Characterization of TCF‐1 and its relationship between CD8+ TIL densities and immune checkpoints and their joint influences on prognoses of lung adenocarcinoma patients

Abstract Background T cell factor‐1 (TCF‐1) + stem‐like tumor‐infiltrating lymphocytes (stem‐like TILs) are important memory cells in the tumor microenvironment. However, their relationship with clinicopathological features, CD8+ TIL densities, immune checkpoint inhibitors (ICs), and prognostic values remain unknown for lung adenocarcinomas (LUADs). In this study, we aimed to characterize TCF‐1+ TILs and their prognostic significance in patients with surgically resected LUADs. Methods Expression of TCF‐1, CD8, and ICs including programmed death‐1 (PD‐1), lymphocyte activating‐3 (LAG‐3), and T cell immunoglobulin and mucin‐domain containing‐3 (TIM‐3) in TILs were estimated using immunohistochemistry of resected LUADs. The association between TCF‐1 expressions and clinicopathological characteristics of patient prognoses were analyzed. Results Positive TCF‐1 expression significantly correlated with advanced pathological stage, tumor grade, CD8+ TILs density, TIM‐3 expression, LAG‐3 expression, and PD‐1 expression. TCF‐1 positivity was significantly associated with a better recurrence‐free survival (RFS), and overall survival (OS). Subgroup analysis revealed that the TCF‐1+/CD8+ group had the best RFS and OS, while the TCF‐1‐/CD8‐ group had the worst RFS and OS. Similarly, patients with TCF‐1 + PD‐1‐ had the best prognoses and patients with TCF‐1‐PD‐1+ had the worst prognoses. Conclusion TCF‐1 had relatively high positive expression and special clinicopathological features in patients with LUAD. TCF‐1+ TILs were related to CD8 density, TIM‐3 expression, LAG‐3 expression, and PD‐1 expression, and were associated with better prognoses in LUAD patients. A combination of TCF‐1 and CD8 densities or PD‐1 expression further stratified patients into different groups with distinct prognoses.


INTRODUCTION
Lung cancer is the most common malignant cancer, which leads to a large number of deaths worldwide. 1pproximately 60% are lung adenocarcinomas, and most are diagnosed at a later stage. 2,35][6][7][8] Recently, immunotherapy has been improved due to the rapid clinical development of effective immune checkpoint inhibitors (ICIs) such as T lymphocyte-associated antigen-4 (CTLA-4), programmed cell death 1 (PD-1), and its ligand (PD-L1). 9,10In addition, other prospective immune checkpoints (ICs) such as lymphocyte activation gene-3 (LAG-3), T cell immunoglobulin and mucin-domain containing 3 (TIM-3), T cell immunoglobulin and ITIM domain (TIGIT), and V-domain Ig suppressor of T cell activation (VISTA) are in preclinical trials.
Favorable outcomes of ICI treatments to a great extent depend on high infiltration of fully functional, cytotoxic effector TILs.Recently, a subset of TCF-1+ stem-like TILs were found to play vital roles in cancer immunotherapy.These progenitor cells sustain self-renewal and proliferation during cancer development, which in turn helps maintain an antitumor response. 11,12Transcription factor T cell factor 1 (TCF-1), encoded by TCF-7, is a critical transcription factor of TIL development.TCF-1 silencing causes T progenitor cells to lose their self-renewing ability, resulting in irreversible differentiation of effector TILs, as confirmed in mouse models. 13n previous studies, TCF-1+ stem-like TILs were associated with an ICI response in murine and human tumors. 14,15The high infiltration of TCF-1+ TILs has been shown to be associated with prolonged progression-free survival (PFS) and overall survival (OS) in melanoma patients receiving checkpoint blockade. 16Ma et al. also reported the prognostic value of TCF-1+ stem-like TILs predicting better survivals in primary small cell carcinomas of the esophagus. 17However, the relationship between TCF-1+ stem-like TILs and clinicopathological characteristics and their prognostic value in patients with surgically resected lung adenocarcinoma is unknown.Moreover, TCF-1+ TILs can undergo massive expansion in response to anti-PD-1 treatment, 14 so the survival of combined TCF-1+ stem-like TILs and PD-1 expression requires further research.In the present study, we therefore quantitatively analyzed clinicopathological features of TCF-1 expression, evaluated its prognostic value, and assessed its associations with PD-1, LAG-3, and TIM-3 expressions, as well as the density of CD8+ TILs.

Patients and clinical pathology
A total of 350 patients with LUADs who were retrospectively treated with tumor resection in the Thoracic Department at The Second Affiliated Hospital of Soochow University from April 2015 to December 2018 were enrolled into the study.Some of the patients in our previous study were included in the cohort. 9The inclusion criteria of patients were as follows: (1) patients pathologically confirmed with primary LUAD according to the eighth edition of the TNM classification, 12,14 (2) patients who had not undergone preoperative neoadjuvant radiochemotherapy or targeted therapy and (3) patients with postoperative adjuvant chemotherapy based on cisplatin.The exclusion terms included: (1) patients lost to medical follow-up, and (2) patients with other malignancies or concurrent multiple primary tumors.According to the criteria, 60 patients who accepted preoperative or postoperative chemoradiotherapy, targeted therapy, or immune therapy and 62 patients who were lost to follow-up were excluded, with the remaining 228 patients enrolled.Immunohistochemistry (IHC) staining was used to detect the expressions of TCF-1, PD-1, LAG-3, and TIM-3 in specimens of these patients.Two experienced pathologists (YQS and HXC) who were blinded to the clinical outcomes, independently analyzed the IHC results.Discussions were made if there was controversy or discordance in the pathological diagnoses, followed by a consensus.LUADs were classified according to the IASLC/ATS/ERS classification 16 and stages were determined according to the eighth edition of the TNM classification for LUAD. 18,19The clinicopathological characteristics of these patients are summarized in Table 1, after an average follow-up duration of 57 months.This study was approved by the Institutional Review Board of The Second Affiliated Hospital of Soochow University.Informed consent of patients was not required due to the retrospective nature of the study.

Statistical analysis
Associations between clinicopathological characteristics were analyzed using the chi-square or Fisher's exact test for categorical variables.In addition, a logistic regression model was used to confirm independent risk factors for the presence of TCF-1.RFS was defined as the time from surgical resection to the first time of recurrence.OS was defined as the time from surgical resection until death from any cause or from the last follow-up.RFS and OS were evaluated using the Kaplan-Meier method, and nonparametric group comparisons were performed using the logrank test.A Cox proportional-hazards regression model was used to identify independent risk factors for RFS and OS.The variables were first examined using univariate analyses, and those with p-values <0.05 were incorporated into a multivariate model.All p-values were based on two-tailed statistical analyses, and p < 0.05 was considered statistically significant.Statistical analyses were conducted using SPSS statistical software for Windows, version 25.0 (SPPS).The survival curves were drawn using Origin 2021 software (OriginLab).

Baseline information
The clinical characteristics of 228 patients are shown in Table 1.The mean age was 63 years (range: 30-79 years).A total of 138 (60.5%) were male and 90 (39.5%) were female.

Relationships between TCF-1 and ICs of LUADs
The relationships between TCF-1 and ICs are shown in Table 1.High expression of TCF-1 was significantly correlated with higher expression of TIM-3 expression in TILs (p = 0.023), LAG-3 expression in TILs (p = 0.001), and PD-1 expression in TILs (p = 0.002).

