1 INTRODUCTION

External apical root resorption (EARR) is a multifactorial disorder with adverse clinical outcomes in orthodontic practices often resulting in significant root shortening.¹ Minor surface resorption on the root can be detected histologically as microscopic lacunae in the absence of EARR radiographic evidence.^{2, 3} This mild resorption can be repaired by cementoblasts. More aggressive progression and exposure of dentin increases the likelihood of permanent root resorption resulting in EARR. Although root resorption may occur in individuals who have never undergone orthodontic treatments, the incidence among treated individuals is quite high, about one third of patients show signs of EARR.^{1, 4}

A number of risk factors have been revealed related to EARR (e.g. genetics, tooth type and root morphology, bone structure, orthodontic therapy, periodontitis, trauma and oral habits); however, the evidence is not clear.^{3, 5} Individuals vary in their susceptibility to EARR with Caucasian and Hispanic patients experiencing significantly more root resorption when compared to Asian orthodontic patients.^{6, 7} This suggests that ethnic differences play an important role in individual's susceptibility to EARR. It is estimated that genetic factors can account for approximately 50%–60% of EARR variations.⁸ Analysis of sib-paired models reported high heritability estimate of 76% for maxillary central incisors.⁹ Furthermore, monozygotic twins have shown concordance estimates for EARR that were approximately twice those seen in dizygotic twins.¹⁰ A growing body of evidence suggests that variations in a number of possible genes contribute to the EARR phenotype. Many of these genes encode for proteins involved in the ATP/P2XR7/IL-1 inflammation modulation pathway.⁸

The purinergic receptor-P2X, ligand-gated ion channel 7 (P2RX7) is activated by the ATP released during compressive orthodontic forces.^11-13 This receptor is found to be involved in both osteoclasts, enabling their apoptosis, and osteoblasts, promoting their normal function. Murine studies have shown that when this receptor is inactivated, root resorption under orthodontic forces is increased.¹⁴ In humans, 3 functional single nucleotide polymorphisms (SNPs) in the P2RX7 gene have been investigated in relation to EARR: rs1718119, rs208294 and rs2230912.^{6, 15-22} Of these the +489 C allele (rs208294) and the +1068G allele (rs1718119) were associated with root resorption.

P2RX7 also promotes the release of the interleukin-1 cytokines (IL-1α and IL-1β) by immune cells.¹⁵ While IL-1α (encoded by the IL-1A gene) is an active pro-inflammatory cytokine released by macrophages during the early stages of inflammation, IL-1β (encoded by the IL-1B gene) is inactive and requires cleavage by caspase 1 (CASP1) to become active and contribute to the inflammatory process that results in removal of the necrotic tissue and eventually leads to root resorption.^{6, 13, 23, 24} However, the action of these cytokines is restrained by the interleukin-1 receptor antagonist (IL-1RN), which bind to the same receptor as IL-1α and IL-1β. The high affinity binding of IL-1RN to the IL-1R1 receptor inhibits the IL-1 signalling activity. On the contrary, stimulation of IL-1R1 activates the IL-1 receptor-associated kinase 1 (IRAK1), a key modulator of the subsequent signal transduction pathways.²⁵

Several SNPs in these genes have been extensively studied in orthodontic patients by various groups.^{6, 12, 15, 17-22, 26-30} However, the results are inconclusive and their association to EARR varies with the population studied.^{13, 31, 32} In some patients, though, the CC genotype of IL-1B rs1143634TT and the TT genotypes of IL-1A and IL-1RN polymorphisms (rs1800587 and rs419598, respectively) have been linked to an increase in root resorption.^{17, 18, 20-22, 33} On the contrary, the CC genotype of IRAK1 polymorphism rs1059703, has been suggested as protective for EARR; although, this variant was investigated only by 1 group.³⁰

In this study, we examined the effect that specific variants within the P2RX7, IL-1A, IL-1B, IL-1RN, CASP1 and IRAK1 genes, involved in ATP/P2XR7/IL-1 pathway, have on the risk of developing EARR in orthodontic patients in X. We also evaluated how other selected patient-related, as well as treatment-related factors may contribute to root resorption in these patients.

