View Article PDF

Globally, approximately 300,000 children are diagnosed with cancer each year,1 a disease which historically had high mortality rates.2,3 Recent improvements in oncology treatment and management has led to significant increases in survival rates.2,3 An important consideration in the management of these patients is the long-term effects of their treatment on their physical and psychosocial health.

The long-term dental effects of childhood oncology treatment are well established, and include tooth agenesis,4–8 microdontia,5,7–10 enamel defects6 and disturbances in root formation.7,10,11 In addition, oncology treatment can affect the surrounding structures of the oral cavity leading to hyposalivation,12,13 periodontal disease,14 dental caries7,15 and less commonly, trismus and malocclusions.16 Consequently, these alterations in the oral cavity can negatively affect the patient’s oral health-related quality of life.17

Chemotherapeutic agents and radiation therapies target neoplastic cells by disrupting the cells’ ability to undergo mitosis.18 However, these therapies lack specificity and thus can affect the developing dentition. Due to the multimodal nature of oncology treatment, including surgery, chemotherapy, radiotherapy and stem cell therapy, it can be difficult to determine causality between specific treatments and the dental effects.

In New Zealand, all children are provided with funded dental treatment, including restorations and preventive treatment up until the age of 18 years. However, the lack of funding for oral rehabilitation including complex restorative, orthodontic treatment and oral surgery, could result in a significant financial burden for long-term childhood cancer survivors and their families.19 While medical treatment injuries in New Zealand are covered by the Accident Compensation Corporation (ACC), dental disturbances are considered a well-recognised complication of childhood oncology treatment and therefore ineligible for Government-funded dental rehabilitation. With approximately 150 children being diagnosed with cancer every year in New Zealand,20 and 10-year survival rates being above 80%,21 this may lead to significant inequalities. To date, there has been no literature of the late dental effects in childhood cancer survivors on New Zealand’s unique and diverse population.

The aim of this study is to determine the prevalence of dental developmental disturbances in long-term survivors of childhood malignancies in New Zealand. Secondarily, this study aims to identify potential specific risk factors such as anti-neoplastic agents, types of treatment and age at diagnosis to help oncologists and dentists understand the likelihood of dental abnormalities.

Methods

Ethics approval for this study was provided by the Northern B Health and Disability Ethics Committee, Health and Disability Ethics Committees, New Zealand.

This cross-sectional study was conducted at multiple sites throughout New Zealand (Auckland, Christchurch, Wellington, Hamilton, Whangarei, Tauranga, New Plymouth, Whanganui, Whakatane, Rotorua, Hastings, Gisborne, Palmerston North, Masterton, Nelson and Dunedin).

The study population consisted of children aged 14–16 years old (inclusive in June 2015) who had been diagnosed with cancer prior to 10 years of age and who attended the National Child Cancer Network Late Effects Assessment Programme (LEAP). Children who successfully completed their cancer treatment in New Zealand are monitored in LEAP. The LEAP database includes information on the treatment-related effects including organ dysfunction and secondary malignancies, and is regularly updated by regional LEAP clinical nurse specialists. This database was used to identify potential participants, and source demographic data and information concerning diagnosis, oncology treatment and related comorbidities. However, this database does not provide any clinical or radiographic dental data. Participants who resided in the Wellington region were contacted via phone to inform them of the study and invite them to participate. Those residing outside of Wellington were sent a letter and also contacted via telephone. Once the clinic venue and dates were finalised, all participants were contacted via phone to arrange an appointment time. This appointment was emailed to the participants, and the day prior to the appointment, participants were called to remind them of the appointment. Consent was obtained in writing from participants aged 16 years and older, and from parents/guardians and participants if under the age of 16.

Each participant in this study underwent a clinical examination, radiographs and a self-reported oral health-related quality of life (OHRQoL) survey. Results from the clinical examination will be reported in a separate paper.

Radiographic examination

Radiographic analysis required an orthopantomogram radiograph (OPG). These were taken in the dental clinics of the respective hospitals or donated by private dental practices with this facility. If the child reported to have had an OPG taken in the past 18 months then attempts were made to source this.

After the completion of all clinical examinations, seven clinicians came together and analysed the radiographs using Holtta’s Defect Index.4 The OPGs were also used to identify any other commonly occurring dental abnormalities that may not have been previously described in this group.

Holtta’s Defect Index (HDI)4 is an established method of quantifying dental defects using OPG radiographs. It comprises of three components—tooth agenesis, microdontia and root-crown ratios (Figures 1 and 2). This study used the HDI to classify the permanent dentition of participants. A tooth was considered missing due to agenesis if it was not present on the OPG and after cross referencing with the patient’s dental records to ensure that the teeth were not extracted.

Figure 1: Holtta’s Defect Index.4

The total HDI was calculated as follows: (nD1x1)+(nD2x2)+(nD3x3)+(nD4x4)+(nD5x5).
Where n is the number of teeth in the respective disturbance category D1, D2, D3, D4, D5.

Figure 2: OPG showing dental defects using the Holtta’s Defect Index.

Oral health-related quality of life

All participants were asked to complete an oral health-related quality of life questionnaire prior to the radiographs. The Oral Health Impact Profile-14 (OHIP-14) is a self-reported, internationally validated measure,22 whereby a questionnaire is used to assess seven different domains of oral health-related quality of life (OHRQoL). It consists of 14 questions graded on a scale from 0–4. The final score is a sum of the 14 scores, ranging from 0 to 56. The prevalence of one or more OHIP-14 impacts was computed by identification of those who responded ‘very often’ or ‘fairly often’ for at least one of the 14 items. A concurrent validity check was used, where the total OHIP-14 value should correlate positively with the global oral health question.

NZDep2013

NZDep2013 23 is an area-based measure of socioeconomic status (SES) in New Zealand using data from the 2013 Census to provide a deprivation score for each meshblock. NZDep was found using the participant’s current address. This address was then cross-referenced using Census meshblock data from 2013 to find the participant’s deprivation score.

Statistical analysis

Data analysis was carried out using IBM SPSS version 25.0 in conjunction with a biostatistician. Results were considered statistically significant if the p-value was less than 0.05.

Chemotherapeutic drugs where the sample size was less than seven (n<7) were excluded from our analyses (asparaginase, cyclosporin, fludarabine, topotecan, vinblastine, hydrocortisone, lomustine, teniposide). Doses for anthracyclines including doxorubicin, daunorubicin, epirubicin and idarubicin were converted into “Doxorubicin equivalent doses” (mg/m2) using the COG LTFU guidelines v5.24 Similarly, doses for alkylating agents (cyclophosphamide, ifosfamide, melphalan, thio-TEPA and busulfan) were converted into “Cyclophosphamide equivalent doses” (mg/m2) as outlined by Green et al 2014.25

A one-way ANOVA was conducted as part of the bivariate analysis, examining mean OHIP-14 score, mean number of teeth missing due to agenesis, and mean total HDI for each independent variable (doxorubicin equivalent dose, cyclophosphamide equivalent dose, dactinomycin, vincristine, cytarabine, methotrexate, etoposide, cisplatin, cyclophosphamide, carboplatin, chemotherapy duration, head and neck radiation therapy, total body irradiation, stem cell therapy, gender, ethnicity, diagnosis, NZDep and age at diagnosis). Outputs that lacked statistical and clinical significance have been omitted from the tables. A binary logistic regression was conducted to determine the likelihood of agenesis (Yes/No). A linear regression was conducted to determine the correlation between oncology treatment and outcome factors on OHIP-14.

Results

As of June 2015, there were 245 patients between the ages of 14 and 16 who were diagnosed with cancer before the age of 10. Twenty patients were not eligible for the study due to being lost to follow up, residing outside of New Zealand, or receiving cancer treatment. Of the remaining 225 patients, 69 are not current LEAP patients—14 are yet to be referred to LEAP, 31 have been discharged from LEAP and 24 did not require any LEAP follow-up. LEAP follow-up is not routinely offered to children with a central nervous system tumour treated with surgery only. Of the 225, 156 children were eligible. Of the 156 eligible children, 68 individuals participated in this study—65 filled out the OHIP-14 questionnaire, and 59 either had OPGs or consented to having OPGs. This gives us a participation rate of 37.8%. The most common reasons for non-participation were lack of interest, unavailability on the day of the clinic or non-response.

Participant sociodemographic characteristics are detailed in Table 1. The mean age of participants at follow up was 14.9 years, and 4.1 years at diagnosis. 54.2% of participants were male. A majority of participants were European (76.3%), followed by Māori (11.9%) and Asian (6.8%). The most common cancers were leukaemia (39.0%), followed by CNS tumours (13.6%) and kidney cancers (11.9%). Other cancers included germ cell tumours (n=2) and hepatoblastomas (n=2).

Table 1: Summary of the sociodemographic characteristics.

All participants had a combination of surgical, chemo-, radio- and/or stem cell therapies, with a summary of the oncology treatment characteristics being found in Table 2. Nine out of 10 participants received chemotherapy with an average of four drugs. The most commonly administered drugs were cyclophosphamide (66.1%), doxorubicin (64.4%) and vincristine (61.0%). One in four participants had head and neck radiation therapy with a mean dose of 35.2Gy, and three participants had total body irradiation with a mean dose of 13.0Gy. In addition, 12 children (20.3%) had stem cell therapy.

Table 2: Oncology treatment characteristics summary.

Details of the prevalence of developmental dental anomalies can be found in Table 3. Using Holtta’s Index, this study found that 15.3% of participants had at least one tooth missing due to agenesis, 6.8% had at least one tooth affected by microdontia, and 32.2% had a decreased root:crown ratio affecting at least one tooth. An inter-examiner reliability was conducted for the radiographical analysis (κ=0.83).

Table 3: Prevalence of late dental effects.

Table 4 shows the impact of potential late dental effects on OHRQoL. Participants who experienced agenesis (7.9) had a higher mean OHIP-14 than those who did not (3.7) (p-value<0.05). Those with microdontia (50.0%) were also more likely to report having 1+ OHIP-14 impacts “Fairly Often” and “Very Often” almost three times as frequently than those who did not (18.2%).

Table 4: The impact of potential late dental effects on oral health-related quality of life.

Table 5 presents an overview of the bivariate analysis. No statistically significant results at the p=0.05 level were found for Doxorubicin equivalent doses, cytarabine, cisplatin, etoposide, methotrexate, chemotherapy duration, total body irradiation, gender, ethnicity, NZDep or age at diagnosis. No statistically significant results were found with the individual OHIP-14 domains.

Table 5: Results of bivariate analysis between oncology treatment and outcome factors.

Six of 25 participants (40%) who received head and neck radiation therapy were missing at least one tooth due to agenesis, whereas only 6.8% who did not have head and neck radiation therapy were missing at least one tooth (p-value=0.006). Participants who had up to and including 20Gy of HNRT were missing on average 2.3 teeth, compared to 0.3 teeth in those who did not have HNRT. Participants who had over 20Gy of HNRT were only missing 1.6 teeth on average. HNRT was also associated with higher mean HDI scores, with participants who did not receive HNRT having a mean HDI of 2.5. Those participants who received <20Gy and >20Gy of HNRT (p-value=0.003) had HDIs of 17.0 and 16.0 respectively. A binary logistic regression was carried out to determine the odds that a participant was likely to have at least one tooth missing due to agenesis (yes/no) (see Table 5). Variables examined in this analysis included dactinomycin (yes/no), head and neck radiation therapy (0Gy, >0–<20Gy, >20Gy), stem cell therapy (yes/no) and cyclophosphamide equivalent dose (0mg/m2, >0–<4,000mg/m2, 4,000–<8,000mg/m2, >8,000mg/m2). We found that HNRT was the strongest correlator for agenesis, with participants who received more than 20Gy having an OR=7.7 (p-value=0.04).

Table 6: Binary logistic regression model for the association between oncology treatment factors and agenesis (Yes/No).

