Abstract Somatic mutations accumulate as we age and are mostly found within the noncoding regions which make up of 98% of the human genome. CRISPR-Cas9, on the other hand, can be designed to induce double strand breaks (DSBs) at specific locations in the genome, and DSBs are known to be cytotoxic. CRISPR-Cas9 requires the presence of a 3-5 nucleotide protospacer adjacent motif (PAM) to serve as a binding signal for Cas9, in which Cas9 would not bind nor cut the target in its absence. We hypothesized that we could identify a large number of somatic 5’-NGG-3’ PAMs that arise from single base substitutions (SBSs) in pancreatic cancers (PCs; e.g. NCG -> NGG), and these novel PAMs could serve as cancer-specific CRISPR-Cas9 targets for selective cell elimination. First, we performed whole genome sequencing (WGS) of three PC cell lines and their corresponding normal cell lines followed by tumor-normal (T-N) subtraction to identify somatic SBSs. PAMfinder, a software designed for somatic PAM discovery, was executed to identify somatic PAMs in each case. On average, 4548 somatic SBSs were found per PC case, in which 417 of them produced somatic PAMs (9.2% of total SBSs). PAMs with >95% variant allele frequency (VAF) and that produced sgRNAs with minimal off-target activity (based on CRISPOR) were further validated through Sanger sequencing, which resulted in an average of 32/33 confirmed targets per cell line. Mutational signature analyses also revealed that clock-like signatures were predominant, suggesting that aging itself could give rise to novel PAMs. We subsequently tested the specificity and selectivity of our patient-specific sgRNAs derived from two PC cases and showed that on-target mutations were only detected in the target cell lines and not in the corresponding normal or other PC cell lines lacking the targets. This result was further substantiated by comprehensive WGS analyses in a negative control cell line which demonstrated no off-target activity. Then, two PC cell lines were co-cultured and transduced with patient-specific sgRNAs identified from one of the PC cell lines, and showed >75% selective cell killing via flow cytometry. Finally, to examine the prevalence of somatic PAMs in different tumor types, we obtained 591 T-N subtracted variant call files from ICGC and performed tumor purity correction and PAMfinder analyses for each sample. We found that in two separate PC projects (APGI-AU and PACA-CA), the number of novel PAMs were 703 (N=44) and 863 (N=130), respectively, on average; in the lung cancer (LUCA-KR) and esophageal cancer (OCCAMS-GB) projects, an average of 3319 (N=29) and 4739 (N=388) novel PAMs were discovered, respectively. Overall, our results showed that adult solid tumors harbor hundreds, if not thousands, of somatic PAMs, and these PAMs could serve as novel CRISPR-Cas9 targets to selectively kill cancer cells. Citation Format: Selina Shiqing K Teh, Kirsten Bowland, Alexis Bennett, Eitan Halper-Stromberg, Alyza Skaist, Aparna Pallavajjala, Jacqueline Tang, Fidel Cai, Hong Liang, Sarah Wheelan, Robert B Scharpf, Nicholas J Roberts, James R Eshleman. Novel PAMs from somatic mutations produce numerous pancreatic cancer-specific CRISPR-Cas9 targets for selective cell elimination: A novel gene therapy approach for cancer [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr LB_B02.