A subgroup analysis of locally advanced nasopharyngeal carcinoma patients with small lymph nodes and large nodes using adaptive radiotherapy

Abstract

Background

To investigate the anatomical and dosimetric differences in target areas and organs at risk (OARs) after adjusting the radiotherapy schedule for nasopharyngeal carcinoma (NPC).

Methods

In total, 34 patients who underwent radiotherapy for NPC were selected. A new localization computed tomography (CT) scan was performed after 25th and before 26th fraction. The target area and organs at risk were redrawn based on the new CT, and the radiotherapy plan was redesigned for the seven remaining rounds of radiotherapy. Plan₁ consists of the old image and old plan, Plan₂ consists of the new image and new plan, and Plan₁₋₂ consists of the new image and old plan. Data on the volume and dose of target areas and organs at risk were collected for two schedule sets.

Results

In the re-scan CT, the volume of Planning Target Volume for Nodal Region Right (PTVnd (R)), Planning Target Volume for Nodal Region Left (PTVnd (L)), left parotid and right parotid decreased significantly (P < 0.05). Additionally, the centroids of the left and right parotid moved towards the centre, abdomen and head. Compared to Plan_1, the D₉₅ and V₁₀₀ of PTVnd (R) and PTVnd (L) of Plan₁₋₂ was significantly lower, also significant differences were observed in the D_mean and D₅₀ of the left and right parotid, and the D_1cc of the spinal cord (P < 0.05). In comparison with Plan₂ and Plan₁₋₂, except for the D_mean and D₉₅ of Planning Target Volume nodal extension (PTVnx) and the D₉₅ of PTVnd (R), significant dosimetric differences were observed in the target area and all organs at risk indicators (P < 0.05). In the re-designed plan, the exposure dose to organs at risk was reduced.

Conclusions

Due to significant anatomical changes in the volume and position of both target volumes and organs at risk during the later fractions of radiotherapy for NPC, the dose coverage of lymph-node regions can no longer meet clinical requirements. Adaptive radiotherapy ensures adequate target coverage while simultaneously reducing the dose to surrounding critical structures.

Introduction

Nasopharyngeal carcinoma (NPC) is highly sensitive to radiation with a high prevalence in the southern region of China [1, 2]. Intensity-modulated radiation therapy (IMRT) is commonly used for the treatment due to the irregular shape of the tumour lesion [3]. NPC Patients may experience regression of the primary tumour and metastatic lymph nodes in the neck during radiotherapy, which may result in alterations to tumour morphology, volume, and location and affecting the surrounding normal tissues’ anatomical structure [4, 5]. Moreover, weight loss and changes in body contour can also lead to the deviation of the target area and surrounding organs of risk (OAR) ’s position from before, eventually impairing the radiotherapy’s accuracy [6].

Adaptive radiotherapy refers to obtaining new images during treatment and adjusting the radiation plan parameters based on the new imaging results [7]. By monitoring changes in the volume and location of the target area and critical organs, and modifying the treatment plan as necessary, we aim to ensures adequate target coverage while simultaneously reducing the dose to surrounding critical structures [8, 9]. Niu et al. investigated the use of neoadjuvant chemotherapy and replanning IMRT for nasopharyngeal carcinoma with intracranial invasion to protect critical normal tissue [4]. Lu et al. [10] found that optimizing radiotherapy after the 25th fraction could ensure the prescribed dose to the target area while reducing the cumulative dose to certain vital organs. Lu et al. [10] found that modifying the treatment plan after the 25th fraction ensured that the prescribed dose was delivered to the target area, while sparing adjacent critical organs from excessive radiation.

However, there is currently no consensus on the timing, triggering mode, and beneficiaries of adaptive radiotherapy.

Our study assesses the morphological changes of target and OARs for patients suffering from NPC after 26 sessions of radiation therapy, to evaluate the necessity of replanning. Moreover, by comparing the dosimetric differences between the initial and re-designed plan, a reference is provided for revising the clinic radiotherapy.