One third of participants who received SCT were more likely to have teeth missing due to agenesis while 1 in 10 participants who did not receive SCT had teeth missing due to agenesis. Those who received SCT had a mean HDI score of 15.2, in contrast to an HDI score of 3.7 for those who did not receive SCT (p-value=0.029).

The mean total HDI score was highest for rhabdomyosarcoma (RMS) with a score of 30.7, followed by retinoblastoma (12.8) and other sarcomas (10.2). Conversely, leukaemia had the lowest mean total HDI score of 0.6.

No statistically significant results were found for counts of “Fairly Often” and “Very Often” in OHIP-14. Participants who had missing teeth due to agenesis were almost twice as likely (33.3%) to report having problems “Fairly Often” or “Very Often” with their OHRQoL than those who did not (18.0%). Participants who received dactinomycin (42.9%) or vincristine (27.8%) reported problems more frequently with their OHRQoL than those who did not (17.3% and 8.7% respectively).

Table 7 outlines the linear regression model used to establish the relationship between variables and OHIP-14 scores. Variables examined included dactinomycin (yes/no), gender (male/female), agenesis (yes/no), vincristine (yes/no) and ethnicity (European/non-European). The use of dactinomycin, with all other factors held equal in the model, correlated with a higher mean OHIP-14 score by 5.8 points (p-value=0.006). There was also a positive association between the presence of missing teeth due to agenesis and mean OHIP-14 score (3.3-point higher), although this finding did not meet the statistically significant threshold (p-value=0.064).

Table 7: Linear regression model for the association between oncology treatment factors and mean OHIP-14 scores.

Discussion

This cross-sectional study of the late dental effects of oncology treatment in childhood cancers in New Zealand found a high prevalence of dental agenesis and microdontia among our sample. This study also identified potential risk factors for agenesis including head and neck radiation therapy, stem cell therapy, cyclophosphamide equivalent doses over 8,000mg/m2, and dactinomycin.

While this study provides useful insights into the late dental effects of childhood oncology treatment, there are some weaknesses. The study had a nonideal participation rate of 37.8%. This may be attributed to the long interval between cancer treatment and the study (on average over 11 years), with the majority of non-participants declining to participate due to lack of interest. In addition to this, participation may have been inconvenient for those who had to travel long distances. The geographical spread of the participants also made the logistical organisation of this study difficult. Furthermore, Pasifika had a proportionally lower participation rate of 20% compared to other ethnicities. This is likely to be a reflection of the inequality in access to healthcare for these individuals.

There are a number of possible implications of the low participation rate. It may be that those with more dental problems had more incentive to participate in this study. Three-quarters of patients who were treated for rhabdomyosarcoma participated in this study and they had the highest mean total HDI score of 30.7, over six times greater than non-RMS diagnoses (4.7). However, only one of the three RMS originated in the head and neck region. This may have biased the results towards more severe dental problems than what is true in the population of interest.

The NZDep, and derived SES, was based on the current address and not the address at time of diagnosis. In addition, the NZDep is based on census data which uses meshblocks of, on average, 81 people.23 It is therefore not based solely on an individual’s SES, but includes the SES of others in that meshblock. While SES was not found to be a risk factor for the late dental effects in this study, this could be attributed to our small sample size.

This retrospective study looked at chronological ages of participants at diagnosis. The current literature shows that late dental effects are correlated with a patients age at treatment.8,27,28 While this study did not find any significant associations with age at diagnosis, it should be acknowledged that chronological age does not always correlate with dental age. This is an especially important consideration in this study as it has been shown that Māori and Pasifika children tend to have an accelerated dental development.29

Another consideration is that children with higher body mass index (BMI) values have also been found to have an accelerated dental development.30–32 High BMI values could also have been a confounding factor in this study as children with high BMI values may have received higher doses of chemotherapy, which is often calculated by body surface area.

Despite these considerations, this study provides useful and insightful information. This is the first study of its kind in New Zealand, and showed strong correlations of clinical importance between specific oncology treatment factors and their effects on the developing dentition. It involved an interprofessional collaboration between paediatric oncologists, nurses and dentists from all over New Zealand. Moreover, this study showed a respectable degree of rigour with seven dentists and dental specialists reviewing the information and arriving at a consensus. Furthermore, the participants in this study had comparable sociodemographic characteristics to the non-participants for age, sex and most ethnicities, allowing us to make some generalisations to our population of interest.

The results of the present study are consistent with the international literature, which shows that agenesis and microdontia are found more commonly among childhood cancer survivors. While this study is cross sectional, we can compare our prevalence data to the general population. This study found the prevalence of agenesis was high at 15.3%, compared with 2.5% to 6.9% in a recent meta-analysis of the general population.33 A previous study conducted in New Zealand found that 0.35% of five-year old children had missing teeth.32 The aetiology of agenesis is thought to involve both genetic and environmental factors.34,35 Another environmental factor to consider is maternal smoking, which has been linked to tooth agenesis and possibly childhood cancers. A study by Al-Ani et al found maternal consumption of 10 or more cigarettes per day during pregnancy was associated with four times the odds of having a child with hypodontia; as maternal smoking was not measured in the present study, this association was not able to be investigated.34,36 A similar comparison was found for microdontia (prevalence = 6.8%), with studies on the general population finding a prevalence of up to 3%.37 However, classification of microdontia can be quite subjective depending on which index is used and inter-observer variability.

Hsieh et al 2011 found a positive correlation between cyclophosphamide doses and total HDI.13 Although not statistically significant because of the small sample size, our study paper consolidates this finding. Our study also showed a dose-response relationship between total cyclophosphamide equivalent dose and number of teeth missing due to agenesis, and total HDI. This dose-response relationship between cyclophosphamide equivalent doses and the likelihood of agenesis persisted with multivariate analysis. While not statistically significant, it suggests that this finding is unlikely due to chance or any external factors.

Our findings corroborated the findings of Das et al,16 which found that head and neck radiation therapy was twice as likely to result in agenesis, and were missing five times as many teeth due to agenesis, than those who did not receive HNRT. In the present study this association persisted using the binary logistic regression model, and showed that HNRT greater than 20Gy increased the odds of agenesis seven-fold (OR=7.7) compared with those who did not have HNRT (p-value=0.04).

The association between specific treatment factors and agenesis can be attributed to the non-specificity of oncology treatments which potentially risk all cells undergoing division. It is thought that this disrupts ectomesenchymal cells involved in tooth formation resulting in agenesis.

Contrary to expectations, the mean OHIP-14 score of 4.3 was quite low in comparison with adolescents in the general population.38 The OHIP-14 has been validated in use for adolescents 15 years and older;39 with the mean age of this study’s participants being 14.9 years a child-specific OHRQoL measure such as the Child Perceptions Questionnaire (CPQ11–14) may have been found to be more responsive.40 The New Zealand Oral Health Survey (NZOHS) 2009 used a global question, similar to the one we used. In the present sample, children reported their OHRQoL to be better, compared with the 2009 NZOHS.26 While a linear regression model was carried out to ascertain correlations with mean OHIP-14 scores, we should acknowledge that by including sociodemographic, treatment factors and outcome variables, there is likely to be some level of confounding. Nevertheless, this model found that dactinomycin and agenesis correlated with a higher OHIP-14 score.

Ideally, the late dental effects of oncology treatment would best be studied longitudinally, thereby reducing bias and optimising participation rates. In addition, the role of dactinomycin should be further investigated for its potential effects on the developing dentition.

Conclusion

Although all children with childhood cancer need regular dental follow ups, particular attention should be paid to children who have had HNRT, SCT and/or cyclophosphamide equivalent doses greater than 8,000mg/m2. Particular attention should be paid to dental caries, agenesis and enamel defects, which have been shown to negatively impact OHRQoL. In addition to this, patients treated for rhabdomyosarcomas have the potential to have severe deleterious effects on their oral health. These are important considerations for paediatric oncologists, as patients who have missing teeth are likely to require complex and costly orthodontic and/or rehabilitative treatments.

This study shows that New Zealand survivors of childhood cancer treatment have a significant prevalence of complex dental anomalies. This justifies a review of primary and tertiary dental services to ensure they and their families have consistent publicly funded access to the orthodontic and advanced oral rehabilitative services which they require.

Summary

Abstract

Aim

The aim of this study was to determine the prevalence of dental developmental disturbances in long-term survivors of childhood malignancies in New Zealand children. This study reports associations with potential risk factors to inform oncologists and dentists of the likelihood of dental abnormalities.

Method

The study population was children aged 14–16 years old who were diagnosed with cancer prior to 10 years of age. A total of 156 children were eligible, of which 59 participated in this study. The indices used in this study were Holtta’s Defect Index (HDI), and Oral Health Impact Profile-14 (OHIP-14).

Results

The prevalence of agenesis was 15.3%, microdontia 6.8% and root abnormalities 32.2%. Cyclophosphamide equivalent doses above 8,000mg/m2, stem cell therapy (SCT), and head and neck radiation therapy (HNRT) were associated with a higher mean number of teeth missing due to agenesis. SCT and HNRT were associated with a higher total HDI. A binary logistic regression was carried out to determine the odds of agenesis and found that HNRT was the main contributing factor (OR=7.7, p-value=0.04). The linear regression model found that dactinomycin and agenesis correlated with the largest mean OHIP-14.

Conclusion

This study found that childhood cancer survivors in New Zealand had a high prevalence of developmental dental abnormalities and it identified potential risk factors related to their cancer treatment. Inequitable access to oral rehabilitation for this patient group argues for a mechanism for consistent improved access to publicly funded dental care across district health boards in New Zealand.

Author Information

Matthew Shum, Associate Dentist, Bishop’s Dental Surgery, Hawkes Bay; Erin Mahoney, Paediatric Dentist and Hon. Clinical Senior Lecturer, Dental Department and Faculty of Paediatrics, Hutt Valley DHB and University of Otago, Wellington; Kate Naysmith, Senior Dentist, Dental Department, Hutt Valley DHB, Wellington, Scott Macfarlane, National Clinical Leader—Child Cancer, National Child Cancer Network, Auckland; Rob Corbett, Paediatric Oncologist, Child Health, Canterbury District Health Board, Christchurch; Manisha Narsinh, Senior Dentist, Auckland Regional Hospital and Specialist Dentistry, Auckland DHB, Auckland; Arun Natarajan, Paediatric Dentist and Hon. Clinical Senior Lecturer, Hospital Dental Service and Department of Paediatrics, Christchurch Hospital, Canterbury DHB and University of Otago, Christchurch; Yaso Ramadas, Paediatric Dentist, Auckland Regional Hospital and Specialist Dentistry, Starship Hospital, Auckland DHB, Auckland; Elizabeth Hitchings, Clinical Leader, Dental and Oral Health, Capital and Coast DHB, Wellington; Heather Anderson, Paediatric Dentist, DayOne Paediatric Dental, Christchurch.

Acknowledgements

First, we would like to acknowledge the New Zealand Ministry of Health and the Australian and New Zealand Society of Paediatric Dentistry, which sponsored this research, and Colgate NZ who donated all of the toothbrushes and toothpastes for the participants to use. We would also like to extend an extensive thank you to the dentists who gave up their time to be calibrated and help perform the clinical and radiographic examinations throughout the country. We would also like to thank Kathy Yallop, Liz Summers, Jo Truscott and Angela Wakelin for their assistance in the organisation and coordination of this study. Finally, we would like to thank Dalice Sim, biostatistician who helped us conduct the statistical analysis and help us with all of our statistics-related questions.

Correspondence

Dr Erin Mahoney, Dental Department, Hutt Valley DHB, Private Bag 31907, Lower Hutt 5040.

Correspondence Email

erin.mahoney@huttvalleydhb.org.nz

Competing Interests

Dr Shum reports grants from University of Otago during the conduct of the study. Dr Naysmith and Dr Mahoney report grants from NZ Ministry of Health during the conduct of the study.

