Personalizing dental screening and prevention protocols in dentulous patients with oropharyngeal cancer undergoing radiotherapy: A retrospective cohort study

Highlights • Post-radiotherapy teeth extractions within 5 years occurred in 31 %.• Main indications radiation caries and periodontal disease.• Dose factors and life-style factors related to radiation caries extractions.• High-dose volumes in mandible related to periodontal disease extractions.


Introduction
Radiotherapy (RT) is an important treatment modality for head and neck cancer.It is associated with various oral side effects, such as dry mouth, dysphagia, dental problems, and osteoradionecrosis (ORN) [1].When treating oropharyngeal cancer (OPSCC) with RT, the proximity of the tumour to the mandible is often associated with substantial exposure to intermediate-to-high radiation doses.This may trigger a pathophysiological process leading to ORN, a severe complication of RT in patients with head and neck cancer [2].According to Wanifuchi et al. (2017), ORN is most frequently observed in the molar region, particularly in the mandible [3].
In addition to other risk factors, such as smoking and dental hygiene, several studies have established an association between post-RT tooth extractions and an increased lifetime risk of developing ORN [1][2][3][4][5][6][7][8].Therefore, patients with head and neck cancer undergo routine dental screenings prior to RT, in order to perform necessary pre-RT extractions to prevent post-RT complications [9].
Despite the implementation of preventive protocols and guidelines, post-RT extractions are frequently needed [8].Currently, little is known about the incidence and baseline factors predictive of post-RT tooth extraction in a population undergoing RT.The purpose of this retrospective cohort study was to address this knowledge gap by reporting the incidence of post-RT tooth extraction and identifying patients at an increased risk of developing ORN.The clinical relevance of this study is that by enhancing our knowledge, we may be able to further personalise dental screening and prevention protocols to prevent post-RT tooth extraction and associated risk of ORN.

Patient cohort
The study protocol was reviewed and approved by the Erasmus MC Medical Ethics Review Committee (EMC17404).This retrospective cohort study reviewed patients diagnosed with OPSCC (T1-4, N0-3, M0) who received (chemo-) RT with curative intent at the Department of Radiotherapy of the Erasmus Medical Center between January 2009 and May 2016.For the current study, we included dentulous patients from this cohort with at least 1 year of follow-up record after the last RT session.Furthermore, we excluded patients who were diagnosed with other malignancies within an approximately 6-month window and those who had previously undergone RT in the head and neck region (more details in Fig. 1 and Table 1).This study complied with the strobe protocol.

