Urinary mevalonate excretion: a specific and sensitive screening test for mevalonate kinase deficiency

Background Urinary mevalonic acid (MVA) analysis is a practical screening test for mevalonate kinase deficiency (MKD) in children with recurrent fever. We assessed its diagnostic accuracy, effects of sampling and preanalytical handling, and MVA dynamics during febrile and afebrile periods, considering the feasibility of urine collection. Methods All urinary MVA analyses performed between January 2004 and July 2025 were reviewed retrospectively; prospective sampling experiments included 12 patients with genetically confirmed MKD. Diagnostic accuracy was assessed at sample and patient levels using only samples pooled from three consecutive urine portions (center protocol). Genetically confirmed cases were compared with clinically and/or genetically negative controls and, separately, with MVK-negative controls. Results The dataset comprised 3,606 samples from 2,520 patients; 153/162 samples from the 22-patient descriptive cohort were MVA-positive. Seventy samples from 12 genetically confirmed cases were positive, and 308 samples from 306 clinical/genetic controls were negative, yielding observed sensitivity of 100% (95% CI 94.9–100.0) and specificity of 100% (95% CI 98.8–100.0). Genetic-reference and patient-level estimates were also 100%, although patient-level sensitivity, based on only 12 cases, had a wider confidence interval (95% CI 73.5–100.0). Because genetic testing was selective, the 100% estimates should be interpreted cautiously and within the context of this cohort. Handling affected concentrations in the one-patient experiment ( p = 0.008) without loss of detection, whereas collection type did not significantly affect febrile concentrations. MVA concentrations varied widely within and between patients, were higher during febrile than afebrile periods ( p < 0.001), and showed no consistent temporal pattern across serial samples. Conclusion Standardized urinary MVA analysis using the three-portion protocol proved reliable and cost-effective screening for MKD in our cohort. Pooling three consecutive urine portions provides a practical alternative to 24-hour collection, particularly in children, with high observed diagnostic accuracy, though confirmation in larger cohorts with complete genetic verification is warranted. MVA concentration is higher during febrile than afebrile periods but varies substantially intra- and inter-individually. Appropriate preanalytical handling is important, particularly when concentrations are close to the detection limit. If clinical suspicion of MKD persists, MVK genetic testing should be performed despite a negative urinary MVA result.