Frailty profiles and symptomatic radiation pneumonitis in patients with lung cancer undergoing radiotherapy: A latent class analysis.

Introduction
Lung cancer is one of the most prevalent malignancies worldwide, according to the latest data from the International Agency for Research on Cancer (IARC), and 2.5 million new cases and 1.8 million deaths were reported worldwide in 2022.1 In China, lung cancer remains the most common cancer, with 1.0606 million new cases and 73.33 million deaths in 2022 alone, posing a serious threat to human health.2
Radiotherapy (RT) is an important treatment for patients with lung cancer3 and enhances patient survival rates and reduces local tumor recurrence.4 Radiation pneumonitis (RP) is one of the common and severe complications in patients with lung cancer receiving radiotherapy, and grade ≥ 2 RP is defined as symptomatic radiation pneumonitis (SRP), with earlier studies estimating its incidence at 15–41.4%.5 Evidence from longitudinal cohort study developing a predictive model for SRP indicated that the acute phase chiefly manifests 2–6 months after radiotherapy that can lead to treatment interruption, significantly impact patients' quality of life and prognosis, and even pose a threat to their lives.6,7
Frailty is a syndrome characterized by reduced physiological reserves and dysfunction across multiple systems, leading to increased vulnerability and diminished stress resistance. It reflects the overall decline in physiological reserve and multi-system function experienced by patients during treatment.8 Patients with lung cancer have increased susceptibility to frailty due to the disease itself, psychological disorders, etc.9 In addition, exposure to radiotherapy further increases their frailty. Our team's previous study found that the incidence of frailty in patients with lung cancer radiotherapy was as high as 55.19%.10 Many studies have indicated that frailty can lead to adverse outcomes such as increased side effects of radiotherapy and chemotherapy and increased mortality.11,12 More importantly, frailty is closely linked to both the susceptibility and severity of pulmonary infection, with manifestations such as diminished physiological reserve and multi-system dysfunction serving as risk factors for SRP.7,13,14 Thus, effective assessment and intervention targeting frailty in lung-cancer patients is critical for reducing the incidence of SRP.
In recent years, with the deepening of research, frailty has been evaluated by various tools such as Clinical Frailty Scale and Fried Frailty Phenotype, etc.9 Many researches have evaluated frailty not only by measuring the frailty score to evaluate the overall level, but also by exploring the subtypes of frailty, risk factors and their association with health outcomes, making it targeted to guide clinical practice. In a cross-sectional survey of 5341 adults aged 60 years and older conducted by Zhang et al.,15 four frailty subtypes were identified via the frailty index: multi-domain, cognitive-functional, psychological, and physical frailty. Old age and low education level are common risk factors across four subtypes. Lin et al.16 analyzed the 11-year follow-up data of 5334 elderly people and identified three subtypes of frailty: energy-based frailty, muscle-based frailty, and mixed-based frailty. They also found that the energy-based frailty subgroup had a lower risk of falls than the muscle-based frailty subgroup, the muscle-based frailty subgroup was less likely to have depression than the energy-based frailty subgroup, and the mixed-based frailty subgroup was more likely to be hospitalized. In addition, previous studies have shown that frailty subtypes have different associations with hearing loss, tinnitus17 and long-term outcomes such as hypertension and diabetes.18 These results indicate that frailty has population heterogeneity, and this heterogeneity may exist in patients with lung cancer radiotherapy, and different types of frailty may be related to different risks of SRP.
However, current researches on the heterogeneity of frailty have focused mostly on older adults in relation to health outcomes or treatment-related complications. Few studies have explored the frailty subgroup of patients with lung cancer radiotherapy, and there is a significant lack of research on the relationship between the frailty subgroup and SRP. Therefore, this study employed latent class analysis (LCA) to address this critical knowledge gap, which is different from the traditional variable-centered statistical analysis method.19 It emphasizes human-centeredness. By establishing a latent class model, it can objectively and accurately identify patient subgroups with heterogeneous characteristics to further improve the effectiveness of symptom management.20 This study aims to explore the inherent characteristics of frailty in lung cancer radiotherapy patients via LCA, analyze the factors influencing frailty classes, and study the relationships between frailty classes and SRP. These findings can provide a theoretical basis and practical reference for personalized and precise interventions for frailty and SRP with lung cancer receiving radiotherapy.

