Longitudinal Changes in Glutamine and Ammonia in Relation to Hyperammonemic Crisis in Urea Cycle Disorders

Abstract

Hyperammonemic crisis (HAC) remains a major risk factor for urea cycle disorders (UCD), and practical outpatient predictors are limited. We tested whether short-term changes in plasma glutamine (ΔGln) and ammonia (ΔNH₃) predict HAC and whether effects differ by onset type. In a retrospective cohort (2014–2024) of 18 patients with UCD (neonatal-onset [NO] nine; late-onset [LO] nine), HAC was defined as ammonia (NH₃) > 150 μg/dL (88.1 μmol/L). For each patient, ΔGln and ΔNH₃ were calculated between sequential outpatient samples. Investigation 1 compared the changes observed between 31–60 days and 8–30 days before HAC, with stable period changes. Investigation 2 compared changes at 61–90 and 31–60 days before HAC with stable period changes. Associations were evaluated using generalized linear mixed-effects models with onset-specific effects. In NO, larger ΔGln during Investigation 1 was associated with higher HAC risk (p < 0.001) whereas ΔNH₃ was not associated with HAC (p = 0.361). The probability of HAC in NO was estimated to reach 67.1% at ΔGln +500 μmol/L. In LO, neither ΔGln nor ΔNH₃ during Investigation 1 showed a significant association with HAC, and the estimated probabilities remained low across the observed ranges. During Investigation 2, no significant associations between biomarkers and HAC were observed in either group. Progressive increases in plasma glutamine levels within the 31–60 and 8–30 days pre-HAC window may serve as early markers of HAC risk in NO-UCD, supporting the utility of longitudinal monitoring. These trends were not associated with LO-UCD, suggesting the need for alternative surveillance strategies tailored to the onset phenotype.