Impact of Consolidation-to-Tumor Ratio on Prognosis After Radiation therapy for Lung Cancer With a Lesion Diameter of 3 cm or Less Authors: Masaki Nakamura, Hidenari Hirata, Yuichi Hiroshima, Kenji Makita, Satoshi Saito, Hiromi Baba, Takashi Someya, Toru Mukohara, Sadamoto Zenda Published: 2026 ABSTRACT AbstractPurposeAccurate risk stratification is essential for optimizing radiation therapy for early stage non-small cell lung cancer (NSCLC). This study evaluated whether the consolidation-to-tumor ratio (CTR) or conventional clinical T stage provides superior prognostic value for patients treated with definitive radiation therapy.Methods and MaterialsThis study included patients with NSCLC with a lesion diameter ≤3 cm, without lymph node or distant metastases, who underwent definitive radiation therapy alone between 2015 and 2022. Patients were categorized into 3 CTR groups: ground-glass opacity (GGO)–dominant (≤0.5), solid-dominant (>0.5 to <1.0), and pure-solid (1.0). Prognostic performance for progression-free survival was compared between CTR-based and T-stage–based models using the Akaike Information Criterion and Concordance Index.ResultsIn total, 265 patients (25 GGO-dominant, 39 solid-dominant, and 201 pure-solid) were analyzed. The median follow-up was 50.2 months. No recurrences or deaths were observed in the GGO-dominant group. The pure-solid group exhibited significantly poorer 3-year outcomes than the containing-GGO group (GGO-dominant + solid-dominant) for overall survival (79.0% vs 96.5%) and progression-free survival (64.5% vs 93.1%), with higher rates of distant metastasis. In multivariate analysis, the CTR-based model demonstrated a lower Akaike Information Criterion (895.6 vs 907.4) and a higher Concordance Index (0.69 vs 0.67) than the T-stage–based model, indicating better goodness-of-fit and predictive ability.ConclusionsCTR is a robust prognostic indicator, superior to T-stage, in early stage NSCLC treated with radiation therapy. Incorporating CTR into clinical decision-making may provide more accurate risk stratification than tumor size alone. Introduction Early stage non-small cell lung cancer (NSCLC) is commonly treated with surgical resection, which remains the standard of care for operable patients.1 However, many patients are medically inoperable because of comorbidities or advanced age. For these patients, definitive radiation therapy, including stereotactic body radiation therapy (SBRT) and proton beam therapy (PBT), is used as a less-invasive alternative. Among nonsurgical modalities, radiation therapy has emerged as an effective and widely adopted treatment option, offering excellent local control with minimal toxicity.2, 3, 4 Accurate staging is essential for selecting an appropriate treatment strategy. In the 8th edition of the International Association for the Study of Lung Cancer TNM classification,5 it was recommended that the T descriptor for lung cancers with ground-glass opacity (GGO) be determined using the invasive (solid) component rather than total lesion diameter; this approach was retained in the 9th edition.6 For example, a 20-mm lesion with a 9-mm solid component is classified as clinical T1a, whereas a pure GGO lesion is classified as Tis (Fig. 1). For small, node-negative, early stage lung cancers, Japanese thoracic surgery research groups have successfully developed treatment approaches by stratifying patients using 2 key radiological parameters: tumor size and consolidation-to-tumor ratio (CTR).7,8 In recent years, CTR has also been reported as an important prognostic factor in radiation oncology.9, 10, 11 Representative examples of clinical T staging and consolidation-to-tumor ratio measurement. Abbreviation: CTR = consolidation-to-tumor ratio. However, it remains unclear whether invasive size defined by TNM staging or CTR provides superior prognostic value for patients undergoing radiation therapy for early stage lung cancer. Given that radiological solid components may reflect distinct biological characteristics such as tumor aggressiveness, growth patterns, and radiosensitivity, clarifying the relative utility