Defining and characterizing sustained remission in patients with rheumatoid arthritis

SEAM-RA is a phase 3, multicenter, randomized, double-blind, controlled, withdrawal study. The study comprises a 30-day screening period, a 24-week open-label run-in period, a 48-week double-blind treatment period, and a 30-day safety follow-up period (Fig. 1). During the run-in period, patients receive open-label etanercept and methotrexate at the same dose they were receiving during screening. At the end of the run-in period, eligible patients are randomly assigned (2:2:1) to etanercept 50 mg weekly by subcutaneous (SC) injection plus oral placebo (planned n = 130), oral methotrexate at 10 to 25 mg weekly plus SC placebo (planned n = 130), or etanercept 50 mg weekly SC plus oral methotrexate 10 to 25 mg weekly (planned n = 65).

Data for this study are not publicly available because the study is ongoing and remains blinded.

Patients are evaluated for eligibility at screening (for participation in the run-in period) and at the end of the run-in period (for participation in the double-blind period). Key inclusion criteria at screening include the following: age ≥ 18 years, history of RA consistent with ACR and EULAR classification criteria, history of moderate to severe RA in the opinion of the investigator, very good disease control for ≥ 6 months in the opinion of the investigator, SDAI score ≤ 3.3 at screening (at the beginning of the run-in period), receiving etanercept at 50 mg weekly for RA for ≥ 6 months prior to the first run-in visit, and receiving a stable dose of methotrexate at 10 to 25 mg weekly for ≥ 6 months.

To be eligible for entry into the double-blind period, patients must have had an SDAI score ≤ 3.3 just prior to randomization at run-in visit 3 and cannot have the following: any clinically significant change in eligibility criteria during the run-in period, SDAI score > 3.3 and ≤ 11 on 2 or more visits during the run-in period, or SDAI score > 11 at any time during the run-in period. This definition would, for example, allow patients to have an SDAI score between 3.3 and 11 at run-in visit 2, with the recognition that some patients might have a short-term flare that was not persistent or have clinical parameters with an SDAI that was close to but did not meet the remission definition (i.e., a “near-miss”).

The primary endpoint of the SEAM-RA trial is SDAI remission (score ≤ 3.3) at week 48 without disease worsening before week 48. Disease worsening is defined as SDAI score > 3.3 and ≤ 11 on 2 consecutive visits at least 2 weeks apart; or SDAI > 3.3 and ≤ 11 on 3 or more separate visits; or SDAI > 11 at any time. Secondary efficacy endpoints include SDAI, DAS-28 with erythrocyte sedimentation rate (DAS-28-ESR), DAS-28-CRP, and Clinical Disease Activity Index (CDAI) scores and changes from baseline at all study time points; SDAI remission at all time points; Boolean remission at all time points; percentage of patients with worsening disease based on SDAI scores; time to disease worsening; and time to recapture remission after rescue therapy is given. Definitions of Boolean remission include 28 joints, 28 joints plus feet and ankles (as recommended by ACR/EULAR for assessments of remission [3]), and 66/68 joints. An additional endpoint is the proportion of patients who are eligible to continue in the study after run-in. Safety endpoints include incidence of all adverse events, serious adverse events, and laboratory parameters.

Although SEAM-RA is still enrolling patients, information from the first patients screened and enrolled in the study, and an examination of patient selection at different study sites are enlightening. Of 141 patients who completed screening and enrolled in the study, 64 maintained remission for the 24-week run-in period and have been randomized. Statistical analyses have so far not revealed any demographic or clinical characteristics that predict successful completion of the run-in period (i.e., maintenance of SDAI remission for 24 weeks). In general, patients who completed run-in and were randomized were not significantly different from the subset of patients who failed to complete the run-in period.

