Published online Nov 24, 2026. doi: 10.5306/wjco.127553
Revised: September 20, 2026
Accepted: October 28, 2026
Published online: November 24, 2026
Processing time: 77 Days and 7.3 Hours
Recurrence risk and overall survival (OS) after curative-intent resection for stage I non-small-cell lung cancer remain incompletely explained by tumor-related variables. Patient-reported outcomes may capture dimensions of postoperative health not represented by conventional clinicopathologic factors, but their prognostic role during survivorship remains insufficiently defined.
To determine whether patient-reported outcomes assessed 1 year after surgery are associated with subsequent recurrence and OS.
We analyzed 349 consecutive patients undergoing resection for pathological stage I non-small-cell lung cancer (NSCLC) between 2006 and 2018. The 1-year landmark cohort comprised 232 patients alive and recurrence-free with available health-related quality of life data. Global health status (QL2), fatigue (FA), and dyspnea (DY) from the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30 were evaluated. Recurrence was the primary endpoint, with death treated as a competing event. Multivariable recurrence analyses used cause-specific Firth Cox and Fine-Gray models; OS was assessed using multivariable Firth Cox regression.
During follow-up, 35 patients (15.1%) developed recurrence and 15 (6.5%) died without prior relapse. Age ≥ 65 years [hazard ratio (HR) = 6.51, 95%CI: 2.37-17.90], higher QL2 (HR = 2.50, 95%CI: 1.15-5.40), and higher FA (HR = 3.48, 95%CI: 1.66-7.27) were independently associated with recurrence, whereas DY was inversely associated (HR = 0.33, 95%CI: 0.16-0.71). Fine-Gray models yielded consistent estimates. For OS, age ≥ 65 years (HR = 4.90, 95%CI: 1.19-20.05) and DY > 0 (HR = 0.26, 95%CI: 0.09-0.81) were independently associated with mortality, whereas FA ≥ 22.2 showed a clinically relevant but borderline association (HR = 2.94, 95%CI: 0.96-9.02). Model discrimination was good for recurrence (C-index 0.79) and OS (C-index 0.73).
Postoperative patient-reported outcomes assessed at 1 year were associated with heterogeneous recurrence and survival trajectories, suggesting complementary prognostic information during survivorship after stage I NSCLC resection.
Core Tip: This landmark study evaluated whether patient-reported outcomes assessed 1 year after curative-intent resection for stage I non-small-cell lung cancer were associated with subsequent recurrence and survival. Among 232 patients alive and recurrence-free at the landmark, global health status, fatigue, and dyspnea showed distinct associations with recurrence, while dyspnea was associated with overall survival and fatigue showed a borderline association in the primary model. These findings suggest that postoperative patient-reported outcomes may provide complementary prognostic information during long-term survivorship.
- Citation: Leuzzi G, Cosentino B, Sabia F, Pastorino U, Pompili C, Dai W, Guerrera F, Prisciandaro E, Lococo F, Ferrari M, Rolli L, Pardolesi A, Stanzi A, Calderoni M, Uslenghi C, Lorenzini D, Fabbri A, Proto C, Lo Russo G, Prelaj A, Marchianò A, Greco G, Solli P. Association of one-year postoperative patient-reported outcomes with recurrence and survival in stage I non-small-cell lung cancer. World J Clin Oncol 2026; 17(11): 127553
- URL: https://www.wjgnet.com/2218-4333/full/v17/i11/127553.htm
- DOI: https://dx.doi.org/10.5306/wjco.127553
Despite generally favorable outcomes after curative-intent resection for stage I non-small-cell lung cancer (NSCLC), a proportion of patients develop recurrence. This variability persists even within the same pathological stage, suggesting that tumor-node-metastasis (TNM) classification does not fully capture prognosis[1,2].
Postoperative risk stratification and surveillance are based on clinicopathologic features and imaging, with very limited integration of patient-centered parameters[3]. In this context, functional status and symptom burden may provide additional prognostic information beyond usual objective evaluation and staging[4].
Patient-reported outcomes (PROs), including health-related quality of life (HRQoL) measures such as global health status (QL2), fatigue (FA), and dyspnea (DY), are increasingly recognized as clinically relevant indicators of patients’ perceived health status and symptom burden. Specifically, these measures capture aspects of physiological reserve, behavioral patterns, and symptom perception that are not fully represented by traditional clinical variables. In advanced NSCLC and during systemic therapy, a substantial body of evidence has demonstrated that HRQoL domains may independently predict overall survival (OS), frequently outperforming classical prognostic factors[5-8].
However, most available data derive from patients with advanced or metastatic disease, in whom HRQoL is typically assessed at baseline or during treatment. Conversely, the prognostic role of HRQoL in patients undergoing surgery for early-stage NSCLC remains insufficiently explored. Moreover, previous studies have predominantly focused on OS only, with limited attention to recurrence patterns in the survivorship phase.
