2012 Update of the 2008 American College of Rheumatology Recommendations for the Use of Disease‐modifying Antirheumatic Drugs and Biologic Agents in the Treatment of Rheumatoid Arthritis

Guidelines and recommendations developed and/or endorsed by the American College of Rheumatology (ACR) are intended to provide guidance for particular patterns of practice and not to dictate the care of a particular patient. The ACR considers adherence to these guidelines and recommendations to be voluntary, with the ultimate determination regarding their application to be made by the physician in light of each patient's individual circumstances. Guidelines and recommendations are intended to promote beneficial or desirable outcomes but cannot guarantee any specific outcome. Guidelines and recommendations developed or endorsed by the ACR are subject to periodic revision as warranted by the evolution of medical knowledge, technology, and practice. The American College of Rheumatology is an independent, professional, medical and scientific society which does not guarantee, warrant, or endorse any commercial product or service. The American College of Rheumatology (ACR) most recently published recommendations for the use of disease-modifying antirheumatic drugs (DMARDs) and biologic agents in the treatment of rheumatoid arthritis (RA) in 2008 (1). These recommendations covered indications for use, monitoring of side effects, assessment of the clinical response to DMARDs and biologic agents, screening for tuberculosis (TB), and assessment of the roles of cost and patient preference in decision making for biologic agents (1). Recognizing the rapidly evolving knowledge in RA management and the accumulation of new evidence regarding the safety and efficacy of existing and newer therapies, the ACR commissioned an update of the 2008 recommendations in select topic areas. The 2012 revision updates the 2008 ACR recommendations in the following areas: 1) indications for DMARDs and biologic agents, 2) switching between DMARD and biologic therapies, 3) use of biologic agents in high-risk patients (those with hepatitis, congestive heart failure [CHF], and malignancy), 4) screening for TB in patients starting or currently receiving biologic agents, and 5) vaccination in patients starting or currently receiving DMARDs or biologic agents (Table 1). Hydroxychloroquine Leflunomide Methotrexate Minocycline Sulfasalazine And, when appropriate, combination DMARD therapy with 2 or 3 DMARDs Hydroxychloroquine Leflunomide Methotrexate Minocycline Sulfasalazine And, when appropriate, combination DMARD therapy with 2 or 3 DMARDs Abatacept Rituximab Anti-TNF Adalimumab Etanercept Infliximab Abatacept Rituximab Tocilizumab Anti-TNF Adalimumab Etanercept Infliximab Certolizumab pegol Golimumab These 2012 recommendations update the 2008 American College of Rheumatology recommendations for the treatment of rheumatoid arthritis (RA). The recommendations cover the use of disease-modifying antirheumatic drugs and biologic agents in patients with RA, including switching between drugs. We address screening for tuberculosis reactivation, immunization, and treatment of RA patients with hepatitis, congestive heart failure, and/or malignancy in these recommendations, given their importance in RA patients receiving or starting biologic agents. We utilized the same methodology as described in detail in the 2008 guidelines (1) to maintain consistency and to allow cumulative evidence to inform this 2012 recommendations update. These recommendations were developed by 2 expert panels: 1) a nonvoting working group and Core Expert Panel (CEP) of clinicians and methodologists responsible for the selection of the relevant topic areas to be considered, the systematic literature review, the evidence synthesis, and creation of “clinical scenarios”; and 2) a Task Force Panel (TFP) of 11 internationally recognized expert clinicians, patient representatives, and methodologists with expertise in RA treatment, evidence-based medicine, and patient preferences who were tasked with rating the scenarios created using an ordinal scale specified in the RAND/University of California at Los Angeles (RAND/UCLA) Appropriateness Method (2-4). This method solicited formal input from this multidisciplinary TFP to make recommendations informed by the evidence. The methods used to develop the updated ACR recommendations are described briefly below. Literature searches for both DMARDs and biologic agents relied predominantly on PubMed searches with medical subject headings and relevant keywords similar to those used for the 2008 ACR RA recommendations (see Supplementary Appendices 1 and 2, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). We included randomized controlled trials (RCTs), controlled clinical trials, quasi-experimental designs, cohort