Methods
The observational study aimed to examine the occurrence of frailty, potential frailty subtypes, the factors influencing these subtypes, and the relationships between frailty subtypes and the SRP in patients with lung cancer receiving radiotherapy. The study was conducted between January and December 2023 in Yunnan Province, China, and was approved by the Ethics Committee of Yunnan Cancer Hospital (Approval No. SLKYLX2023-031).

Study design and sample
Patients with lung cancer who underwent radical thoracic radiotherapy from January 2023 to October 2023 in the radiotherapy department of a tertiary cancer hospital in Yunnan Province were selected by convenience sampling to participate in this study. The inclusion criteria were as follows: (1) age ≥ 18 years old, (2) pathological or cytological confirmation of lung cancer, including non–small cell lung cancer (NSCLC) or limited-stage small cell lung cancer (SCLC),21 (3) the radiotherapy site was confined to the thorax: the target area including the gross tumor volume (GTV), clinical target volume (CTV) and planning target volume (PTV) in the radiotherapy planning system was located in the primary lung tumor and/or involved mediastinal and hilar lymph nodes, (4) the course record or doctor's advice is clearly marked as radical radiotherapy , and meanwhile meets any of the following dosimetry criteria: a. conventional radiotherapy or intensity modulated and volume rotary radiotherapy: the prescription dose is 60–70 Gy / 30–35 times, and b. Postoperative adjuvant thoracic radiotherapy for the purpose of radical cure or reducing the risk of recurrence: the prescription dose is 50–60 Gy / 25–30 times, and (5) at least one imaging evaluation (CT or PET-CT, etc.) had been completed within 6 months after the end of radiotherapy. Exclusion criteria were as follows: (1) absence of complete dosimetry data or clinical follow-up information, (2) treatment intent designated as palliative (explicitly documented as “palliative” in medical advice and progress notes, or prescription dose failing to meet the above radical criteria), (3) primary irradiation site outside the thorax such as bone, brain and other metastases as the main irradiation area, and (4) previous history of radical-dose thoracic radiotherapy.

The sample size estimation
The sample size estimation of this study consisted of two components: LCA and multivariable logistic regression. (1) According to established LCA methodological guidelines, the stability and identifiability of an LCA model depend on the minimum class size. A minimum of 30–50 participants for one latent class and class proportion ≥ 5–10% of the total sample are generally recommended.22,23 In this present study, the smallest latent class accounted for approximately 30% of the total sample (≈61/204), highly above the recommended thresholds, ensuring robust model estimation and convergence. (2) For the binary outcome of symptomatic pneumonia, the sample size was determined via the events per variable (EPV) principle. Following the principle of Peduzziet al.24 and Vittinghoff,25 at least 10–15 events per predictor are required for stable estimation. With 5 independent variables in the study, 50–75 events are needed. The required total sample size was from EPV = 10:50 ÷ 0.40 = 125 to EPV = 15: 75 ÷ 0.40 = 188 participants on the basis of previous symptomatic pneumonia incidence of 40%. After accounting for a 15% loss to follow-up, the adjusted target range was from 147 to 221 participants. The final sample included 204 participants with approximately 82 events (40% × 204), corresponding to EPV≈16.4, exceeding the conservative EPV ≥ 15 criterion. Therefore, the sample size is considered adequate for the planned multivariate model.

