The Hyperintense Acute Reperfusion Marker (HARM) and Gadolinium Leakage in Ocular Structures (GLOS) are pivotal radiological findings in post-contrast fluid-attenuated inversion recovery imaging (pcFLAIR), attesting to gadolinium leakage into the cerebrospinal fluid (CSF) in various neurological disorders. Often observed following acute strokes, HARM and GLOS, however, exhibit considerable variability in their prevalence ranging from 5.5 to 85% and 30-76%, respectively. Given their similarity and association of HARM with poor outcomes in stroke, accurately evaluating these markers may be crucial for advancing our understanding of stroke pathophysiology and improving clinical management. Our work aims to identify the major methodological challenges and confounding factors limiting the understanding of HARM and GLOS in stroke.To address these issues, we thoroughly conducted a literature search in Embase, Scopus, and PubMed until July 2022. Our search yielded 38 stroke studies, with only 6 evaluating GLOS. Guided by major findings, we adapted the Newcastle-Ottawa scale for bias risk assessment. Effect estimates were synthetized considering cohort size, statistical significance, and bias risk.Methodological issues emerged from the lack of time-specific data, omission of differential CSF hyperintensities, imprecise definitions, and overlooking of adjusting variables like assessment timing, contrast dosages or renal function. Discrepancy results mainly arise from an inadequate time window of investigation, and further research should stratify patients based on the timing of gadolinium injection.Our findings emphasize the importance of a more detailed exploration of their timing and localization, rather than simply their binary presence, extent, or severity.