Prognostic factors in early-stage lung adenocarcinoma: a multicenter study prognostic markers in early lung adenocarcinoma

Abstract

Background

Lung adenocarcinoma is the most common subtype of non-small cell lung cancer (NSCLC) and has shown increasing incidence worldwide. Although early-stage lung adenocarcinoma generally has a better prognosis, recurrence occurs in 30–50% of cases. Surgical resection remains the gold standard treatment for early-stage disease. Identifying prognostic factors is essential for guiding postoperative follow-up and determining the need for adjuvant therapy.

Materials and methods

We retrospectively evaluated 1,057 patients who underwent surgery for early-stage lung adenocarcinoma between 2007 and 2020 at two thoracic surgery centers. Descriptive statistics, including means and standard deviations, were calculated for continuous variables. Survival analysis was performed using Kaplan–Meier estimates, and prognostic factors were assessed through Cox proportional hazards regression.

Results

The mean age was 61.2 ± 9.0 years (range: 23–87), with 284 females (26.9%) and 773 males (73.1%). Female patients demonstrated significantly better survival (p < 0.001). Poor survival outcomes were associated with pathological N1 involvement, as well as lymphatic, perineural, vascular, and pleural invasion (p ≤ 0.008 for all). Multivariate analysis identified male sex (p < 0.001, HR: 1.53) as a poor prognostic factor, and perineural invasion (p = 0.005, HR: 0.71), and absence of vascular invasion (p = 0.047, HR: 0.80) as independent predictors of good prognosis.

Conclusion

Male sex, perineural invasion, and vascular invasion were found to be independent poor prognostic factors in early-stage lung adenocarcinoma. Patients with these features should be monitored more closely and considered for adjuvant treatment strategies.

Background

Lung cancer remains the leading cause of cancer-related death and is the second most common malignancy worldwide [1]. Non-small cell lung cancer (NSCLC) accounts for the majority of lung cancer cases, with adenocarcinoma being the most prevalent histologic subtype [2]. Over recent decades, the incidence of lung adenocarcinoma has steadily increased [2].

Lung cancer staging relies on the TNM system, which provides an anatomical framework for tumor assessment [3–5]. Lung adenocarcinomas comprise a broad spectrum ranging from preinvasive lesions to invasive forms [6]. With the widespread use of computed tomography (CT) and the implementation of lung cancer screening programs, early-stage lung adenocarcinomas are being detected more frequently in clinical practice [6–8].

In the 9th edition of the TNM staging system published by the International Association for the Study of Lung Cancer (IASLC) in 2024, the prognostic significance of tumor spread through air spaces (STAS) and nodal involvement (N status) remains controversial [3]. Accurate prognostic estimation is essential to guide follow-up strategies and adjuvant treatment decisions [9, 10]. Several pathological features—such as solid and cribriform growth patterns, N1 disease, STAS, pleural, lymphatic, and vascular invasion, and tumor size—have been associated with worse prognosis and earlier recurrence [9–16].

Although TNM staging remains the cornerstone of lung cancer classification, it reflects only anatomical disease extent. Patients with identical stages and similar demographics may have markedly different survival outcomes [9–16]. Recognizing this limitation, the IASLC has established working groups to integrate additional prognostic markers beyond anatomical staging.

The aim of this study is to identify clinical and pathological prognostic factors associated with survival in patients undergoing surgery for early-stage lung adenocarcinoma.

Materials and methods

This retrospective, multicenter study was conducted at two thoracic surgery centers and included patients who underwent surgical resection for lung adenocarcinoma between 2007 and 2020. Ethics committee approval and institutional permissions were obtained prior to the study (Approval No: E-83045809-604.01.01-712053, dated 15.06.2023).

Study population

Patients who underwent surgery for non-small cell lung carcinoma were initially screened. Those with non-adenocarcinoma histology, pathological stage III–IV, history of neoadjuvant therapy, synchronous tumors, or age under 18 years were excluded. A flow diagram outlining inclusion and exclusion criteria is shown in Fig. 1.

Fig. 1.

Flow diagram showing inclusion and exclusion criteria

Data collection and variables

The following clinical and pathological variables were evaluated: patient demographics, resection type (lobectomy, pneumonectomy, sublobar resection), T and TNM stages, presence of pleural, perineural, lymphatic, and vascular invasion, STAS positivity, N status, and overall survival.

Preoperative evaluation

All patients underwent contrast-enhanced thoracic computed tomography (CT). For tumors ≥ 1 cm, positron emission tomography (PET/CT) and cranial magnetic resonance imaging (MRI) were performed to evaluate for distant metastases. Invasive staging methods, such as endobronchial ultrasound (EBUS) or mediastinoscopy, were utilized for patients with mediastinal lymph nodes ≥ 1 cm or PET-positive hilar/mediastinal nodes, and in adenocarcinoma cases with tumor diameter ≥ 3 cm.

Pulmonary function and diffusion capacity for carbon monoxide (DLCO) were assessed preoperatively. Cardiac evaluation included electrocardiogram (ECG) for all patients and echocardiography as needed. Cardiology consultation was obtained for patients aged ≥ 60 years or with a known cardiac history. The modified Charlson Comorbidity Index (CCI) was used to assess comorbidities.

Surgical decisions were made through multidisciplinary tumor board discussions. Systematic mediastinal lymph node dissection, involving at least three nodal stations, was routinely performed. Right-sided resections included paratracheal and subcarinal nodes, while left-sided resections included subaortic, paraaortic, subcarinal, and other regional lymph nodes.

Postoperative follow-up

Postoperative data included demographic characteristics, histopathologic features, and five-year survival. Information on age, tumor histology, pathological stage, adjuvant treatment, and survival outcomes were obtained from institutional records and the national cancer registry. Pathological staging was performed according to the 9th edition of the TNM classification system [3]. Pathological assessments were based on the system reports from pathologists at the two institutions, and no re-evaluation was performed. Patients with missing pathological data were excluded from the study. The pathology data included the prognostic factors planned for evaluation in the study at both institutions.

Patients were followed with thoracic CT and physical examinations in collaboration with oncology: every 3 months during the first 2 years, every 6 months between years 2 and 5, and annually thereafter. The mean follow-up duration was 9.0 ± 1.0 years. Adjuvant chemotherapy was administered based on postoperative oncologic assessment. Due to the retrospective nature of the study and lack of universal insurance coverage, molecular testing was not available for all patients.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics software (version 22 and 27; IBM Corp., Armonk, NY, USA). Descriptive statistics included means, standard deviations, and ranges for continuous variables, and frequencies and percentages for categorical variables. Chi-square or Fisher’s exact tests were used to assess relationships between categorical variables. Continuous variables were compared using Student’s t-test or Mann–Whitney U test, as appropriate.

Survival analysis was performed using the Kaplan-Meier method, and comparisons were made with the log-rank test. Multivariate survival analysis was conducted using the Cox proportional hazards regression model. A p-value < 0.05 was considered statistically significant.

Results

A total of 1057 patients were included in the study, with a mean age of 61.2 ± 9.0 years (range: 23–87). Among them, 284 (26.9%) were female and 773 (73.1%) were male (Table 1). The mean tumor diameter was 3.0 ± 1.5 cm. Lobectomy was performed in 929 patients (87.9%), sublobar resection in 83 patients (7.9%), and pneumonectomy in 45 patients (4.2%). Right-sided tumors were present in 589 patients (55.7%), while 468 patients (44.3%) had left-sided tumors.

Table 1.

Demographic data

Gender
-Female  -Male	284 (26.9%) 773 (73.1%)
Age	61.2 ± 9.0 (23–87)

Pathological N0 status was observed in 927 patients (87.7%) and N1 in 130 patients (11.3%). Among the 130 patients with pathological N1 status, 12 (9.2%) were clinically staged as N0 but pathologically staged as N1 and were considered as upstaged. The distribution of pathological stages was as follows: 81 patients (7.6%) in stage IA1, 213 (20.2%) in stage IA2, 165 (15.6%) in stage IA3, 235 (22.2%) in stage IB, 98 (9.3%) in stage IIA, and 265 (25.1%) in stage IIB.

Pleural invasion status was classified as PL0 in 729 patients (69.0%), PL1 in 234 (22.1%), PL2 in 40 (3.8%), and PL3 in 54 (5.1%). Perineural invasion (PNI) was present in 204 patients (19.3%), lymphatic invasion in 445 (42.1%), and vascular invasion in 359 patients (34.0%). Among the 587 patients evaluated for spread through air spaces (STAS), 192 (32.7%) were positive.

Survival analysis revealed that female patients had significantly better overall survival compared to males (p < 0.001) (Fig. 2). Worse survival outcomes were significantly associated with pathological N1 status, lymphatic invasion, perineural invasion, vascular invasion, and pleural invasion (p = 0.001, p < 0.001, p < 0.001, p = 0.008) (Figs. 3, 4, 5, 6 and 7). No significant difference in survival was observed with respect to the type of surgical resection (lobar vs. sublobar) or STAS status (p = 0.584, p = 0.849) (Table 2) (Figs. 8 and 9).

Fig. 2.

Survival graph showing the effect of gender on survival

Fig. 3.

Survival graph showing the effect of pathological N status on survival

Fig. 4.