Lung resection represents a major physiological stressor, and postop recovery trajectories vary widely among patients. Previous studies have shown that HRQoL domains, including FA and DY, may follow heterogeneous longitudinal patterns after surgery, showing differences in cardiopulmonary reserve, physical activity, and adaptive processes[9-12]. These observations suggest that selected PROs assessed after recovery may capture an integrated representation of patient condition, instead of isolated symptoms.
The concept of a postoperative functional phenotype may help interpret these measures. Variables such as QL2, FA, and DY may describe a multidimensional state integrating physiological resilience and well-being, adaptive behavior, and symptom perception. DY may be influenced by physical activity and symptom awareness, while response shift may affect self-reporting over time[13]. Paradoxical associations, as in reverse epidemiology, increase the difficulty of inter
To address these issues, a landmark analysis may afford a robust methodological approach by evaluating patients who have completed the early postoperative recovery phase and remain disease-free at a predefined time point. This strategy allows assessment of conditional outcomes while decreasing bias related to early postoperative events[3].
The aim of this study was to determine whether patient-reported HRQoL domains assessed 1 year after surgery can identify distinct recurrence trajectories and to examine their association with long-term OS in patients undergoing curative-intent resection for stage I NSCLC.
This study constitutes a retrospective analysis of prospectively collected clinical and PRO data from a dedicated thoracic oncology database at Fondazione IRCCS Istituto Nazionale dei Tumori (Milan, Italy). The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Fondazione IRCCS Istituto Nazionale dei Tumori (IRB Protocol No. INT 6-25, approved on February 7, 2025). In accordance with the Italian Privacy Code (D.Lgs. 196/2003, Art. 110-bis, par. 4) and the retrospective design of the study, the institutional review board approved the use of existing de-identified data and waived the requirement for individual re-consent. Patient confidentiality was strictly preserved throughout the study.
We included consecutive patients undergoing curative-intent R0 resection with systematic nodal dissection for pathological stage I NSCLC between January 2006 and December 2018. To obtain a clinically homogeneous cohort treated with curative-intent surgery alone, we excluded patients with prior lung cancer, synchronous malignancies, neoadjuvant or adjuvant systemic therapy, or incomplete resection. Patients with inadequate follow-up or unavailable PRO data were also excluded because subsequent outcomes and the prespecified 1-year exposure variables could not be reliably as
Clinical and pathological variables retrieved from the institutional database included patient demographics, smoking history, chronic lung disease, histological subtype, pathological stage, and type of surgical resection.
Patients were managed through a structured follow-up program including scheduled clinical evaluation and imaging, as previously described[3].
HRQoL was prospectively assessed during follow-up using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30 (EORTC QLQ-C30) and the lung cancer-specific Quality of Life Ques
A 1-year landmark analysis was performed. The landmark cohort included patients alive and recurrence-free at 12 months with available HRQoL questionnaires. Time-to-event analyses used the date of the 1-year HRQoL assessment as time zero, reducing bias related to early postoperative events.
The primary endpoint was disease recurrence, defined as locoregional or distant relapse confirmed by radiological and/or pathological assessment. Cumulative incidence of recurrence (CIR) was estimated within the landmark cohort, with death without prior recurrence treated as a competing event. The secondary endpoint was OS, defined from the 1-year landmark assessment to death from any cause or last follow-up.
Postoperative functional status was evaluated using clinically selected domains from the EORTC QLQ-C30 and QLQ-LC13. Eleven domains were identified a priori on the basis of their clinical relevance to postoperative lung cancer survivorship: QL2, physical functioning (PF2), cognitive functioning, FA, nausea and vomiting, pain, DY, insomnia (SL), LC13 DY, coughing, and chest pain.
For categorical analyses, QL2 and FA were dichotomized using previously published thresholds of 66.7 and 22.2, respectively[5]. DY was categorized according to symptom absence (DY = 0) vs presence (DY > 0).
Candidate PRO domains were selected a priori on the basis of their clinical relevance to postoperative recovery and survivorship after lung cancer surgery, and the final models were constructed using a parsimonious approach to limit overfitting.
Baseline characteristics were summarized using medians with interquartile ranges or counts and percentages. Group comparisons were performed using the χ2 test or Fisher’s exact test, as appropriate. CIR was estimated using cumulative incidence functions (CIF), with death without recurrence treated as a competing event; groups were compared using Gray’s test[17].
Associations with recurrence were evaluated using cause-specific Cox regression with Firth penalization[18] and Fine-Gray subdistribution regression[19]. Candidate clinical covariates included age, sex, smoking status, histology, and type of surgical resection; only age was retained alongside the selected PRO domains in final models due to event constraints and clinical relevance. Given the limited number of outcome events, multivariable models were constructed using a parsimonious approach, taking into account clinical relevance, evidence of association in preliminary analyses, and the events per variable (EPV) rule to prevent overfitting and guarantee model stability. No automated stepwise selection procedure was used.
