Impact of TTF-1 Expression on the Prognostic Prediction of Patients with Non–Small Cell Lung Cancer with PD-L1 Expression Levels of 1% to 49%, Treated with Chemotherapy vs. Chemoimmunotherapy: A Multicenter, Retrospective Study
Article information
, Tae Hata1, Tadaaki Yamada,1, Yasuhiro Goto2, Akihiko Amano3, Yoshiki Negi4, Satoshi Watanabe5, Naoki Furuya6, Tomohiro Oba7, Tatsuki Ikoma8, Akira Nakao9, Keiko Tanimura10, Hirokazu Taniguchi11, Akihiro Yoshimura12, Tomoya Fukui13, Daiki Murata14, Kyoichi Kaira15, Shinsuke Shiotsu16, Makoto Hibino17, Asuka Okada18, Yusuke Chihara19, Hayato Kawachi1, Takashi Kijima4, Koichi Takayama1
Abstract
Purpose
Thyroid transcription factor 1 (TTF-1) expression is a useful predictor of treatment efficacy in advanced non-squamous non–small cell lung cancer (NSCLC). This study aimed to evaluate whether TTF-1 could predict the effectiveness of chemotherapy versus chemoimmunotherapy in patients with non-squamous NSCLC with programmed death ligand-1 (PD-L1) expression between 1% and 49%.
Materials and Methods
We conducted a retrospective study of patients with NSCLC who were treated with chemotherapy or chemoimmunotherapy between March 2016 and May 2023. The patients had histologically confirmed NSCLC, stage III-IV or postoperative recurrence, TTF-1 measurements, and PD-L1 expression levels between 1% and 49%. Clinical data were analyzed to evaluate the effect of TTF-1 expression on treatment efficacy.
Results
This study included 283 of 624 patients. TTF-1–positive patients showed longer progression-free survival (PFS) and overall survival (OS) (PFS: 6.4 months [95% confidence interval (CI), 5.0 to 9.4] vs. 4.1 months [95% CI, 2.7 to 6.1], p=0.03; OS: 17.9 months [95% CI, 15.2 to 28.1] vs. 9.4 months [95% CI, 6.3 to 17.0], p < 0.01) in the chemotherapy cohorts (n=93). In the chemoimmunotherapy cohort (n=190), there was no significant difference in PFS and OS between TTF-1–positive and –negative groups (PFS: 7.6 months [95% CI, 6.4 to 11.0] vs. 6.0 months [95% CI, 3.6 to 12.6], p=0.59; OS: 25.0 months [95% CI, 18.0 to 49.2] vs. 21.3 months [95% CI, 9.8 to 28.8], p=0.09).
Conclusion
In patients with NSCLC with PD-L1 expression between 1% and 49%, TTF-1 expression was a predictor of chemotherapeutic, but not chemoimmunotherapeutic, efficacy.
Introduction
Thyroid transcription factor 1 (TTF-1) is a key immunohistochemical marker for diagnosing lung cancer, distinguishing lung and thyroid adenocarcinomas from cancers of other sites, and differentiating lung adenocarcinomas and small-cell lung cancer from squamous cell carcinoma [1]. TTF-1 expression is linked to approximately 30% of fatal lung adenocarcinoma cases [2-4]. The variation in TTF-1 expression may be related to the degree of tumor differentiation within the same tumor category [5]. TTF-1 expression influences treatment response and prognosis in patients with non-squamous non–small cell lung cancer (NSCLC) [6-9]. TTF-1 is also a useful prognostic factor for the risk of postoperative recurrence [10-12] and a predictor of response to cytotoxic chemotherapy in patients with advanced non-squamous NSCLC [6,9,13,14].
