ESMO Congress 2026

TP53 is one of the most central recurrently mutated genes in solid tumor genomic landscapes, a position long established by multi-platform characterization across cancer types: large-scale profiling of breast cancer showed that TP53, PIK3CA and GATA3 are the only three genes mutated at >10% incidence across all breast cancers ^[1]; TP53 mutations reach 96% in high-grade serous ovarian cancer ^[3]; a pan-cancer immunogenomic analysis further found that TP53 mutations correlated with higher leukocyte levels ^[6]; and an epigenomic study of roughly 6,000 cancer samples showed that TP53 mutations co-occur with higher numbers of somatic mutations in tumors with low age acceleration ^[5]. These provide the coordinates for reading the three TP53-linked datasets at this year's meeting: it is both a high-frequency event tied to aggressive biology and one intertwined with the immune microenvironment and mutation burden.

Three TP53-related reads came out of this meeting. First, the NIAGARA analysis of perioperative durvalumab plus chemotherapy in molecular subtypes of MIBC showed the EFS benefit extending across molecularly defined groups: TMB-high HR 0.44, TP53-mutant HR 0.51, FGFR3-mutant HR 0.24, with basal-squamous (B-s) and luminal-Inf tumors showing immune signatures and greater pCR. Second, the biomarker analysis of the evERA trial (giredestrant + everolimus) showed that the PFS benefit was maintained regardless of ctDNA fraction, across ESR1 variants including Y537, and in poor-prognosis subgroups with TP53, GATA3 or RB1 alterations; another live post added that patients with alterations around MAPK, RB, TP53, GATA3 and CDH seemed to derive slightly reduced benefit, though overall giredestrant was still favored. Third, the ctDNA analysis of CROWN presented by Ben Solomon showed that TP53 was consistently a marker of poorer outcomes, that early progressors had a higher number of mutations including TP53, and that inflammatory response was the predominant EP-associated pathway.

Taken together, TP53 played two roles at this meeting. In NIAGARA and evERA, TP53 mutation did not erase treatment benefit — the HR of 0.51 and the preserved benefit across poor-prognosis subgroups indicate that more aggressive biology, consistent with the high-frequency, high-mutation-burden background depicted in ^[1][5], does not mean patients cannot benefit from a given therapy; in CROWN, TP53 once again confirmed its prognostic value. Set against the pan-cancer association between TP53 mutation and higher immune infiltration in ^[6], the same marker clearly needs to be discussed along two separate dimensions — prognostic and predictive — depending on the treatment context. As the evERA discussant noted, current biomarkers mostly tell us whose biology looks more threatening, and whether analyses can move beyond prognostic stratification to a truly predictive signature for endocrine-based combinations after CDK4/6 inhibition remains an open question.

It should be noted that all of the above are exploratory subgroup and biomarker analyses disclosed in live conference posts, with no confidence intervals, p-values, subgroup sample sizes or statistical correction methods provided; the HRs and other figures can only be quoted verbatim from the posts. Neither the consistency of subgroup benefit nor the ctDNA-level associations is sufficient to establish the predictive value of any marker, and none of this constitutes medical or investment advice.