No abstract

Artificial intelligence (AI) has shown promise for diagnosing prostate cancer in biopsies. However, results have been limited to individual studies, lacking validation in multinational settings. Competitions have been shown to be accelerators for medical imaging innovations, but their impact is hindered by lack of reproducibility and independent validation. With this in mind, we organized the PANDA challenge—the largest histopathology competition to date, joined by 1,290 developers—to catalyze development of reproducible AI algorithms for Gleason grading using 10,616 digitized prostate biopsies. We validated that a diverse set of submitted algorithms reached pathologist-level performance on independent cross-continental cohorts, fully blinded to the algorithm developers. On United States and European external validation sets, the algorithms achieved agreements of 0.862 (quadratically weighted κ, 95% confidence interval (CI), 0.840–0.884) and 0.868 (95% CI, 0.835–0.900) with expert uropathologists. Successful generalization across different patient populations, laboratories and reference standards, achieved by a variety of algorithmic approaches, warrants evaluating AI-based Gleason grading in prospective clinical trials.

The Gleason score is the most important prognostic marker for prostate cancer patients, but it suffers from significant observer variability. Artificial intelligence (AI) systems based on deep learning can achieve pathologist-level performance at Gleason grading. However, the performance of such systems can degrade in the presence of artifacts, foreign tissue, or other anomalies. Pathologists integrating their expertise with feedback from an AI system could result in a synergy that outperforms both the individual pathologist and the system. Despite the hype around AI assistance, existing literature on this topic within the pathology domain is limited. We investigated the value of AI assistance for grading prostate biopsies. A panel of 14 observers graded 160 biopsies with and without AI assistance. Using AI, the agreement of the panel with an expert reference standard increased significantly (quadratically weighted Cohen's kappa, 0.799 vs. 0.872; p = 0.019). On an external validation set of 87 cases, the panel showed a significant increase in agreement with a panel of international experts in prostate pathology (quadratically weighted Cohen's kappa, 0.733 vs. 0.786; p = 0.003). In both experiments, on a grouplevel, AI-assisted pathologists outperformed the unassisted pathologists and the standalone AI system. Our results show the potential of AI systems for Gleason grading, but more importantly, show the benefits of pathologist-AI synergy. Members of the ISUP Pathology Imagebase Expert Panel are listed below Acknowledgements.

Prostate adenocarcinoma is the second most prevalent cancer among males in Western countries, with an incidence that increases in direct proportion with age. A large body of evidence indicates that activation of the pathways regulated by androgen receptor plays a central role in the development and malignant progression of prostate cancer.

Deoxynucleoside analogs are used in the treatment of a variety of solid tumors. Their transport across the plasma membrane may determine their cytotoxicity and thus nucleoside transporter (NT) expression patterns may be of clinical relevance. Lack of appropriate antibodies for use in paraffinembedded biopsies has been a bottleneck to undertake highthroughput analysis of NT expression in solid tumors. Here we report the characterization of 2 new antibodies raised against the low-affinity equilibrative NTs, hENT1 and hENT2, suitable for that purpose. These 2 antisera, along with a previously characterized antibody that specifically recognizes the high-affinity Na-dependent concentrative NT, hCNT1, have been used to analyze, using a tissue array approach, NT expression in gynecologic cancers (90 ovarian, 80 endometrial and 118 uterine cervix carcinomas). Human CNT1 was not detected in 33% and 39% of the ovarian and uterine cervix carcinomas, respectively, whereas hENT1 and hENT2 expression was significantly retained in a high percentage of tumors (91% and 96% for hENT1, 84% and 98% for hENT2, in ovarian and cervix carcinomas, respectively). Only a few endometrial carcinomas (15%) were found to be negative for hCNT1, but they all retained hENT1 and hENT2 expression. In ovarian cancer, the loss of all 3 NT proteins was a more common event in the clear cell histologic subtype than in the serous, mucinous and endometrioid histotypes. In uterine cervix tumors, the loss of expression of hCNT1 was significantly associated with the adenocarcinoma subtype. In summary, hCNT1 was by far the isoform whose expression was most frequently reduced or lost in the 3 types of gynecologic tumors analyzed. Moreover, NT expression is related to the type of gynecologic tumor and its specific subtype, hCNT1 protein loss being highly correlated with poor prognosis histotypes. Since hCNT1, hENT1 and hENT2 recognize fluoropyrimidines as substrates, but with different affinities, this study anticipates high variability in drug uptake efficiency in solid tumors.

No abstract

Data availabilitySummary statistics generated by COVID-19 Host Genetics Initiative are available online (https://www.covid19hg.org/results/r6/). The analyses described here use the freeze 6 data. The COVID-19 Host Genetics Initiative continues to regularly release new data freezes. Summary statistics for samples from individuals of non-European ancestry are not currently available owing to the small individual sample sizes of these groups, but the results for 23 loci lead variants are reported in Supplementary Table 3. Individual-level data can be requested directly from the authors of the contributing studies, listed in Supplementary Table 1.

BackgroundTrastuzumab improves survival outcomes in patients with HER2+ metastatic breast cancer. The Long-Her study was designed to identify clinical and molecular markers that could differentiate long-term survivors from patients having early progression after trastuzumab treatment.MethodsData were collected from women with HER2-positive metastatic breast cancer treated with trastuzumab that experienced a response or stable disease during at least 3 years. Patients having a progression in the first year of therapy with trastuzumab were used as a control. Genes related with trastuzumab resistance were identified and investigated for network and gene functional interrelation. Models predicting poor response to trastuzumab were constructed and evaluated. Finally, a mutational status analysis of selected genes was performed in HER2 positive breast cancer samples.Results103 patients were registered in the Long-HER study, of whom 71 had obtained a durable complete response. Median age was 58 years. Metastatic disease was diagnosed after a median of 24.7 months since primary diagnosis. Metastases were present in the liver (25%), lungs (25%), bones (23%) and soft tissues (23%), with 20% of patients having multiple locations of metastases. Median duration of response was 55 months. The molecular analysis included 35 patients from the group with complete response and 18 patients in a control poor-response group. Absence of trastuzumab as part of adjuvant therapy was the only clinical factor associated with long-term survival. Gene ontology analysis demonstrated that PI3K pathway was associated with poor response to trastuzumab-based therapy: tumours in the control group usually had four or five alterations in this pathway, whereas tumours in the Long-HER group had two alterations at most.ConclusionsTrastuzumab may provide a substantial long-term survival benefit in a selected group of patients. Whole genome expression analysis comparing long-term survivors vs. a control group predicted early progression after trastuzumab-based therapy. Multiple alterations in genes related to the PI3K-mTOR pathway seem to be required to confer resistance to this therapy.