Search

Deep learning systems have been proposed to improve the objectivity and efficiency of Ki- 67 PI scoring. The challenge is that while very accurate, deep learning techniques suffer from reduced performance when applied to out-of-domain data. This is a critical challenge for clinical translation, as models are typically trained using data available to the vendor, which is not from the target domain. To address this challenge, this study proposes a domain adaptation pipeline that employs an unsupervised framework to generate silver standard (pseudo) labels in the target domain, which is used to augment the gold standard (GS) source domain data. Five training regimes were tested on two validated Ki-67 scoring architectures (UV-Net and piNET), (1) SS Only: trained on target silver standard (SS) labels, (2) GS Only: trained on source GS labels, (3) Mixed: trained on target SS and source GS labels, (4) GS+SS: trained on source GS labels and fine-tuned on target SS labels, and our proposed method (5) SS+GS: trained on source SS labels and fine-tuned on source GS labels. The SS+GS method yielded significantly (p < 0.05) higher PI accuracy (95.9%) and more consistent results compared to the GS Only model on target data. Analysis of t-SNE plots showed features learned by the SS+GS models are more aligned for source and target data, resulting in improved generalization. The proposed pipeline provides an efficient method for learning the target distribution without manual annotations, which are time-consuming and costly to generate for medical images. This framework can be applied to any target site as a per-laboratory calibration method, for widescale deployment., Comment: Editors: Accepted for publication at MIDL 2023

Introduction: Prostatic adenocarcinoma can occasionally display urothelial carcinoma morphology, which prompts immunohistochemistry (IHC) studies to determine its lineage. Typically, prostate cancer is characterized by the lack of cytokeratin (CK) 7, CK20 and high molecular weight keratin (HMWK) expression, as opposed to bladder cancer., Methods: We report a series of 12 prostatic adenocarcinoma cases with unusual urothelial-like morphology, diagnosed at two academic institutions in Toronto between 2018 and 2023, and analyzed by immunohistochemistry for prostatic, urothelial, and neuroendocrine marker expression. We collected patient age, androgen deprivation therapy (ADT) status, tumour site, histomorphology, Grade group (GG) and results of genetic testing., Results: The median age of the 12 patients included in this case series was 75.5 years (range 41-85). A history of prostatic cancer was noted in 7/12 (58%) patients. Five of nine (56%) patients had elevated serum PSA level at diagnosis. Six of eleven (55%) patients had prior ADT. Tumour sites were prostate (n = 6), bladder (n = 3), liver metastases (n = 2), and lung metastasis (n = 1). GGs of the primary tumours were GG3 (n = 1) and GG5 (n = 8). The observed urothelial-like morphology was diffuse in ten cases, and focal in two cases. CK7 was strong/diffuse in 8/11 tested cases, and focal weak in one case. CK20, HMWK, p63 and GATA3 were patchy/focal/weak/moderate in 3/6, 4/7, 4/8 and 2/9 cases, respectively. Ten (83%) cases were positive for at least one prostatic marker; eight (67%) cases had loss/weak staining of at least one prostatic marker. AR loss was seen in 2/7 (29%) cases. Seven of ten (70%) cases had diffuse/strong expression of at least one neuroendocrine marker. No trend was evident between prior ADT/AR status and any IHC result. Molecular analyses for DNA damage repair (DDR) genes (n = 6) demonstrated one ATM deletion (bladder). In addition, one TMPRSS2:ERG fusion (lung metastasis) was identified., Conclusion: This series comprises high-grade and/or metastatic prostatic adenocarcinoma cases with distinctive urothelial-like morphology and frequent aberrant CK7/CK20/HMWK expression. Their histomorphology, highly suggestive of an urothelial origin, represents a diagnostic pitfall that can lead to considerable management repercussions. The fact that a high proportion of the reported cases had loss/weak expression of at least one of the tested prostatic-specific markers, and occasionally a diffuse positivity for neuroendocrine markers highlights the importance of (1) clinical history and (2) utilization of broad IHC panels to correctly diagnose such unusual prostate cancer cases., Competing Interests: Declaration of Competing Interest The authors report no declarations of interest., (Copyright © 2024 Elsevier GmbH. All rights reserved.)

Gigantomastia is an uncommon benign condition characterized by massive breast enlargement. It is most often due to hormonal imbalance secondary to puberty or pregnancy, or induced by a pharmacological agent but can also be idiopathic. Herein, we report a rare case of idiopathic gigantomastia in a 46-year-old female on antiepileptic multiple-drug therapy who underwent total bilateral mastectomy to relieve associated pain., (© 2021 The Authors. Published by the British Institute of Radiology.)