Association between tooth agenesis and cancer: a systematic review

Abstract The congenital absence of multiple teeth may share the same genetic background of the development of some types of cancer. Objective: This systematic review aimed to investigate the possible association between dental agenesis and cancer, and the perspective of agenesis as an early predictor for cancer risk. Methodology: The electronic databases PubMed, Scopus, Web of Science, Cochrane Library, LILACS, and OpenGrey were searched and the risk of bias was evaluated using the Newcastle-Ottawa tool. The GRADE tool was used to evaluate the certainty of the evidence. Results: Six studies met the eligibility criteria. A positive co-occurrence between ovarian cancer and hypodontia was found in two articles. Three studies evaluated the association between dental agenesis and colorectal cancer and only one showed common genes for these conditions. One paper found individuals with hypodontia had a higher risk of family history of cancer. Five studies had a fair quality and one a good quality. The certainty of evidence was classified as very low. Conclusion: Notwithstanding the limited scientific evidence, there may be a possible association between dental agenesis and cancer due to genes involved in both conditions. Agenesis of multiple teeth could be an early indicator of cancer risk. Nevertheless, studies with a better level of evidence are needed to confirm this possible association.

mutations in these genes may be connected with many types of cancers. 16 The link between dental agenesis and cancer may be elucidated by three factors: (1) there are genes involved in odontogenesis that are present in tumor tissues or cells; 17,18 (2) nucleotide changes on some genes are related with both odontogenesis and cancer 19 and the mutations appear to disturb odontogenesis early in life and later contribute to the carcinogenesis; (3) according to epigenetics, the aberrant methylation of these genes was observed in neoplasm samples. 20 There is still divergence in the literature about this relationship. Some studies showed an association between dental agenesis and cancer, 21,22 while others do not. 23,24 There is a great clinical relevance in this issue since the absence of multiple teeth may be an indicator of cancer. Therefore, this systematic review aims to verify the connection between dental agenesis and cancer, considering a single tooth agenesis or even oligodontia, and the possibility that agenesis is an early indicator for cancer risk.

Methodology
This review was registered at PROSPERO database (CRD42019129901) and performed according to PRISMA guidelines. 25 The process was performed separately by two reviewers. A third reviewer was consulted when there was no agreement between the two reviewers.

Eligibility criteria
The following eligibility criteria were adopted in this systematic review in accordance to the PECOS format:

Information sources
The databases PubMed, Scopus, Web of Science, Cochrane Library, OpenGrey, and LILACS were searched between the 15 th and 21 st of January and the alerts were followed up until the 5 th of September.
A manual search was carried out in the reference list of the included studies for eventual relevant article missed during the searches. No restriction on language or publication date was used.

Search strategy and study selection
The search strategy was created using words associated with the PECOS strategy and these words were combined using Boolean operators. The search strategy for each database is presented in Figure 1.
All relevant references have been imported into the software Endnote (x9 version, Clarivate Analytics, Philadelphia, PA, USA). After duplicate removal, titles, and abstracts were evaluated considering the selection criteria. The included studies were accessed by full text read for further assessment and data extraction.

Study selection
The electronic screening found 827 articles: 273 from PubMed, 367 from Scopus, 146 from Web of Science, two from Cochrane Library, 39 from LILACS, and zero from OpenGrey. After removing duplicates studies, 543 articles were identified. After the authors performed title and abstract screening, 20 articles were assessed by full text. Among them, 14 were excluded for the reasons shown in Figure 2. Finally, six studies were selected for qualitative analysis of risk of bias ( Figure 3).

Characteristics of included articles
The characteristics of the six included articles are presented in Table 1. They were observational and case-control studies. 21,22,24,[28][29][30] Two articles investigated the association between tooth agenesis and ovarian cancer, 21,22 whereas three assessed the interrelation of colorectal cancer with tooth agenesis. 24,29,30 One article investigated the co-occurrence of dental agenesis and family history of cancer. 28 One of the studies that investigated the relationship between agenesis and colorectal cancer also identified family history of cancer; nonetheless, it was not statistically tested or discussed in the manuscript. 30 Therefore, we decided to perform an Odds Ratio to assess this association.
A considerable difference was found in relation to the sample sizes. The sample sizes of the control groups ranged from 44 24 to 4188, 29 while the sample sizes for the case groups ranged from 6 24 to 236. 29 The mean age was only reported by one article. 30 The diagnosis of dental agenesis was made through clinical, 21,22,24,28,30 radiographic examination, 21,22,24,28,30 and a selfreport questionnaire. 24,29 The tooth with the highest percentage of congenital absence were upper lateral incisor, 21,22,24,28 second upper premolars, 21,22,24 second lower premolars, 22,24 and lower central incisors. 22 The diagnosis of cancer was not detailed in the studies, although they report patients were diagnosed and recruited from cancer treatment centers. 21,22,24,29 The family history of cancer was evaluated in the included studies through questionnaires 28 or self-reports. 30 Two studies 28,30 evaluated the relation through genes analysis, which were: AXIN2, FGF3, FGF10, FGFR2, 28 ATF1, DUSP10, CASC8. 30

Results of individual studies
Two studies detected an association between the congenital absence of tooth and ovarian cancer. 21,22 Other two articles did not report an association between colorectal cancer and dental agenesis, 24,29 whereas one showed common genes for both conditions: ATF1, DUSP10, and CASC8. 30 One study found subjects with dental agenesis had a major chance of family history of cancer and associations with AXIN2, FGF3, FGF10, and FGFR2 genes. 28

Synthesis of results
It was not possible to perform a meta-analysis because of the low number of articles investigating the analyzed outcomes. However, an odds ratio was performed for each study individually and for each type of cancer or family history of cancer. It was revealed a statistically significant association between dental agenesis and ovarian cancer, with a chance of a patient with ovarian cancer being diagnosed with tooth agenesis 6.43 higher. No statistically significant association was observed between agenesis and colorectal cancer, which is corroborated by the p-value and the 95% confidence interval. Finally, a statistically significant association was also noticed between family history of cancer and dental agenesis and the results shows a chance 2.71 times greater of the co-occurrence of these two conditions ( Table 2).

Risk of bias assessment
The quality of five studies was classified as fair, 22,24,28-30 and one study as good (Table 3). 21 Limitations were found in the main domains evaluated. The domain "selection of study groups" exhibited deficiencies such as inadequate case definition; 29 poor representativeness of the cases 21-24.28-30 , and lack of information on the selection of controls. [22][23][24]29 The deficiency in the representativeness of the cases was characterized by no description of the recruitment location of control subjects. The domain "comparability of groups" presented limitations in the item "no control of important confounding factors (e.g. gender, age)." 24 Two articles showed inaccurate outcome assessments due to evaluation by self-reporting. 28,30 Level of evidence The GRADE evaluation found a very low certainty of evidence for the three outcomes assessed ( Figure   4). This can be associated with the study design and risk of bias of included articles.
Authors, year, location and type of study    In two included studies, the authors verified the association between dental agenesis and ovarian cancer. 21,22 Chalothorn, et al. 21 (2008) used dental and medical records to assess family history of cancer and tooth agenesis. The dental examination was conducted to detect clinically hypodontia or any phenotype involved with this congenital dysfunction, like microdontia and agenesis. As a result, the authors found an increased prevalence of tooth agenesis in patients with epithelial ovarian cancer. In another study, conducted by Fekonja, Čretnik, and Takač 22 (2014) women diagnosed with epithelial ovarian cancer were evaluated through clinical examination and panoramic radiography to confirm the diagnosis of hypodontia. The results showed a possible association between the two conditions. The OR confirmed a significant association between ovarian cancer and tooth agenesis. The result indicated the chance of a patient with ovarian cancer be diagnosed with a pattern of dental agenesis is 6.43 times greater (Table 2).
The findings of these two studies 21,22 differ from other results in the literature that point to independent causation of these conditions. 23 The authors analyzed the ovarian cancer sample in a cohort study and do not prove that the two conditions are independent from each other, but a genetic connection between them needs more epidemiological studies and molecular analysis to be confirmed. The absence of an adequate control group definition did not allow its inclusion in this systematic review. 23 Regarding the association between dental agenesis and colorectal cancer, one 29 of the three included articles 24,29,30 used a questionnaire to self-report information on congenitally missing teeth. This was a limitation since a dental clinician did not examine the participants, and therefore justified the fair quality rating. The authors concluded the study did not provide scientific evidence strong enough to prove the predisposition of dental agenesis among colorectal cancer patients. The second study 24 which verified this association agrees with the results obtained by Lindor, et al. 29 (2014). The patients with colorectal cancer revealed an increased prevalence of dental agenesis when confronted to patients without history of this cancer, but it was not statistically significant.
Our OR results, as well, demonstrated no statistically significant association between the two conditions, as demonstrated by the p-value and 95% confidence interval ( Table 2).
The major contrast from this study 24 to the Lindor, et al. 's 29 (2014) was the clinical and radiographic analysis of tooth agenesis, which was performed by the same dentist to avoid interexaminer bias, in the first, 24 compared to a self-reported questionnaire of hypodontia in the second. 29 In the third included article, 30 (Table 2).
Dental agenesis is a failure in the odontogenesis process that occurs at the beginning of tooth morphogenesis. 37 It is well known the etiology is related with genetic and environmental factors, 38 and it may be part of a phenotypic expression of a syndrome or occur in isolation. 37 The genes that are often associated with non-syndromic dental agenesis are AXIN2, MSX1, PAX9, EDA, and WNT10. 39,40 It has been reported the association of AXIN2 gene with colorectal cancer. 19,33,41 However, this relationship has not been demonstrated yet, 42,43 which corroborates the results of this systematic review. In consequence, the polymorphism in AXIN2 gene may be considered a biological risk marker for predisposition and prognosis of colorectal cancer. 41 A possible genetic relationship between dental agenesis and colorectal cancer has also been studied by Williams, et al. 30 (2018) which reported the ATF1, DUSP10 and CASC8 genes may be related to colorectal cancer and to odontogenesis.
The hypothesis of tooth agenesis as a risk factor can be considered when evaluated the association with ovarian cancer, helping in its the early detection. In this case, however, it was not found an inherent gene that might be the causal factor responsible for the connection between the two conditions, as recently reported in the literature. 44 The genes BRCA1 e BRCA2 are the strongest recognized genetic risk factors for epithelial ovarian cancer, 45 although some studies show an association with the AXIN2 gene in several cancers, including the ovarian one. 30,46 The epidemiology of the epithelial ovarian cancer requires attention, because it is considered the fifth most common cause of cancer in women and the fourth leading cause of cancer death, 47,48 with a prognosis of approximately 18 months for women with an advanced stage, and 40-50% of overall survival for all ovarian cancer at ten years. 49 It is important the attempt of early establish the co-occurrence between the epithelial ovarian cancer and the dental agenesis as a risk factor, mainly because of the aggressiveness of this type of cancer, that is considered malignantly fatal and silent, therefore to the difficult of diagnosis, 21,50 as the major symptoms are not specific 51 and as the lack of effective screening markers. 46 Some hypothesis that would be useful in the identification of ovarian cancer are to check the family history of this cancer 52