Molecular and Genetic Pathology: Variants, Tests, and Interpretation
Molecular pathology questions test whether you understand what a test measures and how to interpret a result in context. The vocabulary is precise: variant, allele frequency, sensitivity, specificity, and limit of detection.
Concepts to master
- Variant classes: germline vs. somatic; pathogenic, likely pathogenic, variant of uncertain significance (VUS), likely benign, benign. A VUS is not a diagnosis and should not drive management.
- Test types: PCR/Sanger (single site, high accuracy), next-generation sequencing (panels, exome, genome; detects low-frequency variants), FISH (translocations, amplifications), and immunohistochemistry as a surrogate (e.g., mismatch repair proteins, HER2).
- Key paradigms: loss-of-function tumor suppressors (TP53, BRCA1/2, APC), gain-of-function oncogenes (KRAS, EGFR, BRAF V600E), and fusion-driven tumors (BCR-ABL1, EML4-ALK).
- Actionability: a variant is actionable when a targeted therapy exists; resistance mutations (e.g., EGFR T790M) explain acquired resistance.
- Quality concepts: limit of detection, tumor cellularity, and the effect of formalin fixation on DNA quality.
Traps
The most frequent error is treating a detected variant as necessarily pathogenic - frequency in population databases and predicted effect matter. Another is confusing a germline finding (present at ~50% allele fraction in a tumor, or confirmed in blood) with a somatic one. Also, a negative test does not exclude a syndrome if the assay's limit of detection or coverage is inadequate.
Rule of thumb
For any molecular result, ask three questions: (1) Is the variant real and at adequate allele frequency? (2) Is it germline or somatic? (3) Is it actionable - does a therapy or surveillance change? Only then interpret clinically.
Worked example
A lung adenocarcinoma shows an EGFR exon 19 deletion and the patient responds to a tyrosine kinase inhibitor, then progresses. Repeat testing shows EGFR T790M at low allele frequency. This is an acquired resistance mutation, not a new primary, and it is actionable (third-generation inhibitor). The trap is dismissing the low-frequency result as artifact - the clinical context makes it meaningful.