The inclusion criteria for the present study were as follows: (1) epidemiological studies (cross-sectional, case-control, cohort, and clinical trials) that evaluated the association between genetic polymorphisms and EARR in patients treated with fixed orthodontic appliances; (2) studies that used polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis to evaluate genetic polymorphisms, and (3) those that documented pre- and post-orthodontic treatment outcomes. No restrictions were imposed on the type of tooth, EARR measurement criteria, length of treatment, tooth extraction, or age of the participant.

The exclusion criteria were as follows: (1) reviews or systematic reviews, case reports, case series, descriptive studies, opinion articles, abstracts, letters to the editor, laboratory and/or animal studies; (2) studies involving individuals with cleft lip, palate, or both; (3) studies involving individuals with other craniofacial deformities/syndromes; and (4) studies on unrelated topics, such as those not involving genetics or on EARR not caused by orthodontic treatment. Three review articles were excluded from the analysis because they did not include the latest findings or were not systematically reviewed, which is essential for the validity and relevance of a systematic review.

Electronic searches of the following databases were conducted in March 2023, and the search period was restricted to the last decade (up to May 2013): Medline through PubMed (http://www.ncbi.nlm.nih.gov/pubmed/), Cochrane Library (http://www.cochrane.org/), Web of Science (http://www.isiknowledge.com/), Controlled-Trial Database (http://controlled-trial.com), Clinical Trials - US National Institutes of Health (http://www.clinicaltrials.gov), National Institute for Health and Clinical Excellence (http://www.nice.org.uk) and Virtual Health Library (Bireme, Latin America Literature; www.bireme.br). There were no restrictions on the publication date or language.

The following search strategy was used on Web of Science, Medline, and Cochrane Library: (root resorption [Mesh] OR tooth resorption [Mesh], OR root resorption orthodontic, OR external root resorption*, OR tooth movement [Mesh], OR external resorption*, OR apical root resorption*, OR orthodontic treatment*, OR EARR) AND (genetic polymorphism [Mesh], OR single nucleotide polymorphism [Mesh], OR mutation [Mesh], OR genetic*, OR genetic association studies [Mesh], OR genetic research [Mesh], OR genetic variation [Mesh], OR genetic testing [Mesh], OR genes [Mesh]). For Controlled-Trial Database, Clinical Trials–US National Institutes of Health, National Institute for Health and Clinical Excellence–UK National Institute for Health, and Virtual Health Library, the search was performed using combined uniterms ‘polymorphism’ and ‘apical root resorption.’ Grey literature was searched in the following databases for registration of epidemiological studies: Controlled-Trial Database, Clinical Trials–US National Institutes of Health, National Institute for Health and Clinical Excellence–UK National Institute for Health. A manual search was performed of the reference lists of the included studies.

Our electronic search yielded 408 abstracts and titles (Figure 1). We inputted these references into the Reference Manager Software^® (Reference Manager, Thomson Reuters, version 12.0.3; Toronto, ON, Canada). After removing duplicates, 348 abstracts and titles were independently reviewed and analyzed by two calibrated reviewers (ILAL, JDC). The calibration process involved determination of the inclusion and exclusion criteria. The reviewers thoroughly discussed these criteria, which were applied to 20% of the retrieved studies. This exercise was repeated until adequate agreement was obtained based on Cohen’s kappa coefficient (kappa: 0.74).

Of the 348 abstracts reviewed, 30 were selected for full-text analysis. To include all related articles, we contacted one author to obtain the full texts of articles that were not accessible online. After the full-text analysis, 21 studies were included in this systematic review (Figure 1).

Data were independently extracted by the same two reviewers (ILAL, JDC). Inter-examiner conflicts were resolved by consensus. The extraction was performed according to the manufacturer’s instructions.

The forms used for data extraction included the year of publication, location, study design, population, sample size, participants’ age, teeth evaluated, EARR quantification criteria, main results, conclusions, and types of gene polymorphisms (Table 1). Most of the described gene polymorphisms were of the single nucleotide polymorphism (SNP) type, identified by the same “rs” (reference SNP) number. However, one polymorphism was of the variable number tandem repeat (VNTR) type.⁸ Data on genotype distribution and allelic frequency were also included in this systematic review (Table 2).

As different types of polymorphisms have been described, a qualitative analysis was conducted. However, the data were not sufficiently comparable for meta-analysis.