1. Steliarova-Foucher E, Colombet M, Ries LAG, et al. International incidence of childhood cancer, 2001–10: a population-based registry study. Lancet Oncol. 2017; 18:719–31.

2. Jemal A, Ward EM, Johnson CJ, et al. Annual Report to the Nation on the Status of Cancer, 1975–2014, Featuring Survival. J Natl Cancer Inst. 2017; 109.

3. National Center for Health S, Seer P, National Cancer I. Cancer statistics review. Cancer statistics review. 1973.

4. Holtta P, Alaluusua S, Saarinen-Pihkala UM, et al. Long-term adverse effects on dentition in children with poor-risk neuroblastoma treated with high-dose chemotherapy and autologous stem cell transplantation with or without total body irradiation.(Post-Transplant Complications). Bone Marrow Transplant. 2002; 29:121.

5. Cubukcu CE, Sevinir B, Ercan İ. Disturbed dental development of permanent teeth in children with solid tumors and lymphomas. Pediatr Blood Cancer. 2012; 58:80–4.

6. Lauritano D, Petruzzi M. Decayed, missing and filled teeth index and dental anomalies in long-term survivors leukaemic children: a prospective controlled study. Med Oral Patol Oral Cir Bucal. 2012; 17:e977.

7. Nemeth O, Hermann P, Kivovics P, et al. Long-term Effects of Chemotherapy on Dental Status of Children Cancer Survivors. Pediatr Hematol Oncol. 2013; 30:208–15.

8. Kang C-M, Hahn SM, Kim HS, et al. Clinical Risk Factors Influencing Dental Developmental Disturbances in Childhood Cancer Survivors. Cancer Res Treat. 2018; 50:926–35.

9. Hölttä P, Alaluusua S, Saarinen-Pihkala UM, et al. Agenesis and microdontia of permanent teeth as late adverse effects after stem cell transplantation in young children. Cancer. 2005; 103:181–90.

10. Proc P, Szczepańska J, Skiba A, et al. Dental Anomalies as Late Adverse Effect among Young Children Treated for Cancer. Cancer Res Treat. 2016; 48:658–67.

11. Duggal MS. Root surface areas in long-term survivors of childhood cancer. Oral Oncol. 2003; 39:178–83.

12. Cacchillo D, Barker GJ, Barker BF. Late effects of head and neck radiation therapy and patient/dentist compliance with recommended dental care. Spec Care Dentist. 1993; 13:159–62.

13. Hsieh SG, Hibbert S, Shaw P, et al. Association of cyclophosphamide use with dental developmental defects and salivary gland dysfunction in recipients of childhood antineoplastic therapy. Cancer. 2011; 117:2219–27.

14. Kaste SC, Goodman P, Leisenring W, et al. Impact of radiation and chemotherapy on risk of dental abnormalities. Cancer. 2009; 115:5817–27.

15. Thomaz EBAF, Mouchrek JCE, Silva AQ, et al. Longitudinal assessment of immunological and oral clinical conditions in patients undergoing anticancer treatment for leukemia. Int J Pediatr Otorhinolaryngol. 2013; 77:1088–93.

16. Das K, Hazarika M, Krishnatreya M, et al. An observational study of cancer treatment-induced dental abnormalities in paediatric cancer patients.(Original Article)(Report). Journal of Health Specialties. 2015; 3:228.

17. Jensen S, Pedersen A, Vissink A, et al. A systematic review of salivary gland hypofunction and xerostomia induced by cancer therapies: prevalence, severity and impact on quality of life. Support Care Cancer. 2010; 18:1039–60.

18. Malhotra V, Perry MC. Classical Chemotherapy: Mechanisms, Toxicities and the Therapeutc Window. Cancer Biol Ther. 2003; 2:1–3.

19. Al-Ani AH, Antoun JS, Thomson WM, et al. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017; 2017:9378325.

20. Ballantine K, & the NZCCR Working Group. The incidence of childhood cancer in New Zealand 2010–2014: A report from the New Zealand Children’s Cancer Registry. Auckland: National Child Cancer Network., 2017.

21. Ballantine K, & the NZCCR Working Group. Child cancer survival in New Zealand 2005–2014: A report from the New Zealand Children’s Cancer Registry. Auckland: National Child Cancer Network. 2017.

22. Slade GD. Derivation and validation of a short-form oral health impact profile. Community Dent Oral Epidemiol. 1997; 25:284–90.

23. Atkinson J, Salmond C, Crampton P. NZDep2013 Index of Deprivation. Department of Public Health, University of Otago, Wellington, Division of Health Sciences, University of Otago, 2014.

24. Children’s Oncology Group. Long-Term Follow-Up Guidelines for Survivors of Childhood, Adolescent, and Young Adult Cancers Version 5.0. Children’s Oncology Group, 2018.

25. Green DM, Nolan VG, Goodman PJ, et al. The cyclophosphamide equivalent dose as an approach for quantifying alkylating agent exposure: A report from the childhood cancer survivor study. Pediatr Blood Cancer. 2014; 61:53–67.

26. Haisman RJ. Our oral health key findings of the 2009 New Zealand Oral Health Survey. In: Mason K, Holmes E, Templeton R, Weerasekera D, New Zealand. Ministry of H, (eds). Wellington, N.Z.: Wellington, N.Z. : Ministry of Health, 2010.

27. Wilberg P, Kanellopoulos A, Ruud E, et al. Dental abnormalities after chemotherapy in long-term survivors of childhood acute lymphoblastic leukemia 7–40 years after diagnosis. Support Care Cancer. 2016; 24:1497–506.

28. Hutton A, Bradwell M, English M, et al. The oral health needs of children after treatment for a solid tumour or lymphoma. Int J Paediatr Dent. 2010; 20:15–23.

29. Te Moananui R, Kieser JA, Herbison P, et al. Advanced dental maturation in New Zealand Maori and Pacific Island children. Am J Hum Biol. 2008; 20:43–50.

30. Wong HM, Peng S-M, Wen YF, et al. A population survey of early factors associated with permanent tooth emergence: findings from a prospective cohort study. Community Dent Oral Epidemiol. 2016; 44:467–75.

31. Wong HM, Peng S-M, Yang Y, et al. Tooth eruption and obesity in 12-year-old children. J Dent Sci. 2017; 12:126–32.

32. Whittington BR, Durward CS. Survey of anomalies in primary teeth and their correlation with the permanent dentition. N Z Dent J. 1996; 92:4–8.

33. Polder BJ, Van’t Hof MA, Van der Linden FPGM, et al. A meta-analysis of the prevalence of dental agenesis of permanent teeth. Community Dent Oral Epidemiol. 2004; 32:217–26.

34. Al-Ani AH, Antoun JS, Thomson WM, et al. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017; 2017:9378325.

35. Ritwik P, Patterson KK. Diagnosis of Tooth Agenesis in Childhood and Risk for Neoplasms in Adulthood. Ochsner J. 2018; 18:345–50.

36. Al-Ani AH, Antoun JS, Thomson WM, et al. Maternal Smoking during Pregnancy Is Associated with Offspring Hypodontia. J Dent Res. 2017; 96:1014–9.

37. Jeong KH, Kim D, Song YM, et al. Epidemiology and genetics of hypodontia and microdontia: a study of twin families. Angle Orthod. 2015; 85:980–5.

38. Colussi PRG, Hugo FN, Muniz FWMG, et al. Oral Health-Related Quality of Life and Associated Factors in Brazilian Adolescents. Braz Dent J. 2017; 28:113–2.

39. Ravaghi V, Ardakan MM, Shahriari S, et al. Comparison of the COHIP and OHIP- 14 as measures of the oral health-related quality of life of adolescents. Community Dent Health. 2011; 28:82–8.

40. Foster Page L, Gilchrist F, Broder HL, et al. A Comparison of Three Child OHRQoL Measures. Dent J (Basel). 2019; 7(1).1

Contact diana@nzma.org.nz
for the PDF of this article

View Article PDF

Globally, approximately 300,000 children are diagnosed with cancer each year,1 a disease which historically had high mortality rates.2,3 Recent improvements in oncology treatment and management has led to significant increases in survival rates.2,3 An important consideration in the management of these patients is the long-term effects of their treatment on their physical and psychosocial health.

The long-term dental effects of childhood oncology treatment are well established, and include tooth agenesis,4–8 microdontia,5,7–10 enamel defects6 and disturbances in root formation.7,10,11 In addition, oncology treatment can affect the surrounding structures of the oral cavity leading to hyposalivation,12,13 periodontal disease,14 dental caries7,15 and less commonly, trismus and malocclusions.16 Consequently, these alterations in the oral cavity can negatively affect the patient’s oral health-related quality of life.17

Chemotherapeutic agents and radiation therapies target neoplastic cells by disrupting the cells’ ability to undergo mitosis.18 However, these therapies lack specificity and thus can affect the developing dentition. Due to the multimodal nature of oncology treatment, including surgery, chemotherapy, radiotherapy and stem cell therapy, it can be difficult to determine causality between specific treatments and the dental effects.

In New Zealand, all children are provided with funded dental treatment, including restorations and preventive treatment up until the age of 18 years. However, the lack of funding for oral rehabilitation including complex restorative, orthodontic treatment and oral surgery, could result in a significant financial burden for long-term childhood cancer survivors and their families.19 While medical treatment injuries in New Zealand are covered by the Accident Compensation Corporation (ACC), dental disturbances are considered a well-recognised complication of childhood oncology treatment and therefore ineligible for Government-funded dental rehabilitation. With approximately 150 children being diagnosed with cancer every year in New Zealand,20 and 10-year survival rates being above 80%,21 this may lead to significant inequalities. To date, there has been no literature of the late dental effects in childhood cancer survivors on New Zealand’s unique and diverse population.

The aim of this study is to determine the prevalence of dental developmental disturbances in long-term survivors of childhood malignancies in New Zealand. Secondarily, this study aims to identify potential specific risk factors such as anti-neoplastic agents, types of treatment and age at diagnosis to help oncologists and dentists understand the likelihood of dental abnormalities.

Methods

Ethics approval for this study was provided by the Northern B Health and Disability Ethics Committee, Health and Disability Ethics Committees, New Zealand.

This cross-sectional study was conducted at multiple sites throughout New Zealand (Auckland, Christchurch, Wellington, Hamilton, Whangarei, Tauranga, New Plymouth, Whanganui, Whakatane, Rotorua, Hastings, Gisborne, Palmerston North, Masterton, Nelson and Dunedin).

The study population consisted of children aged 14–16 years old (inclusive in June 2015) who had been diagnosed with cancer prior to 10 years of age and who attended the National Child Cancer Network Late Effects Assessment Programme (LEAP). Children who successfully completed their cancer treatment in New Zealand are monitored in LEAP. The LEAP database includes information on the treatment-related effects including organ dysfunction and secondary malignancies, and is regularly updated by regional LEAP clinical nurse specialists. This database was used to identify potential participants, and source demographic data and information concerning diagnosis, oncology treatment and related comorbidities. However, this database does not provide any clinical or radiographic dental data. Participants who resided in the Wellington region were contacted via phone to inform them of the study and invite them to participate. Those residing outside of Wellington were sent a letter and also contacted via telephone. Once the clinic venue and dates were finalised, all participants were contacted via phone to arrange an appointment time. This appointment was emailed to the participants, and the day prior to the appointment, participants were called to remind them of the appointment. Consent was obtained in writing from participants aged 16 years and older, and from parents/guardians and participants if under the age of 16.

Each participant in this study underwent a clinical examination, radiographs and a self-reported oral health-related quality of life (OHRQoL) survey. Results from the clinical examination will be reported in a separate paper.

Radiographic examination

Radiographic analysis required an orthopantomogram radiograph (OPG). These were taken in the dental clinics of the respective hospitals or donated by private dental practices with this facility. If the child reported to have had an OPG taken in the past 18 months then attempts were made to source this.

