Article Text
PDF
Abstract
Background The psoriatic arthritis (PsA) Observational Study of Persistence of Treatment (PRO-SPIRIT) assesses effectiveness and persistence of real-world PsA treatments. Ixekizumab (IXE) is an interleukin (IL)-17A inhibitor (i) (IL-17Ai), approved for the treatment of adult PsA.
Methods The aim of this predefined interim analysis was to report baseline characteristics along with early (3-month) descriptive and comparative real-world effectiveness in patients with PsA prescribed with advanced treatment including IL-17Ai; IXE or secukinumab (SEC), IL-12/23i, IL-23i, tumour necrosis factor (TNFi) or Janus kinase (JAKi).
Results 1192 patients across 6 countries were analysed. At baseline, patients receiving IXE had longer disease duration and higher previous biological/targeted-synthetic disease-modifying antirheumatic drugs experience than patients starting TNFi and SEC 150, and less concomitant conventional-synthetic DMARD use than TNFi and JAKi. Comparative analyses at 3 months showed that: (a) versus TNFi, IXE exhibited similar improvement in clinical Disease Activity in PsA (cDAPSA) but significantly greater improvement in body surface area affected by psoriasis (BSA) and global assessments (physician GA, patient GA (PatGA)); (b) versus IL-12/23i and IL-23i (pooled), IXE showed significantly greater improvement in cDAPSA and PatGA; (c) IXE was as fast as JAKi in improving joint disease activity. Ad hoc analysis indicated that more patients with active psoriasis (BSA ≥3%) achieved minimal disease activity with IXE than JAKi or IL-12/23i. The responses to SEC varied by dosage.
Conclusions This study confirms the rapid 3-month effectiveness of IXE on joint disease activity—as fast as TNFi and JAKi (cDAPSA), and exceeding IL-12/23i and IL-23i—along with clear benefits to skin.
- arthritis, psoriatic
- health-related quality of life
- interleukin-17
- patient reported outcome measures
- tumor necrosis factor inhibitors
Data availability statement
Data are available on reasonable request. Eli Lilly and Company provides access to all individual participant data collected during the trial, after anonymisation, with the exception of pharmacokinetic or genetic data. Data are available to request 6 months after the indication studied has been approved in the USA and EU and after primary publication acceptance, whichever is later. No expiration date of data requests is currently set once data are made available. Access is provided after a proposal has been approved by an independent review committee identified for this purpose and after receipt of a signed data sharing agreement. Data and documents, including the study protocol, statistical analysis plan, clinical study report, blank or annotated case report forms, will be provided in a secure data sharing environment. For details on submitting a request, see the instructions provided at www.vivli.org.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- arthritis, psoriatic
- health-related quality of life
- interleukin-17
- patient reported outcome measures
- tumor necrosis factor inhibitors
WHAT IS ALREADY KNOWN ON THIS TOPIC
Randomised controlled trials (RCTs) have demonstrated the efficacy and safety of ixekizumab (IXE) in the treatment of psoriatic arthritis (PsA); a head-to-head comparison using a composite joint and skin outcome found IXE to be more effective than adalimumab.
However, there is a lack of real-world evidence (RWE) comparing the effectiveness of IXE and other PsA treatments in the heterogeneous, real-world population, especially for treatments with more novel mechanisms of action.
WHAT THIS STUDY ADDS
This RWE study of nearly 1200 patients has confirmed the findings from RCTs and further shown that, at the early and pivotal 3-month timepoint, (1) IXE, as one of the inhibitors of interleukin (IL)-17A, is significantly better in nearly every metric versus a pooled sample of IL-12/23i and IL-23i and (2) IXE is largely similar to inhibitors of tumour necrosis factor (TNFi) and Janus kinase (JAKi) in effectively treating joint disease activity, but leads to better skin outcomes in patients with active psoriasis (baseline body surface area (BSA) >3%), with less concomitant use of conventional-synthetic disease-modifying antirheumatic drugs.
As expected, the study demonstrated better skin-related effectiveness following treatment with IXE compared with TNFi; this distinction appears to have translated into significantly better patient and physician global assessments.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study informs the field on real-world PsA treatment choices as related to baseline clinical manifestations and provides the first-ever comparative early (3-month) effectiveness data for approved PsA treatments in the heterogeneous, real-world population.
The study reflects current real-world practice, indicating that JAKi and TNFi are not being widely prescribed to patients with skin involvement, and that IL-17Ai and IXE specifically are more effective than TNFi on skin, more effective than JAKi at attaining minimal disease activity in patients with active psoriasis (BSA ≥3%) and more effective than IL-12/23i and IL-23i on all outcomes at 3 months.
Our findings highlight the need to consider both the baseline disease activity within each PsA domain as well as the potential for rapid response to treatment in treating patients with PsA.
Introduction
Psoriatic arthritis (PsA) is a heterogeneous inflammatory disease associated with diverse clinical manifestations that affect the musculoskeletal system, skin and nails among others; these manifestations can result in joint damage, reduced physical and psychosocial functioning and generally reduced health-related quality of life (HRQoL).1 EULAR recommends that disease activity be assessed regularly, suggesting between monthly and every 3 months as part of usual care, and PsA treatment adjusted appropriately as to target remission or, alternatively, low disease activity.2–4
The use of biological (b) disease-modifying antirheumatic drugs or targeted synthetic (ts) DMARDs (collectively, b/tsDMARDs) is recommended by EULAR for treating patients with active PsA for whom conventional therapies have failed to achieve the treatment target.2 3 There is currently no prioritisation between tumour necrosis factor (TNF) inhibitor (i) (TNFi), inhibitors of interleukin (IL)-17Ai, IL-12/23i and IL-23i, except that preference is given to IL-17Ai, IL-17A/Fi, IL-23i, IL-12/23i, in patients with clinically relevant skin involvement.2 Inhibitors of Janus kinase (JAKi) are recommended after an inadequate response to at least one bDMARD, unless a bDMARD is not an appropriate choice; for patients with mild disease or for whom neither a bDMARD nor JAKi would be appropriate, phosphodiesterase-4 (PDE4)i may be considered.2 Finally, for patients with clinically relevant (formerly ‘predominant’) axial disease and insufficient response to non-steroidal anti-inflammatory drugs (NSAIDs), therapy with an IL-17Ai, TNFi, IL-17A/Fi or JAKi is recommended.2 3 The Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) also support such individualised treatment.4
Ixekizumab (IXE) is an IgG4 monoclonal antibody that binds with high affinity (<3 pM) and specificity to IL-17A (both IL-17A and IL-17A/F), approved in many jurisdictions for the treatment of adult and paediatric (≥6 years old) psoriasis (PsO), as well as the treatment of adult PsA, radiographic axial spondyloarthritis (axSpA) and non-radiographic axSpA.5 6 Through multiple phase III programme, randomised controlled trials (RCTs) have provided extensive evidence for the efficacy and safety of IXE in treating all clinical manifestations of PsA assessed.7–14 Long-term extension studies SPIRIT-P1 and SPIRIT-P2 have demonstrated continued efficacy of IXE for up to 3 years.9 10
There have been few head-to-head (H2H) trials directly comparing PsA treatments. In the SPIRIT H2H trial, IXE was found superior to the TNFi adalimumab (ADA) in the improvement of joint and skin symptoms (ACR50) as well as total skin clearance (PASI100) at 24 weeks; significantly more patients had also achieved a status of remission (as defined by the Disease Activity in Psoriatic Arthritis (DAPSA) index) or minimal disease activity (MDA) with IXE than with ADA.15 Conversely, the remaining two H2H trials, EXCEED16 and SELECT-PsA1,17 failed to show superiority of the IL-17Ai treatments secukinumab (SEC, in EXCEED) and upadacitinib (in SELECT-PsA1) against ADA in the primary end point (ACR20 at week 52 and week 12, respectively).
The Psoriatic Arthritis Observational Study of Persistence of Treatment (PRO-SPIRIT) is a multinational, prospective study investigating the real-world use of b/tsDMARDs in PsA treatment, with a primary objective of describing persistence at 24 months among patients initiating a new b/tsDMARD. While RCTs are the gold standard for exploring treatment efficacy in the target population, real-world evidence (RWE) is vital for fully capturing the effectiveness of PsA therapies in the heterogeneous PsA population,18 and facilitates the comparison of a greater number of drugs with different mechanisms of action (MoAs).