Treatment and follow-up
During weekly multidisciplinary meetings, the diagnoses and treatments of patients with OPSCC were assessed and discussed.Subsequently, the selected patients were treated with RT following standard clinical protocols.Each patient with head and neck cancer who was scheduled to undergo RT received pre-RT dental care.The assessment of oral hygiene occurred prior to radiation therapy and was conducted by a dental hygienist.The evaluation was based on plaque scores and periodontal conditions, specifically assessing bleeding and swelling.The resulting score categorized oral hygiene as either poor, borderline, or sufficient-good.If necessary, pre-RT dental extractions were performed according to the national guidelines.The current decision pattern outlined in this guidelines concerning removal of dental foci is based on empirical evidence and experience.Useful tools for this purpose include 'dental risk factors' and'malignancy-related risk factors' [10,11].Based on the evaluation of dental and malignancy-related risk factors, the presence of strategic teeth, and overall clinical impression, a decision is made regarding which foci should be removed to prevent post-RT tooth extractions and the associated risk of ORN (Dijkstra et al.Prior to RT, a thermoplastic mask was fitted to immobilise patients during RT.All patients received a total dose of 46 Gy delivered in 2 Gy fractions to the tumour location and indicated elective node regions in the neck area, followed by a boost to the specific tumour location and pathological lymph nodes.The boost was administered as a dose of 24 Gy delivered in 12 fractions of 2 Gy or 16.5 Gy delivered in 3 fractions of 5.5 Gy (a hypofractionated boost), which was only applicable in case of small tumour volumes.The intensity modulated radiotherapy (IMRT) was the selected technique of application.During treatment planning, effort was made to minimize radiation exposure to the organs at risk surrounding the target area.The mean dose to the contralateral parotid gland was maintained below 26 Gy when feasible, which allowed sparing of the parotid gland to preserve its salivary function [12].Additional dose constraints for organs at risk within the oral cavity included mean dose of < 39 Gy and < 50 Gy to the submandibular glands and total oral cavity, respectively, whenever achievable.
Patients with large cT3-T4N0-2 M0 and cT1-4N3M0 tumours underwent a treatment regimen involving IMRT.The macroscopic tumour and pathological lymph nodes received a total dose of 70 Gy, delivered in five to six fractions of 2 Gy per week.In addition, elective lymph node regions were treated with a dose of 46 Gy using IMRT (referred to as the IMRT boost group).
For certain cases meeting specific criteria (cT3-T4 or N + and age ≤ 70), the treatment was further enhanced by including cisplatin or cetuximab.Cisplatin was administered at 100 mg/m 2 on days 1, 22, and 43.Alternatively, cetuximab was administered at an initial dose of 400 mg/m 2 , followed by a weekly dose of 250 mg/m 2 .
In the context of IMRT, the target coverage objective was to achieve a high level of dose distribution.Specifically, the goal was to ensure that a minimum of 98 % of the planning target volume received at least 95 % of the prescribed dose.More details on the treatment can be found elsewhere [13,14].
During RT, all patients regularly visited a dental hygienist and received weekly fluoride treatment using personalised teeth caps.Follow ups consisted of six visits in the first-year post RT.After the first year, follow ups decreased to a minimum of two visits per year for at least 5 years.Patient and their personal dentists were informed after the RT course that in case of indications for dental extractions, teeth had to be preferably extracted at the Oral and Maxillofacial surgery department of the Erasmus Medical Centre.Post-radiation extractions were conducted in accordance with clinical guidelines, with post-radiation tooth extraction mainly indicated when a tooth had a poor prognosis due to various reasons as listed in Table 2.When the radiotherapy dose in the region of the post-RT extraction had exceeded 40 Gy, a prophylactic strategy against osteoradionecrosis (ORN) was started.Antibiotic prophylaxis for extractions in an irradiated jaw included the protracted and heightened administration of amoxicillin (with clavulanic acid) at a dosage of 500 (+125) mg three times daily or clindamycin 600 mg three times daily.Antibiotics were started one day before treatment and continued for 10 days.Furthermore, it was suggested to patients to undergo hyperbaric oxygen therapy as an possible prophylactic strategy.Subsequent follow ups for these extractions was performed 2 weeks after extraction and 2-3 months after completion of RT via clinical examination and panoramic radiography.

Data collection post-RT extractions and risk factors
Required data on post-RT extractions were collected on a separate case record form developed for the current study.This data collection took place without knowledge of the baseline data since these data were already collected earlier for this cohort.Medical records were retrospectively reviewed and radiographic images were compared obtained before and after RT (panoramic radiographs).Data included dates of extractions, the teeth extracted, locations, and indications for extraction.We evaluated patients' data for up to 7 years of follow up, as follow-up data beyond this time point was not available for most patients.Data on the following were collected: radiation-induced caries, periodontal disease, periapical radiolucency, chipped teeth (excluding radiation-induced caries or periodontal disease or radiolucency), caries of a single element without indication of radiation caries, and other relevant factors.To determine the main reason for tooth extraction, the reasons provided in the medical correspondence were reviewed; if no clear reason was mentioned, patient medical records were reviewed to identify the main reason.

Evaluated risk factors
The variables that were tested in this study were based on the existing literature regarding risk factors associated with dental problems and ORN.Notably, little literature is currently available concerning risk factors for post-RT tooth extraction [10,11].The evaluated factors included baseline clinical and dosimetric factors, as well as oral health status, as reported during the pre-RT visit for dental focus screening.The evaluated dose variables included mean dose delivered to the contralateral parotid gland, relative volume of the mandible receiving a dose > 60 Gy / 70 Gy / 80 Gy, and mean dose delivered to the mandible.All dose variables were calculated using an α/β ratio of 3 Gy to account for late effects.

Statistical analyses
The time interval was calculated from the day of the final RT fraction.Follow up was concluded after 7 years (end of the evaluated follow up), at the time of death, entry into a palliative phase, diagnosis of a new primary tumour, or last hospital visit (whichever occurred first).Cox regression was used to evaluate potential risk factors for post-RT tooth extraction.For constructing the multivariable model, all risk factors with p < 0.2 in the univariable analysis were incorporated into a forward conditional selection procedure.The Kaplan-Meier method was used to calculate the cumulative incidences.Statistical significance was set at p < 0.05, and all statistical analyses were performed using the SPSS software (version 25, IBM Corporation, Armonk, NY, USA).

