Introduction

Oncology Letters

1792-1074 1792-1082

D.A. Spandidos

10.3892/ol.2023.13725

OL-25-4-13725

Articles

Diagnostic utility of trefoil factor families for the early detection of lung cancer and their correlation with tissue expression

Minegishi

Kentaro

1 Dobashi

Yoh

2 3 Koyama

Teruhide

4 Ishibashi

Yuko

5 Furuya

Miki

6 Tsubochi

Hiroyoshi

1 Ohmoto

Yasukazu

7 Yasuda

Tomohiko

6 8 Nomura

Sachiyo

1Department of Thoracic Surgery, Saitama Medical Center, Jichi Medical University, Saitama, Saitama 330-8500, Japan 2Department of Medicine, Saitama Medical Center, Jichi Medical University, Saitama, Saitama 330-8500, Japan 3Department of Pathology, School of Medicine, International University of Health and Welfare Hospital, Nasushiobara, Tochigi 329-2763, Japan 4Department of Epidemiology for Community Health and Medicine, Kyoto Prefectural University of Medicine, Kyoto, Kyoto 602-8566, Japan 5Department of Surgery, Breast Surgery, Tokyo Women's Medical University, Adachi Medical Center, Adachi, Tokyo 123-8558, Japan 6Department of Gastrointestinal Surgery, Graduate School of Medicine, The University of Tokyo, Tokyo 113-8655, Japan 7Department of Pharmacokinetics and Biopharmaceutics, Institute of Biomedical Sciences, Tokushima University, Tokushima, Tokushima 770-8505, Japan 8Department of Gastrointestinal Surgery, Nippon Medical School Chiba Hokusoh Hospital, Inzai, Chiba 270-1694, Japan

Correspondence to: Professor Yoh Dobashi, Department of Pathology, School of Medicine, International University of Health and Welfare Hospital, 537-3 Iguchi, Nasushiobara, Tochigi 329-2763, Japan, E-mail: ydobashi2@iuhw.ac.jp

04 2023

21 02 2023

25 4

139

27072022 22122022

2023

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Trefoil factors (TFFs) are upregulated in numerous types of cancer, including those of the breast, the colon, the lung and the pancreas, suggesting their potential utility as biomarkers for screening. In the present study, the clinical relevance of serum or urinary TFFs as biomarkers were comprehensively evaluated and the correlation with TFF expression levels in lung cancer tissue was examined. Serum and urine were collected from 199 patients with lung cancer and 198 healthy individuals. Concentrations of serum and urinary TFF1, TFF2 and TFF3 were measured using ELISA and the potential of TFF levels to discriminate between cancer and non-cancer samples was evaluated. In 100 of the cancer cases, expression of TFF1-3 was analyzed using immunohistochemical staining of paraffin sections. Furthermore, the relationship between TFF levels and clinicopathological factors among these cancer cases was analyzed using immunohistochemistry of tissue specimens, quantified and statistically analyzed. While serum levels of all TFFs measured using ELISA were significantly higher in patients with lung cancer compared with those in healthy individuals, urinary TFFs were lower. Areas under the curve (AUC) of the receiver operating characteristic curves for serum/urinary TFF1, TFF2 and TFF3 were 0.709/0.594, 0.722/0.501 and 0.663/0.665, respectively. Furthermore, the combination of serum TFF1, TFF2, TFF3 and urinary TFF1 and TFF3 demonstrated the highest AUC (0.826). In the clinicopathological analysis, serum TFF1 was higher in the early pathological T-stage (pTis/1/2) compared with the later stage (pT3/4) and TFF2 was higher in the pN0/1 than the pN2 group. With regards to the histological types, urinary TFF1 was higher in squamous cell carcinoma than adenocarcinoma (AC), but TFF2 tended to be higher in AC. Using immunohistochemical analysis, although TFF1 and TFF3 expression showed positive correlation with serum concentrations, TFF2 was inversely correlated. In conclusion, serum and urinary TFF levels are promising predictive biomarkers, and their measurements provide a useful in vivo and non-invasive diagnostic screening tool. In particular, TFF1 and TFF3 could be surrogate markers of clinicopathological profiles of human lung cancer.

TFF lung carcinoma ELISA biomarker immunohistochemistry

Japan Society for the Promotion of Science

17K08727

20K07378

16H06277

22H04923

23501295

Grants-in-Aid for Scientific Research for Priority Areas of Cancer

17015018

Innovative Areas from the Ministry of Education, Culture, Sports, Science and Technology, Japan

221S0001

Ca180066

Ca190058

Ca200066

Ca210033

This work was partially supported by the Japan Society for the Promotion of Science (grant no. 17K08727, 20K07378, 16H06277, 22H04923 and 23501295), Grants-in-Aid for Scientific Research for Priority Areas of Cancer (grant no. 17015018), Innovative Areas from the Ministry of Education, Culture, Sports, Science and Technology, Japan (grant nos. 221S0001, Ca180066, Ca190058, Ca200066 and Ca210033), the Smoking Research Foundation and the Princess Takamatsu Cancer Research Fund.

Introduction

Lung cancer is one of the most common malignancies worldwide and remains the leading cause of cancer-related death (1). The incidence of lung cancer is still increasing in the world, and screening by chest X-ray and computed tomography (CT) has shown a high success rate in diagnosis. Recent advances in targeted therapy for lung cancer have shown benefits for patients with adenocarcinoma (AC) who harbor mutations in driver genes. However, less satisfactory results have been obtained for more complex cases, such as drug-resistant cases, and treatment options for non-AC cases are limited (2,3). Therefore, the overall cure and survival rates in patients with advanced-stage lung cancer remain low (22% 5-year overall survival rate) (4). This compels researchers to find more effective and useful biomarkers for diagnosis, prognosis and further clarification of the unknown pathological aspects of lung cancer. Ideally, the identification and better characterization of more easily accessible and cost-effective biomarkers could revolutionize the diagnoses and therapies for these types of cancer.

Proteins secreted by cancer cells are potentially useful candidates for circulating tumor markers. Although assays, predominantly ELISA, evaluating a number of clinically relevant proteins in serum have been established, several of which have been reported to be useful in the diagnosis and monitoring of advanced cancers, such as CEA, NSE and CYFRA21-1 (5–7), these assays have not been suited to use for mass screening or diagnosis in the early stages of cancer, due to their relatively low sensitivity/specificity and high costs.

The trefoil factor (TFF) family, which comprises gastric peptide pS2/TFF1, spasmolytic peptide/TFF2 and intestinal trefoil factor/TFF3, are expressed and secreted in almost all mucus-secreting cells in mucosal surfaces, predominantly those of the gastrointestinal tract (8,9). TFFs are small (7–12 kDa) protease-resistant proteins with a conserved ‘trefoil’ domain, which contains 3 disulfide bonds that provide functional stability (8,10). TFFs can be detected as monomeric, homo-dimeric or hetero-dimeric forms in secretions and serve an essential role in epithelial restitution and mucosal protection (8,10).

Individual TFFs have been characterized as follows. In vitro experiments using cultured cells have reported that TFF1 serves a pivotal role in enhanced growth, survival, migration and invasion of pancreatic and colonic cancer cells (11–13). Similarly, overexpression of TFF3 has been reported to drive proliferation, migration and invasion, and the angiogenic and antiapoptotic capacities of breast and colonic cancer cells (13–16). Notably, high TFF expression levels have been reported clinically in breast (TFF1 and 3) (17), lung (TFF1 and 3) (18,19), pancreas (TFF2) (20) and colon cancers (TFF3) (21). Moreover, expression of TFF2 has been reported to be predictive of a more aggressive phenotype in gastric cancer (22,23) and TFF3 in gastric (23–25), colorectal (25,26) and breast cancer (25,27). However, loss of TFF1 is carcinogenic in mice (10,28), and the region of chromosome 21q22.3 in which the 3 TFF genes are clustered is frequently deleted in human gastric cancer (10). In cancer cell lines, exogenous TFF1 consistently reduced cell proliferation by delaying the G1 to S phase transition in human colonic and gastric carcinoma cells (29). Similarly, in a previous study, it was shown that TFF1 can suppress cell proliferation, survival, migration and invasion in lung carcinoma cells (30). Notably, the expression of TFF1 is also predictive of improved survival time for patients with breast cancer (27,31). Moreover, TFF3 was also reported to inhibit the growth of colorectal cancer cells, in a cell-type dependent manner (32). In clinical specimens, TFF3 expression was reported to be lower in colorectal cancer compared with normal tissue (33). These observations indicate that TFF1 and TFF3 can also function as tumor suppressor-like proteins.