Methods and materials

Patient selection

We selected 34 patients with III-IVA NPC who received IMRT at Sun Yat-sen University Cancer Center (SYSUCC) from January 2022 to November 2022. We retrospectively analysed 34 patients with NPC who received IMRT from January 2022 to November 2022 at the Radiotherapy Department of Sun Yat-sen University Cancer Center. The median age of the patients was 49 years (range, 16–70 years). The inclusion criteria were: (1) histologically confirmed NPC; (2) did not suffer from other malignant tumours; (3) received neoadjuvant chemotherapy and platinum-based concurrent chemoradiotherapy; and (4) received treatment from linear accelerator (Varian Medical Systems, Palo Alto, CA, USA). The exclusion criteria are as follows: (1) patients diagnosed with recurrent or metastatic NPC. and (2) patients with previous radiotherapy or chemotherapy (Table 1).

Table 1.

Clinical characteristics of patients with nasopharyngeal carcinoma

Characteristic	Value (%)
Age
Median	49
Range	16–70
Gender
Male	23(67.65)
Female	11(32.35)
T stage
T2	1(2.94)
T3	30(88.24)
T4	3(8.82)
N stage
N0	1(2.94)
N1	12(35.30)
N2	17(50.00)
N3	4(11.76)
Clinical stage
III	29(85.29)
IVA	5(14.71)

Radiotherapy and treatment planning

Before radiotherapy, all patients were immobilised with a thermoplastic head-and-neck mask and subsequently underwent CT simulation. Both the CT scanning and reconstruction layers had a thickness of 3 mm. The scanning range extended from the upper edge of the frontal sinus to 2 cm below the lower edge of the clavicle head.

The physician contoured slice-by-slice the target areas, cervical lymph-node drainage areas, and organs at risk (e.g., parotid glands, optic nerves, brainstem, spinal cord, and optic chiasm) on the CT images in the treatment-planning system (TPS). The planning target volume (PTV) forms the boundary of the clinical target volume (CTV). PTV is defined by expanding CTV boundary outward by 3–5 mm. The radiotherapy plan was calculated and optimised on the TPS, while the dose to the target volume and critical organs was evaluated using the dose and volume histogram (DVH) and cross-sectional dose distribution. The treatment plan for each patient was validated before the treatment.

The first CT scan was used for 1 st to 26th fraction, and the second set of CT scans was used for the seven last radiotherapy sessions. The second CT scan was performed according to the original position fixation and reference coordinate points. All target volumes and organs at risk were re-contoured from scratch on the re-planning CT, using the initial contours only as a reference while accounting for anatomical changes.

Data acquisition

Two CT image sets were acquired: the initial planning CT before IMRT, and a re-planning CT after the 25th fraction, both used to assess parotid-gland positional changes. The following parameters were acquired from the DVH maps: the volume of the target and organs at risk, mean dose (D_mean), dose received by 95% of the target area volume (D₉₅), the percentage of volume receiving 100% of the prescribed dose (V₁₀₀) and the dose received by 50% of the volume of the left and right parotid glands (D₅₀). The mean dose received by 1 cm³ of the spinal cord (D_1cc) and the dose received by 1% of the volume of the brainstem, right and left optic nerves, right and left temporal lobes, and optic chiasm (D₁) were also acquired. Information on positional changes in the left and right parotid glands in three directions was acquired, and the similarity coefficient (DSC) of both glands was calculated as DSC = 2×Volume_AB/(Volume_A+Volume_B).

The data were classified as follows: Plan₁ refers to the pre-radiotherapy CT image (old image), and the first plan (old plan). Plan₂ denotes the re-optimised IMRT plan generated on the re-planning CT obtained after the 25th fraction (new images). Plan₁₋₂ represents the use of the old plan on the new image (Fig. 1).

