Abstract
Background:
Acute radiation dermatitis (ARD) is common during head-and-neck radiotherapy (RT) and may impair skin-related quality of life (QoL). Whether intensity-modulated proton therapy (IMPT) improves patient-reported skin outcomes compared with volumetric modulated arc therapy (VMAT) remains uncertain. We compared longitudinal outcomes of skin toxicity between RT modalities and identified significant predictors of ARD and skin-related QoL.
Methods:
This prospective and nonrandomized cohort study enrolled adults with newly diagnosed head-and-neck cancer receiving curative-intent VMAT or IMPT at Kaohsiung Chang Gung Memorial Hospital. Skin-related QoL was assessed using Skindex-16 at baseline, end of RT, and 3 months after RT. ARD was graded weekly using Common Terminology Criteria for Adverse Events version 5.0. Longitudinal outcomes were analyzed using linear mixed-effects models, and predictors of ARD grade ≥2 and ≥10-point worsening in Skindex-16 Global score were evaluated using multivariable regression.
Results:
Among 158 patients, 113 received VMAT and 45 received IMPT; the mean age was 56.5 years, and 130 patients (82.3%) were men. Clinically meaningful Global score worsening occurred in 37.0% of patients at the end of RT and 7.0% at 3 months after RT. Skindex-16 scores worsened significantly at the end of RT and improved by 3 months across all domains (all p < 0.01), with no significant difference between VMAT and IMPT or time-by-modality interactions. Maximum ARD severity was comparable between VMAT and IMPT (grade 2: 30.0% vs 29.5%; grade 3: 7.0% vs 4.8%; p = 0.85). Advanced nodal disease (N2-3) and RT dose ≥66 Gy independently predicted ARD grade ≥2 and Skindex-16 Global score worsening.
Conclusion:
Acute patient-reported skin toxicity during contemporary head-and-neck RT was not associated with RT modality but with nodal extent and RT dose. These findings support risk-adapted planning and proactive supportive care, although interpretation is limited by the nonrandomized design and lack ofskin-specific dosimetry.
Keywords: Intensity-modulated proton therapy, Patient-reported outcomes, Radiation dermatitis, Skindex-16, Volumetric-modulated arc therapy
Lay Summary: Radiotherapy often causes skin reactions that can be painful or irritating for people with head and neck cancer. Newer proton therapy might reduce the dose to healthy tissues compared with modern photon therapy, but it is unclear whether patients feel a difference in skin symptoms. We followed 158 patients who received either photon therapy or proton therapy and asked them to report their skin symptoms before treatment, at the end of treatment, and 3 months later, with doctors also checking their skin weekly. Skin symptoms became worse during treatment and improved afterward in both groups. We found no major differences in patient-reported skin problems between the two radiation types. Instead, patients with more lymph node involvement or higher radiation doses were associated with worse skin reactions. These findings suggest that cancer severity and treatment intensity, rather than radiation type, are more closely related to patients’ skin experiences during therapy.
1. INTRODUCTION
Radiotherapy (RT) is a cornerstone treatment for head and neck cancers (HNC), offering excellent tumor control while aiming to preserve organ function. Modern photon-based techniques such as volumetric modulated arc therapy (VMAT) and proton-based approaches like intensity-modulated proton therapy (IMPT) allow precise dose delivery with improved sparing of organs at risk (OARs) compared to conventional methods.1,2 Dosimetric studies have suggested that IMPT could further reduce unnecessary dose exposure and potentially lessen acute toxicities, although clinical evidence remains mixed.1 Despite these technological advances, acute radiation dermatitis (ARD) remains one of the most common and distressing acute toxicities of RT for HNC patients, with incidence rates approaching 90% to 95% in some series.2 ARD not only causes visible skin changes such as erythema, dry desquamation, or moist desquamation, but also impacts pain, discomfort, and daily functioning. Severe ARD can lead to treatment interruptions, compromised compliance, and reduced quality of life (QoL).3
Patient-reported outcomes (PROs) are increasingly recognized as essential complements to physician-graded toxicity scales such as RTOG or CTCAE.4 Among dermatology-specific instruments, the Skindex-16 is validated for oncology populations and endorsed by a recent international Delphi consensus as a core measure for ARD alongside CTCAE and RISRAS.5 Skindex-16 captures three subdomains—symptoms, emotions, and functioning—and generates a Global score reflecting overall skin-related QoL. The Global score is often selected for primary analyses because it summarizes the overall burden of skin toxicity on daily life, while subscale analyses allow further exploration of specific domains. Recent validation in breast cancer patients with ARD confirmed its reliability, responsiveness, and sensitivity to toxicity severity.6
While comparisons between VMAT and IMPT have primarily focused on physician-reported toxicities and dosimetry, little is known about patient-reported skin-related QoL trajectories between these modalities. In particular, the extent to which ARD severity rather than treatment modality drives QoL changes remains unclear. Therefore, the present study evaluated Skindex-16 in HNC patients treated with VMAT or IMPT at baseline, end of RT, and 3 months post-RT. The objectives were: (1) to characterize temporal changes in skin-related QoL, (2) to assess the impact of ARD severity on PROs, (3) to determine whether treatment modality independently influences skin-related QoL trajectories, and (4) to identify clinical predictors of ARD and/or skin-related QoL.
2. METHODS
2.1. Study design and patients
This single-institution prospective cohort enrolled consecutive adults with HNC treated with curative-intent VMAT or IMPT between January 2023 and December 2024. No randomization or protocol-mandated allocation was used. Insurance coverage and patient financial eligibility were the primary determinants for IMPT access. During the study period, 158 participants were enrolled. Eligibility criteria included (1) newly diagnosed HNC with no prior malignancy or RT, and (2) no chronic inflammatory skin disease of the head and neck (eg, psoriasis, dermatomyositis, scleroderma, eczema). The protocol was approved by the Institutional Review Board (IRB No. 202201291B0), and all participants provided written informed consent.
