Association of symptoms at diagnosis with prognosis in patients with surgically treated stage I non-small cell lung cancer: a retrospective cohort analysis of a prospectively maintained registry

Introduction

Lung cancer remains the most fatal malignancy worldwide, with an estimated 2.5 million new cases and 1.8 million related deaths annually, and imposes a considerable burden on public health (1-3). The National Lung Screening Trial (NLST) established low-dose computed tomography (LDCT) as an effective screening modality, reporting that early detection reduced mortality by 20% among high-risk smokers (4-7). Subsequent adoption in the US Preventive Services Task Force (USPSTF) guidelines resulted in a 3.9% annual rise in stage I diagnoses from 2014 to 2018 and thus a substantial increase in the proportion of patients with early-stage lung cancer (8-10).

In the natural progression of lung cancer, patients with early-stage disease typically exhibit limited and nonspecific symptoms, whereas those with advanced-stage disease have a significantly greater prevalence and severity of symptoms (11-13). In the study by Hu et al., approximately 59% of patients with stage I lung cancer were asymptomatic (14). In contrast, among patients with stage IV lung cancer, only 27.7% remained asymptomatic, while approximately 44% presented with multiple symptoms (11). Evidence from several retrospective analyses on stage I–IV lung cancer cohorts indicates that symptomatic patients experience significantly poorer prognostic outcomes compared with asymptomatic patients (12,13,15-17). However, the clinical relevance of symptoms in patients with resectable early-stage lung cancer remains controversial. Sheel et al. and Raz et al. reported comparable survival outcomes between asymptomatic and symptomatic cohorts following curative resection (18,19). However, in Quadrelli et al.’s study, asymptomatic patients had a higher 5-year overall survival (OS) rate compared to their symptomatic counterparts (66.2% vs. 46%; P=0.001), and subsequent stratified analysis indicated that this survival advantage only remained statistically significant for stage I disease (81.2% vs. 58.6%; P=0.01) (20). These inconsistent findings may be explained by several methodological limitations of previous studies, including heterogeneous study populations, inconsistent definitions and recording of presenting symptoms, inclusion of mixed-stage cohorts, and limited sample sizes of patients with resectable stage I disease. These limitations have made it difficult to determine whether symptom status, particularly individual respiratory symptoms, provides independent prognostic information in patients with surgically treated stage I non-small cell lung cancer (NSCLC).

To address these limitations, this study used a large cohort of patients with pathologically confirmed stage I NSCLC from a prospectively maintained institutional registry with standardized data collection and follow-up. We aimed to evaluate the association between symptom status at diagnosis and long-term survival outcomes after curative-intent resection, and to further determine whether specific presenting symptoms were independently associated with OS and disease-free survival (DFS). By focusing exclusively on surgically treated stage I disease, this study sought to clarify the prognostic value of presenting symptoms in an early-stage population and to provide clinically accessible information for postoperative risk stratification. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0733/rc).

Methods

Data collection

This retrospective cohort study used single-institution data from West China Hospital, Sichuan University, drawn from the Western China Lung Cancer Database, a prospectively maintained institutional registry. Clinical data in the database are collected from electronic medical records, operative records, pathological reports, and structured follow-up records by trained research staff. The database is maintained with standardized data-entry procedures and regular quality checks. Patient information used for research is de-identified before analysis, and access to the database is restricted to authorized study personnel. All data are stored on secure institutional servers in accordance with institutional privacy and data security regulations.

Patients were eligible for inclusion if they met the following criteria: (I) pathologically confirmed NSCLC; (II) pathological stage IA or IB disease; and (III) curative-intent surgical resection at West China Hospital, Sichuan University, between 2011 and 2021. Patients were excluded if they received non-surgical treatment as the primary treatment, underwent non-curative resection, or lacked detailed survival or treatment information. Data from the electronic medical records of the enrolled patients were collected, including basic demographic characteristics (age and gender), diagnostic information (symptom at presentation), past medical and personal history [history of diabetes, coronary artery disease, chronic obstructive pulmonary disease (COPD), family history of cancer, and history of prior surgeries], and surgical details (surgical approach, operative time, and pathological type). In asymptomatic patients, pulmonary nodules were generally identified through routine health examinations, mainly by chest computed tomography. Follow-up data included patient survival outcomes, recurrence, and the time of metastasis. Follow-up was conducted through outpatient visits, review of hospital records, and telephone interviews. Patients were generally followed up every 3 to 6 months during the first 2 years after surgery and every 6 to 12 months thereafter, according to routine institutional practice. Follow-up evaluations included clinical assessment, chest computed tomography, and other examinations when clinically indicated. Recurrence or metastasis was determined based on imaging findings, pathological confirmation when available, and clinical assessment by the treating physicians. Survival status and date of death were obtained from medical records or follow-up contact.