of these parameters is essential for optimizing treatment selection and advancing radiation therapy. Therefore, this study aimed to evaluate the prognostic impact of CTR and T stage in patients with early stage lung cancer treated with radiation therapy, and to identify which parameters more accurately predict clinical outcomes. These findings may help guide individualized treatment strategies and support the development of future therapeutic approaches in radiation oncology. Methods and Materials Patients This study included NSCLC patients with a lesion diameter ≤3 cm, without clinical lymph node or distant metastases, who underwent definitive radiation therapy (SBRT or PBT) at our institute between February 2015 and March 2022. Clinical staging was determined based on imaging studies such as chest and abdominal computed tomography (CT), 18F-fluorodeoxyglucose positron emission tomography–computed tomography (PET-CT), and brain CT/magnetic resonance imaging (MRI). For patients without pathologic confirmation owing to tumor location or prohibitive biopsy risk, the diagnosis of lung cancer was made by a multidisciplinary tumor board based on imaging studies and clinical histories. When lymph node metastasis could not be ruled out on CT and PET-CT, endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) was performed. Patients enrolled in interventional clinical trials were excluded. All study participants provided written informed consent for the comprehensive study protocol before initial treatment. The protocol was approved by the Research Ethics Committee of our institution (reference number: 2017-440) and conducted in accordance with the Declaration of Helsinki (1964) and its subsequent amendments. Radiation therapy For SBRT, the primary tumor was delineated as the gross tumor volume using a CT lung window. The clinical target volume was defined as the gross tumor volume with a 0-mm margin in all directions. The internal target volume was determined using 4-dimensional CT. The planning target volume (PTV) was defined as the internal target volume with a 5-mm setup margin. Respiratory-synchronized irradiation was performed if motion exceeded 5 mm. The dose was prescribed to cover 95% of the PTV. Prescription doses ranged from 42 to 75 Gy in 4 to 25 fractions, while maintaining organ-at-risk tolerance limits. Typically, 42 or 48 Gy in 4 fractions were prescribed for peripheral tumors, 50 or 60 Gy in 8 fractions for central tumors, and 62.5 or 75 Gy in 25 fractions for ultracentral tumors. The general concepts of patient preparation, target delineation, and treatment planning in lung PBT have been previously reported,12 and the same policies and dose schedules were applied to all patients. Follow-up and outcome assessment After treatment, chest-abdominal CT was performed every 3 months during the first year and every 6 months thereafter. When lesions suggestive of recurrence were detected, additional PET-CT and, when feasible, biopsy were performed. Brain MRI was performed when clinical signs of brain involvement were present. Using pretreatment thin-slice CT (slice thickness ≤2 mm), the total lesion diameter and solid component diameter were measured in the lung window setting (W1600/L-600). The CTR was defined as the solid component diameter divided by the total lesion diameter. CTR values were retrospectively confirmed by a single board-certified radiation oncologist. To ensure accuracy, these values were cross-referenced with the official consensus reports of a multidisciplinary team conference attended by thoracic surgeons, medical oncologists, radiation oncologists, and diagnostic radiologists. Overall survival (OS) was defined as the time from the start of radiation therapy to death from any cause, censored at the date of last follow-up for surviving patients. Progression-free survival (PFS) was defined as the time from the start of radiation therapy to the earliest detection of disease progression or death from any cause. Local failure (LF) was defined as tumor enlargement or the appearance of a new lesion within the PTV, confirmed by biopsy or PET-CT. Other