Dosimetrical Correlation between Cervical Site Radiation Dose and Radiation Dermatitis in Advanced Nasopharyngeal Carcinoma: A Retrospective Analysis

1. INTRODUCTION

Nasopharyngeal carcinoma (NPC), originating from the mucous membrane of the nasopharynx, exhibits the highest incidence in southern China. Due to its high sensitivity to ionizing radiation, radiotherapy is the preferred treatment modality [1]. Advances in radiotherapy technologies have significantly reduced the mortality of NPC, with reported 5-year relapse-free survival (RFFS) rates of 90.2% to 98.5% and overall survival (OS) rates of 76.4% to 91.7% [2]. Studies indicate 5-year overall survival rates ranging from 70.5% to 89.4% for patients with stage II to IVB disease [3-5]. Long-term survival among NPC patients underscores the importance of enhancing quality of life, as survivors may face complications such as cervical subcutaneous fibrosis, hearing loss, and skin dystrophy [6, 7].

In this study population, NPC incidence was relatively low, yet most cases were diagnosed at stage III or higher, posing challenges in radiotherapy planning. Addressing these challenges could lead to significant clinical advancements. We therefore investigated adverse events associated with chemoradiotherapy, explored correlations between clinical parameters and radiation dermatitis, and specifically investigated the dosimetric relationship between radiation doses to the neck site and radiation dermatitis in NPC patients. Potential strategies for managing radiation dermatitis are also discussed.

2. MATERIALS AND METHODS

This retrospective case-control study analysis included 109 patients diagnosed with NPC (stage II-IVA, AJCC 8th edition) treated at Anhui Provincial Cancer Hospital from November 2018 to June 2020. The cohort comprised 71 males and 38 females, aged 21 to 75 years, with a median age of 50.4 years. Patient characteristics are presented in Table 1. Eligibility criteria were as follows:

(1) Karnofsky Performance Status (KPS) ≥ 70; (2) sufficient organ function and capacity for regular follow-up; (3) biopsy-confirmed NPC (via endoscopic biopsy or cervical lymph node biopsy); and (4) initial evaluation included MRI/CT scans of the nasopharynx and neck, chest CT, upper abdominal ultrasound, bone ECT scan, and laboratory tests (complete blood count, biochemical profile, and EBV DNA). Exclusion criteria were as follows:

(1) Karnofsky Performance Status (KPS) <70; (2) severe or uncontrolled concurrent medical conditions; (3) pregnancy at diagnosis; (4) prior history of head and neck radiotherapy.

Sample size was calculated using G*Power 3.1 software with the following parameters: effect size = 0.3, α = 0.05, and statistical power (1−β) = 0.8. Incorporating the disease incidence in the study area (where the study population resided), the total required sample size was determined to be no less than 102 cases. Clinical data of the cases were extracted from hospital electronic medical records, and the eligibility criteria for this study were established in accordance with previously published literature [9] [8-10].

3. OUTCOME

4. RESULTS

4.2. Toxicities

Acute adverse events are listed in Table 6. Adverse events were reported by the majority of patients, with the most common ones being neutropenia and gastrointestinal tract reactions. The incidence of grade 1 and 2 neutropenia and gastrointestinal tract reaction were 72.48% and 88.1%, respectively, while grade 3 neutropenia and gastrointestinal tract reaction were observed in 27.52% and 11.9% of patients. Five patients experienced hepatoxicity and 1 developed nephrotoxicity. Overall, adverse events associated with concurrent chemoradiotherapy were generally manageable, with no reports of grade 4 neutropenia, nausea, or vomiting. RD, the most common radiation-induced complication, affected patients with grades 1, 2, 3, and 4 in 0.92%, 40.37%, 53.38%, and 7.33% of all patients, respectively. Moderate to severe RD (grades 3 and 4) was observed in 60.71% of participants.

Table 6.

Event	Grade 1 or 2	Grade 3	Grade4
Radiation dermatitis	45(41.28%)	56(51.38%)	8(7.34%)
Neutropenia	79(72.48%)	30(27.52%)	0
Nausea and Vomiting	96(88.1%)	13(11.9%)	0
Hepatoxicity	5(4.59%)	0	0
Nephrotoxicity	1(0.92)	0	0

In the univariate analysis, node stage and TMN stage demonstrated significant associations with an increased likelihood of grade 3/4 RD. Dosimetric factors, including Dmean to the cervical site and volume of GTV-nd, were analyzed for their relationship with RD severity. Subsequently, multivariate analysis identified Dmean to the cervical site as an independent predictor (OR=1.003; 95%CI=1.001-1.005; P=0.003).

