Introduction

Colorectal cancer (CRC) is one of the most prevalent cancers and accounts for over 8% of all deaths annually worldwide [1]. In the last few years, incidence and mortality of CRC is increasing rapidly in Iran [2]. CRC is a multifactorial disease and both genetic and environmental factors play an important role in development and susceptibility of the cancer. Multiple alternative genetic pathways involve in CRC tumorigenesis. KRAS and BRAF encode a downstream protein that belongs to the RAS-RAF-MEK-ERK signaling pathway. Hyperactivation of this signaling pathway plays a significant role in the proliferation, differentiation, invasiveness, and metastasis of tumor cells. Oncogenic mutations (especially KRAS and BRAF mutations) lead to the persistent activation of this pathway and accelerate the pathogenesis of CRC [3].

Various studies have reported that about 30–45% of CRC tumors harbor a KRAS mutation [4]. Somatic mutations in KRAS are frequently observed in patients with resistance to antiepidermal growth factor receptor (anti-EGFR) therapy and associated with poor prognosis in metastatic or recurrent CRC [5]. Similarly, several recent reports have suggested that the presence of BRAF mutations in about 10% of CRC tumors can also affect the response to anti-EGFR therapy [6]. BRAF mutations are known as an indicator of poor prognosis and negative predictive biomarkers of anti-EGFR therapy in advanced CRC [7]. The frequency of these somatic mutations varies considerably among different populations. Ethnicity, lifestyle and geographical factors seem to affect the frequency and prognosis of mutation [8,9,10]. Today, genotyping of KRAS and BRAF mutations is routinely undertaken as it is an important biomarker used to predict the poor efficacy of anti-EGFR therapy in patients with metastatic colorectal cancer (mCRC).

To date, most studies about association of KRAS and BRAF mutations with survival and clinicopathological features were from developed countries and only limited studies have been reported about prognostic value of KRAS and BRAF mutations in Iranian patients with CRC [11,12,13]. Therefore, it is needed to assess the prognostic value of these mutations and its relationship with clinical/clinicopathological features. The results from previous studies did not reach a consensus and very few studies were performed to show the prognostic value of KRAS and BRAF mutations on overall survival in Iranian CRC patients. Consequently, the present study aimed to identify the frequency of KRAS and BRAF gene mutations in Iranian CRC patients, and investigate the prognostic value of KRAS and BRAF mutations and their associations with clinicopathological features.

Methods

Patient Samples

A hundred formalin-fixed, paraffin-embedded (FFPE) tumor blocks from patients diagnosed with colorectal cancer from February 2012 to August 2015 at the three different hospitals (Afzalipour, bahonar and mehregan hospitals) throughout Kerman province (southeast of Iran) were retrieved. Tumor sections from each FFPE tissue sample were stained with hematoxylin and eosin (H&E) and reviewed by two experienced pathologists (S.D and N.K.K) independently to estimate the percentage of tumor cells. Then, from different FFPE tissue blocks of each patient, the rich-tumor areas (more than 50% tumor cells) with lowest necrosis, hemorrhage, normal colonic cells, stromal cells, and blood-derived leukocytes were selected for genomic DNA extraction. The population study included patients with initial diagnosis of CRC and no patients had accepted adjuvant treatment at the time of sampling. Demographic, clinical, and clinicopathological data were obtained by reviewing the medical records to extract the following information which include age of diagnosis, sex, smoking status, alcohol intake, family history, tumor location (right, left or rectum), differentiation grade (well, moderate or poor), TNM stage (I, II, III, or IV), lymph node metastasis, and distant metastasis. Adjuvant chemotherapy was recommended to CRC patients according to Iranian Ministry of Health guidelines and international guidelines for diagnosis and treatment of CRC. According to the RAS driver variants status, patients were offered targeted agents as an adjunct to systemic chemotherapy. However, due to insurance and economic issues, no patients received anti-EGFR and/or anti-VEGF therapy during the study period. This study was performed under the license from Ethics Committee of Kerman University of Medical Sciences (Approval No. IR.KMU.REC.1397.209), and due to the retrospective nature of the study and unavailability of many patients, informed consent was rejected.

Formalin-Fixed Paraffin-Embedded Tissue DNA Extraction

For Genomic DNA extraction, 5–10-μm-thick sections were cut from FFPE tumor tissue blocks for each case and collected in 1.5 ml tubes. Paraffin was removed by using two washes with 1 ml of absolute xylol and a wash in 1 ml absolute ethanol. After each wash, the sections were vortexed and centrifuged at 13,000 × rpm for 10 min, and the supernatant was then removed. DNA was extracted from FFPE specimens using the QIAamp DNA FFPE Tissue Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s protocol. The concentration and purity of all DNA samples were measured using NanoDrop ND-2000c Spectrophotometer (Thermo Scientific, USA). The mean concentration of DNA was 220.4 ng/μl, range from 40.2 to 358.6, and the ratio of A260/A280 was from 1.6 to 2.1. DNA was finally eluted in 50 μl of ATE buffer and stored at − 20 °C until use.