new recurrences, including distant metastases, were classified as regional/distant failure (RDF). Adverse events (AEs) were graded using the National Cancer Institute-Common Terminology Criteria for Adverse Events, version 5.0, and AEs of grade ≥2 were evaluated. Statistical analysis All statistical analyses were performed using R software, version 4.3.1 (The R Foundation, Vienna, Austria). Cumulative incidence curves of recurrence were calculated and compared using Gray’s test, with death as a competing risk. Survival curves were constructed using the Kaplan-Meier method and compared using the log-rank test and Cox proportional hazards model. Referencing prior surgical series, patients were categorized into 2 or 3 groups based on the CTR for comparison.8,13 Multivariate analysis was performed using the forced-entry method, and prognostic factors were selected based on previous literature.3,14 Because of multicollinearity between CTR and T stage (solid component diameter), these factors were not included in the same multivariate model. Instead, separate models were created. Model performance was compared using the Akaike Information Criterion (AIC) and Concordance Index (C-index). A lower AIC and higher C-index indicated better goodness-of-fit and predictive ability. All tests were 2-sided, and P values<.05 were considered statistically significant. Results A total of 265 patients were included. Patients were categorized into 3 groups based on CTR: GGO-dominant (CTR≤0.5, n=25), solid-dominant (CTR>0.5, and <1.0, n=39), and pure-solid (CTR=1.0, n=201). Patient characteristics are summarized in Table 1. Baseline characteristics differed significantly among groups. The pure-solid group included more males, smokers, and patients with advanced T-stage disease than the GGO-dominant group. Although most patients in the GGO-dominant group were clinically diagnosed (84.0%), the pure-solid group included patients with squamous cell carcinoma. The pure-solid group also showed poorer clinical status, with lower operability (44.3%) and an interstitial lung disease present only in this group (7.5%). Regarding tumor location, pure-solid tumors were more frequently located in the lower lobes (46.3%). The proportion of patients who required additional endobronchial ultrasound-guided transbronchial needle aspiration for lymph node staging was higher in the pure-solid group (10.4%), while staging by CT alone was more prevalent in the GGO-dominant group (20%). Table 1Baseline patient characteristicsTable 1 dummy alt textFactorGGO-dominant(n = 25)Solid-dominant(n = 39)Pure-solid(n = 201)P valueSMDAge, (y)Range73 [40-85]78 [54-89]78 [38-93].130.23Sex, (%)Female13 (52.0)19 (48.7)59 (29.4).010.32ECOG PS, (%)019 (76.0)20 (51.3)84 (41.8).070.5416 (24.0)17 (43.6)102 (50.7)20 (0.0)2 (5.1)13 (6.5)30 (0.0)0 (0.0)2 (1.0)Smoking, (%)Never14 (56.0)22 (56.4)34 (16.9)<.010.6CCI, (%)≥33 (12.0)6 (15.4)53 (26.4).150.25Pathology, (%)Ad4 (16.0)12 (30.8)56 (27.9)<.010.68Sq0 (0.0)0 (0.0)30 (14.9)NSCLC0 (0.0)2 (5.1)14 (7.0)Unknown21 (84.0)25 (64.1)101 (50.2)Lobe, (%)Upper20 (80.0)21 (53.8)97 (48.3).020.52Middle1 (4.0)4 (10.3)11 (5.5)Lower4 (16.0)14 (35.9)93 (46.3)Laterality, (%)Left6 (24.0)17 (43.6)92 (45.8).120.31Right19 (76.0)22 (56.4)109 (54.2)Location, (%)Peripheral22 (88.0)31 (79.5)156 (77.6).600.24Central3 (12.0)7 (17.9)43 (21.4)Ultracentral0 (0.0)1 (2.6)2 (1.0)FEV1.0, (mL)Range2040 [1400-2960]1940 [900-3450]1735 [570-3400].020.35VC, (mL)Range2950 [2210-4260]2480 [1810-4650]2520 [970-4680].020.52T stageis2 (8.0)0 (0.0)0 (0.0)<.012.41a22 (88.0)6 (15.4)14 (7.0)1b1 (4.0)22 (56.4)93 (46.3)1c0 (0.0)11 (28.2)94 (46.8)Lymph node staging, (%)CT5 (20.0)2 (5.1)7 (3.5)<.010.52PET-CT20 (80.0)36 (92.3)173 (86.1)EBUS-TBNA0 (0.0)1 (2.6)21 (10.4)Operable, (%)Yes19 (76.0)30 (76.9)89 (44.3)<.010.47Prior LC, (%)Yes8 (32.0)9 (23.1)67 (33.3).490.15ILD, (%)Present0 (0.0)0 (0.0)15 (7.5).110.27Modality, (%)PBT21 (84.0)26 (66.7)96 (47.8)<.010.54BED10, (%)<100 Gy3 (12.0)4 (10.3)47 (23.4).120.24Definition of central location: a tumor within 2 cm in all directions of any mediastinal critical structure.Abbreviations: Ad = adenocarcinoma; BED10 = biological effective dose with an alpha/beta ratio of 10 Gy; CCI = Charlson comorbidity index; CT = computed tomography; EBUS-TBNA = endobronchial ultrasound-guided transbronchial needle aspiration; ECOG PS = Eastern Cooperative Oncology Group performance status; FEV1.0 = forced expiratory volume in one second; GGO = ground-glass opacity; ILD = interstitial lung disease; LC = lung cancer; NSCLC = non-small cell lung cancer; PBT = proton beam therapy; PET-CT = 18F-fluorodeoxyglucose positron emission tomography–computed tomography; SMD = standardized mean difference; Sq = squamous cell carcinoma; VC = vital capacity.P < .05 is indicated in bold. Baseline patient characteristics Definition of central location: a tumor within 2 cm in all directions of any mediastinal critical structure. Abbreviations: Ad = adenocarcinoma; BED10 = biological effective dose with an alpha/beta ratio of 10 Gy; CCI = Charlson comorbidity index; CT = computed tomography; EBUS-TBNA = endobronchial ultrasound-guided transbronchial needle aspiration; ECOG PS = Eastern Cooperative Oncology Group performance status; FEV1.0 = forced expiratory volume in one second; GGO = ground-glass opacity; ILD = interstitial lung disease; LC = lung cancer; NSCLC = non-small cell lung cancer; PBT = proton beam therapy; PET-CT = 18F-fluorodeoxyglucose positron emission tomography–computed tomography; SMD = standardized mean difference; Sq = squamous cell carcinoma; VC = vital capacity. P < .05 is indicated in bold. The median follow-up duration for all patients was 50.2 months (95% CI, 45.5-55.5), calculated using the reverse Kaplan-Meier method. At last follow-up, 62 deaths had occurred: 40 because of lung cancer and 22 from other causes. There were 73 recurrences (12 local only, 1 local + regional, 2 local + distant, 1 local + regional + distant, 16 regional only, 7 regional + distant, and 34 distant only). The lungs were the most common metastatic site (n = 21), followed by bone (n = 13) and pleura (n = 6). Among the 21 lung metastases, 10 had synchronous extra-parenchymal metastases and 9 had multiple lung nodules. The remaining 2 resected solitary nodules showed identical histology to the primary tumor and were diagnosed as metastases based on imaging, without paired genetic analysis. Kaplan-Meier curves for OS and PFS stratified by CTR are shown in Fig. 2A, B. The GGO-dominant group achieved 3-year OS and PFS rates of 100%, with no recurrence or death (Fig. E1). Therefore, this group was combined with the solid-dominant group, and outcomes were compared between tumors containing GGO and pure-solid tumors. The containing-GGO group showed significantly better outcomes. The 3-year OS rates were 96.5% (95% CI, 86.7-99.1) vs 79.0% (95% CI, 71.7-84.5) (P<.01), and the 3-year PFS rates were 93.1% (95% CI, 82.6-97.3) vs 64.5% (95% CI, 56.6-71.2) (P<.01). Regarding failure patterns, the 3-year cumulative incidence of LF and RDF were 0.0% vs 6.7% (95% CI, 3.5-11.2) (P = .06) and 6.9% (95% CI, 2.2-15.5) vs 24.9% (95% CI, 18.7-31.7) (P<.01), respectively, for the containing-GGO and pure-solid groups (Fig. 3A, B). For T stage, patients were divided into 2 groups, Tis-T1b and T1c, based on a previous report.15 Patients with Tis-T1b tumors showed superior outcomes compared with those with T1c tumors. The 3-year OS and PFS were 86.7% (95% CI, 79.6-91.4) vs 78.5% (95% CI, 67.9-85.9) (P<.01) and 77.5% (95% CI, 69.4-83.7) vs 62.8% (95% CI, 51.9-71.9) (P<.01), respectively (Fig. 2C and D). The 3-year cumulative incidence rates of LF and RDF were 5.2% (95% CI, 2.3-9.9) vs 4.6% (95% CI, 1.5-10.5) (P=.24) and 16.2% (95% CI, 10.5-23.0) vs 26.9% (95% CI, 18.2-36.3) (P=.02), respectively, for the Tis-T1b and T1c groups (Fig. 3C, D). Although T stage provided significant stratification, the separation of the curves appeared less distinct than CTR-based stratification, particularly in the favorable-prognosis group. The distribution of CTR by consolidation size, stratified by event status, is shown in Fig. 4. To compare the PFS predictive value of CTR and T stage, we constructed 2 separate multivariable Cox