2.2. Treatment
VMAT and IMPT techniques followed institutional standards.7 IMPT was delivered with pencil-beam scanning on a Sumitomo Proton Therapy System. Treatment planning for both modalities used RayStation v7 (RaySearch Laboratories, Stockholm, Sweden). All patients were treated with a thermoplastic head-and-neck mask, with attention to minimizing excessive skin pressure. No routine bolus was used in this cohort. Target delineation, dose prescriptions, and fractionation were standardized with 70 Gy to gross disease for definitive cases and 54 to 66 Gy to postoperative/elective regions in 33 to 35 fractions (once daily, 5 d/wk). OARs were constrained according to published consensus guidelines, primarily based on QUANTEC recommendations and contemporary head and neck RT consensus statements, which were adopted as institutional planning standards during the study period.8–10 IMPT plans followed the same OAR dose objectives as VMAT plans, without additional proton-specific constraints. For IMPT, planning typically involves utilizing three-beam directions: left and right anterior obliques, along with posterior fields, utilizing multi-field optimization. Robust optimization was generally used to account for range uncertainties (+3.5%) and positional uncertainties (+3 mm). Range shifters were used when required for shallow targets; air gaps were minimized but not prospectively recorded. A formal skin rind structure (a layer structure of 5 mm inward from the head and neck contour) was optionally outlined and arbitrarily chosen as a constraint with the request of “as small as possible without compromising the coverage of clinical target volume during optimization” in the cohort. Image guidance was used for all IMPT patients and for 65 (57.5%) of VMAT patients. Concurrent chemoradiotherapy was administered in 121 (76.6%) patients with weekly cisplatin 40 mg/m2 for 1 to 7 cycles.
A standardized institutional skin-care protocol was used for all patients regardless of modality, consisting of gentle cleansing and prophylactic moisturizing cream and/or aloe gel during RT, with topical corticosteroids as needed. For ARD grade ≥2, treatment was escalated to topical corticosteroids and/or silver sulfadiazine cream and dressings.
2.3. Skindex-16 assessment
Skindex-16 is a validated dermatology-specific PRO instrument consisting of 16 items grouped into three domains: Symptoms (4 items), Emotions (7 items), and Functioning (5 items).8 Each item is rated on a 7-point Likert scale and linearly transformed to a 0 to 100 scale, with higher scores indicating worse skin-related QoL. Domain scores were calculated as the means of non-missing items within each domain, and a global score was calculated as the mean of all 16 items, with domain or global scores considered missing if more than 50% of the corresponding items were unanswered. The Traditional Chinese Skindex-16 has been validated and used in Taiwanese RT populations.11,12 Questionnaires were administered at baseline (T1), end of RT (T2), and 3 months post-RT (T3); higher scores indicated worse skin-related QoL. A ≥10-point increase from baseline on the 0 to 100 scale was used to define clinically meaningful worsening, consistent with prior Skindex literature.13
2.4. Physician-reported dermatitis
Physician-reported ARD was assessed weekly using CTCAE v5.0 criteria4 by treating radiation oncologists, with support from trained oncology nurses. Grading followed standardized institutional procedures based on CTCAE definitions; formal inter-rater reliability testing was not performed. The maximal grade of ARD was recorded and analyzed both as a dichotomous outcome (grades 0-1 vs grades ≥2) and ordinally to evaluate its relationship with Skindex-16 outcomes.
2.5. Statistical analysis
Baseline characteristics were summarized using descriptive statistics: mean (SD) or median (range) for continuous variables and counts (percentages) for categorical variables. Longitudinal changes in Skindex-16 (Global, Emotional, Functional, Symptom) were analyzed using linear mixed-effects models with patient-specific random intercepts. Fixed effects included treatment modality, time point, their interaction, and prespecified covariates (age, sex, smoking habit, diabetes, education years, body mass index, tumor site, AJCC stage, RT intent, RT dose, image guidance, concurrent weekly cisplatin cycles, and RT modality). Models were estimated using restricted maximum likelihood with an unstructured covariance matrix. Estimated marginal means with SE and adjusted mean differences with 95% CI were derived. To identify predictors of the Skindex-16 Global score at T2 and T3, we applied multivariable linear regression models that included the corresponding baseline Global score and prespecified sociodemographic and treatment covariates; results are presented as β coefficients (β) with 95% CI. Predictors of ARD grade ≥2 and clinically meaningful worsening of Skindex-16 Global (≥10-point increase from baseline) were evaluated using binary logistic regression, reporting odds ratios (OR) with 95% CI. All tests were two-sided with α = 0.05. Analyses were conducted using IBM SPSS Statistics, version 26 (IBM Corp., Armonk, NY).
3. RESULTS
3.1. Patient characteristics
Among the 158 patients included in the study cohort, 113 (71.5%) received VMAT, and 45 (28.5%) received IMPT. Several baseline characteristics differed significantly between treatment groups (Table 1). Compared with VMAT, IMPT patients were younger, had higher educational attainment, were less likely to be smokers, and more frequently had nasopharyngeal tumors (all p < 0.05). IMPT was also more commonly delivered as definitive RT, at higher prescribed doses, and with routine image guidance. Gender, body mass index, diabetes status, AJCC stage, T status, and concurrent cisplatin exposure were comparable between groups. Completion rates for Skindex-16 were high (97.4% at T2 and 89.9% at T3) and did not differ by modality at any time point (Supplementary Table 1, https://links.lww.com/JCMA/A420).
Table 1.