After the completion of all clinical examinations, seven clinicians came together and analysed the radiographs using Holtta’s Defect Index.4 The OPGs were also used to identify any other commonly occurring dental abnormalities that may not have been previously described in this group.

Holtta’s Defect Index (HDI)4 is an established method of quantifying dental defects using OPG radiographs. It comprises of three components—tooth agenesis, microdontia and root-crown ratios (Figures 1 and 2). This study used the HDI to classify the permanent dentition of participants. A tooth was considered missing due to agenesis if it was not present on the OPG and after cross referencing with the patient’s dental records to ensure that the teeth were not extracted.

Figure 1: Holtta’s Defect Index.4

The total HDI was calculated as follows: (nD1x1)+(nD2x2)+(nD3x3)+(nD4x4)+(nD5x5).
Where n is the number of teeth in the respective disturbance category D1, D2, D3, D4, D5.

Figure 2: OPG showing dental defects using the Holtta’s Defect Index.

Oral health-related quality of life

All participants were asked to complete an oral health-related quality of life questionnaire prior to the radiographs. The Oral Health Impact Profile-14 (OHIP-14) is a self-reported, internationally validated measure,22 whereby a questionnaire is used to assess seven different domains of oral health-related quality of life (OHRQoL). It consists of 14 questions graded on a scale from 0–4. The final score is a sum of the 14 scores, ranging from 0 to 56. The prevalence of one or more OHIP-14 impacts was computed by identification of those who responded ‘very often’ or ‘fairly often’ for at least one of the 14 items. A concurrent validity check was used, where the total OHIP-14 value should correlate positively with the global oral health question.

NZDep2013

NZDep2013 23 is an area-based measure of socioeconomic status (SES) in New Zealand using data from the 2013 Census to provide a deprivation score for each meshblock. NZDep was found using the participant’s current address. This address was then cross-referenced using Census meshblock data from 2013 to find the participant’s deprivation score.

Statistical analysis

Data analysis was carried out using IBM SPSS version 25.0 in conjunction with a biostatistician. Results were considered statistically significant if the p-value was less than 0.05.

Chemotherapeutic drugs where the sample size was less than seven (n<7) were excluded from our analyses (asparaginase, cyclosporin, fludarabine, topotecan, vinblastine, hydrocortisone, lomustine, teniposide). Doses for anthracyclines including doxorubicin, daunorubicin, epirubicin and idarubicin were converted into “Doxorubicin equivalent doses” (mg/m2) using the COG LTFU guidelines v5.24 Similarly, doses for alkylating agents (cyclophosphamide, ifosfamide, melphalan, thio-TEPA and busulfan) were converted into “Cyclophosphamide equivalent doses” (mg/m2) as outlined by Green et al 2014.25

A one-way ANOVA was conducted as part of the bivariate analysis, examining mean OHIP-14 score, mean number of teeth missing due to agenesis, and mean total HDI for each independent variable (doxorubicin equivalent dose, cyclophosphamide equivalent dose, dactinomycin, vincristine, cytarabine, methotrexate, etoposide, cisplatin, cyclophosphamide, carboplatin, chemotherapy duration, head and neck radiation therapy, total body irradiation, stem cell therapy, gender, ethnicity, diagnosis, NZDep and age at diagnosis). Outputs that lacked statistical and clinical significance have been omitted from the tables. A binary logistic regression was conducted to determine the likelihood of agenesis (Yes/No). A linear regression was conducted to determine the correlation between oncology treatment and outcome factors on OHIP-14.

Results

As of June 2015, there were 245 patients between the ages of 14 and 16 who were diagnosed with cancer before the age of 10. Twenty patients were not eligible for the study due to being lost to follow up, residing outside of New Zealand, or receiving cancer treatment. Of the remaining 225 patients, 69 are not current LEAP patients—14 are yet to be referred to LEAP, 31 have been discharged from LEAP and 24 did not require any LEAP follow-up. LEAP follow-up is not routinely offered to children with a central nervous system tumour treated with surgery only. Of the 225, 156 children were eligible. Of the 156 eligible children, 68 individuals participated in this study—65 filled out the OHIP-14 questionnaire, and 59 either had OPGs or consented to having OPGs. This gives us a participation rate of 37.8%. The most common reasons for non-participation were lack of interest, unavailability on the day of the clinic or non-response.

Participant sociodemographic characteristics are detailed in Table 1. The mean age of participants at follow up was 14.9 years, and 4.1 years at diagnosis. 54.2% of participants were male. A majority of participants were European (76.3%), followed by Māori (11.9%) and Asian (6.8%). The most common cancers were leukaemia (39.0%), followed by CNS tumours (13.6%) and kidney cancers (11.9%). Other cancers included germ cell tumours (n=2) and hepatoblastomas (n=2).

Table 1: Summary of the sociodemographic characteristics.

All participants had a combination of surgical, chemo-, radio- and/or stem cell therapies, with a summary of the oncology treatment characteristics being found in Table 2. Nine out of 10 participants received chemotherapy with an average of four drugs. The most commonly administered drugs were cyclophosphamide (66.1%), doxorubicin (64.4%) and vincristine (61.0%). One in four participants had head and neck radiation therapy with a mean dose of 35.2Gy, and three participants had total body irradiation with a mean dose of 13.0Gy. In addition, 12 children (20.3%) had stem cell therapy.

Table 2: Oncology treatment characteristics summary.

Details of the prevalence of developmental dental anomalies can be found in Table 3. Using Holtta’s Index, this study found that 15.3% of participants had at least one tooth missing due to agenesis, 6.8% had at least one tooth affected by microdontia, and 32.2% had a decreased root:crown ratio affecting at least one tooth. An inter-examiner reliability was conducted for the radiographical analysis (κ=0.83).

Table 3: Prevalence of late dental effects.

Table 4 shows the impact of potential late dental effects on OHRQoL. Participants who experienced agenesis (7.9) had a higher mean OHIP-14 than those who did not (3.7) (p-value<0.05). Those with microdontia (50.0%) were also more likely to report having 1+ OHIP-14 impacts “Fairly Often” and “Very Often” almost three times as frequently than those who did not (18.2%).

Table 4: The impact of potential late dental effects on oral health-related quality of life.

Table 5 presents an overview of the bivariate analysis. No statistically significant results at the p=0.05 level were found for Doxorubicin equivalent doses, cytarabine, cisplatin, etoposide, methotrexate, chemotherapy duration, total body irradiation, gender, ethnicity, NZDep or age at diagnosis. No statistically significant results were found with the individual OHIP-14 domains.

Table 5: Results of bivariate analysis between oncology treatment and outcome factors.

Six of 25 participants (40%) who received head and neck radiation therapy were missing at least one tooth due to agenesis, whereas only 6.8% who did not have head and neck radiation therapy were missing at least one tooth (p-value=0.006). Participants who had up to and including 20Gy of HNRT were missing on average 2.3 teeth, compared to 0.3 teeth in those who did not have HNRT. Participants who had over 20Gy of HNRT were only missing 1.6 teeth on average. HNRT was also associated with higher mean HDI scores, with participants who did not receive HNRT having a mean HDI of 2.5. Those participants who received <20Gy and >20Gy of HNRT (p-value=0.003) had HDIs of 17.0 and 16.0 respectively. A binary logistic regression was carried out to determine the odds that a participant was likely to have at least one tooth missing due to agenesis (yes/no) (see Table 5). Variables examined in this analysis included dactinomycin (yes/no), head and neck radiation therapy (0Gy, >0–<20Gy, >20Gy), stem cell therapy (yes/no) and cyclophosphamide equivalent dose (0mg/m2, >0–<4,000mg/m2, 4,000–<8,000mg/m2, >8,000mg/m2). We found that HNRT was the strongest correlator for agenesis, with participants who received more than 20Gy having an OR=7.7 (p-value=0.04).

Table 6: Binary logistic regression model for the association between oncology treatment factors and agenesis (Yes/No).

One third of participants who received SCT were more likely to have teeth missing due to agenesis while 1 in 10 participants who did not receive SCT had teeth missing due to agenesis. Those who received SCT had a mean HDI score of 15.2, in contrast to an HDI score of 3.7 for those who did not receive SCT (p-value=0.029).

The mean total HDI score was highest for rhabdomyosarcoma (RMS) with a score of 30.7, followed by retinoblastoma (12.8) and other sarcomas (10.2). Conversely, leukaemia had the lowest mean total HDI score of 0.6.

No statistically significant results were found for counts of “Fairly Often” and “Very Often” in OHIP-14. Participants who had missing teeth due to agenesis were almost twice as likely (33.3%) to report having problems “Fairly Often” or “Very Often” with their OHRQoL than those who did not (18.0%). Participants who received dactinomycin (42.9%) or vincristine (27.8%) reported problems more frequently with their OHRQoL than those who did not (17.3% and 8.7% respectively).

Table 7 outlines the linear regression model used to establish the relationship between variables and OHIP-14 scores. Variables examined included dactinomycin (yes/no), gender (male/female), agenesis (yes/no), vincristine (yes/no) and ethnicity (European/non-European). The use of dactinomycin, with all other factors held equal in the model, correlated with a higher mean OHIP-14 score by 5.8 points (p-value=0.006). There was also a positive association between the presence of missing teeth due to agenesis and mean OHIP-14 score (3.3-point higher), although this finding did not meet the statistically significant threshold (p-value=0.064).

Table 7: Linear regression model for the association between oncology treatment factors and mean OHIP-14 scores.

Discussion

This cross-sectional study of the late dental effects of oncology treatment in childhood cancers in New Zealand found a high prevalence of dental agenesis and microdontia among our sample. This study also identified potential risk factors for agenesis including head and neck radiation therapy, stem cell therapy, cyclophosphamide equivalent doses over 8,000mg/m2, and dactinomycin.

While this study provides useful insights into the late dental effects of childhood oncology treatment, there are some weaknesses. The study had a nonideal participation rate of 37.8%. This may be attributed to the long interval between cancer treatment and the study (on average over 11 years), with the majority of non-participants declining to participate due to lack of interest. In addition to this, participation may have been inconvenient for those who had to travel long distances. The geographical spread of the participants also made the logistical organisation of this study difficult. Furthermore, Pasifika had a proportionally lower participation rate of 20% compared to other ethnicities. This is likely to be a reflection of the inequality in access to healthcare for these individuals.

There are a number of possible implications of the low participation rate. It may be that those with more dental problems had more incentive to participate in this study. Three-quarters of patients who were treated for rhabdomyosarcoma participated in this study and they had the highest mean total HDI score of 30.7, over six times greater than non-RMS diagnoses (4.7). However, only one of the three RMS originated in the head and neck region. This may have biased the results towards more severe dental problems than what is true in the population of interest.

The NZDep, and derived SES, was based on the current address and not the address at time of diagnosis. In addition, the NZDep is based on census data which uses meshblocks of, on average, 81 people.23 It is therefore not based solely on an individual’s SES, but includes the SES of others in that meshblock. While SES was not found to be a risk factor for the late dental effects in this study, this could be attributed to our small sample size.

This retrospective study looked at chronological ages of participants at diagnosis. The current literature shows that late dental effects are correlated with a patients age at treatment.8,27,28 While this study did not find any significant associations with age at diagnosis, it should be acknowledged that chronological age does not always correlate with dental age. This is an especially important consideration in this study as it has been shown that Māori and Pasifika children tend to have an accelerated dental development.29

Another consideration is that children with higher body mass index (BMI) values have also been found to have an accelerated dental development.30–32 High BMI values could also have been a confounding factor in this study as children with high BMI values may have received higher doses of chemotherapy, which is often calculated by body surface area.