Procedures
All patient data were collected by the research team. Prior to enrolment, the researchers screened potential research subjects via the electronic medical record system for radiotherapy, and conducted face-to-face contact during their radiotherapy.
The researchers explained the purpose and significance of the study to the patients in detail, and obtained written informed consent after confirming their willingness to participate voluntarily.
A consecutive sampling method was used to ensure representativeness of the study. All eligible patients with pathologically confirmed lung cancer who underwent radical radiotherapy in the department from January 2023 to October 2023 were consecutively invited to participate until the target sample size was reached.
Data collection was carried out all the radiotherapy period. Paper-based questionnaires were administered in a face-to-face setting to assess frailty and nutritional risk. Each completed form was checked immediately for accuracy and completeness. For participants who were unable to complete the questionnaire independently, a trained researcher provided assistance via a structured question-and-answer format to maintain consistency and reliability of responses.
The diagnosis and grading of SRP were jointly determined by two senior radiation oncologists and one senior diagnostic radiologist on the basis of patients’ clinical symptoms and chest CT results within six months after the completion of radiotherapy. SRP was evaluated in accordance with the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0.
After radiotherapy, treatment-related and dosimetry parameters were extracted from the hospital's radiotherapy information system, including total radiotherapy dose, treatment mode, bilateral lung V20 (%), and bilateral lung mean dose (Dmean, Gy). These variables were analyzed as related factors associated with the occurrence of SRP. Initially, 300 patients were collected. Of these, 89 patients were excluded because of missing post-radiotherapy chest CT follow-up data, and 7 patients were excluded because their questionnaire responses were highly consistent or patterned. Ultimately, 204 patients completed the study, yielding an effective inclusion rate of 68.0%. A flowchart of the data collection process is shown in Fig. 1.

Research instruments

General information questionnaire and radiotherapy parameters
The general questionnaire used in this study was designed through a review of the relevant literature and the inclusion of factors that may affect frailty in patients with lung cancer.26,27 The following information was collected: demographic data (age, sex, occupation, place of residence, educational level, medical payment method, and family average monthly income) and disease-related information (lung cancer cell type, disease duration, tumor-node-metastasis (TNM) stage, number of radiotherapy sessions, comorbidities, treatment history, and the patients’ nutritional risk score). Radiotherapy parameters, including total radiation dose, radiotherapy technique, bilateral lung V20 (%), and bilateral lung mean dose (Dmean, Gy), were extracted from the radiotherapy planning system after completion of treatment to explore factors associated with SRP in patients with lung cancer.

Frailty phenotype
Frailty was assessed via the Fried Frailty Phenotype Scale,8 which includes five indicators: unintentional weight loss, subjective fatigue, decreased grip strength, slow walking speed, and reduced physical activity. Each criterion is scored by a dichotomous scale: 1 point for meeting the criterion, 0 points for not meeting it, with a total score ranging from 0 to 5 points. A score of 0 indicates no frailty, 1–2 points indicate prefrailty, and ≥ 3 points indicate frailty. This scale offers the advantages of simplicity and intuitiveness, with its predictive validity validated across multiple studies, which is widely employed for assessing frailty in elderly individuals and cancer patients. Within mainland China, the scale has been extensively utilized in frailty assessment and risk prediction studies among patients with lung cancer and community-dwelling elderly populations.28 In this study, the internal consistency of the scale was acceptable, with a Cronbach's α coefficient of 0.701.

Classification of radiation pneumonitis
Radiation pneumonitis grading was conducted via version 5.0 of the Common Terminology Criteria for Adverse Events (CTCAE),29 with grade ≥ 2 radiation pneumonitis (RP) defined as significant radiation pneumonitis (SRP). SRP is the endpoint event and refers to SRP occurring within six months after radiotherapy. The diagnosis of SRP was made by two senior radiation oncologists and one senior diagnostic radiologist on the basis of clinical symptoms and chest CT findings.