proportional hazards models using the forced-entry method (Table 2). The CTR model (containing-GGO vs pure-solid), demonstrated a lower AIC (895.6 vs 907.4) and a higher C-index (0.69 vs 0.67) than the T-stage model, indicating better goodness-of-fit and discriminatory ability. Kaplan-Meier estimates of overall survival and progression-free survival (PFS) stratified by CTR and clinical T stage. (A) Overall survival by CTR. (B) PFS by CTR. (C) Overall survival by clinical T stage. (D) PFS by clinical T stage. Abbreviations: CTR = consolidation-to-tumor ratio; GGO = ground-glass opacity; OS = overall survival. Cumulative incidence curves of local failure (LF) and regional/distant failure (RDF) stratified by CTR and clinical T stage. (A) LF stratified by CTR. (B) RDF stratified by CTR. (C) LF stratified by clinical T stage. (D) RDF stratified by clinical T stage. Abbreviations: CTR = consolidation-to-tumor ratio; GGO = ground-glass opacity. Scatter plot of CTR versus consolidation size stratified by event status. Patients are color-coded based on their PFS event status (event vs no event). Abbreviations: CTR = consolidation-to-tumor ratio; PFS = progression-free survival. Table 2Comparison of multivariable models in progression-free survivalTable 2 dummy alt textCTR-based modelT-based modelVariablesHR (95% CI)P valueHR (95% CI)P valueCTR0 to < 1.01.0 (Ref)---13.9 (1.8-8.3)<.01--Tis-1b--1.0 (Ref)-1c--1.7 (1.1-2.6).03Age (y)<781.0 (Ref)1.0 (Ref)-78≤1.0 (0.6-1.5).930.9 (0.6-1.4).63SexFemale1.0 (Ref)-1.0 (Ref)-Male1.0 (0.6-1.6).891.2 (0.8-2.0).35ECOG PS0-11.0 (Ref)-1.0 (Ref)-2≤1.5 (0.7-3.5).311.6 (0.7-3.5).29CCI0-21.0 (Ref)-1.0 (Ref)-3≤1.4 (0.8-2.2).231.4 (0.8-2.3).22PathologyYes1.0 (Ref)-1.0 (Ref)-Unknown0.7 (0.5-1.1).10.7 (0.4-1.1).08OperabilityNo1.0 (Ref)-1.0 (Ref)-Yes1.1 (0.7-1.8).591.0 (0.6-1.7).89Prior LCNo1.0 (Ref)-1.0 (Ref)-Yes1.7 (1.1-2.6).021.7 (1.1-2.7).02ILDNo1.0 (Ref)-1.0 (Ref)-Yes2.5 (1.2-5.1).012.8 (1.4-5.9)<.01BED10<100 Gy1.0 (Ref)-1.0 (Ref)-100 Gy≤0.8 (0.5-1.4).510.7 (0.4-1.3).27AIC895.6907.4C-index0.690.67Abbreviations: AIC = Akaike Information Criterion; BED10 = biological effective dose with an alpha/beta ratio of 10 Gy; CCI = Charlson comorbidity index; C-index = concordance index; CTR = consolidation-to-tumor ratio; ECOG PS = Eastern Cooperative Oncology Group performance status; HR = hazard ratio; ILD = interstitial lung disease; LC = lung cancer.P < .05 is indicated in bold. Comparison of multivariable models in progression-free survival Abbreviations: AIC = Akaike Information Criterion; BED10 = biological effective dose with an alpha/beta ratio of 10 Gy; CCI = Charlson comorbidity index; C-index = concordance index; CTR = consolidation-to-tumor ratio; ECOG PS = Eastern Cooperative Oncology Group performance status; HR = hazard ratio; ILD = interstitial lung disease; LC = lung cancer. P < .05 is indicated in bold. Regarding AEs, no grade 4 or higher AEs occurred. grade ≥2 respiratory-related AEs were observed in 23 patients (8.7%) and grade ≥2 chest wall-related events in 15 patients (5.7%) (Table E1). In a sensitivity analysis restricted to pathologically confirmed patients (n = 118), a significant difference in PFS was observed between the containing-GGO and pure-solid groups (P = .01), whereas no significant difference was found between the Tis-T1b and T1c groups (P = .72) (Fig. E2). Discussion The primary finding of this study is that CTR serves as a superior prognostic indicator compared with conventional clinical T stage in patients with early stage lung cancer treated with radiation therapy alone. Patients with GGO-dominant tumors had excellent outcomes, with no recurrence or death. Furthermore, multivariate analysis revealed that the CTR-based model provided better goodness-of-fit and discriminatory ability for predicting PFS than the T-stage–based model. These results suggest that evaluating tumor density and the proportion of the solid components is more important than tumor size alone for prognostic stratification in this population. Our multivariate analysis demonstrated that the CTR-based model yielded a lower AIC and higher C-index than