Twenty-one studies were included in this systematic review: 15 case-control studies and six cross-sectional studies (Figure 1). In general, studies recruited cases from university services,^{1,8,9,11,14-17,21-23,28,29} with seven studies also including patients from private practice,^{1,9,19-21,24,30} and four studies recruited patients exclusively from private practice.^11,18,25,27 The sample sizes ranged from 54^14,15 to 480²⁹ individuals. The age of participants ranged from 8.4 to 55.4 years.²⁷

The outcomes were based on tooth dimensions (root and crown lengths) measured directly on radiographs obtained before and after treatment. Lateral cephalometric,^{1,9,14,16,18,22,25,26,28,30} panoramic,^{1,9,14,16,18-22,25,26,28-30} periapical,^11,24 and occlusal radiographs were used, in addition to cone beam computed tomography (CBCT).²³ Control groups comprised the following: (1) non-exposed individuals; (2) patients treated with fixed orthodontic appliances; and (3) patients with a history of EARR of < 2 mm. Three categories of control groups were seen in this review: (1) orthodontically treated patients with an EARR of < 2 mm;^{11-18,21-26,28} (2) orthodontically treated patients from the same family (parents or siblings) with an EARR of < 2 mm^1,9 and (3) orthodontically treated patients with resorption lower than the apical third.²² Furthermore, most studies included in this systematic review performed PCR-RFLP analysis for polymorphisms.^{1,9,11-18,22-25,28,30}

Different teeth were evaluated among the articles retrieved: maxillary central incisors,^{1,9,11,14,15,18-21,23-30} lateral maxillary incisors,^{11,14,18-21,26-29} maxillary canine,^19,20 maxillary premolars,^14,16,22 mandibular central incisors and mandibular molars,^1,9,15,18,25 and mandibular first molars.^18,25,30

Not all studies described the same variables. There was considerable clinical and methodological heterogeneity as studies have described different types of polymorphisms and SNPs. One study specifically examined tooth-based samples rather than patient-based samples; in this study, IL1B (rs1143634) was not associated with EARR (P > 0.05).¹⁵ However, another study found a significant association between IL1B (rs1143634) and EARR (95% confidence interval [CI]: 1.9–21.2) after orthodontic treatment.¹ Three other studies examined the association between EARR, gene polymorphisms, and OTM in endodontically treated teeth.^14,16,22 They demonstrated that it was not IL1A (rs1800587) but variations in IL1B SNPs that might contribute to the occurrence of EARR.^14,16 Furthermore, variants of allele 1 of IL1RN (rs419598) are associated with an increased risk of post-orthodontic EARR in endodontically treated teeth.²² In genetics, an SNP is a variation at a single position in the DNA that can influence disease and drug response. A ‘CC’ genotype implies that both alleles at a gene are cytosine, while a ‘T’ genotype has at least one thymine allele, and a ‘TT’ genotype implies that both alleles are thymine. Homozygosity refers to the presence of identical alleles (either CC or TT) in a specific gene. A haplotype is a set of inherited genes that can provide insight into an individual’s genetic predispositions. Microsatellite polymorphisms involve variations in the short DNA sequences used for genetic testing. Linkage disequilibrium describes how certain alleles at different loci are associated more frequently than expected by chance, aiding genetic mapping and disease studies. Only homozygous individuals for the T allele of IL1RN were more prone to orthodontic-induced EARR.²¹ IL1RN and P2RXT analyses have indicated that the presence of SNPs in these genes can predispose individuals to EARR.¹⁸ IL1RN VNTR polymorphism 86-bp in two introns was also related to EARR, specifically in girls.⁸

The STAG2 gene, located on the X chromosome, was found to be associated with an increased risk of EARR, particularly restricted to the male sex.⁹ Another study demonstrated that IRAK1 polymorphism is a protective variant of EARR.²⁰ For genes encoding OPG, RANK, IL1, and IL1RN, the effect changed from protective to deleterious depending on the duration of treatment and the age of the patient.¹⁹

Regarding RANKL polymorphisms, no significant association was found with EARR. Among RANK polymorphisms, only rs12455775 was associated with EARR. Regarding OPG polymorphisms, an association was found between EARR and rs3102724, rs2875845, rs1032128, and rs3102728.²⁴

TNSALP, TNFα, and TNFRSF11A polymorphisms were also investigated. It was concluded that TNFRSF11A (rs1805034) is associated with EARR (P = 0.02), but TNSALP and TNFα are not.⁹ Additionally, SPP1 gene polymorphisms (rs9138 and rs11730582) were related to genetic susceptibility of EARR: rs9138 CC vs. CA/AA (odds ratio: 4.10; 95% CI: 1.03–16.35); rs9138 AA vs. CA/CC (odds ratio: 0.20; 95% CI: 0.05–0.81); rs11730582 CC vs. CT/TT (odds ratio: 11.68; 95% CI: 1.12–121.06); and rs9138 CT vs. CC/TT (odds ratio: 0.035; 95% CI: 0.062–0.90).²⁶ One study observed that the presence of a specific P2RX7 (rs208294) was significantly associated with EARR (P = 0.0028) (Table 1).^27,28

According to Ciurla et al.²⁵ the effect of SNPs in CASP1, TNFRSF11B, and IL 6 genes on EARR was not confirmed however, Guo et al.²³ showed that IL 6 increased the risk of EARR.

