Survival outcomes of radiotherapy alone versus concurrent chemoradiotherapy in T1–2 head and neck cancer with low-volume disease: A multicenter cohort study Authors: Chuanhao Zhang, Zhichao Cheng, Xin Jiang, Genghao Zhao, Yanmei Zhu, Bowen Hou, Yingming Sun, Shanshan Liang, Ye Zhang, Ruoyu Wang, Zhe Wang Published: 2026 ABSTRACT BackgroundThe optimal treatment strategy between definitive radiotherapy and concurrent chemoradiotherapy (CCRT) for patients with T1–2 head and neck cancer with low-volume nodal disease remains unclear due to limited supporting data. This study aimed to compare overall survival (OS) and progression-free survival (PFS) between the two treatment modalities.MethodsPatients with AJCC 7th edition stage T1–2N1–2a head and neck squamous cell carcinoma (HNSCC) were collected from multiple research centers, including the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences (NCC), the University Health Network (UHN), and Quebec, Canada. Inverse probability of treatment weighting (IPTW) was applied to balance baseline differences. Kaplan-Meier analysis and Cox proportional hazards models were used to compare OS and PFS between the two groups.ResultsA total of 240 patients were included, with a median age of 57.8 years (IQR 52.0–65.1); 51 patients (21%) were HPV-negative. After IPTW adjustment, multivariable Cox analysis showed that CCRT did not provide a significant OS benefit compared with radiotherapy alone (hazard ratio [HR], 0.50; 95% CI, 0.20–1.24; P = 0.14). However, CCRT was associated with a significant improvement in PFS (HR, 0.42; 95% CI, 0.19–0.93; P = 0.03). Consistent with the NRG-HN002 trial, which showed a 9.7% absolute reduction in locoregional failure with CCRT relative to radiotherapy alone.ConclusionsThis multicenter study demonstrated that while CCRT conferred limited OS benefit in patients with stage T1–2N1–2a HNSCC, it improved tumor progression control in a subset of patients. Careful evaluation of survival benefit versus treatment-related toxicity is warranted, and larger prospective studies are needed to validate these findings. Introduction Oropharyngeal cancer represents a major subtype of head and neck squamous cell carcinoma (HNSCC), with more than 70% of cases attributable to human papillomavirus (HPV) infection [1]. This distinct etiology has drawn considerable attention and spurred ongoing refinement of therapeutic strategies. Nevertheless, concurrent chemoradiotherapy (CCRT) remains the established standard for locoregionally advanced HNSCC. The pivotal RTOG 94–01 trial demonstrated that CCRT improved overall survival (OS) and locoregional control compared with radiotherapy alone, forming the basis for its adoption in international guidelines as the preferred treatment for patients with locoregionally advanced disease [2], [3], [4], [5]. However, this benefit comes at the cost of substantial treatment-related toxicity, which can profoundly impair quality of life [6], [7]. For patients with low tumor burden but formally advanced disease, such as those with T1–2N1 tumors, the necessity of adding chemotherapy remains uncertain. Notably, current NCCN, ASTRO, and ESTRO guidelines provide no specific recommendations for this subgroup, leaving clinical decision-making largely dependent on physician judgment and institutional practice. Prior analyses using the National Cancer Database (NCDB) suggested an OS advantage with CCRT in HNSCC, particularly among HPV-positive tumors [8], [9], [10]. More recently, however, a study reported no survival benefit from CCRT in patients with T1–2N2a disease, with improved locoregional control observed only in