Background: Tumor Growth Factor β (TGF-β) plays an important role in cancer development. There are three TGF-β isoforms in human: β1, β2, β3. OT-101- an antisense against TGF-β2 – has been shown to be active against glioblastoma and pancreatic cancer and melanoma in various phase 2 clinical trials (NCT00431561, NCT00844064, NCT04862767). In the glioblastoma phase 2 trial, OT-101 single agent was as active as temozolomide (TMZ) (NCT00431561). Methods: To understand the MOA of OT-101, we analyzed the tumor microenvironment including TGF-β and IL-2 through CBioPortal and Km plotter portals to access more than 10,000 pts across multiple clinical trials (Cerami E et al., Cancer Discov (2012) 2: 401; Gao J et al., Sci-Signal. (2013) 6:269; Lanczky A et al., (2012) J Med Internet Res 26:e27633). Results: Surprisingly, only low TGF-β2 was predictive of OS and not TGF-β1 nor TGF-β3. Analysis was performed using quartiles for mRNA level and significance was declared only if the trend was consistent across all quartiles. As OT-101 is the only TGF-β2 specific therapeutic, this would explain the failure of other TGF-β clinical programs. Tumor microenvironment has significant impact on TGF-β2 driven survival including mutation load, regulatory T cell, CD8, Th2 regulatory T-cell, B-cell, macrophage, and MSC. The combination of these factors resulted in improved separation between the low TGF-β2 and the high TGF-β2. In pancreatic cancer (PDAC), macrophage and mutation burden/neoantigen burden were the most important factors. Combined, they drove median survival from 15 to 73 mos (p=7.6e-05). Low TGF-β2 enhanced survival across multiple therapeutics including FOFIRNOX for PDAC; TMZ, Radiation, and Bevacizumab for GBM, and PD-1 for melanoma. For radiation therapy in GBM, there was a steady increase in median survival across the quartiles (25, 73, 93, 95 mos, p=7.68e-08); for TMZ therapy there was a similar trend across the quartiles (25, 50, 114, 95 mos, p=5.25e-10). For PD-1 in melanoma, the ratio of IL-2/TGF-β2 was the most potent, driving upper quartile survival of patients treated with PD-1 from 6.7 to 15 mos (p=5.4e-06) with more than double the long-term survivors normally seen with PD-1 therapy. Conclusions: These findings are supportive of a significant role of TGF-β2, but not TGF-β1 nor TGF-β3, in cellular immunity against the tumor and support the development of OT-101 as an immunotherapeutic agent. There are currently multiple combination trials with OT-101 including IL-2, PD-1, and FOFIRINOX with the pivotal phase 3 trials being planned for PDAC and diffuse midline glioblastoma (DMG).