2 MATERIALS AND METHODS

3 RESULTS

The patients included in this study ranged in age from 9.5 to 51.2 years, with a mean age of 16.2 years. The sample consisted of Caucasians (74%), African Americans (16%), Hispanics (6%), Asians (3%) and South Asians (2%). Of the 195 patients examined, 53 of them were diagnosed with EARR. The interrater agreement on the EARR diagnosis of maxillary incisors was almost perfect with a kappa value of 0.85 and 95% confidence interval of 0.778–1.000.

The demographics for both EARR and the control group are shown in Table 2. The EARR patients, on average, started orthodontic treatment at an earlier age (mean of 14.6 years) compared with the control group (mean of 16.8). However, this was not statistically significant, and the median age for the EARR and control patients were similar (13.1 and 13.3 years, respectively). Although not statistically significant, root resorption seemed to be more prevalent in females (58.5% compared to 54.2% in controls) than males. Furthermore, we found no statistical significance in the relationship between race/ethnicity and EARR diagnosis. There were also no statistically significant associations between EARR and other patient-related factors such as overall dental and skeletal classification, ANB, U1-SN and overjet (Table 3). While we included patients that had been in treatments for at least 9 months, only 3.8% of the EARR group and 7.0% of the controls had received orthodontic treatment for less than 12 months (P > .05). Furthermore, the average treatment times were comparable, 19.8 months and 19 months for EARR and controls, respectively (P > .05). On the other hand, the extraction treatment type displayed significant association with root resorption (P = .01) in the orthodontic patients in this study.

However, as the majority of our sample was comprised of Caucasians and there was not a good representation of the other races/ethnicities, a more in-depth study of the risk factors (including genetic variations) associated with root resorption could only be completed for Caucasians. Of the 144 Caucasian patients analysed, 43 of them displayed root resorption with orthodontic treatment (Table 4). With this small sample size, an 80% power with α = .05 could be obtained to detect an EARR difference of 15% if the proportion of control is 53%. If the proportion of control is 94%, then, the 80% power will detect the difference of 17% in EARR. Furthermore, about 46% of these patients had completed their orthodontic treatment at time of the evaluation, while only 34% of controls had finished their treatment. However, this was not statistically significant (P > .05).

Similar to the data for the overall population, we identified no statistically significant association between EARR diagnosis in Caucasians and their sex, age, skeletal relationship, ANB, U1-SN, overjet and treatment time (P > .05). Furthermore, extractions continued to be a significant risk factor for developing root resorption (P = .02). We found that the odds ratio for root resorption in patients with extraction was 4.19 (95% confidence interval: 1.39–12.65) times the odds for non-extraction (P = .01; Table 5). However, unlike the whole population sample, in Caucasians, dental classification was significantly associated with EARR (P = .03) and we found that patients with Class II or III malocclusion had an increased risk for root resorption (odds ratio of 2.55; P = .02) compared to class I patients (Tables 4 and 5). We also observed that for every one-year increase in starting age for orthodontic treatment, the odds to developing EARR decreased by 8%; although, this was not statistically significant (P = .09; with 95% confidence interval of 0.83–1.101). Finally, logistic regression found no significant interactions between any of the variables mentioned above (P > .05).

The genotype and allele frequencies of the 9 SNPs analysed are shown in Table 6. The IRAK1 rs1059703 genotype distribution in females, as well as the frequencies of all other SNPs followed the Hardy–Weinberg equilibrium. Since IRAK1 is located in chromosome X, males could not be included in this calculation (they do not display a heterozygous genotype). Of all the polymorphisms examined (IL-1A rs1800587; IL-1B rs1143634; IL-1RN rs419598; P2RX7 rs1718119 and rs2230912; IRAK1 rs1059703; CASP1 rs530537, rs580253 and rs554344), only CASP1 rs530537 was associated with EARR (P = .005 and .01 for genotype and allele frequencies, respectively). Specifically, we found that the CC genotype for this variant increased the risk for root resorption with odds ratios of 3.90 and 3.51 times that of the CT and TT genotypes (P = .003 and .01), respectively. However, after Benjamin–Hochberg false discovery rate adjustment, this association was only approaching significance (P = .09).