studies (prospective or retrospective), and case–control studies, with no restrictions on sample size. More details about inclusion criteria are listed below and in Supplementary Appendix 3 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). The 2008 recommendations were based on a literature search that ended on February 14, 2007. The literature search end date for the 2012 update was February 26, 2010 for the efficacy and safety studies and September 22, 2010 for additional qualitative reviews related to TB screening, immunization, and hepatitis (similar to the 2008 methodology). Studies published subsequent to that date were not included. For biologic agents, we also reviewed the Cochrane systematic reviews and overviews (published and in press) in the Cochrane Database of Systematic Reviews to identify additional studies (5-8) and further supplemented by hand checking the bibliographies of all included articles. Finally, the CEP and TFP confirmed that the relevant literature was included in the evidence synthesis. Unless they were identified by the literature search and met the article inclusion criteria (see Supplementary Appendix 3, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658), we did not review any unpublished data from product manufacturers, investigators, or the Food and Drug Administration (FDA) Adverse Event Reporting System. We searched the literature for the 8 DMARDs and 9 biologic agents most commonly used for the treatment of RA. Literature was searched for 8 DMARDs: azathioprine, cyclosporine, hydroxychloroquine, leflunomide, methotrexate, minocycline, organic gold compounds, and sulfasalazine. Similar to 2008, azathioprine, cyclosporine, and gold were not included in the recommendations based on their infrequent use and lack of new data (Table 1). Literature was searched for 9 biologic agents: abatacept, adalimumab, anakinra, certolizumab pegol, etanercept, golimumab, infliximab, rituximab, and tocilizumab. Anakinra was not included in the recommendations due to infrequent use and lack of new data. Details of the bibliographic search strategy are listed in Supplementary Appendix 1 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). With the exception of assessment of TB, hepatitis, and vaccination (see below), studies were included if they met all of the following criteria: 1) original study in English language with an abstract, 2) observational studies (case–control or cohort) or intervention studies, 3) related to the treatment of RA with DMARDs or biologic agents, and 4) study duration of at least 6 months (see Supplementary Appendix 2, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). Studies were excluded if they met any of the following criteria: 1) the report was a meeting abstract, review article, or meta-analysis; 2) the study duration was less than 6 months; and 3) DMARDs or biologic agents were used for non-RA conditions (e.g., psoriatic arthritis, systematic lupus erythematosus) or non–FDA-approved use in health conditions other than RA (e.g., biologic agents in vasculitis) (see Supplementary Appendix 2, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). Two reviewers independently screened the titles and abstracts of the 2,497 potential articles from the PubMed and Cochrane Library searches by applying the above selection method. Any disagreements were resolved by consultation with the lead reviewer (JAS). The lead author also reviewed all titles and abstracts to identify any that might have been overlooked. We identified 149 original articles from the 3 searches for full-text retrieval. After excluding duplicates, 128 unique original articles were identified and the data were abstracted. This included 16 articles focused on DMARDs and 112 on biologic agents (98 on the 6 biologic agents assessed in the 2008 RA recommendations and 14 on certolizumab pegol, golimumab, and tocilizumab, 3 newer biologic agents that had been added since the 2008 recommendations) (see Supplementary Appendix 3, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). A list of all included articles is provided in Supplementary Appendix 4 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). Qualitative reviews of the literature were performed for these 3 topics (completed September 22, 2010). Similar to the strategy for the 2008 recommendations, literature searches were broadened to include case reports and case series of any size, review articles, and meta-analyses, plus inclusion of diseases other than RA. In addition, we included searches on the Centers for Disease Control and Prevention (CDC) web site (www.cdc.gov) for past and current recommendations regarding TB screening and vaccination in immunocompromised patients. The kappa coefficients (agreement beyond chance) for independent selection of articles for full-text