Biomarker selection and reporting architecture in human proximal tubular in vitro models of cisplatin-induced nephrotoxicity: a systematic review

Cisplatin-induced acute kidney injury remains a major dose-limiting complication in oncology. Human proximal tubule in vitro models are widely used to investigate nephrotoxic mechanisms; however, biomarker selection and reporting patterns vary considerably across studies. This PRISMA-guided systematic review aimed to characterize the frequency, biological categorization, and co-reporting patterns of biomarkers used in human proximal tubule in vitro models of cisplatin-induced injury. A systematic search of PubMed, Scopus, and Web of Science identified experimental in vitro studies published between 2014 and 2024 that reported quantifiable biomarkers following cisplatin exposure. Biomarkers were classified into predefined structural injury, regulated cell death, and mechanistic categories. Descriptive frequency analyses were performed, and co-occurrence relationships among the 12 most frequently reported biomarkers were evaluated using a binary presence–absence matrix. Fifty-eight studies met inclusion criteria. Kidney injury molecule-1 was the most frequently reported structural biomarker. Mechanistic endpoints, particularly oxidative stress and intracellular signaling markers, predominated over structural injury markers. Co-reporting analysis revealed recurrent pathway-oriented groupings, suggesting hypothesis-driven panel selection rather than standardized implementation across studies. These findings highlight the need for improved harmonization and structured biomarker strategies to enhance reproducibility, predictive robustness, and translational alignment in vitro nephrotoxicity research.