Here, we report the baseline characteristics of the PRO-SPIRIT cohort and predefined 3-month interim analyses of early responses to treatment across clinical assessments and patient-reported outcomes (PROs) following the initiation of IXE and other b/tsDMARDs locally approved for PsA treatment. The 3-month timepoint is in line with EULAR and GRAPPA recommendations for evaluating the effectiveness of a newly initiated treatment, specifying a decision-making threshold of at least 50% reduction in disease activity within 3 months of treatment.2 4 Furthermore, 3-month early response rates have been shown to predict 6-month and 1-year outcomes in previous PsA studies.19
Methods
Study design and patient population
PRO-SPIRIT (EUPAS31174) is a 24-month, multicentre real-world study being conducted across France, Spain, Italy, Germany, the UK and Canada. We used the Strengthening the Reporting of Observational Studies in Epidemiology cohort reporting guidelines for observational studies.20 Enrolment commenced in December 2019 and ended in June 2022.21 Eligible patients were between 18 and 80 years of age, with a diagnosis of PsA for at least 6 months prior to baseline (as determined by the investigator), for whom the physician decided to initiate a (or switch to a new) b/tsDMARD according to country-specific approved indication. Patients had prior exposure to conventional PsA treatments (eg, corticosteroids, csDMARDs, etc) and were about to be newly prescribed a bDMARD (IL-17Ai (IXE or SEC), IL-12/23i (ustekinumab), IL-23i (guselkumab, risankizumab) or TNFi (ADA, etanercept, certolizumab, golimumab or infliximab, or biosimilars)) or tsDMARD (JAKi (tofacitinib or upadacitinib) or PDE4i (apremilast)). Treatment initiation, as well as any treatment changes during the 24-month observation period, were solely at the discretion of the participating physicians. Any treatment changes were documented, with follow-up continuing until the end of the 24-month observation period. All patients who started treatment 48 hours or more prior to baseline assessment were excluded from analysis (see online supplemental figure S1).
Supplemental material
After patient demographics, clinical assessments and PROs were collected at baseline (treatment initiation), clinical assessments and PROs continued being collected at routine clinical care visits postbaseline (or postfirst dose, if occurring after baseline data collection) at approximately 3 months (±4 weeks), 6 months (±8 weeks), 12 months (±8 weeks), 18 months (±8 weeks) and 24 months (±8 weeks) after treatment initiation, to accommodate the timing of the primary assessment for all bDMARDs and tsDMARDs included in the study, which can range from 12 to 16 weeks according to PsA indication. In Canada and the UK, the visit window for the 6-month visit was extended to −6 and +8 weeks due to the ±6 weeks visit window of the 3-month visit.
Some patients were treated with increased dosage, higher than recommended per label, which is not uncommon in the real-world setting. We have provided an analysis of the European Medicines Agency-approved on-label population, which was generally highly similar to the full sample, as shown in the online supplemental appendix (baseline:online supplemental table S2, 3-month effectiveness: online supplemental table S3).
Objectives and end points of the overall PRO-SPIRIT study
The overall aim of PRO-SPIRIT is to describe the 24-month persistence alongside effectiveness and healthcare resource use among patients with PsA who newly initiated b/tsDMARD treatment. Secondary objectives include both descriptive and comparative analyses of persistence, effectiveness (clinical assessments and PROs) and healthcare resource use among drug classes and between IXE and other treatment groups at 3, 6, 12, 18 and 24 months. Exploratory objectives include assessing the patient-related and disease-related factors associated with persistence, as well as the treatment patterns and outcomes in patients with primary and secondary response failures.
Objectives and end points of the current 3-month interim analysis
The objectives of this interim analysis are to describe the baseline characteristics of the study sample cohort and compare the early responses with treatment captured in clinical assessments and PROs at the 3-month timepoint. Clinical assessmen)ts and PROs include tender (TJC) and swollen joint counts (SJC), the clinical DAPSA (cDAPSA) (ie, without the inclusion of C reactive protein (CRP) measurements,22 due to a high proportion of missing CRP data, which included patients in remission (cDAPSA ≤4), with low disease activity (>4 to ≤13), moderate disease activity (>13 to ≤27) and high disease activity (HDA) (>27)), the extent of body surface area affected by PsO (BSA, %), the proportion of patients in a state of MDA and very low disease activity, the physician global assessment of disease activity (PhysGA) visual analogue scale (VAS), the patient global assessment of disease activity (PatGA) VAS, patient assessment of (joint) Pain VAS and Health Assessment Questionnaire Disability Index (HAQ-DI). Key measures are reported in the main body of the manuscript while the remainder are available in the online supplemental appendix.