OS was analyzed using Kaplan-Meier estimates and multivariable Firth-penalized Cox regression. The same parsimonious modeling approach was applied to the survival analysis. All multivariable analyses were performed in the 1-year landmark cohort (n = 232). No clinical covariate was excluded because of missingness or collinearity.
As a sensitivity analysis, the PRO variables included in the primary multivariable models were additionally evaluated as continuous variables, with effect estimates expressed per 10-point increase, using the same modeling framework as the corresponding categorical analyses. Internal model validation was performed using a non-parametric bootstrap procedure with 1000 resamples generated via simple random sampling with replacement, deriving median hazard ratios (HR) and 95% confidence intervals (CI) from the 2.5th and 97.5th percentiles of the empirical distributions. Model discrimination was evaluated using Harrell’s C-index[20].
Proportional-hazards assumptions were assessed using time-dependent interaction terms for each covariate; no significant violations were detected.
To assess the incremental prognostic value of PRO domains beyond a clinicopathologic base model, nested models were evaluated using the Akaike Information Criterion (AIC). Given the exploratory nature of the analysis and the clinically informed selection of candidate PRO domains, no formal adjustment for multiple comparisons was applied. Follow-up was administratively censored at 8 years from the landmark.
Among 349 patients who underwent curative-intent resection for pathological stage I NSCLC between 2006 and 2018, 232 were alive and recurrence-free at 1 year after surgery and had available HRQoL questionnaire data (landmark cohort). The derivation of the landmark cohort is shown in Figure 1. Baseline clinicopathologic characteristics are summarized in Table 1. Baseline characteristics of patients included in and excluded from the landmark cohort are reported in Supple
| Variable | Value |
| Age, median (IQR) | 67 (61-72) |
| Age < 65 years | 91 (39.2) |
| Age ≥ 65 years | 141 (60.8) |
| Sex | |
| Female | 39 (16.8) |
| Male | 193 (83.2) |
| Smoking status | |
| Never smoker | 37 (15.9) |
| Former smoker | 115 (49.6) |
| Active smoker | 79 (34.1) |
| Missing | 1 (0.4) |
| COPD | |
| No | 68 (29.3) |
| Yes | 164 (70.7) |
| Pack-years, median (IQR) | 36 (10-50) |
| < 40 | 126 (54.3) |
| ≥ 40 | 106 (45.7) |
| Type of resection | |
| Lobectomy | 187 (80.6) |
| Bilobectomy | 1 (0.4) |
| Segmentectomy | 37 (15.9) |
| Wedge Resection | 7 (3.0) |
| Histology | |
| Adenocarcinoma | 179 (77.2) |
| Squamous cell carcinoma | 33 (14.2) |
| Others | 20 (8.6) |
| Stage | |
| IA | 179 (77.2) |
| IB | 53 (22.8) |
| QLQ-C30 scores at 1 year | |
| QL2, median (IQR) | 66.7 (66.7-83.3) |
| QL2 < 66.7 | 110 (47.4) |
| QL2 ≥ 66.7 | 122 (52.6) |
| FA, median (IQR) | 0 (0-33.3) |
| FA < 22.2 | 117 (50.4) |
| FA ≥ 22.2 | 115 (49.6) |
| DY, median (IQR) | 0 (0-33.3) |
| DY = 0 | 103 (44.4) |
| DY > 0 | 129 (55.6) |
The median follow-up time from the 1-year landmark assessment was 6.4 years within the prespecified 8-year analysis horizon. During follow-up, 35 patients (15.1%, 95%CI: 10.48%-19.69%) developed disease recurrence, whereas 15 patients (6.5%, 95%CI: 3.30%-9.63%) died without prior evidence of relapse.
Longitudinal HRQoL trajectories in the overall cohort are shown in Figure 2. At the 1-year landmark assessment, patient-reported domains showed a heterogeneous distribution across the cohort (Table 1, Supplementary Table 2). Detailed comparative analyses of the EORTC QLQ-C30 and QLQ-LC13 domains are reported in Supplementary Tables 3 and 4.
At univariate analysis, age ≥ 65 years, higher QL2, and absence of DY were associated with recurrence, whereas FA showed a non-significant trend (Supplementary Table 5).
The CIR increased over time, reaching 14.7% (95%CI: 10.18%-19.93%) at 5 years when accounting for death as a competing event (Figure 3). CIF curves demonstrated a higher probability of relapse in patients aged ≥ 65 years (5-year CIR 3.4% vs 23.2%, P < 0.0001), in patients without DY (DY = 0; 5-year CIR 10.1% vs 20.0%, P = 0.0070), and in those with higher QL2 scores (QL2 ≥ 66.7; 5-year CIR 10.7% vs 18.1%, P = 0.0466). Differences in relapse trajectories according to FA were less pronounced, although a trend toward higher relapse probability was observed in patients with higher FA scores (FA ≥ 22.2; P = 0.0797). Absolute recurrence risks at 3 and 5 years were 4.9% (95%CI: 1.8%-10.4%) and 10.7% (95%CI: 5.4%-18.0%) for QL2 < 66.7 vs 12.1% (95%CI: 6.9%-18.8%) and 18.1% (95%CI: 11.5%-26.0%) for QL2 ≥ 66.7; 14.1% (95%CI: 8.1%-21.7%) and 20.0% (95%CI: 12.6%-28.7%) for DY = 0 vs 4.2% (95%CI: 1.6%-9.0%) and 10.1% (95%CI: 5.3%-16.7%) for DY > 0; and 6.3% (95%CI: 2.8%-11.9%) and 9.5% (95%CI: 4.8%-16.1%) for FA < 22.2 vs 11.2% (95%CI: 6.1%-18.1%) and 20.0% (95%CI: 12.7%-28.4%) for FA ≥ 22.2, respectively.