Fig. 1.

a Initial planning CT with structures of Plan₁. b Re-planning CT with structures of Plan₂. c Same re-planning CT as b, but with Plan1 structures rigidly copied, illustrating the dose distribution if no re-planning were performed

Statistical analysis

The patient volumes and doses were stated as the mean ± standard deviation (± s). The dosimetric differences between the three groups were analysed using a two-pair sample t-test. The data were analysed using IBM SPSS 27.0 computerised statistical data software. A test level of α = 0.05 (two-tailed) was used as the threshold. A P value < 0.05 was considered statistically significant.

Results

Volume changes of targets and organs of risk

A paired-sample t-test was used to compare the dose–volume parameters derived from the structures that had been (re-)delineated on the initial CT (Plan₁) with those re-delineated on the re-planning CT (Plan₂), revealing statistically significant differences in the volumes of PTVnd (R), PTVnd (L), the left parotid gland, and the right parotid gland (P < 0.05, Table 2). The mean volume of PTVnx in Plan₂ was reduced by 1.2%±4.1%, the mean volume of PTVnd (R) was reduced by 9.6%±16.7%, the mean volume of PTVnd (L) was reduced by 7%±20.3%, and the mean volumes of the left and right parotid glands were reduced by 25.7%±16.3% and 22.4%±13.1%, respectively, compared with those of Plan₁. The percentage reduction in the volume of the parotid glands on both sides was analysed, and the results showed that parotid glands with a large initial volume also had a relatively large volume reduction during radiotherapy (P < 0.001).

Table 2.

Volume changes of targets and organs of risk

Volumes	Plan1 (cm3)	Plan2 (cm3)	t	P
PTVnx	118.20 ± 40.08	115.75 ± 39.90	0.673	0.506
PTVnd (R)	39.91 ± 29.38	32.95 ± 20.74	3.569	0.001
PTVnd (L)	42.50 ± 22.26	37.04 ± 16.03	3.364	0.002
Left parotid gland	29.46 ± 8.03	21.77 ± 7.23	8.93	< 0.001
Right parotid gland	33.83 ± 9.11	26.19 ± 7.83	9.17	< 0.001

Position change of parotid gland

The displacements of the left parotid stroma in the left-right direction (X-axis), head-foot direction (Y-axis), and dorsal-abdominal direction (Z-axis) were 1.72 ± 1.70, 1.32 ± 2.46 and 0. 77 ± 2.78 mm, and the displacements of the right parotid stroma in the left-right, head-foot and dorsal-abdominal directions were − 1.09 ± 1.76, 2.03 ± 1.83 and 0.96 ± 2.15 mm, respectively (Table 3). The centre of mass of the parotid gland on both sides, which is shown in Fig. 2, moved towards the central axis of the body, upwards and anteriorly.

Table 3.

Changes in the displacement of the Parotid glands on the right and on the left in three directions

	DSC	X (mm)	Y (mm)	Z (mm)
Left parotid gland	0.74	1.72 ± 1.70	1.32 ± 2.46	0.77 ± 2.78
Right parotid gland	0.79	−1.09 ± 1.76	2.03 ± 1.83	0.96 ± 2.15

Fig. 2.

Variations in the geometric centre of the parotid gland in three directions. a Left parotid gland b Right parotid gland; The axes indicate the position of the centre point 0. The X-axis represents the left-right direction, the Y-axis represents the head-foot direction, and the Z-axis represents the anterior-posterior direction.The right side of the axes is positive displacement and the left side is negative displacement

Plan₁ vs. Plan₂

A paired sample t-test was performed on the doses to target areas in the two groups. the D_mean and D₉₅ of PTVnx had no significant differences, but all other indicators were statistically significant (P < 0.05). The doses to all organs at risk in Plan₂ were lower than those in Plan₁, and a significant difference was observed (P < 0.05, Table 4).

Table 4.

