All-trans retinoic acid (ATRA) is used in the treatment of Acute Promyelocytic Leukemia (APL) with exceptional results, inducing long-lasting remissions in 78% of the patients. In previous studies (Bolis M et al., Ann Oncol. 2017, 28:611; Centritto F et al., EMBO Mol Med. 2015, 7:950) we demonstrated the therapeutic potential of ATRA in the personalized treatment of breast cancer. The present study is aimed at providing pre-clinical evidence supporting the use of pharmacological strategies based on ATRA in the management of stomach cancer, a very heterogeneous type of tumor, which lacks effective therapeutic options other than surgery. To define the sensitivity of gastric cancer to the anti-proliferative action of ATRA, we used 27 cell-lines, recapitulating the heterogeneity of the tumor. Fourteen of these cell-lines are characterized by a G-DIF phenotype, while the remainders present with a G-INT phenotype. Each cell-line was exposed to increasing concentrations of ATRA and the growth-inhibitory effects of the retinoid were evaluated at 3, 6, and 9 days. The 27 cell-lines were ranked for their sensitivity to the retinoid with the “ATRA-score”, a continuous/quantitative index, which we developed. The results indicate that the cell-lines can be divided into 3 separate groups characterized by high, intermediate and low sensitivity to ATRA. No significant correlation between ATRA-sensitivity and the G-DIF or G-INT phenotype is observed in our panel of cell-lines. To support the results obtained with the cell-lines, we evaluated the anti-proliferative effects exerted by ATRA on short-term tissue slice cultures of primary tumors derived from 13 cases of stomach cancer, which we classified as G-DIF or G-INT following determination of the basal gene-expression profiles with the use of RNA-sequencing studies. The results obtained with the short-term tissue culture model confirm the data generated with the cell-lines. The whole genome RNA-sequencing studies performed in the cell-lines and short-term tissue cultures in basal conditions allowed us to define a model consisting of 42 genes whose expression is quantitatively correlated with ATRA-sensitivity. The model was used to predict the sensitivity of gastric cancer cases to ATRA using the TCGA dataset. The results obtained indicate that approximately 60% of the gastric cancer cases are predicted to be characterized by high/intermediate sensitivity to ATRA, regardless of the G-DIF or G-INT phenotype. In conclusion, our results support the concept that ATRA-based therapeutic strategies are likely to be beneficial in the majority of stomach cancer cases. In addition, we generate a gene-expression tool that may be used for the selection of ATRA-sensitive patients in the clinics.
Less...