Furthermore, logistic regression identified that the GG genotype in IL-1A rs1800587, when compared to GA + AA, interacted with overjet (P = .02). For every unit increase in overjet, the odds ratio to develop EARR for the GG genotype was 1.39 with 95% confidence interval (1.05, 1.85). However, no other significant interactions were found between SNPs and sex, age, dental and skeletal classification, ANB, U1-SN, overjet, treatment type or treatment time. On the other hand, after controlling for significant genetic associations, we found that only in the CASP1 rs530537 did dental classification and treatment type remain significant predictors for EARR.

4 DISCUSSION

External apical root resorption is a complex disorder associated with orthodontic treatment. Since resorption could result in significant shortening of the roots and thus compromise the health of the dentition, understanding the risk factors that contribute to EARR is of great concern for the clinicians. In this study, we evaluated the association of certain patient and treatment-related factors in orthodontic patients in Alabama to the risk of developing EARR.

When looking at racial/ethnic differences, Sameshima and Sinclaire found that Caucasians and Hispanics displayed significantly more root resorption than Asians.³⁵ In our sample, the majority of the patients were Caucasians (74%) and the remaining 26% consisted of a combination of 4 races/ethnicities, which included African Americans (16%), Hispanics (6%), Asian (3%) and South Asians (2%). The number of patients of these later races/ethnicities were insufficient to identify statistical significance association with EARR. However, the trend showed Caucasians, Hispanics and South Asians exhibited more root resorption (30%, 27% and 25%, respectively) than African Americans and Asians (19% and 0%, respectively). The other demographic factors we investigated were sex and age. Similar to other studies, we found that neither one of these factors was a good predictor of the risk of developing EARR with orthodontic treatment.^{5, 36}

Recent reviews of the literature found no evidence that malocclusion factors like dental and skeletal classifications, relative position of maxilla to mandible, maxillary incisor proclination and overjet, are associated with root resorption.^{5, 36} Our studies show similar results for the total patient population. However, unlike the whole population sample, we found that, in the Caucasian patients, deviation from class I molar relationship had a statistically increased risk for EARR compared to those with a class I relationship. While a recent clinical practice guideline for EARR stated that it was not clear if dental classification was associated with root resorption,³⁶ we found only one other study that identified class III patients with significantly more EARR.³⁷

In this study, we also evaluated the length of time the orthodontic patients had been in treatment and the effect of extractions on the risk of EARR. Although the length of treatment has been presented as a risk factor for root resorption by various investigators,^{5, 6, 16, 17, 30, 35, 38, 39} we did not identify any significant differences between the EARR and control groups related to treatment time. Since our study included patients that were in treatment for at least 9 months in our orthodontic clinic, it may be suggested that as these patients continue to be seen, a number of our control may develop EARR later in treatment. Even though, as described earlier, there were no significant differences in the distribution of patients who have completed orthodontic treatment between the EARR and control groups, we recognize this as a limitation of this study. Follow-up studies may give insight into the EARR timeline during orthodontic treatment. Furthermore, as described by Segal et al., prolonged treatment times do not necessarily equate with active orthodontic treatment.⁴⁰ Similarly, Sondeijker et al. concluded that length of treatment was not an independent predictor of amount of root resorption; however, they recommend avoiding prolonged orthodontic treatments.³⁶

On the other hand, our data support previous studies related to premolar extraction being associated with EARR.^{5, 36} We found that root resorption was observed more often in patients with maxillary premolar extractions. However, extraction of maxillary and mandibular premolars may not necessarily display a significant increase in EARR. This may be explained by the fact that an upper premolar extraction treatment protocol is often prescribed for patients with class II division 1 malocclusion, where the maxillary incisors need to be retracted. Thus, the incisor root resorption may be due, in part, to their apical displacement. However, in this study we did not measure the apical root displacements of the incisors.

