A Comparative Study of Patient-Derived Tumor Models of Pancreatic Ductal Adenocarcinoma Involving Orthotopic Implantation

Pan-cancer molecular subtypes of metastasis reveal distinct and evolving transcriptional programs

Molecular mechanisms underlying cancer metastasis span diverse tissues of origin. Here, we synthesize and collate the transcriptomes of patient-derived xenografts and patient tumor metastases, and these data collectively represent 38 studies and over 3,000 patients and 4,000 tumors. We identify four expression-based subtypes of metastasis transcending tumor lineage. The first subtype has extensive copy alterations, higher expression of MYC transcriptional targets and DNA repair genes, and bromodomain inhibitor response association. The second subtype has higher expression of genes involving metabolism and prostaglandin synthesis and regulation. The third subtype has evidence of neuronal differentiation, higher expression of DNA and histone methylation genes and EZH2 transcriptional targets, and BCL2 inhibitor response association. The fourth subtype has higher expression of immune checkpoint and Notch pathway genes. The metastasis subtypes reflect expression differences from paired primaries, with subtype switching being common. These subtypes facilitate understanding of the molecular underpinnings of metastases beyond tissue-oriented domains, with therapeutic implications.

Serum-derived piR-hsa-164586 of extracellular vesicles as a novel biomarker for early diagnosis of non-small cell lung cancer

Non-small cell lung cancer (NSCLC) is a major cause of death in those with malignant tumors. To achieve the early diagnosis of NSCLC, we investigated serum-derived Piwi-interacting RNA (piRNA) of extracellular vesicles to filter diagnostic biomarkers for NSCLC. High-throughput sequencing from cancerous tissues and adjacent noncancerous tissues in patients with NSCLC was first applied to recognize candidate piRNAs as diagnostic biomarkers. These screened piRNAs were further validated in 115 patients (including 95 cases in stage I) and 47 healthy individuals using quantitative real-time PCR (qRT-PCR). We showed that piR-hsa-164586 was significantly upregulated compared with paracancerous tissues and extracellular vesicles from the serum samples of healthy individuals. Moreover, the area under the curve (AUC) value of piR-hsa-164586 was 0.623 and 0.624 to distinguish patients with all stages or stage I of NSCLC, respectively, from healthy individuals. The diagnostic performance of piR-hsa-164586 was greatly improved compared with the cytokeratin-19-fragment (CYFRA21-1). Additionally, piR-hs-164586 was associated with the clinical characteristics of patients with NSCLC. Its expression was associated with the age and TNM stage of patients with NSCLC, indicating that it can serve as an effective and promising biomarker for the early diagnosis of NSCLC.