Late Toxicities after Conventional Radiotherapy for Nasopharyngeal Carcinoma : Incidence and Risk Factors

Background. To determine the incidence and analyze the factors affecting late toxicity for nasopharyngeal carcinomapatients treated with conventional radiotherapy. Patients and Methods. Retrospective analysis was performed on 239 NPC patients treated between 1993 and 2004 in our institution. One hundred and fifty-seven patients were treated with conventional fractionation (2Gy per fraction, 5 fractions per week) and eighty-two patients with hyperfractionated radiotherapy (1.6Gy per fraction twice a day, 5 days per week). One hundred fifty nine patients underwent neoadjuvant cisplatin based chemotherapy. Late toxicity was evaluated according to the RTOG/EORTC score. Results. Xerostomia was the most common related complication (98.7%). Neoadjuvant chemotherapy and hyperfractionated radiotherapy did not increase late toxicities. Multivariate analyses showed that radiation dose was a significant factor for hearing impairment, younger age for trismus, initial node status for neck fibrosis, and initial dental hygiene for dental complications. Female gender was associated with significantly higher incidence of trismus and hearing impairment.Conclusion. Conventional radiotherapy was associated with a high rate of late toxicities which affect patients’ quality of life.With the development of three-dimensional conformal radiotherapy and intensitymodulated radiotherapy, a reduced incidence of radiation related complications could be expected.


Introduction
In Tunisia, nasopharyngeal carcinoma (NPC) is the most common head and neck cancer with an annual incidence rate of 3.8 out of 100000 in South Tunisia.Due to anatomic restrictions and the high degree of radiosensitivity, radiotherapy is the standard treatment for NPC.A total dose of about 70 Gy is generally recommended.Chemotherapy also plays a role in the treatment of advanced NPC.With the anatomic proximity to critical structures, radiotherapy for NPC is associated with late complications.The objective of this study is to determine the incidence of late toxicity after conventional radiotherapy for nasopharyngeal carcinoma and to analyse the risk factors of occurrence of these toxicities.
This study will allow us to compare, in the future, the rates of late toxicities after conventional radiotherapy with those of conformal RT and IMRT.Therefore, we can see the contribution of these new techniques in reduction of late complications incidence.

Patient Characteristics.
From February 1993 to December 2004, 314 patients with nonmetastatic NPC were treated at our institution.Among them, 239 patients were included in this retrospective study.Eligibility criteria for inclusion were regular followup and complete remission at least one year after radiotherapy.
The patients' mean age was 43 years .One hundred and fifty-one were males (63%) and 88 were females (37%).The median followup was 111 months .All patients were staged according to the 1997 American Joint Committee on Cancer (AJCC) staging system after clinical examination and CT and/or MRI imaging.
Seventy-three percent had stage T3-T4 tumors and 60% had N2-N3 nodal stage (Table 1).Abdominal ultrasound, bone scintigraphy, and chest radiology were performed for all patients to eliminate distal metastasis.Before starting treatment, all patients had a dental workup, and fluoroprophylaxis has been recommended for life but not regularly practiced.

Radiotherapy Modalities.
All patients in this study underwent radiotherapy with Cobalt 60 gamma photons using conventional techniques.Two laterally opposed fields were used to treat the nasopharynx and the upper neck and an anterior cervical field with a midline shield was used to treat the lower neck to 38.4-40 Gy.Thermoformable mask was used for contention.The remainder dose of irradiation was delivered via an anterior nasal field which included the nasopharynx and a large anterior cervical field treating all cervical nodes.
External beam radiotherapy was delivered with two modalities.One hundred and fifty-seven patients were treated with conventional fractionation and 82 patients with hyperfractionated radiotherapy (1.6 Gy × 2/day with an interval of 6 hours, 5 days per week).The hyperfractionated schedule was used in a previous randomized phase III trial [1].The total dose to the nasopharynx and involved neck areas was 70-75 Gy and was 50-55 Gy to the remaining cervical areas from level II to level V.
Patients with locally advanced nodal disease (N2-N3) received 2 to 3 cycles of neoadjuvant chemotherapy using a cisplatin regimen.The drugs used in combination with cisplatin were detailed in Table 2.

