Background/Purpose: Methotrexate (MTX) is anti-folate activated to MTX polyglutamates (MTXPGs). MTX metabolism includes multiple enzyme-mediated reactions and genetic polymorphisms in these genes are linked to MTX disposition and effects. The objective of this study was to examine the influences of variant in the dihydrofolate reductase gene (DHFR: rs1382539 G/A) on MTX polyglutamation and clinical efficacy in rheumatoid arthritis (RA).

Methods: Two US based cohorts of consented adult RA were analyzed, a first cohort of 187 patients (median age 65 years under median 15 mg/weekly MTX) evaluated at a single visit for MTXPG levels (>3 months), and a second cohort consisting of 38 patients (mean age 55 years) enrolled in a dose escalation study (starting 7.5 mg/week) for 6 months (206 study visits). RBC MTXPG levels (MTXPG₁ [MTX] up to 5 glutamic residues [MTXPG₅]) were measured using liquid chromatography. MTXPG₃ (the preponderant MTXPG species) was expressed as nmol/L packed RBC; percent long-chain MTXPG_3-5 (over total MTXPG_1-5) and dose normalized MTXPG₃ levels (polyglutamation rate, nmol/L per mg) were also estimated. Real-time PCR was used to genotype the rs1382539 G/A variant in DHFR. Differences in MTXPG accumulation by genotypes were analyzed using Mann-Whitney test. Linear mixed effect with random intercept and fixed slope were used to analyze the impact of the variant on longitudinal changes in MTXPG levels and response (per DAS-28).

Results: In the first cohort, carriers of the rs1382539 A/A genotype (n=15 patients [8%], mean age 64±12 years under 15 mg/week) presented with 15 nmol/L lower MTXPG₃ levels (median 26 nmol/L [IQR: 18-55] vs 41 nmol/L [IQR: 28-60]; p=0.04) than those with G/G or G/A genotypes (total 173 patients mean age 64±12 years under 15 mg/week), respectively. Similarly, 16% lower long-chain MTXPG_3-5 (median 36% [IQR 18-48%] vs median 52% [IQR 40-68%]; p<0.01) and 0.8 nmol/L per mg lower MTX polyglutamation rate (median 2.1 nmol/mg MTX [IQR 1.2-3.1] vs 2.9 nmol/mg MTX [IQR 2.0-4.3]; p=0.02; Figure) were measured in carriers of the A/A genotype vs G/G or G/A genotypes, respectively (Figure, panel A).  In the second cohort, linear mixed effect models revealed that carriers of the rs1382539 A/A genotype (n=4 patients [10%]) also presented with lower RBC MTXPG₃ levels (estimate=-11±5 nmol/L; p=0.03), lower percent long-chain MTXPG_3-5 (estimate=-15±1%; p=0.02) and lower polyglutamation rate (estimate=-0.9±0.4 nmol/mg; p=0.01) than carriers of the G/G or G/A genotype (Figure, panel B). While the rs1382539 variant was not significantly associated with clinical efficacy (p>0.05), lower RBC MTXPG₃ level and percent long-chain MTXPG_3-5 levels associated with higher DAS28 in the cohort (p<0.05). 

Conclusion: Our data indicate that the rs1382539 G/A variant in DHFR impacts MTX polyglutamation in adult RA, and may indirectly contribute to clinical efficacy in some patients.