The Nutrition Risk Screening 2002
The Nutrition Risk Screening 2002 (NRS-2002) comprises three main components: malnutrition assessment (0–3 points) and disease severity assessment (0–3 points). Age assessment (0–1 points for age ≥ 70 years). A total score ≥ 3 indicates malnutrition risk.30

Data analysis
The data were analyzed via SPSS 29.0 and SPSS Mplus 8.7, and origin 2025 software was used to draw separate graphs of frailty categories. In this study, LCA, which usually uses categorical variables, was used to determine the characteristics of frailty in patients receiving lung cancer radiotherapy. According to the scoring rules of the frailty phenotype scale, each item has a score of 1 and no score, and Mplus 8.7 was used to analyze the five dimensions of the frailty scale via a latent class model. The evaluation indices of model fit include the following: (1) Akaike information criterion (AIC), Bayesian information criterion (BIC), and adjusted BIC (aBIC): lower values indicate better model fit.31 (2) Entropy index: This index ranges from 0 to 1, with higher values indicating more accurate classification. (3) Bootstrapped likelihood ratio test (BLRT) and Lo‒Mendell‒Rubin (LMR) test: These tests compare models. A P-value < 0.05 indicates that a model with k classes is preferred over a model with k-1 classes.20 The data were analyzed via SPSS 29.0. Continuous variables are presented as the means ± standard deviations or quartiles, and group comparisons were made via independent sample t tests or analysis of variance. Categorical variables are presented as frequencies and proportions, with comparisons made via chi-square tests or rank-sum tests. Multinomial logistic regression was employed to identify factors of frailty class, and binary logistic regression was used to examine the association between frailty subtypes and the incidence of SRP. A P-value < 0.05 was considered statistically significant.

Results

Latent class analysis of frailty in patients receiving lung cancer radiotherapy
Five models were fitted in this study. For models 1–3, the AIC, BIC, and aBIC decreased, whereas for models 4–5, they increased. Model 3 had the smallest AIC, BIC, and aBIC values. LMR and BLRT were significant (P < 0.05), with an entropy of 0.846 and a classification accuracy over 80%. Thus, a three-class model was chosen (Table 1).
Model 3, the best-fitting model in this study, classifies frailty in patients receiving radiotherapy for lung cancer into three categories on the basis of five assessment criteria (Fig. 2). In our research, fatigue was highly prevalent across all three groups. Specifically, for C1 (prefrailty group), 76 patients (37.3%) experienced fatigue but had low scores on other frailty indicators. This group is termed the “prefrailty group”. C2 (frail physical activity decline group): sixty-five patients (31.9%) presented high rates of both fatigue and reduced physical activity, with the latter being the primary concern. This group is named the “Frail physical Activity Decline Group”. C3 (severe-frail core-strength decline group): sixty-three patients (30.9%) had high scores across all frailty indicators, indicating severe frailty. Notably, they showed the most significant decline in grip strength, reflecting low muscle mass and core strength. This group is termed the “Severe-frail core-strength decline group".

Baseline characteristics and univariate analysis of frailty classes in patients with lung cancer undergoing radiotherapy
Between January and October 2023, a total of 204 patients with lung cancer who received radiotherapy were enrolled in the study. Among them, 174 (85.3%) were male and 30 (14.7%) were female. Regarding age distribution, 114 patients (55.9%) were younger than 60 years, while 90 (44.1%) were aged 60 years or older.
A univariate analysis was performed using the three frailty classes of patients with lung cancer undergoing radiotherapy as the dependent variable, and sociodemographic and disease-related characteristics as independent variables. The results showed that age (H = 7.479, P = 0.006), method of payment for medical expenses (χ2 = 7.786, P = 0.020), nutritional score (H = 8.275, P = 0.004), and number of radiotherapy sessions (χ2 = 30.349, P < 0.001) were significantly associated with frailty class. Detailed baseline characteristics and the results of the univariate analysis are summarized in Table 2.

