Retrospective study of RAS/PIK3CA/BRAF tumor mutations as predictors of response to first-line chemotherapy with bevacizumab in metastatic colorectal cancer patients

This was a retrospective study conducted at a single Japanese institute and approved by the ethics committee of Cancer Institute Hospital of Japanese Foundation for Cancer Research (No.2009-1048). Of the 1001 consecutive patients with histologically confirmed CRC who were examined for tumor RAS, PIK3CA, and BRAF mutations in our institute between November 2006 and December 2013, 90 patients were administered combination chemotherapy with Bmab as the first-line treatment for mCRC. Patients who received neo-adjuvant chemotherapy (NAC) or adjuvant chemotherapy completed less than 6 months before enrollment to this study were excluded. Patients who had undergone surgery for metastatic sites were included if it had been performed more than 4 weeks earlier. Patients were required to have adequate hematologic, hepatic, cardiac, and renal function. Their medical records were reviewed to obtain data on clinicopathologic variables. All patients provided written informed consent before receiving treatment.

The treatment regimen was determined by the physician for each patient. The following regimens were employed: modified FOLFOX6 plus Bmab consisted of a fortnightly course of Bmab (5 mg/kg intravenously over 30 to 90 min on day 1), oxaliplatin (85 mg/m² intravenously over 2 h on day 1) plus l-LV (200 mg/m² intravenously over 2 h on day 1) and 5-fluorouracil (5-FU) (400 mg/m² bolus on day 1, followed by infusion of 2400 mg/m² over 46 h); and CapeOX plus Bmab consisted of oxaliplatin (130 mg/m² intravenously over 2 h on day 1) plus oral capecitabine (1000 mg/m² twice daily for 2 weeks in a 3-week cycle). Bmab (7.5 mg/kg) was administered ahead of oxaliplatin intravenously on day 1 every 3 weeks. FOLFIRI plus Bmab consisted of fortnightly courses of Bmab (5 mg/kg intravenously over 30 to 90 min on day 1), irinotecan (150 mg/m² intravenously over 2 h on day 1) plus l-LV (200 mg/m² intravenously over 2 h on day 1) and 5-FU (400 mg/m² bolus on day 1, followed by infusion of 2400 mg/m² over 46 h).

DNA was extracted from formalin-fixed paraffin-embedded (FFPE) tumor tissue, which was mostly obtained at biopsy. Mutations in KRAS codons 12 and 13 were examined using a kit based on a Luminex assay (MEBGEN KRAS Mutation Detection kit, MBL). A Luminex based kit (GENOSEARCH Mu-PACK, MBL) was also used to detect a total of 36 mutations in KRAS (codons 61 and 146), NRAS (codons 12, 13, and 61), PIK3CA (codons 542, 545, 546, and 1047) and BRAF (codon 600). The concordance of findings based on this newly developed multiplex assay kit with conventional direct sequencing results was confirmed previously [6].

The objective response rate (ORR) was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) ver. 1.1. The progression-free survival (PFS) and overall survival (OS) were calculated using the Kaplan-Meier method. PFS was defined as the duration of survival from the start of chemotherapy to the date of recurrence or death from any cause, whichever occurred first. Patients with no recurrence until the cut-off date were regarded as censored on the last date when no recurrence had been proven by imaging. The disease-progression date was retrospectively re-analyzed by the investigator, and was defined as the date on which progression was first detected using a computed tomography (CT) or fluorodeoxyglucose-positron emission tomography (FDG-PET) scan. If treatments were discontinued before or continued after disease-progression due to adverse events or the patient’s request, they were censored at the time of the last radiological examination. OS was defined as survival from the start of chemotherapy to death from any cause. For patients who were lost to follow-up, data were censored on the date when the patient was last known to be alive. The data cut-off date was August 12, 2015. A one-sided Fisher’s exact test was used to assess the statistical significance of the difference between ORRs according to mutational status at a significance level of 2.5%. Both PFS and OS were estimated using the Kaplan-Meier method and compared using the log-rank test at a significance level of 5%. In addition to RAS, PIK3CA, and BRAF tumor mutations, variables with a P value less than 0.05 in a univariate analysis were included in a multivariate Cox regression analysis. All analyses were performed with EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R software (The R Foundation for Statistical Computing) [7].

The baseline characteristics of the patients are shown in Table 1. Their median age was 63 years (range, 27–79 years). Forty-eight patients (53.3%) were men and 42 patients (46.7%) were women. Almost all of the subjects had a good performance status. Seventy-four patients (82.2%) had colon cancer and 16 (17.8%) had rectal cancer, including right-sided colon cancer (RCC) in 34 patients (37.8%) and left-sided colorectal cancer (LCRC) in 56 patients (62.2%). RCC was defined as a tumor arising from the cecum to the transverse colon, excluding the appendix, while LCRC was defined as a tumor arising from the descending colon to the rectum. Of these 90 patients, the tumors of 43 patients (46.7%) were found to have a mutation in KRAS exon 2. In total 48 patients (53.3%) had a RAS mutation (KRAS/NRAS). Seven patients (8.9%) had a PIK3CA mutation, and another 7 (8.9%) had a BRAF mutation. Thirty-three patients (36.7%) had tumors with no RAS, PIK3CA, or BRAF mutation.

Almost all patients received an oxaliplatin-containing regimen, which was FOLFOX in 34 cases (37.8%) and XELOX in 52 cases (57.8%). Among them, 13 (14.4%) had been administered oxaliplatin prior to this treatment as an adjuvant therapy. The primary tumor was resected in 67 patients (74.4%) and 13 patients (14.4%) underwent a metastatectomy.

The median follow-up period for all eligible patients was 23.5 (0.8–41.4) months, and 51 patients (56.7%) died by the cut-off date. Seventy-seven of the 90 patients had measurable lesions. The overall ORR was 52.6% whilst the ORR of patients with no detected tumor mutations was 64.3%. The ORRs of patients with a PIK3CA or BRAF tumor mutation were very low (28.6%), and more than 40% of patients with a BRAF tumor mutation had confirmed disease progression at the first evaluation (Fig. 1). Although the ORR varied according to the mutational status of the tumor, these differences were not statistically significant (Table 2). The differences in ORR became gradually greater among patients with wild-type tumors as restricting treatment subjects from only wild type with respect to KRAS exon 2, all RAS wild-type toward all wild-type (RAS/PIK3CA/BRAF). The difference in ORR between the whole population and patients with all wild-type tumors was 11.7% (Fig. 2).

The overall median OS and PFS were 27.7 and 13.3 months, respectively. The ORR of patients with KRAS exon 2 wild-type and mutant-type tumors differed by 6.2%, and these populations had nearly identical Kaplan-Meier curves for PFS (Fig. 3). The difference between ORRs (18.4%) was larger when comparing patients with wild-type tumors to those with a tumor carrying any mutation. There was no statistically significant difference in PFS between these groups, although there was a slightly larger difference in Kaplan-Meier curves (Fig. 3).

Univariate analysis revealed that an elevated serum C-reactive protein (CRP) level (>0.05 mg/dl), an unresectable primary tumor, and liver metastases were associated with a significantly shorter PFS (Table 3). Multivariate analysis that included RAS, PIK3CA, and BRAF tumor mutations and baseline prognostic variables revealed that liver metastasis, unresectable primary tumor, RAS and BRAF tumor mutations had independent prognostic value for early progression (Table 3).