review by the 2 reviewers met or exceeded 0.60 (good) for DMARDs, 0.65 (very good) for the 6 biologic agents included in the 2008 ACR recommendations, and 0.84 (excellent) for a combination of certolizumab pegol, golimumab, and tocilizumab (9). The full text of each article was reviewed; data abstraction and entry were performed by reviewers using a standardized Microsoft Access database that was developed and used for data abstraction for the 2008 ACR RA recommendations. Two reviewers were assigned to abstract data on DMARDs (SB, DEF), rituximab (HA, ERV), and the rest of the biologic agents (AB, AJ). To ensure that the error rates were low and abstractions were similar, 26 articles related to biologic agents were dually abstracted by 2 abstractors (AB, AJ). The data entry errors were less than 3%. Entered data were further checked against raw data on biologic agents from the Cochrane systematic reviews (5-8). Following this comprehensive literature review, we developed an evidence report using the data abstracted from the published studies. Clinical scenarios were drafted by the investigators and the CEP, based on the updated evidence report. We used the same key determinant clinical thresholds and treatment decision branch points that were developed for the 2008 ACR RA treatment recommendations (1). Clinical scenarios were constructed based on permutations in the particular therapeutic considerations that reflected: 1) disease duration (early versus established RA), 2) disease activity (low, moderate, or high) (Tables 2 and 3 and Supplementary Appendix 5, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658), 3) current medication regimen, and 4) presence of poor prognostic factors (yes or no, as defined in the 2008 ACR recommendations). An example of a clinical scenario is: “The patient has active established RA and has failed an adequate trial of an Anti-TNF [anti–tumor necrosis factor] biologic because of adverse events. Is it appropriate to switch to another Anti-TNF biologic after failing etanercept?” (see Supplementary Appendix 6, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). Scenarios included both new considerations and questions considered in the 2008 recommendations. Categorized as low, moderate, and high as per validated common scales (Table 3 and Supplementary Appendix 4, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658) or the treating clinician's formal assessment (26-32) For this 2012 update, we used a modified Delphi process and obtained consensus (defined as ≥70% agreement) from the CEP for inclusion of relevant clinical scenarios based on 1) review of each of the previous 2008 scenarios and 2) review of newly developed scenarios to address switching between therapies. We provided CEP members with manuscript abstracts and requested full-text articles to help inform decisions. The CEP members also recommended the following: 1) use of the FDA definitions of “serious” and “non-serious” adverse events (10), 2) exclusion of 3 DMARDs used very infrequently (i.e., cyclosporine, azathioprine, and gold; see above) or without additional relevant new data, and 3) exclusion of 1 biologic agent without additional relevant new evidence and with infrequent use (anakinra). The TFP is referred to as the “panel” in the Methods and the recommendations that follow. For the first round of ratings we contacted panel members by e-mail and provided them with the evidence report, clinical scenarios, and rating instructions. We asked them to use the evidence report and their clinical judgment to rate the “appropriateness” of the clinical scenarios under consideration. The panelists individually rated each scenario permutation using a 9-point Likert appropriateness scale. A median score of 1 to 3 indicated “not appropriate” and 7 to 9 indicated “appropriate” for taking action defined in the scenario (2-4). For all eventual recommendations, the RAND/UCLA appropriateness panel score required a median rating of 7 to 9. Those scenario permutations with median ratings in the 4 to 6 range and those with disagreement among the panelists (i.e., one-third or more TFP members rating the scenario in the 1 to 3 range and one-third or more rating it in the 7 to 9 range) were classified as “uncertain.” At a face-to-face meeting with both the TFP and the CEP members on November 15, 2010, the anonymous first round of ratings by the panel, including dispersion of the scores, ranges, and median scores, was provided to the task force panelists. The task force panelists agreed upon certain assumptions and qualifying statements on which they based their discussion and subsequent ratings of the scenarios (Table 2). A second round of ratings by panel members occurred after extensive in-person discussion of the prior ratings and review of the evidence supporting each