Survival graph showing the effect of lymphatic invasion on survival

Fig. 5.

Survival graph showing the effect of perineural invasion on survival

Fig. 6.

Survival graph showing the effect of vascular invasion on survival

Fig. 7.

Survival graph showing the effect of pleural invasion on survival

Table 2.

10-year survival, median survival, and p-values in univariate survival analysis

		10-Year survival rate (%)	Median survival (months)	95% CI	p Value
Gender	Female	70.1	133.4 ± 5.2	123.3-143.5	< 0.001
	Male	52.5	113.9 ± 3.8	106.6-121.3
N Status	N0	58.9	125.0 ± 3.5	118.1-131.9	0.001
	N1	45.4	89.4 ± 6.5	76.5-102.2
Resection	Lobectomy	57.2	124.1 ± 3.4	117.5-130.7	0.584
	Pneumonectomy	48.9	109.7 ± 12.2	85.8-133.6
	Sublobar Resection	62.7	102.9 ± 10.7	82.0-123.8
Lymphatic Invasion	None	63.6	131.5 ± 4.3	123.0-140.0	< 0.001
	Yes	48.5	110.4 ± 4.7	101.1-119.7
Perineural Invasion	None	59.0	126.7 ± 3.6	119.7-133.8	< 0.001
	Yes	50.0	98.5 ± 8.2	82.4-114.6
Vascular Invasion	None	62.0	130.5 ± 4.1	122.5-138.5	< 0.001
	Yes	47.9	10.2.8 ± 5.3	92.5-113.1
Pleural Invasion	PL0	57.5	122.7 ± 4.0	114.9-130.5	0.008
	PL1	63.7	108.8 ± 4.8	99.3-118.2
	PL2	45.0	98.9 ± 13.4	72.6-125.1
	PL3	35.2	91.5 ± 11.8	68.3-114.7
STAS	None	66.8	95.5 ± 3.3	89.0-102.0	0.849
	Yes	71.4	96.5 ± 4.4	88.0-105.0
pT Stage	pT1a	73.3	129.3 ± 9.3	111.1-147.5	0.008
	pT1b	61.1	124.6 ± 6.4	112.1–137.0
	pT1c	57.4	127.3 ± 7.1	113.4-141.3
	pT2a	58.1	110.5 ± 5.9	99.0-122.1
	pT2b	48.8	115.8 ± 8.3	99.6-131.9
	pT3	45.4	101.6 ± 8.2	85.6-117.6
TNM Stage	1A1	75.3	130.9 ± 9.9	111.6-150.3	< 0.001
	1A2	62.9	127.4 ± 6.8	114.1-140.6
	1A3	57.0	125.9 ± 7.9	110.3-141.4
	1B	61.7	117.3 ± 6.3	105.0-129.6
	2 A	51.0	119.4 ± 9.2	101.3-137.5
	2B	45.7	102.1 ± 5.7	90.9-113.3
Overall		57.2	122.1 ± 3.3	115.6-128.6

Fig. 8.

Survival graph showing the effect of resection types on survival

Fig. 9.

Survival graph showing the effect of STAS status on survival

Significant differences in survival were also found between different T stages and overall TNM stages (p = 0.008, p < 0.001) (Table 2; Figs. 10 and 11).

Fig. 10.

Survival graph showing the effect of T stage on survival

Fig. 11.

Survival graph showing the effect of TNM stage on survival

Multivariate Cox regression analysis identified male gender as an independent poor prognostic factor with a hazard ratio (HR) of 1.53 (p < 0.001, 95% CI: 1.20–1.94). Additionally, the absence of perineural invasion (HR: 0.71, p = 0.005, 95% CI: 0.55–0.90) and vascular invasion (HR: 0.80, p = 0.047, 95% CI: 0.65–0.99) were independently associated with better survival (Table 3).

Table 3.

Multivariate survival analysis results

	p value	Hazard ratio	95% CI
Gender
Female  Male	< 0.001	1 1.53	1.20–1.94
Perineural Invasion
Yes  No	0.005	1 0.71	0.55–0.90
Vascular Invasion
Yes  No	0.047	1 0.80	0.65–0.99

Discussion

In our study, male gender, perineural invasion, and vascular invasion were identified as independent poor prognostic factors in patients undergoing surgery for early-stage lung adenocarcinoma. Consistent with our findings, Behrens et al. demonstrated a correlation between female gender and increased tumor-associated immune cell infiltration in lung adenocarcinomas, which is associated with better prognosis [17]. Several studies in the literature also support female gender as a favorable prognostic factor in lung adenocarcinoma [18, 19].

