Chemotherapy and Radiotherapy Long-Term Adverse Effects on Oral Health of Childhood Cancer Survivors: A Systematic Review and Meta-Analysis

Simple Summary The survival rate for pediatric cancer has increased over the past few decades, short- and long-term complications have been detected and studied, and oral complications have emerged as an important topic of research. Here, we aimed to highlight the importance of oral manifestations that may only become apparent years or even decades after cancer treatment. Childhood cancer survivors presented a higher risk of having dental alterations than control counterparts. Additional analyses reveal possible sex-based differences that should be explored in future studies. These results collectively highlight the importance of oral healthcare and the prevention of disease in childhood cancer survivors. Abstract The survival rate for pediatric cancer has increased over the past few decades, short- and long-term complications have been detected and studied, and oral complications have emerged as an important topic of research. Here, we aimed to highlight the importance of oral manifestations that may only become apparent years or even decades after cancer treatment. This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis. We searched articles using PubMed via the MEDLINE, Web of Science, and LILACS databases until October 2023. Overall, 35 observational studies were included, and the results estimated a pooled prevalence of the following dental anomalies: discoloration, 53%; crown-root malformations and agenesis, 36%; enamel hypoplasia, 32%; root development alterations, 29%; unerupted teeth, 24%; microdontia, 16%; hypodontia, 13%; and macrodontia, 7%. Most childhood cancer survivors have at least one dental sequela. Childhood cancer survivors presented a higher risk of having dental alterations than control counterparts. Additional analyses reveal possible sex-based differences that should be explored in future studies. These results collectively highlight the importance of oral healthcare and the prevention of disease in childhood cancer survivors.


Introduction
Childhood cancer is a leading cause of death, with an estimated 400,000 children and adolescents between the ages of 0 and 19 years diagnosed with cancer [1,2].Because they are generally not prevented or detected by screening, accurate and timely diagnosis is essential to promote clinical success and high survival rates [2].The treatment options for pediatric malignancies include chemotherapy, radiation therapy, surgery, and multimodal approaches [1,2].
Cancers 2024, 16, 110 2 of 22 The survival rates for children with cancer have increased; nevertheless, up to 40% of children present complications later due to cancer treatment [3].Short-and long-term complications have been identified, and oral complications are an important research topic [3].In addition, children are three-times more likely than adults to experience developmental complications, exacerbating the impact of searching for this topic [3].Some oral manifestations may occur early during treatment or years or decades after cancer treatment.Short-term adverse effects may include dental caries, mucositis, bleeding, taste alterations, secondary infections, periodontal disease, trismus and osteoradionecrosis [3,4].Long-term complications were not described until the 1970s because the post-treatment observation period was still short [5,6].More recently, combined anticancer treatments have been identified as being responsible for late oral effects, including craniofacial and dental developmental defects and salivary gland dysfunction, especially when performed at a young age [3,[7][8][9].
This systematic review aimed to summarize the findings of estimating the prevalence of oral short-and long-term adverse effects in pediatric cancer survivors during and after oncologic treatment.We aimed to provide information that will allow for the reinforcement of the role of pediatric oncologists for possible dental abnormalities that have a negative impact on the quality of life of both patients and families.

Protocol and Registration
All authors established the protocol, registered it at the National Institute for Health Research PROSPERO platform (ID Number: CRD42022336369), and reported it according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist [10] (Supplementary Table S1).

Focused Questions and Eligibility Criteria
We developed a protocol to answer two PICO questions: 1.
"What is the prevalence of late oral health adverse effects in childhood cancer survivors with a history of chemotherapy and radiotherapy"? 2.
"Are children who undergo cancer therapy more likely to have late oral health adverse effects when compared with healthy controls counterparts"?
Late oral health adverse effects were defined as late sequelae of oncological treatmentrelated toxicities to dentofacial structures.
The respective statements were as follows: pediatric patients with malignant cancer diagnosed between the ages of 3 and 18 years (P, Participants); patients who had undergone a therapeutic combination of radiotherapy/chemotherapy or not by the age of 18 years and were in the primary/mixed/permanent dentition were included (I, Intervention); the presence or absence of a control group was not a limitation (C, Control); estimated prevalence of the late effects of the oral complications (mucositis, candidiasis, ulcers) and dental structures (microdontia, hypodontia, hypoplasia, malformed teeth, impaired root growth, interrupted root growth, V-shaped roots, taurodontism, premature apical closure, and tooth agenesis) (O, Outcome).
Randomized clinical trials, controlled clinical trials, cohort studies (prospective or retrospective design), and cross-sectional studies were eligible for inclusion.The exclusion criteria were as follows: (1) duplicate studies; (2) abstracts, commentaries, reviews, letters to the editor, consensus, opinions, case studies, and case series; (3) unpublished information; (4) lack of appropriate clinical measures; (5) secondary analysis of data sourced from a previous study; and (6) inclusion of animal studies.There were no restrictions on the year or language of publication.