scenario. After the TFP meeting was complete, recommendations were derived from directly transcribing the final clinical scenario ratings. Based on the ratings, scenario permutations were collapsed to yield the most parsimonious recommendations. For example, when ratings favored a drug indication for both moderate and high disease activity, one recommendation was given, specifying “moderate or high disease activity.” In most circumstances, the recommendations included only positive and not negative statements. For example, the recommendations focused on when to initiate specific therapies rather than when an alternate therapy should not be used. Most of the recommendations were formulated by drug category (DMARD, anti-TNF biologic, non-TNF biologic listed alphabetically within category), since in many instances, the ratings were similar for medications within a drug category. We specifically note instances where a particular medication was recommended but others in its group were not endorsed. Two additional community-based rheumatologists independently reviewed the manuscript and provided comments. CEP and TFP members reviewed and approved all final recommendations. For each final recommendation, the strength of evidence was assigned using the methods from the American College of Cardiology (11). Three levels of evidence were specified: 1) level of evidence A: data were derived from multiple RCTs; 2) level of evidence B: data were derived from a single randomized trial or nonrandomized studies; and 3) level of evidence C: data were derived from consensus opinion of experts, case studies, or standards of care. The evidence was rated by 4 panel experts (JO and JMK, AFK and LWM, where each rated half of the evidence), and discrepancies were resolved by consensus. Level C evidence often denoted a circumstance where medical literature addressed the general topic under discussion but it did not address the specific clinical situations or scenarios reviewed by the panel. Since many recommendations had multiple components (in most cases, multiple medication options), a range is sometimes provided for the level of evidence; for others, the level of evidence is provided following each recommendation. Following construction of the recommendations, the manuscript was reviewed through the regular journal review process and by more than 30 ACR members serving on the ACR Guidelines Subcommittee, ACR Quality of Care Committee, and ACR Board of Directors. Indications for and switching DMARDs and biologic agents: early RA (indications, Figure 1) followed by established RA (indications and switching, Figure 2), along with details of the level of evidence supporting these recommendations (see Supplementary Appendix 7, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658) Use of biologic agents in patients with hepatitis, malignancy, or CHF who qualify for RA management (Table 4) Screening for TB in patients starting or currently receiving biologic agents as part of their RA therapy (Figure 3) Vaccination in patients starting or currently receiving DMARDs or biologic agents as part of their RA therapy (Table 5) 2012 American College of Rheumatology recommendations update for the treatment of early rheumatoid arthritis (RA), defined as a disease duration <6 months. For the level of evidence supporting each recommendation, please see Supplementary Appendix 7 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). DMARD = disease-modifying antirheumatic drug (includes hydroxychloroquine [HCQ], leflunomide [LEF], methotrexate [MTX], minocycline, and sulfasalazine); anti-TNF = anti–tumor necrosis factor. * Definitions of disease activity are discussed in Tables 2 and 3 and Supplementary Appendix 4 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658) and were categorized as low, moderate, or high. † Patients were categorized based on the presence or absence of 1 or more of the following poor prognostic features: functional limitation (e.g., Health Assessment Questionnaire score or similar valid tools), extraarticular disease (e.g., presence of rheumatoid nodules, RA vasculitis, Felty's syndrome), positive rheumatoid factor or anti–cyclic citrullinated peptide antibodies (33-37), and bony erosions by radiograph (38). ‡ Combination DMARD therapy with 2 DMARDs, which is most commonly MTX based, with some exceptions (e.g., MTX + HCQ, MTX + LEF, MTX + sulfasalazine, and sulfasalazine + HCQ), and triple therapy (MTX + HCQ + sulfasalazine) as defined in Table 2. 2012 American College of Rheumatology (ACR) recommendations update for the treatment of established rheumatoid arthritis (RA), defined as a disease duration ≥6 months or meeting the 1987 ACR classification criteria. Depending on a patient's current medication regimen, the management algorithm may begin at an appropriate rectangle in the figure, rather than only at the top of the figure. Disease-modifying antirheumatic drugs (DMARDs) include hydroxychloroquine (HCQ), leflunomide (LEF), methotrexate (MTX), minocycline, and sulfasalazine (therapies are listed alphabetically; azathioprine and cyclosporine were considered but not included). DMARD monotherapy refers to treatment in most instances with HCQ, LEF, MTX, or sulfasalazine; in few instances, where appropriate, minocycline may also be used. Anti–tumor necrosis factor (anti-TNF) biologics include adalimumab, certolizumab pegol, etanercept, infliximab, and golimumab. Non-TNF biologics include abatacept, rituximab, or tocilizumab (therapies are listed alphabetically). For the level of evidence supporting each recommendation, please see Supplementary Appendix 7 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). * Definitions of disease activity are discussed in Tables 2 and 3 and Supplementary Appendix 4 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658) and were categorized as low, moderate, or high. † Features of poor prognosis included the presence of 1 or more of the following: functional limitation (e.g., Health Assessment Questionnaire score or similar valid tools), extraarticular disease (e.g., presence of rheumatoid nodules, RA vasculitis, Felty's syndrome), positive rheumatoid factor or anti–cyclic citrullinated peptide antibodies (33-37), and bony erosions by radiograph (38). ‡ Combination DMARD therapy with 2 DMARDs, which is most commonly MTX based, with few exceptions (e.g., MTX + HCQ, MTX + LEF, MTX + sulfasalazine, sulfasalazine + HCQ), and triple therapy (MTX + HCQ + sulfasalazine). § Reassess after 3 months and proceed with escalating therapy if moderate or high disease activity in all instances except after treatment with a non-TNF biologic (rectangle D), where reassessment is recommended at 6 months due to a longer anticipated time for peak effect. ¶ LEF can be added in patients with low disease activity after 3–6 months of minocycline, HCQ, MTX, or sulfasalazine. # If after 3 months of intensified DMARD combination therapy or after a second DMARD has failed, the option is to add or switch to an anti-TNF biologic. ** Serious adverse events were defined per the US Food and Drug Administration (FDA; see below); all other adverse events were considered nonserious adverse events. †† Reassessment after treatment with a non-TNF biologic is recommended at 6 months due to anticipation that a longer time to peak effect is needed for non-TNF compared to anti-TNF biologics. ‡‡ Any adverse event was defined as per the US FDA as any undesirable experience associated with the use of a medical product in a patient. The FDA definition of serious adverse event includes death, life-threatening event, initial or prolonged hospitalization, disability, congenital anomaly, or an adverse event requiring intervention to prevent permanent impairment or damage. 2012 American College of Rheumatology recommendations update for tuberculosis (TB) screening with biologic agent use. Depending on a patient's current therapy, the management may begin at an appropriate rectangle in the figure, rather than only at the top of the figure. The level of evidence supporting each recommendation for TB reactivation was “C,” except for initiation of biologic agents in patients being treated for latent TB infection, where the level of evidence was “B.” * Anergy panel testing is not recommended. † Interferon-γ–release assay (IGRA) is preferred if the patient has a history of BCG vaccination. ‡ Risk factors for TB exposure are defined based on a publication from the US Centers for Disease Control and Prevention as: close contacts of persons known or suspected to have active TB; foreign-born persons from areas that have a high incidence of active TB (e.g., Africa, Asia, Eastern Europe, Latin America, and Russia); persons who visit areas with a high prevalence of active TB, especially if visits are frequent or prolonged; residents and employees of congregate settings whose clients are at an increased risk for active TB (e.g., correctional facilities, long-term care facilities, and homeless shelters); health care workers who serve clients who are at an increased risk for active TB; populations defined locally as having an increased incidence of latent Mycobacterium tuberculosis infection or active TB, possibly including medically underserved, low-income populations, or persons who abuse drugs or alcohol; and infants, children, and adolescents exposed to adults who are at an increased risk for latent M tuberculosis infection or active TB (14). § If the patient is immunosuppressed and false-negative results are more likely, consider repeating screening test(s) with tuberculin skin test (TST) or IGRA. ¶ Chest radiograph may also be considered when clinically