Multivariate analysis of latent frailty classes in patients receiving lung cancer radiotherapy
In the multinomial logistic regression analysis of the three latent frailty classes in lung cancer radiotherapy patients (C1: prefrailty group, C2: frail physical activity decline group, and C3: severe-frail core-strength decline group), the following variables were used as covariates on the basis of their statistical significance in the univariate analysis: age: < 60 years = 1, ≥ 60 years = 2; Method of payment of medical expenses: Employee medical = 1, Residents' medical = 2; and radiotherapy sessions: < 10 = 1, ≥ 10 and < 20 = 2, ≥ 20 = 3; and nutritional score: < 3 = 1, ≥ 3 = 2). The results indicated that the likelihood ratio test yielded a χ2 value of 99.458 (P < 0.001), while goodness-of-fit tests produced P values > 0.05. The maximum pseudoR2 value was 0.434. Age, number of radiotherapy sessions, and nutritional score were identified as risk factors for patient frailty categories (P < 0.05) (Table 3).

Incidence of radiation pneumonitis in patients
This study included 204 patients with lung cancer who received radiotherapy. Radiation pneumonitis was graded via CTCAE 5.0, with grade ≥ 2 RP (radiation pneumonitis) defined as SRP (significant radiation pneumonitis) within six months after radiotherapy. The incidence of SRP among the patients was 43.6% (Table 4).

Univariate analysis of the incidence of symptomatic radiation pneumonitis in patients undergoing radiotherapy for lung cancer
Using SRP as the dependent variable, univariate analysis was conducted with potential influencing factors including radiotherapy dose, treatment method, bilateral lung V20 (%), and bilateral lung mean dose (Dmean, Gy), frailty latent classes, demographic characteristics, and disease-related information as independent variables. Results indicated statistically significant differences in frailty latent classes (χ2 = 33.331, P < 0.001), V20 (%) (F = −6.126, P < 0.001), and Dmean, (F = −6.120, P < 0.001) (Table 5).

Binary logistic regression analysis of the relationship between frailty latent classes and the incidence of symptomatic radiation pneumonitis in lung-cancer patients undergoing radiotherapy
With the occurrence of SRP as the dependent variable (0 = absent, 1 = present), the variables with statistical significance in the univariate analysis were included in the independent variables for binary logistic regression analysis. The independent variables were assigned as follows: 1 = prefrailty group, 2 = frail physical activity decline group, 3 = severe-frail core-strength decline group, bilateral lung V20 (%), and bilateral lung mean dose (Dmean, Gy) were included in the analysis in the form of continuous variables.
In Model I, the frailty latent class, bilateral lung V20 (%), and bilateral lung mean dose (Dmean, Gy) were included. The results showed that frailty potential category and bilateral lung V20(%) could be used as independent influencing factors for the occurrence of SRP in patients with lung cancer radiotherapy (P < 0.05). The model can explain 47.3% of the variation in the incidence of radiation pneumonitis, suggesting that the patient's frailty class and radiation dosimetry parameters are risk factors for the incidence of SRP in patients (Table 6).
In model II, only the frailty latent class was included to evaluate the independent effect of the frailty class on SRP. The results showed that the frailty class was still a significant predictor (P < 0.05), and the model could explain 21.3% of the variation in the incidence of radiation pneumonitis, suggesting that the frailty class of patients is closely related to the occurrence of SRP even without considering radiation dosimetry factors. With the aggravation of frailty, the risk of radiation pneumonitis gradually increases (Table 7).

Discussion
To the best of our knowledge, this is the first study to investigate the association between latent frailty classes and SRP in patients with lung cancer undergoing radiotherapy. It also represents the inaugural application of LCA to uncover frailty heterogeneity and its determinants in this population. These findings offer new insights for managing frailty and SRP in patients with lung cancer receiving radiotherapy.

