Let-7 miRNA-binding site polymorphism in the KRAS 3^′UTR; colorectal cancer screening population prevalence and influence on clinical outcome in patients with metastatic colorectal cancer treated with 5-fluorouracil and oxaliplatin +/− cetuximab

The prognosis for patients with metastatic colorectal cancer (mCRC) included in clinical trials has increased from approximately 12 months with 5-fluorouracil monotherapy to 20–24 months with the addition of newer chemotherapeutic agents and targeted drugs [1]. Cetuximab and panitumumab, two monoclonal antibodies (moAb) targeting the epidermal growth factor receptor (EGFR), have proven to be effective in combination with chemotherapy or as single agents [2‐7]. KRAS mutation is a negative predictive marker for response to EGFR-targeted therapy and clinical benefit seems to be restricted to patients with KRAS wild-type tumors [2‐5, 8, 9]. In the recent NORDIC-VII study, however, we did not find an improved outcome of adding cetuximab to first-line oxaliplatin-based chemotherapy in KRAS wild-type patients [10]. Similar results were found by the COIN trial [11]. The results of these trials demonstrate the necessity to explore and validate new biomarkers to improve the selection of patients who are likely to benefit from cetuximab treatment [12]. It was recently reported that copy number aberrations (CNA) may provide additional information to mutation status and their use may potentially further improve the selection of mCRC patients for EGFR-targeted therapy [13]. Mekenkamp et al. demonstrated that KRAS copy number loss was associated with good response in both KRAS wild-type and KRAS mutated mCRC patients treated with a cetuximab-containing first-line regimen [13].KRAS copy number gains were associated with poor progression-free survival (PFS) in KRAS wild-type mCRC patients given the same treatment [13].

MicroRNAs (miRNAs) are a class of highly conserved 22-nucleotides single-stranded RNAs that can act as trans-acting factors that suppress translation or induce messenger RNA (mRNA) degradation of target genes [14]. They are global gene regulators implicated in virtually all cancer types studied, where they can function as oncogenes or tumor suppressors [15]. A number of miRNAs have been reported to be involved in CRC development and KRAS regulation, and these may influence the effect of EGFR-targeted therapy [13, 16, 17]. An important miRNA in CRC seems to be miR-143, which, has been described to be downregulated in CRC [18] and to inhibit the translation of KRAS mRNA, thereby altering RAS signaling and inhibiting tumor cell growth [19]. Pichler et al. found that low miR-143 expression was an independent negative prognostic factor for cancer specific survival in mCRC, and they reported a decreased PFS in KRAS wild-type mCRC patients treated with EGFR-targeted agents [16].

The let-7 family of miRNAs plays an important role in many malignant tumors where they mainly function as tumor suppressors. Downregulation of let-7 family members is observed in multiple carcinomas [20], including colon cancer [21]. RAS expression was decreased in colon cancer cell lines after transfection of let-7a-1 miRNA precursor suggesting that let-7 is involved in regulating colon cancer cell growth [22]. Let-7 miRNAs downregulate RAS after binding to specific sites in the 3^′ untranslated region (3^′-UTR) of the KRAS mRNA [23]. A functional single nucleotide polymorphism has been characterized in the let-7 complementary site (LCS6) in the KRAS 3^′-UTR mRNA leading to increased expression of KRAS in vitro and lower let-7 levels in vivo[24].

The LCS6 variant allele is associated with increased risk of non-small cell lung cancer (NSCLC) in moderate smokers [24], triple-negative breast cancer in premenopausal women [25], and ovarian cancer in BRCA negative women from hereditary breast and ovarian cancer syndrome families [26]. Moreover, the LCS6 variant allele is enriched in BRCA negative double primary breast and ovarian cancer patients [27]. One study failed to find an association between the LCS6 variant allele and sporadic or familial ovarian cancer risk [28]. A recent study confirms the importance of the LCS6 variant allele in postmenopausal ovarian cancer patients and demonstrates that it is a biomarker of poor outcome in this disease, probably due to platinum resistance [29]. The LCS6 variant allele is also associated with reduced survival in oral cancer [30]. On the contrary, Smits et al. found that early-stage CRC patients with the LCS6 variant allele had better outcome [31], whereas Ryan et al. recently reported the LCS6 variant allele to be associated with reduced risk of mortality in late-stage CRC [32].

We determined the frequency of the LCS6 variant allele in a Norwegian case–control study, the KAM cohort, consisting of CRC patients, individuals with polyps in the large intestine, and healthy controls, to investigate the feasibility of the variant allele as a risk factor for CRC development.

Three studies have reported conflicting results regarding the relationship between the LCS6 variant allele and clinical outcome in mCRC patients receiving cetuximab [33‐35]. In this work we have studied if the LCS6 variant allele was associated with clinical outcome in NORDIC-VII, a randomized phase III trial where Nordic FLOX (bolus 5-fluorouracil/folinic acid and oxaliplatin) was given with or without cetuximab as first-line treatment in mCRC [10].