Patient characteristics (n = 158)
| Value, % | p | ||||
|---|---|---|---|---|---|
| Total | VMAT | IMPT | |||
| Patient number | 158 (100) | 113 (71.5) | 45 (28.5) | ||
| Age, mean ± SD, y | 56.5 ± 11.1 | 58.1 ± 9.7 | 52.5 ± 13.3 | <0.01 | |
| Gender | |||||
| Female | 28 (17.7) | 18 (15.9) | 10 (22.2) | 0.36 | |
| Male | 130 (82.3) | 95 (84.1) | 35 (77.8) | ||
| Education years | |||||
| ≤12 | 115 (72.8) | 92 (81.4) | 23 (51.1) | <0.01 | |
| >12 | 43 (27.2) | 21 (18.6) | 22 (48.9) | ||
| Smoking habit | |||||
| No | 48 (30.4) | 29 (25.7) | 19 (42.2) | 0.03 | |
| Yes | 110 (69.6) | 84 (74.3) | 26 (57.8) | ||
| Body mass index, mean ± SD, kg/m2, | 24.0 ± 4.7 | 23.9 ± 4.9 | 24.3 ± 4.2 | 0.62 | |
| Diabetes mellitus | |||||
| No | 120 (75.9) | 85 (75.2) | 35 (77.8) | 0.84 | |
| Yes | 38 (24.1) | 28 (24.8) | 10 (22.2) | ||
| Tumor site | |||||
| Nasopharynx | 52 (32.9) | 26 (23.0) | 26 (57.8) | <0.01 | |
| Oral cavity/oropharynx | 68 (43.0) | 58 (51.3) | 10 (22.2) | ||
| Larynx/hypopharynx | 38 (24.1) | 29 (25.7) | 9 (20.0) | ||
| AJCC stage | |||||
| I-II | 40 (25.3) | 26 (23.0) | 14 (31.1) | 0.32 | |
| III-IV | 118 (74.7) | 87 (77.0) | 31 (68.9) | ||
| T status | |||||
| T1-2 | 80 (50.6) | 55 (48.7) | 25 (55.6) | 0.48 | |
| T3-4 | 78 (49.4) | 58 (51.3) | 20 (44.4) | ||
| N status | |||||
| N0-1 | 78 (49.4) | 61 (54.0) | 17 (37.8) | 0.08 | |
| N2-3 | 80 (50.6) | 52 (46.0) | 28 (62.2) | ||
| RT intent | |||||
| Definite | 110 (69.6) | 71 (62.8) | 39 (86.7) | <0.01 | |
| Adjuvant | 48 (30.4) | 42 (37.2) | 6 (13.3) | ||
| RT dose, Gy, median (range) | |||||
| <66 Gy | 44 (27.8) | 38 (33.6) | 6 (13.3) | <0.01 | |
| ≥66 Gy | 114 (72.2) | 75 (66.4) | 39 (86.7) | ||
| Image guidance | |||||
| No | 48 (30.4) | 48 (42.5) | 0 (0.0) | <0.01 | |
| Yes | 110 (69.6) | 65 (57.5) | 42 (100.0) | ||
| Concurrent weekly cisplatin cycles, median (range) | 4 (0-7) | 4 (0-7) | 4 (0-7) | 0.56 | |
| Skindex-16 completed | 0.92 | ||||
| T1 | 158 (100) | 113 (100) | 45 (100) | ||
| T2 | 154 (97.4) | 110 (97.3) | 44 (97.8) | ||
| T3 | 142 (89.9) | 100 (88.5) | 42 (93.3) | ||
AJCC = American Joint Committee on Cancer 8th edition; IMPT = intensity-modulated proton therapy; RT = radiotherapy; T1 = initiation of RT; T2 = end of RT; T3 = 3 mo post-completion of RT; VMAT = volumetric modulated arc therapy.
3.2. ARD grade between VMAT and IMPT
As shown in Fig. 1, the temporal patterns of ARD incidence and severity across eight weekly on-treatment assessments were similar between VMAT (Fig. 1A) and IMPT (Fig. 1B). During the first three weekly on-treatment assessments (corresponding to early cumulative dose delivery), most patients were graded as CTCAE grade 0. From week 4 onward, grade 1 predominated; grades ≥2 first appeared in week 5 and peaked during weeks 6 to 8. The distribution of each patient’s maximum ARD grade over the entire course was comparable between VMAT and IMPT—grade 1: 63.0% vs 65.7%; grade 2: 30.0% vs 29.5%; grade 3: 7.0% vs 4.8%, with no significant difference (overall p = 0.85). No grade 4 ARD was observed.
Fig. 1.
Weekly incidence and severity of acute radiation dermatitis over eight on-treatment assessments in patients with head and neck cancer treated with (A) volumetric modulated arc therapy (VMAT; n = 113) or (B) intensity-modulated proton therapy (IMPT; n = 45), based on grading of CTCAE v5. IMPT = intensity-modulated proton therapy; VMAT = volumetric modulated arc therapy.
3.3. Skindex-16 outcomes between VMAT and IMPT
Across all Skindex-16 domains (Global, Emotional, Functional, and Symptom), skin-related QoL worsened significantly at T2 and improved by T3, with a strong overall time effect observed in all domains (all p for time effect <0.01, Table 2). However, no significant differences were observed between VMAT and IMPT at any time point. Adjusted mean differences between IMPT and VMAT were small and not statistically significant at T1, T2, or T3 across all domains (all p for VMAT vs IMPT >0.20); additionally, there was no significant interaction between time and RT modality for any Skindex-16 domain (all p for interaction >0.50), indicating that the trajectory of skin-related QoL over time was similar between VMAT and IMPT. These findings suggest that temporal changes in patient-reported skin toxicity were driven primarily by RT courses rather than RT modality.
Table 2.