OS was defined as the time from the date of surgery to the date of death or the last follow-up. DFS was defined as the time from the date of surgery to the first recurrence, metastasis, or the last follow-up. For patients without recurrence or metastasis, the date of the last follow-up was considered the censoring point. Patients without detailed survival or treatment information were excluded from the study. This study was approved by the Ethics Committee of West China Hospital, Sichuan University (No. 20251751). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. As this was a retrospective study, a waiver of informed consent was granted.

Statistical analysis

In the statistical analysis, data preprocessing was performed initially to handle missing values. To avoid information loss caused by directly discarding observations with missing covariate data, missing values were imputed using the missForest algorithm. This nonparametric random forest-based method was selected because it can handle mixed continuous and categorical variables and does not require specification of a parametric distribution for the missing variables. For the comparison of baseline characteristics between groups, categorical variables were analyzed with the Chi-squared test, while continuous variables were assessed with the Wilcoxon rank-sum test. Subsequently, univariate Cox regression models were used to preliminarily identify symptoms significantly associated with OS and DFS. Symptoms with statistical significance in the univariate analysis were included in the Kaplan-Meier analysis to calculate survival rates and plot survival curves, and the log-rank test was employed to assess survival differences between groups.

Following this, symptoms with statistical significance in the univariate analysis, along with patients’ baseline characteristics, were incorporated into multivariate Cox regression models to identify independent risk factors for poor OS and DFS outcomes. Hazard ratios (HRs) and their 95% confidence intervals (CIs) were calculated to quantify the risks. To reduce multicollinearity, multivariate Cox models were constructed in two steps. The first model included the binary variable indicating the presence or absence of any symptom but excluded individual symptom variables. In the second model, only the individual symptoms with statistical significance in the univariate analysis were included, while the composite symptom variable was excluded. This two-step approach ensured stability of model estimates and clearer interpretation of symptom-related prognostic effects.

All analyses were conducted with R software version 4.4.1 (The R Foundation for Statistical Computing, Vienna, Austria), with a two-sided P value <0.05 indicating a statistically significant association between variables and survival outcomes.

Results

Study population

A total of 9,205 patients were included in this study, comprising 3,700 (40.2%) males and 5,505 (59.8%) females, with an age range of 15 to 88 years and an average age of 56.62 years. A positive smoking history was identified in 25.0% (2,305 cases) of patients, and 19.0% (1,745 cases) reported a family history of cancer. Comorbidities included COPD in 3.0% (276 cases), diabetes in 7.1% (654 cases), and coronary artery disease in 2.5% (231 cases). Among the surgical approaches, open thoracotomy was performed in 3.8% (353 cases) of patients, while 93.8% (8,631 cases) underwent video-assisted thoracoscopic surgery (VATS). The predominant pathological type was adenocarcinoma, accounting for 91.1% (8,386 cases), followed by squamous cell carcinoma in 5.5% (506 cases), and other types in 3.4% (313 cases) (Table 1). Among the 9,205 included patients, the median follow-up duration was 36 months, and 1,721 patients (18.7%) were followed up for more than 5 years.

Prevalence of symptoms at diagnosis

Among the patients, 28.9% (2,657 cases) presented with symptoms at the initial diagnosis, while 71.1% (6,548 cases) were asymptomatic. Symptomatic patients were significantly older on average (57.69 vs. 56.18 years, P<0.001) and more likely to be male (P<0.001), with a history of smoking (P<0.001), and being diagnosed with comorbid COPD (P<0.001). Open thoracotomy and squamous cell carcinoma were also more frequently observed in the symptomatic group (P<0.001). In contrast, no significant differences were found in the prevalence of diabetes (P=0.55) or coronary heart disease (P=0.08) between the two groups (Table 1). The most common initial symptom was cough, observed in 18.85% (1,735 cases) of patients, followed by chest pain in 5.33% (491 cases), blood-tinged sputum in 4.37% (402 cases), and chest tightness in 2.93% (270 cases). A notable portion of patients (n=1,969, 21.39%) presented with two or more concurrent symptoms. Other less common symptoms included back or limb pain, dizziness, fatigue, and various gastrointestinal complaints (Figure 1 and Table 2). Due to the nonspecific nature of several symptoms, such as dizziness, fatigue, and gastrointestinal discomfort, the subsequent analyses focused primarily on respiratory symptoms.

Figure 1 Distribution of presenting symptoms in patients with stage I resectable lung cancer. The bar chart illustrates the frequency of self-reported symptoms among 2,657 symptomatic patients at the time of diagnosis. The most common symptom was cough (n=1,735), followed by chest pain (n=491), blood-tinged sputum (n=402), and chest tightness (n=270). A substantial proportion of patients (n=1,969) presented with two or more concurrent symptoms. Less common symptoms included musculoskeletal pain, dizziness, fatigue, and gastrointestinal complaints.

Associations with OS

Symptomatic presentation at diagnosis was associated with a significantly worse OS in both the univariate (HR =1.71, 95% CI: 1.44–2.04; P<0.001) and multivariate Cox regression analyses (adjusted HR =1.22, 95% CI: 1.01–1.47; P=0.03), indicating that initial symptoms could independently predict poorer long-term outcomes (Figure 2A).

Figure 2 OS according to symptom presence and individual respiratory symptoms in patients with stage I resectable lung cancer. (A) Kaplan-Meier curves comparing OS between symptomatic (n=2,657) and asymptomatic (n=6,548) patients at the time of diagnosis. (B) Forest plot of univariate Cox regression for OS, with hazard ratios for individual respiratory symptoms including blood-tinged sputum, cough, chest pain, and chest tightness. (C) Forest plot of multivariate Cox regression including symptom variables and baseline covariates. (D) Kaplan-Meier curves comparing OS between patients with (n=402) and without (n=8,803) blood-tinged sputum at diagnosis. COPD, chronic obstructive pulmonary disease; OS, overall survival.

To identify which symptoms contributed to this prognostic difference, individual symptom variables were analyzed. Cox regression analysis for OS revealed that certain symptoms were significantly associated with an increased risk of mortality in univariate analysis (Table 3). Specifically, the variables significantly associated with OS were blood-tinged sputum (unadjusted HR =2.89, 95% CI: 2.28–3.66; P<0.001), the presence of two or more symptoms (unadjusted HR =1.82, 95% CI: 1.53–2.17; P<0.001), and cough (unadjusted HR =1.73, 95% CI: 1.44–2.07; P<0.001), while the associations with chest pain (P=0.13) and chest tightness (P=0.63) were not significant (Figure 2B and Figure S1). Significant variables in the univariate analysis were included in multivariate Cox regression analysis, in which blood-tinged sputum remained an independent prognostic factor for adverse OS outcomes (adjusted HR =1.71, 95% CI: 1.31–2.24; P<0.001) (Figure 2C,2D). Meanwhile, the presence of two or more symptoms (P=0.24) and cough (P=0.99) lost statistical significance after adjustments were made for baseline characteristics (Figure 2C and Table 3). In addition to symptom-related variables, several non-symptom clinicopathological factors were associated with OS in the multivariable model. Older age was associated with an increased risk of death (HR =1.05, 95% CI: 1.04–1.06; P<0.001). Smoking history (HR =1.57, 95% CI: 1.19–2.06; P=0.001), COPD (HR =1.48, 95% CI: 1.11–1.99; P=0.008), and longer duration of surgery (HR =1.002, 95% CI: 1.001–1.004; P=0.008) were also independently associated with poorer OS. Histological subtype was significantly associated with OS, whereas sex, hypertension, diabetes, coronary heart disease, family history of cancer, thoracoscopic approach, and open thoracotomy were not independently associated with OS. Detailed results of the multivariate Cox regression analysis are provided in Table S1.

The results from further analysis of subgroup survival rates (based on the presence or absence of blood-tinged sputum) at 1, 2, 3, 4, and 5 years are presented in Table 4. Among patients without blood-tinged sputum, the 1-, 2-, 3-, 4-, and 5-year survival rates were 99.3% (95% CI: 99.1–99.5%), 98.0% (95% CI: 97.6–98.3%), 96.3% (95% CI: 95.9–96.8%), 94.5% (95% CI: 93.9–95.2%), and 92.3% (95% CI: 91.4–93.1%), respectively; in comparison, the survival rates among patients with blood-tinged sputum were significantly lower, with the 1-, 2-, 3-, 4-, and 5-year survival rates being 95.5% (95% CI: 93.4–97.5%), 91.2% (95% CI: 88.4–94.0%), 87.0% (95% CI: 83.6–90.5%), 83.5% (95% CI: 79.6–87.5%), and 78.3% (95% CI: 73.7–83.1%), respectively. The presence of blood-tinged sputum was thus strongly associated with consistently lower survival rates across all follow-up periods. Furthermore, the survival gap between patients with and without blood-tinged sputum widened over time, suggesting its progressively detrimental impact on long-term survival outcomes.

Associations with DFS

The presence of symptoms at diagnosis was significantly associated with a shorter DFS in both the univariate (HR =1.77, 95% CI: 1.54–2.03; P<0.001) and multivariate analyses (adjusted HR =1.27, 95% CI: 1.09–1.47; P=0.002), indicating that initial symptoms were associated with a higher risk of recurrence or metastasis (Figure 3A).

Figure 3 DFS according to symptom presence and individual respiratory symptoms in patients with stage I resectable lung cancer. (A) Kaplan-Meier curves comparing DFS between symptomatic (n=2,657) and asymptomatic (n=6,548) patients at the time of diagnosis. (B) Forest plot of univariate Cox regression for DFS based on respiratory symptoms. (C) Forest plot of multivariate Cox regression including symptom variables and clinical covariates. (D) Kaplan-Meier curves comparing DFS between patients with (n=402) and without (n=8,803) blood-tinged sputum at diagnosis. COPD, chronic obstructive pulmonary disease; DFS, disease-free survival.

To identify the individual symptoms driving this association, we conducted symptom-level analyses. In the univariate Cox regression analysis for DFS, the symptoms identified as significant risk factors for recurrence were blood-tinged sputum (unadjusted HR =2.91, 95% CI: 2.39–3.55; P<0.001), the presence of two or more symptoms (unadjusted HR =1.80, 95% CI: 1.56–2.07; P<0.001), cough (unadjusted HR =1.75, 95% CI: 1.52–2.03; P<0.001), and chest pain (unadjusted HR =1.40, 95% CI: 1.10–1.78; P=0.006), while chest tightness (P=0.26) was not significant (Figure 3B and Figure S2). In the multivariate Cox regression analysis, after adjustments were made for potential confounding factors, blood-tinged sputum remained an independent adverse prognostic factor for DFS (adjusted HR =1.71, 95% CI: 1.36–2.14; P<0.001) (Figure 3C,3D). However, chest pain (P=0.08), the presence of two or more symptoms (P=0.41), and cough (P=0.53) did not reach statistical significance after adjustment (Figure 3C, Table 3). In addition to symptom-related variables, several non-symptom clinicopathological factors were associated with DFS in the multivariable model. Older age was associated with an increased risk of recurrence or metastasis (HR =1.04, 95% CI: 1.03–1.05; P<0.001). Smoking history (HR =1.52, 95% CI: 1.23–1.89; P<0.001) and COPD (HR =1.50, 95% CI: 1.17–1.92; P=0.001) were also independently associated with poorer DFS. Duration of surgery was statistically associated with DFS, although the effect size per minute was small. Histological subtype was significantly associated with DFS, whereas sex, hypertension, diabetes, coronary heart disease, family history of cancer, thoracoscopic approach, and open thoracotomy were not independently associated with DFS. The detailed results of the multivariate Cox regression analysis are provided in Table S2.

The results from further analysis of subgroup DFS rates at 1, 2, 3, 4, and 5 years (based on the presence or absence of blood-tinged sputum) are presented in Table 4. Among patients without blood-tinged sputum, the 1-, 2-, 3-, 4-, and 5-year DFS rates were 97.3% (95% CI: 97.0–97.7%), 94.8% (95% CI: 94.3–95.3%), 92.1% (95% CI: 91.4–92.7%), 89.9% (95% CI: 89.1–90.7%), and 88.1% (95% CI: 87.2–89.1%), respectively. In comparison, the DFS rates among patients with blood-tinged sputum were significantly lower, with the 1-, 2-, 3-, 4-, and 5-year DFS rates being 90.2% (95% CI: 87.3–93.2%), 82.7% (95% CI: 79.0–86.5%), 77.9% (95% CI: 73.8–82.3%), 72.5% (95% CI: 67.9–77.3%), and 68.9% (95% CI: 64.0–74.2%), respectively. Thus, patients with blood-tinged sputum had consistently and significantly lower DFS rates across all follow-up periods compared to those without blood-tinged sputum, suggesting its strong association with poorer survival outcomes.

Descriptive characteristics of the blood-tinged sputum subgroup

To further characterize patients with blood-tinged sputum at diagnosis, we performed a descriptive subgroup analysis. The proportion of patients with blood-tinged sputum increased with pathological stage, from 1.70% (52/3,114) in the IA1 group to 3.65% (72/1,971) in the IA2 group, reaching 6.75% (52/770) and 6.72% (225/3,350) in the IA3 and in IB groups respectively. A marked histologic gradient was also observed: blood-tinged sputum was most often present in patients with squamous cell carcinoma (116/506, 22.92%), followed by those with other histologies (47/313, 15.02%), and it was least common in patients with adenocarcinoma (239/8,386, 2.85%) (Figure 4).

Figure 4 Distribution of blood-tinged sputum across pathological stage and histologic subtype in patients with stage I resectable lung cancer. (A) Proportion of blood-tinged sputum across pathological stages: IA1, 1.70% (52/3,114); IA2, 3.65% (72/1,971); IA3, 6.75% (52/770); and IB, 6.72% (225/3,350). (B) Proportion of blood-tinged sputum across histologic subtypes: squamous cell carcinoma, 22.92% (116/506); other histologies, 15.02% (47/313); and adenocarcinoma, 2.85% (239/8,386).

Compared with the non-blood-tinged sputum group (n=8,803), the blood-tinged sputum group (n=402) had higher age (57.9 vs. 56.6 years; P=0.02), larger tumors (2.33 vs. 1.65, P<0.001), a longer operative time (median 130 vs. 100 minutes), and a lower thoracoscopic resection rate (77.4% vs. 94.5%), along with a greater proportion of patients who were male (59.2% vs. 39.3%) and had a history of smoking (51.7% vs. 23.8%), COPD (6.0% vs. 2.9%), or open thoracotomy (20.9% vs. 3.1%) (all P values <0.001). Family history of cancer was less common in the blood-tinged sputum group (14.4% vs. 19.2%, P=0.02), while diabetes and coronary heart disease did not differ significantly between the groups (P=0.85 and 0.85 respectively) (Table 5 and Figure S3). Subgroup survival analyses further supported blood-tinged sputum as being a risk factor across key clinical strata (Figures S4,S5).

Discussion

With the expansion of lung cancer screening programs and improved healthcare accessibility, the proportion of patients being diagnosed with early-stage lung cancer has increased (8,9). Although clinical symptoms remain a primary trigger for medical consultation (21), their prognostic value in early-stage disease remains controversial (18,19). In this study of curatively resected stage I lung cancer, we found that the presence of symptoms at diagnosis was significantly associated with a poorer prognosis. Moreover, symptomatic patients were more likely to be older, to be male, to have a history of smoking, and to present with comorbid COPD or squamous cell carcinoma histology. Multivariate analysis specifically identified blood-tinged sputum as the most significant independent predictor of adverse outcomes.

In this cohort, 28.9% (n=2,657) of patients with stage I resectable lung cancer were symptomatic at diagnosis. This prevalence is lower than the 40–50% reported in historical series (14,17,19,20) but aligns with the 31.4% documented by Quadrelli et al. (20). These inconsistencies may reflect the combined impact of expanded lung cancer screening initiatives and increased utilization of thoracic computed tomography during the coronavirus disease 2019 (COVID-19) pandemic, which have facilitated the incidental detection of early-stage asymptomatic malignancies. Beyond this, the higher proportion of adenocarcinoma histology in our cohort may further contribute to the observed low symptomatic rate, as adenocarcinomas—particularly peripheral small nodules—are less frequently symptomatic than the centrally located squamous cell carcinomas. This histology-dependent symptom pattern has been substantiated in several independent studies (14,16,19).

In our analysis, symptomatic presentation at diagnosis was significantly associated with older age, smoking history, male sex, comorbid COPD, and squamous cell carcinoma histology among patients with resectable stage I lung cancer. This risk profile aligns with observations from previous studies on surgical cohorts (14). Such pathophysiological concordance can likely be attributed to certain synergistic mechanisms: patients exhibiting these characteristics typically present with diminished baseline pulmonary functional reserve, lowering the threshold for respiratory symptom manifestation during early tumorigenesis. Concurrently, the predominantly central airway location of squamous cell carcinomas accelerates symptom onset through mechanical bronchial obstruction or vascular erosion (16).

Our finding that presenting symptoms at diagnosis was associated with poorer survival is in line with the analysis of a surgical cohort by Quadrelli et al. but conflicts with the report by Sheel et al., who found that symptoms had no prognostic relevance for resectable disease (18,20). The observed discrepancy is likely due to the heterogeneity in disease stage distribution of the cohorts. Quadrelli et al. used stage-stratified analysis and found that symptoms at diagnosis were significantly associated with survival, specifically among patients with pathological stage I disease, suggesting that symptomatic tumors detected at this earliest stage may harbor a biologically aggressive phenotype. In contrast, among patients with advanced-stage disease, symptoms predominantly reflect direct tumor burden or metastatic complications—including chest pain, weight loss, and neurological deficits—which consistently correlate with poorer survival across studies (13,16,17,22). In operable stage I disease, however, nonspecific manifestations such as cough or sputum, frequently attributable to comorbid cardiopulmonary pathology, have limited inherent prognostic value for patients with lung cancer. Our study represents the largest surgical series exclusively evaluating patients with stage I disease, and the findings from the cohort constitute robust statistical evidence confirming the prognostic significance of symptomatic presentation in this early-stage population.

We evaluated common respiratory symptoms in patients with curatively resected stage I lung cancer in terms of the prognostic value. In the analysis, blood-tinged sputum was the only respiratory symptom independently predictive of adverse oncologic outcomes. Although less prevalent than nonspecific symptoms such as cough, blood-tinged sputum exhibited superior diagnostic specificity for lung cancer, consistent with established epidemiological evidence (23-25). Crucially, while prior research confirmed the prognostic value of blood-tinged sputum in advanced disease (12,26,27), this study provides the first evidence of its association with adverse outcomes in patients with stage I disease following curative-intent resection. This relationship can likely be explained by the direct neoplastic erosion into pulmonary vasculature, which simultaneously causes hemorrhage and facilitates early hematogenous dissemination (28). Consequently, patients with stage I disease presenting with blood-tinged sputum should undergo intensified surveillance for metastatic recurrence. Symptom clusters, including weight loss, pain, and dyspnea, may be present in advanced malignancies and collectively portend poor survival (29), yet they are rare in early-stage disease and thus fall beyond the scope of this stage-specific investigation.

The different prognostic patterns of cough and blood-tinged sputum may be explained by their different clinical specificity. Cough is a common and nonspecific respiratory symptom that may be related to smoking history, COPD, airway inflammation, or other nonmalignant conditions. Therefore, its association with survival was attenuated after adjustment for baseline clinical characteristics. In contrast, blood-tinged sputum is more directly related to tumor-airway or tumor-vascular interaction and may indicate a higher-risk tumor phenotype even in pathologically stage I disease. This interpretation is supported by our finding that patients with blood-tinged sputum had larger tumors and a higher proportion of pathological IB disease. Clinically, blood-tinged sputum may serve as a simple warning sign for postoperative risk stratification and may help identify patients who require closer imaging surveillance after curative resection.

Our descriptive analysis of subgroups indicated that the prevalence of blood-tinged sputum increased as pathological stage increased (from IA1 to IA3/IB). It was significantly more common in the squamous cell carcinoma group than in the adenocarcinoma group, and patients with blood-tinged sputum also had a higher clinical risk and a more adverse perioperative profile. This pattern relating stage, histology, and clinical outcomes suggests that blood-tinged sputum may be indicative of a biologically and anatomically more aggressive phenotype in stage I disease, which is consistent with its associations with adverse OS and DFS. This pattern also supports its use for simple risk stratification among patients with resectable stage I lung cancer.