T1–2N2b patients aged ≤70 years [11]. Importantly, HPV-negative tumors accounted for only 2.9% of that cohort, raising concerns that the findings may have been predominantly driven by HPV-positive disease. In summary, whether concurrent chemotherapy provides a definitive survival benefit in patients with T1–2N1–2a HNSCC remains unresolved, particularly in the context of HPV-related heterogeneity and the lack of subgroup-specific guideline recommendations. To address this evidence gap, we conducted a multicenter analysis, which included a clinically relevant proportion of HPV-negative HNSCC (21.3%), to evaluate real-world outcomes of radiotherapy alone versus concurrent chemoradiotherapy in patients with stage T1–2N1–2a HNSCC. Materials and methods Patient sources All clinical data used in this study were de-identified, and thus informed consent was waived. This study was approved by the Institutional Review Board of Affiliated Zhongshan Hospital of Dalian University. Data from University Health Network (UHN) [12], and Quebec [13] were publicly available in the TCIA database and exempt from additional ethical review. This study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [14]. Patient information We included 240 patients with T1–2N1–2a HNSCC, which was defined as low-volume disease in this study. Patients with missing survival data, unknown HPV status, or insufficient radiotherapy doses were excluded. Demographic and clinical variables collected included age, sex, eastern cooperative oncology group performance status (ECOG PS), pack-years of smoking (Smoking PY), tumor stage, HPV status, radiotherapy details (total dose and fractionation), concurrent chemotherapy details (drug, dose, and cycle), and follow-up information. HPV positivity was defined as p16 positivity, which was determined by strong and diffuse nuclear and cytoplasmic staining in ≥70% of tumor cells. All patients received curative-intent radiotherapy. Radiotherapy alone was administered due to guideline recommendation, medical contraindication, frailty, advanced age, or occasional refusal of chemotherapy. Concurrent chemotherapy consisted of either weekly platinum-based regimens or triweekly high-dose platinum, with some variation across centers. Study endpoints Progression-free survival (PFS) was designated as the primary endpoint, with overall survival (OS) as the secondary endpoint to compare outcomes between radiotherapy alone and concurrent chemoradiotherapy. Statistical analysis Missing data for ECOG PS and Smoking PY (5.8%, 14/240) were imputed using the multiple imputation by chained equations (MICE) method, with random forest algorithms applied for both variables [15], [16]. Inverse probability of treatment weighting (IPTW) was employed to control for known confounders, and standardized mean differences (SMD) were calculated to assess covariate balance between groups. To minimize the influence of extreme weights and assess robustness, weights were truncated at the 1st and 99th percentiles. Categorical variables were analyzed usingχ2tests, and continuous variables were compared using independent t-tests. Kaplan-Meier analysis was used to estimate OS, and Cox proportional hazards models were employed to calculate hazard ratios. Schoenfeld residuals were used to verify the proportional hazards assumption. To evaluate treatment effects across subgroups, Cox models with interaction terms were constructed, including main effects and covariates such as age, sex, ECOG PS, and tumor stage. These covariates were selected a priori based on established clinical relevance and prior literature, rather than being determined through