the T–stage–based model. This superiority likely stems from the fundamental difference between these 2 metrics. Whereas T stage primarily captures the anatomic dimension of tumor burden, CTR serves as a qualitative surrogate for tumor biology and host characteristics. Specifically, a lower CTR (presence of GGOs) is strongly associated with favorable clinicopathological features, such as adenocarcinoma histology, female sex (biological sex at birth, as recorded in the medical records), and nonsmoking status, reflecting an indolent lepidic growth pattern.16 In contrast, a “pure-solid” status (CTR = 1.0) often correlates with squamous cell carcinoma, heavy smoking history, interstitial lung disease, and a predilection for the lower lobes, as observed in our cohort and previous studies.16,17 Although the 9th edition of the TNM classification improved staging by measuring the solid component size, it remains a dimensional metric that may not fully capture the biological aggressiveness inherent to the “solid” phenotype. Our results suggest that CTR offers more comprehensive risk stratification by integrating the degree of pathologic invasiveness with these underlying biological features, which a simple size–based measurement alone (T stage) cannot fully encompass. A notable observation in this cohort is the complete absence of recurrence or death in the GGO-dominant group (CTR≤0.5) during follow-up. Onishi et al18 similarly reported no local or regional lymph node recurrence after SBRT in 84 tumors with CTR<0.5. A secondary analysis of the JCOG0802 trial comparing segmentectomy and lobectomy showed higher local recurrence after segmentectomy in solid tumors, whereas no difference was observed for part-solid tumors.13 In addition, Zhang et al19 demonstrated that adenocarcinomas with CTR≤0.5 and maximum lesion diameter ≤3 cm have an extremely low likelihood of lymph node metastasis. These findings support the use of radiation therapy targeting only the primary tumor in this population. Moreover, given the highly favorable prognosis of GGO-dominant tumors, active surveillance may be an appropriate strategy in selected patients, a concept currently being evaluated in ongoing confirmatory clinical trial.20 Conversely, the pure-solid group exhibited significantly poorer outcomes, with higher rates of distant metastasis, particularly to the lungs and bones. This subgroup suggests a clearly defined high-risk population that may benefit from intensified therapy. Ongoing confirmatory trials of dose escalation21 and immunotherapy add-on strategies22 are expected to improve outcomes for such high-risk patients. No grade 4 or 5 AEs were observed. The incidences of grade ≥2 respiratory-related toxicity (8.7%) and chest wall-related toxicity (5.7%) were comparable to or lower than those reported in previous prospective trials.23,24 To improve survival while maintaining the minimally invasive profile of radiation therapy, identifying appropriate candidates for intensified therapy is crucial. Our findings suggest that CTR may serve as a useful imaging biomarker for this purpose. Furthermore, because CTR is calculated using a 1-dimensional diameter ratio, it may not accurately reflect the true proportion of solid to ground-glass components, particularly in irregularly shaped tumors; area- or volume-based approaches might offer greater accuracy. However, while volumetric measurements require dedicated software, diameter-based measurements are straightforward and readily feasible in routine clinical practice,25 making CTR an easily implementable imaging biomarker. This study has several limitations. First, it was a single institution retrospective analysis, introducing potential selection bias. Second, the GGO-dominant group was relatively small (n = 25), which limited statistical power for detailed subgroup comparisons, although the signal of “zero events” remains clinically relevant. Third, pathologic confirmation was lacking in most patients in the GGO-dominant group (84.0% clinically diagnosed), which is a common challenge in SBRT cohorts of medically inoperable patients. However, the