Patient-reported outcomes data in patients with psoriatic arthritis from a randomised trial of etanercept and methotrexate as monotherapy or in combination

Introduction

Psoriatic arthritis (PsA) is a progressive, chronic inflammatory disease with both joint and skin-related manifestations. Patients with PsA can experience a multitude of symptoms including peripheral and axial joint inflammation, enthesis inflammation (enthesitis; inflammation of the areas where tendons or ligaments attach to bones), finger or toe inflammation (dactylitis), nail disease and psoriasis.1 In patients with psoriasis, about 30% develop PsA with an annual incidence of 1%–3%.2–5 Effective disease management requires not only timely diagnosis and treatment but also recognition that PsA can adversely affect patient-reported outcomes (PROs). Previous research suggests that, in general, patients with PsA report worse PRO scores compared with either the general population or patients with psoriasis.6–9 PsA is associated with considerable disease burden including bodily pain and physical function below that of age-matched and gender-matched norms,10 fatigue, increased absenteeism from work, and decreased work productivity.11–14 When assessing the overall clinical value of PsA treatments, PRO measures should be included since health-related quality of life (HRQoL), physical function and global assessments are affected by treatment efficacy, safety, tolerability, dosing frequency and route of administration. Disease-modifying antirheumatic drugs (DMARDs) used to treat PsA such as methotrexate (MTX) and tumour necrosis factor inhibitors (TNFis) can improve PROs related to physical ability, joint pain and mental ability in patients with PsA.9 15 Although MTX, as a monotherapy or in a combination regimen, is widely used to treat PsA, the evidence supporting its use in this disease setting is limited. Further, few randomised controlled trials (RCTs) in PsA have examined PROs when directly comparing MTX with a TNFi or in combination with a TNFi. These include a study that compared MTX monotherapy versus placebo in patients with psoriasis16 and a study that compared MTX monotherapy versus golimumab and MTX combination therapy in patients with early PsA.17

The Study of Etanercept And Methotrexate in Combination or as Monotherapy in Subjects with Psoriatic Arthritis (SEAM-PsA)18 reported that both etanercept monotherapy and combination therapy were statistically significantly more effective than MTX monotherapy by the percentage of patients who were American College of Rheumatology 20% responders (ACR20, primary endpoint) and had Minimal Disease Activity (MDA) (key secondary endpoint) at week 24. Week-48 safety outcomes indicated that no deaths occurred in the study and that the incidences of adverse events and serious adverse events were similar across the three treatment groups.18 Published PRO results with the Health Assessment Questionnaire-Disability Index (HAQ-DI) and the Medical Outcomes Study Short Form-36 Questionnaire (SF-36) showed similar improvements in HAQ-DI and Mental Component Summary (MCS) scores from baseline to week 24 in all three treatment groups.18 However, the SF-36 Physical Component Summary (PCS) score showed greater mean changes from baseline at week 24 in both etanercept groups compared with MTX monotherapy.

Here, we report additional PRO measures from the SEAM-PsA RCT examining the effect of MTX monotherapy, etanercept monotherapy, and MTX plus etanercept combination therapy on PROs that evaluated HRQoL, physical function and global assessments of disease activity, including: HAQ-DI, Patient Global Assessment of disease activity (PtGA), Patient Global Assessment of Joint Pain (PtGAJP) and SF-36 summary and domain scores. By evaluating changes across the 3 groups at week 24 and comparing scores to age-matched and gender-matched normative values (based on the general population without chronic disease), we aimed to examine improvements in PROs.

Methods

Results

Change in PRO scores

An examination of changes in PRO scores from baseline to week 24 indicated that improvements generally occurred across all three treatment groups. For most SF-36 domains, changes from baseline at week 24 were numerically higher in the etanercept groups compared with MTX monotherapy (table 2).

Table 2

View inline•Open as popup

Change in patient-reported outcome scores from baseline at week 24

Treatment differences between MTX monotherapy versus the two etanercept groups by changes in PRO scores at week 24 after adjusting for baseline BMI status and prior non-bDMARD use are shown in table 2. Patients receiving etanercept monotherapy reported greater improvements from baseline in PtGA (p<0.001), PtGAJP (p=0.032), SF-36 PCS (p=0.033) and the domains of Bodily Pain (p=0.034) and General Health (p=0.01) compared with MTX monotherapy. Patients receiving combination therapy reported greater improvements from baseline in PtGA (p=0.012), PtGAJP (p=0.014), SF-36 PCS (p=0.015), Bodily Pain (p=0.007), and Vitality (p=0.05) domains compared with MTX monotherapy.