Multivariable analysis using cause-specific Cox (Table 2, Supplementary Figure 1) and Fine-Gray models (Table 3), with 35 events and 4 predictors hence 8.75 EPV, showed that age ≥ 65 years (HR = 6.51, 95%CI: 2.37-17.90), higher QL2 (HR = 2.50, 95%CI: 1.15-5.40), and higher FA (HR = 3.48, 95%CI: 1.66-7.27) were independently associated with increased recurrence risk, whereas DY was inversely associated (HR = 0.33, 95%CI: 0.16-0.71). Primary model discrimination was good (C-index 0.79, 95%CI: 0.73-0.86). Results were consistent in Fine-Gray models [subdistribution HR (sHR) = 6.78, 95%CI: 2.37-19.39; sHR = 2.44, 95%CI: 1.23-4.83; sHR = 3.47, 95%CI: 1.72-6.99; sHR = 0.34, 95%CI: 0.17-0.69, respectively]. Internal validation via 1000 bootstrap resamples demonstrated high stability of these estimates across all recurrence models (Tables 2 and 3).
| Variable | Primary model HR (95%CI) | Bootstrap model median HR (95%CI) |
| Age ≥ 65 years | 6.51 (2.37-17.90) | 6.05 (3.48-15.53) |
| QL2 ≥ 66.7 | 2.50 (1.15-5.40) | 2.41 (1.47-4.30) |
| Fatigue ≥ 22.2 | 3.48 (1.66-7.27) | 3.57 (2.20-7.05) |
| Dyspnea > 0 | 0.33 (0.16-0.71) | 0.33 (0.17-0.55) |
| Model discrimination | ||
| C-index | 0.79 (0.73-0.86) | 0.79 (0.74-0.85) |
| Variable | Primary model sHR (95%CI) | Bootstrap model median sHR (95%CI) |
| Age ≥ 65 years | 6.78 (2.37-19.39) | 6.67 (3.60-21.58) |
| QL2 ≥ 66.7 | 2.44 (1.23-4.83) | 2.42 (1.45-4.51) |
| Fatigue ≥ 22.2 | 3.47 (1.72-6.99) | 3.50 (2.12-6.97) |
| Dyspnea > 0 | 0.34 (0.17-0.69) | 0.34 (0.17-0.57) |
To evaluate the incremental prognostic value of PROs, nested models using AIC criterion showed a progressive decrease from 342.852 (baseline age model) to 327.511 (full model: Age + QL2 + FA + DY), confirming the incremental prognostic performance of PRO domains for disease recurrence.
In sensitivity analyses treating PRO domains as continuous variables, internal validation via bootstrap showed that FA remained positively associated with recurrence (HR per 10-point increase 1.21, 95%CI: 1.02-1.40) and DY remained inversely associated (HR = 0.81, 95%CI: 0.62-0.95), whereas QL2 showed no significant linear association (HR = 1.01, 95%CI: 0.89-1.40) (Supplementary Table 6).
The estimated conditional 5-year OS rate was 95.35%. Univariate analyses are reported in Supplementary Table 7. Older age was associated with worse OS, whereas DY > 0 showed a borderline association (Figure 4). In multivariable Firth-penalized Cox models, with 15 events and 3 predictors hence 5 EPV, age ≥ 65 years (HR = 4.90, 95%CI: 1.19-20.05) and DY > 0 (HR = 0.26, 95%CI: 0.09-0.81) were independently associated with mortality. High FA (≥ 22.2) showed a strong clinical association with increased mortality risk, although it narrowly missed statistical significance due to the small sample size (HR = 2.94, 95%CI: 0.96-9.02). Primary model discrimination was good (C-index 0.73, 95%CI: 0.58-0.88) (Supplementary Table 8). Internal validation via 1,000 bootstrap resamples confirmed overall model consistency and supported the prognostic relevance of FA (median HR = 2.88, 95%CI: 1.38-7.31). Nested model evaluation via AIC demonstrated a progressive drop from 141.927 in the baseline age model to 136.039 in the full multivariable model (age + DY + FA), supporting the incremental prognostic value of PROs for survival.