Despite these considerations, this study provides useful and insightful information. This is the first study of its kind in New Zealand, and showed strong correlations of clinical importance between specific oncology treatment factors and their effects on the developing dentition. It involved an interprofessional collaboration between paediatric oncologists, nurses and dentists from all over New Zealand. Moreover, this study showed a respectable degree of rigour with seven dentists and dental specialists reviewing the information and arriving at a consensus. Furthermore, the participants in this study had comparable sociodemographic characteristics to the non-participants for age, sex and most ethnicities, allowing us to make some generalisations to our population of interest.

The results of the present study are consistent with the international literature, which shows that agenesis and microdontia are found more commonly among childhood cancer survivors. While this study is cross sectional, we can compare our prevalence data to the general population. This study found the prevalence of agenesis was high at 15.3%, compared with 2.5% to 6.9% in a recent meta-analysis of the general population.33 A previous study conducted in New Zealand found that 0.35% of five-year old children had missing teeth.32 The aetiology of agenesis is thought to involve both genetic and environmental factors.34,35 Another environmental factor to consider is maternal smoking, which has been linked to tooth agenesis and possibly childhood cancers. A study by Al-Ani et al found maternal consumption of 10 or more cigarettes per day during pregnancy was associated with four times the odds of having a child with hypodontia; as maternal smoking was not measured in the present study, this association was not able to be investigated.34,36 A similar comparison was found for microdontia (prevalence = 6.8%), with studies on the general population finding a prevalence of up to 3%.37 However, classification of microdontia can be quite subjective depending on which index is used and inter-observer variability.

Hsieh et al 2011 found a positive correlation between cyclophosphamide doses and total HDI.13 Although not statistically significant because of the small sample size, our study paper consolidates this finding. Our study also showed a dose-response relationship between total cyclophosphamide equivalent dose and number of teeth missing due to agenesis, and total HDI. This dose-response relationship between cyclophosphamide equivalent doses and the likelihood of agenesis persisted with multivariate analysis. While not statistically significant, it suggests that this finding is unlikely due to chance or any external factors.

Our findings corroborated the findings of Das et al,16 which found that head and neck radiation therapy was twice as likely to result in agenesis, and were missing five times as many teeth due to agenesis, than those who did not receive HNRT. In the present study this association persisted using the binary logistic regression model, and showed that HNRT greater than 20Gy increased the odds of agenesis seven-fold (OR=7.7) compared with those who did not have HNRT (p-value=0.04).

The association between specific treatment factors and agenesis can be attributed to the non-specificity of oncology treatments which potentially risk all cells undergoing division. It is thought that this disrupts ectomesenchymal cells involved in tooth formation resulting in agenesis.

Contrary to expectations, the mean OHIP-14 score of 4.3 was quite low in comparison with adolescents in the general population.38 The OHIP-14 has been validated in use for adolescents 15 years and older;39 with the mean age of this study’s participants being 14.9 years a child-specific OHRQoL measure such as the Child Perceptions Questionnaire (CPQ11–14) may have been found to be more responsive.40 The New Zealand Oral Health Survey (NZOHS) 2009 used a global question, similar to the one we used. In the present sample, children reported their OHRQoL to be better, compared with the 2009 NZOHS.26 While a linear regression model was carried out to ascertain correlations with mean OHIP-14 scores, we should acknowledge that by including sociodemographic, treatment factors and outcome variables, there is likely to be some level of confounding. Nevertheless, this model found that dactinomycin and agenesis correlated with a higher OHIP-14 score.

Ideally, the late dental effects of oncology treatment would best be studied longitudinally, thereby reducing bias and optimising participation rates. In addition, the role of dactinomycin should be further investigated for its potential effects on the developing dentition.

Conclusion

Although all children with childhood cancer need regular dental follow ups, particular attention should be paid to children who have had HNRT, SCT and/or cyclophosphamide equivalent doses greater than 8,000mg/m2. Particular attention should be paid to dental caries, agenesis and enamel defects, which have been shown to negatively impact OHRQoL. In addition to this, patients treated for rhabdomyosarcomas have the potential to have severe deleterious effects on their oral health. These are important considerations for paediatric oncologists, as patients who have missing teeth are likely to require complex and costly orthodontic and/or rehabilitative treatments.

This study shows that New Zealand survivors of childhood cancer treatment have a significant prevalence of complex dental anomalies. This justifies a review of primary and tertiary dental services to ensure they and their families have consistent publicly funded access to the orthodontic and advanced oral rehabilitative services which they require.

Summary

Abstract

Aim

The aim of this study was to determine the prevalence of dental developmental disturbances in long-term survivors of childhood malignancies in New Zealand children. This study reports associations with potential risk factors to inform oncologists and dentists of the likelihood of dental abnormalities.

Method

The study population was children aged 14–16 years old who were diagnosed with cancer prior to 10 years of age. A total of 156 children were eligible, of which 59 participated in this study. The indices used in this study were Holtta’s Defect Index (HDI), and Oral Health Impact Profile-14 (OHIP-14).

Results

The prevalence of agenesis was 15.3%, microdontia 6.8% and root abnormalities 32.2%. Cyclophosphamide equivalent doses above 8,000mg/m2, stem cell therapy (SCT), and head and neck radiation therapy (HNRT) were associated with a higher mean number of teeth missing due to agenesis. SCT and HNRT were associated with a higher total HDI. A binary logistic regression was carried out to determine the odds of agenesis and found that HNRT was the main contributing factor (OR=7.7, p-value=0.04). The linear regression model found that dactinomycin and agenesis correlated with the largest mean OHIP-14.

Conclusion

This study found that childhood cancer survivors in New Zealand had a high prevalence of developmental dental abnormalities and it identified potential risk factors related to their cancer treatment. Inequitable access to oral rehabilitation for this patient group argues for a mechanism for consistent improved access to publicly funded dental care across district health boards in New Zealand.

Author Information

Matthew Shum, Associate Dentist, Bishop’s Dental Surgery, Hawkes Bay; Erin Mahoney, Paediatric Dentist and Hon. Clinical Senior Lecturer, Dental Department and Faculty of Paediatrics, Hutt Valley DHB and University of Otago, Wellington; Kate Naysmith, Senior Dentist, Dental Department, Hutt Valley DHB, Wellington, Scott Macfarlane, National Clinical Leader—Child Cancer, National Child Cancer Network, Auckland; Rob Corbett, Paediatric Oncologist, Child Health, Canterbury District Health Board, Christchurch; Manisha Narsinh, Senior Dentist, Auckland Regional Hospital and Specialist Dentistry, Auckland DHB, Auckland; Arun Natarajan, Paediatric Dentist and Hon. Clinical Senior Lecturer, Hospital Dental Service and Department of Paediatrics, Christchurch Hospital, Canterbury DHB and University of Otago, Christchurch; Yaso Ramadas, Paediatric Dentist, Auckland Regional Hospital and Specialist Dentistry, Starship Hospital, Auckland DHB, Auckland; Elizabeth Hitchings, Clinical Leader, Dental and Oral Health, Capital and Coast DHB, Wellington; Heather Anderson, Paediatric Dentist, DayOne Paediatric Dental, Christchurch.

Acknowledgements

First, we would like to acknowledge the New Zealand Ministry of Health and the Australian and New Zealand Society of Paediatric Dentistry, which sponsored this research, and Colgate NZ who donated all of the toothbrushes and toothpastes for the participants to use. We would also like to extend an extensive thank you to the dentists who gave up their time to be calibrated and help perform the clinical and radiographic examinations throughout the country. We would also like to thank Kathy Yallop, Liz Summers, Jo Truscott and Angela Wakelin for their assistance in the organisation and coordination of this study. Finally, we would like to thank Dalice Sim, biostatistician who helped us conduct the statistical analysis and help us with all of our statistics-related questions.

Correspondence

Dr Erin Mahoney, Dental Department, Hutt Valley DHB, Private Bag 31907, Lower Hutt 5040.

Correspondence Email

erin.mahoney@huttvalleydhb.org.nz

Competing Interests

Dr Shum reports grants from University of Otago during the conduct of the study. Dr Naysmith and Dr Mahoney report grants from NZ Ministry of Health during the conduct of the study.

1. Steliarova-Foucher E, Colombet M, Ries LAG, et al. International incidence of childhood cancer, 2001–10: a population-based registry study. Lancet Oncol. 2017; 18:719–31.

2. Jemal A, Ward EM, Johnson CJ, et al. Annual Report to the Nation on the Status of Cancer, 1975–2014, Featuring Survival. J Natl Cancer Inst. 2017; 109.

3. National Center for Health S, Seer P, National Cancer I. Cancer statistics review. Cancer statistics review. 1973.

4. Holtta P, Alaluusua S, Saarinen-Pihkala UM, et al. Long-term adverse effects on dentition in children with poor-risk neuroblastoma treated with high-dose chemotherapy and autologous stem cell transplantation with or without total body irradiation.(Post-Transplant Complications). Bone Marrow Transplant. 2002; 29:121.

5. Cubukcu CE, Sevinir B, Ercan İ. Disturbed dental development of permanent teeth in children with solid tumors and lymphomas. Pediatr Blood Cancer. 2012; 58:80–4.

6. Lauritano D, Petruzzi M. Decayed, missing and filled teeth index and dental anomalies in long-term survivors leukaemic children: a prospective controlled study. Med Oral Patol Oral Cir Bucal. 2012; 17:e977.

7. Nemeth O, Hermann P, Kivovics P, et al. Long-term Effects of Chemotherapy on Dental Status of Children Cancer Survivors. Pediatr Hematol Oncol. 2013; 30:208–15.

8. Kang C-M, Hahn SM, Kim HS, et al. Clinical Risk Factors Influencing Dental Developmental Disturbances in Childhood Cancer Survivors. Cancer Res Treat. 2018; 50:926–35.

9. Hölttä P, Alaluusua S, Saarinen-Pihkala UM, et al. Agenesis and microdontia of permanent teeth as late adverse effects after stem cell transplantation in young children. Cancer. 2005; 103:181–90.

10. Proc P, Szczepańska J, Skiba A, et al. Dental Anomalies as Late Adverse Effect among Young Children Treated for Cancer. Cancer Res Treat. 2016; 48:658–67.

11. Duggal MS. Root surface areas in long-term survivors of childhood cancer. Oral Oncol. 2003; 39:178–83.

12. Cacchillo D, Barker GJ, Barker BF. Late effects of head and neck radiation therapy and patient/dentist compliance with recommended dental care. Spec Care Dentist. 1993; 13:159–62.

13. Hsieh SG, Hibbert S, Shaw P, et al. Association of cyclophosphamide use with dental developmental defects and salivary gland dysfunction in recipients of childhood antineoplastic therapy. Cancer. 2011; 117:2219–27.

14. Kaste SC, Goodman P, Leisenring W, et al. Impact of radiation and chemotherapy on risk of dental abnormalities. Cancer. 2009; 115:5817–27.

15. Thomaz EBAF, Mouchrek JCE, Silva AQ, et al. Longitudinal assessment of immunological and oral clinical conditions in patients undergoing anticancer treatment for leukemia. Int J Pediatr Otorhinolaryngol. 2013; 77:1088–93.

16. Das K, Hazarika M, Krishnatreya M, et al. An observational study of cancer treatment-induced dental abnormalities in paediatric cancer patients.(Original Article)(Report). Journal of Health Specialties. 2015; 3:228.

17. Jensen S, Pedersen A, Vissink A, et al. A systematic review of salivary gland hypofunction and xerostomia induced by cancer therapies: prevalence, severity and impact on quality of life. Support Care Cancer. 2010; 18:1039–60.

18. Malhotra V, Perry MC. Classical Chemotherapy: Mechanisms, Toxicities and the Therapeutc Window. Cancer Biol Ther. 2003; 2:1–3.

19. Al-Ani AH, Antoun JS, Thomson WM, et al. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017; 2017:9378325.

20. Ballantine K, & the NZCCR Working Group. The incidence of childhood cancer in New Zealand 2010–2014: A report from the New Zealand Children’s Cancer Registry. Auckland: National Child Cancer Network., 2017.