Data Search Strategy and Study Selection
We searched PubMed through MEDLINE, Web of Science, and LILACS for all relevant articles published until October 2023.Grey literature was also searched for using OpenGrey Two independent examiners performed, in duplicate, the assessment of titles and/or abstracts of retrieved studies independently (J.P.L. and L.B.L.).For measurement reproducibility, inter-examiner reliability following full-text assessment was calculated using kappa statistics.Any disagreements were resolved by discussion with a third author (M.M.).

Data Extraction Process and Data Items
Data extraction was performed by two reviewers independently and in duplicate (J.P.L. and L.B.L.).Any paper deemed potentially eligible by one of the reviewers was independently reviewed.All disagreements were resolved by discussion with a third reviewer (VM).The following information was collected: general description, research characteristics, methodology, and outcome measures.The following standard information was extracted from each eligible study: the first author's name, year of publication, country and place of sampling, study period, sample size (male/female), case definition setting, observation setting, sampling strategy, cancer type, treatment (chemotherapy and/or radiotherapy), adverse oral health effects, study funding, and risk of bias.

Risk of Bias (RoB) Assessment
The methodological quality of the eligible studies was assessed using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist [11].This tool allowed for analysis in eight domains, presented in the form of questions as follows: (1) Were the criteria for inclusion in the sample clearly defined?(2) Were the study participants and settings described in detail?(3) Was exposure measured in a valid and reliable way? (4) Were objective and standard criteria used for the measurement of the condition?(5) Were confounding factors identified?(6) Were the strategies to deal with the confounding factors stated?(7) Were the outcomes measured in a valid and reliable manner?(8) Was appropriate statistical analysis used?Each item was scored as Y (i.e., yes)-reported and adequate, N (i.e., no)-not reported, and U (i.e., unclear)-reported inadequately.Any disagreements between examiners were resolved through discussion with a third author.Only studies with all items scored with "Y" were considered to be of high quality, studies with at least one item "N" were of low quality, and, finally, for those which presented at least one "U" item and all the others "Y" were of unclear quality.The Risk-Of-Bias VISualization (ROBVIS) tool was used to analyze the risk of bias [12].

Summary Measures and Synthesis of Results
Standard spreadsheet software (Microsoft Excel for Mac, version 16.50.Microsoft, Redmond, WA, USA) was used for data extraction.Frequencies and percentages were used to describe categorical variables, whereas continuous variables were reported as mean ± standard deviation (SD) and range.Random-effects meta-analysis and forest plots of prevalence were calculated in R version 3.4.1 (R Studio Team 2018) using the 'meta' package [13], through the DerSimonian-Laird random-effects meta-analysis.A meta-analysis was performed to calculate dental anomalies in pediatric cancer survivors.A risk ratio (RR) with a 95% confidence interval (CI) was used to describe the dental disharmonies of cancer survivors compared to healthy children.The RR was pooled using a random-effects model in R version 3.4.1 (R Studio Team 2018), using the 'readxl' package and pairwise random-effects meta-analysis, and p-values less than 0.05 were considered statistically significant.The chi-square (χ 2 ) test was used to calculate overall homogeneity, and substantial heterogeneity was considered when I 2 statistics exceeded 50% [14].To explore potential sources of heterogeneity, we performed a subgroup analysis according to the methodological quality of the included studies and the female/male ratio.Publication bias was considered when the meta-analysis included at least 10 studies [14].

