British Journal of Cancer

doi: 10.1038/s41416-023-02169-2pmid: 36737658

BackgroundInflammatory, insulin and oestrogenic pathways have been linked to breast cancer (BC). We aimed to examine the relationship between pre-diagnostic dietary patterns related to these mechanisms and BC survival.MethodsThe diabetes risk reduction diet (DRRD), inflammatory score of diet (ISD) and oestrogen-related dietary pattern (ERDP) were calculated using dietary data from the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Cox proportional hazards models were used to assess associations between dietary patterns and overall mortality and competing risk models for associations with BC-specific mortality.ResultsWe included 13,270 BC cases with a mean follow-up after diagnosis of 8.6 years, representing 2340 total deaths, including 1475 BC deaths. Higher adherence to the DRRD score was associated with lower overall mortality (HR1–SD 0.92; 95%CI 0.87–0.96). Greater adherence to pro-inflammatory diets was borderline associated with 6% higher mortality HR1–SD 1.06; 95%CI 1.00–1.12. No significant association with the oestrogen-related dietary pattern was observed. None of the dietary patterns were associated with BC-specific mortality.ConclusionsGreater adherence to an anti-diabetic and anti-inflammatory diet prior to diagnosis is associated with lower overall mortality among BC survivors. Long-term adherence to these dietary patterns could be a means to improve the prognosis of BC survivors.

BackgroundWe aimed to develop and validate a plasma extracellular vesicle circular RNA (circRNA)-based signature that can predict overall survival (OS) in first-line abiraterone therapy for metastatic castration-resistant prostate cancer (mCRPC) patients.MethodsIn total, 582 mCRPC patients undergoing first-line abiraterone therapy from four institutions were sorted by three phases. In the discovery phase, 30 plasma samples from 30 case-matched patients with or without early progression were obtained to generate circRNA expression profiles using RNA sequencing. In the training phase, differentially expressed circRNAs were examined using digital droplet PCR in a training cohort (n = 203). The circRNA signature was constructed using a least absolute shrinkage and selection operator Cox regression to predict OS. In the validation phase, the prognostic ability of this signature was prospectively validated in two external cohorts (Cohort I, n = 183; Cohort II, n = 166).ResultsWe developed a five-circRNA signature, based on circCEP112, circFAM13A, circBRWD1, circVPS13C and circMACROD2, which successfully stratified patients into high-risk and low-risk groups. The prognostic ability of this signature was prospectively validated in two external cohorts (P < 0.0001, P < 0.0001). Patients with high-risk scores had shorter OS than patients with low-risk scores.ConclusionThis five-circRNA signature is a reliable predictor of OS for mCRPC patients undergoing abiraterone.

doi: 10.1038/s41416-023-02157-6pmid: 36717674

BackgroundColorectal cancer (CRC) primary tumours are molecularly classified into four consensus molecular subtypes (CMS1–4). Genetically engineered mouse models aim to faithfully mimic the complexity of human cancers and, when appropriately aligned, represent ideal pre-clinical systems to test new drug treatments. Despite its importance, dual-species classification has been limited by the lack of a reliable approach. Here we utilise, develop and test a set of options for human-to-mouse CMS classifications of CRC tissue.MethodsUsing transcriptional data from established collections of CRC tumours, including human (TCGA cohort; n = 577) and mouse (n = 57 across n = 8 genotypes) tumours with combinations of random forest and nearest template prediction algorithms, alongside gene ontology collections, we comprehensively assess the performance of a suite of new dual-species classifiers.ResultsWe developed three approaches: MmCMS-A; a gene-level classifier, MmCMS-B; an ontology-level approach and MmCMS-C; a combined pathway system encompassing multiple biological and histological signalling cascades. Although all options could identify tumours associated with stromal-rich CMS4-like biology, MmCMS-A was unable to accurately classify the biology underpinning epithelial-like subtypes (CMS2/3) in mouse tumours.ConclusionsWhen applying human-based transcriptional classifiers to mouse tumour data, a pathway-level classifier, rather than an individual gene-level system, is optimal. Our R package enables researchers to select suitable mouse models of human CRC subtype for their experimental testing.