Comparison of doses to target areas and organs exposed in plan ₁ and plan ₂, plan ₁ vs. plan ₁₋₂ and plan ₂ vs. plan ₁₋₂

Target-area dose		Plan1	Plan2	Plan1−2	Plan1 Vs Plan2		Plan1 Vs Plan1−2		Plan2 Vs Plan1−2
					T	P	T	P	t	P
PTVnx	Dmean (Gy)	70.40 ± 0.40	70.49 ± 0.35	70.41 ± 0.78	−1.917	0.064	−0.172	0.864	0.646	0.523
	D95 (Gy)	68.84 ± 0.41	68.80 ± 0.45	68.82 ± 0.78	0.918	0.365	0.155	0.878	−0.206	0.838
	V100 (%)	0.993 ± 0.011	0.983 ± 0.015	0.990 ± 0.009	2.458	0.019	1.487	0.147	−2.07	0.046
PTVnd (R)	Dmean (Gy)	69.82 ± 1.27	69.25 ± 1.34	69.65 ± 1.51	4.291	< 0.001	1.14	0.262	−2.403	0.022
	D95 (Gy)	68.48 ± 1.17	67.60 ± 1.34	67.47 ± 2.74	6.066	< 0.001	2.239	0.032	0.273	0.786
	V100 (%)	0.985 ± 0.045	0.981 ± 0.024	0.933 ± 0.072	4.625	< 0.001	3.498	0.001	−5.957	< 0.001
PTVnd (L)	Dmean (Gy)	69.86 ± 0.99	69.41 ± 1.33	69.76 ± 1.24	4.081	< 0.001	0.793	0.433	−2.309	0.027
	D95 (Gy)	68.46 ± 0.95	67.62 ± 1.73	67.93 ± 1.68	3.74	0.001	2.183	0.036	−2.081	0.045
	V100 (%)	0.990 ± 0.011	0.972 ± 0.038	0.930 ± 0.069	5.01	< 0.001	5.154	< 0.001	−4.864	< 0.001
Organ at risk
Left parotid gland	Dmean (Gy)	41.73 ± 3.99	39.36 ± 4.36	42.30 ± 3.87	11.5	< 0.001	−2.643	0.012	−10.101	< 0.001
Right parotid gland	D50 (Gy)	39.33 ± 7.59	36.60 ± 7.23	40.71 ± 7.45	11.278	< 0.001	−3.92	< 0.001	−9.421	< 0.001
Brainstem	Dmean (Gy)	41.75 ± 3.22	38.99 ± 3.57	42.35 ± 3.41	12.188	< 0.001	−2.581	0.014	−10.996	< 0.001
Spinal cord	D50 (Gy)	39.08 ± 5.90	36.00 ± 5.80	40.92 ± 6.28	11.368	< 0.001	−5.027	< 0.001	−10.858	< 0.001
	Dmean (Gy)	34.28 ± 3.59	31.96 ± 3.21	34.20 ± 3.61	15.184	< 0.001	0.644	0.524	−11.867	< 0.001
Left temporal lobe	D1 (Gy)	55.37 ± 5.22	53.30 ± 4.76	55.71 ± 4.68	7.208	< 0.001	−1.652	0.108	−9.483	< 0.001
	Dmean (Gy)	25.40 ± 2.94	23.30 ± 2.43	25.30 ± 2.98	14.041	< 0.001	0.664	0.511	−10.812	< 0.001
Right temporal lobe	D1cc (Gy)	41.00 ± 3.31	40.22 ± 2.53	41.63 ± 3.24	3.788	< 0.001	−2.95	0.006	−5.217	< 0.001
	Dmean (Gy)	16.68 ± 3.52	15.09 ± 3.33	16.60 ± 3.53	22.471	< 0.001	0.748	0.46	−13.021	< 0.001
Left optic nerve	D1 (Gy)	65.87 ± 3.54	63.86 ± 5.15	65.76 ± 3.64	5.373	< 0.001	0.534	0.597	−5.085	< 0.001
	Dmean (Gy)	17.42 ± 3.81	15.66 ± 3.59	17.34 ± 3.81	26.993	< 0.001	0.756	0.455	−14.155	< 0.001
Right optic nerve	D1 (Gy)	65.12 ± 2.65	62.69 ± 3.86	64.91 ± 2.85	7.878	< 0.001	0.957	0.346	−7.023	< 0.001
	Dmean (Gy)	39.15 ± 11.37	33.16 ± 10.01	38.91 ± 11.70	20.008	< 0.001	0.676	0.504	−11.804	< 0.001
Optic chiasma	D1 (Gy)	57.63 ± 8.20	49.76 ± 8.42	57.31 ± 8.35	14.75	< 0.001	0.702	0.488	−12.568	< 0.001
	Dmean (Gy)	39.11 ± 11.21	33.22 ± 9.96	38.75 ± 11.29	19.905	< 0.001	0.98	0.334	−12.022	< 0.001
Organ at risk	D1 (Gy)	58.29 ± 8.81	50.67 ± 9.05	57.69 ± 8.90	13.793	< 0.001	1.318	0.197	−10.957	< 0.001
Left parotid gland	Dmean (Gy)	46.34 ± 10.55	39.15 ± 9.87	45.88 ± 10.81	19.493	< 0.001	1.079	0.289	−11.74	< 0.001
Right parotid gland	D1 (Gy)	59.10 ± 7.85	51.12 ± 8.95	58.52 ± 8.06	12.173	< 0.001	1.162	0.253	−10.432	< 0.001