univariate Cox regression screening. Sensitivity analyses were performed by excluding patients with incomplete covariate data and those who received altered fractionation schedules. All statistical tests were two-sided, with P < 0.05 considered statistically significant. Analyses were conducted using R version 4.3.2 (R Foundation for Statistical Computing). Results Patient characteristics A total of 240 patients with T1–2N1–2a HNSCC met the inclusion criteria (eFig. 1), including 51 HPV-negative (21%) and 189 HPV-positive (79%) cases, consistent with current epidemiological trends [1]. Compared with the radiotherapy-alone group, patients receiving CCRT were younger at diagnosis (mean age, 55.5 vs. 61.1 years; P < 0.001). Notably, the proportion of patients with ECOG PS 1–2 was comparable between groups (31.6% vs. 40.5%; P = 0.19). The median follow-up was 48.2 months (IQR 30.0–70.4), during which 49 patients experienced disease progression or death. Detailed baseline characteristics are presented in Table 1.Table 1Baseline Characteristics of Head and Neck Cancer.No. (%) of PatientsCharacteristicOverall(N = 240)RT alone(N = 126)CCRT(N = 114)P valueAge, yMean (SD)58.4(9.7)61.1 (10.3)55.5 (8.0)<0.001Sex0.11Female54 (22.5)34 (27.0)20 (17.5)Male186 (77.5)92 (73.0)94 (82.5)ECOG PS0.19ECOG 0153 (63.8)75 (59.5)78 (68.4)ECOG 180 (33.3)44 (34.9)36 (31.6)ECOG 27 (3.0)7 (5.6)0 (0)Smoking PY, nMedian (IQR)10.0 (0.0–30.0)20.0 (0.0–30.0)7.0 (0.0–33.8)0.09Subsite0.22Hypopharynx3 (1.3)3 (2.4)0 (0)Larynx3 (1.3)2 (1.6)1 (0.9)Oropharynx234 (97.5)121 (96.0)113 (99.1)T0.04T1 (T1N1–2a)89 (37.1)55 (43.7)34 (29.8)T2 (T2N1–2a)151 (62.9)71 (56.4)80 (70.2)N<0.001N1145 (60.4)92 (73.0)53 (46.5)N2a95 (39.6)34 (27.0)61 (53.5)HPV0.14HPV positive189 (78.8)94(74.6)95 (83.3)HPV negative51 (21.3)32 (25.4)19 (16.7) Table 1Baseline Characteristics of Head and Neck Cancer.No. (%) of PatientsCharacteristicOverall(N = 240)RT alone(N = 126)CCRT(N = 114)P valueAge, yMean (SD)58.4(9.7)61.1 (10.3)55.5 (8.0)<0.001Sex0.11Female54 (22.5)34 (27.0)20 (17.5)Male186 (77.5)92 (73.0)94 (82.5)ECOG PS0.19ECOG 0153 (63.8)75 (59.5)78 (68.4)ECOG 180 (33.3)44 (34.9)36 (31.6)ECOG 27 (3.0)7 (5.6)0 (0)Smoking PY, nMedian (IQR)10.0 (0.0–30.0)20.0 (0.0–30.0)7.0 (0.0–33.8)0.09Subsite0.22Hypopharynx3 (1.3)3 (2.4)0 (0)Larynx3 (1.3)2 (1.6)1 (0.9)Oropharynx234 (97.5)121 (96.0)113 (99.1)T0.04T1 (T1N1–2a)89 (37.1)55 (43.7)34 (29.8)T2 (T2N1–2a)151 (62.9)71 (56.4)80 (70.2)N<0.001N1145 (60.4)92 (73.0)53 (46.5)N2a95 (39.6)34 (27.0)61 (53.5)HPV0.14HPV positive189 (78.8)94(74.6)95 (83.3)HPV negative51 (21.3)32 (25.4)19 (16.7) Baseline Characteristics of Head and Neck Cancer. Survival outcomes after IPTW adjustment Given baseline differences between the radiotherapy-alone and CCRT groups, IPTW was applied to balance potential confounders. Post-weighting, baseline characteristics were well balanced between groups (Table 1, eTable 1 and eFigure 2). Survival analysis demonstrated that CCRT significantly improved PFS compared with radiotherapy alone, whereas no significant difference in OS was observed (Figure 1). These findings remained consistent after further adjustment for confounding variables (Table 2).Fig. 1Survival Differences between Radiotherapy Alone and Concurrent Chemoradiotherapy after IPTW.Overall (A) and progression-free survival (B) outcomes for all patients.Fig. 1Table 2Multivariable COX Regression Model Adjusted for Each Factor.Overall survivalProgression-free survivalMultivariableMultivariableCovariatePatients,No.HR (95% CI)P valueHR (95% CI)P valueAge, y (Mean ± SD)58.4 ± 9.71.07 (1.01–1.14)0.031.05 (1.01–1.10)0.02SexFemale541 [Reference]NA1 [Reference]NAMale1861.02 (0.34–3.06)0.971.27 (0.48–3.39)0.63ECOG PSECOG 01531 [Reference]NA1 [Reference]NAECOG 1–2871.09 (0.43–2.77)0.851.05 (0.50–2.20)0.90Smoking PY (Mean ± SD)19.0 ± 21.11.03 (1.01–1.04)<0.0011.01 (1.00–1.03)0.14TT1 (T1N1–2a)891 [Reference]NA1 [Reference]NAT2 (T2N1–2a)1512.06 (0.71–5.92)0.182.98 (1.13–7.88)0.03NN11451 [Reference]NA1 [Reference]NAN2a950.63 (0.24–1.63)0.340.50 (0.23–1.08)0.08HPVHPV positive1891 [Reference]NA1 [Reference]NAHPV negative512.54 (1.02–6.33)0.052.39 (1.12–5.10)0.02TreatmentRT alone1261 [Reference]NA1 [Reference]NACCRT1140.50 (0.20–1.24)0.140.42 (0.19–0.93)0.03 Survival Differences between Radiotherapy Alone and Concurrent Chemoradiotherapy after IPTW. Overall (A) and progression-free survival (B) outcomes for all patients. Table 2Multivariable COX Regression Model Adjusted for Each Factor.Overall survivalProgression-free survivalMultivariableMultivariableCovariatePatients,No.HR (95% CI)P valueHR (95% CI)P valueAge, y (Mean ± SD)58.4 ± 9.71.07 (1.01–1.14)0.031.05 (1.01–1.10)0.02SexFemale541 [Reference]NA1 [Reference]NAMale1861.02 (0.34–3.06)0.971.27 (0.48–3.39)0.63ECOG PSECOG 01531 [Reference]NA1 [Reference]NAECOG 1–2871.09 (0.43–2.77)0.851.05 (0.50–2.20)0.90Smoking PY (Mean ± SD)19.0 ± 21.11.03 (1.01–1.04)<0.0011.01 (1.00–1.03)0.14TT1 (T1N1–2a)891 [Reference]NA1 [Reference]NAT2 (T2N1–2a)1512.06 (0.71–5.92)0.182.98 (1.13–7.88)0.03NN11451 [Reference]NA1 [Reference]NAN2a950.63 (0.24–1.63)0.340.50 (0.23–1.08)0.08HPVHPV positive1891 [Reference]NA1 [Reference]NAHPV negative512.54 (1.02–6.33)0.052.39 (1.12–5.10)0.02TreatmentRT alone1261 [Reference]NA1 [Reference]NACCRT1140.50 (0.20–1.24)0.140.42 (0.19–0.93)0.03 Multivariable COX Regression Model Adjusted for Each Factor. Subgroup analysis suggested that the PFS benefit of CCRT was more pronounced in T2N1–2a patients (eFigure 3B), whereas no similar trend was observed in T1N1–2a patients (eFigure 3 A). However, interaction testing indicated that the difference in treatment effect between T1N1–2a and T2N1–2a groups did not reach statistical significance (HR, 0.76, 95% CI 0.14–4.15; P = 0.76). Results from NRG-HN002 demonstrated that, among patients with T1N1–2a disease, CCRT did not confer a meaningful reduction in locoregional failure compared with radiotherapy alone. In contrast, a 9.7% absolute improvement in locoregional control was observed in patients with T2N1–2a disease (Table 3). The findings of the present study are in line with those observed in the NRG-HN002 trial (eFigure 4 and eTable 2).Table 3Comparison of locoregional failure between radiotherapy alone and concurrent chemoradiotherapy in the NRG-HN002 trial.FailedRT alone(N = 48)CCRT(N = 45)Total(N = 93)T1N1–2a2/18 (11.1%)2/17 (11.8%)4/35 (11.4%)T2N1–2a4/30 (13.3%)1/28 (3.6%)5/58 (8.6%) Table 3Comparison of locoregional failure between radiotherapy alone and concurrent chemoradiotherapy in the NRG-HN002 trial.FailedRT alone(N = 48)CCRT(N = 45)Total(N = 93)T1N1–2a2/18 (11.1%)2/17 (11.8%)4/35 (11.4%)T2N1–2a4/30 (13.3%)1/28 (3.6%)5/58 (8.6%) Comparison of locoregional failure between radiotherapy alone and concurrent chemoradiotherapy in the NRG-HN002 trial. Sensitivity analysis Cox regression analyses, before and after IPTW adjustment, showed similar OS between treatment groups, while radiotherapy alone was associated with inferior PFS (Table 4). Excluding patients with incomplete covariate data or altered fractionation schedules did not materially change the results, confirming the robustness of the findings (Table 4).Table 4Sensitivity analyses of concurrent chemoradiotherapy versus radiotherapy alone before and after IPTW.CategoryMultivariableHR (95% CI)P valueOverall survivalAll patients (n = 240)before IPTW0.42 (0.17–1.00)0.05after IPTW0.50 (0.20–1.24)0.14Excluding patients with missing covariates (n = 226)before IPTW0.40 (0.16–1.03)0.06after IPTW0.51 (0.18–1.46)0.21Excluding patients altered fractionation schedules (n = 231)before IPTW0.44 (0.18–1.12)0.08after IPTW0.48 (0.17–1.33)0.16Progression-free survivalAll patients (n = 240)before IPTW0.40 (0.19–0.81)0.01after IPTW0.42 (0.19–0.93)0.03Excluding patients with missing covariates (n = 226)before IPTW0.34 (0.16–0.73)0.005after IPTW0.33 (0.14–0.78)0.01Excluding patients altered fractionation schedules (n = 231)before IPTW0.39 (0.18–0.83)0.01after IPTW0.41 (0.18–0.95)0.04 Table 4Sensitivity analyses of concurrent chemoradiotherapy versus radiotherapy alone before and after IPTW.CategoryMultivariableHR (95% CI)P valueOverall survivalAll patients (n = 240)before IPTW0.42 (0.17–1.00)0.05after IPTW0.50 (0.20–1.24)0.14Excluding patients with missing covariates (n = 226)before IPTW0.40 (0.16–1.03)0.06after IPTW0.51 (0.18–1.46)0.21Excluding patients altered fractionation schedules (n = 231)before IPTW0.44 (0.18–1.12)0.08after IPTW0.48 (0.17–1.33)0.16Progression-free survivalAll patients (n = 240)before IPTW0.40 (0.19–0.81)0.01after IPTW0.42 (0.19–0.93)0.03Excluding patients with missing covariates (n = 226)before IPTW0.34 (0.16–0.73)0.005after IPTW0.33 (0.14–0.78)0.01Excluding patients altered fractionation schedules (n = 231)before IPTW0.39 (0.18–0.83)0.01after IPTW0.41 (0.18–0.95)0.04 Sensitivity analyses of concurrent chemoradiotherapy versus radiotherapy alone before and after IPTW. Discussion In this multicenter study, we observed that adding concurrent chemotherapy to radiotherapy did not confer an OS benefit in patients with T1–2N1–2a HNSCC, but it was associated with a significant improvement in PFS. Subgroup analysis suggested that the PFS benefit was primarily seen in T2N1–2a patients, whereas no clear benefit was observed in T1N1–2a patients. Data from the NRG-HN002 trial provide further evidence supporting these observations. Previous studies have largely focused on nodal stage, with higher nodal burden considered indicative of increased risk and a potential need for additional chemotherapy, often overlooking the prognostic value of T stage [10], [17], [18]. Current NCCN guidelines similarly base recommendations on nodal size and number when deciding between radiotherapy alone or CCRT. Lorenz et al. [17] reported that, among HPV-positive stage I oropharyngeal cancer, T2 tumors were associated with significantly higher mortality compared with T1 tumors (HR, 1.8; 95% CI, 1.4–2.2; P < 0.001). Consistent with this, our study found that T2 disease significantly increased the risk of tumor progression compared with T1 (HR, 2.98; 95% CI, 1.13–7.88; P = 0.03). Optimal treatment for T1–2N1–2a patients remains controversial (eTable 3). Our findings align with those of Huang and Adrian [11], [19], who reported no OS benefit with the addition of concurrent chemotherapy. Zumsteg et al. [8] noted that T2N1 patients had higher mortality than T1N1 patients, suggesting that the former may derive more benefit from chemotherapy. Data from the NRG-HN002 trial further support this observation, showing that radiotherapy alone was inferior to CCRT in PFS, primarily due to higher locoregional recurrence rates [20]. Specifically, among T2N1–2a patients, locoregional recurrence rates were 13.3% (4/30) with radiotherapy alone versus 3.6% (1/28) with CCRT; in T1N1–2a patients, the rates were 11.1% (2/18) versus 11.8% (2/17), respectively. These findings are consistent with our analysis, indicating that