Study Selection
The online search strategy identified 3601 potentially relevant publications.After removing duplicates, 3029 articles were assessed against the eligibility criteria, and 2950 were excluded after title and/or abstract review.Of the 79 articles assessed for eligibility for full-paper review, 44 were excluded, with the respective reasons for exclusion detailed in Supplementary Table S2.As a result, a final number of 35 observational studies were included for qualitative synthesis; a PRISMA diagram is shown in Figure 1.The interexaminer reliability of the full-text screening was considered very high (kappa score = 0.915, 95% CI: 0.895-0.925).
pairwise random-effects meta-analysis, and p-values less than 0.05 were considered statistically significant.The chi-square (χ 2 ) test was used to calculate overall homogeneity, and substantial heterogeneity was considered when I 2 statistics exceeded 50% [14].To explore potential sources of heterogeneity, we performed a subgroup analysis according to the methodological quality of the included studies and the female/male ratio.Publication bias was considered when the meta-analysis included at least 10 studies [14].

Study Selection
The online search strategy identified 3601 potentially relevant publications.After removing duplicates, 3029 articles were assessed against the eligibility criteria, and 2950 were excluded after title and/or abstract review.Of the 79 articles assessed for eligibility for full-paper review, 44 were excluded, with the respective reasons for exclusion detailed in Supplementary Table S2.As a result, a final number of 35 observational studies were included for qualitative synthesis; a PRISMA diagram is shown in Figure 1.The interexaminer reliability of the full-text screening was considered very high (kappa score = 0.915, 95% CI: 0.895-0.925).
Some research highlights themes in a unique way, such as the regularity of dental attendance and type of dentist visited [37], number of erupted teeth relative to age [30], oral mucositis and ulceration, candidiasis, herpes and herpetic gingivo-stomatitis, oral petechiae, facial pain [15], already [21] addressed facial asymmetry and jaw hypoplasia, as well as trismus.Hutton 2010 mentioned traumatized teeth and [35] calculated the root surface areas of mandibular teeth.
Furthermore, studies have been conducted in 16 countries worldwide.Notably, no studies have been performed in Oceania or Africa.

Methodological Quality of the Included Studies
Most studies were categorized with high methodological quality (n = 21, 60%), while six had unclear methodological quality, and eight were of low methodological quality (Table 2).
Y-Yes; U-Unclear; N-No.Items description: 1-Were the criteria for inclusion in the sample clearly defined?; 2-Were the study subjects and the setting described in detail?; 3-Was the exposure measured in a valid and reliable way?; 4-Were objective, standard criteria used for measurement of the condition?;5-Were confounding factors identified?; 6-Were strategies to deal with confounding factors stated?; 7-Were the outcomes measured in a valid and reliable way?; 8-Was appropriate statistical analysis used?
No statistically significant differences were found on craniofacial growth among the controls and cancer survivors [26,28].
Hyposalivation in childhood cancer survivors is relatively high [25,41,43], with more significant alterations found in stimulated salivary flow [27,43].Ref. [39] was the only study reporting no alterations in saliva flow rates.Studies did not find alterations in salivary buffer capacity [41,43], but a salivary microbial flora shift in patients who received radiation therapy was found, with an increased number of mutans streptococci and lactobacilli in saliva [27,41].

Additional Analysis
We further assessed, through sensitivity analyses, whether the risk of bias (Table 5) and the female-male ratio (Table 6) could interfere with the estimates.Risk of bias only proved to be significant in the root development alteration (p < 0.0001).Female-male ratio showed a significant effect in the estimates concerning root development alteration (p < 0.0001), enamel hypoplasia (p = 0.0001), discoloration (p = 0.047), and microdontia (p = 0.0204), unveiling a possible sex-based difference.The results of the present systematic review estimated the pooled prevalence of the following dental anomalies as long-term dental sequelae in patients who had undergone cancer therapy during early childhood: discoloration, 53%; crown-root malformations and agenesis, 36%; enamel hypoplasia, 32%; root development alterations, 29%; unerupted teeth, 24%; microdontia, 16%; hypodontia, 13%; and macrodontia, 7%.Compared with controls, these dental anomalies were significantly more prevalent in cancer survivors and pediatric patients.Root development alterations were 591%, microdontia was 518%, discoloration was 468%, agenesis was 349%, and enamel hypoplasia was 95% more likely to be found in cancer survivors than in controls.