Assessment and Statistical Analysis.
Patients were followed every 3 months during the first 2 years, every 6 months for the next 3 years, and then every year until death.The overall 5-and 10-year survivals were, respectively, 75% and 62%.Biological tests of serum FT4 and TSH were performed annually.Late toxicity was recorded on every visit and evaluated with the RTOG/EORTC late radiation morbidity scoring criteria for xerostomia and fibrosis.Quality of life and professional activity were evaluated with a subjective questionary: (1) poor satisfaction; (2) satisfactory; (3) good satisfaction.
All events were measured from the date of start of radiotherapy.The SPSS program 15.0 was used for statistical evaluation.The frequency of toxicities was compared with  2 test.The actuarial rates were calculated using the Kaplan Meier method, and the differences were compared with the log-rank test.A multivariate analysis of different risk factors was conducted using the Cox proportional hazard model.
Multivariate analysis further confirmed the independent value of gender and radiation dose.Trismus was noted in 104 patients; the 5-year-CFS rate was 57%.Patients younger than 20 years, females, and advanced T stage had a higher incidence of trismus.Multivariate analysis confirmed that gender ( = 0.01) and age ( = 0.0001) were independent predictors for trismus.
A total of 143 patients had fibrosis of the neck (Figure 2).It was grade 1 in 29.7%, grade 2 in 23%, and grade 3 in 7.1%.The 5-year-CFS rate was 45%.Patients with nodal involvement had a higher incidence of fibrosis (66.5% versus 41.3;  = 0.0001).Females and patients treated with associated chemotherapy had also developed more neck fibrosis.Only initial node status was demonstrated as independent predictor factor for neck fibrosis.
Dental complications were observed in 51.7% of cases, represented by pains (26%), breaks (12%), falls and mobility (14%), caries (35%), and dental and alveolar radionecrosis (5%).Their occurrences were often late with average delays greater than 4 years.The rate of dental complications increases with time and rose from 16% to 74% at 1 year to 7 years (Figure 3).Initial poor dental hygiene, female gender, and use of chemotherapy were significant risk factors of higher incidence of dental toxicity.Multivariate analysis retained the independent value of initial dental hygiene.Forty-nine patients (20.6%) developed bone necrosis.It was mainly mandibular and maxillary radionecrosis, but we noted skull base, sphenoid, and cervical (C1, C2) radionecrosis.Patients with an initial poor dental hygiene had a greater rate of bone necrosis (8.5% versus 2.4%;  = 0.04).Multivariate analysis showed that none of the risk factors had a statistically significant effect.Neurological complications were observed in 21 patients (8.8%) such as temporal necrosis in nine cases, brain stem necrosis in five, optic nerve atrophy in two, and myelitis in one case (Table 3).Hyperfractionated radiotherapy was associated with a higher incidence of neurotoxicity (14.6% versus 5.7%;  = 0.02).The rate of temporal lobe necrosis was significantly greater in younger patients (9.8% versus 2.5%;  = 0.04).On multivariate analysis, hyperfractionation and age were insignificant.
Thirty-six patients (15%) developed erysipelas.Four patients (1.6%) developed radioinduced cancer.The mean period was 60 months (36-122 months) after the beginning of radiotherapy and the pathological diagnosis included glioblastoma, basal cell carcinoma of the internal angle of Nasal mucosa complications were found in 77% of cases, represented mainly by nasal obstruction, nasal voice, nasal discharge, and epistaxis.Other complications include a case of nasal reflux of liquids and a case of soft palate necrosis.Oral pharyngeal mycosis was observed in 45% of cases after a mean period of 33 months.The oral hygiene prior to initial radiotherapy was found as a significant factor in occurrence of fungal infection;  = 0.034.
Satisfactory professional activity after the end of treatment was found in 30% of cases and quality of life was satisfactory in 45% of cases.In multivariate analysis, factors that affect the quality of life are age, xerostomia, and dental complications.
The results of uni-and multivariate analyses of late toxicity risk factors are summarised in Table 4.