We generated a spydergram (figure 1) to illustrate how mean baseline and week 24 SF-36 domain scores across treatment groups compared against USA age-matched and gender-matched normative scores matched to the protocol population.27 Published normative values are: HAQ-DI: ≤0.25, SF-36 PCS and MCS: ≥50, and domain scores listed in the figure legend, with missing values not imputed. Figure 1 illustrates how the largest differences between MTX monotherapy and etanercept treatment groups were reported in Bodily Pain, General Health, and Vitality Domains.

Figure 1

Request permission

Spydergram of domain scores from baseline to week 24 versus age-matched and gender-matched normative scores. The spydergram was generated using domain raw scores (range: 0–100); age-gender norms in the USA were matched to the study population. The age-matched and gender-matched normative values for the SF-36 domains were: 82.01 for PF, 82.55 for RP, 72.86 for BP, 70.31 for GH, 58.89 for VT, 85.11 for SF, 88.12 for RE and 75.94 for MH. Missing values were not imputed. A higher SF-36 score indicates a more positive health state, so greater positive changes from baseline. P values in red indicate p≤0.05; p values in blue indicate p≤0.01. A, age; BL, baseline; BP, Bodily Pain; ETN, etanercept; G, gender; GH, General Health; MH, Mental Health; MTX, methotrexate; PF, Physical Function; RE, Role Emotional; RP, Role Physical; SF, Social Function; VT, Vitality; Wk, week.

Threshold analyses

Patients reporting improvements in PRO scores at week 24 ≥ MCID

We examined the percentage of patients in each treatment group who reported changes in PRO scores at week 24≥MCID in HAQ-DI, PtGA, PtGAJP and SF-36 PCS, MCS and domains. The etanercept monotherapy and combination therapy groups generally included a higher percentage of patients reporting clinically meaningful changes in PRO scores at week 24 (figure 2) compared with MTX monotherapy. The highest percentages of clinically meaningful improvements were reported in PtGA (66.7% with MTX monotherapy vs 78.3% with etanercept monotherapy, and 75.1% with combination therapy). Analyses comparing across treatment groups generally showed similar proportions of patients reporting changes≥MCID at week 24. Exceptions were higher percentages in both etanercept groups in PtGA (66.7% with MTX monotherapy vs 78.3% with etanercept monotherapy; p=0.005) and PtGAJP (65.1% with MTX monotherapy vs 73.9% with combination therapy; p=0.038; figure 2).

Figure 2

Request permission

Percentages of patients reporting PRO improvements≥MCID week 24. *P values are descriptive and are for comparison with methotrexate monotherapy. Only p values ≤0.05 are shown. MCID values were change from baseline ≤ –0.35 for HAQ-DI, ≤ –10.0 for PtGA, ≤ –10.0 for PtGAJP, ≥2.5 for SF-36 PCS, ≥2.5 for SF-36 MCS, and ≥5.0 for each SF-36 domain. BP, Bodily Pain; GH, General Health; HAQ-DI, Health Assessment Questionnaire-Disability Index; MCID, minimal clinically important difference; MCS, Mental Component Summary; MH, Mental Health; PCS, Physical Component Summary; PF, Physical Function; PRO, patient-reported outcome; PtGA, Patient Global Assessment; PtGAJP, Patient Global Assessment of Joint Pain; RE, Role Emotional; RP, Role Physical; SF, Social Function; SF-36, Medical Outcomes Study Short Form-36 Questionnaire; VT, vitality.

Patients reporting PRO scores ≥ age-matched and gender-matched normative values at week 24

Across PROs, the percentages of patients reporting scores at week 24≥normative values ranged from 20.8% (General Health domain) to 51.0% (SF-36 MCS) with MTX monotherapy, 30.9% (Role Physical domain) to 48.8% (SF-36 MCS) with etanercept monotherapy, and 30.6% (General Health domain) to 52.3% (SF-36 MCS) with combination therapy (figure 3). Higher percentages of patients in both etanercept groups reported scores at week 24≥normative values (p≤0.05) in HAQ-DI, SF-36 PCS, and Role Physical, Bodily Pain, and General Health domains. Numerically, more patients receiving combination therapy reported scores at week 24≥normative values in Vitality and Mental Health domains compared with MTX monotherapy (p≤0.05; figure 3).

Figure 3

Request permission

Percentages of patients reporting PRO scores ≥normative values at week 24. *P values are descriptive and are for comparison with methotrexate monotherapy at week 24. Only P values ≤0.05 are shown. Normative values were ≤0.25 for HAQ-DI; ≥50 for SF-36 PCS, and ≥50 for SF-36 MCS. Age/gender normalised values for the SF-36 domains were 82.01 for PF, 82.55 for RP, 72.86 for BP, 70.31 for GH, 58.89 for VT, 85.11 for SF, 88.12 for RE, and 75.94 for MH. BP, Bodily Pain, GH, General Health; HAQ-DI, Health Assessment Questionnaire-Disability Index; MCS, Mental Component Summary; MH, Mental Health; PCS, Physical Component Summary; PF, Physical Function; PRO, patient-reported outcome; RE, Role Emotional; RP, Role Physical; SF, Social Function; SF-36, Medical Outcomes Study Short Form-36 Questionnaire; VT, Vitality.

Discussion

Our analyses of the PRO data collected from the SEAM-PsA RCT indicate that patients receiving MTX monotherapy, etanercept monotherapy and MTX plus etanercept all reported improvements in PRO scores at week 24 compared with baseline. Both the etanercept monotherapy and MTX plus etanercept groups reported greater improvements from baseline compared with MTX in PtGA, PtGAJP, SF-36 PCS and Bodily Pain domain. Additionally, the etanercept monotherapy group reported greater improvement in General Health and the combination therapy group greater improvement in Vitality compared with MTX monotherapy. Improvements in PROs reported with etanercept monotherapy and MTX plus etanercept were numerically similar, although this was not statistically evaluated as the trial was not powered for statistical comparison between these two groups.

Reported results indicated that treating patients with PsA with etanercept either monotherapy or in combination with MTX results in greater, more clinically meaningful improvements in PROs compared with MTX monotherapy. Results from the primary and secondary endpoints of SEAM-PsA, for which the RCT analyses were powered, indicate that etanercept monotherapy and combination therapy were statistically significantly more effective than MTX monotherapy by ACR20 responses and MDA at week 24.18 PRO results from these analyses most closely aligned with ACR responses and MDA at week 24 are those examining changes from baseline at week 24 in PtGA, PtGAJP, SF-36 PCS, and General Health domain. These PROs may be most sensitive for capturing how patients with PsA experience their disease.

The internal validity of our study’s results is strengthened due to SEAM-PsA being an active-comparator RCT that directly compared PROs in patients treated with MTX, etanercept and combination therapy. In addition, the active comparator included patients receiving MTX at currently recommended doses (~20 mg/week).28

Results from our study are different from those observed in a phase 3 trial of golimumab in combination with MTX versus MTX monotherapy,17 in which golimumab did not show significant improvements in physical PCS at week 22 (SF-36 PCS score of 50.1 vs 50.7; p=0.543). In our study, etanercept monotherapy or in combination with MTX showed significant improvements in PCS scores at week 24 vs MTX monotherapy (least squares mean change from baseline in SF-36 PCS score of 7.38 and 7.62, respectively, vs 5.68; p≤0.05) for both comparisons. However, similar to observations in our study, the golimumab trial did not find any significant improvements in MCS scores.17

This study has a few limitations. There was no placebo group in the trial. This is important because including a placebo arm could help contextualise the results of this trial with prior RCTs in PsA. It is well known that active comparator trials, without placebo, result in higher responses as all patients have higher ‘expectation bias’, knowing that they are all receiving active treatment.29 30 However, there are ethical issues regarding prolonged placebo exposure in patients with active, especially early disease. Another limitation is that the SEAM-PsA trial was not powered to detect differences between etanercept groups and thus they could not be compared. The most important limitation was the generalisability of our results, as the treatment-naïve population enrolled in this RCT may not reflect the experiences of typical patients with severe or moderate PsA. However, our results are likely most relevant to patients with relatively early, active PsA disease.

Future research in PsA should include active comparators and PROs in their design. In addition, direct comparisons between different treatments could help determine which treatments are more effective in patients with PsA. Considering results from this RCT, PROs should be incorporated in discussions between patients and clinicians regarding their choice of treatment, to incorporate the patient’s perspective. Improvements in PROs reported with etanercept may inform decision-making between clinicians and patients with PsA.