Study population
Among the 479 patients diagnosed with OPSCC who were scheduled for RT at our department between January 2009 and May 2016, 424 completed RT with curative intent and were eligible according to our general set of exclusion criteria (Fig. 1).Within this group, we selected patients who had at least 1-year disease-free follow-up available and those who were dentate following RT, resulting in a final study group including 174 patients (Fig. 1).The median follow-up period was 4.9 years.Demographic and baseline characteristics of the study population are presented in Table 1.The mean age was 59.5 year; 66 % of the patients were men, and 43 % of them had undergone partial pre-RT extractions.Poor oral hygiene at baseline (at the visit for pre-RT dental screening) was noted in 23 % of the patients.

Risk factors
In the univariable analysis, we identified several baseline risk factors significantly associated with an increased risk of post-RT tooth extraction (Table 3).These risk factors included alcohol abuse (hazard ratio [HR] = 2.27), poor oral hygiene (HR = 2.52), and relative volume of the mandible receiving a dose > 60 Gy (V60, HR = 1.36 per 10 % volume increase; V70, HR = 2.66 per 10 % volume increase; and V80, which indicated the presence of small volumes of very high dose hotspots, HR = 2.00 per 1 % volume increase).
When evaluating post-RT extractions related specifically to radiation caries, we identified the following risk factors: active smoking (HR = 2.95), alcohol abuse (HR = 4.60), poor oral hygiene (HR = 3.95), mean dose to the contralateral spared parotid gland (HR = 1.85 per 10 Gy), and all evaluated dose parameters of the mandible (highest HR of 6.18 observed for V70) (Table 3).For post-RT extractions associated with periodontal disease, we observed a correlation with the hotspot variable V80 when evaluating extractions that included the mandible.This correlation was observed in 16 of the 24 patients who underwent extraction owing to periodontal disease.
In multivariable analysis focusing on the endpoint of radiation caries, the following factors remained significant in the model (Table 4): V70 (HR = 8.30), active smoking (HR = 2.33), and poor oral hygiene (HR = 5.49).

Osteoradionecrosis
We observed 26 cases of ORN for which follow-up data were available (n = 24, mandibles; n = 2, maxilla).Among them, 14 patients developed ORN specifically after pre-RT extractions, of whom 6 also underwent post-RT extractions later (in two patients, the post-RT extractions were related to the previously developed ORN).Seven patients developed ORN after post-RT extraction, and two of them underwent both pre-and post-RT extractions.Five patients developed ORN with no previous history of pre-or post-RT extractions, and four of these patients later underwent post-RT extractions (in one patient, post-RT extraction was mainly related to ORN).