Marañón-Vásquez et al.³⁰ showed that individuals with the AA genotype of VDR rs2228570 had a higher risk of EARR than those with the AG and GG genotypes. In addition, the incidence of EARR in heterozygous rs2228570 was lower than that in homozygotes. In addition, participants with the CCG haplotype (rs1544410-rs7975232-rs731236) in the VDR showed a lower EARR than those who did not carry this haplotype. Regarding CYP27B1 rs4646536, participants with at least one G allele showed a lower risk of EARR than those with homozygote AA. Although these results were not significant after multiple test adjustments, potential associations may still be suggested.³⁰

The present review presents no bias related to the language or year of publication. The search strategy identified only articles published in English^1,9,11,14-30 between 2003 and 2023. Although there is a possibility of bias due to the small sample sizes of some reports, this is a relatively new field of investigation, and the retrieved studies presented high methodological quality.

All case-control studies presented an adequate definition of EARR through radiographic evaluation.^{8,11,14-17,22,26,27} The cases analyzed were from the same family or clinic, and the study participants were controlled for age.^{8,11,14-17,22,26,27} The control group comprised orthodontically treated patients with an EARR of < 2 mm,^11-17,22 orthodontically treated patients from the same family (parents or siblings) with an EARR of < 2 mm, and orthodontically treated patients with resorption limited to the apical third.²⁷ Individuals presenting with EARR of ≥ 2 mm^11,14-17 or resorption equal to or more than the apical third²⁷ were considered as cases. In addition, the same method of ascertainment (PCR-RFLP analysis) was employed for both cases and controls.

All cross-sectional studies in this systematic review included clinically diagnosed cases of EARR.^1,9 Secure records with no indication of EARR at the start of the study were presented and radiographs of the patients after the end of treatment were used to evaluate the presence of EARR. Therefore, the follow-up was considered adequate. The details of blind processes have rarely been reported.

In 1975, EARR was described as familial,³¹ and in 1997, a hypothesis of genetic influence on EARR was proposed using a sib-pair model with an estimated heritability of approximately 70%.³² The current understanding is that EARR has a multifactorial etiology influenced by a combination of environmental and host factors. Genetic predisposition to an exaggerated inflammatory response may contribute to EARR in some patients during OTM. Accordingly, many studies have attempted to demonstrate the association between gene polymorphisms involved in inflammatory responses and EARR. In this regard, the IL1 gene is the most frequently described, as it may be an early critical mediator in inflammatory processes.³³ IL1 plays a critical role in the inflammatory response related to orthodontic treatment; it recruits large immune cells and affects transmigration to the tooth movement area. Moreover, IL1 directly or indirectly amplifies osteoclast differentiation,³⁴ affecting root resorption. Elevated IL1β levels in gingival crevicular fluid and gingival tissues during OTM further support its role in tooth resorption.³⁵

Some studies have investigated IL1 polymorphisms in the context of EARR; however, conflicting results have been reported, mainly associated with ethnic populations. These discrepancies may arise due to the geographic restrictions of each study.¹⁵ Tomoyasu et al.¹⁵ demonstrated that IL1B (rs1143634) was not associated with a predisposition to EARR. Accordingly, a meta-analysis revealed that IL1B polymorphisms were not found to be associated with the susceptibility to EARR.³⁶ In contrast, other studies found an association between IL1 polymorphisms and EARR following OTM.^1,17 Studies report that the presence of haplotype +3,954 and –889 with “T” allele increases the production of IL1B and IL1A, respectively,^8,15,17 suggesting an association with EARR.^1,11,14 A link between EARR and IL1B SNPs has also been described.²⁵ On the contrary, other studies found that the presence of “CC” genotype increase the risk of EARR.^1,11,27 This result agrees with a previous study showing that the absence of IL1B gene causes root resorption in mouse molar.¹³

IL1A (rs1800587) and IL1B (rs1143634) polymorphisms were analyzed in four case-control studies.^8,14,18,21 Linhartova et al.⁸ found no association between IL1 (rs1800587 and rs1143634) and EARR, while Iglesias-Linares et al.¹⁴ reported a relationship between IL1B (rs1143634) and EARR. Although both studies used similar methods, the number and age of the participants could explain these differences. The study by Linhartova et al.⁸ included 106 participants (mean age: 15.0 ± 4.1 years in EARR group and 15.2 ± 5.3 years in the control group), whereas Iglesias-Linares et al.¹⁴ included 54 patients (mean age: 22.9 ± 6.3 years). Silva et al.¹⁹ showed that for IL1, the effect changed from protective to deleterious depending on the duration of treatment and the age of the patient.