Statistical analyses
A target sample size of 1600 patients was predefined in the study protocol. Missing data were primarily due to loss of visit (ie, data not collected). In order to retain statistical power, we performed a single imputation using the expectation maximisation algorithm for baseline variables, under the assumption that the data are missing at random.23 Missing data occurring during follow-up visits were imputed in a multistage approach using the Markov Chain Monte Carlo algorithm.24 25 For variables collected during the follow-up unrelated to treatment discontinuation, lost in follow-up, the intermittent missingness was imputed through a multiple imputation strategy.24 25 At the end of this stage, a monotone missing pattern was obtained for 100 datasets. In the second stage, a monotone single imputation within each of the 100 datasets was performed. Variables with >40% missing data were not imputed. All analyses were conducted using SAS V.9.4 (North Carolina) and the threshold for the type 1 error (alpha risk) was 0.05. Results were aggregated using SAS Software Proc MI to obtain final, pooled results, with no multiplicity adjustments performed.
Descriptive statistics were used to summarise baseline patient characteristics and the adjusted and unadjusted changes from baseline (CFB) in key effectiveness measures (mean TJC, SJC, cDAPSA and BSA, and proportion MDA) using mixed model for repeated measures (MMRM); these outputs are presented by treatment group and reported as either mean and SD/SE or median and range for continuous variables, and as proportions and percentages for categorical variables. The MMRM was predefined to be used for obtaining baseline-adjusted results included treatment and visit as factors, baseline score of the response outcome as covariate and treatment-by-visit and baseline-by-visit as interactions. The preferred covariance structure to model within-patient errors was unstructured, followed by Toeplitz, compound symmetry and variance components, in cases where the model did not converge. The Kenward-Roger approximation was used to estimate denominator df.
Inverse probability of treatment weighting (IPTW) was predefined to be used to balance measured patient characteristics between treatment groups for all comparative analyses (ie, adjusted comparisons). The variables used to compute these weights were derived from a propensity score based on logistic regression. The independent variables included in the model were country, age, body mass index (BMI; missing, <25, 25 to <30, ≥30), sex, education status (missing, no formal education, minimum mandatory required level of education, more than minimum required level of education but less than the university, university education), smoking status (never, current, former), time between initial PsA diagnosis and baseline, TJC, SJC, PhysGA, enthesitis, dactylitis, nail PsO, HAQ-DI, PatGA, joint pain, diagnosis of PsO at baseline, BSA ≥3% or <3%, axial manifestation, total number of b/tsDMARDs used prior to and including the time of enrolment (0, 1, 2, 3+), use of corticosteroid treatment and concomitant csDMARD use at baseline. Comparative analyses were performed using IPTW-weighted MMRM for continuous effectiveness outcomes (TJC, SJC, cDAPSA, BSA, PatGA, PhysGA, Pain VAS and HAQ-DI) and weighted logistic regression for categorical effectiveness outcomes (MDA). Continuous outcomes are thus presented using least squares mean (LSM) differences, while categorical outcomes are presented using ORs. The following comparisons were predefined and considered of interest: (1) IL-17Ai versus TNFi, (2) IL-17Ai versus IL-12/23i and IL-23i (pooled due to their low individual sample sizes), (3) IL-17Ai versus JAKi, (4) IXE versus SEC (150 mg and 300 mg pooled due to their low individual sample sizes, ‘SEC TOTAL’), (5) IXE versus TNFi and (6) IXE versus JAKi, to which (7) IXE versus IL-12/23i and IL-23i (pooled) was added for completeness. Analyses were only conducted on groups in which at least approximatively 10% of the overall study population was present. PDE4i comparative analyses were not performed due to low sample size.
In addition to these predefined analyses, on the basis of meaningful differences in group-level characteristics observed at baseline, we conducted an exploratory ad hoc analysis for patients with higher disease burden to fully understand the nuances within our initial dataset. Specifically, we investigated the proportion of patients with MDA at the 3-month timepoint across treatment groups, separately between those who exhibited either ≥3% or <3% BSA at baseline. Given the small sample sizes and the exploratory nature of this analysis, we only report the estimations and associated 95% CIs, but not statistical significance.
Results
Descriptive baseline characteristics (imputed)
This study initially enrolled 1358 patients; after excluding those who did not meet criteria or exhibited protocol deviations, the final sample comprised 1192 patients from 175 study sites (online supplemental figure S1). Patients were enrolled from Canada (n=133), France (n=164), Germany (n=278), Italy (n=278), Spain (n=188) and the UK (n=151). Between December 2019 and June 2022, they initiated treatment with either a bDMARD (IL-17Ai (n=507, 42.5%), IL-12/23i (n=35, 2.9%), IL-23i (n=56, 4.7%) or TNFi (n=436, 36.6%)), or a tsDMARD (JAKi (n=124, 10.4%) or PDE4i (n=32, 2.7%), or other (no treatment assigned at baseline; n=2)). IXE was initiated in 341 patients (28.6%), SEC 150 in 87 patients (7.3%) and SEC 300 in 79 patients (6.6%) (pooled in ‘SEC TOTAL’, n=166, 13.9%).
