Context: Detection of KRAS mutation is mandatory to predict response to anti-epidermal growth factor receptor monoclonal antibodies in patients with metastatic colorectal cancers.
Objective: To demonstrate challenges posed to pathologists in the clinical detection of KRAS mutations in colorectal cancers.
Design: In this retrospective analysis for quality assessment of the pyrosequencing assay, we survey the characteristics of 463 formalin-fixed, paraffin-embedded neoplastic tissues submitted for KRAS mutation detection during a 26-month period.
Results: The KRAS mutation was detected in 39.2% of tumors. This included 2 tumors with complex pyrograms (GGT>GAG at codon 12 and GGC>GTT at codon 13, as resolved by a Pyromaker software program) and 3 tumors with an indeterminate percentage of mutant alleles (defined as 4% to 5% and confirmed by a next-generation sequencing platform). Among the 25 specimens (5.5%) with fewer than 20% tumor cells, 22 were resected after chemotherapy/radiation. Significant depletion of tumor cells was observed in rectal cancers resected after neoadjuvant therapy (31.0%) versus those without previous treatment (0%) (P = .01). We also explore other specimens with low tumor cellularity and potential causes of discrepancy between the estimated tumor cell percentage and detected mutant allele frequency, such as intratumor heterogeneity of KRAS mutation.
Conclusions: Neoadjuvant therapy may deplete tumor cells and confound the molecular diagnosis of KRAS mutations. Accurate detection of specimens with poor tumor cellularity requires the appropriate selection of neoplastic tissues, evaluation of tumor cellularity, use of assays with high sensitivity, and prospective quality assessment.