Correlation between KRAS and NRAS mutational status and clinicopathological features in 414 cases of metastatic colorectal cancer in Morocco: the largest North African case series

Colorectal cancer (CRC) is the second most common cause of cancer deaths [1]. Approximately 20% of patients with CRC present with metastases at the time of diagnosis [1]. Metastatic and recurrent diseases are associated with poor survival [2]. Advances in molecular biology have improved understanding of the molecular features of carcinogenesis and progression of CRC. Thus, the identification of prognostic and predictive biomarkers made it possible to refine prognosis in order to prescribe targeted therapies as part of a personalized medicine. In this context, it has been demonstrated that Epidermal growth factor receptor (EGFR) signaling is an important intracellular signaling pathway in CRC development and progression. RAS (for rat sarcoma virus) proteins are downstream signaling molecules of EGFR [3, 4]. There are three isoforms of RAS: KRAS, NRAS and HRAS [3, 4]. The efficacy of anti-EGFR depends upon the RAS mutational status. Mutations in RAS are associated with resistance to anti-EGFR therapy [3, 5‐8]. HRAS is rarely mutated in CRC [3]. Treatment guidelines now recommend that all metastatic CRC (mCRC) patients should be tested for KRAS and NRAS if anti-EGFR therapy is available [9‐11]. Only few studies were conducted in low middle income countries due to the high costs related to tumor mutation tests. Our first experience showed no statistically significant relationship between mCRC patient characteristics and KRAS/NRAS mutations. However due to low samples no conclusion has been made. The aim of this study was to report KRAS and NRAS mutation status frequencies in Moroccan patients with mCRC and investigate the associations of these mutations with clinicopathological characteristics in a larger study.

This study represents the second report of our pilot study in Morocco using the Idylla™ platform: a fully automated, real-time PCR based molecular diagnosis system. Short-term laboratory-based training of technicians for one day only was necessary.

This study showed RAS mutation in 54.6% of cases, a low rate of invalid cases despite being a pilot study, significant correlation between NRAS mutation and age and no correlations between RAS mutations and other clinicopathological characteristics.

Previous studies have demonstrated highly variable frequencies of RAS mutations in North Africa ranging from 16 to 75.2% [12‐25] (80% for one study but with 20 patients only) [26].

RAS mutation is an early event in colorectal carcinogenesis. In this study, most of tissues (87.20%) were from primary tumor. This high frequency of primary tumor on specimens for RAS study is probably due to habitudes of our institute to not systematically re-biopsy for metastases or rapid progression of disease. Although a high concordance exists between primary CRC and liver and lung metastases [27, 28], metastatic or recurrent CRC tissues are the preferred specimens for RAS (and other biomarkers) testing [29]. Indeed, acquired KRAS mutations during progression of CRC metastases can occur [30]. In absence of metastatic and recurrent CRC tissues, primary tumor tissue is an acceptable alternative [29, 30].

We obtained 20 cases with invalid RAS results: 7 (1.7%) for KRAS and 13 (3.1%) for NRAS. We think that this low rate of invalid cases was achieved through respect to pre-analytical factors: cold ischemia time and formalin fixation. All invalid cases corresponded to prepared outside of the laboratory paraffin blocks. We think that a probable non-optimal pre-analytical phase may have caused irreversible DNA and RNA damage. This made their amplification impossible.

For correlation of primary tumor localization and RAS mutation, literature also was controversial. KRAS was significantly more frequent in left-sided CRC in three studies [20‐22]. In our study, no correlation between primary tumor site and RAS mutation was observed. Data for association between RAS status and histological type and tumor differentiation were also inconclusive. A summary of frequency and association between RAS status and clinicopathological features in North Africa are shown in Table 6.

Our preliminary results did not show a statistically significant relationship between clinicopathological features and KRAS and NRAS mutations (n = 186) [31]. In the present study, there was a trend towards older patients (> 60 years) for NRAS mutations. Indeed, only one patient with NRAS mutation has ≤ 60 years (n = 1/12).

Worldwide, distribution of RAS mutations is uneven. The KRAS mutation rate is 44.7% in Western Europe, 35.8% in Eastern Europe [32], 41% in Indonesia [33] and 36.1% in China [34] (for Asia) and 19.5% in the Middle East [32]. In many studies, KRAS mutation incidence was higher in right-sided than in left-sided colon tumors [34‐36] and higher in women [34, 36, 37] but not for NRAS mutation [34, 35]. Other studies did not find differences in location of primary tumor [37, 38] and patients’ gender [38] for KRAS mutation prevalence. Interestingly, KRAS was mutated significantly more often in the primary tumors of patients with lung metastases in one study [27].

This study has several limitations:

1) The small number of samples compared to the published literature from Western countries. However, it represents the largest North African case series.

2) The use of the Idylla™ platform. Indeed, Next Generation Sequencing (NGS) is the current diagnostic gold-standard for RAS mutational analysis in CRC. Several studies have demonstrated that Idylla™ testing is highly accurate with excellent concordance with NGS [39‐41]. In addition, Idylla™ testing is less than 2 min hands-on time and rapid (2 h) compared with NGS. Costs are lowest for Idylla™ making molecular testing affordable (no molecular pathology platform is needed). This can be very useful for low- and middle-income countries, especially for a small number of cases to test. Our laboratory is in the process of setting up a molecular pathology platform. Our perspective is to retest all 414 specimens with NGS to determine accuracy of our technique.

3) The main objective of the study did not include the analysis of survival and recurrence but this is consistent with this first pilot study of feasibility and assessment of RAS mutations. With more hindsight we will be able to report survival and recurrence of this case series.

We report the largest North African analysis of NRAS and KRAS status in colorectal metastatic patients. This study showed the ability in low middle income countries to perform a high rate of valid tests and the unusual trend towards older patients for NRAS mutations.