The tumor microenvironment plays a crucial role in disease progression and prognosis prediction in lung adenocarcinoma [20, 21]. Perineural and vascular invasions are markers of aggressive tumor behavior within the microenvironment and correlate with poor survival outcomes [20, 21]. PNI is defined as the invasion of tumor cells into the nerve sheath or involvement of more than one-third of the nerve circumference [22, 23]. In the literature, PNI has been associated with aggressive tumor behavior and has been linked to recurrence, poor prognosis, and extrathoracic metastases [22, 23]. Liu et al. further demonstrated in meta-analyses that PNI is an independent poor prognostic factor in NSCLC [23].

In our study, PNI was also identified as an independent poor prognostic factor in patients with early-stage lung adenocarcinoma. This finding suggests that the identification of additional adverse prognostic factors, such as PNI, even in early-stage disease, may inform updates to adjuvant therapy decisions and follow-up protocols in the future. Moreover, prognostic factors that are not yet fully incorporated into TNM staging may potentially be integrated into the 10th edition of TNM with accumulating evidence.

Vascular invasion is another key factor influencing tumor microenvironment and aggressive behavior and has been associated with poor prognosis in early-stage NSCLC [24]. Locke et al. demonstrated the negative prognostic impact of vascular invasion in stage I NSCLC and the potential contribution of adjuvant therapy to survival [24]. Consistently, our study identified vascular invasion as a poor prognostic factor, suggesting that even in early-stage disease, the presence of such factors may guide the development of future adjuvant treatment and follow-up strategies.

Although clinical staging may categorize patients as early-stage, pathological upstaging to N1 or N2 status can occur, emphasizing the importance of thorough mediastinal lymph node sampling for accurate staging and management [25].

In our study, while pathological N1 status, pleural invasion, and lymphatic invasion were significantly associated with worse survival in univariate analysis, they did not retain independent prognostic significance in multivariate analysis. This suggests complex interactions between T stage, N stage, and tumor microenvironment factors [12–16]. Therefore, prognostic assessment in early-stage lung adenocarcinoma should not rely solely on TNM staging but also incorporate additional pathological and microenvironmental factors.

According to the literature, patients classified as high-risk NSCLC due to the presence of various prognostic factors are associated with higher recurrence rates [26–29]. Features such as visceral pleural invasion, lymphovascular invasion, STAS positivity, and larger tumor size are considered markers of high-risk NSCLC [26–29]. Once long-term follow-up is completed, our study is planned to evaluate this patient group in terms of recurrence outcomes.

The identification of prognostic factors helps explain why patients within the same stage may have different survival outcomes and highlights the importance of tumor behavior patterns and microenvironment interactions [30]. These prognostic factors can guide individualized patient follow-up and adjuvant treatment strategies. However, instead of considering a single prognostic factor, evaluating multiple effective factors together and creating a scoring system may be more beneficial [30]. Therefore, we propose that in addition to TNM staging, additional prognostic factors should be incorporated. Artificial intelligence assisted models for identifying prognostic factors can support clinicians in staging, surgical decision-making, and adjuvant therapy selection [31, 32]. The integration of these approaches into future clinical practice will be valuable for clinicians [31, 32].

Identifying these prognostic factors helps explain survival differences among patients within the same stage and guides personalized follow-up and adjuvant treatment strategies. Thus, we recommend integrating both TNM staging and additional prognostic markers in clinical decision-making.

The retrospective design limited our ability to evaluate adenocarcinoma subtypes comprehensively. Additionally, since STAS evaluation has only recently been incorporated into routine pathology reports, the available data on STAS were limited and may not fully reflect its prognostic value. Additionally, molecular biomarkers could not be analyzed in our study. In the future, incorporating and evaluating molecular biomarkers such as EGFR, ALK, and PD-L1 for subgroup-specific prognostic molecular profiling will be important. In this context, our study could be further strengthened using a prospective or international database.

Conclusions

Our findings emphasize that male gender and the presence of perineural and vascular invasion are significant indicators of poor prognosis in early-stage lung adenocarcinoma. These patients may benefit from closer surveillance and consideration of adjuvant systemic therapies to improve outcomes.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Gizem Özçıbık Işık, Dilekhan Kizir, Esra Yamansavcı Şirzai and Akif Turna. The first draft of the manuscript was written by Gizem Özçıbık Işık, Celal Buğra Sezen, Esra Yamansavcı Şirzai, Özkan Saydam and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

The authors declare that no funds, grants, or other support was received during the preparation of this manuscript.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by The Istanbul University and Yedikule Chest Disease and Thoracic Surgery Education and Research Hospital. All procedures performed in studies involving human participants were in accordance with the 1964 Helsinki Declaration and later versions.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.