Colorectal cancer is known to have abundant aberrant methylations, but its clinical implication for treatment of metastatic disease has been rarely studied. This study aimed to assess the association of clinical benefit from irinotecan-containing chemotherapy for metastatic colorectal cancer (mCRC) with CpG island methylation phenotype (CIMP) and promotor methylation of CHFR, SRBC, OCTN2, WRN, and SULF2.
We retrospectively reviewed for clinical information concerning time to progression (TTP) through medical records of 100 patients (pts) who underwent palliative or curative surgery and were treated for mCRC with irinotecan-containing chemotherapy, which had to be given with palliative intent for measurable or non-measurable lesion. Their surgical specimens including tumor and peritumoral normal tissue were tested for aberrant methylation of 5 CIMP-specific markers (CACNA1G, IGF2, NEUROG1, RUNX3, and SOCS1), CHFR, SRBC, WRN, OCTN2 and SULF2 with methylation-specific PCR method.
The pts' median age was 55 (range 30-77) years and 33% were female. 63% were synchronous metastatic disease. 67%/14%/19% were chemo-naÃ¯ve, fluoropyrimidine-treated, fluoropyrimidine and oxaliplatin-treated, respectively. 24 were given irinotecan monotherapy while 76 were treated with fluoropyrimidine and irinotecan. Among the tested 10 markers, methylation of CHFR was seen in 16% and significantly associated with better TTP with irinotecan-based chemotherapy (median TTP: 133 days vs 263 days in unmethylated and methylated, respectively, p = 0.012). WRN was hypermethylated in 24% and also marginally associated with better TTP (p = 0.05). 10 pts with methylation of CHFR and WRN had longest TTP while pts with one of them methylated (n = 20) and those with both unmethylated (n = 70) showed similar TTP (330 days, 167 days, 130 days, respectively. P = 0.038).
This pilot study suggests combination of CHFR and WRN methylation could be helpful for selection of pts with mCRC who are most likely to benefit from irinotecan-based chemotherapy.
Clinical trial identification
All authors have declared no conflicts of interest.