Frailty class characteristics in lung-cancer patients receiving radiotherapy
The frailty in lung cancer radiotherapy patients can be categorized into three latent classes: 37.3% in the prefrailty group, 31.9% in the frail physical activity decline group, and 30.9% in the severe-frail core-strength decline group. The frailty category was consistent with Béland et al.32 in elderly patients, but the incidence of fatigue in this study was higher than that in previous studies.33 The possible reasons are as follows: First, the development of frailty is gradual.34 Simple fatigue is often a signal of prefrailty. When fatigue and physical activity decline at the same time, it suggests that frailty has entered a more serious and faster stage. Secondly, advanced tumor stage and anti-tumor treatment are important factors for the high incidence of fatigue.35,36 In this study, most of patients (91.67%) received anti-tumor treatment, and all patients were stage III or IV, which partly explained the high incidence of fatigue.
In the prefrailty group, the incidence of other indicators except fatigue is low. The patients are in the early stage of frailty and have strong reversibility. Frailty management should focus on prevention. Exercise therapy is one of the effective methods to prevent frailty.37,38 Medical staffs and their families can urge patients to maintain regular exercise through health education to delay or prevent the occurrence of frailty.
In the frail physical activity decline group, the reduction of physical activity was more prominent in addition to fatigue. Helping patients recover or maintain basic exercise is the main measure to delay the process of frailty. However, this study found that some elderly patients still adhere to the traditional belief of “bed rest” or “quiet recuperation,” and because of the limitations of objective factors such as intravenous infusion catheters and drainage tubes, they often dare not exercise. Therefore, in addition to health education, individualized exercise prescription including exercise mode and exercise volume should be formulated according to the actual situation of patients. Meanwhile, it is worth exploring the application of mobile internet and artificial intelligence technology in the formulation and implementation of the exercise plan to improve the feasibility and compliance of the exercise plan.
The patients in the severe frailty with core strength decline group showed higher performance in all frailty indicators, often accompanied by reduced activity, weight loss and severe fatigue, and the quality of life decreased significantly. Management at this stage should rely on comprehensive, multidisciplinary interventions.39 Fundamental nursing care and symptom control are essential, and rehabilitation and nutrition specialists can form individualized exercise and dietary plans to mitigate frailty manifestations, improve overall functional status, and enhance quality of life.
Notably, the Just-in-Time Adaptive Intervention (JITAI) model has been validated in the field of mental health and health management.40,41 If JITAI is introduced into the frailty management of patients with lung cancer radiotherapy, remote, dynamic and individualized intervention can be achieved. Specifically, after the JITAI system is embedded in the medical end, it interacts with the patients' mobile device, and the system can regularly identify the patients' debilitating state and category, thereby providing targeted immediate intervention. This model will improve the management efficiency of frailty and promote the whole, timely and accurate intervention of frailty in patients with lung cancer radiotherapy.

Influencing factors and clinical implications
This study further found that patient age, number of radiotherapy sessions and nutritional risk score were important influencing factors of frailty. Among them, age and nutritional risk are consistent with previous findings,29,42,43 and number of radiotherapy sessions is a new finding. Compared with patients with reduced physical activity or severe frailty, patients who received less than 10 radiotherapy sessions were more likely to be in prefrailty, whereas patients who received 10–19 times of radiotherapy were more likely to have frailty based on reduced physical activity. Radiotherapy is the cornerstone of lung cancer treatment. During the course of treatment, patients often have side effects such as fatigue, radiation esophagitis-related pain, loss of appetite and malnutrition especially around the 10th time, and gradually increase with the increase of treatment times and doses 44, 45, 46, 47. These adverse reactions are important risk factors for frailty, and the more sessions of radiotherapy in patients with moderate to severe frailty in this study also confirms this conclusion. Therefore, in the frailty management of patients with lung cancer receiving radiotherapy, active control of radiotherapy-related symptoms is a key measure to delay the progression of frailty in addition to early identification of risk factors. For patients receiving more than 10 times of radiotherapy, it is recommended to routinely assess the severity of side effects of radiotherapy every week and give timely intervention to reduce symptom distress. Moreover, future studies may consider applying the growth mixture model (GMM) to explore the dynamic trajectory of frailty in patients with lung cancer at different stages of radiotherapy, as well as the category characteristics of different populations. This will provide a more solid theoretical basis for early identification and individualized management of frailty.48

