Introduction

Lung cancer is one of the most prevalent cancers globally, with high death rates in both genders. The majority of lung cancers are attributed to non-small cell lung cancer (NSCLC), causing the most cancer-related deaths and ranking as the second most prevalent cancer globally [1,2,3]. In recent years, despite great progress in multidisciplinary treatment including surgery, radiotherapy, chemotherapy and immunotherapy, the prognosis of patients with NSCLC remains unsatisfactory. Hence, it is crucial to identify significant prognostic biomarkers for NSCLC to improve the clinical management of patients.

Tertiary lymphoid structure (TLS) is an abnormal lymphoid organ that closely resembles the secondary lymphoid organ (SLO) [4]. Under normal circumstances, TLS does not typically occur in the body, instead, it is found in non-lymphoid tissues where chronic inflammation is present [5]. TLS can develop in different pathophysiological conditions such as autoimmune diseases, infectious diseases, and tumors, leading to various effects that are influenced by the environment [6]. Recently, high levels of TLS have been proven to be linked to improved prognosis in various types of cancer [7, 8]. In patients with breast cancer, elevated levels of TLS is strongly related to a positive outlook for their prognosis [9]. In cases of gastrointestinal tumors, TLS can serve as a valuable prognostic indicator for gastrointestinal cancer and help direct the use of cancer immunotherapy [Data extraction and quality assessment

From each of the studies that were included, we extracted the following information: authors, year of publication, country, study design, sample size, treatment, TLS detection methods, TLS location, cut-off criteria of TLS, follow-up time and survival outcomes. In addition, we collected the association of TLS with the age, gender, pathologic staging, smoking, Tumor Node Metastasis (TNM) staging and neutrophil-lymphocyte ratio (NLR) for patients. In the end, we extracted the hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) for overall survival (OS), disease-free survival (DFS)/recurrence-free survival (RFS), disease-specific survival (DSS) from each study. If a study conducts both univariate and multivariate analysis of variance, the results of the multivariate analysis will be used in further meta-analysis.

The quality of the included studies was evaluated using the Newcastle-Ottawa Quality Assessment Scale (NOS) [16]. Studies with scores equal to or higher than 6 points can be used for further meta-analysis. Two authors (Luyuan Ma and Rongyang Li) independently appraised the quality of each study, and all disagreements were resolved by consulting other researchers.

Statistical analysis

The prognostic significance of TLS in patients with NSCLC was evaluated by calculating the aggregated HRs and 95% CIs, and the association between TLS and clinicopathological features in patients was evaluated by the aggregated odds ratio (ORs) and 95% CIs. In cases where studies displayed Kaplan-Meier curves but did not provide HRs or 95% CIs, we determined the HRs and 95% CIs by analyzing the survival curves with Engauge Digitizer V4.1 (Markmitch, Goteborg, Sweden) [17]. To reduce possible bias, a random effects model was used to calculate the overall effect size. The degree of heterogeneity was measured using the Cochrane Q test and I² statistics, where I² values exceeding 50% were deemed to indicate significant heterogeneity. Subgroup analysis was performed to identify the source of heterogeneity. Potential publication bias was assessed by Egger’s and Begg’s test. In order to confirm the stability of the combined results, we conducted a sensitivity analysis to assess how each study influenced the overall estimate by omitting individual studies in turn. A bilateral P value less than 0.05 was deemed statistically significant. All statistical analyses were executed by Stata software (version 15.1; Stata Corp., College Station, Texas, USA) and Review Manager software (RevMan version 5.3, the Nordic Cochrane Center, the Cochrane Collaboration, 2014).

Results

Literature search

Through the literature search scheme, 237 documents with potential research value were retrieved, including 70 PubMed citations, 157 EMBASE citations, 9 Cochrane Library citations, and 1 relevant study yielded from the reference list. After eliminating duplicate publications, there were 171 studies left. By sifting through the titles and abstracts of each study, 31 studies remained. Finally, we carefully read the full text of the remaining articles, and 10 studies with 1,451 patients were included in our meta-analysis. A diagram illustrating the literature search process is shown in Fig. 1.

Fig. 1
figure 1

PRISMA flow diagram of literature search. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Full size image

Characteristics of the included studies

Table 1 describes the baseline characteristics and methodological assessments of each included study. There are ten retrospective studies published range from 2008 to 2023 from various regions of the globe. Four were published in China, two in Japan, two in America, one in Greece and one in Spain. The sample size of the study ranged from 59 to 490. It must be mentioned that all the eligible studies focused on intratumoral TLS, thus we mainly discuss the influence of intratumoral TLS on the prognosis of NSCLC patients. Of the included studies, patients in eight studies received surgical treatment only, and patients in three studies received neoadjuvant chemoimmunotherapy (NCIT) and surgical treatment. Notably, the study of Sun et al. examined both surgery-only and surgery with NCIT patients in relation to TLS, so we analyzed it as two studies [18]. In these ten studies, four evaluated the correlation between TLS and OS [19,20,21,22], seven evaluated the correlation between TLS and DFS/RFS [2, and Supplementary Fig. 1).

Fig. 2
figure 2

Forest plot of the correlation between TLS and (A) disease-free survival/recurrence-free survival, (B) overall survival, and (C) disease-specific survival in non-small cell lung cancer patients. TLS, tertiary lymphoid structures; HR, hazard ratio; CI, confidence interval

Full size image
Table 2 Subgroup analyses of DFS/RFS in non-small cell lung cancer
Full size table

Four studies appraised the association between intratumoral TLS and OS in 422 patients [19,20,21,22]. The pooled analysis revealed that high TLS was associated with preferable OS (HR = 0.48, 95% CI: 0.35–0.66, p < 0.001). The heterogeneity of the studies was low (I2 = 0.0%, p = 0.554) (Fig. 2B). Only two studies have appraised the association between intratumoral TLS and DSS in NSCLC patients [19, 25]. The results indicate that high TLS is closely related to batter DSS (HR = 0.45, 95% CI: 0.30–0.68, p < 0.001), with low heterogeneity (I2 = 26.3%, p = 0.244) (Fig. 2C).

