Potential impact of non-exonuclease <i>POLE</i> and <i>POLD1</i> mutations on high and ultrahigh TMB across cancers.

Abstract

e14558 Background: DNA polymerases ε ( POLE ) and δ1 ( POLD1 ) encode enzymes involved in DNA replication and proofreading. Somatic POLE / D1 mutations occur in 1-3% of cancers. Loss-of-function alterations typically reside within the POLE / D1 exonuclease domain and can result in high tumor mutation burden (TMB). Ultrahigh TMB (≥100 mut/Mb) is often associated with POLE/D1 mutations, and results in near-universal responses to immune checkpoint inhibitors (ICI). Deciphering the contribution of POLE / D1 mutations to TMB can be challenging, as these often co-occur with microsatellite instability (MSI-H) which also drives TMB. Methods: Samples were analyzed by NGS (NextSeq or NovaSeq) (Caris Life Sciences, Phx, AZ). TMB (totaled somatic mutations found per megabase pair) was evaluated in patients with POLE / POLD1 / D2 / D3 / D4 -altered cancers. MSI-H or microsatellite stable (MSS) status was determined by IHC and NGS. Cohorts were divided into pathogenic and likely pathogenic variants (i.e. exonuclease domain mutations) versus VUS alterations (i.e. variants outside the exonuclease domain). Differences in TMB between cohorts were assessed using Fisher’s exact test. Results: Pan-cancer, TMB ≥20 / ≥50 was identified in 10.9% / 3.4% of POLE VUS MSS cancers compared to 3.1% / 0.4% of MSS cancers with POLE / D1 WT status ( p &lt; 0.0001). 0.94% of POLE VUS cancers had ultrahigh TMB versus 0.05% of MSS POLE / D1 WT cancers ( p &lt; 0.0001). Similar rates were observed in MSS cancers with POLD1 VUSs (Table). MSS cancers with pathogenic POLE variants exhibited variable rates of ultrahigh TMB depending on cancer type, ranging from 52.6% (ovarian) to 62.9% (endometrial) to 80.0% (colorectal). Notably, ultrahigh TMB occurred in 0.53%, 0.44%, and 0.44% of MSS patients with POLD2 , POLD3 , and POLD4 mutations ( p &lt; 0.0001), after excluding coexisting POLE / D1 mutations. Conclusions: We propose that certain non-exonuclease POLE / D1 mutations can result in loss-of-function and ultrahigh TMB in MSS patients. Further, tumor type appears to influence the impact of established POLE / D1 driver mutations on TMB. Finally, POLD2 - 4 mutations (not previously studied) may also contribute to genome instability and mutational load. These findings argue for deeper characterization of DNA polymerase mutations to identify patients that may benefit from ICI. POLE / D1 WT + MSS[n (%)] POLE VUS Mutation + MSS[n (%)] POLE Pathogenic Mutation + MSS[n (%)] POLD1 VUS Mutation + MSS[n (%)] POLD1 Pathogenic Mutation + MSS[n (%)] Total 139,842 5,658 557 1,793 11 TMB ≥20 4,378(3.1%)[ref] 614(10.9%) p &lt;0.0001 529(95.0%) p &lt;0.0001 168(9.4%) p &lt;0.0001 5(45.5%) p &lt;0.0001 TMB ≥50 586(0.42%)[ref] 193(3.4%) p &lt;0.0001 467(83.8%) p &lt;0.0001 79(4.4%) p &lt;0.0001 2(18.2%) p =0.0009 TMB ≥100 73(0.05%)[ref] 53(0.94%) p &lt;0.0001 343(61.6%) p &lt;0.0001 28(1.6%) p &lt;0.0001 1(9.1%) p =0.0058 Evaluated by Fisher’s exact test comparing to respective POLE/D1 WT + MSS cohort.