Rheumatology · Medical Literature
Subcutaneous methotrexate compared with oral methotrexate in rheumatoid arthritis: a systematic review and meta-analysis.
Abstract
To integrate evidence from randomized controlled trials (RCTs) comparing subcutaneous (SC) and oral methotrexate (MTX) for rheumatoid arthritis (RA) to provide optimal clinical treatment strategies. This meta-analysis conducted comprehensive search of PubMed, Web of Science, Embase and Cochrane Library, retrieving all relevant RCTs published up to June 18, 2025. Random-effects models were utilized to calculate the relative risk (RR), mean difference (MD) and their 95% confidence intervals (CIs), to evaluate efficacy and safety between subcutaneous injections and oral administration of MTX. A total of 1,034 articles were retrieved, and 9 RCTs that met the criteria were ultimately included, involving 974 RA patients in total. In the primary random-effects analyses, compared to oral MTX, subcutaneous MTX increased the ACR20 response rate (RR = 1.15; 95%CI: 1.05, 1.25), increased the ACR50 response rate (RR = 1.14; 95%CI: 1.01, 1.29), and reduced the incidence of gastrointestinal (GI)-related adverse event (AE) (RR = 0.58; 95%CI: 0.40, 0.83) and diarrhea (RR = 0.42; 95%CI: 0.21, 0.84). Fixed-effect sensitivity analyses supported the ACR20 and GI-related safety findings, but the ACR50 result was attenuated. Subcutaneous MTX did not show statistically significant differences in ACR70 response rate, DAS28-ESR score, bioavailability area under the curve, Cmax, or the incidence of other AE. Compared with oral administration, subcutaneous MTX was associated with a higher ACR20 response and lower GI-related AE and diarrhea in the primary random-effects analyses. These findings suggested that subcutaneous MTX may be an effective and generally well-tolerated option, particularly for patients with inadequate response or poor gastrointestinal tolerability to oral MTX, while evidence for ACR50, ACR70, DAS28-ESR and bioavailability remains less certain.
Clinical relevance
In adults with rheumatoid arthritis, does subcutaneous (SC) methotrexate produce superior clinical response rates and a more favorable gastrointestinal safety profile compared with oral methotrexate when evaluated across randomized controlled trials? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis of randomized controlled trials (RCTs); random-effects and fixed-effect models applied","setting":"Multi-database search (PubMed, Web of Science, Embase, Cochrane Library); literature search through June 18, 2025; international RCT populations","sampleSize":"n=974 RA patients across 9 included RCTs (intervention and control arm sizes not individually reported in abstract)","population":"Adults with rheumatoid arthritis enrolled in RCTs comparing SC versus oral MTX; specific eligibility criteria (disease duration, disease activity thresholds, MTX-naive vs. MTX-experienced status) not detailed in abstract","intervention":"Subcutaneous (SC) methotrexate (doses and frequency not specified in abstract)","comparator":"Oral methotrexate (doses and frequency not specified in abstract)","primaryOutcome":"ACR20 and ACR50 response rates; incidence of gastrointestinal (GI)-related adverse events including diarrhea","followUp":"Not reported","fundingSource":"Not reported"}
Key findings
In the primary random-effects meta-analysis of 9 RCTs (n=974), SC MTX was associated with a statistically significant improvement in ACR20 response rate compared with oral MTX (RR=1.15; 95% CI: 1.05–1.25), representing a 15% relative increase in the probability of achieving an ACR20 response. Similarly, SC MTX increased the ACR50 response rate (RR=1.14; 95% CI: 1.01–1.29), a result that is statistically significant but reflects a narrower margin of benefit, with the lower bound of the confidence interval barely crossing 1.0, indicating some uncertainty. Absolute risk differences for ACR20 and ACR50 cannot be calculated without event rates per arm, which are not reported in the abstract. On the safety side, SC MTX significantly reduced GI-related adverse events overall (RR=0.58; 95% CI: 0.40–0.83) and diarrhea specifically (RR=0.42; 95% CI: 0.21–0.84), suggesting a 42% relative reduction in GI AEs and a 58% relative reduction in diarrhea with SC versus oral administration. The abstract notes that fixed-effect sensitivity analyses supported the ACR20 and GI AE findings (full sensitivity analysis results for all endpoints not provided in abstract). P-values, NNT, NNH, and heterogeneity statistics (I²) are not reported in the abstract.
