Novel Urinary Glycan Biomarkers Predict Cardiovascular Events in Patients With Type 2 Diabetes: A Multicenter Prospective Study With 5-Year Follow Up (U-CARE Study 2).

Kazuyuki Mise ,Shinichiro Ando,Chigusa Higuchi,Tomokazu Nunoue,Kazutoshi Murakami,Jun Wada,Atsuhito Tone,Katsuhiro Miyashita,Mayu Watanabe,Akihiro Katayama,Ikki Shimizu,Sanae Teshigawara,Atsuko Nakatsuka,Shinji Kamei,Takashi Matsuoka,Haruhito A Uchida,Masao Yamada,Kenichi Shikata,Tatsuaki Nakato,Jun Eguchi,Mariko Imamura,Satoshi Yamaguchi,Kazuyuki Hida,Sadaaki Miyamoto ,Michihiro Yoshida

doi:10.3389/fcvm.2021.668059

Abstract

Background: Although various biomarkers predict cardiovascular event (CVE) in patients with diabetes, the relationship of urinary glycan profile with CVE in patients with diabetes remains unclear.Methods: Among 680 patients with type 2 diabetes, we examined the baseline urinary glycan signals binding to 45 lectins with different specificities. Primary outcome was defined as CVE including cardiovascular disease, stroke, and peripheral arterial disease.Results: During approximately a 5-year follow-up period, 62 patients reached the endpoint. Cox proportional hazards analysis revealed that urinary glycan signals binding to two lectins were significantly associated with the outcome after adjustment for known indicators of CVE and for false discovery rate, as well as increased model fitness. Hazard ratios for these lectins (+1 SD for the glycan index) were UDA (recognizing glycan: mixture of Man5 to Man9): 1.78 (95% CI: 1.24–2.55, P = 0.002) and Calsepa [High-Man (Man2–6)]: 1.56 (1.19–2.04, P = 0.001). Common glycan binding to these lectins was high-mannose type of N-glycans. Moreover, adding glycan index for UDA to a model including known confounders improved the outcome prediction [Difference of Harrel's C-index: 0.028 (95% CI: 0.001–0.055, P = 0.044), net reclassification improvement at 5-year risk increased by 0.368 (0.045–0.692, P = 0.026), and the Akaike information criterion and Bayesian information criterion decreased from 725.7 to 716.5, and 761.8 to 757.2, respectively].Conclusion: The urinary excretion of high-mannose glycan may be a valuable biomarker for improving prediction of CVE in patients with type 2 diabetes, and provides the rationale to explore the mechanism underlying abnormal N-glycosylation occurring in patients with diabetes at higher risk of CVE.Trial Registration: This study was registered with the University Hospital Medical Information Network on June 26, 2012 (Clinical trial number: UMIN000011525, URL: https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000013482).

Highlights

Cardiovascular disease (CVD) is a global burden especially in low- and middle-income countries and the leading cause of disability and mortality [1]
We found that the urinary glycan binding signals to Sambucus nigra (SNA), Ricinus communis (RCA120), Dolichos biflorus (DBA), Agaricus bisporus (ABA), Artocarpus integrifolia (Jacalin), and Amaranthus caudatus (ACA) improved the prediction of renal outcome in the models employing the known risk factors [16]
Adding of either glycan indexes to the multivariate model significantly improved the ability of discrimination and reclassification such as area under the receiver operating characteristic (AUROC) and net reclassification improvement (NRI) [difference in AUROC: 0.031 for Urtica dioica (UDA), 0.027 (0.001–0.053, P = 0.040) for Calsepa, category-free NRI: 0.368 (0.045–0.692, P = 0.026) for UDA, and 0.388 (0.099–0.677, P = 0.008) for Calsepa], whereas either of the two glycan indexes did not significantly improve integrated discrimination [integrated discrimination improvement (IDI): 0.024 (−0.009–0.056, P = 0.16) for UDA

Summary

Introduction

Cardiovascular disease (CVD) is a global burden especially in low- and middle-income countries and the leading cause of disability and mortality [1]. Chronic kidney disease (CKD) is an emerging global health burden with prevalence of ∼15% of adult populations and is independently associated with increased cardiovascular event (CVE) including stroke and peripheral arterial disease (PAD) besides the traditional risk factors [3, 4]. In type 2 diabetes, the CVE risk prediction is potentially improved by novel biomarkers involved in the biological process, not explained by the traditional risk factors [7]. The area under the receiver operating characteristic (AUROC) or c-index is a measurement for discrimination capacity of classification model, while the net reclassification improvement (NRI) is a commonly used measure for the prediction increment by the addition of new biomarkers. Various biomarkers predict cardiovascular event (CVE) in patients with diabetes, the relationship of urinary glycan profile with CVE in patients with diabetes remains unclear

Methods

Results

Discussion

Conclusion