Discussion
In the present retrospective cohort study, we evaluated the incidence and risk factors for post-RT extractions in dentate patients diagnosed with OPSCC who had previously undergone curative-intent RT.We observed a cumulative incidence of 31 % at 5 years of follow up.
Exclusively, the OPSCC group was intentionally chosen to ensure homogeneity within the study population.Significant divergences in treatment modalities and chronological sequences exist between OPSCC and oral cavity cancer.This methodological selection aims to enhance internal validity.The primary reasons for extraction were radiation caries and periodontal disease.The personalised risk profile of patients who had undergone post-RT extractions related to radiation caries differed considerably from the risk profile for extractions related to periodontal disease.For radiation caries, we identified multiple clinical and dosimetric variables, whereas for periodontal disease, only the presence of hotspots in the mandible correlated with this endpoint.To our knowledge, this is the first study to investigate the risk factors for post-RT tooth extraction classified by the cause of dental failure.
The prognostic value of small high-dose volumes for dental late toxicity has been previously reported in this study cohort regarding osteoradionecrosis (ORN).Baker et al and Verduijn et al reported that these high-dose volumes, particularly present in patients treated with stereotactic hypofractionated boost (SBRT), were predictive for ORN [13,15].In our current study, we also find this variable to be prognostic for post-RT extractions.However, the variable SBRT itself did not emerge as a significant univariable predictive factor.This can be partially attributed to the fact that selected SBRT patients are treated for small tumours, in contrast to non-SBRT patients who do not meet the criteria for small tumour eligibility for SBRT.Verduijn et al similarly found SBRT not to be a univariable predictive factor for ORN due to this phenomenon.Therefore, we conclude that high-dose volumes associated with SBRT increase the risks of late dental toxicities in a selectively treated patient population with relatively small tumour volumes and relatively low baseline risks for late dental toxicities.
The national guidelines recommend pre-RT oral assessment and associated extractions based on dental and malignancy-related risk factors to minimise the likelihood of post-RT tooth extraction, thereby preventing the risk of developing ORN and other associated complications (Dijkstra et al. 2004).In our study, approximately one-third of the dentulous patients required post-RT teeth extraction which was not initially considered at risk before RT.The findings of this study, along with current guidelines, could be valuable to inform patients about the increased risk of post-RT extraction at an early stage, support them with specific precautionary measures, and aid in making decisions regarding pre-RT tooth extraction for individual cases.In our cohort, seven patients developed ORN after post-RT extraction, suggesting that incorporating personalised and patient-specific risk factors into the current dental screening protocols might have prevented these cases.
Post-RT tooth extraction was primarily indicated for periodontal disease and radiation-induced caries.We observed correlations between risk factors, including active smoking, alcohol abuse, poor oral hygiene, a higher mandibular dose, a higher mean dose to the contralateral spared parotid gland, and an increased risk of extractions caused by radiation caries.Recently, Brennan et al. (2022) reported the incidence and risk factors of tooth failure (with 24 months of follow-up) in a prospective cohort study including 572 patients with head and neck cancer [16].They reported a cumulative incidence of 18 %, which closely aligns with 17 % observed in our study at 2 years.They identified various associated risk factors, including a low number of teeth at baseline, older age, active caries at baseline, proton therapy as a preventive factor, RT dose, oral hygiene compliance, and salivary flow.Notably, they did not specify the different types of tooth failure as we did in our study, nor did they examine specific dose levels in the parotid glands and mandible.A relationship between the mean dose to the spared parotid gland and radiation caries has been previously reported in patients with head and neck cancer patients by Hey et al. and Gomez et al. [17,18].In agreement with these results, we observed a significant relationship between the mean dose to the spared parotid gland and risk of post-RT extraction due to radiation-induced caries.Gomez et al. also reported that high-dose levels in the mandible are associated with post-RT dental extractions [17].A relationship between alcohol abuse and periodontal disease has been reported in literature [19]; however, no specific association has been observed for radiation-induced caries, as observed in our study.In our study, alcohol abuse was identified a significant factor in the univariable analysis for post-RT extraction due to radiation-induced caries; however, it was not included in the multivariable model, possibly because of its high correlation with active smoking and poor oral hygiene.
Patients who receive RT in the head and neck area are at an increased risk of developing periodontal disease compared to the general population 1 .RT causes hyposalivation and changes in oral microbiome, which may lead to a loss of salivary protection, leading to periodontal disease [1,10].In our study, we observed that the molars were the teeth most frequently extracted after RT, with periodontal decay being the main reason for their extraction.This location also carries a greater risk # For all patients with radiation caries, elements in the mandible were removed * Only extractions in the mandible were used for analysis of mandibular dose variables (n = 16 events).Abbreviations: HR=Hazard Ratio, RT=radiotherapy, SBRT=stereotactic boost radiotherapy, 95% CI=95% confidence interval.

Table 4
Results of multivariable regression analysis for the endpoint of post-radiation teeth extraction indicated by radiation caries (N = 18 events).
Among the 174 patients, 49 underwent post-RT tooth extraction, totalling 62 sessions (n = 37, one session; n = 11, two sessions; n = 1, three sessions).Characteristics of the extraction sites are listed in Table 2.Among these patients, 12 underwent extraction of one tooth, 16 underwent extraction of two or three teeth, 7 underwent extraction of four to nine teeth, and 14 underwent extraction of more than ten teeth, including 11 cases where the entire dentition was extracted.Most dental extractions were performed during the first 4 years of RT (Fig. 2).The estimated cumulative incidence of post-RT tooth extraction at 5 years was 30.7 % (3.8 % 1SE).

Table 1
Baseline characteristics of the study population (N = 174).

Table 2
Details of post-RT tooth extractions within 7 year of follow-up (N = 49 patients).

Table 3
Univariable regression analysis of baseline factors for post-radiation teeth extraction.