Practice implications
This meta-analysis provides pooled RCT-level evidence that SC MTX offers modest but statistically significant advantages over oral MTX in achieving ACR20 and ACR50 responses, and is associated with fewer GI adverse events in RA. These findings align with the pharmacokinetic rationale: SC MTX has higher and less variable bioavailability than oral MTX at equivalent doses, a difference that is most pronounced at medium-to-high doses (>15 mg/week). However, several important caveats limit the strength of inference. First, only an abstract is available for this review; complete methodology, heterogeneity data (I² values), risk-of-bias assessments, and subgroup analyses cannot be evaluated. Second, the aggregate sample size of 974 patients across 9 RCTs is modest for a meta-analysis in a condition as heterogeneous as RA. Third, absolute risk differences and NNT/NNH cannot be derived from the reported data. Fourth, the ACR50 finding (RR=1.14; 95% CI: 1.01–1.29) is statistically marginal. Fifth, the meta-analysis does not characterize which patient subgroups benefit most — for example, MTX-naive patients versus those switching from oral due to intolerance or inadequate response. Finally, the comparative tolerability advantage of SC MTX must be weighed against practical considerations such as patient injection burden, cost, and access. Whether results should immediately change practice depends on individual clinical context; SC MTX is already preferred or guideline-supported in many centers for patients with suboptimal oral response or GI intolerance, and this meta-analysis reinforces that practice pattern. Routine first-line use of SC over oral MTX in all RA patients is not clearly established by this evidence alone, particularly in the absence of ACR70 data, DAS28 remission rates, radiographic outcomes, drug retention rates, or long-term safety data. __LIMITATIONS__ Only the abstract is available; full methodology, PRISMA adherence, risk-of-bias assessments, and heterogeneity statistics (I²) cannot be independently evaluated.,Small aggregate sample size (n=974 across 9 RCTs) limits statistical power, particularly for subgroup analyses.,MTX doses, dose escalation strategies, and treatment durations across included trials are not reported in the abstract, making it impossible to assess dose-response effects or applicability to specific dosing regimens.,ACR50 confidence interval barely excludes the null (lower bound 1.01), suggesting fragility of this finding and potential sensitivity to single-study influence.,Absolute event rates, absolute risk differences, NNT, and NNH are not calculable from the reported data, limiting direct clinical translation.,Heterogeneity of included patient populations (MTX-naive vs. experienced, early vs. established RA, varying disease activity at baseline) is not characterized in the abstract.,Outcomes critical to clinical decision-making — including ACR70, DAS28 remission, radiographic progression, drug retention, and serious adverse events (hepatotoxicity, myelosuppression) — are not addressed in the abstract.,Funding source is not reported; potential conflicts of interest among study authors or included RCT sponsors cannot be assessed.,Publication bias assessment (e.g., funnel plot analysis) is not referenced in the abstract.,Generalizability is uncertain: patient demographics, disease duration, concomitant DMARDs, and geographic/practice setting variation across included RCTs are not described. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ This meta-analysis of 9 RCTs (n=974) demonstrates that SC methotrexate produces modestly higher ACR20 and ACR50 response rates and significantly fewer GI adverse events compared with oral methotrexate in rheumatoid arthritis. These findings support the existing clinical rationale for preferring SC MTX in patients with inadequate oral response or GI intolerance, consistent with current guideline-aligned practice in many rheumatology centers. However, the modest absolute sample size, absence of full publication details, and lack of data on ACR70, remission, radiographic, and long-term safety outcomes limit the ability to advocate for universal first-line SC MTX across all RA patients without further confirmatory evidence.