the presence of either polymorphic SLC47A1 with wild‐type SLC22A2 or wild‐type SLC47A1 with polymorphic SLC22A2 exhibited increased metformin Cmax and exposures in the presence of trimethoprim. Hence, the combination of transporter genetics may result in observed differences with potent MATE inhibitors.
3.7.2 Metformin Efficacy
3.7.2.1 Diabetes
The first large study to report an association between SLC47A1 polymorphisms and changes in glucose control, as noted by hemoglobin A1c, in diabetics treated with metformin was informed from the European Rotterdam Study (n = 116) [108]. The rs2289669 G > A polymorphism was significantly associated with metformin response. For each minor A allele, hemoglobin A1c reduction was 0.3% greater. Subsequently, relationships between OCT1 (SLC22A1; rs622342 A > C) and SLC47A1 (rs2289669; G > A) polymorphisms and effects on glucose lowering were investigated by the Rotterdam investigators [109]. The study used multiplicative interaction analyses to evaluate changes in HbA1c in patients newly prescribed metformin (n = 98). The decrease in hemoglobin A1c levels was greatest for each SLC22A1 A allele (wild type) and SLC47A1 A allele (variant). This study demonstrated the interaction between polymorphisms in two transporter genes in determining efficacy of treatment with metformin. Despite these findings associating glycemic control with polymorphisms in MATE and OCT2 transport genes in patients undergoing therapy with metformin, a large meta‐analysis in Europeans failed to detect these associations in 7,968 subjects in the Metformin Genetics Consortium (MetGen) [110].
Two studies by separate investigators evaluated associations between human genetic variants in SLC47A1 and SLC47A2 and MATE1/MATE2/2‐K transport function [111, 112]. Subsequently, the influence of promoter variants in SLC47A1 (−66 T > C, rs2252281) and SLC47A2 (−130 G > A, rs12943590) on metformin disposition and response was reported [113]. Healthy volunteers (n = 57) receiving metformin exhibited increased renal and secretory clearances if they were carriers of the variant SLC47A2 and wild‐type for SLC47A1. Significant differences in oral glucose tolerance responses were demonstrated, with carriers of SLC47A1 variants having increased and SLC47A2 variants having decreased responses to metformin. The authors also reported better HbA1c responses in patients with diabetes (n = 145) carrying the SLC47A1 promoter variant suggesting improved pharmacotherapeutic outcomes. Similarly, promoter variants (n = 8) in SLC47A2 have also been characterized for their effects on metformin pharmacodynamics in a diverse US population (n = 272) [114]. Likewise, coding variants 485 C > T (rs19556609) and 1177 G > A (19525563) had significantly lowered metformin uptake and reduced protein expression in vitro. Addition of the −130 G > A promoter polymorphism to SLC47A2 haplotypes increased luciferase activities and reduced binding to the transcriptional repressor myeloid zinc finger 1. Similarly, diabetic patients who were homozygous for the −130 G > A variant (rs19560590) had a significantly reduced relative change in hemoglobin A1c compared with homozygous wild type or heterozygous individuals.
A large study in diabetic patients (800 cases and 400 controls) of Pakistani descent evaluated the influence of SLC47A2/MATE2‐K (rs138244461) on the efficacy of metformin [115]. Efficacy of diabetes was reported through hemoglobin A1c, random blood glucose, and fasting blood glucose measurements. Efficacy of metformin on cholesterol (total, HDL, LDL) and triglyceride profiles were also assessed. Study results showed that subjects homozygous for the wild‐type genotype (C/C) had statistically significant improvements of their diabetes versus patients who were heterozygous or homozygous for T/T.