When FA and DY were modeled continuously, internal validation via bootstrap showed that FA remained associated with OS (HR per 10-point increase 1.29, 95%CI: 1.07-1.67), whereas DY showed no statistically significant linear association (HR = 0.80, 95%CI: 0.50-1.02) (Supplementary Table 6). Additional proportional-hazards diagnostics, nested-model AIC analyses, longitudinal questionnaire availability, and absolute recurrence risks are reported in Supplementary Tables 9-12.
This study provides novel insights into the association between postoperative functional status and disease recurrence in patients undergoing curative-intent resection for stage I NSCLC. Specifically, older age, higher FA, and higher QL2 were associated with increased recurrence, whereas higher DY was associated with lower recurrence risk and improved OS. These findings suggest that postoperative symptom profiles reflect multidimensional patient states beyond a purely deficit-based framework.
A substantial body of evidence demonstrates the prognostic value of PROs in oncology, particularly in advanced NSCLC and during systemic therapy, where higher HRQoL and lower symptom burden consistently correlate with improved survival[5-8]. However, these associations have primarily been described in relation to mortality, surgical complications, and outcomes, with minimal exploration of recurrence trends in early-stage, surgically treated populations. In particular, data linking postoperative PROs assessed after completion of early recovery to subsequent recurrence in resected stage I NSCLC remain very limited. The present findings show that the relationship between patient-reported domains and clinical outcomes may differ substantially when relapse, rather than survival, is considered as the primary endpoint.
In our study, the observed patterns do not support a straightforward interpretation. Higher QL2, indicating better perceived HRQoL, was associated with an increased risk of recurrence. Similarly, greater FA was associated with increased recurrence risk. In contrast, the presence of DY, despite reflecting greater symptom burden, was associated with a lower risk of recurrence. These discordant associations suggest that the individual PRO domains may capture different components of postoperative patient status. Rather than representing isolated symptoms, QL2, FA, and DY may describe a broader patient-reported functional profile integrating perceived well-being, symptom burden, behavioral patterns, and adaptation during the survivorship phase. However, given the observational design, these counterintuitive associations should be regarded as hypothesis-generating rather than as evidence of causal relationships.
This interpretation is supported by evidence from symptom science and survivorship research, indicating that symptom burden, functional capacity, and global health perception are interrelated and reflect underlying biological and behavioral states. In particular, the concept of allostatic load describes cumulative physiological stress influencing HRQoL trajectories and systemic resilience beyond single-organ function[21]. Similarly, models of symptom “causal cascade” suggest that multiple QOL domains may converge through FA, supporting the interpretation of PROs as an integrated network rather than independent variables[22]. In this context, functional resilience further captures the ability to maintain or recover functional capacity despite disease-related stressors, potentially linking symptom burden to long-term outcomes[23].
Within the present framework, the inverse association between DY and outcomes is counterintuitive and should be interpreted cautiously. Possible explanations include differences in physical activity and symptom perception[11], stable postoperative respiratory sequelae[9], psychological adaptation and response shift[13], and differences in clinical phenotype, including smoking-related factors[24]. Reverse epidemiology has also been proposed as a conceptual framework for paradoxical symptom-outcome associations[14]. However, none of these mechanisms was directly measured in our study; therefore, they remain speculative, and the observed DY association should be considered hypothesis-generating.
Conversely, the association between higher QL2 and increased relapse risk requires careful interpretation. This finding differs from the well-established relationship between higher HRQoL and improved survival in advanced disease. Postoperative adaptation and response shift may influence self-reported QL2[13], but these mechanisms were not directly assessed here. Accordingly, the QL2 association should be regarded as exploratory and requires external validation.
FA and DY showed potentially relevant associations with survival in our cohort, although the limited number of deaths requires cautious interpretation. Cancer-related FA is associated with lowered physical activity, frailty, and diminished physiological reserve[25,26]. In surgical populations, FA may reflect the relationship between cardiopulmonary fitness, muscle function, and postoperative recovery. As well, exercise capacity and physical performance have repeatedly been shown to modify outcomes in lung cancer patients. For example, perioperative exercise capacity has been associated with HRQoL and rehabilitation pathways after lung cancer surgery[11]. Similarly, rehabilitation programs have been shown to improve functional status and symptom burden in NSCLC survivors[12]. As a result, these observations suggest that patient-reported FA may serve as a surrogate marker of global physiological reserve rather than an isolated symptom. In this context, the association between FA and survival observed in our study may reflect underlying differences in “functional resilience” between patients who have successfully completed curative surgical treatment.
