Prevalence of KRAS, PIK3CA, BRAF and AXIN2 gene mutations in colorectal cancer and its relationship with dental agenesis: a systematic review

Introduction: The study of allelic and genotypic frequencies contributes to determining the distribution of genetic variants in different populations and their possible association with biomarkers. This knowledge could improve the decision-making process regarding the management of some diseases such as colorectal cancer (CRC), in which the detection of clinical biomarkers such as dental agenesis could be crucial in clinical practice.
Objective: To evaluate the available scientific evidence on the prevalence of KRAS, PIK3CA, BRAF and AXIN2 mutations and their possible association with dental agenesis in people with CRC.
Materials and methods: A systematic search was conducted in PubMed, EMBASE and Cochrane Library databases using the following search strategy: type of studies: observational studies reporting the prevalence of KRAS, PIK3CA, BRAF and AXIN2 mutations in people diagnosed with CRC and their possible association with dental agenesis; publication language: English and Spanish; publication period: 2010-2020; search terms: “Genes”, “RAS”, “Kras”, “PIK3CA”, “BRAF”, “AXIN2”, “Mutation”, “Polymorphism”, “Colorectal Neoplasms”, “Colorectal Cancer”, used in different combinations (“AND” and “OR”).  
Results: The initial search yielded 403 records, but only 30 studies met the eligibility criteria. Of these, 11, 5, 5 and 1 only reported the prevalence of PIK3CA, KRAS, BRAF and AXIN2 mutations, respectively; while 8 reported the prevalence of more than one of these mutations in patients with CRC. The prevalence of KRAS (p.Gly12Asp), PIK3CA (p.Glu545Lys), and BRAF (p.Val600Glu) mutations ranged from 20.5% to 54%, 3.5% to 20.2%, and 2.5% to 12.1%, respectively. There were no findings regarding the association between the occurrence of these mutations and dental agenesis.  
Conclusions: KRAS mutations were the most prevalent; however, there is no evidence on the association between dental agenesis and the occurrence of KRAS, PIK3CA and BRAF germline mutations in individuals with CRC.

Objetivo. Evaluar la evidencia científica sobre la prevalencia de mutaciones KRAS, PIK3CA, BRAF y AXIN2 y su posible asociación con la agenesia P u b l i c a c i ó n a n t i c i p a d a .
R e v . F a c . M e d . dental en individuos con CCR.

Introduction
Colorectal cancer (CRC) is one of the leading causes of morbimortality worldwide (1). According to demographic projections and temporal profiles, its global incidence is expected to increase by 60%, leading to more than 2.2 million new cases and 1.1 million deaths by 2030 (2). The cancer pathogenesis is complex and has not been completely understood.
However, genetic factors reportedly play a critical role in tumorigenesis The main treatment approach for CRC is surgery and chemotherapy, which has an effectiveness of approximately 75%. However, roughly 30% of treated patients could develop new neoplastic polyps (8), suggesting that this treatment is not totally effective for this disease (9). Thus, new therapies have been proposed, developing anti-EGFR monoclonal antibodies drugs, considering that the Epidermal Growth Factor Receptor (EGFR) is the major therapeutic target in colorectal cancer (10). However, therapeutic effectiveness is affected in the presence of KRAS, BRAF and PIK3CA mutations (11,12). Thus, these genes constitute important biomarkers for CRC. In addition, it has been reported that mutations in P u b l i c a c i ó n a n t i c i p a d a .
R e v . F a c . M e d .
AXIN2 could act as a diagnostic biomarker for CRC associated to dental agenesis, so this gene variant and the presence of this oral developmental anomaly have been proposed as a predictive factor for this malignant disease (13).
Non-syndromic dental agenesis, the most common human anomaly (14), is the congenital absence of one or more permanent teeth due to alterations during early stages of dental development (15). Genetically, it has been described as the association between AXIN2 gene and teeth development in mice, suggesting its possible participation in human dental development (16). Lammi et al. reported the association between dental agenesis and the predisposition to CRC by a nonsense mutation According to the above statements, it is important for all the healthcare professionals to know about the prevalence of these genetic mutations and their possible association with clinical biomarkers such as dental agenesis to improve the diagnosis, prognostic and early treatment of cancer. Particularly for dentistry practitioners, who have an important role in detecting dental agenesis, this process could promote possible medical diagnoses of CRC. Therefore, the aim of this study was to evaluate the available scientific evidence on the prevalence of KRAS, PIK3CA, BRAF, AXIN2 mutations and their possible association with dental agenesis in people diagnosed with CRC.
The search equation used in each database is shown in Appendix 1.