Skindex-16 outcome for patients treated by VMAT (n = 113) vs IMPT (n = 45) at different time points
| Domain | Timepoint | VMAT (mean ± SE)a | IMPT (mean ± SE)a | Adjusted mean differenceb (95% CI) | p (VMAT vs IMPT) | p (Time effect)c | p (Time × RT modality)d |
|---|---|---|---|---|---|---|---|
| Global | T1 | 2.2 ± 1.2 | 1.2 ± 1.5 | −0.3 (−4.2, 3.7) | 0.89 | <0.01 | 0.94 |
| T2 | 14.0 ± 1.6 | 13.2 ± 2.4 | 0.0 (−4.0, 4.0) | 0.98 | |||
| T3 | 3.1 ± 1.3 | 1.7 ± 1.7 | −0.9 (−5.1, 3.3) | 0.66 | – | ||
| Emotional | T1 | 2.4 ± 0.9 | 1.5 ± 1.1 | 0.0 (−4.0, 4.0) | 0.99 | <0.01 | 0.96 |
| T2 | 14.8 ± 1.7 | 13.1 ± 2.6 | −0.7 (−4.8, 3.4) | 0.72 | |||
| T3 | 3.5 ± 1.0 | 2.1 ± 1.4 | −0.8 (−5.0, 3.5) | 0.73 | – | ||
| Functional | T1 | 1.7 ± 1.4 | 0.4 ± 1.7 | −0.2 (−4.8, 4.4) | 0.92 | <0.01 | 0.57 |
| T2 | 14.5 ± 1.8 | 16.4 ± 2.5 | 3.0 (−1.6, 7.7) | 0.20 | |||
| T3 | 0.6 ± 0.7 | 0.4 ± 1.1 | −0.6 (−5.4, 4.3) | 0.82 | – | ||
| Symptom | T1 | 1.8 ± 0.6 | 0.7 ± 0.7 | −0.7 (−4.8, 3.4) | 0.74 | <0.01 | 0.81 |
| T2 | 11.5 ± 1.8 | 9.0 ± 2.8 | −2.1 (−6.2, 2.1) | 0.33 | |||
| T3 | 2.5 ± 0.8 | 0.7 ± 1.2 | −1.5 (−5.9, 2.8) | 0.49 | – |
Linear mixed-effects models were used for analysis with controlling age, gender, smoking habit, diabetes, education years, body mass index, tumor site, AJCC stage, RT intent, RT dose, image guidance, concurrent weekly cisplatin cycles, and RT modality. Bold values represent statistically significant results (p < 0.05).
IMPT = intensity-modulated proton therapy; RT = radiotherapy; T1 = initiation of RT; T2 = end of RT; T3 = 3 mo post-completion of RT; VMAT = volumetric modulated arc therapy.
Estimated marginal mean values with SE were derived.
IMPT minus VMAT.
p value: T2 as reference vs T1 or T3.
p value >0.05 means no interaction.
3.4. ARD grade related to Skindex-16 outcomes
At T2, higher physician-reported ARD severity was strongly associated with worse patient-reported skin outcomes, with significantly higher adjusted Skindex-16 Global, Emotional, Functional, and Symptom scores observed in patients with ARD grades 2 or 3 compared with grade 1, demonstrating a clear severity-dependent relationship during the acute treatment phase (Table 3). In contrast, at T3, Skindex-16 scores had markedly decreased across all domains, and no significant differences were observed by maximal ARD grade (Table 4), indicating substantial recovery of skin-related QoL regardless of acute ARD severity. Together, these findings suggest that while physician-reported ARD severity closely reflects patient-reported skin burden during active treatment, its impact on PROs largely resolves within 3 months following RT.
Table 3.
Estimated marginal mean values for Skindex-16 scores at the end of radiotherapy (T2) stratified by maximal grade of acute radiation dermatitis during radiotherapy (n = 154)
| G1 (n = 98) | G2 (n = 46) | G3 (n = 10) | p | |
|---|---|---|---|---|
| Global | 14.2 ± 3.4 | 27.3 ± 3.8 | 29.59 ± 6.6 | <0.05a,b |
| Emotion | 15.1 ± 3.3 | 26.5 ± 3.8 | 30.2 ± 6.5 | <0.05a,b |
| Function | 13.3 ± 3.8 | 30.4 ± 4.3 | 33.9 ± 7.4 | <0.05a,b |
| Symptom | 14.2 ± 3.7 | 25.1 ± 4.3 | 24.2 ± 7.4 | <0.05a |
Grade (G) of acute radiation dermatitis was based on CTCAE v5.0; Estimated marginal mean values ± SE were obtained from linear mixed-effects models adjusting baseline, age, gender, smoking habit, diabetes, education years, body mass index, tumor site, AJCC stage, RT intent, RT dose, image guidance, concurrent weekly cisplatin cycles, and RT modality.
AJCC = American Joint Committee on Cancer 8th edition; RT = radiotherapy; T2 = end of RT.
p < 0.05 while G2 vs G1
p < 0.05 while G3 vs G1.
Table 4.
Estimated marginal mean values for Skindex-16 scores at 3 mo post-completion of radiotherapy (T3) stratified by maximal grade of acute radiation dermatitis during radiotherapy (n = 142)
| G1 (n = 92) | G2 (n = 41) | G3 (n = 9) | p | |
|---|---|---|---|---|
| Global | 2.4 ± 1.4 | 3.5 ± 1.5 | 3.6 ± 2.7 | 0.72 |
| Emotion | 3.2 ± 1.4 | 4.4 ± 1.5 | 5.0 ± 2.7 | 0.57 |
| Function | 1.9 ± 1.5 | 2.9 ± 1.7 | 2.8 ± 2.9 | 0.79 |
| Symptom | 2.1 ± 1.5 | 2.8 ± 1.6 | 2.8 ± 2.8 | 0.15 |
Grade (G) of acute radiation dermatitis was based on CTCAE v5.0; Estimated marginal mean values ± SE were obtained from linear mixed-effects models adjusting for baseline, age, gender, smoking habit, diabetes, education years, body mass index, tumor site, AJCC stage, RT intent, RT dose, image guidance, concurrent weekly cisplatin cycles, and RT modality.
AJCC = American Joint Committee on Cancer 8th edition; RT = radiotherapy; T3 = 3 mo post-completion of RT.
3.5. Predictors of ARD
In multivariable logistic regression analysis, advanced nodal disease (N2-3) and higher prescribed RT dose (≥66 Gy) were independently associated with an increased risk of developing ARD grades ≥2 (Table 5). Patients with N2-3 disease had more than 3-fold higher odds of these grades compared with those with N0-1 disease (OR: 3.2, 95% CI, 1.5-6.8; p < 0.01), while receipt of ≥66 Gy was associated with a similarly elevated risk (OR: 3.5, 95% CI, 1.1-16.5; p = 0.02). In contrast, demographic factors, diabetes, tumor site, RT intent, image guidance, concurrent cisplatin cycles, and RT modality (IMPT vs VMAT) were not significantly associated with these grades ≥2. These findings suggest that disease extent and RT intensity, rather than RT modality, are the primary determinants of clinically significant ARD.
