In recent years, deep reinforcement learning is one of the hot topics of artificial intelligence, which can be applied in many fields. However, it also faces many problems and challenges, such as insufficient sample data, large sample space, complex action space and so on. The emergence of AlphaGo solved the problem of large sample space very well. After that, artificial intelligence such as AlphaGo Zero and AlphaStar were released. These intelligent frameworks can be applied in various scenarios. Differentiating from other works, in this paper, we focus on the in-depth analysis of the internal connections of Alpha series from the perspective of the problems and challenges solved to give an insight for the future development of reinforcement learning.

Background: Exosomal mRNAs secreted from cancer cells are regarded as important messengers in communication system between cells. In PC, some exosomal RNAs from patients’ plasma were expected as biomarkers of PC. However, exosomes in the blood of cancer patients are not derived only from cancer cells. Relationship of RNAs produced in cancer cells and those secreted into exosomes is still unclear. In this study, we evaluated relationships between mRNAs in PC cell lines and those of exosomes secreted into supernatant and determined candidates for PC biomarkers from cancer specific exosomal mRNAs. Material and Method: Three PC cell lines (PANC-1, MIAPaCa-2 and Capan-1) were used. Three studies were performed for each cell line. After 48h serum-free culture, the supernatants were collected, and exosomes were isolated by ultracentrifugation. The remaining cells in the culture dish were also collected and mRNAs were extracted. RNA-seq was performed for mRNAs extracted from cell and exosomal RNAs extracted from supernatant. Data was analyzed by R (version 4.2.1). Results: In MIAPaCa-2, expression of cellular mRNAs and corresponding exosomal mRNAs showed strong correlation (R was higher than 0.7). In the other 2 cell lines, they showed moderate correlation (R was higher than 0.5). Then, expression of mRNA in the cells was approximately reflected in those in the exosomes. However, some mRNAs showed high expression in exosome and low expression in the cell. We focused on these mRNAs, and then, performed a differential analysis of the whole raw RNA data by using TCC package of R and found differentially expressed mRNAs. Conclusion: Comparing with data of GEO database showing exosomal mRNA extracted from healthy volunteers, we finally find candidates of cancer specific exosomal mRNAs of PC. Citation Format: Yuhan Rong, Naoki Okubo, Risa Fukui, Akihiro Suzuki, Motokazu Sato, Motohiko Tokuhisa, Yuma Takeda, Noritoshi Kobayashi, Itaru Endo, Yasushi Ichikawa. Exosomal mRNAs secreted from pancreatic cancer (PC) cell lines might include good candidates of biomarkers of PC patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 879.

No abstract

Background: Most of CRC were well or moderately differentiated adenocarcinoma and P-ad-ca is very rare. In Japan, P-ad-ca was reported to be only 4-6% of all CRC. NEC is also rare. The incidence of that were reported to be less than 1%. One of their points of common is morphological poorly differentiated appearance. In this study, we evaluated the incidence of NEC contaminated in the CRC given a confirmed diagnosis as P-ad-ca. Material and Method: From 2009 to 2019, 816 CRCs were resected in department of Yokohama City University Hospital and 74 cases showing P-ad-ca or having a small region of P-ad-ca with well or moderately differentiated ad-ca (W-ad-ca, M-ad-ca). Among the 74 cases, there were 8 of P-ad-ca, 4 of W-ad-ca with P-ad-ca, 44 of M-ad-ca with P-ad-ca, and 18 W-ad-ca and M-ad-ca with P-ad-ca. Immunohistochemistry (IHC) using neuroendocrine markers (NM) including chromogranin A, synaptophysin and insulinoma-associated protein 1 (INSM1) was performed to identified NEC. Recently, a high incidence of Rb protein loss was reported in NEC of several organs, then IHC using Rb monoclonal antibody was also performed. The cancers showing immunohistochemically positive of at least one of the three NM and negative of Rb were identified as true NEC in this study. Results: At least one of the three NM was positive in the 39 cases. In these 39 cases, 3 cases were also Rb negative. Two of the three cases were region of P-ad-ca. These 2 P-ad-ca regions showed positive of NM. Interestingly, in the other one case, region of W-ad-ca showed positive of NM and negative of Rb. Region of P-ad-ca of the case showed positive of all three NM and Rb. From morphological appearance, this case was quite unlikely to true NEC. Conclusion: In 74 P-ad-ca, 2 cases might be true NEC in our small evaluation. NEC might be contaminated in the P-ad-ca of CRC which is difficult to be distinguished by only pathological findings of HE staining. Citation Format: Yuhan Rong, Yuma Takeda, Naoki Okubo, Akihiro Suzuki, Motohiko Tokuhisa, Noritoshi Kobayashi, Atsushi Ishibe, Ikuma Kato, Shoji Yamanaka, Satoshi Fujii, Itaru Endo, Yasushi Ichikawa. In colorectal cancer (CRC), neuroendocrine carcinoma (NEC) is contaminated in the cancer given a confirmed diagnosis of poorly differentiated adenocarcinoma (P-ad-ca) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5252.