21. Ballantine K, & the NZCCR Working Group. Child cancer survival in New Zealand 2005–2014: A report from the New Zealand Children’s Cancer Registry. Auckland: National Child Cancer Network. 2017.

22. Slade GD. Derivation and validation of a short-form oral health impact profile. Community Dent Oral Epidemiol. 1997; 25:284–90.

23. Atkinson J, Salmond C, Crampton P. NZDep2013 Index of Deprivation. Department of Public Health, University of Otago, Wellington, Division of Health Sciences, University of Otago, 2014.

24. Children’s Oncology Group. Long-Term Follow-Up Guidelines for Survivors of Childhood, Adolescent, and Young Adult Cancers Version 5.0. Children’s Oncology Group, 2018.

25. Green DM, Nolan VG, Goodman PJ, et al. The cyclophosphamide equivalent dose as an approach for quantifying alkylating agent exposure: A report from the childhood cancer survivor study. Pediatr Blood Cancer. 2014; 61:53–67.

26. Haisman RJ. Our oral health key findings of the 2009 New Zealand Oral Health Survey. In: Mason K, Holmes E, Templeton R, Weerasekera D, New Zealand. Ministry of H, (eds). Wellington, N.Z.: Wellington, N.Z. : Ministry of Health, 2010.

27. Wilberg P, Kanellopoulos A, Ruud E, et al. Dental abnormalities after chemotherapy in long-term survivors of childhood acute lymphoblastic leukemia 7–40 years after diagnosis. Support Care Cancer. 2016; 24:1497–506.

28. Hutton A, Bradwell M, English M, et al. The oral health needs of children after treatment for a solid tumour or lymphoma. Int J Paediatr Dent. 2010; 20:15–23.

29. Te Moananui R, Kieser JA, Herbison P, et al. Advanced dental maturation in New Zealand Maori and Pacific Island children. Am J Hum Biol. 2008; 20:43–50.

30. Wong HM, Peng S-M, Wen YF, et al. A population survey of early factors associated with permanent tooth emergence: findings from a prospective cohort study. Community Dent Oral Epidemiol. 2016; 44:467–75.

31. Wong HM, Peng S-M, Yang Y, et al. Tooth eruption and obesity in 12-year-old children. J Dent Sci. 2017; 12:126–32.

32. Whittington BR, Durward CS. Survey of anomalies in primary teeth and their correlation with the permanent dentition. N Z Dent J. 1996; 92:4–8.

33. Polder BJ, Van’t Hof MA, Van der Linden FPGM, et al. A meta-analysis of the prevalence of dental agenesis of permanent teeth. Community Dent Oral Epidemiol. 2004; 32:217–26.

34. Al-Ani AH, Antoun JS, Thomson WM, et al. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017; 2017:9378325.

35. Ritwik P, Patterson KK. Diagnosis of Tooth Agenesis in Childhood and Risk for Neoplasms in Adulthood. Ochsner J. 2018; 18:345–50.

36. Al-Ani AH, Antoun JS, Thomson WM, et al. Maternal Smoking during Pregnancy Is Associated with Offspring Hypodontia. J Dent Res. 2017; 96:1014–9.

37. Jeong KH, Kim D, Song YM, et al. Epidemiology and genetics of hypodontia and microdontia: a study of twin families. Angle Orthod. 2015; 85:980–5.

38. Colussi PRG, Hugo FN, Muniz FWMG, et al. Oral Health-Related Quality of Life and Associated Factors in Brazilian Adolescents. Braz Dent J. 2017; 28:113–2.

39. Ravaghi V, Ardakan MM, Shahriari S, et al. Comparison of the COHIP and OHIP- 14 as measures of the oral health-related quality of life of adolescents. Community Dent Health. 2011; 28:82–8.

40. Foster Page L, Gilchrist F, Broder HL, et al. A Comparison of Three Child OHRQoL Measures. Dent J (Basel). 2019; 7(1).1

Contact diana@nzma.org.nz
for the PDF of this article

View Article PDF

Globally, approximately 300,000 children are diagnosed with cancer each year,1 a disease which historically had high mortality rates.2,3 Recent improvements in oncology treatment and management has led to significant increases in survival rates.2,3 An important consideration in the management of these patients is the long-term effects of their treatment on their physical and psychosocial health.

The long-term dental effects of childhood oncology treatment are well established, and include tooth agenesis,4–8 microdontia,5,7–10 enamel defects6 and disturbances in root formation.7,10,11 In addition, oncology treatment can affect the surrounding structures of the oral cavity leading to hyposalivation,12,13 periodontal disease,14 dental caries7,15 and less commonly, trismus and malocclusions.16 Consequently, these alterations in the oral cavity can negatively affect the patient’s oral health-related quality of life.17

Chemotherapeutic agents and radiation therapies target neoplastic cells by disrupting the cells’ ability to undergo mitosis.18 However, these therapies lack specificity and thus can affect the developing dentition. Due to the multimodal nature of oncology treatment, including surgery, chemotherapy, radiotherapy and stem cell therapy, it can be difficult to determine causality between specific treatments and the dental effects.

In New Zealand, all children are provided with funded dental treatment, including restorations and preventive treatment up until the age of 18 years. However, the lack of funding for oral rehabilitation including complex restorative, orthodontic treatment and oral surgery, could result in a significant financial burden for long-term childhood cancer survivors and their families.19 While medical treatment injuries in New Zealand are covered by the Accident Compensation Corporation (ACC), dental disturbances are considered a well-recognised complication of childhood oncology treatment and therefore ineligible for Government-funded dental rehabilitation. With approximately 150 children being diagnosed with cancer every year in New Zealand,20 and 10-year survival rates being above 80%,21 this may lead to significant inequalities. To date, there has been no literature of the late dental effects in childhood cancer survivors on New Zealand’s unique and diverse population.

The aim of this study is to determine the prevalence of dental developmental disturbances in long-term survivors of childhood malignancies in New Zealand. Secondarily, this study aims to identify potential specific risk factors such as anti-neoplastic agents, types of treatment and age at diagnosis to help oncologists and dentists understand the likelihood of dental abnormalities.

Methods

Ethics approval for this study was provided by the Northern B Health and Disability Ethics Committee, Health and Disability Ethics Committees, New Zealand.

This cross-sectional study was conducted at multiple sites throughout New Zealand (Auckland, Christchurch, Wellington, Hamilton, Whangarei, Tauranga, New Plymouth, Whanganui, Whakatane, Rotorua, Hastings, Gisborne, Palmerston North, Masterton, Nelson and Dunedin).

The study population consisted of children aged 14–16 years old (inclusive in June 2015) who had been diagnosed with cancer prior to 10 years of age and who attended the National Child Cancer Network Late Effects Assessment Programme (LEAP). Children who successfully completed their cancer treatment in New Zealand are monitored in LEAP. The LEAP database includes information on the treatment-related effects including organ dysfunction and secondary malignancies, and is regularly updated by regional LEAP clinical nurse specialists. This database was used to identify potential participants, and source demographic data and information concerning diagnosis, oncology treatment and related comorbidities. However, this database does not provide any clinical or radiographic dental data. Participants who resided in the Wellington region were contacted via phone to inform them of the study and invite them to participate. Those residing outside of Wellington were sent a letter and also contacted via telephone. Once the clinic venue and dates were finalised, all participants were contacted via phone to arrange an appointment time. This appointment was emailed to the participants, and the day prior to the appointment, participants were called to remind them of the appointment. Consent was obtained in writing from participants aged 16 years and older, and from parents/guardians and participants if under the age of 16.

Each participant in this study underwent a clinical examination, radiographs and a self-reported oral health-related quality of life (OHRQoL) survey. Results from the clinical examination will be reported in a separate paper.

Radiographic examination

Radiographic analysis required an orthopantomogram radiograph (OPG). These were taken in the dental clinics of the respective hospitals or donated by private dental practices with this facility. If the child reported to have had an OPG taken in the past 18 months then attempts were made to source this.

After the completion of all clinical examinations, seven clinicians came together and analysed the radiographs using Holtta’s Defect Index.4 The OPGs were also used to identify any other commonly occurring dental abnormalities that may not have been previously described in this group.

Holtta’s Defect Index (HDI)4 is an established method of quantifying dental defects using OPG radiographs. It comprises of three components—tooth agenesis, microdontia and root-crown ratios (Figures 1 and 2). This study used the HDI to classify the permanent dentition of participants. A tooth was considered missing due to agenesis if it was not present on the OPG and after cross referencing with the patient’s dental records to ensure that the teeth were not extracted.

Figure 1: Holtta’s Defect Index.4

The total HDI was calculated as follows: (nD1x1)+(nD2x2)+(nD3x3)+(nD4x4)+(nD5x5).
Where n is the number of teeth in the respective disturbance category D1, D2, D3, D4, D5.

Figure 2: OPG showing dental defects using the Holtta’s Defect Index.

Oral health-related quality of life

All participants were asked to complete an oral health-related quality of life questionnaire prior to the radiographs. The Oral Health Impact Profile-14 (OHIP-14) is a self-reported, internationally validated measure,22 whereby a questionnaire is used to assess seven different domains of oral health-related quality of life (OHRQoL). It consists of 14 questions graded on a scale from 0–4. The final score is a sum of the 14 scores, ranging from 0 to 56. The prevalence of one or more OHIP-14 impacts was computed by identification of those who responded ‘very often’ or ‘fairly often’ for at least one of the 14 items. A concurrent validity check was used, where the total OHIP-14 value should correlate positively with the global oral health question.

NZDep2013

NZDep2013 23 is an area-based measure of socioeconomic status (SES) in New Zealand using data from the 2013 Census to provide a deprivation score for each meshblock. NZDep was found using the participant’s current address. This address was then cross-referenced using Census meshblock data from 2013 to find the participant’s deprivation score.

Statistical analysis

Data analysis was carried out using IBM SPSS version 25.0 in conjunction with a biostatistician. Results were considered statistically significant if the p-value was less than 0.05.

Chemotherapeutic drugs where the sample size was less than seven (n<7) were excluded from our analyses (asparaginase, cyclosporin, fludarabine, topotecan, vinblastine, hydrocortisone, lomustine, teniposide). Doses for anthracyclines including doxorubicin, daunorubicin, epirubicin and idarubicin were converted into “Doxorubicin equivalent doses” (mg/m2) using the COG LTFU guidelines v5.24 Similarly, doses for alkylating agents (cyclophosphamide, ifosfamide, melphalan, thio-TEPA and busulfan) were converted into “Cyclophosphamide equivalent doses” (mg/m2) as outlined by Green et al 2014.25

A one-way ANOVA was conducted as part of the bivariate analysis, examining mean OHIP-14 score, mean number of teeth missing due to agenesis, and mean total HDI for each independent variable (doxorubicin equivalent dose, cyclophosphamide equivalent dose, dactinomycin, vincristine, cytarabine, methotrexate, etoposide, cisplatin, cyclophosphamide, carboplatin, chemotherapy duration, head and neck radiation therapy, total body irradiation, stem cell therapy, gender, ethnicity, diagnosis, NZDep and age at diagnosis). Outputs that lacked statistical and clinical significance have been omitted from the tables. A binary logistic regression was conducted to determine the likelihood of agenesis (Yes/No). A linear regression was conducted to determine the correlation between oncology treatment and outcome factors on OHIP-14.

Results

As of June 2015, there were 245 patients between the ages of 14 and 16 who were diagnosed with cancer before the age of 10. Twenty patients were not eligible for the study due to being lost to follow up, residing outside of New Zealand, or receiving cancer treatment. Of the remaining 225 patients, 69 are not current LEAP patients—14 are yet to be referred to LEAP, 31 have been discharged from LEAP and 24 did not require any LEAP follow-up. LEAP follow-up is not routinely offered to children with a central nervous system tumour treated with surgery only. Of the 225, 156 children were eligible. Of the 156 eligible children, 68 individuals participated in this study—65 filled out the OHIP-14 questionnaire, and 59 either had OPGs or consented to having OPGs. This gives us a participation rate of 37.8%. The most common reasons for non-participation were lack of interest, unavailability on the day of the clinic or non-response.

