This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Circulating Thioredoxin 1 as an Adjunct to Mammography for Breast Cancer Detection: A Multicenter Clinical Validation Study
by
, , , , ,
Hye Mi Ko
1
,
Jee Ye Kim
2,
Songhak Kim
3,
Jungchan Shin
3,
Xiaoguang Yang
3,
Jong Am Song
3,
Ji Yeon Kim
1,
Sang Il Lee
1,
Jeong Eun Lee
4,
Bo Bae Choi
5,
Jin Man Kim
6,
Jin Gyu Jung
7,
Je Ryong Kim
1,
Ji Young Sul
1,
Eun Heui Jin
8,
Jang Hee Hong
8,
Choong Sik Lee
9,
Kyoung Hoon Suh
3,
Seung Il Kim
2 and
Jin Sun Lee
1,* 1
Department of Surgery, Chungnam National University Hospital & Chungnam National University School of Medicine, Daejeon 35015, Republic of Korea
2
Department of Surgery, Severance Hospital, Yonsei University College of Medicine, Seoul 03722, Republic of Korea
3
E&S Healthcare, Daejeon 34002, Republic of Korea
4
Department of Internal Medicine, College of Medicine, Chungnam National University, Daejeon 35015, Republic of Korea
5
Department of Radiology, Chungnam National University Hospital, Daejeon 35015, Republic of Korea
6
Department of Pathology, Chungnam National University Hospital, Daejeon 35015, Republic of Korea
7
Department of Family Medicine, Chungnam National University Hospital, Daejeon 35015, Republic of Korea
8
Department of Pharmacology, Chungnam National University College of Medicine, Daejeon 35015, Republic of Korea
9
Department of Pathology, Konyang University Hospital, Daejeon 35365, Republic of Korea
*
Author to whom correspondence should be addressed.
Cancers 2026, 18(15), 2416; https://doi.org/10.3390/cancers18152416
Submission received: 25 June 2026
/
Revised: 16 July 2026
/
Accepted: 23 July 2026
/
Published: 27 July 2026
(This article belongs to the Section Cancer Biomarkers)
Simple Summary
Breast cancer is commonly detected using mammography, but the accuracy of imaging can be reduced in women with dense breast tissue and in patients with very small tumors or inconclusive imaging findings. Blood-based biomarkers may provide additional information and improve diagnostic assessment in these situations. In this study, we evaluated thioredoxin 1, a protein associated with breast cancer biology, as a complementary blood biomarker in a multicenter cohort of 1901 serum samples. We found that circulating thioredoxin 1 distinguished breast cancer from non-cancer controls and showed strong diagnostic performance across different tumor stages, tumor sizes, molecular subtypes, and breast-density categories. When combined with mammographic assessment, thioredoxin 1 improved diagnostic performance and demonstrated greater net benefit in decision curve analysis, particularly in women with dense breasts and in patients with indeterminate imaging findings. These findings suggest that thioredoxin 1 may serve as a clinically useful adjunct to imaging-based breast cancer evaluation.
Abstract
Background/Objectives: Mammography remains central to breast cancer diagnosis, yet clinically meaningful uncertainty persists in women with dense breasts and in patients with very small lesions or indeterminate imaging findings. We evaluated whether circulating thioredoxin 1 (Trx1) could provide biological information that complements imaging-based assessment. Methods: This multicenter clinical validation study evaluated circulating Trx1 in 1901 serum samples across four predefined cohorts. Diagnostic performance was assessed using receiver operating characteristic (ROC) and precision–recall analyses, together with evaluation of integration with mammographic assessment and decision curve analysis (DCA). A predefined Trx1 cutoff established in prior clinical investigations was applied. Results: Circulating Trx1 concentrations were significantly elevated in breast cancer compared with controls (p < 0.001). Trx1 showed excellent diagnostic discrimination (AUC 0.985; 95% CI, 0.974–0.996), with sensitivity and specificity of 96.4% and 97.3%, respectively. Diagnostic performance was consistent across disease stages, tumor-size categories, molecular subtypes, mammographic categories, and breast-density groups. Trx1 sensitivity remained high in dense breasts (98.9% in BI-RADS density grades C/D), whereas mammographic sensitivity was substantially lower (67.0%). Integration of Trx1 with mammographic assessment improved diagnostic discrimination (AUC 0.980) and provided greater net benefit in decision curve analysis, particularly among women with low-suspicion or indeterminate imaging findings. Conclusions: Circulating Trx1 demonstrated robust diagnostic performance across disease stages, tumor-size categories, and clinically relevant patient subgroups while providing biologically independent information that complements mammographic assessment. Although this retrospective multicenter study requires prospective validation in independent diagnostic populations, Trx1 may serve as a clinically useful adjunct to imaging-based breast cancer evaluation, particularly in diagnostically uncertain settings.