Detection of BRAFV600E Mutation

Detection of BRAFV600E mutation was performed using the therascreen BRAF RGQ PCR Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s protocol. This kit utilizes a combination of ARMS (Amplification Refractory Mutation System) and Scorpion probes technology for mutation-specific amplification and detection of PCR products, respectively. This method allows detection of four different mutation including V600E, V600K, V600D, and V600R. Amplification, detection, and data analysis were performed on a Corbett Rotor-Gene 6000 Real-time PCR instrument (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions.

Detection of KRAS Mutations

The mutational analysis of KRAS (exon 2 and 3) was performed using PCR products and bidirectional sequencing from DNA samples. The primers used to evaluate exon 2 [14] and 3 [15] of KRAS were as previously described. PCR was performed on a Bio-Rad T100 Thermal Cycler in a total volume of 30 μl, containing 40–100 ng of template DNA, 15 μl 2× Taq DNA Polymerase Master Mix RED (Ampliqon, Denmark), and 10 mM of each primer. Amplification was carried out with the following condition: 95 °C for 10 min (first cycle); 35 cycle of 95 °C for 15 s, 56 °C for 15 s, 72 °C for 15 s; and final extension cycle at 72 °C for 10 min. PCR products were purified with a PCR product purification kit (Roche, Germany), according to the manufacturer’s instructions. Forward and reverse strands sequenced by the BigDye Terminator v3.1 kit (Applied Biosystems, Foster City, USA) on an ABI 3130xl genetic analyzer (Applied Biosystems, Foster City, USA) and the sequence data were analyzed using Sequencing Analysis software v5.0 (Applied Biosystems, Foster City, USA). Visual evaluation of each chromatogram was done by two independent investigators and all abnormal or ambiguous sequences were confirmed by re-sequencing. Sequences were compared by the BLAST tool (www.ncbi.nim.nih.gov/BLAST/).

Statistical Analysis

The continuous data were presented in mean ± SD, analyzed by independent student t test. Normality of data was analyzed by Kolmogorov–Smirnov and Shapiro–Wilk test. Categorical variable data analysis was conducted using the χ2 or Fisher’s exact test. The χ2 test or Fisher’s exact test was used to compare the proportion of mutations among patients with different clinicopathological data. Logistic regression models were used to analyze the association based on the estimation of the odds ratios (ORs) and 95% confidence intervals (CIs). Overall survival (OS) was defined since the date of diagnosis up to the date of death or last of follow-up visit. The overall survival was plotted and analyzed by Kaplan–Meier (log-rank test). All statistical analyses were conducted by using SPSS 22.0 statistical package (SPSS Inc., Chicago, IL, USA). All p values were two-sided. The statistical significance was considered if the p value < 0.05.

Results

Characteristics of CRC Patients

In this study, we retrieved 100 FFPE tissue blocks from 3 different centers of Kerman province. Table 1 summarizes demographic data and clinicopathological characteristics of CRC patients. Briefly, the prevalence of CRC was 64% in males and 36% in females. The average patient age was 59.60 ± 15.24 years (range from 19 to 85 years), and patients who aged younger than 60 years and older than 60 years respectively represented in 46% and 54% of the patients. The proportion of familial history of CRC was 15%. The tumor size of the patients ranged approximately from 2 to 10 cm (median size was 5.85 ± 3.4). The tumors were located at the right side of the colon (29%), including cecum and transverse colon; left side of the colon (30%), including sigmoid colon and splenic flexure; and rectum (41%). Regarding the tumor differentiation, 8% of tumors were well differentiated, 78% were moderately differentiated and 14% were poorly differentiated. In total, the percentage of patients in stages I to IV were 11, 17, 59, and 13, respectively. The liver metastasis (10%) had the most frequently metastatic site, followed by nonregional lymph node and vessels (each 7%). There was no clinical information regarding previous chemotherapy.