Over half of all screened subjects failed to meet eligibility criteria. The high number of screen failures due to SDAI score may be a reflection of physicians being unfamiliar with SDAI criteria and the depth of SDAI remission compared to DAS-28 remission. No single component of the SDAI score consistently contributed to elevations in SDAI score at the time of screen failure. The overall screen failure rates for study sites did not seem to improve over time, and screen failures due to increases in SDAI score were higher in sites active > 12 months compared with those that were active ≤ 12 months, suggesting that sites were exhausting the best candidates for the trial (i.e., those with the deepest and most sustained remission) within the first year that the site was open, and few additional patients newly attained remission over time given the stringent criteria required for SEAM-RA. Additionally, while SDAI score ≤ 3.3 is the remission definition proposed by ACR/EULAR, it may not be widely used by rheumatologists as a part of regular disease assessments for RA patients, and this lack of familiarity may also be a contributing factor to the screen failure rate based on SDAI. For future real-world trials that might be more permissive and mirror the types of RA patients for whom a rheumatologist would reasonably consider stopping therapy and where a validated remission definition is not required for potential regulatory claims, a more permissive and liberal inclusion may be used (e.g., SDAI score < 11 with SJC ≤ 1).

Although patients are allowed to have minor disease fluctuations that result in elevation of SDAI scores above 3.3 and up to 11 during the first two run-in visits, allowing some natural variability in disease activity, this laxity is not permitted at run-in visit 3. An SDAI score above 3.3 at the third run-in visit results in termination of the patient from the study. For the patients who failed because of an elevated SDAI score at run-in visit 3, the SDAI scores ranged between 3.4 and 9.5; notably, all values were within the range of LDA (SDAI score < 11). Indeed, no patients with “near-misses” for SDAI remission who initially met remission criteria at run-in visit 1 were ultimately excluded. However, reassuringly, even for the patients who failed SDAI remission criteria before run-in visit 3, almost half (48%) regained remission and continued in the trial, supporting the utility of following these patients over time since many were again in remission at their next visit.

The SEAM-RA study design was informed by prior studies of etanercept or methotrexate reduction or discontinuation after achieving good disease control on the combination. Studies including PRIZE [17], PRESERVE [18], and DOSERA [19] have characterized the role of etanercept in inducing and maintaining LDA in RA patients. Importantly, these studies demonstrated that a significantly greater proportion of patients on combination etanercept plus methotrexate therapy maintains low disease states compared to patients taking methotrexate alone. The DOSERA study demonstrated that most patients whose disease activity worsens after drug withdrawal are able to return to LDA [19]. Other studies provided insights regarding the role of etanercept as monotherapy in the maintenance of low disease states. In the CAMEO study, patients who achieved LDA (DAS-28-ESR < 3.2) at month 6 had similar disease activity at 24 months whether they discontinued methotrexate or remained on combination therapy. Combination treatment led to lower disease activity scores compared to etanercept monotherapy in patients with moderate to high disease activity at randomization (DAS-28-ESR > 3.2) [20]. In the COMET study [21], 50% of patients who discontinued methotrexate from combination therapy with methotrexate plus etanercept had achieved DAS-28 remission after 2 years, which was similar to the 57% rate of DAS-28 remission in patients who remained on combination therapy. Collectively, these studies have shown that etanercept may be required to maintain disease control and that etanercept may be sufficient as monotherapy in maintaining good disease control. Etanercept is particularly well suited to be used as monotherapy because of the lack of observed neutralizing antibodies [22]. Sustained disease remission with etanercept alone could potentially lead to a reduction in polypharmacy and undesired side effects.

Most of the studies that have examined DMARD discontinuation in patients in remission used DAS-28 to assess disease activity [9‐15]. While the DAS-28 is an easy and useful tool, the weight of the indicator of inflammation (ESR or CRP) is high in the DAS-28 calculation, which may exaggerate the rate of remission in patients receiving therapies that interfere with the acute phase response [16]. Additionally, DAS-28 is not as sensitive to detect changes in disease activity in the lower ranges, i.e., LDA and remission, whereas SDAI is sensitive to changes in the lower ranges. The primary endpoint in our study is the percentage of patients who achieve remission based on SDAI criteria; other methods of assessing disease severity (DAS-28-ESR, DAS-28-CRP, CDAI), as well as remission via Boolean criteria will also be evaluated. SDAI and CDAI are considered to have the most stringent criteria for remission [16]. SEAM-RA therefore uses the most stringent clinical remission criteria, which most closely represent a true remission of disease.

