OBJECTIVES: Compare the benefits and harms of corticosteroids, oral and biologic disease-modifying antirheumatic drugs (DMARDs) for adults with rheumatoid arthritis.
DATA SOURCES: English-language articles from 1980 to February 2011 identified through PubMed, Embase, Cochrane Library, and International Pharmaceutical Abstracts; unpublished literature including dossiers from pharmaceutical companies.
METHODS: Two people independently selected relevant head-to-head trials of any sample size, prospective cohort studies with at least 100 participants, and relevant good- or fair-quality meta-analyses that compared benefits or harms of 14 drug therapies. Retrospective cohort studies were also included for harms. For biologic DMARDs, placebo-controlled, double-blind RCTs were also included. We required trials and cohort studies to have a study duration of at least 12 weeks. Literature was synthesized qualitatively within and between the two main drug classes (oral and biologic DMARDs). Network meta-analysis also was performed to examine the relative efficacy of biologic DMARDs and comparing withdrawal rates from placebo controlled trials.
RESULTS: Head-to-head trials showed no clinically important differences in efficacy among oral DMARD comparisons (methotrexate, sulfasalazine, leflunomide). The only head-to-head trial comparing biologic DMARDs (abatacept vs. infliximab) found no clinically important differences. Combination therapy of biologic DMARDs plus methotrexate improved clinical response rates and functional capacity more than monotherapy with methotrexate. Network meta-analyses found higher odds of reaching ACR 50 response for etanercept compared with most other biologic DMARDs (abatacept, adalimumab, anakinra, infliximab, rituximab, tocilizumab) for methotrexate-resistant patients with active rheumatoid arthritis. Similar overall tolerability profiles were found among oral and biologic DMARDs, but short-term adverse events were more common with biologic DMARDs. Adjusted indirect comparisons of biologic DMARDs found that certolizumab had the most favorable overall withdrawal profile, followed by etanercept and rituximab. Certolizumab had lower relative withdrawal rates due to lack of efficacy than adalimumab, anakinra, and infliximab. Certolizumab and infliximab had more, while etanercept had fewer withdrawals due to adverse events than most other drugs. Evidence was insufficient to assess comparative risk of serious adverse events among biologic DMARDs. Combinations of biologic DMARDs have higher rates of serious adverse events than biologic DMARD monotherapy. Limited data existed for subgroups.
CONCLUSIONS: Limited head-to-head comparative evidence does not support one therapy over another for adults with rheumatoid arthritis. Network meta-analyses from placebo-controlled trials of biologics suggest some differences, including higher odds of reaching ACR 50 response, but strength of evidence was low.
Síntesis amplia/ Revisión panorámica de revisiones sistemáticas
ANTECEDENTES: Los fármacos antirreumáticos modificadores de la enfermedad (FARMEs) biológicos son muy efectivos para el tratamiento de la artritis reumatoide (AR), aunque faltan estudios de comparaciones directas. OBJETIVOS: Comparar la eficacia y la seguridad de abatacept, adalimumab, anakinra, etanercept, infliximab y rituximab en los pacientes con AR. RESULTADOS PRINCIPALES: De las seis revisiones Cochrane disponibles se obtuvieron datos de siete estudios sobre abatacept, ocho sobre adalimumab, cinco sobre anakinra, cuatro sobre etanercept, cuatro sobre infliximab y tres sobre rituximab. Las estimaciones de las comparaciones indirectas mostraron una eficacia similar para el resultado primario de eficacia para todos los tratamientos biológicos con tres excepciones. La anakinra fue menos eficaz que el etanercept con una proporción de CR (IC del 95%; valor de p) de 0,44 (0,23 a 0,85; p = 0,014); la anakinra fue menos eficaz que el rituximab 0,45 (0,22 a 0,90; p = 0,023) y de manera similar el adalimumab fue más eficaz que la anakinra, 2,34 (1,32 a 4,13; P = 0,003).
En cuanto a la seguridad fue más probable que el adalimumab diera lugar a retiros comparado con el etanercept, con una proporción de OR de 1,89 (1,18 a 3,04; p = 0,009); la anakinra tuvo más probabilidades que el etanercept 2,05 (1,27 a 3,29; p = 0,003) y de manera similar el etanercept tuvo menos probabilidades que el infliximab 0,37 (0,19 a 0,70; P = 0,002). CONCLUSIONES DE LOS AUTORES: Según las comparaciones indirectas la anakinra pareció menos eficaz que etanercept, adalimumab y rituximab, y el etanercept pareció provocar menos retiros debido a eventos adversos que adalimumab, anakinra e infliximab. La heterogeneidad significativa en las características de las poblaciones de los ensayos implica que estos hallazgos se deben interpretar con cuidado. Estos hallazgos pueden informar a los médicos y los pacientes con respecto a la selección del tratamiento biológico para el tratamiento de la AR.
