Thus age maturation affects speed of isoniazid acetylation b

Thus, age (maturation) affects speed of isoniazid acetylation by modifying effect of genotype, a non-linear interaction, however after 5.3years the maturation is complete. Table 4 shows the BFs for NAT2 affinity for isoniazid, or Km. A child\'s height was the primary predictor and had the highest variable importance score. Height however could be substituted by age, given the tight link between height and age (Pearson r=0.917; p<0.001), a tightly linked relationship frequently used in the clinic. Thus age was the main predictor of Km as well; in fact, the coefficient of determination (R2) value of age alone was 31%. However, MARS also identified that isoniazid peak concentration and genotype as important predictors (Table 4). Combined, these factors explained 80% of the variance in NAT2 enzyme affinity, as given by equation for Km and its predictors: Table 4 shows that BF4 is a special type of a basis function, described by an interaction of BF1 (age or height) and genotype. In other words, age constrained the effect of genotype on enzyme affinity. Table 5 shows the BFs for isoniazid elimination constant (KINH). The elimination rate constant is a composite measure reflecting both acetylation and elimination of parent beta adrenergic receptors via loss in urine. The main predictors were isoniazid dose and NAT2 genotype. In this case, however BF2 was an interaction based on the modification of effect of genotype on the effect of isoniazid dose (i.e., isoniazid concentration), an example of higher order interaction among predictors. This relationship is shown in Eq. (4): This was confirmed by standard regression methods, which showed that for each unit increase in isoniazid dose above 40mg, KINH increased by an additional 2.82×10 per hour for subset 1 versus 1.05×10 per hour for subset 2 genotypes, a 2.69-fold change in slope. As regards to the ratio VAINH/fAINH, the R2 for the relationship was only 0.55 in the best model. We thus considered that our potential predictors such as clinical and demographic factors were not as explanatory of this parameter, and thus no further explorations were made.
Discussion In pharmacometrics, primary pharmacokinetic parameters are the observed values such as elimination rate constant (KINH), volume of distribution (both central and peripheral), and in the current study also Km and Vmax. These observations quantify rates and extents of underlying physicochemical processes. In other words, these parameters, including the enzyme reaction constants, are the biological phenotypes. On the other hand, in pharmacogenomics, inferred acetylation status is commonly termed the “acetylation phenotype.” Acetylation phenotype is a grouping of drug clearances (into 3 overlapping Gaussian curves). Moreover, commonly in clinical practice, acetylation status is assigned using 2-h post-dose isoniazid concentrations. However, the isoniazid elimination rate itself is a product of at least two physicochemical processes, acetylation and the loss of unmetabolised (parent) compound in urine. “Phenotyping” methods such as the 2-h concentration or concentration-time profiles or AUC0–24 are even further removed from the primary process of N-acetylation: they are products of several primary biological processes, namely absorption rate constant (and closely related time to maximum concentration), volume of distribution, and elimination rate constants. Only the last of these is related to acetylation, and only partially. The primary biological process of acetylation itself (i.e., the phenotype), sensu strictiore, is defined by Km and Vmax values, which are the “observed” values based on calculations using the Michaelis-Menten Eq. (1). These values are thus not inferred from genotypes, but are the “observed”. Thus, effect of genotype should be inferred based on these phenotypes, and not vice-versa. Here, we simply asked: what are the physiological and clinical factors that could affect these quantitative measures of phenotype, if at all, and which factors affect them the most? In terms of the genes, what groupings of NAT2 genes categorize the phenotypes of isoniazid NAT2 affinity and maximum velocity in children, during periods of highest growth? MARS, in which predictors do not take a pre-specified form or groupings (e.g. inferred acetylation status), is excellent at such an agnostic function, and utilizes only the presented data for its own categorization.