Table 5.
Multivariable logistic regression model of predictors of acute radiation dermatitis ≥ G2
| Parameter | ≥G2 | |
|---|---|---|
| OR (95% CI) | p | |
| Age, continuous (per 10 y) | 0.9 (0.6, 1.2) | 0.43 |
| Gender (male vs female) | 0.9 (0.3, 2.5) | 0.81 |
| Body mass index (per kg/m2) | 1.0 (0.9, 1.1) | 0.55 |
| Education years (>12 y vs ≤12) | 1.0 (0.4, 2.3) | 0.98 |
| Smoking habit (yes vs no) | 1.7 (0.7, 4.6) | 0.25 |
| Diabetes (yes vs no) | 1.1 (0.5, 2.7) | 0.77 |
| Tumor site (nasopharynx vs others) | 1.2 (0.5, 2.8) | 0.76 |
| N status (2-3 vs 0-1) | 3.2 (1.5, 6.8) | <0.01 |
| RT intent (definite vs adjuvant) | 1.6 (0.4, 6.7) | 0.55 |
| RT dose (≥66 Gy vs <66 Gy) | 3.5 (1.1, 16.5) | 0.02 |
| RT modality (IMPT vs VMAT) | 0.6 (0.2, 1.5) | 0.27 |
| Image guidance (yes vs no) | 1.1 (0.5, 2.3) | 0.82 |
| Concurrent weekly cisplatin (per cycle) | 1.0 (0.8, 1.2) | 0.81 |
Grade (G) of acute radiation dermatitis was based on CTCAE v5.0. Bold values show the statistical significance.
IMPT, intensity-modulated proton therapy; OR, odds ratio; RT, radiotherapy; VMAT, volumetric modulated arc therapy.
3.6. Predictors of Skindex-16 global score
Multivariable analyses of Skindex-16 Global scores demonstrated a consistent pattern across continuous and dichotomized outcomes (Tables 6 and 7). Clinically meaningful ≥10-point worsening occurred in 37.0% of patients at T2 but only 7.0% at T3. At T2, worse patient-reported skin outcomes were independently associated with advanced nodal disease (N2-3) and higher prescribed RT dose (≥66 Gy), both as higher Global scores in linear regression and as increased odds of a clinically meaningful ≥10-point worsening. Additional factors, including gender, education years, smoking status, and definitive RT intent, were associated with acute Global score severity at T2 in linear models, with female sex and higher education also predicting ≥10-point worsening. In contrast, RT modality (IMPT vs VMAT) was not significantly associated with either Global score severity or ≥10-point worsening at any time point. At T3, baseline Global score and N status showed modest associations with continuous Global scores, whereas no clinical or treatment-related factors predicted persistent ≥10-point worsening, reflecting substantial recovery of skin-related QoL. Together, these findings indicate that acute PRO of skin toxicity is driven primarily by disease extent and RT intensity, with clinically meaningful deterioration largely resolving after treatment completion.
Table 6.
Multivariable linear regression of predictors associated with Skindex-16 Global score at T2 and T3
| Parameter | Global score at T2 | Global score at T3 | ||
|---|---|---|---|---|
| β (95% CI) | p | β (95% CI) | p | |
| Global score at baseline (T1) | 0.0 (−0.6, 0.6) | 0.97 | 0.4 (0.1, 0.7) | <0.01 |
| Age, continuous (per 10 y) | 0.1 (−2.4, 2.7) | 0.93 | 0.6 (−0.5, 1.7) | 0.27 |
| Gender (male vs female) | −13.7 (−21.4, −6.0) | <0.01 | −1.2 (−4.4, 2.0) | 0.47 |
| Body mass index (per kg/m2) | 0.0 (−0.6, 0.6) | 0.92 | −0.2 (−0.4, 0.0) | 0.10 |
| Education years (>12 vs ≤12) | 11.3 (4.7, 18.0) | <0.01 | 1.1 (−1.7, 3.9) | 0.44 |
| Smoking habit (yes vs no) | 7.3 (0.4, 14.3) | 0.04 | 1.3 (−1.6, 4.2) | 0.39 |
| Diabetes (yes vs no) | 2.5 (−4.0, 9.1) | 0.45 | 0.1 (−2.6, 2.9) | 0.92 |
| Tumor site (nasopharynx vs others) | 0.9 (−6.1, 8.0) | 0.79 | −0.4 (−3.4, 2.5) | 0.78 |
| N status (2-3 vs 0-1) | 4.2 (1.1, 10.4) | 0.02 | 2.6 (0.6, 5.1) | 0.04 |
| RT intent (definite vs adjuvant) | 13.8 (3.1, 24.4) | 0.01 | 3.2 (−1.2, 7.6) | 0.16 |
| RT dose (≥66 Gy vs <66 Gy) | 18.80 (7.8, 29.9) | <0.01 | 3.0 (−1.5, 7.6) | 0.18 |
| RT modality (IMPT vs VMAT) | −5.0 (−11.6, 1.6) | 0.14 | −0.9 (−3.8, 1.9) | 0.51 |
| Image guidance (yes vs no) | −0.3 (−6.0, 5.4) | 0.92 | 1.9 (−0.5, 4.4) | 0.13 |
| Concurrent weekly cisplatin (per cycle) | 1.2 (−0.3, 1.6) | 0.11 | −0.1 (−0.7, 0.5) | 0.81 |
Bold values represent statistically significant results (p < 0.05).
β = β coefficient; IMPT = intensity-modulated proton therapy; RT = radiotherapy; T1 = initiation of RT; T2 = end of RT; T3 = 3 mo post-completion of RT; VMAT = volumetric modulated arc therapy.
Table 7.