Table 1 Demographics and Clinicopathological features according to KRAS and BRAF gene mutation status in 100 colorectal cancer patient
Full size table

Distribution of KRAS and BRAFV600E Mutations in CRC Patients

The prevalence and distribution of KRAS and BRAF mutations in the CRC patients is showed in Table 2. KRAS exon 2 and 3 amplification was done using polymerase chain reaction and detected in the presence of 293 bp and 289 bp fragments on 2% agarose gel electrophoresis, respectively. An example of representative electropherogram of KRAS mutant (exon 2 and 3) is shown in Fig. 1. KRAS mutation was identified in 29 (29%) of all the patient samples. Among 29 KRAS mutants, 27 (27%) had mutations in exon 2 and 2 (2%) in exon 3. Within KRAS exon 2, 21 (77.8%) of the mutations were identified in codon 12, and 6 (22.2%) were in codon 13. The most frequently observed mutation (13 of 27) was a 35G>A transition in codon 12 (G12D), followed by 38G>A (G13D) and 35G>T (G12V) (6 of 27 each). Also, 2 mutations were detected in codon 61 of exon 3 (183A>C, Q61H). Of the 100 tumor samples, 7 samples (7%) harbored a mutation in codon 600 of the BRAF gene (1799T>A, V600E). No other recurrent forms of BRAFV600 mutations (including V600K, V600D, and V600R) were identified in current study. Figure 2a–d shown the distributions of all three tumor subgroups (KRAS-mutant, BRAF-mutant, and negative) with respect to tumor sites, tumor differentiation, TNM stage, and distant metastasis. Additionally, Fig. 2e shown the distribution of KRAS and BRAF mutations by specific mutation in all included patients.

Table 2 Frequency and distribution of observed KRAS and BRAF mutations in 100 Iranian CRC patients
Full size table
Fig. 1
figure 1

Partial direct sequencing electropherogram of KRAS gene (exon 2 and 3): a exon 2, G12D mutation; b exon 3, Q61H mutation

Full size image
Fig. 2
figure 2

Distribution of tumor samples for KRAS and BRAF mutation status with respect to: a tumor site, b tumor differentiation, c TNM stage, d distant metastasis, and e type of mutation

Full size image

Correlation of KRAS and BRAF Gene Mutations with Clinical and Clinicopathological Characteristics

We analyzed the correlation between KRAS or BRAF mutations and the clinicopathological characteristics of CRC samples. A summary of the relationships between KRAS or BRAF mutations and different clinicopathological characteristics is provided in Table 3. Statistical analysis of the various characteristics variables showed a significant correlation between KRAS mutations and the older age (≥ 60 years) (OR 1.045; 95% CI 1.018–1.093,p = 0.044), but there were no statistically significant correlation between KRAS mutations, and sex (OR 0.934; 95% CI 0.328–2.659,p = 0.898), smoking (OR 2.863; 95% CI 0.92–8.85,p = 0.410), alcohol intake (OR 2.225; 95% CI 0.447–11.075, p = 0.329), family history (OR 0.924; 95% CI 0.233–3.662, p = 0.910), tumor location (p = 0.759), tumor size (OR 1.119; 95% CI 0.589–2.093, p = 0.729), tumor differentiation (p = 0.728), TNM stage (p = 0.733), lymph node metastasis (p = 0.543), and distant metastasis (OR 0.929; 95% CI 0.289–2.982, p = 0.901). Although, no correlation was observed between specific mutations in codon 12 and 13 and different clinicopathological parameters (data not shown).

Table 3 Multivariate logistic regression analysis of the relationship between KRAS and BRAF mutations and clinical/clinicopathological features in CRC patient
Full size table

The BRAFV600E mutant tumors in clinical samples were associated with older age (≥ 60 years) (OR 2.947; 95% CI 0.283–8.761, p = 0.031) and poor differentiation (OR 3.162; 95% CI 1.131–14.437, p = 0.017). There were no statistically significant correlation between BRAFV600E mutation and sex (OR 1.129; 95% CI 0.468–2.724, p = 0.778), smoking (OR 0.507; 95% CI 0.582–4.437, p = 0.533), alcohol intake (OR 0.935; 95% CI 0.087–0.987, p = 0.488), family history (OR 1.731; 95% CI 0.273–10.998, p = 0.533), tumor location (p = 0.671), tumor size (OR 1.195; 95% CI 0.638–2.240, p = 0.578), TNM stage (p = 0.688), lymph node metastasis (p = 0.943), and distant metastasis (OR 1.457; 95% CI 0.612–3.454, p = 0.390).

