Purpose. Pancreatic cancer (PC) is a common, highly lethal cancer with a low survival rate. Autophagy is involved in the occurrence and progression of PC. This study aims to explore the feasibility of using an autophagy-related long noncoding RNA (lncRNA) signature for assessing PC patient survival. Methods. We obtained RNA sequencing and clinical data of patients from the TCGA website. Autophagy genes were obtained from the Human Autophagy Database. The prognostic model, generated through univariate and multivariate Cox regression analyses, included 10 autophagy-related lncRNAs. Receiver operating characteristic (ROC) curves and forest plots were generated for univariate and multivariate Cox regression analyses, to examine the predictive feasibility of the risk model. Gene set enrichment analysis (GSEA) was used to screen enriched gene sets. Results. Twenty-eight autophagy-related lncRNAs were filtered out through univariate Cox regression analysis ( P < 0.001 ). Ten autophagy-related lncRNAs, including 4 poor prognosis factors and 6 beneficial prognosis factors, were further screened via multivariate Cox regression analysis. The AUC value of the ROC curve was 0.815. GSEA results demonstrated that cancer-related gene sets were significantly enriched. Conclusion. A signature based on ten autophagy-related lncRNAs was identified. This signature could be potentially used for evaluating clinical prognosis and might be used for targeted therapy against PC.

Multidrug combination therapy for pancreatic cancer is widely applied in clinical practice. In this study, we used phenylboronic acid and polyethylene glycol as materials of nanomicelles, loaded with the PI3K/mTORC1 dual inhibitor PF04691502 (PF) to inhibit the resistance and metastasis of pancreatic cancer and increase the sensitivity of doxorubicin (DOX). We prepared the PPD nanoparticles (NPs) with a small PDI and a uniform morphology by controlling the DOX substitution degree (size of 164.8 ± 3.6  nm and zeta potential of − 16.9 ± 0.4  mV). We determined the rates of PF and materials through the combination experiment of free drugs and the obtained PF@PPD NPs (size of 200.8 ± 2.6  nm and zeta potential of − 13.9 ± 0.3  mV). The drug loadings of DOX and PF in the nanomicelle were 14.8 ± 0.4 % and 9.5 ± 0.3 % , respectively. And the drug release in vitro was slow ( 29.17 ± 2.00 % for DOX and 39.22 ± 2.49 % for PF). The cell assay showed that the NPs had a good curative effect and migration on BxPC-3 cells, and it could be continuously taken up by cells. The PF@PPD NPs displayed a dose-dependent cytotoxicity with less cell viability ( 20.38 ± 1.11 % ) and higher uptake in BxPC-3 cells compared with the free drug. The combined medication or PF@PPD NPs reduced tumor metastasis, indicating that PF@PPD NPs had the potential for clinical application.