Discussion
High doses of radiotherapy to the nasopharyngeal tumor and involved nodes are necessary to obtain locoregional control of the disease.Because of the close anatomic proximity to critical structures, NPC is one of the most difficult and challenging cancer sites for radiation oncologists.In fact, radiation treatment is associated with several acute and late toxicities.A few studies were interested in an exhaustive analysis of late toxicities after radiotherapy for NPC [2][3][4].The majority of authors reported only one pattern of these toxicities [5][6][7][8][9][10][11].Our study presents one of the most exhaustive analyses of all late toxicities observed after conventional radiotherapy for NPC and their risk factors.
Xerostomia is the most common toxicity after radiotherapy for nasopharyngeal carcinoma [2][3][4].Yeh et al. showed that the 5-year-xerostomia-free rate was 9.7% and elderly patients had a significantly higher incidence of xerostomia [4].In our study, incidence of xerostomia was 98.7% and no risk factor was showed.
Xerostomia can last for months or even years after irradiation.The recovery of salivary function is variable and depends on the volume of salivary glands irradiated, total dose, and individual susceptibility.The association with neoadjuvant chemotherapy may increase the effects of ionizing radiation on the salivary glands [12].However, the effect of concomitant chemotherapy on late salivary toxicity was not assessed, due to the lack of distance.In our series, improved salivary function was observed in 10% of cases.Sex and fractionation have been identified as risk factors in univariate analysis.The oral dryness is responsible for difficulties in mastication, digestion, swallowing, and speech and for dental damage.These complications severely affect quality of life of these patients [13].
Several therapeutic approaches have been proposed to protect the salivary glands, prevent oral dryness, and improve the quality of life of irradiated patients.Amifostine is a radio protector which has been widely used with radiation therapy of head and neck cancers.Brizel et al. [14] showed in a randomized phase III trial that conventional radiotherapy combined with amifostine significantly reduced the rate of late xerostomia 23% compared with radiotherapy alone with no detrimental effect on local control of the disease, progressionfree survival, and overall survival.This benefit was valid both for low (≤40 Gy) and high radiation doses (>40 Gy) [14].In a randomized phase III trial including 132 patients with head and neck cancer, Buentzel et al. conclude that reducing the rate of late xerostomia (grade ≥ 2) was not significant between the combination of concurrent radiochemotherapy with amifostine and radiochemotherapy alone (37% versus 24%) [15].A meta-analysis, including 14 randomized trials (1451 patients), compared the results of radiotherapy alone with radiotherapy and amifostine [16].The use of amifostine in head and neck cancer significantly reduces (67%) the risk of grades 2 and 3 late xerostomia ( < 0.00001) [14,17].The benefit of amifostine was retained for patients receiving radiotherapy alone.It reduces the risk of toxicity and improves tumor response rate.However, the benefit of amifostine during concomitant chemoradiotherapy was not retained.
The role of pilocarpine in the preservation of salivary function is controversial [18,19].A randomized trial including 170 patients with head and neck cancer showed that concomitant administration of pilocarpine with radiotherapy does not reduce the incidence or severity of xerostomia.Assessment of salivary function was mostly clinical (partly subjective) using the most commonly used scales (SOMA-LENT or EORTC/RTOG).However, the methods of objective measurement of salivary function have been developed in recent years by scintigraphy [20] or sialography [21,22].Some studies have used scintigraphy of the salivary glands (submandibular and/or parotid) as a method of quantitative measurement of the change in salivary function (secretory and excretory) before and after radiotherapy for head and neck cancer [20,23].The comparison of subjective clinical data (RTOG score) and scintigraphic (objective data) found no significant correlation [20,24,25].
Conformational radiotherapy with intensity modulation (IMRT) can increase the dose gradient between the target volume (CTV) and salivary glands.Dose tolerance of the parotid gland varies from 25 to 30 Gy [26].Indeed, Pow et al. [27] showed in a randomized trial including 51 patients with nasopharyngeal cancer, that IMRT is superior to conventional radiotherapy in terms of preservation of salivary function and improved quality of life.These results are consistent with several other tests [28,29].Lee et al. reported a decreased incidence of late oral dryness grade greater than or equal to 2 in the group of patients treated with IMRT compared to those treated with conformal radiotherapy (67% versus 12%); the difference was statistically significant [28].
