Polymorphisms correlated with the clinical outcome of locally advanced or metastatic colorectal cancer patients treated with ALIRI vs. FOLFIRI

Leucovorin-modulated 5-fluorouracil (5-FU) plus irinotecan (FOLFIRI) is the most common treatment of metastatic colorectal cancer (CRC). 5-FU inhibits thymidylate synthase (TS) and irinotecan topoisomerase I, leading to inhibition of DNA replication and repair. FOLFIRI efficacy suggested that other TS inhibitors might synergize with irinotecan, and Phase I/II studies for second-line treatment showed promising results of combinations with the multitargeted antifolate pemetrexed (PMX), which exerts its effects primarily via TS inhibition. However, a randomized Phase II trial of PMX + irinotecan (ALIRI) showed similar efficacy and safety, but significantly shorter progression-free survival (PFS) compared with FOLFIRI in locally advanced or metastatic CRC. In our previous aCGH study, we evaluated genomewide copy number variations, whereas in the current study we evaluated relationships between functional polymorphisms and PFS. Candidate polymorphisms were studied by restriction fragment length polymorphismpolymerase chain reaction (RFLP-PCR) (TSER-2R/3R) or Taqman-PCR ( MTHFR-1958G > A, MTR-2756A > G, MTHFR-1298A > C, SHMT1-1420C > T, ATIC-347C > G, AMPD-134C > T, MTRR-66A > G and SLC-19A180G > A ) in 84 patients (40 FOLFIRI, 44 ALIRI). The Kaplan-Meier method was used to plot PFS, and the log-rank test to compare curves. At univariate analysis the homozygous variants of both MTR-2756A > G and SHMT1-1420C > T were associated with significantly shorter PFS. Conversely, a significantly longer PFS (7.3 months) was observed when ATIC-347C > G CC + CG genotypes were grouped vs. GG . At multivariate analysis the genotypes MTR-2756A > G AA + AG, SHMT1-1420C > T TT + CT and ATIC-347C > G CC + CG emerged as significant predictors for PFS. Because MTR, SHMT1 and ATIC are all involved in folate pathways, we further explored the effect of a combination of their risk genotypes on PFS, showing that patients carrying two risk genotypes had a significantly shorter PFS (3.9 months, p < 0.001). The correlations of polymorphisms in genes with clinical outcome underscore the importance of a candidate gene-based approach. Ultimately, the validation of the role of these polymorphisms in prospective multicenter trials might optimize currently available treatments in selected CRC patients (e.g., FOLFIRI) or PMX-based treatments in other tumor types.

• View at Publisher
• View at Scopus
• View at Web of Science