Overall Survival Analysis

Median of survival time for KRAS mutant was 42.6 months and for KRAS wild-type patients was 45.9 months (overall = 44.9 months). The Kaplan–Meier survival curves (5-year OS) based on KRAS and BRAF status has been shown in Figs. 3 and 4, respectively. The analysis showed no difference of OS between KRAS mutant patients and KRAS wild-type patients in 5-year (62.1% vs. 71.8%; log rank p = 0.543) OS. Similarly, patients with BRAF mutations have no significant association with overall survival rate. Median of survival time in patients with BRAF mutation was 39.7 and in BRAF wild-type patients was 44.5 months (overall = 44.3 months). As shown in Fig. 3, the 5-year OS rate in BRAF-mutant and BRAF wild-type patients was 57.1% and 67.7%, respectively (log rank p = 0.673).

Fig. 3
figure 3

Kaplan–Meier curves for 5-year OS based on KRAS status (log-rank p value = 0.543)

Full size image
Fig. 4
figure 4

Kaplan–Meier curves for 5-year OS based on BRAF status (log-rank p value = 0.673)

Full size image

Discussion

In this study, we evaluated KRAS and BRAFV600E mutations frequencies in 100 CRC FFPE tissue samples from three different hospitals located in Kerman, Iran. The present study utilized direct sequencing and real-time PCR methods to analyze the mutational status of KRAS and BRAFV600E in Iranian CRC patients, respectively. According to our findings, the prevalence of KRAS and BRAFV600E mutations were 29% and 7%, respectively. The remaining 64% of patients had no mutation in any type of genes analyzed.

KRAS is the most studied gene of RAS-RAF-MAPK pathway in CRC. It triggers the downstream cascades including the PIK3-AKT pathway, which may affect the cell proliferation, differentiation, and invasion [16]. Mutations in KRAS negatively predict the response to anti-EGFR therapies in patients with metastatic CRC. According to previous studies, the prevalence of KRAS mutation in CRC patients varies from 12.9 to 66.1% (mostly 30–45%) across the globe [17]. To our findings, the frequency of Iranian CRC patients with KRAS mutant tumors was 29%, which was similar to published reports from Asia and Europe [18,19,20]. It means that, if KRAS mutant profiling is applied to select candidates for anti-EGFR treatment, the number of Iranian patients that would be excluded is similar to that of other populations. In CRC patients, most of the KRAS mutations occur in codon G12 or G13 (about 90%), and the G12D (35G>A) is the most common mutation which results in an amino acid substitution (from glycine to aspartic acid) in KRAS codon 12 [21, 22]. Our data were consistent with these reports. In comparison with Iranian studies, these data is similar to other published data from various regions of Iran [11, 23,24,25,26].

Several studies have reported the associations of KRAS mutation with different clinical and/or clinicopathological features, including female gender [27,28,29,30,31], age at diagnosis (> 50 years) [28], location of tumor (right side) [18, 27, 31], tumor differentiation (well/moderately differentiated) [18, 31,32,33], TNM stage [30], and microsatellite-stable phenotype [33] in Caucasian and Asian CRC populations, while others did not report any association [34, 35]. In the current study, our findings showed the association of KRAS mutation with older age (≥ 60 years), which is consistent with a recent report of Iranian patients [12]. This condition can be due to increased genetic alterations of tumors with age. However, this finding was contrary to the findings of a study conducted by Nazemalhosseini-Mojarad et al. in Iranian CRC patients [11]. Other clinicopathologic features did not correlate significantly with the occurrence of KRAS mutations.

According to several earlier reports, there is no convincing evidence that KRAS mutations are independent prognostic biomarker for poor OS in CRC patients [5, 34, 36, 37]. Among these, Abubaker et al. [36] and Richman et al. [5] found that KRAS mutations are associated with a poorer overall survival. Additionally, a meta-analysis conducted by Qiu et al. [38] showed that OS was significantly shorter in KRAS mutant patients compared with that in KRAS wild-type patients. In contrast, a recent study on 353 Chinese CRC patients revealed that KRAS mutations were not associated with OS, but BRAF mutations were associated with poorer OS [34]. Conversely to many published reports, we identified no significant association between KRAS mutations and OS in Iranian patients with CRC and it can be due to our sample selection or low sample size. Similar to our findings, recent Iranian studies did not find any association between KRAS mutation and OS [11, 13].