Serum TFFs (sTFFs) have been sporadically reported as possible biomarkers in cancer screening: TFF1 and TFF3 for breast (17), gastric (34,35), colorectal (36) and endometrial cancer (37), and all TFFs for prostate cancer (38). Moreover, sTFF1 and sTFF3 have been described as possible prognostic markers in endometrial (37), gastric (39) and colorectal cancer (40,41). However, the utility of TFF3, as suggested by its high serum concentration in patients with lung cancer regardless of histological type (42), has yet to be established as a clinical diagnostic tool. Moreover, the potential correlation between serum and/or urinary TFFs (uTFFs) and tissue expression has yet to be assessed.

To extend previously reported in vitro analysis (30) into potential in vivo clinical analyses, a comprehensive survey of all TFFs in both serum and urine was conducted in the present study. uTFF analysis was of particular interest as this is a completely non-invasive diagnostic method and is potentially suitable for mass screening. Overall data were evaluated clinicopathologically and compared to protein expression levels of TFFs in cancer tissues.

Materials and methods <sec> <title>Patients

Patients' clinicopathological profiles, including tumor stage by TNM classification (43) were presented in Table I and further details were presented in Table SI. The study population consisted of 199 patients according to the following inclusion criteria; those with lung cancer, without history of malignant tumor nor history of other malignant tumor, who had undergone surgery at the Saitama Medical Center at Jichi Medical University (Saitama, Japan) from January 2017 to Jul 2019. These patients included of 142 AC, 45 squamous cell carcinoma (SCC), 4 large cell carcinoma (LCC), 4 small cell carcinoma (SmCC) and 4 pleomorphic carcinoma cases. Serum and urine were collected from 198 healthy individuals (Table SII), whose ages (range, 38–75 years old; mean, 67.5 years old) and sex (127 males and 71 females) were matched as closely as possible with the patients with lung cancer (range, 32–93 years old; mean, 69.3 years old; 128 males and 71 females). Patients or healthy individuals were excluded from the study if they had a history of malignancy, pre-operative chemotherapy and/or diabetes mellitus. Individuals with a Helicobacter pylori infection were also excluded from the study as this infection has been reported to enhance sTFF levels (35).

All patients and healthy individuals were confirmed to have normal urinary creatinine levels, and all uTFFs were normalized to urinary creatinine to avoid variations in urinary protein concentration.

ELISA

A total of 0.5 ml of serum and 2 ml of morning urine were collected from each patient after overnight fasting. The samples were centrifuged at 5,000 × g for 15 min at 0°C and the supernatant collected. All serum and urine samples were frozen at −80°C before use. Polyclonal anti-sera were raised in our laboratory, in rabbits immunized with recombinant human TFF1, TFF2 and TFF3 which had also been prepared in our laboratory as previously described (32). The IgG fraction was affinity purified using protein A into a concentrated solution of 1 mg/ml. Each antibody solution was diluted with PBS to a 1/200 working solution and was coated onto 96-well microtiter plates. Concentrations of serum or urine TFF1, TFF2 and TFF3 were measured using these in-house ELISA plates generated by coating with antibodies against TFF1, TFF2 or TFF3, as previously reported (17,35). Detection sensitivities were 7.0 pg/ml for TFF1, 30.0 pg/ml for TFF2 and 30.0 pg/ml for TFF3. It has been previously reported that each TFF antibody reacted specifically and no cross reactivity with other TFFs was reported (17,35). Appropriate sample dilutions were determined to place samples within the linear range of the calibration curve, generated using recombinant human TFF1, 2 or 3 as previously described (35). Absorbance at 450 nm was measured on a DTX-880 plate reader (Beckman Coulter, Inc.) and sample TFF concentrations were calculated from the calibration curves. There was no difference in the detection quality between serum and urine TFFs.

Tissues and immunohistochemistry (IHC)

Of the 199 samples of primary lung carcinoma obtained by surgery in the Saitama Medical Center, 100 samples which had been preserved and fixed at a high quality, and demonstrated no necrosis or degeneration and positive staining in bronchial gland cells, were prepared for IHC by paraffin section. These 100 samples consisted of 73 AC, 23 SCC, 1 LCC, 2 SmCC and 1 pleomorphic carcinoma case (Table I). Tissues from pathological lesions, together with adjacent normal tissues were fixed in 10% formaldehyde at room temperature for 24 h. Tissues were sliced into small pieces, embedded in paraffin and cut into 3.5 µm sections. Heat-induced epitope was performed using PT Link (cat. no. PT100/PT101; Agilent Technologies, Inc.) and EnVision^TM FLEX Target Retrieval Solution, High pH (pH 9.0; cat. no. S2367; Agilent Technologies, Inc.) for TFF1 and TFF3, and Low pH (pH 6.0; cat. no. #S2031; Agilent Technologies, Inc.) for TFF-2 at 95°C for 30 min. As the antibodies used for ELISA did not react in paraffin-embedded tissues, commercialized primary antibodies as follows were used: Anti-TFF1 rabbit polyclonal (1:350; cat. no. PA5-31863; Invitrogen; Thermo Fisher Scientific, Inc.), anti-TFF2, rabbit polyclonal (1:50; cat. no. ab131147; Abcam) and anti-TFF3 rabbit monoclonal (1:300; cat. no. ab108599; Abcam). Sections were incubated with the blocking solution included in the kit for 20 min at room temperature, with primary antibodies at 4°C overnight and with secondary antibodies for 20 min at room temperature and visualization was performed with a Catalyzed Signal Amplification System II kit which included secondary antibodies (cat. no. K1497 and K1501; Agilent Technologies, Inc.). A total of 5 high magnification fields of view were randomly selected from each section and 200 cells were counted per field. Images were digitally captured using a BX51 bright-field microscope (Olympus Corporation) equipped with a DP-72 digital camera in conjunction with the cellSens imaging software (version 1.18, Olympus Corporation). Adobe Photoshop was used for image processing (CS4 Extended; Adobe Systems Europe, Ltd.)

IHC score was determined semi-quantitatively by multiplication of the ‘positive fraction’ with the ‘intensity-score’ according to the following system. Firstly, the positive fraction was sub-divided as follows: i) 0, no staining; ii) 1, <10% staining; iii) 2, 10–50% staining; and iv) 3, >50% staining of cells with intensity score >0. Secondly, intensity-score was determined as follows: i) 0 if there was no staining or staining was weaker than that in non-neoplastic cells; ii) 1 if the staining was the same as in non-neoplastic cells; and iii) 2 if there was more intense staining than in the non-neoplastic cells. An IHC score >0 was defined as ‘positive’ (44). Staining was evaluated by two observers and discordance was resolved by discussion.

Statistical analysis

Statistical analyses were performed using the JMP software package (version 11; SAS Institute, Inc.) and StatMate IV (GraphPad Software; Dotmatics). Sufficient sample sizes were confirmed and statistics were calculated with a 0.80 power at the 0.05α level using EZR 1.54 (https://www.softpedia.com/get/Science-CAD/EZR.shtml).

The Receiver operating characteristics (ROC) sensitivity and specificity of serum or urine TFFs were plotted and the figures were generated using EZR 1.54 (https://www.softpedia.com/get/Science-CAD/EZR.shtml). Area under the curve (AUC) was calculated to determine the marker cut-off points and the discriminatory potential of each TFF, as well as each combination of TFFs. Logistic regression modeling was used to estimate the odds ratio with 95% confidence interval (CI) to construct a final composite score. Validation of analysis was performed using Fisher's exact test in all sample groups, including those from patients with lung cancer and from healthy individuals, with the cut-off value (threshold) showing the highest AUC (combination of sTFF1/2/3 + uTFF1/3). Observers' agreement in the evaluation of IHC results was analyzed using κ statistics as follows: i) 0, poor agreement; ii) 0<κ<0.2, slight agreement; iii) 0.2<κ<0.4, fair agreement; iv) 0.4<κ<0.6, moderate agreement; v) 0.6<κ<0.8, substantial agreement; and vi) 0.8<κ<1.0, almost perfect agreement. The Spearman's rank correlation coefficient (ρ) was used as a measure of correlation between levels of each TFF assessed using ELISA or between IHC score and TFFs levels. Results were categorized as follows: i) no correlation, ρ=0; ii) equivocal, ρ<0.2; iii) low, 0.2<ρ<0.4; iv) substantial, 0.4<ρ<0.7; v) high, 0.7<ρ<1.0; and vi) complete, ρ=1.0. The difference in TFF levels among clinicopathological variables were analyzed using Mann-Whitney U test between two groups or Kruskal-Wallis test followed by Dunn's test among multiple groups. All tests were two-sided and P<0.05 was considered to indicate a statistically significant difference.