indicated in patients with risk factors, even with a negative repeat TST or IGRA. # Obtain respiratory (e.g., sputum, bronchoalveolar lavage fluid) or other samples as clinically appropriate for acid-fast bacilli (AFB) smear and culture and consider referral to a TB specialist for further evaluation and treatment. ** In a patient diagnosed with latent or active TB, consider referral to a specialist for the recommended treatment. †† Patients who test positive for TST or IGRA at baseline often remain positive for these tests even after successful treatment of TB. These patients need monitoring for clinical signs and symptoms of recurrent TB disease, since repeating tests will not allow help in diagnosis of recurrent TB. In the figures, decision points are shown as diamonds and actions to be taken by the health care provider are shown as rectangles. The recommendations in the text below and in Tables 4 and 5 represent the results of the 2012 update only, whereas Figures 1-3 also incorporate some of the 2008 ACR RA recommendations that did not change (1). Areas of uncertainty by the panel (that did not lead to recommendations) are noted in Supplementary Appendix 8 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). We first describe a recommendation targeting remission or low disease activity in RA (section 1A). This is followed by recommendations for DMARD or biologic agent use in early RA (section 1B). Next, we provide recommendations for initiating and switching between DMARDs and biologic agents in established RA (section 1C). The panel recommends targeting either low disease activity (Table 3) or remission (Table 2) in all patients with early RA (Figure 1; level of evidence C) and established RA (Figure 2; level of evidence C) receiving any DMARD or biologic agent. In patients with early RA, the panel recommends the use of DMARD monotherapy both for low disease activity and for moderate or high disease activity with the absence of poor prognostic features (Figure 1; level of evidence A–C) (details are shown in Supplementary Appendix 7, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). In patients with early RA, the panel recommends the use of DMARD combination therapy (including double and triple therapy) in patients with moderate or high disease activity plus poor prognostic features (Figure 1; level of evidence A–C). In patients with early RA, the panel also recommends the use of an anti-TNF biologic with or without methotrexate in patients who have high disease activity with poor prognostic features (Figure 1; level of evidence A and B). Infliximab is the only exception and the recommendation is to use it in combination with methotrexate, but not as monotherapy. The remainder of panel recommendations regarding indications for DMARDs and biologic agents are for patients with established RA. The 3 subsections below define recommendations for initiating and switching therapies in established RA (Figure 2). Where the prognosis is not mentioned, the recommendation to use/switch to a DMARD or a biologic agent applies to all patients, regardless of prognostic features. If after 3 months of DMARD monotherapy (in patients without poor prognostic features), a patient deteriorates from low to moderate/high disease activity, then methotrexate, hydroxychloroquine, or leflunomide should be added (rectangle A of Figure 2; level of evidence A and B). If after 3 months of methotrexate or methotrexate/DMARD combination, a patient still has moderate or high disease activity, then add another non-methotrexate DMARD or switch to a different non-methotrexate DMARD (rectangle B of Figure 2; level of evidence B and C). If a patient has moderate or high disease activity after 3 months of methotrexate monotherapy or DMARD combination therapy, as an alternative to the DMARD recommendation just noted above, the panel recommends adding or switching to an anti-TNF biologic, abatacept, or rituximab (rectangles C and D of Figure 2; level of evidence A–C). If after 3 months of intensified DMARD combination therapy or after a second DMARD, a patient still has moderate or high disease activity, add or switch to an anti-TNF biologic (rectangle C of Figure 2; level of evidence C). If a patient still has moderate or high disease activity after 3 months of anti-TNF biologic therapy and this is due to a lack or loss of benefit, switching to another anti-TNF biologic or a non-TNF biologic is recommended (rectangles F and G of Figure 2; level of evidence B and C). If a patient still has moderate or high disease activity after 6 months of a non-TNF biologic and the failure is due to a lack or loss of benefit, switch to another non-TNF biologic or an anti-TNF biologic (rectangles F and G of Figure 2; level of evidence B and C). An assessment period of 6 months was chosen rather than 3 months, du