Studies screening and selection process
The titles and abstracts of the records retrieved in the searches were in concomitance with other cancers were excluded.
Disagreements were resolved by consensus, and when necessary, a third reviewer (MM) participated in the discussion until an agreement was reached.

Methodological quality assessment
The methodological quality of the selected studies was evaluated independently by two appraisers following the criteria previously reported (19). Briefly, the evaluated criteria were a) question/aim of research (1 item), b) participants (5 items), c) comparability between groups studied (4 items), d) definition and measurement of the main variables (4 items), e) statistical analysis and confusion (4 items), global assessment of internal validity, f) results (4 items), g) conclusions, external validity and applicability of results (4 items), h) conflict of interest (1 item). Each item was assessed as "very good", "good", "regular", "bad", "not reported", "not apply", and global assessment of study quality, being qualified "high", "medium" and "low".
Two authors (FS and FG) determined a grading system with the scores "very good", "good", "regular", "bad", "not reported", "not apply", or "5", "4", "3", "2", "1", "0", respectively, with the highest total score being 135 and minimum of 27. The score articles within the interval 81 -107 and regular internal validity were categorized with a "median" methodological quality, while a score over 108 represented a "high" methodological quality. However, in the articles that did not apply for the evaluation of criterion "c) comparability between the groups studied", the maximum accumulated score was 115 and the minimum was 23. The score articles within the interval 69-91 and regular internal validity were categorized with a "medium" methodological quality, while a score of over 92 represented P u b l i c a c i ó n a n t i c i p a d a .
a "high" methodological quality.

Data extraction and analysis
The following information was extracted for each study: author, publication year, geographic region in which the study was conducted, sample size, general prevalence of the mutation, mutation prevalence by sex (male and female), mutation (changes in amino acids) and sequencing techniques.
The information was gathered in tables by genes.

Selection and characterization of studies
The selection process is presented in Figure 1. In total, 30 articles were included in this systematic review for full analysis.  to 2299, and from 17 to 1110 participants, respectively. Besides, crosssectional studies were the most frequent type of study (100%) ( Table 1).

Methodological quality assessment
Twenty-three studies scored "high" methodological quality, while 7 studies scored "medium" quality using the critical appraisal of epidemiological cross-sectional studies instrument ( Table 1). The domains with the highest score were "participants" and "results", whereas "statistical analysis and confusion" domain obtained the lowest score.

KRAS
The higher overall prevalence was 54% in a sample of 447 individual and the lowest was 20.5% in a sample of 1323 individuals. The higher prevalence by sex was 66% and 50% for males and females, respectively.
Moreover, 46% of the articles carried out direct sequencing and 23%, next generation sequencing (NGS) to identify mutations in KRAS. The most frequent mutation consisted of an amino acid change from glycine P u b l i c a c i ó n a n t i c i p a d a .
R e v . F a c . M e d .

AXIN2
Only one study (48) was selected with a general prevalence of 21.7% in a sample from Taiwan. The variant p.A603P was the most frequently detected through a next generation sequencing platform (Table 2).