Logistic regression of predictors of ≥10-point worsening from baseline (T1) in Skindex-16 Global score at T2 and T3
| Parameter | Global score at T2 | Global score at T3 | ||
|---|---|---|---|---|
| OR (95% CI) | p | OR (95% CI) | p | |
| Global score at baseline (T1) | 0.9 (0.9, 1.0) | 0.21 | 1.1 (0.9, 1.3) | 0.45 |
| Age, continuous (per 10 y) | 1.0 (0.7, 1.4) | 0.88 | 3.3 (0.5, 20.5) | 0.20 |
| Gender (male vs female) | 0.3 (0.1, 0.9) | 0.04 | 1.2 (0.0, 61.0) | 0.93 |
| Body mass index (per kg/m2) | 1.0 (0.9, 1.1) | 0.83 | 0.7 (0.4, 1.1) | 0.11 |
| Education years (>12 vs ≤12) | 4.1 (1.6, 10.6) | <0.01 | 6.2 (0.3, 52.8) | 0.22 |
| Smoking (yes vs no) | 2.4 (0.9, 6.7) | 0.08 | 1.1 (0.0, 29.5) | 0.98 |
| Diabetes (yes vs no) | 1.3 (0.5, 3.2) | 0.63 | 0.2 (0.0, 5.6) | 0.33 |
| Tumor site (nasopharynx vs others) | 1.4 (0.5, 3.6) | 0.52 | 2.2 (0.0, 26.9) | 0.73 |
| N status (2-3 vs 0-1) | 1.9 (1.1, 3.7) | 0.02 | 1.9 (0.1, 90.4) | 0.91 |
| RT intent (definite vs adjuvant) | 3.8 (0.8, 18.9) | 0.10 | 1.7 (0.1, 61.2) | 0.71 |
| RT dose (≥66 Gy vs <66 Gy) | 7.2 (1.4, 30.9) | 0.02 | 1.5 (0.0, 66.4) | 0.92 |
| RT modality (IMPT vs VMAT) | 0.6 (0.2, 1.5) | 0.29 | 1.4 (0.0, 76.2) | 0.84 |
| Image guidance (yes vs no) | 1.2 (0.6, 2.8) | 0.60 | 7.0 (0.6, 77.1) | 0.26 |
| Concurrent weekly cisplatin (per cycle) | 1.2 (0.9, 1.4) | 0.16 | 1.1 (0.5, 2.4) | 0.87 |
At T2, n = 57/154 (37%) patients experienced a ≥10-point worsening; at T3, n = 10/142 (7%) patients experienced a ≥10-point worsening. Wide confidence intervals at T3 reflect the limited number of events. Bold values represent statistically significant results (p < 0.05).
IMPT = intensity-modulated proton therapy; OR = odds ratio; RT = radiotherapy; T1 = initiation of RT; T2 = end of RT; T3 = 3 mo post-completion of RT; VMAT = volumetric modulated arc therapy.
4. DISCUSSION
In this prospective comparative cohort study, treatment modality (VMAT vs IMPT) did not independently determine the acute patient-reported burden of skin toxicity. Skindex-16 outcomes were highly concordant between modalities, being low at baseline, peaking at the end of RT, then improving over 3 months, with no difference in maximum ARD grade distribution. Instead, disease extent (higher N status) and prescription dose ≥66 Gy emerged as the principal drivers of both physician-reported ARD (≥G2) and worse Skindex-16 Global scores at T2. These findings suggest that, in contemporary practice, field geometry and total dose outweigh beam modality in determining acute cutaneous burden when plans are already well-optimized.
Although proton therapy has theoretical advantages in reducing entrance dose to superficial tissues, the lack of a clinically meaningful difference in ARD observed in this study suggests that such advantages may not translate into improved patient-reported skin outcomes in routine practice. Several potential technical factors, including the use of range shifters, air gaps between the range shifter and patient surface, beam arrangement, and the absence of explicit skin rind optimization, might contribute to this observation14–16; however, these explanations should be regarded as speculative, as detailed skin dosimetry and delivery-specific parameters were not prospectively collected or analyzed in the present study.
In particular, range shifters and unavoidable air gaps in pencil-beam scanning proton therapy can increase surface dose through secondary particle scatter, potentially offsetting the expected reduction in skin dose.16–18 Similarly, without routine contouring and constraining of a superficial skin rind during treatment planning, skin dose might not be explicitly minimized.19–21 Because these factors were not quantitatively evaluated, the present findings cannot establish causal relationships between specific proton delivery characteristics and acute skin toxicity.
Our multivariable results point to pragmatic planning levers. Because higher N status typically entails larger, shallower, and more wrap-around fields, planners should anticipate elevated skin dose and consider explicit skin objectives during optimization.22,23 A practical approach is to contour a 2 to 3 mm skin structure over the target-projected OARs and include soft constraints on dose–volume endpoints while maintaining target robustness.22 For IMPT, minimizing range-shifter use, reducing air gaps, and selecting beam angles that avoid long skin traversals can lower surface dose.15,16,24 For VMAT, avoiding arcs that skim the skin and managing buildup/bite-block influences can be beneficial.17 Prospective documentation of dose–volume metrics at surface skin alongside ARD/Skindex outcomes would help define site-specific dose–response relationships and inform constraint development.19,23,25
The patient-reported perspective adds nuance. At T2, Skindex-16 Global, Emotional, Functional, and Symptom scores scaled with ARD grade, validating ARD as a clinically meaningful surrogate of patient experience.3 By T3, most patients improved, but residual Global/Emotional differences remained for those who had grade 2 ARD at T2, indicating a short-term psychosocial and symptom tail even after visible healing, while PROs capture symptom burden not fully reflected by physician-reported grading.6,26 Female gender and smoking were associated with higher Global scores at T2, consistent with prior reports of sex-related symptom sensitivity and smoking-related skin vulnerability27–29; higher education also correlated with higher scores, possibly reflecting expectation and reporting effects.30,31 In clinical practice, integrating a brief PRO (eg, Skindex-16 Global) into on-treatment visits can sharpen triage, complementing CTCAE.32
Methodologically, several choices in the study strengthen internal validity: adjustment for baseline PRO, analysis of both continuous outcomes and a clinically meaningful worsening threshold (≥10 points), and inclusion of N status and dose in the multivariable framework. Still, limitations remain: nonrandomized design with residual confounding, limited power for modality contrasts and for grade 3 events, lack of skin-specific dosimetry,16 and three discrete PRO timepoints rather than weekly capture.33 These factors might obscure subtle modality effects and preclude firm dose–response inferences.