Participant sociodemographic characteristics are detailed in Table 1. The mean age of participants at follow up was 14.9 years, and 4.1 years at diagnosis. 54.2% of participants were male. A majority of participants were European (76.3%), followed by Māori (11.9%) and Asian (6.8%). The most common cancers were leukaemia (39.0%), followed by CNS tumours (13.6%) and kidney cancers (11.9%). Other cancers included germ cell tumours (n=2) and hepatoblastomas (n=2).

Table 1: Summary of the sociodemographic characteristics.

All participants had a combination of surgical, chemo-, radio- and/or stem cell therapies, with a summary of the oncology treatment characteristics being found in Table 2. Nine out of 10 participants received chemotherapy with an average of four drugs. The most commonly administered drugs were cyclophosphamide (66.1%), doxorubicin (64.4%) and vincristine (61.0%). One in four participants had head and neck radiation therapy with a mean dose of 35.2Gy, and three participants had total body irradiation with a mean dose of 13.0Gy. In addition, 12 children (20.3%) had stem cell therapy.

Table 2: Oncology treatment characteristics summary.

Details of the prevalence of developmental dental anomalies can be found in Table 3. Using Holtta’s Index, this study found that 15.3% of participants had at least one tooth missing due to agenesis, 6.8% had at least one tooth affected by microdontia, and 32.2% had a decreased root:crown ratio affecting at least one tooth. An inter-examiner reliability was conducted for the radiographical analysis (κ=0.83).

Table 3: Prevalence of late dental effects.

Table 4 shows the impact of potential late dental effects on OHRQoL. Participants who experienced agenesis (7.9) had a higher mean OHIP-14 than those who did not (3.7) (p-value<0.05). Those with microdontia (50.0%) were also more likely to report having 1+ OHIP-14 impacts “Fairly Often” and “Very Often” almost three times as frequently than those who did not (18.2%).

Table 4: The impact of potential late dental effects on oral health-related quality of life.

Table 5 presents an overview of the bivariate analysis. No statistically significant results at the p=0.05 level were found for Doxorubicin equivalent doses, cytarabine, cisplatin, etoposide, methotrexate, chemotherapy duration, total body irradiation, gender, ethnicity, NZDep or age at diagnosis. No statistically significant results were found with the individual OHIP-14 domains.

Table 5: Results of bivariate analysis between oncology treatment and outcome factors.

Six of 25 participants (40%) who received head and neck radiation therapy were missing at least one tooth due to agenesis, whereas only 6.8% who did not have head and neck radiation therapy were missing at least one tooth (p-value=0.006). Participants who had up to and including 20Gy of HNRT were missing on average 2.3 teeth, compared to 0.3 teeth in those who did not have HNRT. Participants who had over 20Gy of HNRT were only missing 1.6 teeth on average. HNRT was also associated with higher mean HDI scores, with participants who did not receive HNRT having a mean HDI of 2.5. Those participants who received <20Gy and >20Gy of HNRT (p-value=0.003) had HDIs of 17.0 and 16.0 respectively. A binary logistic regression was carried out to determine the odds that a participant was likely to have at least one tooth missing due to agenesis (yes/no) (see Table 5). Variables examined in this analysis included dactinomycin (yes/no), head and neck radiation therapy (0Gy, >0–<20Gy, >20Gy), stem cell therapy (yes/no) and cyclophosphamide equivalent dose (0mg/m2, >0–<4,000mg/m2, 4,000–<8,000mg/m2, >8,000mg/m2). We found that HNRT was the strongest correlator for agenesis, with participants who received more than 20Gy having an OR=7.7 (p-value=0.04).

Table 6: Binary logistic regression model for the association between oncology treatment factors and agenesis (Yes/No).

One third of participants who received SCT were more likely to have teeth missing due to agenesis while 1 in 10 participants who did not receive SCT had teeth missing due to agenesis. Those who received SCT had a mean HDI score of 15.2, in contrast to an HDI score of 3.7 for those who did not receive SCT (p-value=0.029).

The mean total HDI score was highest for rhabdomyosarcoma (RMS) with a score of 30.7, followed by retinoblastoma (12.8) and other sarcomas (10.2). Conversely, leukaemia had the lowest mean total HDI score of 0.6.

No statistically significant results were found for counts of “Fairly Often” and “Very Often” in OHIP-14. Participants who had missing teeth due to agenesis were almost twice as likely (33.3%) to report having problems “Fairly Often” or “Very Often” with their OHRQoL than those who did not (18.0%). Participants who received dactinomycin (42.9%) or vincristine (27.8%) reported problems more frequently with their OHRQoL than those who did not (17.3% and 8.7% respectively).

Table 7 outlines the linear regression model used to establish the relationship between variables and OHIP-14 scores. Variables examined included dactinomycin (yes/no), gender (male/female), agenesis (yes/no), vincristine (yes/no) and ethnicity (European/non-European). The use of dactinomycin, with all other factors held equal in the model, correlated with a higher mean OHIP-14 score by 5.8 points (p-value=0.006). There was also a positive association between the presence of missing teeth due to agenesis and mean OHIP-14 score (3.3-point higher), although this finding did not meet the statistically significant threshold (p-value=0.064).

Table 7: Linear regression model for the association between oncology treatment factors and mean OHIP-14 scores.

Discussion

This cross-sectional study of the late dental effects of oncology treatment in childhood cancers in New Zealand found a high prevalence of dental agenesis and microdontia among our sample. This study also identified potential risk factors for agenesis including head and neck radiation therapy, stem cell therapy, cyclophosphamide equivalent doses over 8,000mg/m2, and dactinomycin.

While this study provides useful insights into the late dental effects of childhood oncology treatment, there are some weaknesses. The study had a nonideal participation rate of 37.8%. This may be attributed to the long interval between cancer treatment and the study (on average over 11 years), with the majority of non-participants declining to participate due to lack of interest. In addition to this, participation may have been inconvenient for those who had to travel long distances. The geographical spread of the participants also made the logistical organisation of this study difficult. Furthermore, Pasifika had a proportionally lower participation rate of 20% compared to other ethnicities. This is likely to be a reflection of the inequality in access to healthcare for these individuals.

There are a number of possible implications of the low participation rate. It may be that those with more dental problems had more incentive to participate in this study. Three-quarters of patients who were treated for rhabdomyosarcoma participated in this study and they had the highest mean total HDI score of 30.7, over six times greater than non-RMS diagnoses (4.7). However, only one of the three RMS originated in the head and neck region. This may have biased the results towards more severe dental problems than what is true in the population of interest.

The NZDep, and derived SES, was based on the current address and not the address at time of diagnosis. In addition, the NZDep is based on census data which uses meshblocks of, on average, 81 people.23 It is therefore not based solely on an individual’s SES, but includes the SES of others in that meshblock. While SES was not found to be a risk factor for the late dental effects in this study, this could be attributed to our small sample size.

This retrospective study looked at chronological ages of participants at diagnosis. The current literature shows that late dental effects are correlated with a patients age at treatment.8,27,28 While this study did not find any significant associations with age at diagnosis, it should be acknowledged that chronological age does not always correlate with dental age. This is an especially important consideration in this study as it has been shown that Māori and Pasifika children tend to have an accelerated dental development.29

Another consideration is that children with higher body mass index (BMI) values have also been found to have an accelerated dental development.30–32 High BMI values could also have been a confounding factor in this study as children with high BMI values may have received higher doses of chemotherapy, which is often calculated by body surface area.

Despite these considerations, this study provides useful and insightful information. This is the first study of its kind in New Zealand, and showed strong correlations of clinical importance between specific oncology treatment factors and their effects on the developing dentition. It involved an interprofessional collaboration between paediatric oncologists, nurses and dentists from all over New Zealand. Moreover, this study showed a respectable degree of rigour with seven dentists and dental specialists reviewing the information and arriving at a consensus. Furthermore, the participants in this study had comparable sociodemographic characteristics to the non-participants for age, sex and most ethnicities, allowing us to make some generalisations to our population of interest.

The results of the present study are consistent with the international literature, which shows that agenesis and microdontia are found more commonly among childhood cancer survivors. While this study is cross sectional, we can compare our prevalence data to the general population. This study found the prevalence of agenesis was high at 15.3%, compared with 2.5% to 6.9% in a recent meta-analysis of the general population.33 A previous study conducted in New Zealand found that 0.35% of five-year old children had missing teeth.32 The aetiology of agenesis is thought to involve both genetic and environmental factors.34,35 Another environmental factor to consider is maternal smoking, which has been linked to tooth agenesis and possibly childhood cancers. A study by Al-Ani et al found maternal consumption of 10 or more cigarettes per day during pregnancy was associated with four times the odds of having a child with hypodontia; as maternal smoking was not measured in the present study, this association was not able to be investigated.34,36 A similar comparison was found for microdontia (prevalence = 6.8%), with studies on the general population finding a prevalence of up to 3%.37 However, classification of microdontia can be quite subjective depending on which index is used and inter-observer variability.

Hsieh et al 2011 found a positive correlation between cyclophosphamide doses and total HDI.13 Although not statistically significant because of the small sample size, our study paper consolidates this finding. Our study also showed a dose-response relationship between total cyclophosphamide equivalent dose and number of teeth missing due to agenesis, and total HDI. This dose-response relationship between cyclophosphamide equivalent doses and the likelihood of agenesis persisted with multivariate analysis. While not statistically significant, it suggests that this finding is unlikely due to chance or any external factors.

Our findings corroborated the findings of Das et al,16 which found that head and neck radiation therapy was twice as likely to result in agenesis, and were missing five times as many teeth due to agenesis, than those who did not receive HNRT. In the present study this association persisted using the binary logistic regression model, and showed that HNRT greater than 20Gy increased the odds of agenesis seven-fold (OR=7.7) compared with those who did not have HNRT (p-value=0.04).

The association between specific treatment factors and agenesis can be attributed to the non-specificity of oncology treatments which potentially risk all cells undergoing division. It is thought that this disrupts ectomesenchymal cells involved in tooth formation resulting in agenesis.

Contrary to expectations, the mean OHIP-14 score of 4.3 was quite low in comparison with adolescents in the general population.38 The OHIP-14 has been validated in use for adolescents 15 years and older;39 with the mean age of this study’s participants being 14.9 years a child-specific OHRQoL measure such as the Child Perceptions Questionnaire (CPQ11–14) may have been found to be more responsive.40 The New Zealand Oral Health Survey (NZOHS) 2009 used a global question, similar to the one we used. In the present sample, children reported their OHRQoL to be better, compared with the 2009 NZOHS.26 While a linear regression model was carried out to ascertain correlations with mean OHIP-14 scores, we should acknowledge that by including sociodemographic, treatment factors and outcome variables, there is likely to be some level of confounding. Nevertheless, this model found that dactinomycin and agenesis correlated with a higher OHIP-14 score.

Ideally, the late dental effects of oncology treatment would best be studied longitudinally, thereby reducing bias and optimising participation rates. In addition, the role of dactinomycin should be further investigated for its potential effects on the developing dentition.

Conclusion

Although all children with childhood cancer need regular dental follow ups, particular attention should be paid to children who have had HNRT, SCT and/or cyclophosphamide equivalent doses greater than 8,000mg/m2. Particular attention should be paid to dental caries, agenesis and enamel defects, which have been shown to negatively impact OHRQoL. In addition to this, patients treated for rhabdomyosarcomas have the potential to have severe deleterious effects on their oral health. These are important considerations for paediatric oncologists, as patients who have missing teeth are likely to require complex and costly orthodontic and/or rehabilitative treatments.

