Assessing the Genomic Landscape of Cervical Cancers: Clinical Opportunities and Therapeutic Targets

Authors

Friedman CF

Ravichandran V

Miller K

Vanderbilt C

Zhou Q

Iasonos A

Vivek M

Mishra P

Leitao MM Jr

Broach V

Sonoda Y

Kyi C

Zamarin D

O’Cearbhaill RE

Konner J

Berger MF

Weigelt B

Momeni Boroujeni A

Park KJ

Aghajanian C

Solit DB

Donoghue MTA

Doi

10.1158/1078-0432.CCR-23-1078

PMID: 37643132 · DOI: 10.1158/1078-0432.CCR-23-1078 · Journal: Clinical Cancer Research (2023)

TL;DR

Friedman et al. prospectively sequenced 177 cervical cancers across multiple histologic subtypes using MSK-IMPACT and integrated the genomic data with clinical demographics, treatment history, and outcomes. The most prevalent genomic alterations were mutations or amplifications in PIK3CA (25%), ERBB2 (12%), KMT2C (10%), KMT2D (9%), and KRAS (12%). Thirty-seven percent of cases had at least one potentially actionable alteration (OncoKB level 3B), and 17% of patients were enrolled on therapeutic clinical trials, including 10% matched to a trial based on their MSK-IMPACT results.

Cohort & data

  • N = 177 patients with cervical cancer (CESC) sequenced at Memorial Sloan Kettering Cancer Center from February 2014 through May 2019.
  • Histologic subtypes: squamous cell carcinoma (n=69, 39%), usual endocervical adenocarcinoma (n=50, 28%), gastric-type endocervical adenocarcinoma (n=22, 12%), adenosquamous carcinoma (n=21, 12%), and other (n=15, 9%).
  • 60% of samples were from primary tumors and 40% from metastatic sites.
  • 80% of tumors were positive for high-risk HPV.
  • Sequenced using MSK-IMPACT panels (341, 410, or 468 genes) with average 648x coverage.
  • Dataset: cervix_msk_2023. Compared against cesc_tcga_pan_can_atlas_2018.
  • Actionability annotated using OncoKB version 3.1.4.

Key findings

  • The most frequently altered genes were PIK3CA (44/177, 25%), ERBB2 (22/177, 12%), KRAS (21/177, 12%), KMT2C (17/177, 10%), and KMT2D (16/177, 9%).
  • KRAS mutations were significantly enriched in the MSK cohort vs. TCGA (12% vs. 5%, P=0.019).
  • 13% of patients (23/177) had TMB-high (>10 mut/Mb) tumors; 3 of these were also MSI-H.
  • 37% of cases had at least one potentially actionable alteration at OncoKB level 3B, most commonly PIK3CA (25%) and ERBB2 (9.6%).
  • ERBB2 and KRAS alterations were mutually exclusive in squamous and adenocarcinoma subtypes.
  • Oncogenic TP53 mutations and HPV positivity were mutually exclusive: only 3.5% (5/142) of HPV-positive cases harbored oncogenic TP53 alterations, compared with 53.8% (7/13) of non-gastric-type HPV-negative cases (P=0.00007).
  • APOBEC mutational signature was detected in 46% of SCCs, 30% of UEAs, 9% of GEAs, and 33% of adenosquamous carcinomas.
  • 30 patients (17%) were enrolled on therapeutic clinical trials; 18 (10%) were matched to a trial based on MSK-IMPACT results.
  • 2 patients with TMB-low/MSS tumors achieved durable complete responses (>5 years) to dual checkpoint immunotherapy.
  • Median OS for the MSK survival cohort (n=97) was 59.2 months (95% CI: 46.7–89.4), with significant differences by stage (P=0.027) but not by histologic subtype (P=0.099).

