Mette Marie Hougaard Christensen

The pharmacogenetics of metformin and its impact on plasma metformin steady-state levels and glycosylated hemoglobin A1c

OBJECTIVE The aim of this study was to evaluate the effect of genetic variations in OCT1, OCT2, MATE1, MATE 2, and PMAT on the trough steady-state plasma concentration of metformin and hemoglobin A1c (Hb1Ac). METHOD The South Danish Diabetes Study was a 2 x 2 x 2 factorial, prospective, randomized, double-blind, placebo-controlled, multicentre study. One hundred and fifty-nine patients received 1 g of metformin, twice daily continuously, and 415 repeated plasma metformin measurements were obtained after 3, 6, and 9 months of treatment. RESULTS The mean trough steady-state metformin plasma concentration was estimated to be 576 ng/ml (range, 54–4133 ng/ml, p = 0.55) and correlated to the number of reduced function alleles in OCT1 (none, one or two: 642, 542, 397 ng/ml; P = 0.001). The absolute decrease in Hb1Ac both initially and long term was also correlated to the number of reduced function alleles in OCT1 resulting in diminished pharmacodynamic effect of metformin after 6 and 24 months. CONCLUSION In a large cohort of type 2 diabetics, we either confirm or show for the first time: (a) an enormous (80-fold) variability in trough steady-state metformin plasma concentration, (b) OCT1 activity affects metformin steady-state pharmacokinetics, and (c) OCT1 genotype has a bearing on HbA1c during metformin treatment.

A gene-gene interaction between polymorphisms in the OCT2 and MATE1 genes influences the renal clearance of metformin.

Objective The aim of this study was to determine the association between the renal clearance (CLrenal) of metformin in healthy Caucasian volunteers and the single-nucleotide polymorphism (SNP) c.808G>T (rs316019) in OCT2 as well as the relevance of the gene–gene interactions between this SNP and (a) the promoter SNP g.-66T>C (rs2252281) in MATE1 and (b) the OCT1 reduced-function diplotypes. Methods Fifty healthy volunteers genotyped for the c.808G>T were enrolled in the study. The distribution was 25 GG, 20 GT, and 5 TT volunteers. The pharmacokinetics of a 500 mg single oral dose of metformin was studied. Results When analyzed alone, the c.808 (G>T) affected neither the CLrenal nor the secretory clearance (CLsec) of metformin. However, both CLrenal and CLsec were increased for the volunteers with minor alleles in c.808 (G>T) who were also homozygous for the reference variant g.-66T>C: CLrenal: GG, GT, and TT: 28.1, 34.5, and 44.8 l/h (P=0.004), respectively and CLsec: GG, GT, and TT: 21.4, 27.8, and 37.6 l/h (P=0.005), respectively. In the volunteers with minor alleles in c.808 (G>T) who were also heterozygous for g.-66T>C, both CLrenal and CLsec were found to be reduced (P<0.028) when compared with volunteers with minor alleles in c.808 (G>T) carrying the g.-66T>C reference genotype. Conclusion We report counteracting effects of the c.808 (G>T) and g.-66T>C on the renal elimination of metformin. When adjusted for the genetic variation g.-66T>C, our results suggest that c.808 (G>T) could have a dominant genotype to phenotype correlation.

The role of genetic variants in CYP2C8, LPIN1, PPARGC1A and PPARγ on the trough steady-state plasma concentrations of rosiglitazone and on glycosylated haemoglobin A1c in type 2 diabetes.