Several limitations to this study should be acknowledged. First, although we used data from a prospectively maintained follow-up cohort, the retrospective study design might have introduced residual confounding and limited the generalizability of our conclusions. In addition, patients lacking detailed survival or treatment information were excluded from the analytic cohort. Because complete baseline data for these excluded patients were not available in the final analytic dataset, we could not formally compare their characteristics with those of the included cohort. This may have introduced selection or attrition bias, and the direction of this potential bias could not be fully determined. Another limitation is that tumor size and detailed surgical resection type, such as lobectomy versus segmentectomy, were not included in the multivariable Cox models. Therefore, residual confounding by tumor burden, pathological T category, and surgical management cannot be fully excluded, and blood-tinged sputum should be interpreted as a marker of a higher-risk clinicopathological phenotype. Furthermore, symptom data were recorded in binary form, precluding analyses of symptom duration and severity. Finally, the presence of blood-tinged sputum was chiefly determined according to initial history-taking records. However, in some patients, it was identified only after further questioning and not as the chief complaint, and some episodes might have been the result of nontumor airway or upper-airway mucosal bleeding. These factors might have caused symptom misclassification and biased the effect estimates. Prospective studies with standardized symptom definitions are needed to validate these associations.

Conclusions

In this study, we found that the presence of symptoms at diagnosis—particularly blood-tinged sputum—was independently associated with a significantly worse OS and DFS. These findings suggest that symptomatic presentation, even in patients with early-stage disease, may reflect a more aggressive tumor phenotype or unfavorable anatomical characteristics. Blood-tinged sputum, as an individual symptom, warrants closer clinical attention and may serve as a valuable prognostic indicator in risk stratification and postoperative surveillance strategies.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0733/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0733/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0733/prf