We assessed the Dmean to the neck in 109 patients, ranging from 4425 cGy to 5942 cGy, with an average of 5330 cGy. To clarify in detail the toxicity levels of radiation-induced skin injury associated with different radiation doses, patients were categorized into four groups based on Dmean: <5000 cGy (group 1, n = 9), 5001-5200 cGy (group 2, n = 25), 5201-5500 cGy (group 3, n = 51), and >5500cGy (group 4, n = 24). We compared the incidence of RD between mild (grade 1 and 2) and moderate/severe (grade 3 and 4) groups across these Dmean categories: 6 vs. 3 (mild vs moderate/severe) cases in group 1, 18 vs. 7 cases in group 2, 17 vs. 34 cases in group 3, and 4 vs. 20 cases in group 4 (Table 7). As anticipated, there were significantly more cases of severe RD in higher Dmean groups (group 2 vs. 3, P=0.03; group 2 vs. 4, P<0.01), with no statistical significance between groups 3 and 4. Further stratification into ≤5200 cGy (N=34) and >5200 cGy (N=76) groups revealed a notably higher incidence of grade 3 or 4 RD in the high Dmean group (Dmean>5200 cGy, RD grade≤2 vs. RD grade≥3, P<0.01). Receiver operating characteristic (ROC) curve analysis indicated that a Dmean >52.06 Gy was predictive of grade ≥3 RD (AUC=0.733%; P<0.01; 95%CI= 0.637-0.829) (Figs. 1-3).

Table 7.

Radiothemities	Grade 1	Grade 2	Grade 3	Grade 4	Grade 1+2	Grade 3+4	P-value
<5000cGy	0	6	3	0	3	6	0.763
5000~5200cGy	0	18	6	1	7	18	0.003
5201~5500cGy	1	16	29	5	34	17	0.173
>5500cGy	0	4	18	2	20	4	-
<5200cGy	0	24	9	1	10	24	-
≥5200cGy	1	20	47	7	54	21	0.000*

5. DISCUSSION

Radiation dermatitis, ranging from erythema to severe desquamation, results from epithelial cell depletion due to radiation exposure [11], and is assessed according to the RTOG criteria for severity. It is reported that 41.4% of head and neck cancer patients suffer grade 3 or 4 dermatitis [12], if untreated, can lead to severe complications, even death. Our investigation highlights the development of severe acute radiation dermatitis in patients undergoing concurrent chemoradiotherapy, correlating with specific dosimetric factors. By comparing dermatitis incidence across different mean neck dose groups, we identified that a mean cervical dose exceeding 5200 cGy may induce severe dermatitis. A prospective clinical study [12] on acute radiotherapy side effects in nasopharyngeal carcinoma demonstrated that advanced radiation-focusing techniques can prevent skin irradiation and reduce skin damage. However, the study did not address factors such as radiotherapy dose and target volume. Bonomo P [12] concluded that the volume of the 2-mm skin ring receiving 5000 cGy and 6000 cGy as the optimal dosimetric predictors, with corresponding values of 19.9 cc and 5.8 cc, respectively (AUC = 0.61 for both). The development framework based on the dose-toxicity surface histogram (DTH) enables severity-based classification and predictive dose assessment of RD, with a predicted dose of 5430 cGy for Grade 3 RD [12]. This study reviewed and summarized parameters in radiation plans for patients with nasopharyngeal carcinoma, with a specific focus on analyzing cervical skin doses. This approach allowed for a more direct and comprehensive assessment of the correlation between radiation dermatitis development and cervical radiation exposure. We found that Dmean=5206 cGy was highly consistent with the findings of previous studies. In a case study reported by Donghayun Kim [13], the mean dose to level I-VII neck nodes was found to be 5860 cGy, further validating the reliability and clinical relevance of the results in the present study.

The mechanism of RD remains unclear, and standardized treatments for this condition are lacking. Preventive interventions can significantly reduce RD-related damage, with optimized radiotherapy dose serving as a key measure to mitigate RD [14, 15]. While prior studies have established an association between cervical radiation dose and dermatitis severity [16], detailed dose-specific data regarding RDremain scarce. This study aims to clarify in detail the toxicity levels of such skin damage across different radiation doses and provides precise dosimetric insights. Future efforts will focus on reducing cervical radiation exposure and exploring dermatitis management strategies from existing studies.

