Linking Genomic and Clinical Data Enhances Understanding of Cancer Patient Outcomes

Researchers have linked genomic data and clinical data to identify genes that affect prognosis in cancer patients. The researchers reported these findings in Nature Medicine.

For this study, the researchers used data from the cancer program of the 100,000 Genomes Project. They performed whole-genome sequencing (WGS) on 13,880 tumor samples encompassing 33 cancer types, as well as 13,616 matched normal samples.

The most common tumor types included were invasive breast carcinoma (n=2925), colon adenocarcinoma (n=1948), sarcoma (n=1617), and clear cell renal cell carcinoma (n=1163).

Clinically relevant genetic variants, as indicated by their presence in the National Genomic Test Directory for Cancer (NGTDC), were seen in more than 50% of tumor samples for glioblastoma, low-grade glioma, cutaneous melanoma, head and neck squamous cell carcinoma, colon and rectal adenocarcinoma, and lung adenocarcinoma.

Clinically relevant variants were also found in 20% to 49% of invasive breast carcinomas, high-grade serous ovarian carcinomas, mesotheliomas, sarcomas, endometrial cancers, urothelial carcinomas, and squamous cell lung carcinomas. Less than 20% of pancreatic, prostate, esophageal, and stomach adenocarcinomas had mutations in genes listed in the NGTDC.

Somatic and Germline Variants

TP53 was the most frequently mutated gene, with mutations found in 5411 samples (39.0%). TP53 was mutated in more than 70% of cases of uterine corpus endometrial serous carcinoma, high-grade serous ovarian carcinoma, lung squamous cell carcinoma, rectal adenocarcinoma, esophageal adenocarcinoma, and esophageal squamous cell carcinoma.

The second most frequently mutated gene was PIK3CA, with mutations found in 2750 patients (19.8%). Mutations in this gene occurred most often in uterine corpus endometrial carcinoma (53.5%), ovarian endometrioid adenocarcinoma (49.0%), invasive breast carcinoma (42.2%), uterine corpus endometrial serous carcinoma (38.1%), and colon adenocarcinoma (26.5%).

The highest prevalence of actionable germline variants was seen in high-grade serous ovarian carcinoma, in which 13% of patients had variants in BRCA1 and BRCA2. Patients with these variants tended to have a younger age at diagnosis.

Similarly, patients with germline variants in mismatch-repair genes had earlier age of onset for colon cancer. Earlier age of onset was also seen in patients with high-grade serous ovarian carcinoma or invasive breast carcinoma who had germline mutations in homologous recombination repair genes, and in clear cell renal cell carcinoma patients with germline variants in the VHL gene.

Biomarkers and Mutational Signatures

The researchers noted significant variations in tumor mutational burden (TMB) across cancer types, with cutaneous melanoma and lung adenocarcinoma having the highest average TMB.

In addition, certain mutational signatures were associated with certain cancer types. APOBEC signatures 2 and 13 were associated with invasive breast carcinoma, head and neck squamous cell carcinoma, urothelial carcinoma, and lung adenocarcinoma. Smoking signatures 4 and 92 were associated with lung cancers, and ultraviolet signatures 7a-d were associated with cutaneous melanoma.

The researchers also found the highest prevalence of homologous recombination deficiency (HRD) in high-grade serous ovarian carcinoma (40%), but HRD was seen in other tumors as well.

Clinical Outcomes

By linking the WGS data with longitudinal clinical data, the researchers evaluated treatment outcomes, stratified according to pangenomic markers.

In patients who received platinum therapy (n=189), for example, the presence of HRD was associated with a lower risk of death (hazard ratio [HR], 0.37; 95% CI, 0.23-0.61; P <.001), particularly in patients with invasive breast carcinomas or high-grade serous ovarian carcinomas.

In patients with cutaneous melanoma, those with low TMB had significantly shorter overall survival (OS) than those with high TMB (HR, 2.34; 95% CI, 1.14-4.80; P =.015). A similar difference was not seen in lung cancer (P =.72).

“This may indicate that the level of TMB is relevant in prognosis and supports the need for further refining of pangenomic biomarkers as both prognostic and predictive for immunotherapy response, as highlighted in previous studies,” the researchers noted.

The researchers also evaluated OS in all 33 cancer types, stratified according to the presence or absence of mutations in 40 NGTDC-indicated genes. This analysis revealed 15 genes that affect OS.

The gene that most affected OS was CDKN2A (HR, 2.3; 95% CI, 2.0-2.6; P <1×10⁻¹⁰), “which corresponds to its association with high-grade disease and poor prognosis in some cancer subtypes, such as glioma and soft-tissue sarcoma,” according to the researchers.

Other results were as expected, with the presence of KRAS mutations in colorectal cancer and KRAS and TP53 mutations in non-small cell lung cancer indicating poor prognosis, and PIK3CA mutations associated with favorable outcomes.

The researchers predicted that additional prognostic implications of WGS will be established as this dataset grows and is integrated with data from patients’ life course as well as data from other sequencing modalities, such as cell-free DNA and single-cell sequencing.

“Our findings demonstrate the utility of linking genomic and real-world clinical data to enable survival analysis to identify cancer genes that affect prognosis and advance our understanding of how cancer genomics impacts patient outcomes,” the researchers concluded.

Disclosures: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of disclosures.

This article originally appeared on Cancer Therapy Advisor