(P 360) TP53 ALTERATIONS IN SARCOMAS: AN ANALYSIS FROM THE AACR-GENIE DATABASE

Objective: TP53 is the most commonly mutated gene in cancer, and it plays a significant role in sarcomas. Various structural alterations of TP53 (somatic and germline) have been observed in sarcomas. Effects on disease-free survival and responsiveness with certain therapeutics have been reported in sarcomas(anthracyclines).[ PMID: 34282771] In addition, the development of p53 reactivators in preclinical studies and early phase clinical trials are demonstrating signs of efficacy. [PMID: 35948006] The aim with this study is to perform a comprehensive analysis on the frequency, mutation type, histologic distribution, and evaluation of other alterations associated with TP53-mutated sarcomas.

Methods: GENIE cohort (version 13.1) was evaluated for de-identified sarcoma files, which include limited clinical information and genomic data. cBioPortal and R software was utilized for data visualization and mutation mapping. GraphPad Prism (version 9) and Tableau Desktop 8.1 were utilized for descriptive statistics. OncoKB (oncokb.org) was utilized to annotate the type and function of gene alterations and to determine actionability of each mutational event. Gene ontology analysis (GOA) was performed on genes with mutations detected at > 2.5% frequency with PANTHER 17.0 (pantherdb.org). Fisher’s exact test was performed for enrichment analysis.

Results: 5,222 sarcoma samples from 4,875 patients were identified. 251 patients had 2 or more samples each (5.1%). Primary tumors represented 3,298 of sample (63%) and metastatic sites accounted for 1,791 of samples (34%). Of these, 2,092 samples harbored TP53 mutations. The most common TP53 mutations in sarcomas included p.R175H (n=80 patients), p.R248Q (n=72), p.R273H (n=63), p.R273C (n=51), p.R248W (n=49), p.Y220C (n=42), and p.R342* (n=40). With exception of p.R342*, the other most common hotspot mutations occur in the p53 DNA binding domain. By OncoKB, p.R175H, p.R273H, p.Y220C, and p.R342* were considered oncogenic, loss-of-function; p.R248Q, p.R273C, and p.R248W were only likely oncogenic. The p.Y220C variant was associated with more TP53 CNAs compared to the other most common hotspot TP53 mutations. Uterine carcinosarcomas presented the most common histology across these 7 hotspot mutations. Missense mutations (n=1430), frame shift deletions (n=264), nonsense mutations (n=249), and splice site variants (n=149) were the most common mutation types. Compared to wild type TP53 (median 3, range 1 – 293), sarcomas with TP53 alterations (median 5, range 1 – 120, p< 0.0001) harbored a higher mutation count. A GOA demonstrated a >100-fold over enrichment for telomerase activities.

Conclusion: This dataset represents the largest sample of TP53 alterations in sarcomas. The most common hotspot mutations of TP53 were oncogenic with loss-of-functionality. TP53 mutated sarcomas had higher mutation counts than sarcomas without TP53 alterations. Regulation of telomerase activities were over-represented by GOA.