Genomic Eligibility Matching

The core engine that maps a patient's specific genomic variants (e.g., EGFR exon 19 deletion, BRAF V600E) directly to a trial's molecular inclusion criteria. This process involves parsing structured variant call format (VCF) files and comparing them against a machine-readable library of required biomarkers. The system must handle complex logic, such as requiring a mutation but excluding a specific co-mutation, ensuring that only patients with the precise molecular profile are matched.

Resolves the semantic gap between how a biomarker is described in a protocol versus how it is reported in a genomic lab result. For example, a trial may require 'HER2 overexpression,' while a pathology report states 'ERBB2 amplification.' This feature uses medical ontology alignment to map synonymous terms, gene aliases, and protein names to a unified concept, preventing false negatives caused by simple string mismatches.

Evaluates intricate logical expressions that combine multiple genomic requirements. A trial might specify: '(EGFR mutation OR ALK rearrangement) AND NOT KRAS G12C'. The system parses this logic tree and evaluates it against the patient's full genomic profile. This goes beyond simple presence/absence checks to handle nested AND/OR/NOT conditions with high precision, automating what is typically a manual, error-prone curation step.

Distinguishes between clinically actionable pathogenic variants and benign polymorphisms. A trial requiring a 'deleterious BRCA1 mutation' should not match a patient with a benign BRCA1 variant of unknown significance. This feature integrates with knowledge bases like ClinVar to automatically filter variants by their clinical significance, ensuring that only patients with truly relevant mutations are flagged for screening.

Extends matching beyond single nucleotide variants (SNVs) to include copy number variations (CNVs) and gene fusions. A trial for HER2-targeted therapy requires identifying patients with ERBB2 amplification, not just a sequence mutation. The system interprets quantitative metrics like ploidy and log2 ratios from sequencing data to determine if a patient meets the amplification threshold defined in the protocol's eligibility criteria.

Automates the assessment of a patient's Tumor Mutational Burden (TMB) score against a trial's numeric cutoff (e.g., TMB-High ≥ 10 mut/Mb). The system extracts the calculated TMB value from the structured genomic report and performs a direct numeric comparison. This feature is critical for immunotherapy trials where TMB is a pan-tumor eligibility biomarker, enabling rapid identification of patients who qualify based on this quantitative metric.

A trial recruitment method that prioritizes patient identification based on the presence or level of a specific biological marker rather than disease type alone.

• Key Markers: EGFR T790M, KRAS G12C, MSI-H/dMMR, PD-L1 TPS
• Mechanism: Replaces histology-based screening with molecular profiling
• Example: A basket trial enrolling all solid tumors with an NTRK fusion, regardless of organ origin
• Data Source: Structured pathology reports and next-generation sequencing (NGS) output files

The process of enriching raw genomic coordinates with functional context and mapping variant descriptions to a standard nomenclature to enable consistent computational comparison.

• Annotation: Adding gene name, protein change, and population frequency to a variant call
• Normalization: Converting 'EGFR L858R' and 'c.2573T>G' into a single HGVS-compliant identifier
• Tools: Ensembl VEP, SnpEff, and ClinVar for clinical significance
• Challenge: Resolving legacy variant names and non-standard lab reporting formats

A machine-processable definition of a clinical condition expressed as logical expressions and data queries, used to identify patient cohorts from electronic health records.

• Structure: Combines diagnosis codes, lab values, medication orders, and genomic findings
• Example: [EGFR Exon 19 Deletion == TRUE] AND [NSCLC Diagnosis == TRUE] AND [Prior TKI Therapy == FALSE]
• Execution: Translated into SQL or FHIR API calls against a clinical data warehouse
• Role in Genomics: Bridges structured variant data with clinical narrative for holistic eligibility assessment

The standardized HL7 FHIR resources and operations designed specifically for the exchange of structured molecular and genomic data between healthcare systems.

• Key Resources: MolecularSequence, Observation-genetics, and DiagnosticReport-genomics
• Operations: $find-subject-variants and $find-population-variants for cohort discovery
• Purpose: Enables EHRs to transmit structured variant data to trial matching services without custom integrations
• Adoption: Increasingly required by precision medicine initiatives and clinical research networks

The process of converting parsed, structured eligibility criteria into executable database queries to screen patient repositories for matching genomic profiles.

• Input: A structured representation of 'EGFR T790M mutation positive'
• Output: A SQL or GraphQL query against a variant warehouse filtering for the specific cDNA change
• Complexity: Must handle zygosity, allele frequency thresholds, and co-occurring mutation requirements
• Optimization: Query planning to efficiently join variant tables with demographic and diagnosis data at scale

The systematic review of reasons why pre-screened patients failed to meet trial eligibility, used to optimize recruitment strategies and refine protocol inclusion criteria.

• Genomic Failures: Patient harbors the target mutation but fails due to a co-mutation exclusion (e.g., KRAS mutation excluding EGFR TKI trial)
• Data Quality Failures: Incomplete or ambiguous variant reporting preventing definitive matching
• Feedback Loop: Aggregated failure data informs protocol amendments and site selection
• Metric: Screen failure rate by biomarker category to identify overly restrictive molecular criteria