From a clinical perspective, these data have potential relevance for postoperative survivorship assessment in early-stage NSCLC. Current standard follow-up protocols are based on imaging schedules and clinicopathologic features, with limited integration of patient-centered variables. Our results show that incorporating PROs into postop assessment may provide additional information regarding relapse risk. In particular, identifying patient-reported functional profiles at 1 year after surgery in stage I patients may refine risk stratification beyond histology or TNM staging, supporting their potential role as complementary tools in postop assessment. This is consistent with our previous work demonstrating the value of landmark-based strategies for “tailoring” follow-up in early-stage NSCLC[3]. At present, these findings do not support modification of imaging surveillance schedules; rather, PRO assessment may help identify patients who warrant focused clinical evaluation, rehabilitation referral, or symptom-directed management. In the context of follow-up outpatient clinics, functional status represents a potentially modifiable domain that can be easily identified by thoracic surgeons, pulmonologists, or oncologists. Interventions targeting physical activity, symptom control, and perioperative optimization have been shown to improve HRQoL and functional capacity in patients undergoing lung cancer surgery[12]. As a result, early identification of patients with unfavorable functional profiles (based on QL2, FA, and DY) may help guide supportive interventions to improve HRQoL, functional recovery, and symptom control. Interestingly, this concept is consistent with the emerging survivorship framework recently proposed for early-stage NSCLC, in which functional recovery is viewed as a multidimensional endpoint integrating physiological reserve, symptom burden, physical activity, resilience, and long-term autonomy beyond conventional oncologic outcomes[27]. Longitudinal studies have further shown that recovery after lung cancer resection is heterogeneous across DY, physical function, pulmonary function, and QOL domains, with clinically relevant impairments persisting beyond the early postoperative period[28,29]. More recent evidence has also highlighted persistent symptom burden and rehabilitation needs during survivorship, as well as distinct patient-reported recovery trajectories after lung cancer surgery[30,31]. Within this framework, PROs may help identify vulnerable patient phenotypes and characterize different patterns of functional recovery during postoperative follow-up.
The present study has several limitations. Given the single-center design, our findings require validation in larger multicenter cohorts to improve generalizability. Inclusion in the landmark cohort required availability of the EORTC QLQ-C30 and QLQ-LC13 questionnaires at 1 year. Although most baseline clinicopathologic characteristics were comparable between included and excluded patients, differences in smoking status and chronic obstructive pulmonary disease were observed (Supplementary Table 1), and selection bias related to questionnaire availability and completion cannot be excluded. Second, with only 15 deaths observed, the survival analyses remain exploratory despite the use of Firth-penalized Cox regression to address small event numbers. Third, although the QL2 and FA thresholds were derived from previous literature, these cutoffs have not been validated specifically for postoperative outcome prediction in resected stage I NSCLC, while DY was categorized according to symptom absence or presence. Continuous sensitivity analyses only partially reproduced the categorical findings: FA and DY remained associated with recurrence, whereas QL2 did not; for OS, FA remained associated, whereas DY did not show a statistically significant linear association. These findings therefore remain exploratory and hypothesis-generating. Fourth, no formal adjustment for multiple comparisons was applied; therefore, the possibility of type I error cannot be excluded. Fifth, the landmark design selects patients surviving recurrence-free at 1 year, limiting generalizability to the full surgical population. Finally, PROs are influenced by behavioral and psychological factors, including physical activity, symptom awareness, and adaptation. Despite these limitations, this study considers a clinically relevant and underexplored area in thoracic oncology. Evidence regarding the role of postoperative PROs in predicting recurrence in stage I NSCLC remains scarce, particularly when PROs are assessed after completion of early postoperative recovery. By focusing on relapse rather than survival, this analysis presents a complementary perspective on disease dynamics during the survivorship phase.
In conclusion, PROs evaluated 1-year after surgery may provide complementary prognostic information in resected stage I NSCLC. The findings of the present study support the potential relevance of a patient-reported functional profile that reflects the patient's integrated physiological and perceptive condition. FA and DY showed clinically relevant associations with long-term survival during follow-up, although the limited number of deaths and the sensitivity analyses support cautious interpretation. The analyses presented here reinforce the growing recognition of PROs as valuable tools for understanding survivorship trajectories and highlight the need for additional research to clarify the processes linking symptom perception and long-term outcomes in early-stage lung cancer. Additional multicenter studies with larger cohorts are needed to better understand the impact of direct interventions on symptom burden, outcomes, and recurrence trajectories in resected stage I NSCLC.
We would like to thank Mrs. Chiara Banfi for her valuable assistance in the collection and management of the HRQoL questionnaires.