Objective The aim of this study was to examine the effect of single nucleotide polymorphisms in CYP2C8, LPIN1, PPARGC1A and PPAR&ggr; on rosiglitazone’s (i) trough steady-state plasma concentration (Css,min), (ii) on glycosylated haemoglobin A1c (HbA1c) and (iii) the risk of developing adverse events, mainly oedema, in patients with type 2 diabetes mellitus (T2D). Methods The data used in this study were obtained from the South Danish Diabetes Study including 371 T2D patients with a focus on the 187 patients who were treated with rosiglitazone. The study was a placebo-controlled, partly blinded and multicentre clinical trial. The Css,min of rosiglitazone and HbA1c was determined and the genotype of the patients was identified. Results The mean Css,min of rosiglitazone was 21.3 ng/ml (95% confidence interval 18.8; 24.2 ng/ml), with observations ranging from 1 to 296 ng/ml. Carriers of CYP2C8*3 (n=32) (rs10509681 and rs11572080) had a statistically significantly lower mean Css,min than wild types (n=106), and they also had a statistically significantly lower mean absolute difference in HbA1c during rosiglitazone treatment. Finally, the carriers of CYP2C8*3 had a lower odds ratio of developing oedema. Conclusion We showed that CYP2C8*3 was associated with lower plasma levels of rosiglitazone and hence a reduced therapeutic response but also a lower risk of developing oedema during treatment with rosiglitazone. Individualized treatment with rosiglitazone on the basis of the CYP2C8 genotype may therefore be possible.

In Vivo Imaging of Human 11C-Metformin in Peripheral Organs: Dosimetry, Biodistribution, and Kinetic Analyses

Metformin is the most widely prescribed oral antiglycemic drug, with few adverse effects. However, surprisingly little is known about its human biodistribution and target tissue metabolism. In animal experiments, we have shown that metformin can be labeled by 11C and that 11C-metformin PET can be used to measure renal function. Here, we extend these preclinical findings by a first-in-human 11C-metformin PET dosimetry, biodistribution, and tissue kinetics study. Methods: Nine subjects (3 women and 6 men) participated in 2 studies: in the first study, human radiation dosimetry and biodistribution of 11C-metformin were estimated in 4 subjects (2 women and 2 men) by whole-body PET. In the second study, 11C-metformin tissue kinetics were measured in response to both intravenous and oral radiotracer administration. A dynamic PET scan with a field of view covering target tissues of metformin (liver, kidneys, intestines, and skeletal muscle) was obtained for 90 (intravenous) and 120 (oral) min. Results: Radiation dosimetry was acceptable, with effective doses of 9.5 μSv/MBq (intravenous administration) and 18.1 μSv/MBq (oral administration). Whole-body PET revealed that 11C-metformin was primarily taken up by the kidneys, urinary bladder, and liver but also to a lesser extent in salivary glands, skeletal muscle, and intestines. Reversible 2-tissue-compartment kinetics was observed in the liver, and volume of distribution was calculated to be 2.45 mL/mL (arterial input) or 2.66 mL/mL (portal and arterial input). In the kidneys, compartmental models did not adequately fit the experimental data, and volume of distribution was therefore estimated by a linear approach to be 6.83 mL/mL. Skeletal muscle and intestinal tissue kinetics were best described by 2-tissue-compartment kinetics and showed only discrete tracer uptake. Liver 11C-metformin uptake was pronounced after oral administration of the tracer, with tissue-to-blood ratio double what was observed after intravenous administration. Only slow accumulation of 11C-metformin was observed in muscle. There was no elimination of 11C-metformin through the bile both during the intravenous and during the oral part of the study. Conclusion: 11C-metformin is suitable for imaging metformin uptake in target tissues and may prove a valuable tool to assess the impact of metformin treatment in patients with varying metformin transport capacity.

Intake of St John's wort improves the glucose tolerance in healthy subjects who ingest metformin compared with metformin alone

AIMS Our objective was to investigate the steady-state pharmacokinetic and pharmacodynamic interaction between the antidepressive herbal medicine St Johns wort and the antidiabetic drug metformin. METHODS We performed an open cross-over study in 20 healthy male subjects, who received 1 g of metformin twice daily for 1 week with and without 21 days of preceding and concomitant treatment with St Johns wort. The pharmacokinetics of metformin was determined, and a 2 h oral glucose tolerance test was performed. RESULTS St Johns wort decreased the renal clearance of metformin but did not affect any other metformin pharmacokinetic parameter. The addition of St Johns wort decreased the area under the glucose concentration-time curve [702 (95% confidence interval, 643-761) vs. 629 min*mmol/L (95% confidence interval, 568-690), P = 0.003], and this effect was caused by a statistically significant increase in the acute insulin response. CONCLUSIONS St Johns wort improves glucose tolerance by enhancing insulin secretion independently of insulin sensitivity in healthy male subjects taking metformin.

