THE TRYPTOPHAN LOAD AS A TEST FOR PYRIDOXINE DEFICIENCY IN HOSPITALIZED PATIENTS

Abstract

Measurement of urinary excretion of xanthurenic acid after administration of an oral dose of tryptophan has been the technique most widely used for detection of subclinical pyridoxine deficiency. Although alterations in urinary excretion of this and other metabolites of tryptophan have been described in a wide variety of conditions and diseases,’ comparison of results obtained by different investigators has been difficult because of variability in dose and kind of tryptophan, urine collections, method of analysis and definitions of “abnormality.” Recently, as a result of the development of more accurate methods by Price and associates2 and the plea for standardization made by C o ~ r s i n , ~ greater agreement has been reached concerning proper methods of study. For the past ten years our laboratory has been concerned with the investigation of the prevalence of biochemical evidence of deficiency of several of the watersoluble vitamins in il l patients in our hospital. Several years ago we began to assess the relative “sensitivity” of several methods proposed for the detection of pyridoxine deficiency in normal adults placed upon a pyridoxine-deficient diet supplemented with vitamin-free casein. Studies included serial determinations of absolute lymphocyte count, serum and erythrocyte glutamic-oxalacetic transnminase and glutamic-pyruvic transaminase, taurine excretion after a cysteine load, estimation of phosphorylated and nonphosphorylated derivatives of pyridoxine in plasma, urinary excretion of 4-pyridoxic acid and 24-hour urinary excretion of xanthurenic acid and hydroxykynurenine after a 5 g dose of Ltryptophan. Afte’r the study of three patients, we arrived at the same conclusions as those in the then just published papers of Yess, Swan and co l leagi ie~~-~ that the earliest and most consistent abnormalities observed were increases in excretion of hydroxykynurenine and xanthurenic acid after a tryptophan load. At this time, Coursin made his plea for standardization of methods and for utilization of a 2 g dose of ~t ryptophan .~ For this reason and also because we had observed considerable interindividual variation in excretion of tryptophan metabolites in our normal subjects tested during control periods before and after the period of deficiency, a comparison of the 2 g and 5 g dose of L-tryptophan was made in sixteen healthy adults both prior to and following administration of pyridoxine.’ Urinary hydroxyanthranilic acid was also determined, since the ratio of excretion of hydroxyknurenine to that of hydroxyanthranilic acid has been proposed as a helpful adjunct in the assessment of pyridoxine deficiency.8 In order to clarify the rationale for this approach, FIGURE 1 presents a diagram of the metabolic pathway of tryptophan metabolism of importance in the tryptophan load test. Tryptophan pyrrolase, which catalyzes the first reaction in this pathway, is inducible by substrate (tryptophan) and by adrenocortical steroids. The vitamin B6-containing enzymes are not subject to adaptive change (enzyme induction),9 and therefore a relative or absolute decrease in activity of these enzymes could bring about alterations in tryptophan metabolism. Since the