doi: 10.1038/s41416-023-02141-0pmid: 36690721

BackgroundRisk-assessment of endometrial cancer (EC) is based on clinicopathological factors and molecular subgroup. It is unclear whether adding hormone receptor expression, L1CAM expression or CTNNB1 status yields prognostic refinement.MethodsParaffin-embedded tumour samples of women with high-risk EC (HR-EC) from the PORTEC-3 trial (n = 424), and a Dutch prospective clinical cohort called MST (n = 256), were used. All cases were molecularly classified. Expression of L1CAM, ER and PR were analysed by whole-slide immunohistochemistry and CTNNB1 mutations were assessed with a next-generation sequencing. Kaplan–Meier method, log-rank tests and Cox’s proportional hazard models were used for survival analysis.ResultsIn total, 648 HR-EC were included. No independent prognostic value of ER, PR, L1CAM, and CTNNB1 was found, while age, stage, and adjuvant chemotherapy had an independent impact on risk of recurrence. Subgroup-analysis showed that only in NSMP HR-EC, ER-positivity was independently associated with a reduced risk of recurrence (HR 0.33, 95%CI 0.15–0.75).ConclusionsWe confirmed the prognostic impact of the molecular classification, age, stage, and adjuvant CTRT in a large cohort of high-risk EC. ER-positivity is a strong favourable prognostic factor in NSMP HR-EC and identifies a homogeneous subgroup of NSMP tumours. Assessment of ER status in high-risk NSMP EC is feasible in clinical practice and could improve risk stratification and treatment.

doi: 10.1038/s41416-023-02143-ypmid: 36717673

BackgroundFast and accurate diagnostics are key for personalised medicine. Particularly in cancer, precise diagnosis is a prerequisite for targeted therapies, which can prolong lives. In this work, we focus on the automatic identification of gastroesophageal adenocarcinoma (GEA) patients that qualify for a personalised therapy targeting epidermal growth factor receptor 2 (HER2). We present a deep-learning method for scoring microscopy images of GEA for the presence of HER2 overexpression.MethodsOur method is based on convolutional neural networks (CNNs) trained on a rich dataset of 1602 patient samples and tested on an independent set of 307 patient samples. We additionally verified the CNN’s generalisation capabilities with an independent dataset with 653 samples from a separate clinical centre. We incorporated an attention mechanism in the network architecture to identify the tissue regions, which are important for the prediction outcome. Our solution allows for direct automated detection of HER2 in immunohistochemistry-stained tissue slides without the need for manual assessment and additional costly in situ hybridisation (ISH) tests.ResultsWe show accuracy of 0.94, precision of 0.97, and recall of 0.95. Importantly, our approach offers accurate predictions in cases that pathologists cannot resolve and that require additional ISH testing. We confirmed our findings in an independent dataset collected in a different clinical centre. The attention-based CNN exploits morphological information in microscopy images and is superior to a predictive model based on the staining intensity only.ConclusionsWe demonstrate that our approach not only automates an important diagnostic process for GEA patients but also paves the way for the discovery of new morphological features that were previously unknown for GEA pathology.[graphic not available: see fulltext]

doi: 10.1038/s41416-023-02180-7pmid: 36759721

BackgroundTrop-2 and Nectin-4 are transmembrane proteins overexpressed in many tumours and targets of antibody–drug conjugates (ADC). In metastatic colorectal cancer (mCRC), the role of Trop-2 and Nectin-4 has been poorly investigated.MethodsTumour samples of patients randomised in the phase III TRIBE2 were assessed for Trop-2 and Nectin-4 expression.ResultsThree hundred eighty-six tumours were assessed for Trop-2 expression. 90 (23%), 115 (30%) and 181 (47%) were Trop-2 high, medium and low, respectively. Patients with low Trop-2 tumours achieved longer PFS (12 versus 9.9 months, p = 0.047) and OS (27.3 versus 21.3 months, p = 0.015) than those with high/medium Trop-2 tumours. These findings were confirmed in multivariate analysis (p = 0.022 and p = 0.023, respectively). A greater OS benefit from treatment intensification with FOLFOXIRI/bevacizumab was observed in patients with high/medium Trop-2 tumours (p-for-interaction = 0.041).Two hundred fifty-one tumours were assessed for Nectin-4 expression. Fourteen (5%), 67 (27%) and 170 (68%) were high, medium and low, respectively. No prognostic impact was observed based on Nectin-4 expression and no interaction effect was reported between Nectin-4 expression groups and treatment arm.ConclusionsIn mCRC, expression levels of Trop-2 and Nectin-4 are heterogeneous, suggesting a target-driven development of anti-Trop2 and anti-Nectin-4 ADCs. Medium/high Trop-2 expression is associated with worse prognosis and higher benefit from chemotherapy intensification.