An additional feature of this study is the duration of remission: only patients in good disease control for ≥ 6 months before enrollment (based on clinician judgment) and SDAI remission throughout a 24-week run-in period are randomized, for a total of at least 1 year of stable therapy and good disease control before withdrawal of any therapy. This approach was used to reflect clinical practice. Previous trials assessing withdrawal of therapy have used short periods of treatment duration before withdrawal of either etanercept or methotrexate. Given that the duration of time in remission has been shown in some studies to predict successful withdrawal [23, 24], this clinical trial attempts to take this into account and consolidates or withdraws therapy only after a patient has had at least 1 year of stable treatment with combination methotrexate plus TNFi and very good disease control. Maintenance of remission for the 48-week treatment period without disease flare will be assessed for the primary endpoint (SDAI remission). In contrast to previous studies in which remission was assessed at a single time point at the end of the study, this study will assess remission every 12 weeks and at every unscheduled disease assessment visit. Patients must remain in remission throughout the course of the study and at the final week 48 time point to be considered as maintaining remission in this study.

An alternative to discontinuing a bDMARD is to discontinue the csDMARD from combination therapy. Methotrexate has been a mainstay of RA therapy for decades, with established efficacy and safety profiles. However, a recent study showed that most patients with RA had predominantly negative implicit attitudes toward methotrexate [25], and patients may therefore prefer discontinuing methotrexate over a TNFi medication. This study will characterize the adverse events associated with methotrexate use that could be reduced or eliminated after methotrexate withdrawal, such as fatigue, malaise, oral ulcers, alopecia, and liver and gastrointestinal toxicities [26, 27]. Discontinuation of methotrexate may be difficult when used in combination with some TNFi medications, such as anti-TNF monoclonal antibodies, as it has been shown to reduce the incidence of anti-drug antibodies, which can compromise bDMARD efficacy [28]. The development of neutralizing anti-drug antibodies is associated with loss of efficacy, and if clinicians increase the dose to overcome the loss of efficacy, patients may face an increased risk of adverse events [29] and increased cost. Also, the addition of methotrexate to the anti-TNF monoclonal antibody infliximab delays the decline in serum concentrations of infliximab [30].

Etanercept does not require coadministration with methotrexate to reduce the incidence of anti-etanercept antibodies [22] or delay declines in etanercept serum concentration. Anti-drug antibodies that develop with the use of etanercept have no known clinical significance [22, 31], and neutralizing anti-drug antibodies have not been reported with the use of etanercept [22]. Therefore, discontinuation of methotrexate from combination therapy of etanercept plus methotrexate may not reduce the efficacy or serum concentrations of etanercept or result in the development of clinically significant anti-etanercept antibodies.

Across RA clinical trials, there is variability in the definitions of disease flares, the number of flares allowed, and whether patients are discontinued from the study after a flare. In SEAM-RA, patients are allowed to have flares of a specified number and severity, but are not considered to be in remission if they exceed this narrow definition. SEAM-RA only allows complete withdrawal of either etanercept or methotrexate, without considering tapering doses. SEAM-RA includeds a run-in period to ensure that patients are in stable remission prior to withdrawal of any therapy. The appropriate duration of the run-in period and the most efficient number of run-in visits have not been established, but will be informed by the results of SEAM-RA.

The ability to achieve remission with new targeted therapies in many patients has been an exciting and recent development in the treatment of RA. Because rates of remission were low with early therapies, there has been little need to refine definitions and assessments of remission until recently. The optimal approach to consolidating therapy after achieving stable remission remains to be adequately characterized. The results of SEAM-RA will provide important information of practical consideration for determining therapy for maintenance of remission in clinical practice.