Dataset Information

Machine learning-based tissue of origin classification for cancer of unknown primary diagnostics using genome-wide mutation features.

ABSTRACT: Cancers of unknown primary (CUP) origin account for ∼3% of all cancer diagnoses, whereby the tumor tissue of origin (TOO) cannot be determined. Using a uniformly processed dataset encompassing 6756 whole-genome sequenced primary and metastatic tumors, we develop Cancer of Unknown Primary Location Resolver (CUPLR), a random forest TOO classifier that employs 511 features based on simple and complex somatic driver and passenger mutations. CUPLR distinguishes 35 cancer (sub)types with ∼90% recall and ∼90% precision based on cross-validation and test set predictions. We find that structural variant derived features increase the performance and utility for classifying specific cancer types. With CUPLR, we could determine the TOO for 82/141 (58%) of CUP patients. Although CUPLR is based on machine learning, it provides a human interpretable graphical report with detailed feature explanations. The comprehensive output of CUPLR complements existing histopathological procedures and can enable improved diagnostics for CUP patients.

SUBMITTER: Nguyen L

PROVIDER: S-EPMC9273599 | biostudies-literature | 2022 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning-based tissue of origin classification for cancer of unknown primary diagnostics using genome-wide mutation features.

Nguyen Luan L Van Hoeck Arne A Cuppen Edwin E

Nature communications 20220711 1

Cancers of unknown primary (CUP) origin account for ∼3% of all cancer diagnoses, whereby the tumor tissue of origin (TOO) cannot be determined. Using a uniformly processed dataset encompassing 6756 whole-genome sequenced primary and metastatic tumors, we develop Cancer of Unknown Primary Location Resolver (CUPLR), a random forest TOO classifier that employs 511 features based on simple and complex somatic driver and passenger mutations. CUPLR distinguishes 35 cancer (sub)types with ∼90% recall a ...[more]

PMID: 35817764

Similar Datasets

Project description:BackgroundIn ∼3%-5% of patients with metastatic disease, tumor origin remains unknown despite modern imaging techniques and extensive pathology work-up. With long diagnostic delays and limited and ineffective therapy options, the clinical outcome of patients with cancer of unknown primary (CUP) remains poor. Large-scale genome sequencing studies have revealed that tumor types can be predicted based on distinct patterns of somatic variants and other genomic characteristics. Moreover, actionable genomic events are present in almost half of CUP patients. This study investigated the clinical value of whole genome sequencing (WGS) in terms of primary tumor identification and detection of actionable events, in the routine diagnostic work-up of CUP patients.Patients and methodsA WGS-based tumor type 'cancer of unknown primary prediction algorithm' (CUPPA) was developed based on previously described principles and validated on a large pan-cancer WGS database of metastatic cancer patients (>4000 samples) and 254 independent patients, respectively. We assessed the clinical value of this prediction algorithm as part of routine WGS-based diagnostic work-up for 72 CUP patients.ResultsCUPPA correctly predicted the primary tumor type in 78% of samples in the independent validation cohort (194/254 patients). High-confidence predictions (>95% precision) were obtained for 162/254 patients (64%). When integrated in the diagnostic work-up of CUP patients, CUPPA could identify a primary tumor type for 49/72 patients (68%). Most common diagnoses included non-small-cell lung (n = 7), gastroesophageal (n = 4), pancreatic (n = 4), and colorectal cancer (n = 3). Actionable events with matched therapy options in clinical trials were identified in 47% of patients.ConclusionsGenome-based tumor type prediction can predict cancer diagnoses with high accuracy when integrated in the routine diagnostic work-up of patients with metastatic cancer. With identification of the primary tumor type in the majority of patients and detection of actionable events, WGS is a valuable diagnostic tool for patients with CUP.

Project description:BackgroundCarcinoma of unknown primary origin (CUP) accounts for 2%-5% of newly diagnosed advanced malignancies, with chemotherapy as the standard of care. CUPISCO (NCT03498521) is an ongoing randomized trial using comprehensive genomic profiling (CGP) to assign patients with CUP to targeted or immunotherapy treatment arms based on genomic profiling. We performed a retrospective analysis of CUP cases referred for CGP to determine how many were potentially eligible for enrollment into an experimental CUPISCO arm.Materials and methodsCentrally reviewed adenocarcinoma and undifferentiated CUP specimens in the FoundationCore database were analyzed using the hybrid capture-based FoundationOne CDx assay (mean coverage, >600×). Presence of genomic alterations, microsatellite instability (MSI), tumor mutational burden (TMB), genomic loss of heterozygosity (gLOH), and programmed death-ligand 1 (PD-L1) positivity were determined.ResultsA total of 96 of 303 patients (31.7%) could be matched to an experimental CUPISCO arm. Key genomic alterations included ERBB2 (7.3%), PIK3CA (6.3%), NF1 (5.6%), NF2 (4.6%), BRAF (4.3%), IDH1 (3.3%), PTEN, FGFR2, EGFR (3.6% each), MET (4.3%), CDK6 (3.0%), FBXW7, CDK4 (2.3% each), IDH2, RET, ROS1, NTRK (1.0% each), and ALK (0.7%). Median TMB was 3.75 mutations per megabase of DNA; 34 patients (11.6%) had a TMB ≥16 mutations per megabase. Three patients (1%) had high MSI, and 42 (14%) displayed high PD-L1 expression (tumor proportion score ≥50%). gLOH could be assessed in 199 of 303 specimens; 19.6% had a score of >16%.ConclusionsThirty-two percent of patients would have been eligible for targeted therapy in CUPISCO. Future studies, including additional biomarkers such as PD-L1 positivity and gLOH, may identify a greater proportion potentially benefiting from CGP-informed treatment. Clinical trial identification number. NCT03498521 IMPLICATIONS FOR PRACTICE: The findings of this retrospective analysis of carcinoma of unknown primary origin (CUP) cases validate the experimental treatment arms being used in the CUPISCO study (NCT03498521), an ongoing randomized trial using comprehensive genomic profiling to assign patients with CUP to targeted or immunotherapy treatment arms based on the presence of pathogenic genomic alterations. The findings also suggest that future studies including additional biomarkers and treatment arms, such as programmed death-ligand 1 positivity and genomic loss of heterozygosity, may identify a greater proportion of patients with CUP potentially benefiting from comprehensive genomic profiling-informed treatment.

Dataset Information

Machine learning-based tissue of origin classification for cancer of unknown primary diagnostics using genome-wide mutation features.

Publications

Machine learning-based tissue of origin classification for cancer of unknown primary diagnostics using genome-wide mutation features.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets