suggested. At present, the diagnosis is still based on pathological examination after cyst removal.

No abstract

Thirteen melanocytic skin neoplasms with a consultation diagnosis by A. Bernard Ackerman were submitted to immunohistochemistry for HMB-45, Ki67, cyclin D1, e-cadherin, and p16; 9/13 cases underwent fluorescence in situ hybridization (FISH) test targeting 6p25 (RREB1), 6q23 (MYB), centromere 6 (Cep6), and 11q13 (CCND1), as well as the centromere 7 (Cep7). A "consensus diagnosis" among 3 experts was also advocated both before and after morphomolecular information. Three neoplasms with a consultation diagnosis of Spitz nevus showed at least 3 abnormal immunohistochemical patterns; 2 of these cases were also FISH-positive for CCND1 gain, but none of them had a final consensus diagnosis of melanoma. Two neoplasms with a consultation diagnosis of congenital nevus received a consensus diagnosis of melanoma. Molecular morphology techniques can highlight the atypical features of melanocytic neoplasms and support existence of a morphobiologic "spectrum": This should be mirrored in the final report by abandoning the dichotomic (benign vs malignant) diagnostic approach.

Serum HER-2/neu concentrations were evaluated in 172 healthy subjects, 176 primary and 55 metastatic breast cancer patients, employing a new automated assay (Bayer Immuno 1 serum HER-2/neu). Using 13 ng/mL as the cutoff, abnormal HER-2/neu serum levels were found in 8% (14/176) of primary and 50.9% (28/55) of metastatic breast cancer patients. Both in primary and metastatic breast cancer a significant relationship was found with the stage of the disease when serum HER-2/neu was considered as a categorized variable (p=0.0003 and p=0.02, respectively), but not when it was taken as a continuous variable (p=0.247 and p=0.146, respectively). Moreover, we evaluated the correlation between Immuno 1 HER-2/neu and Oncogene Research Products ELISA assay in 53 normal subjects, 46 primary and 34 metastatic breast cancer patients. The correlation was relatively good (p<0.0001), although substantial differences could be found in single cases. The Immuno 1 assay was also evaluated for the first time in breast cancer tissue. The method, which showed good performance both in terms of imprecision and linearity, was used to measure HER-2/neu protein in 140 cytosol samples from primary breast cancer tissue and in homogenates from 40 matched cases. The correlation between the two matrixes was very close (p<0.0001). By contrast, no correlation was found between serum and matched cytosol (p=0.101) or homogenate samples (p=0.511).