Plan₁ vs. Plan₁₋₂

Comparing the doses to the target areas between Plan ₁ and Plan₁₋₂, the results showed significant differences at D₉₅ and V₁₀₀ only in PTVnd (R) and PTVnd (R) (P < 0.05, Table 4). Based on the comparison of the doses to the organs at risk in the two groups, the left and right parotid doses increased in Plan₁₋₂, in which the mean dose to the left parotid was higher than that in Plan1 in 71% of patients, the left parotid D₅₀ increased in 76% of patients, the right parotid D_mean increased in 82% of patients, the right parotid D₅₀ increased in 85% of patients, the spinal cord D_1cc increased in 65% of patients, and the differences were statistically significant (P < 0.05, Table 4).

Plan₂ vs. Plan₁₋₂

The comparison between Plan ₂ and Plan ₁₋₂ shows that the changes in dosimetric parameters in the plan reflect the advantages replanning. Based on the comparison of the dose to target areas in the two groups, no difference was observed in the PTVnx D_mean, D₉₅ and PTVnd (R) D₉₅, and the differences in all other indicators were statistically significant (P < 0.05). Comparing the dose of critical organs in the two groups, the dose of each index in Plan₁₋₂ was significantly higher than that in Plan₂ (P < 0.05) (Table 4). A decrease was observed in V100 (%) of PTVnd (R) and PTVnd (L), whereas all other parameters increased.

Discussion

The volumes of the NPC target area, cervical lymph nodes, and surrounding normal tissues, as well as changes in their dose and conformity, were systematically monitored throughout the course of IMRT for NPC [11–13]. Based on the analysis of rescan CT and replanning, the cervical lymph-node volumes underwent varying degrees of shrinkage, in which the mean volume of PTV for lymph nodes (PTVnd) shrunk by 7 ~ 9.6% compared with the pre-radiotherapy period. Chophy et al. reported the similar result of PTVnx and PTVnd, however, we found the mean volume of PTV for nasopharyngeal tumour (PTVnx) only shrunk by 1.2%. Maybe it was due to all patients enrolled in this study were locally advanced nasopharyngeal carcinoma [14]. Among the organs at risk, the parotid glands are particularly vulnerable to radiation damage due to their high radiosensitivity. The mean volume of both the left and right parotid glands was reduced by 25.7% and 22.4%, respectively. In a study by Wang et al. [15], anatomical changes were observed in 15 patients with locally progressed nasopharyngeal carcinoma, and they noted the displacement of parotid glands of both sides to the midline by an average of 4.8 and 4.3 mm, respectively. In the present study, the stroma of the left parotid gland was displaced medially and upward by 1.72 ± 1.70 and 1.32 ± 2.46 mm, respectively, while the right parotid stroma was displaced medially and upward by 1.09 ± 1.76 and 2.03 ± 1.83 mm, respectively. These displacements were similar to the above-mentioned results for the left and right parotid gland. Overall, patients with nasopharyngeal carcinoma exhibit a tendency of external contour retraction during radiotherapy.