A single-center retrospective study, which compared the change in salivary function after conventional radiotherapy with or without amifostine to intensity modulated radiation therapy in 75 patients with head and neck cancer, showed better preservation of the function saliva with IMRT ( < 0.05).Assessment of salivary function (secretory and excretory) was made objectively by scintigraphy of the parotid.The IMRT is currently considered the best therapeutic option for the protection of salivary function.New therapeutic approaches are being evaluated for the combination of IMRT and amifostine [23].
Postirradiation neck fibrosis is one of the most common late sequelae for NPC patients.The incidence ranges from 40 to 70% in the literature and is correlated with the dose in the neck itself (less than or equal to 70 Gy), dependent on the initial size of positive lymph nodes [2].In our study, 60% of patients had neck fibrosis, which was associated with positive neck lymph nodes as independent factor.Neoadjuvant chemotherapy was associated with a significant higher incidence of neck fibrosis.In fact, many drugs, particularly Bleomycin, were incriminating to increase fibrosis rate.The study of Yeh et al. showed that 38% of patients had neck neck fibrosis.Gender, initial node status, and radiation dose (>62 Gy) were further demonstrated as independent predictor factors [4].Several authors have shown that the combination of concurrent chemoradiation increases the risk of fibrosis [30].Hyperfractionated radiotherapy does not increase the rate of cervical sclerosis in our series, which is consistent with other publications [31].
Hypersensitivity of teeth is a common side effect of radiotherapy.The postradiation dental lesions are not due to a direct action of ionizing radiation on the teeth but rather are secondary to oral dryness and changes in the bacterial flora of the oral cavity.This can be explained by the decrease in salivary flow and changes in the composition of saliva and its pH [32].Postradiation-dental-complications are dominated by caries that cause dental falls and osteoradionecrosis.Darkening and dental fractures were also described.To our knowledge, the incidence of dental injuries after radiotherapy for nasopharyngeal cancer was not assessed in the literature.The etiology of postradiation caries is multifactorial: the radiomucositis, the oral dryness, poor dental status before treatment, and the lack of fluoride protection are the main risk factors for dental caries [33].In our series, the complication rate was 51.7% dental.Thirty percent of our patients had poor dental status before treatment and oral hygiene unsatisfactory.Only 20% used their fluoride gel trays door on a regular basis after the end of irradiation.This could explain the high rate of dental complications.In multivariate analysis, dental status before radiation therapy was identified as independent risk factor for dental complications.The influence of chemotherapy in the occurrence of dental complications was not assessed in the literature.However, a significant increase in the incidence of dental complications was observed in our study ( = 0.03).Oral radiation must be preceded by a preradiotherapy dental care and the use of fluoride gel [34,35].Dental extractions should be done 2-3 weeks (minimum 10 days) before the start of radiation therapy to allow for cicatrisation [12].The care plan should include frequent teeth brushing with a soft brush and use of toothpaste or a fluoride gel on a daily basis to prevent accumulation of plaque and caries [36].Rudat et al. showed, in a retrospective study on 35 patients, the protective role of amifostine on dental health after radiotherapy for head and neck cancer [37].
Osteoradionecrosis is a severe late complication of radiation therapy of the nasopharyngeal cancer.The hypo vascularization, hypoxia, and hypocellularity induced by ionizing radiation result in a reduced capacity of bone turnover which promotes the risk of necrosis [38].The incidence of mandibular and jaw osteoradionecrosis after nasopharyngeal carcinoma irradiation varies from 1.8 to 22% [39,40].In our series, the incidence was 13%.The lack of oral hygiene and daily use of fluoride gel are the predisposing factors for osteoradionecrosis [38].The high dose of radiation was considered by several authors as a risk factor for osteoradionecrosis [41,42].Indeed, beyond 70 Gy, the risk increases with the dose [42,43].Time between the end of radiotherapy and dental extractions has not been clearly defined.The extraction will be done under antibiotic coverage, alveolectomy, and suture gum [44].