The incidence of symptomatic radiation pneumonitis
In this study, the incidence of SRP in patients within 6 months after radiotherapy was 43.63%, which was similar to 41.4% reported by Wang et al.49 in mainland China, but significantly higher than the 20% to 37% reported in the wider literature.50,51 This difference may be due to the differences in the study population and treatment characteristics. All the patients in this study received intensity modulated radiation therapy (IMRT). Although this treatment method improved the concentration of the tumor target dose, the patients received more radiotherapy, and the cumulative dose was greater. In addition, 91.67% of patients in this study had received anti-tumor therapy, which may further lead to a decline in physiological reserve and impaired multi-system function, thereby increasing the risk of SRP.

The relationship between frailty latent class and symptomatic radiation pneumonitis
In Model I, frailty latent class, bilateral lung V20 (%) and Dmean (Gy) were included. The results showed that both the frailty latent class and the bilateral lung V20 (%) were independent risk factors for SRP (P < 0.05), and the model could explain 47.3% of the variation in the incidence of SRP (Table 6). In model II, only the frailty latent class was included. The results showed that the frailty category was still a significant predictor (P < 0.05), and the model could explain 21.3% of the variation in the incidence of SRP (Table 7). This shows that even if the radiation dosimetry factors are not considered, the degree of frailty is still closely related to the occurrence of SRP. With the aggravation of frailty, the risk of SRP gradually increases, suggesting that frailty can be used as a non-dosimetry early clinical risk signal to predict the occurrence of SRP.
This relationship may be explained by multiple mechanisms. Firstly, patients with severe frailty frequently present with malnutrition and compromised immune function, diminishing the body's capacity to repair radiotherapy-induced pulmonary tissue damage.52 Secondly, cumulative adverse reactions during radiotherapy such as fatigue, esophagitis, and reduced appetite may accelerate declines in physical strength and pulmonary function, thereby increasing susceptibility to SRP.7,14,53 Furthermore, patients with severe frailty frequently experience activity limitations and inadequate respiratory reserve, which may exacerbate pulmonary inflammatory responses.15,54,55 In addition to the above factors, current research shows that frailty can be regarded as a systemic biological state characterized by chronic low-grade inflammation, oxidative stress, mitochondrial dysfunction, and endothelial damage. These pathological processes together weaken the body's tissue repair capacity and also increase the risk of radiation-induced oxidative damage and fibrosis.56, 57, 58, 59, 60 Relevant studies suggest that incorporating frailty assessment into the routine nursing care for patients undergoing radiotherapy for lung cancer should be considered one of the measures for preventing and managing SRP. For patients with prefrailty, health education and early intervention should be strengthened to delay the progression of frailty. For patients with decreased physical activity, rehabilitation exercise and nutritional intervention should be combined to improve body tolerance. For patients with severe frailty, multidisciplinary cooperation is needed, not only to comprehensively intervene in frailty symptoms, but also to focus on early identification and intervention of SRP to reduce the incidence of complications. Medical staff can help patients delay or reverse the state of frailty, thereby reducing the incidence of SRP. Meanwhile, future investigations should adopt longitudinal study designs and employ growth mixture modeling (GMM) to delineate dynamic frailty trajectories across radiotherapy courses and explore the temporal relationship between frailty evolution and SRP occurrence. These approaches could help construct a scientifically grounded, individualized SRP risk-prediction model, providing stronger evidence for precision nursing and personalized supportive care in radiotherapy practice.