Clinically, these data support a risk-adapted rather than modality-driven approach to ARD. In the present cohort, a standardized skin-care protocol was applied uniformly across VMAT and IMPT. Based on the observed associations with nodal extent and dose (≥66 Gy), future practice could consider additional risk-adapted measures for higher-risk patients, including (i) proactive skin-sparing planning (eg, skin rind objectives, careful arc or beam selection, and robust surface checks)15,16,19; (ii) earlier and more standardized ARD prophylaxis beginning in week 1 (eg, topical corticosteroids, friction control, and early dressings in high-shear areas)2,34; and (iii) structured weekly PRO collection to facilitate timely symptom-based scenarios.32,33 These strategies are modality-independent and target the key determinants identified in this study—field extent and dose.
Future work should pursue (1) propensity-matched or randomized VMAT–IMPT comparisons powered for both ARD grades ≥ 2 and Skindex-16 endpoints; (2) incorporation of skin dosimetry and beamline parameters (range-shifter use, air gap, number of fields) to model cutaneous dose–response; (3) weekly ePROs to map symptom kinetics and detect transient modality differences33; and (4) adaptive workflows (plan re-optimization with weight change or mask loosening) to avoid late-course skin hotspots when ARD risk peaks.2 These steps will clarify when, and for whom, protons can meaningfully reduce cutaneous burden beyond best-in-class VMAT—and how planning plus supportive care can work synergistically to improve the patient experience without compromising tumor control.
In conclusion, in modern head-and-neck RT, ARD severity, closely linked to nodal extent and dose, explains acute patient-reported skin burden more than beam modality. Focusing on skin-aware planning and proactive, PRO-guided supportive care is likely to yield the greatest clinical benefit across both VMAT and IMPT.
ACKNOWLEDGMENTS
This study was supported by the Chang Gung Memorial Hospital and Cheng Shiu University (grant CORPG8N0131 and CORPG8P0041); and the National Science and Technology Council, Taiwan (NSTC 114-2637-B-230-001). The funding agencies had no role in study design, data collection, data analysis, and interpretation of data.
APPENDIX A. SUPPLEMENTARY DATA
Supplementary data related to this article can be found at https://links.lww.com/JCMA/A420.
Supplementary Material
Footnotes
Author contributions: Dr. Wen-Ling Tsai and Dr. Ko-Chun Fang contributed equally to this work.
Conflicts of interest: The authors declare that they have no conflicts of interest related to the subject matter or materials discussed in this article.
REFERENCES
- 1.Jain S, Kumar P, Kumar P, Nigam J, Silambarasan NS, Sachan R. Intensity-modulated radiotherapy versus volumetric modulated arc therapy in head and neck cancers: a comparative analysis of compliance, toxicities and dosimetric parameters. Cureus 2025;17:e86143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Jeans EB, Shiraishi S, Manzar G, Morris LK, Amundson A, McGee LA, et al. An comparison of acute toxicities and patient-reported outcomes between intensity-modulated proton therapy and volumetric-modulated arc therapy after ipsilateral radiation for head and neck cancers. Head Neck 2022;44:359–71. [DOI] [PubMed] [Google Scholar]
- 3.Manzar GS, Lester SC, Routman DM, Harmsen WS, Petersen MM, Sloan JA, et al. Comparative analysis of acute toxicities and patient reported outcomes between intensity-modulated proton therapy (IMPT) and volumetric modulated arc therapy (VMAT) for the treatment of oropharyngeal cancer. Radiother Oncol 2020;147:64–74. [DOI] [PubMed] [Google Scholar]
- 4.Behroozian T, Goldshtein D, Ryan Wolf J, van den Hurk C, Finkelstein S, Lam H, et al. ; Multinational Association of Supportive Care in Cancer (MASCC) Oncodermatology Study Group Radiation Dermatitis Guidelines Working Group. MASCC clinical practice guidelines for the prevention and management of acute radiation dermatitis: part 1) systematic review. EClinicalMedicine 2023;58:101886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rzepecki A, Birnbaum M, Ohri N, Daily J, Fox J, Bodner W, et al. Characterizing the effects of radiation dermatitis on quality of life: a prospective survey-based study. J Am Acad Dermatol 2022;86:161–3. [DOI] [PubMed] [Google Scholar]
- 6.Freites-Martinez A, Santana N, Arias-Santiago S, Viera A. Using the common terminology criteria for adverse events (CTCAE - Version 5.0) to evaluate the severity of adverse events of anticancer therapies. Actas Dermosifiliogr (Engl Ed) 2021;112:90–2. [DOI] [PubMed] [Google Scholar]
- 7.Rahman SM, Patel P, Finkelstein S, Lee SF, Wong H, Chan AW, et al. ; Multinational Association of Supportive Care in Cancer MASCC Oncodermatology Study Group Radiation Dermatitis Guidelines Working Group. Delphi consensus-based recommendations of tools and outcomes for the assessment of radiation dermatitis severity. Radiother Oncol 2025;206:110846. [DOI] [PubMed] [Google Scholar]
- 8.Day M, Kuszaj O, Zhang L, Wong H, Lee SF, Kwan J, et al. Validation of the Skindex-16 questionnaire for assessing patient-reported outcomes in radiation dermatitis among breast cancer patients. Support Care Cancer 2025;33:192. [DOI] [PubMed] [Google Scholar]