If mandibular and jaw radionecroses have been widely described in the literature, few studies have reported skull base osteoradionecrosis after radiotherapy for nasopharyngeal cancer [9,45].It was 7.1% in our population after an average period of 81 months.
Radionecrosis of the skull base is often asymptomatic.Headache and epistaxis may be revealing signs [10].The systematic radiological examinations during the monitoring (CT or MRI) away from the irradiation of the nasopharynx could find asymptomatic radionecrosis.The lesions are most often at the sphenoid bone and clivus.The diagnosis of osteoradionecrosis is based on a set of arguments: clinical, radiological, and endoscopic [10].A comparative reading of nasopharyngeal imaging before and after treatment must always be made to ensure that bone lesions did not exist before the start of treatment and the patient is in remission from his tumor.This is the case for our patients.Most authors require histological examination to confirm the diagnosis and rule out possible tumor recurrence [10,45].
Neurotoxicity is serious late sequelae in NPC patients treated with radiotherapy.For patients treated by hyperfractionated radiotherapy, more neurological damages were noted in our experience.Teo et al. had reported in a randomised trial a significant increase of neurological complications especially temporal lobe injury after an accelerated radiation therapy in patients with NPC (49.4% versus 23.2%;  = 0.001) [31].These concerns were also reported by Jen et al. [46].Yeh et al. showed that higher external beam radiation dose (>72 Gy) was associated with the development of temporal necrosis for NPC patients [4].These results are also found by Lee et al. in their retrospective study [47].However, in our study, the differences between the radiation doses 75 and 70 Gy were not significant.Younger patients had a higher incidence of temporal necrosis.It had not been reported in the literature, and more work should be made to prevent neurotoxicity in these patients.
Ototoxicity, particularly sensorineural hearing loss, is a known complication after radiotherapy for NPC.Using chemotherapy, particularly cisplatin regimen, has been considered as risk factor for otologic complications [3].Kwong et al. showed that older age, male gender, and the development of postradiotherapy serious otitis were associated with a significantly greater risk of hearing loss [48].Radiation dose and use of chemotherapy were insignificant.In contrast, Grau et al. reported that cochlea dose >50 Gy had a significant greater incidence of sensorineural hearing loss [49].In a recent study of Lee et al., multivariate analyses showed that, in addition to chemotherapy, patient age and cochlea dose were independent factors for hearing loss [3].In our study, we found that radiation dose and female gender were independently poor prognostic factors for otological complications, but use of chemotherapy was insignificant.
In our study, we notice that female gender was associated with higher incidence of late toxicity such as dental complications, trismus, neck fibrosis, hearing loss, and endocrine dysfunction.This finding has been reported previously by Yeh et al. for otologic complications only.It can be explained by a higher hypersensitivity of women.
Lee et al. concluded that age and chemotherapy were significant factors in multivariate analysis for all toxicities after radiotherapy for NPC [3].But, in our study, we found that chemotherapy increases the risk of dental complications and neck fibrosis only in univariate analysis.Younger age was associated with higher risk of trismus, neck fibrosis, and temporal necrosis.

Conclusion
This study enabled us to evaluate our results in term of late toxicity after conventional radiotherapy for nasopharyngeal carcinoma.We showed that radiotherapy was associated with a high rate of late toxicities which affect patients' quality of life.The improvement of our results requires a better definition of radiation modality.With the development of three-dimensional conformal radiotherapy and intensity modulated radiotherapy, a reduced incidence of radiation related complications could be expected.In fact, this study will allow us to compare, in the future, the rates of late toxicities after conventional radiotherapy with those of conformal RT and IMRT.Therefore, we can see the contribution of these new techniques in reduction of late complications incidence.

Figure 3 :
Figure 3: Incidence of dental toxicity after radiotherapy.

Table 4 :
Uni-and multivariate analyses of late toxicity risk factors.