This study shows that New Zealand survivors of childhood cancer treatment have a significant prevalence of complex dental anomalies. This justifies a review of primary and tertiary dental services to ensure they and their families have consistent publicly funded access to the orthodontic and advanced oral rehabilitative services which they require.

Summary

Abstract

Aim

The aim of this study was to determine the prevalence of dental developmental disturbances in long-term survivors of childhood malignancies in New Zealand children. This study reports associations with potential risk factors to inform oncologists and dentists of the likelihood of dental abnormalities.

Method

The study population was children aged 14–16 years old who were diagnosed with cancer prior to 10 years of age. A total of 156 children were eligible, of which 59 participated in this study. The indices used in this study were Holtta’s Defect Index (HDI), and Oral Health Impact Profile-14 (OHIP-14).

Results

The prevalence of agenesis was 15.3%, microdontia 6.8% and root abnormalities 32.2%. Cyclophosphamide equivalent doses above 8,000mg/m2, stem cell therapy (SCT), and head and neck radiation therapy (HNRT) were associated with a higher mean number of teeth missing due to agenesis. SCT and HNRT were associated with a higher total HDI. A binary logistic regression was carried out to determine the odds of agenesis and found that HNRT was the main contributing factor (OR=7.7, p-value=0.04). The linear regression model found that dactinomycin and agenesis correlated with the largest mean OHIP-14.

Conclusion

This study found that childhood cancer survivors in New Zealand had a high prevalence of developmental dental abnormalities and it identified potential risk factors related to their cancer treatment. Inequitable access to oral rehabilitation for this patient group argues for a mechanism for consistent improved access to publicly funded dental care across district health boards in New Zealand.

Author Information

Matthew Shum, Associate Dentist, Bishop’s Dental Surgery, Hawkes Bay; Erin Mahoney, Paediatric Dentist and Hon. Clinical Senior Lecturer, Dental Department and Faculty of Paediatrics, Hutt Valley DHB and University of Otago, Wellington; Kate Naysmith, Senior Dentist, Dental Department, Hutt Valley DHB, Wellington, Scott Macfarlane, National Clinical Leader—Child Cancer, National Child Cancer Network, Auckland; Rob Corbett, Paediatric Oncologist, Child Health, Canterbury District Health Board, Christchurch; Manisha Narsinh, Senior Dentist, Auckland Regional Hospital and Specialist Dentistry, Auckland DHB, Auckland; Arun Natarajan, Paediatric Dentist and Hon. Clinical Senior Lecturer, Hospital Dental Service and Department of Paediatrics, Christchurch Hospital, Canterbury DHB and University of Otago, Christchurch; Yaso Ramadas, Paediatric Dentist, Auckland Regional Hospital and Specialist Dentistry, Starship Hospital, Auckland DHB, Auckland; Elizabeth Hitchings, Clinical Leader, Dental and Oral Health, Capital and Coast DHB, Wellington; Heather Anderson, Paediatric Dentist, DayOne Paediatric Dental, Christchurch.

Acknowledgements

First, we would like to acknowledge the New Zealand Ministry of Health and the Australian and New Zealand Society of Paediatric Dentistry, which sponsored this research, and Colgate NZ who donated all of the toothbrushes and toothpastes for the participants to use. We would also like to extend an extensive thank you to the dentists who gave up their time to be calibrated and help perform the clinical and radiographic examinations throughout the country. We would also like to thank Kathy Yallop, Liz Summers, Jo Truscott and Angela Wakelin for their assistance in the organisation and coordination of this study. Finally, we would like to thank Dalice Sim, biostatistician who helped us conduct the statistical analysis and help us with all of our statistics-related questions.

Correspondence

Dr Erin Mahoney, Dental Department, Hutt Valley DHB, Private Bag 31907, Lower Hutt 5040.

Correspondence Email

erin.mahoney@huttvalleydhb.org.nz

Competing Interests

Dr Shum reports grants from University of Otago during the conduct of the study. Dr Naysmith and Dr Mahoney report grants from NZ Ministry of Health during the conduct of the study.

1. Steliarova-Foucher E, Colombet M, Ries LAG, et al. International incidence of childhood cancer, 2001–10: a population-based registry study. Lancet Oncol. 2017; 18:719–31.

2. Jemal A, Ward EM, Johnson CJ, et al. Annual Report to the Nation on the Status of Cancer, 1975–2014, Featuring Survival. J Natl Cancer Inst. 2017; 109.

3. National Center for Health S, Seer P, National Cancer I. Cancer statistics review. Cancer statistics review. 1973.

4. Holtta P, Alaluusua S, Saarinen-Pihkala UM, et al. Long-term adverse effects on dentition in children with poor-risk neuroblastoma treated with high-dose chemotherapy and autologous stem cell transplantation with or without total body irradiation.(Post-Transplant Complications). Bone Marrow Transplant. 2002; 29:121.

5. Cubukcu CE, Sevinir B, Ercan İ. Disturbed dental development of permanent teeth in children with solid tumors and lymphomas. Pediatr Blood Cancer. 2012; 58:80–4.

6. Lauritano D, Petruzzi M. Decayed, missing and filled teeth index and dental anomalies in long-term survivors leukaemic children: a prospective controlled study. Med Oral Patol Oral Cir Bucal. 2012; 17:e977.

7. Nemeth O, Hermann P, Kivovics P, et al. Long-term Effects of Chemotherapy on Dental Status of Children Cancer Survivors. Pediatr Hematol Oncol. 2013; 30:208–15.

8. Kang C-M, Hahn SM, Kim HS, et al. Clinical Risk Factors Influencing Dental Developmental Disturbances in Childhood Cancer Survivors. Cancer Res Treat. 2018; 50:926–35.

9. Hölttä P, Alaluusua S, Saarinen-Pihkala UM, et al. Agenesis and microdontia of permanent teeth as late adverse effects after stem cell transplantation in young children. Cancer. 2005; 103:181–90.

10. Proc P, Szczepańska J, Skiba A, et al. Dental Anomalies as Late Adverse Effect among Young Children Treated for Cancer. Cancer Res Treat. 2016; 48:658–67.

11. Duggal MS. Root surface areas in long-term survivors of childhood cancer. Oral Oncol. 2003; 39:178–83.

12. Cacchillo D, Barker GJ, Barker BF. Late effects of head and neck radiation therapy and patient/dentist compliance with recommended dental care. Spec Care Dentist. 1993; 13:159–62.

13. Hsieh SG, Hibbert S, Shaw P, et al. Association of cyclophosphamide use with dental developmental defects and salivary gland dysfunction in recipients of childhood antineoplastic therapy. Cancer. 2011; 117:2219–27.

14. Kaste SC, Goodman P, Leisenring W, et al. Impact of radiation and chemotherapy on risk of dental abnormalities. Cancer. 2009; 115:5817–27.

15. Thomaz EBAF, Mouchrek JCE, Silva AQ, et al. Longitudinal assessment of immunological and oral clinical conditions in patients undergoing anticancer treatment for leukemia. Int J Pediatr Otorhinolaryngol. 2013; 77:1088–93.

16. Das K, Hazarika M, Krishnatreya M, et al. An observational study of cancer treatment-induced dental abnormalities in paediatric cancer patients.(Original Article)(Report). Journal of Health Specialties. 2015; 3:228.

17. Jensen S, Pedersen A, Vissink A, et al. A systematic review of salivary gland hypofunction and xerostomia induced by cancer therapies: prevalence, severity and impact on quality of life. Support Care Cancer. 2010; 18:1039–60.

18. Malhotra V, Perry MC. Classical Chemotherapy: Mechanisms, Toxicities and the Therapeutc Window. Cancer Biol Ther. 2003; 2:1–3.

19. Al-Ani AH, Antoun JS, Thomson WM, et al. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017; 2017:9378325.

20. Ballantine K, & the NZCCR Working Group. The incidence of childhood cancer in New Zealand 2010–2014: A report from the New Zealand Children’s Cancer Registry. Auckland: National Child Cancer Network., 2017.

21. Ballantine K, & the NZCCR Working Group. Child cancer survival in New Zealand 2005–2014: A report from the New Zealand Children’s Cancer Registry. Auckland: National Child Cancer Network. 2017.

22. Slade GD. Derivation and validation of a short-form oral health impact profile. Community Dent Oral Epidemiol. 1997; 25:284–90.

23. Atkinson J, Salmond C, Crampton P. NZDep2013 Index of Deprivation. Department of Public Health, University of Otago, Wellington, Division of Health Sciences, University of Otago, 2014.

24. Children’s Oncology Group. Long-Term Follow-Up Guidelines for Survivors of Childhood, Adolescent, and Young Adult Cancers Version 5.0. Children’s Oncology Group, 2018.

25. Green DM, Nolan VG, Goodman PJ, et al. The cyclophosphamide equivalent dose as an approach for quantifying alkylating agent exposure: A report from the childhood cancer survivor study. Pediatr Blood Cancer. 2014; 61:53–67.

26. Haisman RJ. Our oral health key findings of the 2009 New Zealand Oral Health Survey. In: Mason K, Holmes E, Templeton R, Weerasekera D, New Zealand. Ministry of H, (eds). Wellington, N.Z.: Wellington, N.Z. : Ministry of Health, 2010.

27. Wilberg P, Kanellopoulos A, Ruud E, et al. Dental abnormalities after chemotherapy in long-term survivors of childhood acute lymphoblastic leukemia 7–40 years after diagnosis. Support Care Cancer. 2016; 24:1497–506.

28. Hutton A, Bradwell M, English M, et al. The oral health needs of children after treatment for a solid tumour or lymphoma. Int J Paediatr Dent. 2010; 20:15–23.

29. Te Moananui R, Kieser JA, Herbison P, et al. Advanced dental maturation in New Zealand Maori and Pacific Island children. Am J Hum Biol. 2008; 20:43–50.

30. Wong HM, Peng S-M, Wen YF, et al. A population survey of early factors associated with permanent tooth emergence: findings from a prospective cohort study. Community Dent Oral Epidemiol. 2016; 44:467–75.

31. Wong HM, Peng S-M, Yang Y, et al. Tooth eruption and obesity in 12-year-old children. J Dent Sci. 2017; 12:126–32.

32. Whittington BR, Durward CS. Survey of anomalies in primary teeth and their correlation with the permanent dentition. N Z Dent J. 1996; 92:4–8.

33. Polder BJ, Van’t Hof MA, Van der Linden FPGM, et al. A meta-analysis of the prevalence of dental agenesis of permanent teeth. Community Dent Oral Epidemiol. 2004; 32:217–26.

34. Al-Ani AH, Antoun JS, Thomson WM, et al. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017; 2017:9378325.

35. Ritwik P, Patterson KK. Diagnosis of Tooth Agenesis in Childhood and Risk for Neoplasms in Adulthood. Ochsner J. 2018; 18:345–50.

36. Al-Ani AH, Antoun JS, Thomson WM, et al. Maternal Smoking during Pregnancy Is Associated with Offspring Hypodontia. J Dent Res. 2017; 96:1014–9.

37. Jeong KH, Kim D, Song YM, et al. Epidemiology and genetics of hypodontia and microdontia: a study of twin families. Angle Orthod. 2015; 85:980–5.

38. Colussi PRG, Hugo FN, Muniz FWMG, et al. Oral Health-Related Quality of Life and Associated Factors in Brazilian Adolescents. Braz Dent J. 2017; 28:113–2.

39. Ravaghi V, Ardakan MM, Shahriari S, et al. Comparison of the COHIP and OHIP- 14 as measures of the oral health-related quality of life of adolescents. Community Dent Health. 2011; 28:82–8.

40. Foster Page L, Gilchrist F, Broder HL, et al. A Comparison of Three Child OHRQoL Measures. Dent J (Basel). 2019; 7(1).1

Contact diana@nzma.org.nz
for the PDF of this article

Subscriber Content

The full contents of this pages only available to subscribers.

LOGINSUBSCRIBE