Computational pathology, utilizing whole slide images (WSIs) for pathological diagnosis, has advanced the development of intelligent healthcare. However, the scarcity of annotated data and histological differences hinder the general application of existing methods. Extensive histopathological data and the robustness of self-supervised models in small-scale data demonstrate promising prospects for developing foundation pathology models. Here we show BEPH (BEiT-based model Pre-training on Histopathological image), a foundation model that leverages self-supervised learning to learn meaningful representations from 11 million unlabeled histopathological images. These representations are then efficiently adapted to various tasks, including patch-level cancer diagnosis, WSI-level cancer classification, and survival prediction for multiple cancer subtypes. By leveraging the masked image modeling (MIM) pre-training approach, BEPH offers an efficient solution to enhance model performance, reduce the reliance on expert annotations, and facilitate the broader application of artificial intelligence in clinical settings. The pre-trained model is available at https://github.com/Zhcyoung/BEPH.

Mediation analysis with high-dimensional DNA methylation markers is important in identifying epigenetic pathways between environmental exposures and health outcomes. There have been some methodology developments of mediation analysis with high-dimensional mediators. However, high-dimensional mediation analysis methods for time-to-event outcome data are still yet to be developed. To address these challenges, we propose a new high-dimensional mediation analysis procedure for survival models by incorporating sure independent screening and minimax concave penalty techniques for variable selection, with the Sobel and the joint method for significance test of indirect effect. The simulation studies show good performance in identifying correct biomarkers, false discovery rate control, and minimum estimation bias of the proposed procedure. We also apply this approach to study the causal pathway from smoking to overall survival among lung cancer patients potentially mediated by 365,307 DNA methylations in the TCGA lung cancer cohort. Mediation analysis using a Cox proportional hazards model estimates that patients who have serious smoking history increase the risk of lung cancer through methylation markers including cg21926276, cg27042065, and cg26387355 with significant hazard ratios of 1.2497(95%CI: 1.1121, 1.4045), 1.0920(95%CI: 1.0170, 1.1726), and 1.1489(95%CI: 1.0518, 1.2550), respectively. The three methylation sites locate in the three genes which have been showed to be associated with lung cancer event or overall survival. However, the three CpG sites (cg21926276, cg27042065 and cg26387355) have not been reported, which are newly identified as the potential novel epigenetic markers linking smoking and survival of lung cancer patients. Collectively, the proposed high-dimensional mediation analysis procedure has good performance in mediator selection and indirect effect estimation. In this research, we established an efficient procedure for high-dimensional mediation analysis with time-to-event data to select DNA methylation markers and estimate the mediation effects. To evaluate the performance of the proposed procedure, we conducted extensive simulation studies and analyzed a lung cancer data set using our method. We demonstrated the validity and utility of our method under a variety of scenarios.

CA125 is a prevalently used serum biomarker for detecting ovarian cancer over the last three decades. However, it has a significant deficiency in screening for early-stage cancer. With the purpose of exploring an effective approach to improve its performance in early diagnosis, we investigated the postprandial fluctuation pattern of cancer-derived CA125 and the underlying mechanism. In two medical centers, 551 patients sonographically diagnosed with ovarian (adnexal) cysts (< 5 cm in diameter) were enrolled and divided into five disease groups (pelvic inflammatory cysts, retention cysts, endometrioma, benign/borderline cystadenoma and malignant cysts). The subtle differences in 1-h postprandial serum CA125 increases were compared between disease groups. A support vector machine (SVM)-based algorithm was used for refining the performance of CA125 postprandial increment. Ovarian cancer xenograft animal and cancer cell models were used to recapitulate the clinical findings and reveal the molecular basis of postprandial blood glucose and insulin in invoking the synthesis/secretion/re-absorption of CA125. Patients with ovarian cancer presented the highest postprandial increment 13.3 ± 0.7% (mean ± standard deviation) among the five disease groups. Using a CA125 increment ≥ 10% criterion, the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) reached 83.3, 96.3, 61.1 and 98.8%, respectively, for early-stage ovarian cancer. This performance was further improved by the SVM-based CA125-increment algorithm, which exhibited 91.7% sensitivity, 99.2% specificity, 89.2% PPV and 99.4% NPV. Both modalities manifested diagnostic advantages over the traditional CA125 test (75.0% sensitivity, 25.4% specificity, 6.6% PPV and 93.6% NPV at the cut-off of 35 U/mL). Regarding the molecular basis, the postprandial blood glucose and insulin-invoked overexpression of Mucin 16 (encoding CA125) were demonstrated in animal and cancer cell models, which were mediated by the PI3K-Akt pathway. Nevertheless, a Mesothelin-based CA125 re-absorption behavior was noted in the treated cancer cells, which contributed to the over-drop following the postprandial peak of serum CA125. Cancer-derived serum CA125 possesses a unique and distinctive postprandial pattern, that distinguishes it from the common CA125 elevation in a benign disease condition. The dynamic measurement/assessment strategy can achieve a discriminatory power superior to that of a static test.

Polyunsaturated fatty acids (PUFAs), especially long-chain PUFAs (LCPUFAs), are crucial for both the structural and functional integrity of cells. PUFAs have been reported to be insufficient in schizophrenia, and the resulting cell membrane impairments have been hypothesized as an etiological mechanism. However, the impact of PUFA deficiencies on the onset of schizophrenia remain uncertain. We investigated the associations between PUFAs consumption and schizophrenia incidence rates through correlational analyses and conducted Mendelian randomization analyses to reveal the causal effects. Using dietary PUFA consumption and national schizophrenia incidence rates in 24 countries, we found that incidence rates of schizophrenia were inversely correlated with arachidonic acid (AA) and ω-6 LCPUFA consumption (r_AA = −0.577, p < 0.01; r_{ω-6 LCPUFA} = −0.626, p < 0.001). Moreover, Mendelian randomization analyses revealed that genetically predicted AA and gamma-linolenic acid (GLA) were protective factors against schizophrenia (OR_AA = 0.986, OR_GLA = 0.148). In addition, no significant relationships were observed between schizophrenia and docosahexaenoic acid (DHA) or other ω-3 PUFAs. These findings show that the deficiencies of ω-6 LCPUFAs, especially AA, are associated with schizophrenia risk, which sheds novel insight into the etiology of schizophrenia and a promising diet supplementation for the prevention and treatment of schizophrenia.