Results <sec> <title>Serum and urine TFFs

Concentrations of all TFFs in serum and urine were measured using ELISA. All raw results from patients with lung cancer and from healthy individuals are presented in Tables SI and SII, respectively. sTFF1, sTFF2 and sTFF3 concentrations in patients with lung cancer were all significantly higher compared with those in healthy individuals (P<0.0001; Fig. 1A). uTFF1 and uTFF3 values normalized using the creatinine level were lower in patients with lung cancer (P=0.0012 and P<0.0001, respectively; Fig. 1B).

Based on the results of upregulated sTFFs and downregulated uTFFs in patients with lung cancer, the precise diagnostic potential of either source alone or in combination was explored to evaluate how well these levels in patients with lung cancer or healthy individuals could be differentiated. The ability to detect and segregate patients with lung cancer by ROC analysis was assessed, which generated a graphical plot of the sensitivity vs. specificity of TFF levels. AUC and sensitivity or specificity at the optimal cut-offs are presented in Fig. 2. For serum, the AUC/cut-off values for TFF1, TFF2 and TFF3 were 0.709/0.837, 0.722/3.702 and 0.663/7.211 ng/ml, respectively (Fig. 2A). With these cut-off values, the sensitivity/specificity for discriminating between patients with lung cancer and healthy individuals were calculated as 0.677/0.702 in TFF1, 0.582/0.778 in TFF2 and 0.572/0.768 in TFF3. For urine, the AUC/cut-off values for TFF1, TFF2 and TFF3 were 0.594/1.287, 0.501/18.174 and 0.665/4.162, respectively (Fig. 2B). The resultant sensitivity/specificity values were 0.688/0.495 for TFF1, 0.642/0.455 for TFF2 and 0.547/0.737 for TFF3.

The diagnostic potential of TFF data in combination was also evaluated. Among the AUCs calculated, the highest value was for the combination of sTFF1 + 2 + 3 with uTFF1 + 3, which was 0.826 with a cut-off value of 0.443 (Fig. 2A). Therefore, while any single TFF discriminated patients with lung cancer, this combination of TFFs demonstrated an improved diagnostic performance.

Validation of TFFs for their potential in diagnostic screening

Instead of dividing the clinical sample groups into training and validation sets, the utility of TFFs as a biomarker for baseline screening was evaluated in all 397 cases of both patients with lung cancer and healthy individuals (Tables SI and SII). A cut-off value of 0.443 was used, which showed the highest AUC by the combination of sTFF1 + 2 + 3 and uTFF1 + 3. Cases were divided into 4 categories as shown in the cross-tabulation table (Table II) and the differences were statistically analyzed. With this cut-off value, the sensitivity and the specificity of patients with lung cancer and healthy individuals were 0.804 and 0.742, respectively, and the positive and negative predictive values were 0.758 and 0.790, respectively. The difference was determined to be statistically significant using Fisher's exact test (P<0.0001).

Expression of TFFs by IHC

In parallel to the ELISA, TFF expression by IHC was also analyzed. Inter-observer agreement in the evaluation of staining results ranged from ‘substantial agreement’ to ‘almost perfect agreement’ (TFF1, κ=0.837; TFF2, κ=0.791; TFF3, κ=0.830). Representative IHC results are presented in Fig. 3, with staining results detailed in Table III, and the results of individual patients presented in Table SIII. All TFFs were expressed in the cytoplasm of bronchial glands and occasionally in bronchial epithelial cells, but not in the alveolar epithelial cells or interstitial cells in normal tissue. Among the 100 cancer cases analyzed, TFF1, TFF2 and TFF3 were positively stained in 41 (41.0%), 37 (37.0%) and 50 cases (50.0%), respectively. TFF1 was expressed in the cytoplasm of 29/73 (39.7%) and 12/23 (52.2%) cases of AC and SCC, respectively, but in no cases of SmCC, LCC or pleomorphic carcinoma. IHC scores of TFF1 staining were categorized into the following results: score 6, 3 cases; score 4, 8 cases; score 3, 10 cases; score 2, 13 cases; score 1, 7 cases; score 0, 59 cases. TFF2 was expressed in the cytoplasm of 25/73 (34.2%), 10/23 (43.5%), 2/2 (100%) and 0/2 cases (0%) of AC, SCC, SmCC and others (pleomorphic carcinoma and large cell carcinoma), respectively. TFF3 was expressed in the cytoplasm of 38/73 (52.1%), 12/23 (52.2%) cases of AC and SCC, respectively and none (0/4) of others (Table III). Overall, 83 out of 100 (83.0%) cases were positive for at least one of TFF1, TFF2 and TFF3 when assessed by IHC staining. There were no specific microscopic staining patterns, such as co-expression or reciprocal expression, among these three proteins as a whole.

Correlation between sTFFs and uTFFs

Correlations between serum and urine TFFs levels were determined using Spearman's rank correlation test. As shown in Table IV, the correlation coefficient (ρ) was generally higher between sTFF1 and sTFF2 in both the patients with lung cancer and the healthy individuals. The ρ between sTFF1 and sTFF2 was 0.494 in patients and 0.473 in healthy individuals (deemed ‘substantial correlation’; both P<0.001). Furthermore, there was a significant positive correlation between sTFF1 and sTFF2 in all groups of pT pathological tumor stage (pT) or pathological nodal stage (pN) categories (except pT4) as well as in AC, which demonstrated ρ-values ranging from 0.448-0.717. Most of the ρ-values were within the range of ‘substantial correlation’, while the pN1 group had a score of 0.717, which was categorized as a ‘high correlation’. In healthy individuals, all uTFF1, uTFF2 and uTFF3 values were correlated with each other in contrast to the patients with lung cancer. In the patients with lung cancer, the positive correlation between TFF1 and TFF2 was weaker in urine and the ρ-values between uTFF1 and uTFF2 were within the range of ‘substantial correlation’ in pT2, pT4 (marginally) and pN1-2 groups. uTFF1 and uTFF3 showed ‘high correlation’ in pT4 and ‘substantial correlation’ in pN1/2 groups of lung cancer cases, but uTFF2 and uTFF3 correlated only in pT4 with ρ=0.736, although it indicated a ‘high correlation’. Between sTFF and uTFF, a ‘substantial correlation’ was observed for all TFFs (TFF1, 0.516; TFF2, 0.463; TFF3, 0.560; P<0.0001 for all; Table V). No inverse correlation was found between each sTFF or uTFF.

Correlation of sTFFs and uTFFs with IHC scores

The correlations between sTFF and uTFF levels and IHC-determined TFF expression levels in lung cancer tissues were examined, the results of which are presented in Table V. A statistically significant correlation between sTFF1-3 levels and IHC scores was demonstrated. Using Spearman's correlation test, sTFF1 and sTFF3 showed ρ-values of 0.563 and 0.813 with the IHC scores, respectively (‘substantial’ or ‘high correlation’, respectively; P<0.0001), whereas TFF2 had a ρ-value of −0.657 (‘substantial inverse correlation’; P<0.0001). Moreover, analysis using Kruskal-Wallis test followed by Dunn's test, groups with a higher IHC score generally showed higher sTFF1 and sTFF3 at statistically significant levels (P<0.0001; Table VI). However, a significant inverse association was confirmed for sTFF2. Analysis of urine data using Spearman's test, demonstrated no correlation with IHC score (TFF1, ρ=0.197; TFF2, ρ=−0.281; TFF3, ρ=0.342; Table V).

Association of sTFFs and uTFFs with clinicopathological features

Next, sTFF and uTFF levels were compared with clinicopathological variables (Tables VI and VII, respectively). In the pT category, the pTis/1/2 group showed a higher sTFF2 mean value than the pT3/4 group, at a statistically significant level (P=0.018). Furthermore, the pN0/1 group had a higher mean sTFF1 value than the pN2 group, at a statistically significant level (P=0.033). Histologically, although SCC cases had a significantly higher mean uTFF1 value than AC (P=0.027), AC had a higher mean level of uTFF2, but not at the level of statistical significance (P=0.062). No significant difference in TFF levels was demonstrated for any other factors.

Discussion

The identification of novel diagnostic markers is valuable not only because it enables the early detection of disease, but also as it may allow post-operative or post-chemotherapeutic monitoring of patients.

The involvement of TFFs in cancer has been previously described, and their normal biological functions have been shown to be quite diverse. Aberrant expression of TFFs, in particular, of TFF1 and TFF3, has been observed in numerous types of clinical cancer, including those of the breast and the lung (8,45). Furthermore, higher expression of TFF1 and/or TFF3 has been reported to be associated with chemoresistance (46,47), lymph node/distant metastasis, high Tumor-Node-Metastasis stage and a poor prognosis in colorectal (13,48), gastric (24) and endometrial cancer (37).

sTFFs have been sporadically reported as cancer biomarkers in the literature (17,34,36–38,49,50). In particular, sTFF1 and sTFF3 have been described as predictive markers for lymph node metastasis and for worse prognosis in prostatic (38), gastric (39) and colorectal cancer (46,48,51). Consistently, decreased sTFF3 levels were significantly associated with positive responses to chemotherapy in both gastric and colorectal cancer (52).

In lung cancer, sTFF1 has been described as a marker of AC, and the level of sTFF1 was reported to have normalized following the removal of tumors in a small sample group, although these values did not reach levels of significance as a screening tool for tumors (53). However, sTFF3 has been reported to be elevated in all histological types of lung cancer (42). Overall, as there have not been many reports on the involvement of TFFs in lung cancer, conclusions are somewhat equivocal. In particular, changes in sTFF and uTFF levels have not previously been investigated in precise detail. Therefore, the utility of TFFs in diagnosing lung cancer is still controversial. In the present study, the utility of TFFs as serum or urinary biomarkers were evaluated with additional analyses of the relationship between TFF expression in lung cancer tissue and clinicopathological factors.

Firstly, all sTFF levels were significantly higher in patients with lung cancer (P<0.0001), while uTFFs were lower. In the latter, the results were statistically significant for uTFF1 and 3, but not for uTFF2. Using ROC analysis, TFFs successfully discriminated patients with lung cancer from healthy individuals and an analysis of combined sTFFs and uTFFs gave the highest AUC of 0.826. Currently established markers for lung cancer have shown satisfactorily high AUCs, with carcinoembryonic antigen (CEA) providing an AUC up to 0.850 in AC (5), CYFRA21-1 and ‘SCC-antigen’ providing AUCs up to 0.930 and 0.780, respectively in SCC (5), and neuron-specific enolase (NSE) and pro-gastrin-releasing peptide providing AUCs up to 0.890 and 0.950, respectively in SmCC (7). However, the AUCs for these markers evaluated against all histological types as a whole were far lower. The combination of CYFRA/CEA/NSE/‘SCC-antigen’ for combined lung cancer cases demonstrated an AUC of 0.854 compared with non-cancer cases (6). the AUC obtained in the present study by combining 5 values from 3 proteins, 2 of which were from urine, was equivalent to that value regardless of histological type. Indeed, the screening and diagnostic potential of sTFF1 + 2 + 3 with uTFF1 + 3 was verified in this sample group, when assessed as a whole (Table II).

As another detection tool, low-dose helical CT (LDCT) is a sensitive screening method and demonstrates high performance in the detection of pulmonary nodules and could therefore detect a high proportion of small cancers at an early stage in baseline screening. For the detection of cancer, the highest reported sensitivity was 96%, but with the lowest sensitivity at 49% (54). However, the highest reported specificity was 95%, but with the lowest sensitivity at 55% (55) for the detection of cancer among all pulmonary nodules, including non-neoplastic lesions. The sensitivity/specificity in the ELISA system used in the present study was 74.2/80.9%; both values being in the middle of those previously reported highest and lowest scores. Moreover, LCDT has a consistently high false-positive rate (≤98%) in single baseline examination, depending on the entry and imaging diagnostic criteria (56). For example, inflammatory scarring, hamartoma and fungal granuloma are major factors that could contribute to a high false-positive rate (56,57). This is a major limitation of LDCT as a method for screening. Accordingly, repeated examinations are necessary with this method and in diagnoses, baseline examination must be re-evaluated by further investigation, such as follow-up with detailed (high resolution, thin section) CT, usually repeated 3–4 times over 2–4 years (54,57). Furthermore, results of randomized trials have not yet confirmed that LDCT screening has a measurable effect in decreasing mortality (56,58). Overall, the ELISA screening method utilized in the present study is more convenient and thus, has the potential to be more prognostic at baseline screening than results reported using LDCT, to date.

In the patients with lung cancer analyzed in the present study, TFF levels were higher in serum but lower in urine. It was hypothesized that TFFs may be present in a modified form in serum, including as homo- or heterodimers, and thus, are not rapidly excreted into the urine through normal turnover in patients with lung cancer. In healthy individuals, all TFFs are present in un-modified form in serum and are rapidly excreted into the urine through normal turnover. Supporting the aforementioned hypothesis, all uTFF levels showed ‘substantial’ to ‘high’ correlation with each other in healthy individuals, i.e., all TFF were coordinately excreted. However, such a correlation was not found among patients with lung cancer as a whole (Table IV).

Secondly, the observation that sTFF2 levels were higher in the pTis/1/2 group compared with the pT3/4 group and sTFF1 was higher in the pN0/1 group compared with the pN2 group suggested that upregulation of TFF1and TFF2 was an early and transient phenomenon during tumor growth and nodal metastasis in lung carcinoma. Moreover, patients with SCC showed higher uTFF1 levels, but patients with AC showed higher uTFF2 levels. However, the significance and the mechanism of such differences depending on the histological type are unclear.

Thirdly, TFF1, TFF3 and, less intensely, TFF2 demonstrated positive IHC staining in normal bronchial epithelial cells and bronchial glands in the lung in the present study, as has been previously described (59,60). All TFFs were detected in cancer tissues of AC, SCC and SmCC, with a slightly higher frequency in SCC: i.e., 43.5-52.2% in SCC vs. 34.2-52.1% in AC and none in the other subtypes, except for SmCC in which both cases expressed TFF2 (IHC score of 2). Although AC, SCC and a minor fraction of SmCC have been reported to express TFF1 (18,61), the present study was, to the best of our knowledge, the first report of IHC expression of all TFFs in all histological types.

Fourthly, in a previous report on breast carcinoma, TFF1, but not TFF3, was reported to show a positive correlation between serum concentration and IHC staining (17). sTFF2 was lower in patients with cancer in that study, however, a statistical analysis could not be performed due to the absence of TFF2-positive cases as assessed using IHC (17). However, in the present study, TFF1 and TFF3 were demonstrated to show a positive correlation between serum concentrations and IHC score, and a ‘substantial’ inverse correlation for TFF2 by Spearman's test. These results for TFF1 and TFF3 were also demonstrated to be statistically significant, ie, the concentration of sTFF1 and sTFF3 was higher in groups having a higher IHC score. There are numerous possible mechanisms to explain the inverse correlation between high sTFF2 and low IHC score. Aberrantly overexpressed TFF2 may be secreted into the bronchial lumen, but not into serum. For example, in well differentiated breast cancer, polarized TFF3 expression leads to secretion into the ductal lumen, but not into serum (27). Alternatively, a major fraction of elevated sTFF2 may not derive directly from cancer, but from the mucus-secreting cells of bronchial glands. Lastly, TFF2 may be exhausted in cancer cells after increased secretion into the serum even if it is aberrantly overexpressed. Collectively, sTFFs may be regulated at numerous levels in a complex manner, such as expression in the cells, secretion into the lumen or the serum and excretion into the urine.

Functionally, TFF1 in the stomach has been described as being involved in the formation of a protective mucous layer (62) and stabilization of cell junctions by binding to MUC2, MUC5AC (8,63) and gastrokine-2 (GKN2) (64). TFF2 also binds to GKN2 (64) and MUC6 (65), and maintains the inner layer of the gastric mucus (66). TFF3 is a constituent of goblet cells in both the small and large intestine, but not in the stomach, and co-localizes with MUC5B and MUC8 in bronchial gland mucous cells (10,60), and with MUC5AC in respiratory goblet cells (60,67). Therefore, in the lungs, TFFs bind with mucin from the bronchial gland, in a manner similar to the stomach, and thereby may maintain cell adhesion and support restitution. In cancer, it is not surprising that TFFs are secreted from the bronchial glands and protect the adjacent bronchial mucosa from deleterious damage as a safe-guard mechanism. Indeed, the induction of mucinous metaplasia was reported in transgenic mice with high TFF3 expression (68). Therefore, by physical and/or chemical stimuli, TFF actively serves a role in enhancing the production of mucin, which contains its binding partners.

Previous studies suggested mutual regulation among TFFs. TFF2^KO mice were reported to have shown high TFF3 expression in the gastric antrum (69). In a rat model of colitis, TFF1 expression was elevated and TFF3 was decreased in the colon during the acute phase of disease (70), whereas the opposite pattern of expression was demonstrated in the restitution phase (70,71). Hence, a coordinated regulation of TFF expression to ensure mucosal protection and restitution depending on the stage of the disease may exist. In the present study, although sTFF1 and sTFF2 levels showed ‘substantial’ to ‘high’ correlation in patients with lung cancer as a whole or in each pTN category, concurrent relative suppression, i.e., inverse correlation, of sTFF3 was not found. Moreover, although the staining frequencies and intensities of TFF1 and TFF3 were predominant compared with TFF2, no specific pattern of staining, such as exact overlapping or reciprocal staining of each TFF, was observed.

In contrast with other cancer-specific markers, TFFs are produced not only by cancer cells, but also by non-neoplastic cells as components of tissue-protective mucin (8) or regulators of inflammation (72). Therefore, although the clinical utility of circulating and excreting TFFs as cancer biomarkers for the diagnostic screening of lung cancer may be cautiously evaluated, increased TFFs levels also reflects epithelial damage. Regardless of the underlying mechanism, results from the present study demonstrated that discrimination between patients with lung cancer and healthy individuals using TFFs was possible, and that all TFFs assessed provided unique biomarker information that could contribute to a panel of markers having significant diagnostic potential. However, a limitation of the present study was the short period of patient follow-up, and thus the long-term prognostic power of TFFs could not be precisely evaluated.

Based on the results of the present study, it is proposed that this combined serum/urine assay of TFFs represents a novel, easily monitored tool for accurately diagnosing lung cancer, with TFF1 and TFF2 levels in particular showing good correlation with clinicopathological factors. This assay may permit larger-scale cancer screening, particularly in the early stages of disease. Considering the limitations of the present work, further studies to clarify how sTFF and uTFF levels change after surgery and to clarify any correlation with relapse and metastases, as well as detailed analysis in relation to smoking history are required. If these efforts are successful, this s/uTFFs analysis may further contribute to cancer management and provide a tool to monitor the course of the disease after surgery and to evaluate the effects of adjuvant therapy.

Supplementary Material

Supporting Data

Acknowledgements

The authors would like to thank Ms Emi Kimura (Department of Pathology, International University of Health and Welfare Hospital, Nasushiobara, Japan), for their assistance with manuscript preparation.

Availability of data and materials

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

Authors' contributions

YD and SN designed the study. KM and YD performed immunohistochemical staining and evaluated the results. KM, YI, MF and TY performed sample preparation and ELISA and contributed to the quantification. TK, HT, YO and SN collected the clinical samples. KM, TK and HT statistically analyzed the data and undertook its interpretation. YI, MF and YO established the ELISA experimental system and provided technical support. KM, YD, TK, HT and SN drafted and completed the manuscript. KM, YD and TK confirm the authenticity of all the raw data. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

This study protocol adhered to the principles of the Declaration of Helsinki and was approved by the Institutional Ethics Committee in Saitama Medical Center, Jichi Medial University (Saitama, Japan; approval no. S17-035, S20-137), the University of Tokyo (Tokyo, Japan; approval no. 11414-[1,2]) and Kyoto Prefectural University of Medicine (Kyoto, Japan; approval no. ERB-C-1239-1,2). Written informed consent was obtained from all participants.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Abbreviations

TFF

trefoil factor

adenocarcinoma

SCC

squamous cell carcinoma

LCC

large cell carcinoma

SmCC

small cell carcinoma

References 1

Siegel

Miller

Jemal

Cancer statistics, 2019

CA Cancer J Clin697342019

10.3322/caac.21551

30620402

Sharma

Bell

Settleman

Haber

Epidermal growth factor receptor mutations in lung cancer

Nat Rev Cancer71691812007

10.1038/nrc2088

17318210

Arcila

Rekhtman

Sima

Zakowski

Pao

Kris

Miller

Ladanyi

Riely

Analysis of tumor specimens at the time of acquired resistance to EGFR-TKI therapy in 155 patients with EGFR-mutant lung cancers

Clin Cancer Res19224022472013

10.1158/1078-0432.CCR-12-2246

23470965

Siegel

Miller

Fuchs

Jemal

Cancer statistics, 2022

CA Cancer J Clin727332022

10.3322/caac.21708

35020204

Chen

Huang

Zhu

Assessment of seven clinical tumor markers in diagnosis of non-small-cell lung cancer

Dis Markers201898451232018

10.1155/2018/9845123

30647803

Wen

Huang

Ling

Chen

Wei

Tang

Wen

Deng

Lack of efficacy of combined carbohydrate antigen markers for lung cancer diagnosis

Dis Markers202047167932020

10.1155/2020/4716793

33488842

Lamy

Grenier

Kramar

Pujol

Pro-gastrin-releasing peptide, neuron specific enolase and chromogranin A as serum markers of small cell lung cancer

Lung Cancer291972032000

10.1016/S0169-5002(00)00113-6

10996422

Kjellev

The trefoil factor family-small peptides with multiple functionalities

Cell Mol Life Sci66135013692009

10.1007/s00018-008-8646-5

Thim

Trefoil peptides: From structure to function

Cell Mol Life Sci538889031997

10.1007/s000180050108

Xiao

Ling

Lan

Liu

Yang

Trefoil factors: Gastrointestinal-specific proteins associated with gastric cancer

Clin Chim Acta4501271342015

10.1016/j.cca.2015.08.004

Arumugam

Brandt

Ramachandran

Moore

Wang

May

Westley

Hwang

Logsdon

Trefoil factor 1 stimulates both pancreatic cancer and stellate cells and increases metastasis

Pancreas408158222011

10.1097/MPA.0b013e31821f6927

21747314

Rodrigues

Rodrigue

Attoub

Flejou

Bruyneel

Bracke

Emami

Gespach

Induction of the adenoma-carcinoma progression and Cdc25A-B phosphatases by the trefoil factor TFF1 in human colon epithelial cells

Oncogene25662866362006

10.1038/sj.onc.1209665

16715141

Yusufu

Shayimu

Tuerdi

Fang

Wang

TFF3 and TFF1 expression levels are elevated in colorectal cancer and promote the malignant behavior of colon cancer by activating the EMT process

Int J Oncol557898042019

Kannan

Kang

Kong

Tang

Perry

Mohankumar

Miller

Liu

Mertani

Zhu

Trefoil factor 3 is oncogenic and mediates anti-estrogen resistance in human mammary carcinoma

Neoplasia12104110532010

10.1593/neo.10916

Pandey

Zhang

Zhu

Lobie

Trefoil factor 3 promotes metastatic seeding and predicts poor survival outcome of patients with mammary carcinoma

Breast Cancer Res164292014

10.1186/s13058-014-0429-3

25266665

Taupin

Kinoshita

Podolsky

Intestinal trefoil factor confers colonic epithelial resistance to apoptosis

Proc Natl Acad Sci USA977998042000

10.1073/pnas.97.2.799

10639160

Ishibashi

Ohtsu

Ikemura

Kikuchi

Niwa

Nishioka

Uchida

Miura

Aikou

Gunji

Serum TFF1 and TFF3 but not TFF2 are higher in women with breast cancer than in women without breast cancer

Sci Rep748462017

10.1038/s41598-017-05129-y

28687783

Ruibal

Nunez

del Rio

Garcia Diez

Rodriguez

Alvarez De Linera

Cytosolic pS2 levels in 154 non-small-cell lung carcinomas. Correlation with other clinical and biological parameters

Rev Esp Med Nucl211091142002(In Spanish)

10.1016/S0212-6982(02)72044-2

Amelung

Buhrens

Beshay

Reymond

Key genes in lung cancer translational research: A meta-analysis

Pathobiology7753632010

10.1159/000278292

20332665

Ohshio

Suwa

Kawaguchi

Imamura

Yamaoka

Yamabe

Matsumoto

Yoshioka

Hashimoto

Takeda

Differential expression of human spasmolytic polypeptide (trefoil factor family-2) in pancreatic carcinomas, ampullary carcinomas, and mucin-producing tumors of the pancreas

Dig Dis Sci456596642000

10.1023/A:1005471005289

John

El-Rouby

Tomasetto

Rio

Karam

Expression of TFF3 during multistep colon carcinogenesis

Histol Histopathol227437512007

Dhar

Wang

Maruyama

Udagawa

Kubota

Fuji

Tachibana

Ono

Otani

Nagasue

Expression of cytoplasmic TFF2 is a marker of tumor metastasis and negative prognostic factor in gastric cancer

Lab Invest83134313522003

10.1097/01.LAB.0000085190.99749.15

Dhar

Wang

Tabara

Tonomoto

Maruyama

Tachibana

Kubota

Nagasue

Expression of trefoil factor family members correlates with patient prognosis and neoangiogenesis

Clin Cancer Res11647264782005

10.1158/1078-0432.CCR-05-0671

16166422

Yamachika

Werther

Bodian

Babyatsky

Tatematsu

Yamamura

Chen

Itzkowitz

Intestinal trefoil factor: A marker of poor prognosis in gastric carcinoma

Clin Cancer Res8109210992002

12006524

Yang

Lin

Bei

Guo

Pathological and therapeutic roles of bioactive peptide trefoil factor 3 in diverse diseases: Recent progress and perspective

Cell Death Dis13622022

10.1038/s41420-022-00849-1

35039476

Babyatsky

Lin

Yio

Chen

Zhang

Zheng

Twyman

Bao

Schwartz

Thung

Trefoil factor-3 expression in human colon cancer liver metastasis

Clin Exp Metastasis261431512009

10.1007/s10585-008-9224-9

18979216

Ahmed

ARH

Griffiths

Tilby

Westley

May

FEB

TFF3 is a normal breast epithelial protein and is associated with differentiated phenotype in early breast cancer but predisposes to invasion and metastasis in advanced disease

Am J Pathol1809049162012

10.1016/j.ajpath.2011.11.022

Buache

Etique

Alpy

Stoll

Muckensturm

Reina-San-Martin

Chenard

Tomasetto

Rio

Deficiency in trefoil factor 1 (TFF1) increases tumorigenicity of human breast cancer cells and mammary tumor development in TFF1-knockout mice

Oncogene30326132732011

10.1038/onc.2011.41

21358676

Bossenmeyer-Pourie

Kannan

Ribieras

Wendling

Stoll

Thim

Tomasetto

Rio

The trefoil factor 1 participates in gastrointestinal cell differentiation by delaying G1-S phase transition and reducing apoptosis

J Cell Biol1577617702002

10.1083/jcb200108056

Minegishi

Dobashi

Tsubochi

Hagiwara

Ishibashi

Nomura

Nakamura

Ohmoto

Endo

TFF-1 functions to suppress multiple phenotypes associated with lung cancer progression

Onco Targets Ther14476147772021

10.2147/OTT.S322697

34531663

Thompson

Elton

Hawkins

Chetty

Steel

PS2 mRNA expression adds prognostic information to node status for 6-year survival in breast cancer

Br J Cancer774924961998

10.1038/bjc.1998.78

9472649

Uchino

Kataoka

Itoh

Hamasuna

Koono

Overexpression of intestinal trefoil factor in human colon carcinoma cells reduces cellular growth in vitro and in vivo

Gastroenterology11860692000

10.1016/S0016-5085(00)70414-8

10611154

Espinoza

Agarwal

Reddy

Shenoy

Subramony

Sakiyama

Fair

Poosarla

Zhou

Shannon Orr

Expression of trefoil factor 3 is decreased in colorectal cancer

Oncol Rep452542642021

10.3892/or.2020.7829

33210724

Huang

Zhang

Wang

Zhang

Ding

Serum trefoil factor 3 is a promising non-invasive biomarker for gastric cancer screening: A monocentric cohort study in China

BMC Gastroenterol14742014

10.1186/1471-230X-14-74

Aikou

Ohmoto

Gunji

Matsuhashi

Ohtsu

Miura

Kubota

Yamagata

Seto

Nakajima

Tests for serum levels of trefoil factor family proteins can improve gastric cancer screening

Gastroenterology141837845e172011

10.1053/j.gastro.2011.05.040

21699780

Vocka

Langer

Petrtyl

Vockova

Hanus

Kalousova

Zima

Petruzelka

Trefoil factor family (TFF) proteins as potential serum biomarkers in patients with metastatic colorectal cancer

Neoplasma624704772015

10.4149/neo_2015_056

Bignotti

Ravaggi

Tassi

Calza

Rossi

Falchetti

Romani

Bandiera

Odicino

Pecorelli

Santin

Trefoil factor 3: A novel serum marker identified by gene expression profiling in high-grade endometrial carcinomas

Br J Cancer997687732008

10.1038/sj.bjc.6604546

18682706

Vestergaard

Borre

Poulsen

Nexo

Torring

Plasma levels of trefoil factors are increased in patients with advanced prostate cancer

Clin Cancer Res12(3 Pt 1)8078122006

10.1158/1078-0432.CCR-05-1545

16467092

Yoo

Jung

Choi

Yoo

Kang

Reduced expression of TFF1 and increased expression of TFF3 in gastric cancer: Correlation with clinicopathological parameters and prognosis

Int J Med Sci101331402013

10.7150/ijms.5500

Casado

Garcia

Sanchez

Gomez Del Pulgar

Feliu

Maurel

Castelo

Moreno Rubio

Lopez

Garcia-Cabezas

Upregulation of trefoil factor 3 (TFF3) after rectal cancer chemoradiotherapy is an adverse prognostic factor and a potential therapeutic target

Int J Radiat Oncol Biol Phys84115111582012

10.1016/j.ijrobp.2012.01.083

22516806

Wang

Fang

Serum Trefoil Factor 3 as a protein biomarker for the diagnosis of colorectal cancer

Technol Cancer Res Treat164404452017

10.1177/1533034616674323

Yang

Xiao

Increased trefoil factor 3 levels in the serum of patients with three

Oncol Rep27127712832012

10.3892/or.2012.1627

22246423

Amin

Greene

Edge

Compton

Gershenwald

Brookland

Meyer

Gress

Byrd

Winchester

The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more ‘personalized’ approach to cancer staging

CA Cancer J Clin6793992017

10.3322/caac.21388

28094848

Dobashi

Tsubochi

Matsubara

Inoue

Inazawa

Endo

Ooi

Diverse involvement of isoforms and gene aberrations of Akt in human lung carcinomas

Cancer Sci1067727812015

10.1111/cas.12669

Ribieras

Tomasetto

Rio

The pS2/TFF1 trefoil factor, from basic research to clinical applications

Biochim Biophys Acta1378F61F771998

Xue

Zhang

Huang

Sheng

Lai

Identification of serum biomarkers for colorectal cancer metastasis using a differential secretome approach

J Proteome Res95455552010

10.1021/pr9008817

19924834

Chen

Chiou

Chong

Poh

Tan

Zhang

Zhu

Pandey

BasappaKumar

Lobie

Pharmacological Inhibition of TFF3 Enhances Sensitivity of CMS4 Colorectal Carcinoma to 5-Fluorouracil through Inhibition of p44/42 MAPK

Int J Mol Sci2062152019

10.3390/ijms20246215

Yusup

Huji

Fang

Wang

Dadihan

Wang

Upur

Expression of trefoil factors and TWIST1 in colorectal cancer and their correlation with metastatic potential and prognosis

World J Gastroenterol231101202017

10.3748/wjg.v23.i1.110

Xie

Guo

Tian

Yang

Wang

Guo

Diagnostic value evaluation of trefoil factors family 3 for the early detection of colorectal cancer

World J Gastroenterol23215921672017

10.3748/wjg.v23.i12.2159

Kaise

Miwa

Tashiro

Ohmoto

Morimoto

Kato

Urashima

Ikegami

Tajiri

The combination of serum trefoil factor 3 and pepsinogen testing is a valid non-endoscopic biomarker for predicting the presence of gastric cancer: A new marker for gastric cancer risk

J Gastroenterol467367452011

10.1007/s00535-011-0396-8

Huang

Wang

Zhang

Aberrant expression of trefoil factor 3 is associated with colorectal carcinoma metastasis

J Cancer Res Ther93763802013

10.4103/0973-1482.119308

24125969

Xiao

Liu

Xiao

Ren

Guleng

Serum TFF3 may be a pharamcodynamic marker of responses to chemotherapy in gastrointestinal cancers

BMC Clin Pathol14262014

10.1186/1472-6890-14-26

Higashiyama

Doi

Kodama

Yokuchi

Inaji

Tateishi

Estimation of serum level of pS2 protein in patients with lung adenocarcinoma

Anticancer Res16235123551996

8694568

Swensen

Jett

Hartman

Midthun

Sloan

Sykes

Aughenbaugh

Clemens

Lung cancer screening with CT: Mayo Clinic experience

Radiology2267567612003

10.1148/radiol.2263020036

12601181

Sone

Yang

Honda

Maruyama

Takashima

Hasegawa

Kawakami

Kubo

Haniuda

Yamanda

Results of three-year mass screening programme for lung cancer using mobile low-dose spiral computed tomography scanner

Br J Cancer8425322001

10.1054/bjoc.2000.1531

11139308

Swensen

Jett

Sloan

Midthun

Hartman

Sykes

Aughenbaugh

Zink

Hillman

Noetzel

Screening for lung cancer with low-dose spiral computed tomography

Am J Respir Crit Care Med1655085132002

10.1164/ajrccm.165.4.2107006

11850344

Manser

Irving

de Campo

Abramson

Stone

Pedersen

Elwood

Campbell

Overview of observational studies of low-dose helical computed tomography screening for lung cancer

Respirology10971042005

10.1111/j.1440-1843.2005.00590.x

15691245

Pastorino

Lung cancer screening

Br J Cancer102168116862010

10.1038/sj.bjc.6605660

20424610

dos Santos Silva

Ulrich

Doring

Botzenhart

Gott

Trefoil factor family domain peptides in the human respiratory tract

J Pathol1901331422000

10.1002/(SICI)1096-9896(200002)190:2<133::AID-PATH518>3.0.CO;2-B

Wiede

Jagla

Welte

Kohnlein

Busk

Hoffmann

Localization of TFF3, a new mucus-associated peptide of the human respiratory tract

Am J Respir Crit Care Med159(4 Pt 1)133013351999

10.1164/ajrccm.159.4.9804149

10194185

Wang

Johnston

Liu

Sloan

Buchanan

Expression of a breast-cancer-associated protein (pS2) in human neuro-endocrine tumours

Int J Cancer742702741997

10.1002/(SICI)1097-0215(19970620)74:3<270::AID-IJC6>3.0.CO;2-W

9221803

Ruchaud-Sparagano

Westley

May

The trefoil protein TFF1 is bound to MUC5AC in human gastric mucosa

Cell Mol Life Sci61194619542004

10.1007/s00018-004-4124-x

Otto

Thim

Trefoil factor family-interacting proteins

Cell Mol Life Sci62293929462005

10.1007/s00018-005-5482-8

Kouznetsova

Laubinger

Kalbacher

Kalinski

Meyer

Roessner

Hoffmann

Biosynthesis of gastrokine-2 in the human gastric mucosa: Restricted spatial expression along the antral gland axis and differential interaction with TFF1, TFF2 and mucins

Cell Physiol Biochem208999082007

10.1159/000110450

17982272

Sturmer

Muller

Hanisch

Hoffmann

Porcine gastric TFF2 is a mucus constituent and differs from pancreatic TFF2

Cell Physiol Biochem338959042014

10.1159/000358662

24713603

Hoffmann

Trefoil factor family (TFF) peptides and their diverse molecular functions in mucus barrier protection and more: Changing the paradigm

Int J Mol Sci2145352020

10.3390/ijms21124535

Koh

Shin

Lee

Hwang

Lee

Kim

Park

Lee

Choi

Kim

Gastric-type gene expression and phenotype in non-terminal respiratory unit type adenocarcinoma of the lung with invasive mucinous adenocarcinoma morphology

Histopathology768989052020

10.1111/his.14077

31985086

Gardner

Rulifson

Wang

Xiong

Belouski

Cao

Tang

Trefoil Factor 3 (TFF3) is regulated by food intake, improves glucose tolerance and induces mucinous metaplasia

PLoS One10e01269242015

10.1371/journal.pone.0126924

26083576

Baus-Loncar

Schmid

Lalani

el-N

Rosewell

Goodlad

Stamp

Blin

Kayademir

Trefoil factor 2 (TFF2) deficiency in murine digestive tract influences the immune system

Cell Physiol Biochem1631422005

10.1159/000087729

16121031

Itoh

Tomita

Uchino

Kobayashi

Kataoka

Sekiya

Nawa

cDNA cloning of rat pS2 peptide and expression of trefoil peptides in acetic acid-induced colitis

Biochem J318(Pt 3)9399441996

10.1042/bj3180939

8836141

Xian

Howarth

Mardell

Cool

Familari

Read

Giraud

Temporal changes in TFF3 expression and jejunal morphology during methotrexate-induced damage and repair

Am J Physiol277G785G7951999

Hoffmann

Trefoil factor family (TFF) peptides and their links to inflammation: A Re-evaluation and new medical perspectives

Int J Mol Sci2249092021

10.3390/ijms22094909

Figure 1.

Box-and-whisker plots of serum and urinary levels of TFFs in patients with lung cancer and healthy individuals. Samples were collected from 199 patients with lung cancer and 198 healthy individuals. (A) Mean concentration of all serum TFFs was elevated in patients with carcinomas compared with healthy individuals (control) (P<0.0001). (B) In patients with carcinomas, mean urinary TFF1 and TFF3 values normalized to urinary creatinine were significantly lower compared to healthy individuals (control), while that for TFF2 was not. Data are presented as mean ± standard deviation. cre, creatinine; TFF, trefoil factor.

Figure 2.

ROC curves. ROC curves were generated for each TFF and the sTFF1 + 2 + 3 + uTFF1 + 3 combination. Cut-off values and their sensitivity/specificity, as well as AUC and 95% CI are indicated in tables below the curves. (A) sTFFs and the combination of sTFF1 + 2 + 3 and uTFF1 + 3. (B) uTFFs. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic; sTFF, serum trefoil factor; uTFF, urine TFF.

Figure 3.

TFF immunohistochemical staining of patient tissues. Representative TFF1, TFF2 and TFF3 staining, immunohistochemical score and mean concentrations of serum TFF. (A) Squamous cell carcinoma, (B-D) Adenocarcinoma. In each panel, the upper row presents the lower power view, scale bar=200 mm, and the lower row presents the higher power view of a portion of the lower power view, scale bar=100 mm. TFF, trefoil factor.

Table I.

Patients and tumor characteristics.

	ELISA		IHC

Characteristics	Patients (n=199)	Control (n=198)	Patients (n=100)
Sex, n
Male	128	127	65
Female	71	71	35
Age, years
<70, n	95	95	47
>70, n	104	103	53
Range	32-93	38-75	32-93
Mean, median	69.3, 71	67.5, 71	68.6, 70
Histology, n
AC	142		73
SCC	45		23
LCC	4		1
SmCC	4		2
Pleomorphic	4		1
pT factor, n
Tis	6		2
T1mi	19		12
T1a	30		15
T1b	37		18
T1c	16		6
T2a	42		22
T2b	18		9
T3	20		10
T4	11		6
pN factor, n
N0	148		70
N1	23		12
N2	28		18
pStage, n
I	125		63
II	41		14
IIIa	33		23

AC, Adenocarcinoma; LCC, large cell carcinoma; IHC, immunohistochemistry; SCC, squamous cell carcinoma; SmCC, small cell carcinoma; pT, tumor stage; pN, node stage.

Table II.

Case-distribution of all participants separated with the cut-off (0.443) by the combination of sTFF1 + 2 + 3 with uTFF1 + 3 (n=397).

Patients	>cut-off	<cut-off	Statistical result
Patients with lung cancer (n=199)	160	39	Sensitivity=0.804
Healthy individuals (n=198)	51	147	Specificity=0.742
Statistical result	Positive predictive value=0.758	Negative predictive value=0.790	P=7.09×10⁻²⁹

Statistical analysis was performed using Fisher's exact test.

Table III.

Results of immunohistochemical staining (n=100).

	TFF1						TFF2						TFF3

IHC score	0	1	2	3	4	6	0	1	2	3	4	6	0	1	2	3	4	6
Total number of cases	59	7	13	10	8	3	63	5	16	7	6	3	50	11	20	4	6	9
Histological type (cases), n
AC (73)	44	5	11	6	5	2	48	4	9	6	5	1	35	9	15	3	6	5
SCC (23)	11	2	2	4	3	1	13	1	5	1	1	2	11	2	5	1	0	4
LCC (1)	1	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	0
SmCC (2)	2	0	0	0	0	0	0	0	2	0	0	0	2	0	0	0	0	0
Pleomorphic (1)	1	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	0

TFF, trefoil factor; IHC score, immunohistochemical score; AC, adenocarcinoma; SCC, squamous cell carcinoma; LCC, large cell carcinoma; SmCC, small cell carcinoma.

Table IV.

Correlation coefficients (ρ) between each TFF and P-values analyzed by Spearman's rank correlation test.

	Serum						Urine

	TFF1 vs. TFF2		TFF1 vs. TFF3		TFF2 vs. TFF3		TFF1 vs. TFF2		TFF1 vs. TFF3		TFF2 vs. TFF3

Characteristic	ρ	P-value	ρ	P-value	ρ	P-value	ρ	P-value	ρ	P-value	ρ	P-value
Healthy individuals (n=198)	0.473*	<0.001	0.294	-	0.352	-	0.528*	<0.001	0.756**	<0.001	0.545*	<0.001
Patients (n=199)	0.494*	<0.001	0.348	-	0.349	-	0.345	-	0.339	-	0.162	-
pTis/T1	0.497*	<0.001	0.326	-	0.443*	<0.001	0.265	-	0.304	-	0.108	-
pT2	0.486*	<0.001	0.458*	<0.001	0.230	-	0.471*	<0.001	0.352	-	0.172	-
pT3	0.567*	0.0091	0.322	-	0.316	-	0.393	-	0.322	-	0.319	-
pT4	0.582	0.066	0.418	0.203	0.109	-	0.609*	0.052	0.718**	0.016	0.736**	0.013
pN0	0.448*	<0.001	0.289	-	0.344	-	0.283	-	0.286	-	0.120	-
pN1	0.717**	<0.001	0.707**	<0.001	0.743**	<0.001	0.592*	0.0029	0.420*	0.046	0.356	-
pN2	0.602*	<0.001	0.459*	0.014	0.273	-	0.535*	0.0033	0.484*	0.009	0.270	-
Adenocarcinoma	0.570*	<0.001	0.345	-	0.353	-	0.374	-	0.312	-	0.154	-
Squamous cell	0.334	-	0.187	-	0.396	-	0.392	-	0.269	-	0.096	-
carcinoma

P-values for all groups were only indicated for categories showing ρ>0.4. *substantial correlation, **high correlation. pN, node stage; pT, tumor stage; TFF, trefoil factor.

Table V.

Correlation Coefficients (ρ) between sTFF, uTFF and IHC score by Spearman's rank correlation test in patients with lung cancer.

Factor	sTFF1	sTFF2	sTFF3	IHC score
uTFF (n=199)
ρ	0.516*	0.195	0.139	0.197
P-value^a	<0.000	-	-	-
uTFF2 (n=199)
ρ	0.128	0.463*	−0.008	−0.281
P-value^a	-	<0.0001	-	-
uTFF3 (n=199)
ρ	0.169	0.234	0.560*	0.342
P-value^a	-	-	<0.0001	-
IHC score (n=100)
ρ	0.563*	−0.657*	0.813**	-
P-value^a	<0.0001	<0.0001	<0.0001	-

P-values are indicated only for categories showing ρ>0.4. *substantial correlation, **high correlation. sTFF, serum trefoil factor; uTFF, urinary TFF; IHC, immunohistochemistry.

Table VI.

Association between sTFF levels and clinicopathological variables.

		sTFF1		sTFF2		sTFF3

Clinicopathological factors (n=199)	Number of cases	Mean + SD	P-value	Mean + SD	P-value	Mean + SD	P-value
Age, years
<70, n	95	1.351±1.192	0.0764	4.472±3.058	0.0728	8.243±5.923	0.1932
>70, n	104	1.479±1.007		4.859±2.604		8.817±5.049
Sex
Male	128	1.396±1.003	0.7101	4.579±2.570	0.738	8.959±6.051	0.216
Female	71	1.458±1.259		4.846±3.258		7.793±5.419
pT factor
pTis/T1/T2	168	1.408±1.108	0.928	4.801±2.862	0.018	8.311±4.662	0.729
pT3/T4	31	1.469±1.060		3.987±2.582		9.803±8.662
pN factor
pN0/1	171	1.473±1.129	0.033	4.770±2.912	0.2521	8.227±4.690	0.3174
pN2	28	1.082±0.825		4.087±2.216		10.215±8.835
pStage
I	125	1.481±1.167	0.085	4.814±2.933	0.093	8.419±4.883	0.618
II	41	1.492±1.064		4.145±2.390		7.870±4.182
IIIa	33	1.086±0.794		3.971±2.134		9.850±8.303
Histological type
Adenocarcinoma	142	1.448±1.160	0.5596	4.880±3.007	0.154	8.644±5.814	0.3939
Squamous cell carcinoma	45	1.418±0.9820		4.380±2.500		8.764±4.558
Small cell carcinoma	4	1.241±1.043		2.996±0.906		5.809±3.558
Others^a	8	1.004±0.493		3.683±0.885		7.251±4.986
IHC score	100
0		38.746	<0.0001	63.873	<0.0001	29.460	<0.0001
1		47.000		67.600		46.727
2		56.385		26.125		70.250
3		75.700		22.000		77.500
4		81.125		11.000		91.333
6		98.667		16.667		88.889

Others are large cell carcinoma and pleomorphic carcinoma. P-values were obtained by Mann-Whitney U test for two groups and by Kruskal-Wallis test for the comparison of three or more groups. sTFF, serum trefoil factor; IHC, immunohistochemistry, pN, node stage; pT, tumor stage.

Table VII.

Association between uTFF levels and clinicopathological variables.

		uTFF1		uTFF2		uTFF3

Clinicopathological) factors (n=199	Number of cases	Mean + SD	P-value	Mean + SD	P-value	Mean + SD	P-value
Age, years
<70	95	1.397±1.696	0.620	19.522±16.988	0.603	5.411±5.452	0.813
>70	104	1.295±1.223		17.198±10.064		5.341±4.780
Sex
Male	128	1.366±1.448	0.281	16.538±9.463	0.126	5.454±5.194	0.734
Female	71	1.304±1.505		21.460±19.023		5.230±4.956
pT factor
pTis/T1/T2	168	1.334±1.500	0.760	18.647±14.232	0.174	5.394±5.138	0.841
pT3/T4	31	1.397±1.279		16.466±11.366		5.266±4.960
pN factor
pN0/1	171	1.397±1.542	0.309	18.531±14.202	0.623	5.389±5.107	0.943
pN2	28	1.018±0.797		16.958±11.346		5.285±5.139
pStage
I	125	1.376±1.240	0.399	16.355±9.641	0.243	5.277±4.886	0.98
II	41	1.558±1.510		18.561±21.203		4.958±3.929
IIIa	33	1.171±1.035		14.864±8.759		4.793±4.437
Histological type
Adenocarcinoma	142	1.273±1.503	0.027	19.512±14.974	0.062	5.483±5.172	0.444
Squamous cell carcinoma	45	1.704±1.467		16.094±10.861		5.343±5.158
Small cell carcinoma	4	0.780±0.383		9.874±5.352		2.347±1.978
Others^a	8	0.971±0.786		14.395±4.235		5.509±4.914
IHC score (n=100)
0		45.169	-	56.302	-	41.120	-
1		65.571		48.400		55.727
2		49.692		42.563		57.750
3		74.000		44.857		42.750
4		54.500		25.833		75.333
6		34.667		37.000		67.000

Others are large cell carcinoma and pleomorphic carcinoma. P-values were obtained by Mann-Whitney U test for two groups and by Kruskal-Wallis test for the comparison of three or more groups uTFF, urinary trefoil factor; IHC, immunohistochemistry.