| 1. | Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71:209-249. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 76817] [Cited by in RCA: 71803] [Article Influence: 14360.6] [Reference Citation Analysis (115)] |
| 2. | Rami-Porta R, Nishimura KK, Giroux DJ, Detterbeck F, Cardillo G, Edwards JG, Fong KM, Giuliani M, Huang J, Kernstine KH Sr, Marom EM, Nicholson AG, Van Schil PE, Travis WD, Tsao MS, Watanabe SI, Rusch VW, Asamura H; Members of the IASLC Staging and Prognostic Factors Committee and of the Advisory Boards, and Participating Institutions. The International Association for the Study of Lung Cancer Lung Cancer Staging Project: Proposals for Revision of the TNM Stage Groups in the Forthcoming (Ninth) Edition of the TNM Classification for Lung Cancer. J Thorac Oncol. 2024;19:1007-1027. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 299] [Cited by in RCA: 297] [Article Influence: 148.5] [Reference Citation Analysis (0)] |
| 3. | Leuzzi G, Sabia F, Calderoni M, Uslenghi C, Pastorino U, Marchianò A, Ferrari M, Pardolesi A, Lorenzini D, Lo Russo G, Proto C, Prelaj A, Solli P. Leveraging Landmark Analysis for Tailored Surveillance in Stage I Non-Small-Cell Lung Cancer. Cancers (Basel). 2026;18:367. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 4. | Jovanoski N, Bowes K, Brown A, Belleli R, Di Maio D, Chadda S, Abogunrin S. Survival and quality-of-life outcomes in early-stage NSCLC patients: a literature review of real-world evidence. Lung Cancer Manag. 2023;12:LMT60. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 17] [Reference Citation Analysis (0)] |
| 5. | Movsas B, Moughan J, Sarna L, Langer C, Werner-Wasik M, Nicolaou N, Komaki R, Machtay M, Wasserman T, Bruner DW. Quality of life supersedes the classic prognosticators for long-term survival in locally advanced non-small-cell lung cancer: an analysis of RTOG 9801. J Clin Oncol. 2009;27:5816-5822. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 146] [Cited by in RCA: 153] [Article Influence: 9.0] [Reference Citation Analysis (0)] |
| 6. | Servetto A, Di Maio M, Salomone F, Napolitano F, Paratore C, Di Costanzo F, Viscardi G, Santaniello A, Formisano L, Bianco R. Analysis of phase III clinical trials in metastatic NSCLC to assess the correlation between QoL results and survival outcomes. BMC Med. 2023;21:234. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 7] [Reference Citation Analysis (0)] |
| 7. | Badaoui S, Shahnam A, McKinnon RA, Abuhelwa AY, Sorich MJ, Hopkins AM. The predictive utility of patient-reported outcomes and performance status for survival in metastatic lung cancer patients treated with chemoimmunotherapy. Transl Lung Cancer Res. 2022;11:432-439. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1] [Cited by in RCA: 27] [Article Influence: 5.4] [Reference Citation Analysis (0)] |
| 8. | Husson O, de Rooij BH, Kieffer J, Oerlemans S, Mols F, Aaronson NK, van der Graaf WTA, van de Poll-Franse LV. The EORTC QLQ-C30 Summary Score as Prognostic Factor for Survival of Patients with Cancer in the "Real-World": Results from the Population-Based PROFILES Registry. Oncologist. 2020;25:e722-e732. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 63] [Cited by in RCA: 192] [Article Influence: 27.4] [Reference Citation Analysis (0)] |
| 9. | Pompili C, Tariq J, Dalmia S, Harle A, Gilbert A, Valuckiene L, Brunelli A. Cohort study investigating evolution and factors associated with dyspnoea after anatomic lung resection. J Thorac Dis. 2024;16:113-122. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 6] [Reference Citation Analysis (0)] |
| 10. | Zhang S, Deng Y, Xiang X, Xu Q, Hu L, Xia M, Liu L. Postoperative symptom network analysis in non-small cell lung cancer patients: a cross-sectional study. BMC Pulm Med. 2025;25:244. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1] [Cited by in RCA: 4] [Article Influence: 4.0] [Reference Citation Analysis (0)] |
| 11. | Oikawa M, Hanada M, Nagura H, Takeuchi R, Miyazaki T, Doi R, Sato S, Matsumoto K, Kozu R. Association between health-related quality of life and perioperative exercise capacity in older postoperative patients with non-small cell lung cancer. J Thorac Dis. 2024;16:8300-8308. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 3] [Reference Citation Analysis (0)] |
| 12. | Cavalheri V, Granger CL. Exercise training as part of lung cancer therapy. Respirology. 2020;25 Suppl 2:80-87. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 10] [Cited by in RCA: 60] [Article Influence: 10.0] [Reference Citation Analysis (0)] |
| 13. | Sprangers MA, Schwartz CE. Integrating response shift into health-related quality of life research: a theoretical model. Soc Sci Med. 1999;48:1507-1515. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1430] [Cited by in RCA: 1496] [Article Influence: 55.4] [Reference Citation Analysis (0)] |
| 14. | Kalantar-Zadeh K, Block G, Horwich T, Fonarow GC. Reverse epidemiology of conventional cardiovascular risk factors in patients with chronic heart failure. J Am Coll Cardiol. 2004;43:1439-1444. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 460] [Cited by in RCA: 498] [Article Influence: 22.6] [Reference Citation Analysis (1)] |