BRAF is a downstream member of the RAS-RAF-MAPK signaling pathway and its mutation is the most commonly observed gene alteration after KRAS mutation in colorectal cancer. In the present study, the prevalence of BRAFV600E mutation was 7% (7/100), one of the highest prevalence reported in the Iranian CRC patients. Notably, most Iranian studies reported low frequency of BRAFV600E (mostly no BRAF mutation) in CRC patients [39,40,41,42]. In 2008, Brim et al. reported a very low frequency of BRAF mutation (2%) among Iranian CRC patients [42]. Additionally, Nazemalhosseini-Mojarad et al. [11] who aimed to explore the BRAFV600E mutation in 258 Iranian CRC patients also demonstrated that the prevalence of BRAFV600E mutation was 5.8%. According to best of our knowledge, this is the first Iranian study that used real-time PCR method in reporting the frequency of BRAF mutation among CRC patients, which can be a reason for the higher frequency of BRAF mutation. Although, a study conducted by Mohammadi-Asl et al. [43] showed that 46.25% (37/80) of the patients with colorectal cancer had BRAFV600E mutation. This high frequency of BRAFV600E mutation in CRC from Ahwaz city (southwest of Iran) could be due to different sample selections, different methodologies, different ethnicity of people, geographical and environmental features of this region, and common lifestyle. In addition, the frequency of BRAFV600E mutation is higher than several Asian populations (China, Japan, Korea, Taiwan, and Malaysia), but similar to that of western countries (France, UK, USA, Greece, and Italy [18, 20, 27, 33, 44,45,46,47,48,49,50,51,52]. According to published data from a distinct ancestral populations study, BRAF mutations occurring at a higher frequency in European patients with colorectal cancer with (17%) versus Asian (4%) [53]. BRAFV600E mutation frequency ranged from 0 to 22% within CRC patients from various geographical regions across the globe [17, 18, 40, 47, 54]. A recent meta-analysis conducted by Lowe et al. in 2019 [55] estimated that the global prevalence of BRAF mutation in CRC patients was 7.1%; however, two previous meta-analysis reported that it was 10.8% and 11.1% [56, 57]. These data indicate a slight reduction in the prevalence of BRAF mutation across the globe. The exact reason of this variation is still not clear, but the racial and/or environmental factors might contribute to the difference. Several studies have reported that BRAF mutation existed only in KRAS wild-type tumoral tissues, which is consistent with our results [17, 18, 32, 58]; however, few studies reported concurrent mutation of KRAS and BRAF mutation in CRC patients [27]. In this context, tumor heterogeneity may play a role [59].

We found that BRAF mutations occurred more frequently in older patients (age ≥ 60 years). BRAF mutations were also more common in poorly differentiated tumors, which is inconsistent with previous reports from different regions of Iran [11, 39]. Recent studies in Iranian populations did not find any correlations between BRAF mutations and clinical and/or clinicopathological features [11, 39]; however, most studies did not investigate the association of BRAF mutation with clinical and/or clinicopathological features [26, 40,41,42]. In contrast, a recent study reported association of BRAFV600E mutation with mucinous characteristics in Iranian CRC patients [11]. In several case series and meta-analysis studies, BRAF mutation was either strongly correlated with clinicopathological features such as older age, female gender, tumor location, poorly differentiated tumor, or at least showed a trend toward such correlations [27, 33, 46, 56, 57, 60].

Several studies widely reported that there is a correlation between BRAFV600E mutation and worse OS in CRC patients [34, 61]. Although, our findings showed that BRAFV600E mutation was not correlated with OS. Similarly, a previous research from Iran has reported that BRAF mutations are not associated with worse OS [11]. Anyway, this findings requires further confirmation in a larger Iranian CRC population and multicenter setting given the relatively low frequencies of BRAFV600E mutation (n = 7) in our study.

There were some limitations of this study. Part of our study limitations were small size and bias sample selection, incomplete information on clinicopathological data, observational retrospective nature of research, incomplete follow-up time such as recurrence and OS, and absence of epigenetic or MSI status of tumor tissues. Additionally, some hotspot mutations such as in exon 4 of the KRAS gene and in exon 11 and 15 of the BRAF gene were not screened. We collected tumoral tissues from three different hospitals in Kerman province (southeast of Iran) and evaluated the frequency of KRAS and BRAF mutations, associations with clinicopathological data, and correlations of these mutations with poor prognosis in Iranian CRC patients. These aspects make our findings more representative and prognostic for new CRC patients in Iran or southeast of Iran at least.

Conclusion

In this study, we identified KRAS and BRAFV600E mutations in 29 (29%) and 7 (7%) of the Iranian CRC patients, respectively. Distinctively, our study revealed a higher prevalence of BRAFV600E mutation in Iranian CRC patients. The prevalence of KRAS mutations was higher in older age CRC patients. Also, the BRAFV600E mutation was more common in older age patients, and patients who showed poor differentiation in clinical samples. This study adds to the evidence that KRAS and BRAF mutations in Iranian colorectal cancer patients occur at a similar status to that of other populations; however, prevalence of BRAF mutation is higher in this study than in previous Iranian studies. Our findings open the field to further studies investigating how these mutations can be variable in frequency in different populations. However, large-scale clinical studies are needed to confirm this finding in Iranian CRC patients.