Síntesis amplia/ Revisión panorámica de revisiones sistemáticas/ Revisión sistemática
ANTECEDENTES: Hemos tratado de comparar los beneficios y la seguridad de 6 productos biológicos (abatacept, adalimumab, anakinra, etanercept, infliximab y rituximab) en pacientes con artritis reumatoide.
MÉTODOS: En este meta-análisis de la red, se incluyeron todos completos y actualizados revisiones Cochrane sobre tratamientos biológicos para la artritis reumatoide. Se incluyeron los datos de todos los ensayos controlados con placebo que utilizaron regímenes de dosis estándar. Los principales resultados fueron los beneficios (que se define como una mejora del 50% en el paciente y el médico-reporte de los criterios del American College of Rheumatology [ACR50]) y la seguridad (determinado por el número de retiros relacionados con eventos adversos). Se utilizó efectos mixtos de regresión logística para llevar a cabo una comparación indirecta de los efectos del tratamiento entre los biológicos.
RESULTADOS: En comparación con el placebo, los agentes biológicos se asociaron con un mayor índice de importancia clínica ACR50 (odds ratio [OR] 3,35, intervalo de confianza del 95% [IC]: 2.62-4.29) y un número necesario a tratar para el beneficio de 4 (95% CI 4 -6). Sin embargo, los productos biológicos se asociaron con más retiros relacionados con eventos adversos (OR 1,39, IC 95%: 1,13 a 1,71), con un número necesario a tratar para el daño de 52 (IC 95%: 29-152). La anakinra fue menos eficaz que todos los otros productos biológicos, aunque esta diferencia fue estadísticamente significativa sólo para la comparación con adalimumab (OR 0,45, IC 95% 0,21 a 0.99) y etanercept (OR 0,34, IC 95% 0.14-0.81). El adalimumab, anakinra e infliximab tuvieron más probabilidades de etanercept para liderar a los retiros relacionados con eventos adversos (OR 1,89 adalimumab, IC 95% 1.18-3.04; anakinra OR 2,05, IC 95% 1,27-3,29, y OR infliximab 2,70, IC del 95%: 1,43 -5,26).
INTERPRETACIÓN: Dadas las limitaciones de las comparaciones indirectas, anakinra fue menos eficaz que el etanercept y adalimumab, etanercept y era más seguro que adalimumab, anakinra e infliximab. Este resumen de la prueba ayudará a los médicos ya los pacientes a tomar decisiones basadas en la evidencia sobre productos biológicos para el tratamiento de la artritis reumatoide.
Compare the benefits and harms of corticosteroids, oral and biologic disease-modifying antirheumatic drugs (DMARDs) for adults with rheumatoid arthritis.
DATA SOURCES:
English-language articles from 1980 to February 2011 identified through PubMed, Embase, Cochrane Library, and International Pharmaceutical Abstracts; unpublished literature including dossiers from pharmaceutical companies.
METHODS:
Two people independently selected relevant head-to-head trials of any sample size, prospective cohort studies with at least 100 participants, and relevant good- or fair-quality meta-analyses that compared benefits or harms of 14 drug therapies. Retrospective cohort studies were also included for harms. For biologic DMARDs, placebo-controlled, double-blind RCTs were also included. We required trials and cohort studies to have a study duration of at least 12 weeks. Literature was synthesized qualitatively within and between the two main drug classes (oral and biologic DMARDs). Network meta-analysis also was performed to examine the relative efficacy of biologic DMARDs and comparing withdrawal rates from placebo controlled trials.
RESULTS:
Head-to-head trials showed no clinically important differences in efficacy among oral DMARD comparisons (methotrexate, sulfasalazine, leflunomide). The only head-to-head trial comparing biologic DMARDs (abatacept vs. infliximab) found no clinically important differences. Combination therapy of biologic DMARDs plus methotrexate improved clinical response rates and functional capacity more than monotherapy with methotrexate. Network meta-analyses found higher odds of reaching ACR 50 response for etanercept compared with most other biologic DMARDs (abatacept, adalimumab, anakinra, infliximab, rituximab, tocilizumab) for methotrexate-resistant patients with active rheumatoid arthritis. Similar overall tolerability profiles were found among oral and biologic DMARDs, but short-term adverse events were more common with biologic DMARDs. Adjusted indirect comparisons of biologic DMARDs found that certolizumab had the most favorable overall withdrawal profile, followed by etanercept and rituximab. Certolizumab had lower relative withdrawal rates due to lack of efficacy than adalimumab, anakinra, and infliximab. Certolizumab and infliximab had more, while etanercept had fewer withdrawals due to adverse events than most other drugs. Evidence was insufficient to assess comparative risk of serious adverse events among biologic DMARDs. Combinations of biologic DMARDs have higher rates of serious adverse events than biologic DMARD monotherapy. Limited data existed for subgroups.
CONCLUSIONS:
Limited head-to-head comparative evidence does not support one therapy over another for adults with rheumatoid arthritis. Network meta-analyses from placebo-controlled trials of biologics suggest some differences, including higher odds of reaching ACR 50 response, but strength of evidence was low.