Implications for nursing practice
This study indicated that there is population heterogeneity in frailty in patients with lung cancer radiotherapy, and there are significant differences in the incidence of SRP in patients with different frailty categories. This association may be mediated by decreased physiological reserve, multiple system dysfunction, and impaired inflammation and tissue repair processes, so frailty can be considered as an interventional risk factor for SRP. This finding not only provides a new entry point for the prevention and management of SRP in patients with lung cancer radiotherapy, but also further highlights the importance of frailty management. Nursing staff should pay attention to early screening, timely identify frailty and its risk factors, and implement personalized and multidisciplinary comprehensive intervention according to individual characteristics to delay and even reverse the frailty process, further reducing the risk of SRP. Future studies should verify these findings in a larger and more diverse patient study, and evaluate the effectiveness of nursing-led targeted interventions in preventing frailty progression and reducing the risk of SRP to promote the organic combination of cancer care and quantifiable clinical outcomes.

Limitations
This study has several limitations. Firstly, this is a single-center study. It only included patients with lung cancer undergoing radiotherapy within this specific setting due to time constraints and limitations in radiotherapy resources. The relatively small sample size may limit the generalizability of the findings. Future investigations could be conducted across multiple centers, with larger samples and broader geographical coverage to strengthen these results. Secondly, lung cancer itself exhibits high heterogeneity. Although our model adjusted for key clinically relevant variables, it did not control for the influence of other confounding factors such as clinical case classification or history of respiratory disease. Future studies can conduct in-depth analysis of lung cancer subgroups to further elucidate the association between frailty and radiotherapy. Finally, this study lacks an independent external validation cohort. Consequently, the derived latent categories and their number may be influenced by specific sample characteristics with potential instability. Future prospective cohort studies can validate these findings, track the dynamic trajectory of frailty, and lay the groundwork for developing scientifically grounded and dynamic intervention strategies.

Conclusions
This study firstly identified distinct frailty subtypes among patients with lung cancer undergoing radiotherapy via LCA. Three frailty classes were identified: prefrailty, frailty with physical activity decline, and severe frailty with core strength decline. Frailty classification was independently associated with radiotherapy sessions, nutritional status, and age, and served as a significant predictor of SRP.
The novelty of this study lies in integrating LCA derived frailty phenotypes with SRP risk, providing new insights into how heterogeneity in frailty influences treatment-related toxicity. These findings offer a theoretical and practical basis for precision nursing assessments and individualized frailty management in patients undergoing radiotherapy.
Medical and nursing teams should design targeted prevention and intervention strategies on the basis of frailty subtypes to delay frailty progression and reduce SRP risk. Future studies may explore the application of Just-in-time Adaptive Interventions (JITAI) to achieve real-time, remote monitoring and personalized management of frailty, potentially improving clinical outcomes in this population.

CRediT authorship contribution statement
Jiang Zhang and Xijuan Zhao conceived the research idea, developed the study framework and plan, and wrote the manuscript. Jiawei Liao and Yanyan Fei provided a grading diagnosis of radiation pneumonitis in patients. Lu Xu and Jiang Wu participated in data collection. Song Li and Qiongyao Guan provided valuable feedback and contributed to the final draft. All the authors contributed to and approved the final version of the manuscript.

Ethics statement
This study was approved by the Ethics Committee of Yunnan Cancer Hospital (Approval No. SLKYLX2023-031) and was conducted in accordance with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. All participants provided written informed consent.

Data availability statement
The data that support the findings of this study are available from the first author (Email address: 1120209154@qq.com), upon reasonable request.

Declaration of generative AI and AI-assisted technologies in the writing process
No AI tools/services were used during the preparation of this work.

Funding
This study was supported by the Scientific Research Fund of the Yunnan Provincial Department of Education (No. 2023Y0757 and 2024J0363) and the Chinese Nursing Association's 2023 Established Scientific Research Project (No. ZHKYQ202312). The funders had no role in considering the study design or in the collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication.

Declaration of competing interest
The authors declare no conflict of interest.