Funding: The work received funding from the Key R&D Project of Sichuan Provincial Department of Science and Technology (Nos. 2024YFFK0235 and 2024YFFK0213).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0733/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of West China Hospital, Sichuan University (No. 20251751). As this was a retrospective study, a waiver of informed consent was granted.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
3. Leiter A, Veluswamy RR, Wisnivesky JP. The global burden of lung cancer: current status and future trends. Nat Rev Clin Oncol 2023;20:624-39. [Crossref] [PubMed]
4. National Lung Screening Trial Research Team. Lung Cancer Incidence and Mortality with Extended Follow-up in the National Lung Screening Trial. J Thorac Oncol 2019;14:1732-42. [Crossref] [PubMed]
5. de Koning HJ, van der Aalst CM, de Jong PA, et al. Reduced Lung-Cancer Mortality with Volume CT Screening in a Randomized Trial. N Engl J Med 2020;382:503-13. [Crossref] [PubMed]
6. National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365:395-409. [Crossref] [PubMed]
7. Pastorino U, Silva M, Sestini S, et al. Prolonged lung cancer screening reduced 10-year mortality in the MILD trial: new confirmation of lung cancer screening efficacy. Ann Oncol 2019;30:1162-9. [Crossref] [PubMed]
8. Potter AL, Rosenstein AL, Kiang MV, et al. Association of computed tomography screening with lung cancer stage shift and survival in the United States: quasi-experimental study. BMJ 2022;376:e069008. [Crossref] [PubMed]
9. US Preventive Services Task Force. Screening for Lung Cancer: US Preventive Services Task Force Recommendation Statement. JAMA 2021;325:962-70. [Crossref] [PubMed]
10. Li J, Xu HL, Li WX, et al. Prognostic factors of survival in patients with lung cancer after low-dose computed tomography screening: a multivariate analysis of a lung cancer screening cohort in China. BMC Cancer 2025;25:646. [Crossref] [PubMed]
11. Ruano-Raviña A, Provencio M, Calvo de Juan V, et al. Lung cancer symptoms at diagnosis: results of a nationwide registry study. ESMO Open 2020;5:e001021. [Crossref] [PubMed]
12. Athey VL, Walters SJ, Rogers TK. Symptoms at lung cancer diagnosis are associated with major differences in prognosis. Thorax 2018;73:1177-81. [Crossref] [PubMed]
13. In KH, Kwon YS, Oh IJ, et al. Lung cancer patients who are asymptomatic at diagnosis show favorable prognosis: a korean Lung Cancer Registry Study. Lung Cancer 2009;64:232-7. [Crossref] [PubMed]
14. Hu XL, Xu ST, Wang XC, et al. Prevalence of and risk factors for presenting initial respiratory symptoms in patients undergoing surgery for lung cancer. J Cancer 2018;9:3515-21. [Crossref] [PubMed]
15. Kim HC, Jung CY, Cho DG, et al. Clinical Characteristics and Prognostic Factors of Lung Cancer in Korea: A Pilot Study of Data from the Korean Nationwide Lung Cancer Registry. Tuberc Respir Dis (Seoul) 2019;82:118-25. [Crossref] [PubMed]
16. Kocher F, Lunger F, Seeber A, et al. Incidental Diagnosis of Asymptomatic Non-Small-Cell Lung Cancer: A Registry-Based Analysis. Clin Lung Cancer 2016;17:62-7.e1. [Crossref] [PubMed]
17. Polanco D, Pinilla L, Gracia-Lavedan E, et al. Prognostic value of symptoms at lung cancer diagnosis: a three-year observational study. J Thorac Dis 2021;13:1485-94. [Crossref] [PubMed]
18. Sheel AR, McShane J, Poullis MP. Survival of patients with or without symptoms undergoing potentially curative resections for primary lung cancer. Ann Thorac Surg 2013;95:276-84. [Crossref] [PubMed]
19. Raz DJ, Glidden DV, Odisho AY, et al. Clinical characteristics and survival of patients with surgically resected, incidentally detected lung cancer. J Thorac Oncol 2007;2:125-30. [Crossref] [PubMed]
20. Quadrelli S, Lyons G, Colt H, et al. Clinical characteristics and prognosis of incidentally detected lung cancers. Int J Surg Oncol 2015;2015:287604. [Crossref] [PubMed]
21. Friedemann Smith C, Whitaker KL, Winstanley K, et al. Smokers are less likely than non-smokers to seek help for a lung cancer ‘alarm’ symptom. Thorax 2016;71:659-61. [Crossref] [PubMed]
22. Renaud S, Casabianca P, Diez-Andreu P, et al. Epidemiology, patient management, and survival outcomes in resected patients with non-metastatic non-small cell lung cancer: a nationwide real-world study. BMC Cancer 2025;25:966. [Crossref] [PubMed]
23. Shim J, Brindle L, Simon M, et al. A systematic review of symptomatic diagnosis of lung cancer. Fam Pract 2014;31:137-48. [Crossref] [PubMed]
24. Shapley M, Mansell G, Jordan JL, et al. Positive predictive values of ≥5% in primary care for cancer: systematic review. Br J Gen Pract 2010;60:e366-77. [Crossref] [PubMed]
25. Okoli GN, Kostopoulou O, Delaney BC. Is symptom-based diagnosis of lung cancer possible? A systematic review and meta-analysis of symptomatic lung cancer prior to diagnosis for comparison with real-time data from routine general practice. PLoS One 2018;13:e0207686.
26. Fukushima T, Tateishi K, Hanaoka M, et al. Clinical outcomes in patients with small cell lung cancer in a single institute: Comparative analysis of radiographic screening with symptom-prompted patients. Lung Cancer 2015;88:48-51. [Crossref] [PubMed]
27. Dregan A, Møller H, Charlton J, et al. Are alarm symptoms predictive of cancer survival?: population-based cohort study. Br J Gen Pract 2013;63:e807-12. [Crossref] [PubMed]
28. Salvayre R, Hanotin C, Parrot A, et al. Severe hemoptysis associated with lung cancer in the ICU: recurrence and outcome. Ann Intensive Care 2025;15:33. [Crossref] [PubMed]
29. Walter FM, Rubin G, Bankhead C, et al. Symptoms and other factors associated with time to diagnosis and stage of lung cancer: a prospective cohort study. Br J Cancer 2015;112:S6-13. [Crossref] [PubMed]

(English Language Editor: J. Gray)