Perrier et al.,³⁷ demonstrated that allele 2 (two repeats) of IL1RN (VNTR 86bp) polymorphism downregulates interleukin-1 receptor antagonist (IL1Ra) production in saliva, increasing IL1 expression. One study that evaluated the role of IL1RN polymorphism VNTR 86bp in EARR⁸ showed that the allele 2 variant was specifically related to EARR and OTM in girls.⁸ Individuals that are homozygous for the T allele of IL1RN rs419598 were more prone to EARR during orthodontic treatment.²¹ There is further evidence that “TT” genotype of IL1RN rs419598 increases the risk of EARR in endodontically treated teeth.²² In addition, IL1RN rs419598 and VNTR 86bp haplotypes are in linkage disequilibrium, affecting IL1Ra and IL1B production.²² Another study¹⁸ concluded that the presence of IL1RN SNP (rs419598) may predispose individuals to EARR.

Al-Qawasmi et al.,¹ Bastos Lages et al.,¹¹ and Iglesias-Linares et al.^14,16,22 suggested a possible link between IL1A (rs1800587), IL1B (rs1143634), and EARR. Moreover, Ciurla et al.²⁵ showed that IL1B increased the odds of developing EARR four-fold. Although Al-Qawasmi et al.¹ and Bastos Lages et al.¹¹ used similar methods, the ethnic diversity of the samples may have influenced the results. They found that the “CC” genotype increases the risk of EARR, although the presence of “T” allele 4-fold increased the level of IL1B.⁹ In contrast, Iglesias-Linares et al.^14,16,22 evaluated endodontically treated teeth, which may have been a confounding factor.

TNFSF11A, involved in RANK encoding, plays a crucial role in osteoclastogenesis, a major mechanism involved in mineral matrix destruction. An important association between TNFRSF11A rs1805034 microsatellite polymorphism (D18S64) and EARR has been discussed based on linkage disequilibrium between them.⁹ Although TNSALP is important for cementum formation and mineralization and TNFα is critical in bone remodeling, no evidence of association or linkage disequilibrium was found between them and EARR.⁹

SPP1, which regulates the release of pro- and anti-inflammatory mediators (mainly IL1B) by human monocytes, has also been studied. Iglesias-Linares et al.²⁶ found that “C” allele of SPP1 gene polymorphisms “rs9138 and rs11730582” increase the risk of EARR, and they are associated with elevated osteonectin expression.

The purinergic receptor (P2X) is a member of the cationic channel family, in which receptor P2RX7 plays an important role in the immune response, mainly in bone metabolism, inflammation (upstream of IL1B maturation and release), cell proliferation, and the central nervous systems.³⁸ Association between EARR and P2RX7 gene polymorphisms “rs208294 and rs1718119” have been reported; the “T” allele was associated with a protective role while “G” allele increased the risk for EARR.²⁷ Borilova Linhartova et al.²⁸ also found that variability in P2RX7 may be an important factor contributing to the etiopathogenesis of EARR. Furthermore, an association of caspase 1—involved with IL1B maturation and release—with EARR was investigated; the “T” allele of rs530537 increases the risk of EARR.

Altogether, existing evidence suggests that IL1 is associated with EARR, however, pinpointing the specific IL1 allele (IL1A, IL1B, or IL1RN) implicated in EARR remains challenging. Several factors contribute to this uncertainty. First, the retrieved studies exhibited significant heterogeneity in gene polymorphism analyses. Second, variations in radiographic techniques—ranging from periapical and occlusal radiographs to CBCT—yield varying measurement accuracies of EARR. This could represent bias in the diagnosis of EARR. Additionally, the lack of important data related to the sample, such as sex and age, are often absent.

Moreover, isolated studies have explored other candidate genes. These include TNFRSF11A, TNFRSF11B, SPP1, IL6, STAG2, VDR, IRAK1, and IL17. However, further research is needed to confirm the possible association of these genes with EARR and OTM.