Looking at other patient-related factors, the evidence linking genetic predisposition to EARR is continuing to grow. Several studies have investigated how variations in a number of genes that participate in 1 of the 2 pathways suggested to control osteoclasts activation during resorption: the ATP/P2XR7/IL-1 inflammation modulation pathway; or the OPG/RANK/RANKL (RANK ligand) pathway, contribute to the EARR phenotype.^{6, 8, 12, 15, 17-22, 26-30} In this study, we analysed polymorphisms in genes that encode for proteins involved in the ATP/P2XR7/IL-1 inflammation modulation pathway. This included genotyping for 9 SNPs from 6 genes (IL-1A rs1800587; IL-1B rs1143634; IL-1RN rs419598; P2RX7 rs1718119 and rs2230912; IRAK1 rs1059703; CASP1 rs530537, rs580253 and rs554344).

Members of the interleukin 1 family cytokines were some of the earliest molecules investigated in conjunction with root resorption. In some studies, the CC genotype of IL-1B rs1143634 and the TT genotypes of IL-1A rs1800587 and IL-1RN rs419598 polymorphisms have been linked to an increase risk of EARR.^{17, 18, 20-22, 33} However, we did not identify any significant associations with EARR for any of these3 cytokines with one exception. In our Caucasian patients, an increase in overjet was associated with a higher risk for root resorption in individuals carrying the GG genotype for rs1800587 variant of IL-1A. In fact, this was the only SNP that significantly interacted with any of the other patient or treatment-related variables.

Concerning the P2RX7 polymorphisms, 3 previous studies have examined their association with EARR.^{6, 15, 16} Two of them found the GG genotype of the rs1718119 variant to be a significant risk for root resorption.^{15, 16} However, the second SNP (rs2230912) was not linked to EARR.⁶ In this study, we did not find that any of the P2RX7 polymorphisms tested were predictive of root resorption. Similarly, we discovered no association between the IRAK1 rs1059703 polymorphism and EARR. This contrasts the study by Pereira et al., who identified the CC genotype (C in males) as protective against root resorption.³⁰

Finally, we evaluated 3 SNPs of the CASP1 gene (rs530537, rs580253 and rs554344). Similar to Sharab et al. study, we saw no association between EARR and the rs580253 and rs554344 polymorphisms.⁶ On the contrary, our study demonstrated that the CC genotype for the rs530537 variant may increase the risk for root resorption. Furthermore, after controlling for significant genetic associations, we found that in patients with this SNP, dental classification and treatment type remained significant predictors for EARR. To our knowledge, this is the first study to demonstrate a potential link between a CASP1 polymorphism and the EARR disorder. Further studies are needed to elucidate the possible mechanism of how this polymorphism may affect the function of CASP1 and, consequently, the activation of pro-inflammatory cytokines like IL-1B and the signalling pathway leading to root resorption.

5 CONCLUSION

Both patients and treatment factors contribute to the development of EARR. In this study we showed that maxillary extraction treatment protocol was a risk factor for EARR. We also demonstrated that, in Caucasians, class II and III dental classifications and class III skeletal relationship was associated with root resorption. Furthermore, we found that the GG genotype of rs1800587 variant of IL-1A predisposed individuals with an increased overjet to root resorption. Finally, we showed that the rs530537 polymorphism of CASP1 gene may contribute to the EARR phenotype and should be included as a candidate SNP in future studies.

AUTHOR CONTRIBUTIONS

Nellie N. Baghaei participated in data collection and assisted in manuscript preparation. Guihua Zhai performed the statistical analysis and assisted in manuscript editing. Ejvis Lamani contributed to the design, methodology, formal analysis, writing the final version of the manuscript. All authors read and approved the final manuscript.

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest to declare.