Genetic Polymorphisms in Organic Cation Transporter 1 Attenuates Hepatic Metformin Exposure in Humans

Metformin has been used successfully to treat type 2 diabetes for decades. However, the efficacy of the drug varies considerably from patient to patient and this may in part be due to its pharmacokinetic properties. The aim of this study was to examine if common polymorphisms in SLC22A1, encoding the transporter protein OCT1, affect the hepatic distribution of metformin in humans. We performed noninvasive 11C‐metformin positron emission tomography (PET)/computed tomography (CT) to determine hepatic exposure in 12 subjects genotyped for variants in SLC22A1. Hepatic distribution of metformin was significantly reduced after oral intake in carriers of M420del and R61C variants in SLC22A1 without being associated with changes in circulating levels of metformin. Our data show that genetic polymorphisms in transporter proteins cause variation in hepatic exposure to metformin, and it demonstrates the application of novel imaging techniques to investigate pharmacogenetic properties in humans.

Metformin-associated lactic acidosis (MALA): Moving towards a new paradigm

Although metformin has been used for over 60 years, the balance between the drugs beneficial and adverse effects is still subject to debate. Following an analysis of how cases of so‐called “metformin‐associated lactic acidosis” (MALA) are reported in the literature, the present article reviews the pitfalls to be avoided when assessing the purported association between metformin and lactic acidosis. By starting from pathophysiological considerations, we propose a new paradigm for lactic acidosis in metformin‐treated patients. Metformin therapy does not necessarily induce metformin accumulation, just as metformin accumulation does not necessarily induce hyperlactatemia, and hyperlactatemia does not necessarily induce lactic acidosis. In contrast to the conventional view, MALA probably accounts for a smaller proportion of cases than either metformin‐unrelated lactic acidosis or metformin‐induced lactic acidosis. Lastly, this review highlights the need for substantial improvements in the reporting of cases of lactic acidosis in metformin‐treated patients. Accordingly, we propose a check‐list as a guide to clinical practice.

Quantitation of Metformin in Human Plasma and Urine by Hydrophilic Interaction Liquid Chromatography and Application to a Pharmacokinetic Study

Abstract: We describe an analytical method for the quantification of the widely used antihyperglycemic agent, metformin, in human plasma and urine. The separation was performed using isocratic hydrophilic interaction liquid chromatography on a Luna hydrophilic interaction liquid chromatography column (125 × 4.6 mm, 3 &mgr;m). The sample preparation was accomplished by solid-phase extraction. Validation of the method was performed in the range 10–2000 ng/mL for plasma and 5–30 mcg/mL for urine. The methods were linear within the investigated range (r2 > 0.988). Within-day repeatability ranged from 3.1% to 7.5% in plasma and 1.6% to 6.2% in urine. Between-day reproducibility ranged from 2.9% to 5.3% in plasma and 0.6% to 1.8% in urine. The inaccuracy expressed as bias ranged from −3.1% to 1.9% in plasma and from −7.2% to 0.7% in urine. The lower limit of quantification for metformin in plasma was 5 ng/mL and in urine was 40 ng/mL. The method was therefore considered to be precise, accurate, reproducible, and sensitive enough to be appropriate for pharmacokinetic studies of metformin. The applicability of the method for human pharmacokinetic studies was demonstrated by dosing a healthy male volunteer with 500-mg metformin hydrochloride as a single oral dose; plasma and urine concentrations were measured for 24 hours.