The toxicity of RDvaries based on factors such as total radiation dose, fractionation schedule, and volume of organ or tissue treated [17]. Strategies aimed at reducing doses to the cervical site include minimizing doses to GTV-nd and CTV, as well as optimizing target volumes. Some studies have shown that elective neck irradiation can mitigate RD and enhance the quality of life [18], particularly for patients with N0-N1 nasopharyngeal carcinoma. However, this approach lacks robust, high-level evidence [19], partly due to reported in-field failures (64%) as the primary pattern of regional recurrence [20, 21]. Safely defining target volume boundaries remains crucial. Metastasis from nasopharyngeal carcinoma typically follows a predictable path through cervical lymph node levels, necessitating adequate target dose delivery for efficacy [22, 23]. Consequently, directly reducing prescription doses and volumes to the cervical site to minimize RD remains a topic of controversy.

Reducing unnecessary target volumes is a potential method to lower cervical site doses. Radiation therapy affects not only cancer cells but also nearby tissues and organs, emphasizing the importance of optimizing treatment delivery to balance tumor control with sparing normal tissues [24]. Incorporating PET/CT imaging into IMRT planning for nasopharyngeal carcinoma offers improved precision in delineating GTVnd volumes and elective nodal targets, enhancing treatment accuracy, uniformity, and efficiency [25].

Efforts to mitigate cervical site radiation doses remain limited, necessitating a comprehensive review of clinical approaches to reduce skin toxicity. Variations in radiosensitivity and toxicity profiles across different chemotherapy regimens are often overlooked in current chemoradiotherapy protocols [26]. Specifically, higher incidences of acute grade 3 skin toxicity have been associated with 5-fluorouracil use [27], suggesting avoidance of this agent during radiotherapy. Radioresistance poses a significant challenge in managing nasopharyngeal carcinoma [28], prompting exploration into combination therapies to potentially overcome this issue. Notably, immunotherapy, particularly in cases with high PD-L1 expression, offers a promising avenue for enhancing treatment efficacy [29], despite potential immune-related adverse events affecting the skin and appendages [30]. Clinical trials evaluating anti-PD-L1 antibodies and other immune checkpoint inhibitors in various treatment contexts are ongoing, with further investigation needed on their impact on radiation dermatitis [31].

While establishing definitive guidelines for minimizing cervical site dose and volume remains challenging, clinicians can implement interventions to manage skin toxicity during chemoradiotherapy. Topical corticosteroids reduce the production of inflammatory mediators, modulate inflammatory/immune responses, cell proliferation, and vasoconstriction, thereby lowering the risk of RD [32, 33]. An antiulcer oral mucosal protectant creates a microenvironment that promotes the self-healing of damaged mucosal epithelial cells, enabling the prevention and treatment of oral mucositis and mucosal ulcers [34]. Epidermal growth factor (EGF) maintains homeostasis by regulating the proliferation, growth, and migration of epithelial cells. It also induces angiogenesis and provides nutritional support for tissues [35]. These interventions have additionally been shown to benefit the treatment of RD. A silicone-based film-forming gel dressing, which creates a barrier to reduce mechanical friction and transepidermal water loss (TEWL), is recommended for the prevention of radiation dermatitis [36]. Advancements in radiotherapy techniques, such as TOMO therapy, show promise in improving therapeutic ratios for treating nasopharyngeal carcinoma [37]. However, further clinical data and studies are necessary to justify broader application, considering economic considerations and treatment efficacy.

6. LIMITATIONS OF THE STUDY

Given the current gaps in evidence, ongoing challenges in nasopharyngeal carcinoma treatment and clinical management, and the inherent limitations of our study in terms of sample size and duration of follow-up, future research should involve larger cohorts and long-term quality of life assessments. Such efforts are essential to advance the management of radiation dermatitis in nasopharyngeal carcinoma treatment protocols.

CONCLUSION

Based on the findings of this study, maintaining a Dmean <5200 cGy for the neck site may mitigate severe skin toxicities and delay the onset of moderate/severe radiation dermatitis. One should carefully consider reducing cervical doses to protect neck skin integrity, ensuring precise and adequate delivery to target volumes in patients with locally advanced nasopharyngeal carcinoma. Based on current evidence and literature, we recommend early topical treatments (such as EGF and corticosteroids) for patients with Dmean ≥5200 cGy during radiotherapy.

The correlation between RD and whole cervical site dose remains inadequately explored, with no consensus on optimal radiation doses for this region in nasopharyngeal carcinoma patients. Our research contributes to understanding the clinical characteristics of radiation dermatitis and proposes a potential strategy to mitigate skin damage.

REFERENCES