Genes & alterations

  • PIK3CA — Mutations/amplifications in 25% of cases; enriched in squamous (38%) vs. gastric type (5%). Most common OncoKB level 3B alteration.
  • ERBB2 — Alterations in 12% (4% amplifications, 9% oncogenic mutations); enriched in UEA (28%). One patient with D769N mutation had exceptional response to HER kinase inhibitor (>5 years).
  • KRAS — Mutations in 12%; enriched in gastric type (27%) and UEA (20%). Included 1 G12C hotspot mutation. Significantly enriched vs. TCGA (P=0.019).
  • TP53 — Mutations in 11% overall; predominant in gastric type (55%). Mutually exclusive with HPV positivity.
  • KMT2C — Altered in 10%; enriched in squamous (14%) vs. gastric type (0%).
  • KMT2D — Altered in 9%; enriched in squamous (14%) vs. gastric type (0%).
  • PTEN — Altered in 7%; enriched in adenosquamous (14%).
  • STK11 — Altered in 14% of gastric type and 33% of adenosquamous carcinoma.
  • ARID1A — Mutations absent in adenosquamous carcinoma; present in mesonephric carcinoma (2/3 cases).
  • FBXW7 — Altered in 13% of squamous cases.
  • TERT — Promoter mutations in 12% of squamous cases.
  • MYC — Amplifications seen only in UEA (4%) and in small cell carcinoma (2/5 cases with concurrent TP53 mutations).
  • BRCA1/BRCA2 — Pathogenic somatic alterations in 3 patients (1 BRCA1, 2 BRCA2).
  • SMAD4, CDKN2A — Genomic drivers in gastric-type adenocarcinoma resembling pancreatobiliary tumors.

Clinical implications

  • Prospective genomic profiling identified actionable alterations in 37% of cervical cancer patients (OncoKB level 3B), supporting routine clinical sequencing.
  • ERBB2 is a likely important therapeutic target, especially in UEA (28% altered). Neratinib showed 25% ORR and 7.0-month median PFS in HER2-mutant cervical cancer; trastuzumab deruxtecan is under investigation (NCT04482309).
  • TMB-H and MSI-H status (13% combined) are FDA-recognized biomarkers for pembrolizumab, though ORR with single-agent PD-1 blockade remains modest (~10–15%).
  • Two exceptional responders to dual checkpoint blockade had TMB-low/MSS but PD-L1-positive, HPV-positive tumors, questioning the utility of TMB as a sole immunotherapy biomarker in cervical cancer.
  • KRAS G12C inhibitors (adagrasib, sotorasib) and pan-KRAS G12X inhibitors may be relevant for the ~12% of patients with KRAS mutations.
  • Gastric-type adenocarcinoma harbored genomic drivers (KRAS, TP53, SMAD4, CDKN2A) resembling pancreatobiliary tumors, suggesting potential for cross-indication treatment strategies.

Limitations & open questions

  • Relatively small overall sample size (N=177) limits power for subtype-specific and stage-specific genomic comparisons.
  • Retrospective clinical data analysis at a single tertiary referral center with overrepresentation of rare histologies and advanced disease.
  • Clinical trial enrollment was influenced by institutional access to a diverse trial portfolio, limiting generalizability.
  • The utility of TMB as a predictive biomarker for immunotherapy response in cervical cancer requires further investigation given exceptional responses in TMB-low/MSS patients.
  • HPV detection by off-target MSK-IMPACT reads had 68% recall; the subsequently added on-target HPV probes may improve sensitivity.
  • No NTRK or RET fusions or BRAF V600E mutations were identified, limiting assessment of tumor-agnostic biomarker utility in this cohort.

Citations from this paper used in the wiki

  • “The most prevalent genomic alterations were somatic mutations or amplifications in PIK3CA (25%), ERBB2 (12%), KMT2C (10%), and KMT2D (9%).” (Abstract)
  • “Furthermore, 13% of patients had high tumor mutational burden (TMB>10 mut/Mb), 3 of which were also microsatellite instability-high (MSI-H).” (Abstract)
  • “Thirty-seven percent of cases had at least one potentially actionable alteration designated as a level 3B mutational event according to the FDA-recognized OncoKB tumor mutation database.” (Abstract)
  • “A total of 30 patients (17%) were enrolled on a therapeutic clinical trial, including 18 (10%) who were matched with a study based on their MSK-IMPACT results.” (Abstract)
  • “KRAS: 12% (21/177) vs. 5% (15/297), P=0.019” (Results, comparison with TCGA)
  • “only 3.5% (5/142) HPV-positive cases harboring oncogenic TP53 alterations… contrasts with 53.8% (7/13) of non-GEA, HPV-negative cases with concurrent oncogenic mutations in TP53 (P=0.00007).” (Results)
  • “The median OS for this cohort (n=97) was 59.2 months (95% CI: 46.7–89.4)” (Results)

This page was processed by crosslinker on 2026-05-04.