- 9.Liao KC, Huang YJ, Tsai WL, Lee CH, Fang FM. Longitudinal assessment of quality of life in nasopharyngeal cancer patients treated with intensity-modulated proton therapy and volumetric modulated arc therapy at different time points. Cancers (Basel) 2024;16:1217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lee N, Harris J, Garden AS, Straube W, Glisson B, Xia P, et al. Intensity-modulated radiation therapy with or without chemotherapy for nasopharyngeal carcinoma: radiation therapy oncology group phase II trial 0225. J Clin Oncol 2009;27:3684–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bentzen SM, Constine LS, Deasy JO, Eisbruch A, Jackson A, Marks LB, et al. Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC): an introduction to the scientific issues. Int J Radiat Oncol Biol Phys 2010;76(3 Suppl):S3–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Marks LB, Yorke ED, Jackson A, Ten Haken RK, Constine LS, Eisbruch A, et al. Use of normal tissue complication probability models in the clinic. Int J Radiat Oncol Biol Phys 2010;76(3 Suppl):S10–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.He Z, Lu C, Chren MM, Zhang Z, Li Y, Ni X, et al. Development and psychometric validation of the Chinese version of Skindex-29 and Skindex-16. Health Qual Life Outcomes 2014;12:190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hsu RT, Pan CY, Tsai WN, Liao XP, Su MH, Lin CW, et al. Key insights and implementation of a patient-centered education video for managing acute radiation dermatitis in breast cancer: a single-center pilot study. Adv Radiat Oncol. 2025;10:101730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Braun H, Yeung H, Chen SC. Research techniques made simple: developing and validating QOL outcome measures for skin diseases. J Invest Dermatol 2020;140:1888–92. [DOI] [PubMed] [Google Scholar]
- 16.Shen J, Liu W, Anand A, Stoker JB, Ding X, Fatyga M, et al. Impact of range shifter material on proton pencil beam spot characteristics. Med Phys 2015;42:1335–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lin H, Shi C, Huang S, Shen J, Kang M, Chen Q, et al. Applications of various range shifters for proton pencil beam scanning radiotherapy. Radiat Oncol 2021;16:146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kern A, Bäumer C, Kröninger K, Wulff J, Timmermann B. Impact of air gap, range shifter, and delivery technique on skin dose in proton therapy. Med Phys 2021;48:831–40. [DOI] [PubMed] [Google Scholar]
- 19.Hunte SO, Clark CH, Zyuzikov N, Nisbet A. Volumetric modulated arc therapy (VMAT): a review of clinical outcomes-what is the clinical evidence for the most effective implementation? Br J Radiol 2022;95:20201289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Penoncello GP, Ding GX. Skin dose differences between intensity-modulated radiation therapy and volumetric-modulated arc therapy and between boost and integrated treatment regimens for treating head and neck and other cancer sites in patients. Med Dosim 2016;41:80–6. [DOI] [PubMed] [Google Scholar]
- 21.Chang L, Shaaban SG, Gogineni E, Page B, Quon H, Li H, et al. Daily head and neck treatment assessment for optimal proton therapy planning robustness. Cancers (Basel). 2023;15:3719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nguyen ML, Cantrell JN, Ahmad S, Henson C. Intensity-modulated proton therapy (IMPT) versus intensity-modulated radiation therapy (IMRT) for the treatment of head and neck cancer: a dosimetric comparison. Med Dosim 2021;46:259–63. [DOI] [PubMed] [Google Scholar]
- 23.Mizuno T, Tomita N, Takaoka T, Tomida M, Fukuma H, Tsuchiya T, et al. Dosimetric comparison of helical tomotherapy, volumetric-modulated arc therapy, and intensity-modulated proton therapy for angiosarcoma of the scalp. Technol Cancer Res Treat 2021;20:1533033820985866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bonomo P, Talamonti C, Desideri I, Marrazzo L, Pezzulla D, Rampini A, et al. Analysis of skin dose distribution for the prediction of severe radiation dermatitis in head and neck squamous cell carcinoma patients treated with concurrent chemo-radiotherapy. Head Neck 2020;42:244–53. [DOI] [PubMed] [Google Scholar]
- 25.Basu S, Chatterjee S, Chatterjee K, Samanta S, Saha S, Hossain ST, et al. Correlation of degree of acute radiation dermatitis (RD) with skin dose distribution in head and neck squamous cell carcinoma patients treated with definitive concurrent chemoradiation. Rep Pract Oncol Radiother 2024;29:579–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Shirey RJ, Wu HT. Quantifying the effect of air gap, depth, and range shifter thickness on TPS dosimetric accuracy in superficial PBS proton therapy. J Appl Clin Med Phys 2018;19:164–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kawamura M, Yoshimura M, Asada H, Nakamura M, Matsuo Y, Mizowaki T. A scoring system predicting acute radiation dermatitis in patients with head and neck cancer treated with intensity-modulated radiotherapy. Radiat Oncol 2019;14:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rajeswaran T, Gojsevic M, Zhang L, Kennedy SKF, Karam I, Ding K, et al. Correlation of CTCAE and patient-reported breast radiation dermatitis symptom scores. Support Care Cancer 2024;32:801. [DOI] [PubMed] [Google Scholar]
- 29.Spałek M. Chronic radiation-induced dermatitis: challenges and solutions. Clin Cosmet Investig Dermatol 2016;9:473–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Unger JM, Vaidya R, Albain KS, LeBlanc M, Minasian LM, Gotay CC, et al. Sex differences in risk of severe adverse events in patients receiving immunotherapy, targeted therapy, or chemotherapy in cancer clinical trials. J Clin Oncol 2022;40:1474–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rades D, Zwaan I, Janssen S, Yu NY, Schild SE, Idel C, et al. Evaluation of the impact of smoking and alcohol consumption on toxicity and outcomes of chemoradiation for head and neck cancer. Anticancer Res 2023;43:823–30. [DOI] [PubMed] [Google Scholar]
- 32.Corish S, Fulton BA, Galbraith L, Coltart K, Duffton A. Impact of patient information format on the experience of cancer patients treated with radiotherapy. Tech Innov Patient Support Radiat Oncol 2024;30:100252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.ElMokhallalati Y, Alaloul E, Shatat M, Shneewra T, El Massri S, Shaer O, et al. The symptom burden and quality of life in cancer patients in the Gaza Strip, Palestine: a cross-sectional study. PLoS One 2022;17:e0262512. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Basch E, Reeve BB, Mitchell SA, Clauser SB, Minasian LM, Dueck AC, et al. Development of the National Cancer Institute’s patient-reported outcomes version of the common terminology criteria for adverse events (PRO-CTCAE). J Natl Cancer Inst 2014;106:dju244. [DOI] [PMC free article] [PubMed] [Google Scholar]