PIVKA-II concentrations in patients with cystic fibrosis.

Montalembert et al report that PIVKA-II was detected in 33% of patients with cystic fibrosis, while vitamin K, plasma concentrations were normal.' It is astonishing that despite daily supplements of 5-10 mg vitamin K,, PIVKA-II was detectable in these patients. The authors conclude that PIVKAII is not associated with vitamin K deficiency, but with the use of antibiotics. There is some doubt as to whether the assay for PIVKA-II used by the authors is reliable. Widdershoven et al compared different methods for measuring PIVKA-II and reported that techniques involving adsorption of normal factor II may result in false positive values, because the carboxylated prothrombin may not be removed completely.2 Determination of PIVKA-II by monoclonal antibody was found to be the most specific and sensitive method.2 We did not detect PIVKA-II in any of eight patients with cystic fibrosis who were supplemented with vitamin K, (4-30 mg/day).3 In only one out of 16 unsupplemented patients with cystic fibrosis was PIVKA-II found (0-16 AU/ml). This patient took antibiotics, had a low vitamin K, concentration of 0-06 ,ug/l and a Thrombotest of 56%.3 The authors do not mention vitamin K,. Except for vitamin K,, vitamin K, must be accounted for when assessing vitamin K status. Antibiotics may disturb vitamin K, production by intestinal flora and hence reduce the amount of total vitamin K available for the carboxylation of PIVKA-Il to functional factor II. A correlation between subnormal coagulation tests and antibiotics in cystic fibrosis was reported by Komp and Selden.4 As there was no information on concentrations of vitamin K, it is impossible to establish normal values for vitamin K,. PIVKA-II, however, is a direct reflection of the availability of total vitamin K in the liver and hence is associated with vitamin K deficiency. In our study PIVKA-II was found in only one unsupplemented patient with cystic fibrosis, and hence we conclude that vitamin K deficiency occurs infrequently in cystic fibrosis.' EAM CORNELISSEN AFVAN LIEBURG CGVAN OOSTROM LAH MONNENS Departnent of Paediatrics, University Hospital Nijmegen, PO Box 9101, NM-6500 HB Nijmegen, The Netherlands