Could anatomical changes cause actual dose redistribution during radiotherapy? Our study compared the plan applied to the initial and re-scanned image, and found no statistically significant difference in the dose coverage of the nasopharyngeal tumour. However, the coverage rate of the lymph-node targets was below 95% (PTVnd(L): 93.0%, PTVnd(R):93.3%), which may increase the likelihood of cancer recurrence. If radiotherapy for nasopharyngeal cancer is not re-planned, it could lead to a noteworthy rise in the radiation dose to crucial organs, including the parotid gland, brainstem and spinal cord [3, 16, 17]. In the present study, the original plans applied to the initial and second CT images were compared. It was shown that the critical organs such as the right and left parotid gland D_mean and D₅₀ and brainstem D₁ and spinal cord D_1cc in Plan₁₋₂ were statistically significantly higher than Plan₁. Although the dosimetric parameters of OARs were still within the safe threshold range, an increased risk of complications, similar to Kumar’s report [18], might be avoidable. It is worth noting that the higher radiation dose to the parotid gland is due to its reduced volume and internal contraction, but the D_1cc of brainstem and spinal cord with little volume change was also significantly increased. The objective triggering conditions for adaptive radiotherapy not only consider anatomical changes, but also dosimetric changes, especially for serial organs.

In this study, PTV rather than GTV was chosen as the primary volume for dosimetric evaluation based on several considerations. First, international multicenter clinical trials such as RTOG 0615 and NRG-HN001 commonly use the PTV as the benchmark for dose coverage. The PTV typically includes a 3–5 mm margin beyond the gross tumor volume (GTV) and is generally required to receive at least 95% of the prescribed dose [19, 20]. Second, although the GTV may undergo considerable changes during the course of radiotherapy, as long as these changes occur within the original PTV, dose coverage is usually maintained and adaptive re-planning is not necessary [21, 22]. In contrast, using the GTV as the evaluation volume may increase the risk of false-positive findings.

In addition, considering the anatomical asymmetry commonly observed in nasopharyngeal carcinoma, we divided the parotid glands by both left–right laterality and ipsilateral–contralateral orientation for subgroup analysis. Both grouping strategies revealed similar trends in dose reduction [22, 23], and for consistency, we adopted the conventional left–right grouping approach in this study.

Regarding the Plan1–2 scheme, this was a simulation in which the original treatment plan was applied to the second CT scan to estimate the potential changes in dose distribution that might have occurred if adaptive re-planning had not been performed. While not implemented clinically, this simulation approach is commonly used in adaptive radiotherapy research to evaluate theoretical dose variations [21].

Finally, all patients were treated using Varian linear accelerators to reduce variability arising from equipment differences. Previous studies have demonstrated that, under comparable imaging and planning protocols, modern intensity-modulated radiotherapy platforms produce consistent and generalizable dosimetric outcomes [23]. Therefore, the findings of this study are broadly applicable.