T2N1–2a patients may benefit from CCRT in terms of PFS, while OS remains unaffected. Chemotherapy heterogeneity should also be acknowledged. Previous studies have shown that weekly cisplatin regimens administered at doses of 20 mg/m2 or 30 mg/m2 were associated with inferior clinical outcomes despite potential reductions in treatment-related toxicity [21], [22]. In contrast, a meta-analysis of six randomized trials found no significant differences in overall survival or progression-free survival between weekly and three-weekly cisplatin schedules [23]. Furthermore, the phase II/III JCOG1008 trial demonstrated that weekly cisplatin at 40 mg/m2 was non-inferior to the standard three-weekly high-dose regimen in patients with locally advanced head and neck squamous cell carcinoma, while providing a more favorable toxicity profile [24]. Therefore, although variations in cisplatin scheduling existed across participating centers, particularly between weekly 40 mg/m2 and three-weekly high-dose regimens, current evidence suggests that these differences are unlikely to have substantially affected the oncologic outcomes observed in our study. It is important to note that concurrent chemotherapy carries significant toxicity. Previous studies have documented increased rates of ototoxicity, nausea, nephrotoxicity, and hematologic toxicity, with leukopenia incidence approximately 41% higher than with radiotherapy alone [6], [7], [25], [26], [27]. Therefore, although T2N1–2a patients may gain PFS benefit, clinicians must carefully balance treatment efficacy against potential toxicities. Our study draws on multicenter clinical data, which may improve the representativeness of the study population. Nevertheless, differences in systemic therapy protocols across countries or regions should be acknowledged, although all conformed to local clinical guidelines. Additionally, a subset of patients in the radiotherapy-alone group received altered fractionation schedules, which may have improved locoregional control, approximating the efficacy of CCRT. Importantly, sensitivity analyses excluding these patients confirmed the robustness of our findings. Furthermore, given the retrospective nature of this study, no a priori sample size or power calculation was performed. The relatively limited number of progression events may have affected the precision of the estimated treatment effects. In addition, toxicity data were not systematically collected, which precludes comprehensive assessment of treatment-related adverse events. Finally, potential residual confounding cannot be excluded, underscoring the need for prospective randomized trials to validate these results. Conclusion In summary, T1N1–2a patients are unlikely to derive PFS or OS benefit from the addition of concurrent chemotherapy, whereas T2N1–2a patients may experience improved PFS but not OS. These findings highlight the importance of carefully weighing therapeutic benefit against treatment-related toxicity when selecting management strategies for patients with T1–2N1–2a HNSCC. Ethical statement This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the Affiliated Zhongshan Hospital of Dalian University, with a waiver of informed consent (KY2025–107-1). Funding statement This study was supported by 10.13039/501100001809National Natural Science Foundation of China (82172822), Revitalizing Liaoning Talents Program-Medical Experts Project (YXMJ-JC-10), Liaoning Provincial International Science and Technology Cooperation Project (2024JH2/101900006), and Liaoning Provincial Science and Technology Plan-Joint Program-Key Technology Research Project (2024JH2/102600067).