Early Lung Cancer Action Project and the Pan-Canadian Early Detection of Lung Cancer Study demonstrated that part-solid nodules were more likely to be malignant than solid nodules,26,27 and strict diagnostic criteria based on serial CT imaging likely minimized the inclusion of benign nodules. Fourth, the heterogeneity in treatment modalities, particularly the proportion of PBT, and dose-fractionation schedules across the groups is a limitation of this study. However, in routine practice, dose-fractionation schedules are primarily dictated by tumor centrality to maintain organ-at-risk dose constraints, rather than by the CTR status. Crucially, our baseline data demonstrated no significant differences in tumor centrality among the groups. Furthermore, a previous meta-analysis comparing PBT with SBRT reported no significant differences in clinical outcomes despite dosimetric variations.28 Therefore, we believe that this treatment heterogeneity had a minimal impact on the survival differences observed in our study. Finally, the follow-up period was relatively short for assessing late recurrence in indolent GGO tumors, warranting long-term observations. In conclusion, CTR is a robust prognostic factor in patients with early stage lung cancer treated with SBRT. The presence of any GGO component was associated with significantly better survival outcomes than pure-solid tumors. The CTR-based model outperformed the T–stage–based model in predicting PFS, suggesting that incorporating CTR into clinical decision-making provides more accurate risk stratification than T stage alone. Future prospective studies are needed to validate these findings and explore personalized treatment strategies based on CTR. Disclosures MN received a research grant from Illumina, Inc, and personal fees from AstraZeneca, Illumina, and Sumitomo Heavy Industries. TM received research grants from MSD K.K, AstraZeneca K.K., Eisai Co, Ltd, Sanofi K.K., Novartis Pharma K.K, Bayer Yakuhin, Ltd, ONO Pharmaceutical Co, Ltd, Daiichi Sankyo Co, Ltd, Chugai Pharmaceutical Co, Ltd, Pfizer Japan Inc, Gilead Sciences, Inc, Merck Biopharma Co, Ltd, Seagen Inc, Astellas Pharma Inc, Sysmex Corporation, and Exact Sciences K.K.; received personal fees from TERUMO CORPORATION., Bayer Yakuhin, Ltd, MICIN, Inc, Kyowa Kirin Co, Ltd, Daiichi Sankyo Co, Ltd, Eli Lilly Japan K.K., MI-Force, Inc, AstraZeneca K.K., Eisai Co, Ltd, TAIHO Pharmaceutical Co, Ltd, Chugai Pharmaceutical Co, Ltd, Shionogi & Co, Ltd, Pfizer Japan Inc, and ONO Pharmaceutical Co, Ltd; and serves on the advisory board for Pfizer and AstraZeneca. SZ received personal fees from Eisai Co Ltd, Maruho Co Ltd, Maruishi Pharmaceutical Co Ltd, CMIC HOLDINGS Co Ltd, Shin Nippon Yakugyo Co Ltd, DAIICHI SANKYO COMPANY LIMITED, AYUMI Pharmaceutical Corporation, TAIHO PHARMACEUTICAL Co Ltd, Merck Biopharma Co Ltd, Bristol-Myers Squibb K.K., ONO PHARMACEUTICAL Co Ltd, SANWA KAGAKU KENKYUSHO Co Ltd, Zeria Pharmaceutical Co Ltd, Bayer Yakuhin Ltd, MSD K.K., NUTRI Co Ltd, Rakuten Medical, Inc, and Stratpharma AGAll remaining authors declare no conflicts of interest in the writing of this article. Acknowledgments Acknowledgments The authors would like to thank Editage (http://www.editage.jp) for their assistance in the English language editing process. Declaration of AI and AI-Assisted Technologies in the Writing Process During the preparation for this study, the authors used Gemini (https://gemini.google.com) to assist in language refinement during the drafting process. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Footnotes Sources of support: This work was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI [Grant Numbers 24K10926] and the Japan Agency for Medical Research and Development (AMED) [Grant Number JP25ck0106956, JP26ck0106924]. Research data are stored in an institutional repository and will be shared upon request to the corresponding author. Supplementary material associated with this article can be found in the online version at doi:10.1016/j.adro.2026.102122. Supplementary materials Supple References Associated Data Supplementary Materials Supple