The anatomical and dosimetric changes that occur during the course of radiation therapy are influenced by multiple underlying mechanisms. Anatomical and dosimetric changes during radiation therapy are driven by tumor shrinkage, normal tissue response, and patient-specific factors like weight loss. These changes, particularly in the nasopharyngeal region, can significantly alter dose distribution, risking underdosing of the target or overdosing of nearby tissues. The most rapid changes usually occur within the first two to three weeks, especially in radiosensitive tumors, necessitating timely treatment adjustments to maintain efficacy and minimize toxicity. Optimal re-planning should occur around the third to fourth week when most anatomical changes have stabilized but before dose deviations compromise treatment. Early re-planning may be necessary if significant changes are detected. Re-imaging with nasopharyngeal MRI, which offers superior soft tissue contrast, can enhance the delineation of anatomical changes and refine treatment plans to better protect critical structures such as the temporal lobes, brainstem, and optic nerves. Adaptive radiation therapy (ART) is valuable for managing these changes, though greater caution is needed for the nasopharynx due to its complex anatomy. Decisions on adaptation should be based on anatomical changes, their impact on dose distribution, and the risk to critical structures.

Sectional radiotherapy mode may be a good solution to correct the dose uncertainty caused by patients’ anatomical changes. It was reported that re-planning during IMRT treatment in patients with NPC resulted in a dose reduction of brainstem, spinal cord and optic nerve. These findings indicate that replanning leads to a significant enhancement in dose–volume parameters for organs at risk [24, 25]. Our study performed CT re-scan on 34 patients with NPC who underwent 26 radiotherapy treatments. We involved re-studying the plan design by comparing the new plan (Plan2) with the synthetic plan obtained by superimposing the original plan onto the new CT (Plan1-2). It was shown that the re-optimization improved the dose coverage of the lymph nodes while significantly reducing the dosimetric parameters of most OARs (Table 4), which was similar to Huang’s report [24].

It is recommended to perform CT re-scanning and treatment plan re-design before the 26th fraction of radiotherapy for patients with locally advanced nasopharyngeal carcinoma, particularly those at N2 and N3 stages. However, several limitations of our study need to be acknowledged. Our study did not analyse the trends of anatomical changes in the target and OARs. Additionally, this study did not assess the radiation therapy-related toxicities and side effects. Future research could address this aspect to gain a more comprehensive understanding of treatment outcomes. Moreover, compared to a unified re-scanning time, it was more reasonable to start re-planning based on individual differences of patients, such as lateral neck diameter [26].

Conclusions

Due to anatomical changes observed from fraction 26 to 33 (the last seven fractions) of radiotherapy for NPC, the overall contraction change will make the dose coverage of lymph nodes unable to meet clinical requirements. Adaptive radiotherapy can ensure the adequate dose coverage of target and reduce the dose of peripheral organs at risk.

Abbreviations

NPC

Nasopharyngeal carcinoma

OARs

Target areas and organs at risk

CT

Computed tomography

IMRT

Intensity-modulated radiation therapy

PTV

Planning target volume

CTV

Clinical target volume

DVH

Dose volume histogram

Authors’ contributions

Senkui Xu: Conceptualization, Methodology, Software, Funding; Weixiong Xia: Writing- Reviewing and Editing, Funding, Software; Jiang Hu: Conceptualization, Methodology, Software; Wenyan Yao: Data curation, Writing- Original draft preparation; Shuhui Dong: Investigation, Writing- Original draft preparation; Peixun Xu: Visualization, Investigation, Writing- Original draft; Ziyue Zhong: Visualization, Investigation; Zongtai Li: Visualization, Investigation; Yupeng Zhou: Supervision; Yongwen Fang: Software, Validation; Zhiyue Lin: Writing- Reviewing and Editing; Mengxue He: Visualization, Investigation. All authors read and approved the final manuscript.

Funding

This research was funded by the Medical Science and Technology Research Foundation of Guangdong Province (No. A2020621); the Medical Science and Technology Research Foundation of Guangdong Province (No. A2017613); and the Natural Science Foundation of Guangdong Province (No. 2022A1515010673).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the ethics committee of Institutional Review Board of Sun Yat-Sen University Cancer Center (SL-B2023-498-01) in accordance with the Declaration of Helsinki. Written informed consent was obtained from all individual patients included in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.