Purpose/Objective(s): To determine the discordance rate in epidermal growth factor receptor (EGFR) mutation status between primary and paired distant metastatic tumors in non-small-cell lung cancer (NSCLC).

Materials/Methods: We searched MEDLINE database for eligible studies and performed a meta-analysis. We calculated the percentages of discordance in EGFR mutation status and 95% confidence intervals (CIs) for each study. We performed subgroup analyses for EGFR mutation status of primary tumor (mutant or wildtype), site of distant metastasis (brain, bone, liver or lung/ pleural), type of testing (direct sequencing or allele-specific testing) and timing of metastasis relative to the diagnosis of primary tumor (synchronous or metachronous). We used the random effects model to calculate the pooled percentages. Heterogeneity across studies was assessed using I squared (I²) test. Comparison of subgroups was performed using chi-square test.

Results: We identified 19 eligible studies including 422 patients. The overall discordance rate in EGFR mutation status was 13% (95% CI, 6% to 22%; I² = 83%). The EGFR discordance was statistically significantly higher in bone compared with brain (45% versus 16%; chi-square P value (P)= 0.0005) and lung/ pleural (45% versus 13%; P = 0.0032) metastases. Subgroup analyses did not show any significant effect modification on the discordance rates by EGFR mutation status of primary tumor (mutant, 20%; wildtype, 14%; P = 0.11), type of testing (direct sequencing, 14%; allele-specific testing, 21%; P= 0.12) and timing of metastasis relative to the diagnosis of primary tumor (synchronous, 13%; metachronous, 25%; P= 0.58).

Conclusion: The discordance rates in EGFR mutation status between primary and paired distant metastatic tumors in NSCLC varied largely between studies. Discordance occurred more commonly in bone compared with brain and lung/ pleural metastatic sites. Future researches assessing the impact of discordance in mutation on treatment efficacy and survival are needed.