DISCUSSION
The past decade has witnessed the rapid development of immunotherapy for the treatment of cancer. 24Among these immunotherapies, use of coinhibitory immune ICIs, including PD-1, PD-L1, and CTLA-4 monoclonal antibodies (mAbs), have become the most promising clinical treatments. 25The next generation of ICIs such as for LAG-3, TIM-3, TIGIT, VISTA, B7 homolog 3 protein, and B and T cell lymphocyte attenuators are now in preclinical trials.However, the responses of anti-PD-1/PD-L1mAb or anti-CTLA-4 mAb is still far from satisfactory. 26Studies of novel  ICIs have been ongoing.Meanwhile, quantitative detection of new and meaningful immunity-associated proteins may help predict which patients will benefit from immunotherapy.
During progression of CD8+ T cell differentiation, a small subset of CD8+ T cells retain the potential for lymphoid recirculation and the ability of self-renewal, resulting in the production of more differentiated effector TILs.These cells are defined by, and depend on, expression of the transcription factor, TCF-1.This key transcription factor is essential for the generation of stem-like TILs during cancer immunity. 27,28Ablation of intratumoral TCF-1+ TILs showed that TCF-1-TILs lacked expansion capacity and restricted responses to immunotherapy.As a result, residual TILs lost their robust capability to maintain tumor control.In the past year, there have been reports documenting the presence of TCF-1+ TILs in human cancers, 17 and also reported results suggesting that the frequency of these cells was associated with their clinical outcomes.Miller and Sade-Feldman reported that melanoma patients with high stem-like TIL infiltration had a longer PFS and OS. 16,29ccordingly, Ma et al. reported that primary small cell LUAD of esophagus patients with high infiltration of TCF-1+ TILs had a longer OS and low infiltration ( p = 0.009; HR: 0.506). 17In contrast, coexpression of LEF-1 and TCF-1 proteins in patients with nasopharyngeal carcinomas were positively correlated with lymph node metastasis (p = 0.001 and p = 0.020, respectively), advanced clinical stage (p < 0.003 and p = 0.027, respectively), and poor survival status (p < 0.001 and p = 0.004, respectively).In some other malignant tumors, TCF-1 was overexpressed in osteosarcoma tissues, when compared with matched adjacent normal tissues. 30,31Similar results have also been reported in renal cell carcinomas. 32The opposite role of TCF-1 may depend on its location, which determines whether it is expressed in TILs or tumor cells.In the present study, TCF-1+ TILs were more frequently found in LUAD patients with a higher tumor grade ( p < 0.001), advanced pathological stage (p = 0.045), and greater CD8 expression ( p = 0.036), but not age, sex, smoking history, tumor location, CEA level, or VPI.Further survival analyses revealed that LUAD patients with high TCF-1 TIL infiltration had a higher RFS (p = 0.001) and OS (p = 0.011).Based on these results, we propose that infiltration of TCF-1+ stem-like TILs is a positive prognostic biomarker in LUAD patients.
Because TCF-1 acts as an important regulator in TIL stemness, studies have investigated the detailed gene axis and possible regulatory mechanisms.Chatterjee and Xu 34,35 reported that TCF-1 was expressed in multiple isoforms in TILs, in which the long isoforms interacted with β-catenin through an N-terminal domain, while TCF-1 short isoforms supported developing thymocytes to traverse through maturation steps to regulate most TCF-1 target genes.Chemical inhibition of β-catenin/TCF-1 interactions improves longterm self-renewal and enhances functional pluripotency with increased Nanog expression.Man et al. 33 showed that in CD8+ T cells, IRF4, BATF, and NFAT were recruited to adjacent binding sites, and binding of all three factors was significantly enriched among the core group of proteins related to exhaustion, including PD-1, LAG-3, HAVCR2, TIGIT, and CTLA-4.Furthermore, Wu et al. 36 reported that TCF-1 acted upstream of the Bcl6-Blimp1 axis in TCF-1 high CD8 T cells, and that these TCF-1 high CD8 T cells also expressed lower Abbreviations: CEA, carcinoembryonic antigen; CI, confidence interval; HR, hazard ratio; LAG-3, lymphocyte activation gene-3; PD-1, programmed cell death-1; TCF1, T cell factor 1; VPI, visceral pleural invasion; TIM-3, T cell immunoglobulin and mucin-domain containing-3; vs., versus.
levels of canonical TH1 markers, including Blimp1 and Il2ra. 37IL-2 signaling can repress TIL differentiation via the signal transducer and activator of transcription 5 (STAT5) pathways. 38In the present study, we found that positive TCF-1 expression significantly correlated with expressions of PD-1, LAG-3, and TIM-3, as well as high CD8 + TILs density.It is therefore possible that an internal connection between high expressions of coinhibitory ICIs and IRF4, BATF, and NFAT gene pathways in LUADs exists.Furthermore, the role of variable isoforms of TCF-1 in promoting coinhibitory ICI expression should be investigated.
In conclusion, this is the first study to characterize TCF-1 expression in TILs, and its prognostic significance in patients with surgically resected LUADs.However, there were some limitations in our study.First, performance and selection bias were inevitable because of the retrospective nature of the study.Second, we only included patients from a single institution.A prospective study and a larger cohort of patients with LUAD are therefore needed.Also, the patients lost to follow-up appear high.Third, we did not investigate PD-L1, an established predictive marker for immunotherapy.Additional multicenter studies with larger patient cohorts may address these limitations.
In conclusion, TCF-1+ TILs had relatively high positive infiltrations and specific clinicopathological features in patients with LUADs.TCF-1+ TILs significantly correlated with pathological stage, tumor grade, CD8+ TILs density, and PD-1, LAG-3, and TIM-3 expression levels in TILs.TCF-1+ TILs were significantly associated with a better RFS and OS.Furthermore, the combination of TCF-1+ TILs and PD-1 or CD8 expression in TILs further stratified patients into distinct groups with different prognoses.
Conceptualization, methodology and writing original draft: K. J, B. Q, Data curation and writing original draft preparation: Y. C, Visualization and investigation: S. L, Resources and visualization: Z. X, Software and validation: H. C, Writing-reviewing and editing: Z. X, Validation formal analysis: Y. S, Resources: D. C, Investigation: W. X, Software formal analysis: Q. D, Conceptualization and supervision project administration: M. S.
Correlation between TCF1 expression and clinicopathologic parameters.

Table 1
Multivariate logistic regression model for TCF-1 expression in patients with lung adenocarcinoma.
T A B L E 2 T A B L E 3 Cox proportional-hazards regression model for recurrence-free survival and overall survival in all patients.
Note: Smoking: current smoker or former smoker (yes) and never smoker (no).Variables with p-value <0.05 in univariate models were analyzed in multivariate analysis model.