Nearly Every Tumor Carries a Broken TP53: The Genomic Landscape of 585 Ovarian Cancers, Mapped

A multi-omic atlas of high-grade serous ovarian carcinoma reveals near-universal TP53 disruption, distinct molecular subtypes with divergent survival, and a handful of rare microsatellite-unstable tumors that may respond to immunotherapy.

Ovarian cancer kills more American women than any other gynecologic malignancy — roughly 13,000 a year — and the deadliest form, high-grade serous ovarian carcinoma, is also the most common. It tends to announce itself late, after the tumor has already seeded the peritoneum, and it tends to come back after chemotherapy, sometimes within months. To understand why, and to find molecular handles for better treatment, The Cancer Genome Atlas profiled 585 serous ovarian cystadenocarcinomas with every tool in the modern genomics arsenal: whole-exome sequencing, copy-number arrays, mRNA expression, DNA methylation, and reverse-phase protein arrays. The resulting dataset, updated as part of the PanCancer Atlas, is one of the most deeply characterized cancer cohorts in existence. This dashboard walks through what those 585 tumors reveal — from the chromosomal chaos that defines the disease to the rare molecular exceptions that might change how a small subset of patients is treated.

Who Is in the Cohort

The 585 patients are overwhelmingly what epidemiology would predict for high-grade serous disease: women diagnosed in their fifties and sixties, with the distribution peaking around age 60 and a modest left tail of younger patients — some below 40 — who are likely to carry inherited BRCA1 or BRCA2 mutations. Almost all tumors are grade 3, the highest histologic grade, consistent with the aggressive biology of high-grade serous carcinoma. The cohort is predominantly White and of European genetic ancestry, a well-documented limitation of TCGA that restricts the power of ancestry-stratified analyses. Self-reported race and computationally inferred ancestry labels are available for cross-referencing, but the small numbers in non-European groups mean that any race-specific findings should be treated as exploratory.

Follow-Up Maturity and Pre-Analytic Checks

Before any survival curve can be trusted, the follow-up data has to be long enough to capture late events. The TCGA-OV follow-up distribution is right-skewed: many patients were tracked for fewer than 1,000 days, but a meaningful tail extends beyond 3,000 days, anchoring the late end of survival estimates. The cohort draws specimens from multiple tissue source sites across the country — a strength for generalizability but a known source of batch effects in methylation and protein data. Most samples were primary tumors collected at debulking surgery, with very few recurrent specimens. Neoadjuvant chemotherapy — treatment given before surgery — was uncommon, meaning the vast majority of molecular profiles reflect treatment-naïve disease, which simplifies interpretation of the somatic landscape.

Chromosomal Chaos Is the Rule

High-grade serous ovarian carcinoma is one of the most chromosomally unstable cancers in the human body, and the aneuploidy scores in this cohort confirm it. Aneuploidy — the gain or loss of entire chromosome arms — is measured by counting how many arms deviate from the normal diploid state. Scores in this cohort are shifted far to the right, with medians around 15 to 20 on a scale where zero means no arm-level changes. Very few tumors have low aneuploidy scores; those that do deserve a second look, because they may represent misclassified low-grade tumors or samples contaminated with normal tissue.

Microsatellite instability, a separate axis of genomic instability caused by defective DNA mismatch repair, is the mirror image: almost nonexistent. MSIsensor scores cluster near zero, confirming that these tumors are microsatellite stable. The rare exceptions — any specimen scoring above 3.5 — are candidates for immune checkpoint therapy, an option typically associated with colorectal or endometrial cancers but increasingly explored in gynecologic oncology.

High-grade serous ovarian carcinoma is one of the most chromosomally unstable cancers in the human body, and the aneuploidy scores in this cohort confirm it.

The Driver Landscape: TP53 Everywhere, BRCA in a Critical Minority

The OncoPrint — a gene-by-patient matrix that integrates mutations and copy-number changes into a single visual — tells the story of this disease in one glance. TP53, the gene encoding the p53 tumor suppressor, is altered in virtually every patient. That near-universality is the defining molecular signature of high-grade serous ovarian carcinoma and distinguishes it from low-grade serous, endometrioid, and clear-cell subtypes, which have different drivers entirely.

Beyond TP53, the most clinically consequential alterations are in BRCA1 and BRCA2. Patients whose tumors carry BRCA mutations — whether somatic or inherited — have defective homologous recombination repair, a vulnerability that can be exploited with PARP inhibitors like olaparib and niraparib. The OncoPrint also reveals patterns of mutual exclusivity: BRCA alterations and CCNE1 amplification, for instance, tend not to co-occur in the same tumor, suggesting distinct oncogenic programs that may require different therapeutic strategies.

The variant classification spectrum is dominated by missense mutations, with a notable fraction of truncating variants — nonsense and frameshift — consistent with the tumor-suppressor-driven biology of this cancer. Point mutations (SNPs) vastly outnumber insertions and deletions, as expected for a microsatellite-stable disease.

Treatment Response and the Swimmer's View

Platinum-based chemotherapy remains the backbone of first-line treatment for high-grade serous ovarian cancer, and the treatment-response breakdown reflects that reality. A large fraction of patients achieved a complete response — their tumors shrank below the threshold of detection — consistent with the known platinum sensitivity of HGSOC. But a substantial minority showed only partial response, stable disease, or outright progression, marking tumors with primary platinum resistance.

The swimmer plot anchors these response categories to time. Each patient's follow-up is rendered as a horizontal lane, color-coded by initial treatment response, with death events marked. Complete responders tend to have longer lanes, but the correlation is imperfect: some patients who initially responded well relapsed and died within two years, while a few with only partial responses survived far longer. The visual makes a point that aggregate statistics obscure — individual trajectories vary enormously, and molecular subtyping alone does not fully explain who lives and who does not.

What 585 tumors make plain is that high-grade serous ovarian carcinoma is, at its core, a disease of chromosomal anarchy governed by a single broken gene — TP53 — with a handful of actionable molecular exceptions scattered across the cohort. The BRCA-mutant subset already has targeted therapy in PARP inhibitors. The vanishingly rare MSI-high cases may have a path to immunotherapy. The mesenchymal subtype's hypoxic biology points toward anti-angiogenic strategies. But for the majority of patients — those with platinum-sensitive disease who respond initially and then relapse — the molecular portrait has yet to deliver a decisive new therapeutic handle. The open question is not what drives this cancer; it is why, knowing what drives it, we still cannot reliably prevent its return. The next experiments will likely come from integrating these genomic layers with single-cell resolution, spatial profiling of the tumor microenvironment, and longitudinal sampling across lines of therapy — following the disease not just at diagnosis, but as it evolves under treatment pressure.