Presented at the Neonatal Society 2004 Spring Meeting.

Harrington DJ¹, Clarke P², Card DJ¹, Savidge GF¹, Mitchell SJ³, Soper R⁴, Hodges SJ⁴, Shearer MJ¹

¹ The Centre for Haemostasis and Thrombosis, St. Thomas’ Hospital, London, UK
² Neonatal Unit, Hope Hospital, Salford, UK
³ Neonatal Medical Unit, St Mary’s Hospital, Manchester, UK
⁴ Institute of Hepatology, Liver Failure Group, Dept. of Medicine, University College London, UK

Background: Vitamin K deficiency bleeding (VKDB) in infancy is preventable by adequate vitamin K prophylaxis. There remains little consensus on optimal regimes for VKDB prophylaxis, particularly for preterm infants (1). Possible consequences of extremely high plasma vitamin K concentrations after prophylaxis are unknown. Conversely suboptimal regimes may render infants at risk of bleeding. Study of vitamin K status in this vulnerable population has hitherto been problematic due to limitations in available techniques and reliance on invasive tests. Measurement by HPLC of the two major urinary catabolites of vitamin K (the 5C and 7C catabolites) offers a novel non-invasive marker of total vitamin K status (2).

Aim: To assess the cumulative excretion of vitamin K catabolites in the urine of term and preterm infants in the period 24-48 hours following vitamin K1 prophylaxis.

Methods: We attempted to collect all urine voided from birth and during the 48 hour period following vitamin K1 prophylaxis. Term infants received 1000µg of vitamin K1 intramuscularly (IM); preterm infants were randomised to 200µg IM, 500µg IM or 200µg intravenously (IV). The 5C and 7C catabolites of vitamin K1 were measured using HPLC.2 The ratio of urinary catabolite excretion was compared with that reported in healthy adults, in whom the ratio of 5C:7C metabolites is 4:1. Additionally, serum vitamin K1 2,3-epoxide, a diagnostic marker of vitamin K1 2,3-epoxide reductase activity, was determined by HPLC in all preterm infants 5 days post partum.
Results: Thirty term and twenty-two preterm infants were enrolled. Complete urine collections were obtained in the study period for five term and thirteen preterm infants. The table shows cumulative urinary excretion of vitamin K by gestational age and regime:

In infants from the IM group, catabolite excretion was proportional to the administered dose of vitamin K1 (Spearman’s correlation coefficient r= 0.597, P = 0.031). In fifteen infants excretion of the 5C catabolite predominated and the relative 5C:7C catabolite excretion reflected that of normal adults. Vitamin K1 2,3-epoxide was undetectable (<0.05 mg/L) in these infants. In three preterm infants excretion of the 7C catabolite predominated and elevated levels of vitamin K 2,3-epoxide were detected (Median [range] 50.62 [16.03 – 130.92] mg/L).

Conclusion: These are the first data to be reported for the urinary excretion of the two major vitamin K catabolites in term and preterm infants. The relative excretion of the catabolites is generally comparable to that of healthy adults. The biochemical pathways involved in the hepatic metabolism of vitamin K may be saturated in infants who predominately excrete the 7C catabolite and have elevated serum levels of vitamin K1 2,3-epoxide. Measurement of urinary excretion of vitamin K catabolites may offer further insights into neonatal vitamin K metabolism and the potential for non-invasive monitoring of vitamin K status.

References
1. Clarke P, Mitchell S. J Thromb Haemost 2003;1:384-6
2. Harrington DJ, Soper R, Edwards C, Savidge GF, Hodges S, Shearer M.J. Bone 2002; 30 (suppl.): 28S