Since the advent of immunotherapy, the survival rate of patients with advanced NSCLC has increased significantly. Chemoimmunotherapy, which combines chemotherapy and immunotherapy, has become the first-line treatment for patients with advanced NSCLC. Therefore, predicting the efficacy of this combination therapy is crucial. Although intratumoral programmed death ligand-1 (PD-L1) expression is a robust predictor of response to immunotherapy and chemoimmunotherapy and is employed in clinical practice, it is not comprehensive. This study aimed to evaluate whether tumor expression of TTF-1, a known predictor of chemotherapy response, could also predict the efficacy of chemoimmunotherapy in patients with advanced lung cancer and PD-L1 expression ranging between 1% and 49%.
Materials and Methods
1. Patients and study design
We conducted a retrospective study of patients with NSCLC who received chemotherapy or chemoimmunotherapy at 19 institutions in Japan between March 2016 and May 2023. We included patients with (1) histologically diagnosed NSCLC; (2) histologically certified stage III-IV according to TNM staging (American Joint Committee on Cancer 8th edition) or postoperative recurrence; (3) available TTF-1 measurements; and (4) PD-L1 expression between 1% and 49%. Subsequently, we excluded the following patients: (1) those histologically diagnosed with squamous cell carcinoma, (2) those not evaluated for TTF-1 expression in tumors, and (3) those with driver mutations. Clinical data were collected from the electronic medical records. PD-L1 scores were analyzed using PD-L1 immunohistochemistry with 22C3 pharmDx antibody (clone 22C3, Dako North America, Inc.), and TTF-1 was analyzed using mouse 84 monoclonal anti-TTF-1 antibody clone 8G7G3/1, mouse monoclonal anti-TTF-1 antibody clone 85 SPT24, and rabbit monoclonal anti– TTF-1 antibody clone SP141.
2. Assessment outcome
We examined the effect of TTF-1 expression on treatment efficacy in patients receiving chemotherapy and chemoimmunotherapy, including the effect of pemetrexed-containing regimens in TTF-1–negative and TTF-1–positive patients. Progression-free survival (PFS) was defined as the duration from the date of treatment initiation to the date of tumor progression or death. Overall survival (OS) was defined as the duration from the date of initiation of first-line treatment to the date of all-cause death. The follow-up cutoff date was January 5, 2024, and the median follow-up duration was 21.3 months.
3. Statistical analysis
Binomial variables were analyzed using Fisher’s exact test. Differences in continuous variables between the two groups were assessed using the Mann-Whitney U test. PFS and OS were evaluated using the Kaplan-Meier method and compared using the log-rank test. Cox proportional hazard models were used to assess predictive and prognostic factors for PFS and OS. All analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) and the R graphical user interface (R Foundation for Statistical Computing), with statistical significance set at p < 0.05 [15].
Results
1. Patient characteristics
We first identified 624 patients with advanced or recurrent NSCLC with PD-L1 expression between 1% and 49% who received chemotherapy or chemoimmunotherapy between August 2014 and June 2023 at 19 institutions in Japan. Of these, 295 patients without TTF-1 expression evaluation were excluded, followed by the exclusion of 45 patients with actionable driver mutations, such as epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK), as well as one patient who received pembrolizumab monotherapy. Finally, two hundred eighty-three patients were included in the study (S1 Fig.). The patient characteristics are presented in Tables 1 and 2. The median age of the patients in the chemotherapy cohort (n=93) was 68 years (range, 39 to 81 years). Most patients were male (76.3%) and had an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0-1 (88.2%). Additionally, most tumors were adenocarcinomas (73.1%) or initial stage IV (73.1%), with the majority having PD-L1 expression of 1% or more but less than 25% (68.8%). We classified patients into TTF-1–positive or negative groups, and the TTF-1–positive group had a significantly higher percentage of adenocarcinomas (83.0% vs. 60.0%, p=0.02). The median patient age in the chemoimmunotherapy cohort (n=190) was 69 years (range, 36 to 85 years). Similar to the chemotherapy cohort, most patients were male (75.6%) and had an ECOG PS of 0-1 (92.1%). Most tumors were adenocarcinomas (93.7%) or initial stage IV tumors (84.2%), with the majority having PD-L1 expression of 1% or more but less than 25% (72.1%). In both the chemotherapy and chemoimmunotherapy cohorts, the TTF-1–positive group had a significantly lower proportion of non-adenocarcinoma NSCLC than the TTF-1–negative group (chemotherapy cohort: 17.0% vs. 40.0%, p=0.02, chemoimmunotherapy cohort: 6.3% vs. 38.1%, p < 0.01).
Patient characteristics: chemotherapy cohort
Patient characteristics: chemoimmunotherapy cohort
2. Impact of TTF-1 expression on the efficacy of chemotherapy
In the chemotherapy cohort, the median follow-up period was 16.4 months (range, 12.6 to 20.8 months), and 81 of 93 patients (87.1%) experienced progressive disease. Regarding PFS, the TTF-1–negative group had a significantly shorter PFS than the TTF-1–positive group (4.1 months [95% confidence interval (CI), 2.7 to 6.1] vs. 6.4 months [95% CI, 5.0 to 9.4], p=0.03) (Fig. 1A). Multivariate analysis revealed that TTF-1 expression (adjusted hazard ratio [HR], 0.58; 95% CI, 0.36 to 0.93; p=0.02) and ECOG PS (adjusted HR, 0.41; 95% CI, 0.19 to 0.88; p=0.02) were independent predictive factors for PFS (Table 3). Similarly, in terms of OS, the TTF-1–negative group had significantly shorter OS than that of the positive group (9.4 months [95% CI, 6.3 to 17.0] vs. 17.9 months [95% CI, 15.2 to 28.1], p < 0.01) (Fig. 1B). In the multivariate analysis of OS, TTF-1 expression (adjusted HR, 0.61; 95% CI, 0.37 to 1.03; p=0.06) and histology (adjusted HR, 0.60; 95% CI, 0.35 to 1.04; p= 0.07) were identified as important factors. However, they were not statistically significant (Table 3).
Comparison of treatment efficacy between thyroid transcription factor-1 (TTF-1)–positive and negative groups in the chemotherapy and chemoimmunotherapy cohorts. Kaplan-Meier curves for progression-free survival (PFS) (A) and overall survival (OS) (B) in the chemotherapy cohorts. Kaplan-Meier curves for PFS (C) and OS (D) in the chemoimmunotherapy cohort. CI, confidence interval.
Predictive factors for progression free survival and overall survival in the chemotherapy cohort
3. Impact of TTF-1 expression on the efficacy of chemoimmunotherapy
In the chemoimmunotherapy cohort, the median follow-up period was 24.1 months (range, 18.0 to 28.8 months), and 139 of 190 patients (73.2%) experienced progressive disease. Regarding PFS, there were no significant differences between the TTF-1–negative and positive groups (6.0 months [95% CI, 3.6 to 12.6] vs. 7.6 months [95% CI, 6.4 to 11.1], p=0.59) (Fig. 1C). In the multivariate analysis of PFS, ECOG PS (adjusted HR, 0.51; 95% CI, 0.26 to 0.98; p=0.04) was an independent predictor of PFS, while TTF-1 expression did not show significance (adjusted HR, 0.86; 95% CI, 0.57 to 1.30; p=0.48) (Table 4). Regarding OS, there was a discernible tendency towards enhanced performance in the chemoimmunotherapy cohort. However, no discernible discrepancy was observed between the TTF-1–negative and positive groups (21.3 months [95% CI, 9.8 to 28.8] vs. 25.0 months [95% CI, 18.0 to 49.2], p=0.09) (Fig. 1D). In the multivariate analysis of OS, TTF-1 was also not a prognostic factor (adjusted HR, 0.79; 95% CI, 0.50 to 1.25; p=0.31) (Table 4).
Predictive factors for progression free survival and overall survival in the chemoimmunotherapy cohort
4. Comparison of OS between pemetrexed-based and non-pemetrexed–based regimens within the chemotherapy and chemoimmunotherapy cohorts, stratified by TTF-1 expression status
The chemotherapy and chemoimmunotherapy cohorts were divided into TTF-1–positive and negative groups, and chemotherapy in each group was classified as follows to evaluate OS: pemetrexed-based chemotherapy group (PC group)/chemoimmunotherapy group (CI with PEM group) and non-pemetrexed–based chemotherapy group (non-PC group)/chemoimmunotherapy group (CI without PEM group). In the chemotherapy cohort, the OS was significantly longer in the PC group than in the non-PC group within the TTF-1–positive group (23.9 vs. 16.4 months, p=0.02) (Fig. 2A). However, in the TTF-1–negative group, there was no significant difference between the two groups (9.4 months vs. 10.0 months, p=0.42) (Fig. 2C). In the chemoimmunotherapy cohort, regardless of TTF-1 expression, there was no significant difference in OS between the CI with and without PEM groups (TTF-1–positive: 27.5 months vs. 22.6 months, p=0.25 [Fig. 2B], TTF-1–negative: 21.3 months vs. 18.4 months, p=0.53 [Fig. 2D]).
Comparison of overall survival (OS) between pemetrexed (PEM)-based and non-PEM–based regimens within the chemotherapy and chemoimmunotherapy cohorts stratified by thyroid transcription factor-1 (TTF-1) expression status. We compared the OS between the PEM-based and non-PEM–based regimen groups in the chemotherapy and chemotherapy cohorts stratified by TTF-1 expression status. Kaplan-Meier curves are shown for the chemotherapy (A) and chemoimmunotherapy (B) groups in the TTF-1–positive population and for the chemotherapy (C) and chemoimmunotherapy (D) groups in the TTF-1–negative population. CI, confidence interval.
5. Comparison of treatment efficacy between chemoimmunotherapy and chemotherapy stratified by TTF-1 expression
We compared the efficacy of chemoimmunotherapy and chemotherapy stratified by TTF-1 expression. In both TTF-1–positive and negative groups, PFS tended to be longer with chemoimmunotherapy than with chemotherapy (TTF-1–positive: 7.6 months vs. 6.4 months, p=0.054 [S2A Fig.]; TTF-1–negative: 6.0 months vs 4.1 months, p=0.007 [S2C Fig.]). However, for OS, no significant difference was observed between the two treatment modalities in either group (TTF-1–positive: 25.0 months vs 17.9 months, p=0.28 [S2B Fig.]; TTF-1–negative: 21.3 months vs 9.4 months, p=0.13 [S2D Fig.]).
Discussion
This is the most extensive study to evaluate the influence of TTF-1 expression on the therapeutic effectiveness of chemoimmunotherapy and chemotherapy in patients with NSCLC with PD-L1 levels between 1 and 49%. Therefore, the results suggest that while TTF-1 was a prognostic factor in the chemotherapy cohort, it was not a prognostic factor in the chemoimmunotherapy cohort, indicating that immunotherapy may compensate for the prognosis of TTF-1–negative NSCLC.
Prior to the clinical adoption of immunotherapy, previou studies consistently reported that TTF-1 expression influenced the therapeutic efficacy and prognosis of cytotoxic chemotherapy [6,7,9,14,16]. Among the groups treated with cytotoxic chemotherapy in this study, the TTF-1–positive group tended to have a longer PFS and OS than the TTF-1–negative group. Multivariate analysis also suggested that TTF-1 could serve as an independent predictor of PFS and OS, as shown in previous studies. Several factors contribute to the worse prognosis of patients with TTF-1–negative NSCLC compared to those who are TTF-1–positive. Among these, we focused on the relationship between TTF-1 expression and epithelial-mesenchymal transition (EMT). EMT has been reported to contribute to resistance to treatment, including cytotoxic anticancer agents, cancer progression, and malignancy [17]. Saito et al. [18] demonstrated that, in NSCLC, TTF-1 suppresses EMT by inhibiting the expression of Snail and Slug, which are downstream molecules of transforming growth factor β (TGF-β). Furthermore, TGF-β signaling has been suggested to suppress TTF-1 expression. Based on these preclinical findings, it is expected that patients with TTF-1–negative NSCLC may exhibit lower treatment responsiveness to chemotherapy and worse prognosis due to EMT induction.
Since the clinical adoption of immunotherapy, the effect of TTF-1 expression on patient outcomes in NSCLC treated with immunotherapy or chemotherapy remains unclear. This study found that among patients with NSCLC and PD-L1 expression between 1% and 49%, TTF-1 did not affect the treatment response or prognosis in the chemoimmunotherapy group. However, recent studies suggest that TTF-1 might predict therapeutic response to chemotherapy and conventional chemotherapy [19-24]. Our previous study found that TTF-1 was a predictor of chemoimmunotherapy, which contradicts the results of this study [25]. These conflicting findings may stem from various factors; however, the association between TTF-1 and PD-L1 expression is particularly significant. Our previous study reported that significantly more patients in the TTF-1–negative NSCLC group had low or negative PD-L1 expression than those in the TTF-1–positive group [25]. As a result, TTF-1–negative patients are less likely to benefit from the therapeutic effects of programmed death-1 (PD-1)/PD-L1 inhibitors, which may have resulted in inferior PFS and OS. Several other studies have also reported this association between PD-L1 and TTF-1 [26,27]. Therefore, in this study, limiting patients with NSCLC to those with a PD-L1 expression rate of 1% to 49% would eliminate the imbalance in PD-L1 expression between the TTF-1–positive and negative groups, suggesting that TTF-1 expression does not affect treatment efficacy in chemoimmunotherapy. Supporting this hypothesis, our previous study showed no significant difference in treatment efficacy between patients who are TTF-1–positive and those who are negative, even in patients with NSCLC with PD-L1 expression ≥ 50%, whether they received PD-1/PD-L1 monotherapy or chemoimmunotherapy [28]. Therefore, it can be inferred that the influence of TTF-1 expression is minimal, at least during periods when the effects of PD-1/PD-L1 inhibitors are more pronounced.
We compared survival in pemetrexed-based and non-pemetrexed-based regimen groups stratified by TTF-1 expression in both chemotherapy and chemoimmunotherapy cohorts. The chemotherapy cohort indicated that OS may vary by regimen depending on TTF-1 expression, whereas the chemoimmunotherapy cohort showed no significant difference in OS according to regimen type, regardless of TTF-1 expression [6,7,9,14,16]. This study showed that OS was not affected by regimen type in the chemoimmunotherapy cohort, regardless of TTF-1 expression. OS tends to increase with chemotherapy, and the impact of post-treatment chemotherapy is expected to be significant. Additionally, the small sample size in this study may have contributed to these results. Therefore, further large-scale studies are warranted.
This study has some limitations. First, although this was the largest cohort study of this population, it was a retrospective Japanese study with a relatively small sample size. Specifically, comparing the treatment regimens for each TTF-1 expression group, especially chemoimmunotherapy, involved a small number of patients. Second, the selection of treatment regimens may vary owing to factors other than TTF-1, potentially influencing treatment efficacy in each population. Third, this study did not have information on driver gene abnormalities in NSCLC other than EGFR and ALK, which may indicate the presence of other mutations in the study population. Consequently, it cannot be ruled out that these unassessed abnormalities may have influenced the study results. Finally, we did not collect data on tumor differentiation in this study, and it is possible that differences in differentiation could have influenced treatment outcomes.
In conclusion, in patients with NSCLC with PD-L1 expression between 1% and 49%, TTF-1 expression was a predictor of cytotoxic chemotherapeutic but not chemoimmunotherapeutic efficacy. The addition of immunotherapy may attenuate the adverse effects of TTF-1 expression on the therapeutic effects of chemoimmunotherapy. Further prospective studies with larger sample sizes are required to verify these findings.
Electronic Supplementary Material
Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).
Notes
Ethical Statement
This study was conducted in compliance with the principles of the Declaration of Helsinki. This study complied with the Declaration of Helsinki and followed protocols approved by the Ethics Committee of the Kyoto Prefectural University of Medicine (approval No. ERB-C-2934), covering all participating centers through a centralized review process. Informed consent for the use of personal medical data was obtained using the opt-out method as described in the disclosure document.
Author Contributions
Conceived and designed the analysis: Nishioka N, Yamada T.
Collected the data: Nishioka N, Hata T, Goto Y, Amano A, Negi Y, Watanabe S, Furuya N, Oba T, Ikoma T, Nakao A, Tanimura K, Taniguchi H, Yoshimura A, Fukui T, Murata D, Kaira K, Shiotsu S, Hibino M, Okada A, Chihara Y, Kawachi H.
Contributed data or analysis tools: Nishioka N, Yamada T.
Performed the analysis: Nishioka N, Yamada T.
Wrote the paper: Nishioka N, Yamada T.
Supervision: Kijima T, Takayama K.
Conflict of Interest
N. Nishioka received personal fees from Chugai Pharmaceutical Co. Ltd., AstraZeneca KK., Eli Lilly Japan KK, and MSD KK, outside the purview of the submitted work. Tadaaki Yamada received research grants from Ono Pharmaceutical, Janssen, AstraZeneca, and Takeda Pharmaceutical and has received speaking honoraria from Eli Lilly and Chugai-Roshe outside the purview of the submitted work. Satoshi Watanabe received grants from Boehringer Ingelheim and Nippon Kayaku and has received honoraria for speakers’ bureaus from Lilly, Novartis Pharma, Chugai Pharma Bristol-Myers, Ono Pharmaceutical, Daiichi Sankyo, Taiho Pharmaceutical, Nippon Kayaku, Kyowa Kirin, Merck, Takeda Pharmaceutical, Celltrion, and AstraZeneca outside the purview of the submitted work. Hirokazu Taniguchi has received lecture fees from AstraZeneca and Chugai Pharma. Tomoya Fukui received personal fees from AstraZeneca K.K., Boehringer-Ingelheim Japan Inc., Chugai Pharmaceutical Co. Ltd., Eli Lilly Japan KK, Nippon Kayaku Co. Ltd., Novartis Pharma K.K., Ono Pharmaceutical Co. Ltd., and Pfizer Japan Inc., outside the scope of the submitted work. Kyoichi Kaira has received speaker honorariums from Ono Pharmaceutical Company, Chugai Pharmaceutical, Bristol-Myers Company, Boehringer Ingelheim, and AstraZeneca, and research grants from AstraZeneca. Asuka Okada has received personal fees from Chugai-Roshe, Kyowa Kirin, MSD KK, AstraZeneca, Takeda Pharmaceutical, Boehringer Ingelheim, Eli Lilly, Japan, Nippon Kayaku, and Bristol-Myers Squibb outside the purview of the submitted work. Hayato Kawachi received personal fees from Bristol-Myers Squibb, Ono Pharmaceutical Co. Ltd., Chugai Pharmaceutical Co. Ltd., AstraZeneca KK, Taiho Pharmaceutical Co. Ltd., Eli Lilly Japan KK, and MSD KK, outside the purview of the submitted work. Takashi Kijima received personal fees from Chugai Pharmaceutical Co., Ltd., Bristol-Myers Squibb, Ono Pharmaceutical Co. Ltd., and MSD KK, outside the purview of the submitted work. Koichi Takayama received research grants from Chugai Pharmaceutical Co. Ltd. and Ono Pharmaceutical and personal fees from AstraZeneca, Chugai Pharmaceutical Co. Ltd., MSD-Merck, Eli Lilly, Boehringer Ingelheim, and Daiichi-Sankyo outside the purview of the submitted work. The other authors declare no potential conflicts of interest.