Correlation between TLS and clinicopathological characteristics in NSCLC

The correlation analysis and evaluation results between TLS and various clinicopathological features are shown in Table 3. Overall, we examined the patients’ age (elder vs. young), gender (male vs. female), histological type (adenocarcinoma vs. squamous cell carcinoma), tumor size (large vs. small), smoking history (ever vs. never), TNM stage (II-IV vs. I), and NLR levels (high vs. low). After careful investigation, we determined that the relationship between TLS and TNM stage (OR = 0.71, 95% CI: 0.51-1.00, p < 0.05) and NLR level (OR = 0.33, 95% CI: 0.17–0.62, p < 0.001) was significant. TLS did not show any notable correlation with the patient’s age (OR = 1.11, 95% CI: 0.71–1.76, p = 0.64), gender (OR = 0.81, 95% CI: 0.61–1.08, p = 0.15), tumor classification (OR = 0.97, 95% CI: 0.73–1.30, p = 0.85), tumor size(OR = 0.97, 95% CI: 0.55–1.72, p = 0.92), or smoking status(OR = 1.01, 95% CI: 0.67–1.51, p = 0.97) (Fig. 3).

Table 3 Correlations of clinicopathological characteristics in patients with non-small cell lung cancer
Full size table
Fig. 3
figure 3

Forest plot of the correlation between TLS and clinicopathological characteristics in patients with NSCLC. (A) age; (B) TNM stage; (C) gender; (D) tumor size; (E) smoke; (F) NLR; (G) histological type. TLS, tertiary lymphoid structure; OR, odds ratio; CI, confidence interval; TNM, Tumor Node Metastasis; NLR, Neutrophil-lymphocyte ratio

Full size image

Sensitive analysis and publication bias

We conducted a sensitivity analysis by excluding the studies one by one. The HRs calculated from the combined results of the remaining studies in each analysis did not go beyond the expected range, as illustrated in Supplementary Fig. 2 and Supplementary Fig. 3. There is no significant difference between the revised overall estimate and the primary combined estimate, indicating that the meta-analysis is reliable. Begg’s and Egger’s tests are employed to identify any potential publication bias. The meta-analysis did not show any clear publication bias on TLS with respect to OS (Egger’s p = 0.369, Begg’s p = 0.308) and DFS/RFS (Egger’s p = 0.117, Begg’s p = 0.117).

Discussion

In the past few years, as researchers have delved deeper into the tumor microenvironment (TME) and the workings of tumor immunotherapy, TLS has emerged as a significant biological structure that hinders tumor growth by stimulating the activation of immune cells near the tumor [4, 26, 27]. Numerous research studies have investigated the significance of TLS in treating individuals with cancers, and the majority indicating that elevated TLS levels are a crucial indicator of a positive prognosis for various solid tumors [28, 29]. However, the prognostic value of TLS in NSCLC remains controversial. This meta-analysis integrated prognostic data and clinical characteristics of 1,451 NSCLC patients from 10 studies and conducted subgroup analysis. Following a thorough quantitative analysis of prognostic data, we determined that elevated levels of TLS were strongly associated with improved OS, DSS, and DFS/RFS. Additionally, high TLS levels were found to be closely linked to the tumor TNM stage and NLR. This meta-analysis represents the most up-to-date and extensive investigation regarding the correlation between TLS and prognosis, as well as relevant clinicopathological characteristics in individuals diagnosed with NSCLC.

Although only three studies in this meta-analysis focused on the relationship between TLS and patient outcomes in those who underwent immunotherapy before surgery, our findings indicated that individuals with increased TLS levels who received immunotherapy before surgery had a more favorable prognosis compared to those who underwent surgery alone [30]. TLS is an essential component of the tumor immune microenvironment (TIME) and includes T cells, B cells, fibroblast reticular cells (FRC) networks, high endothelial venules (HEV), and follicular dendritic cells (FDC) [4, 31, 32]. Within the TME, TLS serves as a site where immune cells can proliferate and interact. This area is primarily made up of an internal region of CD20+ B cells and a surrounding region of CD3+ T cells [33, 34]. Additionally, there is a significant presence of dendritic cells (DC) surrounding the immune cells, all of which congregate in this space to collectively suppress tumor growth. In this cluster of immune cells, DC displays the surface antigen of nearby tumor tissue to T cells via TLS [35, 36]. The activated T cells then produce memory helper T cells and effector memory cytotoxic cells to aid in the destruction of tumor cells through phagocytosis [37, 38]. Furthermore, this cluster supports the activation and growth of B cells, facilitating the development, activation, and growth of memory B cells and plasma cells [12, 13, 18]. Finally, our meta-analysis focused solely on the presence of TLS within the tumor itself, rather than its presence outside the tumor, which may not fully represent its impact on tumor prognosis. Given these constraints, it is essential to conduct numerous multi-center prospective studies to validate our findings before implementing them in clinical settings.

Conclusion

TLS plays a crucial role in the treatment of NSCLC. Elevated TLS levels are strongly related to positive survival outcomes such as OS, DSS, and DFS/RFS in NSCLC. Additionally, TLS expression levels are closely associated with certain clinicopathological factors of NSCLC patients. Therefore, TLS has the potential to serve as a biomarker for predicting the prognosis of NSCLC patients and may influence clinical treatment decisions. Nevertheless, further prospective studies are necessary to validate the prognostic significance of TLS in NSCLC patients before its clinical application.