A study was conducted to determine the allelic frequencies of the SLC47A1 (816 G > A; rs2289669) and SLC47A2 (−130 G > A; rs12943590) polymorphism in a South Indian population (n = 102) and found highly different prevalence than Chinese, Nigerian, and Northern and Western European ancestry [116]. Subsequently, the authors conducted a study evaluating SLC47A1 (816 G > A; rs2289669) and SLC47A2 (−130 G > A; rs12943590) on changes in fasting blood glucose, hemoglobin A1c, dosage requirements, and toxicities in type 2 diabetes mellitus patients (n = 105) from South India receiving metformin monotherapy [117]. The differences in fasting blood glucose and hemoglobin A1c between the G/G and AG/AA genotype groups for both SLC47A1 and SLC47A2, while greater, were not considered significant. Metformin dosing requirements and adverse events were also similar across genotype groups. By comparison, another study conducted in a South Indian population of metformin‐naïve Type 2 diabetic patients (n = 221) [118]. SLC47A2 (rs12943590) across a codominant model was significantly associated with better hemoglobin A1c response. An interaction analysis between the SLC22A1 (rs622342; A > C) and SLC47A1 (rs2289669, rs8065082) polymorphisms and between SLC22A2 (rs316019) and SLC47A2 and metformin response demonstrated hemoglobin A1c changes were greatest for SLC22A2 G/G genotype and SLC47A2 G/A genotype. The results demonstrate the power of assessing genetics of several transporters in determining associations with metformin response.
Chinese patients (n = 53) who were recently diagnosed with Type 2 diabetes and treated with metformin monotherapy were evaluated for associations between SLC22A1 (rs594709; G > A) and SLC47A1 (rs2289669) polymorphisms and treatment effects (glucose, lipids, insulin sensitivity) [119]. While SLC22A1 rs594709 G/G patients had larger increases in fasting insulin and greater decreases in insulin sensitivity than A/A or A/G genotypes, patients with SLC47A1 rs2289669 A/A had decreased fasting blood glucose and post‐prandial insulin than A/G or G/G. SLC47A1 rs2289669 G/G patients had a greater decrease in total cholesterol and low‐density lipoprotein cholesterol than A/G or A/A genotypes. These data suggested the interplay between polymorphisms in SLC22A1 and SLC47A1 and beneficial metabolic responses to metformin. Another study in patients (n = 291) receiving monotherapy with metformin were assessed for changes in hemoglobin A1c and basal glucagon‐like peptide (GLP‐1) according to the rs2289669 genetic polymorphism in SLC47A1 [120]. Carriers of the variant A allele (G/A and A/A) exhibited better hemoglobin A1c reductions and had larger increases in basal GLP‐1. After adjustment for multiple confounding variables, a significant association between the polymorphism and changes in hemoglobin A1c was observed. The authors also concluded that differential responses (increases) in basal GLP‐1, while not directly associated with changes to hemoglobin A1c, may contribute to inter‐individual responses to metformin.
The SLC47A1 variant (rs2289669; 816 G > A) can also impact metformin responses in Iranian subjects (n = 71) [121] and Chinese subjects [122]. After 6 months of therapy with metformin, patients who were heterozygous (A/G) for the noncoding polymorphism showed the greatest reduction in hemoglobin A1c. The study genetic demographics reported only 10 patients with the A/A genotype, which may have limited the ability to detect the full homozygous effect of the variant. An additional Chinese study studied the influence of SLC47A1 (rs2289669) on metformin pharmacokinetics and pharmacologic effects [122]. Newly diagnosed Type 2 diabetes patients (n = 220) were divided into SLC47A1 genotype groups (G/G, G/A, A/A). A subgroup of patients (10 per group) was evaluated for metformin pharmacokinetics. HbA1c level decrease was significantly greater in A/A (−2.32% vs G/A −1.16% vs G/G −1.07%, P < 0.05). Associations existed after adjustments for age, sex, body mass index, baseline HbA1c, and diabetes duration (P < 0.05). Patients with SLC47A1 A/A had higher metformin exposures (17,312 μg h/l vs 11,731 μg h/l vs 10,962 μg h/l) and lower renal secretory clearance (25.80 l/h vs 51.43 l/h vs 42.52 l/h) than other genotype groups. The results from metformin pharmacokinetics and hemoglobin A1c outcomes and alignment with the incorporation of the SCL47A1 genetics data suggest strong evidence for the effects of the polymorphism in the clinical environment in Chinese patients.
3.7.2.2 Nondiabetes Indications
The impact of SLC22A1/OCT1 and SLC47A1/MATE1 polymorphisms on clinical response to metformin in patients with castration‐resistant prostate cancer was evaluated [123]. Patients (n = 36) received metformin until disease progression or toxicity (diarrhea, bloating, anorexia, nausea, fatigue). Disease progression