Implications for Practice and Research
As previously mentioned, the late side effects of chemotherapy and radiotherapy on the stomatognathic system in pediatric cancer survivors are numerous, which challenges clinical care and management in the dental setting.Regarding cancer types, it is perceived that the most prevalent cancers in children worldwide are leukemias, with the highest rate, followed by tumors of the central nervous system, then lymphomas, and others [4,16,23].Thus, most patients receive chemotherapy without radiotherapy, but they may receive radiotherapy alone or in combination with chemotherapy.Radiation therapy to the head and/or neck area can range from 27 to 70 Gray [4].And we also know that dental development or odontogenesis is a complex process that occurs over a long period of time, starting in intrauterine life and ending at 14-15 years of age [4,16].Thus, each tooth goes through different stages of development, which when subjected to extrinsic or intrinsic factors, can result in the appearance of dental development defects.Depending on the stage of odontogenesis that is affected, different changes may occur; that is, if any changes occur during histodifferentiation, the structure of enamel and dentin may be altered.In turn, if they occur during morphodifferentiation, they may cause shape and size abnormalities of the teeth, and if the disturbances persist, they can damage root formation, resulting in a shortened or tapered root, which, in turn, can impair tooth eruption and occlusion.The first signs are expected after one to two years of anticancer treatment [4].Some antineoplastics inhibit odontogenesis and eruption and can induce qualitative and quantitative changes in dental tissues.Regarding radiotherapy treatments, exposure to radiotherapy doses greater than 20 Gy has been shown to contribute to a greater risk of developing dental anomalies [28].
Therefore, alterations in root development, microdontia, discoloration, agenesis, and enamel hypoplasia, which were the most common alterations recorded, had an impact on the quality of life.
With all this in mind, it is our understanding that, given the possibility of the presence of dental abnormality and increased caries risk as a consequence of cancer treatment, the most acceptable course of action should be to assume that the quality of life and oral implications are real, so the normal dental therapy scenario may increase the level of clinical priority for preventive screening and early screening.
Despite all the included oral manifestations, crown malformation, prevalence of unerupted teeth, and macrodontia, craniofacial growth was not statistically significant between controls and cancer survivors.However, a higher level of caries in the primary dentition has only been reported once [17], as well as alterations in saliva flow rates [39].

Strengths and Limitations
This study was conducted following PRISMA, a strict and widely advised guideline that has increased robustness and decreased reporting errors.Furthermore, a comprehensive literature search was conducted using a meticulous predefined protocol.Nevertheless, there are some limitations that need to be discussed.It is possible to see that there are several studies that address late health effects; however, in a non-systematic way and on multiple distinct points, this leads to a low sample size of children with cancer, where it is essential to obtain consistent results.
Dental abnormalities have been addressed, but studies on saliva alterations are scarce and have different objectives, making them inconclusive, [25,27,39,41,43], as well as and despite reports of a higher prevalence of caries in the permanent dentition [3,27,31,37,38,42,45].In the deciduous dentition, the results showed that there were no differences between the groups [24,36,39], with only one reporting the opposite [17].One point that was not mentioned was malocclusion and occlusal disharmonies, which would be interesting given the high prevalence of changes in number and tooth development.
Thus, future studies should focus on data representativeness and method standardization to ensure more homogeneous evidence-based results.This information is extremely relevant to pediatric oncologists and to raise awareness among oral health professionals regarding the possible and predictable problems they are facing.

Conclusions
Childhood cancer survivors presented a higher risk of having dental alterations than control counterparts.Also, this group of people also presents considerable prevalence of such conditions.Additional analyses reveal possible sex-based differences that should be explored in future studies, as well as more longitudinal studies, as this is the only way to assess and understand the oral consequences of antineoplastic agents.These results collectively highlight the importance of oral healthcare and the prevention of disease in childhood cancer survivors.

Table 1 .
Characteristics of the included studies.

Table 2 .
Results from the methodological appraisal using JBI Critical Appraisal Checklist.

Table 3 .
Prevalence data on dental anomalies in pediatric cancer patient survivors.

Table 4 .
Risk ratio on dental anomalies in cancer survivor pediatric patients.

Table 5 .
Sensitivity analysis of risk of bias on prevalence using meta-regressions.

Table 6 .
Sensitivity analysis of female and male ratio on prevalence data on dental anomalies in cancer survivor pediatric patients.