Discussion
The fact that cancer cells contain multiple genetic mutations suggests that the development and progression of tumors could be in part caused by mutagenesis. Additionally, these events can contribute to develop resistance to conventional oncological therapies, such as chemotherapy (49). Currently, scientific evidence shows that therapy against cancer is   Regarding the prevalence of mutations in PIK3CA, an interval of 3.5% to 20.2% was obtained (7,(20)(21)(22)(23)28,29,35,36,37,40,41,45,47). Considering the geographic distribution of these prevalences, in America, 20.2% was the highest prevalence in a studied sample of 2299 individuals in the United States (20). Conversely, the lowest (11%) was reported in a sample of 377 in individuals from the United States and Canada (40).
Regarding sex, 10.4% and 9.7% were the highest prevalences for males and females, respectively (20). In Asian countries, the highest prevalence was reported in India with 16% in a sample of 112 (47), and the lowest was 3.5% in Chinese subjects (7). Similarly, 65.4% and 34.6% were the highest prevalences for male and female, respectively (35). In the Middle East, only one article was selected, reporting a prevalence of 13% (45). Regarding sex, 13% for males and 12% for females, in a sample of 99 individuals from Middle Eastern countries (45). In western Europe, 17.4% was the highest prevalence (21), and regarding sex, the reported prevalences were 16% and 19% for males and females, respectively, in a P u b l i c a c i ó n a n t i c i p a d a .
R e v . F a c . M e d .
sample of 384 Italian individuals (21). In Oceania, one study was chosen, reporting a prevalence of 14% (28). Regarding sex, a prevalence of 50% was found for both sexes in a sample of 757 Australian individuals (28).
The most frequent variant in the PIK3CA leads the change of glutamic acid into lysine (p.Glu545Lys) due to alterations in exon 9. However, a high mutation index was also reported in exon 20, resulting in a change of histidine by leucine (p.His1047Leu) and histidine by arginine (p.His1047Arg).
BRAF intervenes in the proliferation, differentiation, and cell apoptosis pathways (75) According to sex, the highest prevalence for males was 3.3%, and 5.1% for females (45). In Western Europe, the highest prevalence reported P u b l i c a c i ó n a n t i c i p a d a .
R e v . F a c . M e d . The differences between general prevalences of mutations and prevalences according to sex in the same geographical area could be attributed to differences in the size of the samples studied and to the sensitivity of the molecular techniques used, which have been shown to influence the frequency mutation detection (81,82). Other factors that could have an influence are the quality and quantity of the DNA obtained, the heterogeneity in the tumor and possible environmental exposures not controlled or unknown by the authors (83,84).
Although the relationship between the prevalence of the studied genetic mutations and sex was not statistically significant in most of the articles reported in the present study, this relationship was evaluated because it has been reported that genetic associations with sex could provide information about the pathogenesis of diseases (85), and also because of the established differences in susceptibility and incidence of cancer among men and women around the world (86), attributable to environmental causes and genetic differences (87,88). Therefore, it can be thought that differences in the prevalence of mutations between males and females could condition the development of colorectal cancer more frequently in individuals of a specific sex.
The incidence and mortality of colorectal cancer has increased over the past 10 years (89), therefore, the need to identify and implement early diagnostic strategies for this malignant pathology has increased, including the analysis of molecular and clinical biomarkers. Recently, a gene that could be considered a molecular biomarker, AXIN2, has been studied and a possible association of variants in this gene, as well as the phenotype of colorectal cancer and dental agenesis have been reported (16). The P u b l i c a c i ó n a n t i c i p a d a .
R e v . F a c . M e d .
present systematic review aims to report the prevalence of variants in AXIN2 in individuals with primary and/or metastatic colorectal cancer around the world. However, only one study was identified. In addition, due to the knowledge of the possible relations of the AXIN2 gene with both phenotypes, we attempted to identify if some of the higher prevalent genes in colorectal cancer such as KRAS, PIK3CA and BRAF were also related to dental agenesis, however, no studies were identified that reported such an association.
AXIN2 is known for its tumor suppressing activity by negatively regulating the WNT pathway by the intracellular degradation of β-catenine (90,91).
In mice, AXIN2 is expressed during odontogenesis in dental mesenchyme, enamel knot, dental papilla and mesenchymal odontoblast. It is reasonable to hypothesize that a loss in function of this gene could affect the development of molars and incisors, leading to dental agenesis (92). In addition, there is evidence that the expression of AXIN2 in colorectal tissue

Conclusions
Our findings suggest that there is a wide and diverse distribution of KRAS, PIK3CA and BRAF gene mutations in individuals with CRC worldwide; KRAS being the most prevalent. Moreover, this study highlights that there is no evidence on the association between dental agenesis and KRAS, PIK3CA and BRAF germline gene mutations in people with CRC. AXIN2 is the unique gene in which the association with both phenotypes has been well stated, but population studies focused on AXIN2 mutations prevalence are limited.

Conflict of interest:
The authors declare no conflict of interest.
R e v . F a c . M e d .