Linkage disequilibrium between the CYP2C19*17 allele and other clinically important CYP2C allelic variants in a healthy Scandinavian population

In a previous study, in three Nordic populations [1], we determined the genotype and allele frequencies of the five most relevant CYP2C allelic variants (CYP2C8*3, CYP2C9*2, CYP2C9*3, CYP2C19*2, CYP2C19*17) and inferred the most likely haplotypes. CYP2C19*17 was found at a frequency of 15–22% and inferred to be a haplo-type marker not present together with the four defective alleles and thereby predicting efficient CYP2C substrate metabolism. Subsequently, Suarez-Kurtz [2] reported in populations of African descent a haplotype containing CYP2C19*17 and the reduced-activity allele CYP2C8*2, which opposes CYP2C19*17 as a marker for extensive CYP2C8 substrate metabolism. The present study was conducted in order to prospectively validate the previous findings and determine Nordic haplotypes containing the clinically important allelic variants CYP2C8*2 and CYP2C8*4. From a newly established biobank, 394 healthy Possible deviation from Hardy-Weinberg equilibrium was tested for each SNP with Pearsons chi-squared test with a level of significance of 5%, and the individual haplotypes were inferred by use of the software package PHASE version 2.1 [3, 4]. All seven CYP2C allelic variants were present in the Scandinavian population and tested to be in Hardy-Weinberg equilibrium. The allele frequencies with binomial confidence intervals were as follows: CYP2C8*3: 0.090 which were similar to our previous findings [1]. The allele frequencies of the two added CYP2C8 alleles were CYP2C8*2: 0.0051 (0.0014–0.013) and CYP2C8*4: 0.043 (0.030–0.060). As shown in Table 1, the apparent CYP2C wild-type haplotype was the most frequent (47%). The second most frequent haplotype (no. 2), with a frequency of 18%, was composed of CYP2C19*17 and apparent CYP2C8 and CYP2C9 wild-type alleles, and represented 97% (144 out of 148) of all predicted CYP2C19*17-containing haplotypes. Four rare CYP2C8*2 alleles were detected and inferred to be exclusively in haplotype with the remaining four CYP2C19*17 alleles. The 34 detected CYP2C8*4 alleles were inferred in two different CYP2C haplotypes: 16 CYP2C*4 alleles were inferred to represent a haplotype without any of the 6 other clinically important CYP2C SNPs (no. 8), and 18 CYP2C*4 alleles were inferred together with CYP2C19*2 (no. 7). The third haplotype composed of two CYP2C allelic variants (no. 4) also confirms previous findings [5] showing that CYP2C8*3 is present almost exclusively in haplotypes with CYP2C9*2 (70 of 71 alleles: 99%), and CYP2C9*2 appears most often in haplotypes with CYP2C8*3 (70 of 89 alleles: 79%). This study supports our previous findings that CYP2C19*17 is a frequent allelic variant in Scandinavia not in haplotype with the clinically …

Metformin, but not rosiglitazone, attenuates the increasing plasma levels of a new cardiovascular marker, fibulin-1, in patients with type 2 diabetes

OBJECTIVE The extracellular matrix protein fibulin-1 is upregulated in the arterial wall in type 2 diabetes (T2D) and circulates in increased concentrations in diabetes. Metformin is an antidiabetic drug with beneficial cardiovascular disease effects in diabetes. We hypothesized that metformin would influence the increased level of plasma fibulin-1 in diabetes. RESEARCH DESIGN AND METHODS After a 4-week run-in period, 371 eligible patients with T2D were randomized to treatment groups in a factorial design including insulin alone (control), +metformin, +rosiglitazone, or +both metformin and rosiglitazone. Plasma fibulin-1 was analyzed at the beginning of the study and after 18 and 24 months. RESULTS Plasma fibulin-1 increased in all groups throughout the 2-year period; however, the increase was strongly attenuated among patients treated with metformin. A highly significant difference was observed when the mean change in plasma fibulin-1 was compared between metformin- and non–metformin-treated individuals both at 18 and 24 months of treatment, but rosiglitazone had no effect. Metformin and rosiglitazone alone reduced the HbA1c levels to comparable levels and in combination even further. CONCLUSIONS Metformin attenuates the increase in plasma fibulin-1 concentrations in T2D, independently of glycemic effects. Changes in fibulin-1 may reflect an important element in diabetic arteriopathy that can be influenced by metformin.