The frequency of the LCS6 variant allele in a Norwegian cohort was similar in healthy controls, individuals with polyps, and CRC patients. These results do not support the LCS6 variant allele as a candidate risk factor for development of colorectal polyps or CRC. The frequency of the LCS6 variant allele varies across geographic populations, with European populations exhibiting the variant allele most frequently [24, 31, 32, 39, 40]. The frequency of the LCS6 variant allele was 16% in 535 genotyped patients in the NORDIC-VII cohort, which is consistent with a recent study which analyzed 734 CRC cases from the Netherlands [31]. Interestingly, the frequency of the variant allele in mCRC patients in the NORDIC-VII cohort is significantly lower than in the CRC patients from the KAM cohort (P = 0.02). A possible explanation could be that the NORDIC-VII cohort consists of a more heterogeneous population from the Nordic countries compared to the KAM cohort, which consists of a homogenous Norwegian population. Another possible explanation could be that CRC patients with the LCS6 wild-type have a greater tendency to metastasize.

A recently published study found that early-stage CRC cases with the LCS6 variant had improved survival [31]. Another study reported that the LCS6 variant was associated with a reduced risk of mortality in late-stage CRC [32]. Numerically increased median PFS and OS were found in patients with the LCS6 variant allele when compared to LCS6 wild-type carriers, but the differences were not statistically significant at the 5% level and we cannot conclude that mCRC patients with the variant allele belong to a favorable prognostic group. Furthermore, the difference in treatment response of adding cetuximab was larger, albeit not statistically significant, in patients with the LCS6 variant. There was a non-significant trend of increased response rate for patients with the LCS6 variant allele when treated with 5-fluorouracil/oxaliplatin and cetuximab compared to 5-fluorouracil/oxaliplatin alone. The trend of numerically increased PFS, OS and response rate was also observed independent of KRAS mutation status and in KRAS and BRAF wild-type patients, but none of the findings proved statistically significant. Thus, any potential predictive effect of the LCS6 variant allele is likely to be too small to be demonstrated with the patient sample available from the NORDIC-VII study. The NORDIC-VII cohort has limitations for studies of biomarkers predictive of cetuximab effect, as cetuximab did not add significant benefit to the Nordic FLOX regimen. Also, the number of patients with the LCS6 variant allele is relatively small, and the analyses of LCS6 as a predictive marker thus have low power.

Zhang et al. demonstrated that of 67 KRAS wild-type mCRC patients, there was a higher response rate and a trend of longer PFS and OS in patients with LCS6 variant allele (N=12) compared to patients with LCS6 wild-type (N=55) when treated with cetuximab monotherapy [34]. Contrary, another study on 121 BRAF wild-type mCRC patients who underwent salvage cetuximab – irinotecan therapy, of which 58 were KRAS mutated, reported that patients with LCS6 variant allele (N=34) had shorter PFS and OS compared to LCS6 wild-type (N=87) [33]. Similar results were reported by Winder et al. who found mCRC patients with mutant KRAS and LCS6 variant allele to have shorter PFS when treated with irinotecan and cetuximab [35]. The conflicting results in these studies suggest that the chemotherapy backbone may play a role, and that the LCS6 variant allele have different predictive values in mCRC patients treated with cetuximab alone or in combination with 5-fluorouracil/oxaliplatin than in patients treated with cetuximab in combination with irinotecan [41, 42].

Ragusa et al. demonstrated that cetuximab treatment induced miRNA transcriptome changes in drug-sensitive and drug-resistant CRC cell lines [17]. The set of differentially expressed miRNAs in the two cell lines (one sensitive and the other resistant) was almost entirely not overlapping. These data suggest that different responses to cetuximab are associated with different sets of miRNAs and thereby different molecular signaling. Interestingly, 67% of the differentially expressed miRNAs were involved in cancer, including CRC, whereas 19 miRNA targets had previously been reported to be involved in the cetuximab pathway and CRC. Based on their results, they suggest downregulation of let-7b and let-7e and the upregulation of miR-17* to be associated with cetuximab resistance [17]. Although these miRNAs were generated from cell studies, they illustrate that miRNAs may be promising predictive markers of cetuximab response to be further studied in mCRC patients.

We have only investigated one miRNA binding site polymorphism in this study, representing a small piece in a large puzzle of polymorphisms in the miRNA pathway. Future research on miRNA pathway polymorphisms as potential prognostic and/or predictive markers in mCRC should ideally include an integrated approach using bioinformatical tools combined with biological data to get a comprehensive understanding of the role and functions of miRNA polymorphisms and cetuximab response.

The LCS6 variant allele does not seem to be a risk factor for development of colorectal polyps or CRC. No prognostic effect of the LCS6 variant allele was demonstrated in the NORDIC-VII cohort. No predictive effect of the LCS6 variant allele on response rate, PFS or OS when cetuximab is given in combination with 5-fluorouracil/oxaliplatin in the first-line treatment of mCRC patients could be proven. Although our study has a larger sample size than previously published studies, the sample size in the subgroup with the variant allele is still too low to obtain sufficient power to reliably conclude in subgroups based on treatment and genotype. Studies with larger sample size and with stratification of molecular subgroups and chemotherapy backbone are necessary to reveal if the LCS6 variant allele could be clinically useful when treating mCRC patients in the future.