| 15. | Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, Filiberti A, Flechtner H, Fleishman SB, de Haes JC. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993;85:365-376. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 12937] [Cited by in RCA: 12112] [Article Influence: 367.0] [Reference Citation Analysis (19)] |
| 16. | Bergman B, Aaronson NK, Ahmedzai S, Kaasa S, Sullivan M. The EORTC QLQ-LC13: a modular supplement to the EORTC Core Quality of Life Questionnaire (QLQ-C30) for use in lung cancer clinical trials. EORTC Study Group on Quality of Life. Eur J Cancer. 1994;30A:635-642. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 585] [Cited by in RCA: 700] [Article Influence: 21.9] [Reference Citation Analysis (0)] |
| 17. | Gray RJ. A Class of $K$-Sample Tests for Comparing the Cumulative Incidence of a Competing Risk. Ann Statist. 1988;16:1141-1154. [DOI] [Full Text] |
| 18. | Firth D. Bias reduction of maximum likelihood estimates. Biometrika. 1993;80:27-38. [DOI] [Full Text] |
| 19. | Fine JP, Gray RJ. A Proportional Hazards Model for the Subdistribution of a Competing Risk. J Am Stat Assoc. 1999;94:496-509. [RCA] [DOI] [Full Text] [Cited by in Crossref: 8085] [Cited by in RCA: 10237] [Article Influence: 379.1] [Reference Citation Analysis (0)] |
| 20. | Harrell FE Jr, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the yield of medical tests. JAMA. 1982;247:2543-2546. [PubMed] |
| 21. | Chen X, Zhuang J, Chen Z, Liu S, Xu Y, Chen C, Yi J, Yu F, Xie B, He F. The relationship between allostatic load levels and time to deterioration of health-related quality of life in non-small cell lung cancer patients. J Cancer Surviv. 2026;20:892-903. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 3] [Cited by in RCA: 4] [Article Influence: 4.0] [Reference Citation Analysis (0)] |
| 22. | Stull DE. Causal Cascade of Symptoms on Patient Functioning and Health-Related Quality of Life in Non-Small Cell Lung Cancer and Metastatic Breast Cancer. JCO Clin Cancer Inform. 2026;10:e2500127. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 23. | Presley CJ, Tang J, Benedict J, Grogan M, Reisinger S, Janse S, Wong ML, Arrato NA, Davenport A, Shields PG, Andersen BL. Functional resilience and overall survival in adults treated for advanced non-small-cell lung cancer. Lung Cancer. 2024;196:107953. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 4] [Cited by in RCA: 6] [Article Influence: 3.0] [Reference Citation Analysis (0)] |
| 24. | Murphy C, Pandya T, Swanton C, Solomon BJ. Lung Cancer in Nonsmoking Individuals: A Review. JAMA. 2025;334:1836-1845. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 29] [Reference Citation Analysis (0)] |
| 25. | Liu C, Wang H, Wen J, Xu T, Wang H, Wei X, Zhai Y. Trajectory and predictive factors of cancer-related fatigue in hospitalized elderly patients with non-small cell lung cancer undergoing chemotherapy. Medicine (Baltimore). 2026;105:e45277. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 26. | Ochi E, Fukushima T, Katsushima U, Yamashita T, Nakano J. Effects of exercise on people living with advanced lung cancer: a systematic review and meta-analysis. Support Care Cancer. 2026;34:179. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 2] [Reference Citation Analysis (0)] |
| 27. | Leuzzi G, Lococo F, Cosentino B, Sabia F, Ferrari M, Pardolesi A, Stanzi A, Brandolini J, Rolli L, Calderoni M, Uslenghi C, Solli P. Functional Recovery as a Survivorship Endpoint in Early-Stage NSCLC. Cancers (Basel). 2026;18:1958. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 28. | Heiden BT, Subramanian MP, Liu J, Keith A, Engelhardt KE, Meyers BF, Puri V, Kozower BD. Long-term patient-reported outcomes after non-small cell lung cancer resection. J Thorac Cardiovasc Surg. 2022;164:615-626.e3. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 17] [Cited by in RCA: 15] [Article Influence: 3.8] [Reference Citation Analysis (0)] |
| 29. | Shin S, Kong S, Kang D, Lee G, Cho JH, Shim YM, Cho J, Kim HK, Park HY. Longitudinal changes in pulmonary function and patient-reported outcomes after lung cancer surgery. Respir Res. 2022;23:224. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 29] [Cited by in RCA: 44] [Article Influence: 11.0] [Reference Citation Analysis (0)] |
| 30. | Kaasgaard M, Borg M, Kaltoft JP, Alstrup G, Nielsen AO, Kristiansen P, Bhatia LS, Jensen JS, Petersen RH, Jakobsen E, Hilberg O, Bodtger U. Patient-reported outcomes after surgical resection for non-small cell lung cancer and perceptions of postsurgical rehabilitation programmes. BMJ Open Respir Res. 2026;13:e003865. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 2] [Reference Citation Analysis (0)] |
| 31. | Fei S, Zhu T, Li S, Zhao H. Development and validation of a nomogram for predicting heterogeneous patient-reported outcome recovery trajectories after lung cancer surgery: a retrospective study. J Cardiothorac Surg. 2026;21:353. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |