Persistent acute kidney injury biomarkers: A systematic review and meta-analysis
Keran Shi,Wei Jiang,Lin Song,Xianghui Li,Chuanqing Zhang,Luanluan Li,Yunfan Feng,Jiayan Yang,Tianwei Wang,Haoran Wang,Lulu Zhou,Jiangquan Yu,Ruiqiang Zheng
DOI: https://doi.org/10.1016/j.cca.2024.119907
2024-08-08
Abstract:Background: Various biomarkers reportedly predict persistent acute kidney injury (AKI) despite their varying predictive performance across clinical trials. This study aims to compare the accuracy of various biomarkers in predicting persistent AKI in different populations and regions. Methods: In this meta-analysis, we searched for urinary C-C motif chemokine ligand 14 (CCL14), Tissue inhibitor of metalloproteinase-2&insulin-like growth factor-binding protein-7 (TIMP-2&IGFBP7), Neutrophil Gelatinase-Associated Lipocalin (NGAL), plasma Cystatin C (pCysC), Soluble urokinase plasminogen activator receptor (suPAR), Proenkephalin (PenK) and urinary dickkopf-3:urinary creatinine (uDKK3:uCr) from various databases including Medline, PubMed, Embase, and Cochrane. This was geared towards predicting persistent AKI in adults (>18 years). Hierarchically summarized subject work characteristic curves (HSROC) and diagnostic odds ratio (DOR) values were used to summarize the diagnostic accuracy of the biomarkers. Further, meta-regression and subgroup analyses were carried out to identify sources of heterogeneity as well as evaluate the best predictive biomarkers in different populations and regions. Results: We screened 31 studies from 2,356 studies and assessed the diagnostic value of 7 biomarkers for persistent AKI. Overall, CCL14 had the best diagnostic efficacy with an AUC of 0.79 (95 % CI 0.75-0.82), whereas TIMP-2 & IGFBP7, NGAL, and pCysC had diagnostic efficacy of 0.75 (95 % CI 0.71-0.79),0.71 (95 % CI 0.67-0.75), and 0.7007, respectively. Due to a limited number of studies, PenK, uDKK3:uCr, and suPAR were not subjected to meta-analysis; however, relevant literature reported diagnostic efficacy above 0.70. Subgroup analyses based on population, region, biomarker detection time, AKI onset time, and AKI duration revealed that in the intensive care unit (ICU) population, the AUC of CCL14 was 0.8070, the AUC of TIMP-2 & IGFBP7 was 0.726, the AUC of pCysC was 0.72, and the AUC of NGAL was 0.7344; in the sepsis population, the AUC of CCL14 was 0.85, the AUC of TIMP-2&IGFBP7 was 0.7438, and the AUC of NGAL was 0.544; in the post-operative population, the AUC of CCL14 was 0.83-0.93, the AUC of TIMP-2&IGFBP7 was 0.71, and the AUC of pCysC was 0.683. Regional differences were observed in biomarker prediction of persistent kidney injury, with AUCs of 0.8558 for CCL14, 0.7563 for TIMP-2 & IGFBP7, and 0.7116 for NGAL in the Eurasian American population. In the sub-African population, TIMP-2 & IGFBP7 had AUCs of 0.7945, 0.7418 for CCL14, 0.7097 for NGAL, and 0.7007 for pCysC. for TIMP-2 & IGFBP7 was 0.7945, AUC for CCL14 was 0.7418, AUC for NGAL was 0.7097, and AUC for pCysC was 0.7007 in the sub-African population. Duration of biomarker detection, AKI onset, and AKI did not influence the optimal predictive performance of CCL14. Subgroup analysis and meta-regression of CCL14-related studies revealed that CCL14 is the most appropriate biomarker for predicting persistent stage 2-3 AKI, with heterogeneity stemming from sample size and AKI staging. Conclusion: This meta-analysis discovered CCL14 as the best biomarker to predict persistent AKI, specifically persistent stage 2-3 AKI.