Key baseline demographics and clinical characteristics of the full sample population (n=1192) are reported in table 1, while the remainder, plus the full baseline characteristics of those receiving on-label treatment (n=1159, after adjudication), are available in online supplemental tables S1 and S2, respectively. Across treatment groups, patients were of similar mean age (range from 50.5 to 54.8 years) and were more often female (54.3%–66.7%). Groups exhibited similar degrees of joint involvement, although the patients prescribed an IL-12/23i or IL-23i had less axial involvement. IL-17Ai and IL-23i medications were more commonly prescribed to patients with relevant skin involvement. We observed these and other meaningful between-group differences in both baseline characteristics as well as 3-month effectiveness, discussed below. We report the adjusted descriptive and comparative effectiveness results for the overall population in the main text, while the on-label population and remaining unadjusted analyses can be found in the online supplemental appendix.
Baseline demographics and clinical characteristics of patients with psoriatic arthritis initiating real-world treatment with ixekizumab or other DMARDs (b/tsDMARDs) (n=1192), with meaningful between-group differences highlighted for side-by-side comparison
Within the IL-17Ai class (IXE and SEC)
Baseline characteristics of the IL-17Ai treatment groups were similar between IXE and SEC 300 but differed from the SEC 150 group. The latter had a smaller proportion of b/tsDMARD-experienced patients (54.0%) compared with both IXE (70.4%) and SEC 300 (79.7%). The SEC 150 group had shorter disease duration (8.4 years) compared with IXE (9.3 years). The SEC 150 group was also more moderate at baseline overall, with a smaller proportion of patients presenting with baseline BSA ≥3% (SEC 150, 37.9%; IXE, 41.1%; SEC 300, 43.0%), as well as lower mean BSA scores (SEC 150, 4.0±9.5; IXE, 5.5±10.8; SEC 300, 7.8±16.0), and lower baseline joint disease activity (mean cDAPSA: SEC 150, 25.6±1.5; IXE, 27.8±0.9; SEC 300, 30.5±2.1) (table 1).
At 3 months, the MMRM-adjusted descriptive analysis revealed that IXE and SEC TOTAL (pooled dosage) exhibited similar real-world effectiveness across joint disease activity (TJC, SJC, cDAPSA), skin and PROs (figures 1, 2, table 2). Notably, the responses to SEC varied by dosage. Despite being a more b/tsDMARD-naïve and moderate population, SEC 150 performed similarly to IXE in terms of improving joint disease activity (mean adjusted CFB in cDAPSA: IXE, −10.7±0.7; SEC 150, –11.0±1.3), as well as the skin in terms of adjusted CFB in BSA (IXE, −2.9±0.3; SEC 150, –2.7±0.6) and the proportion who shifted from a BSA ≥3% to <3% (IXE: 56.4%; SEC 150, 48.5%) (table 2). By contrast, while the baseline profile was similar to IXE, the SEC 300 group exhibited lower effectiveness in terms of adjusted CFB in cDAPSA (IXE, −10.7±0.7; SEC 300, –8.0±1.4) (table 2) but similar improvement in terms of adjusted CFB in BSA (IXE, −2.9±0.3; SEC 300, –2.7±0.7), although with a smaller proportion of patients shifting from ≥3% to <3% BSA with SEC 300 (41.2%) than IXE (56.4%) (table 2). Lastly, none of the IPTW-adjusted comparative analyses revealed significant differences between IXE and SEC TOTAL in effectiveness at the early 3-month timepoint (figure 3A–I).
Baseline-adjusted change from baseline in mean cDAPSA scores of the full study sample (n=1192) (descriptive, adjusted for baseline differences in mean cDAPSA). All data presented as mean (95% CI). Adjusted for between-group comparison through the use of MMRM to account for baseline differences in mean cDAPSA. SEC 150 (n=87; mean (95% CI)=−10.99% (−13.58% to −8.40%)) and SEC 300 (n=79; −8.01% (−10.75% to −5.27%)) pooled together for data visualisation purposes. cDAPSA, clinical disease activity in psoriatic arthritis; i, inhibitor; IL, interleukin; IXE, ixekizumab; JAKi, Janus kinase inhibitor; MMRM, mixed model for repeated measures; PsA, psoriatic arthritis; SEC, secukinumab; TNF, tumour necrosis factor.
Baseline-adjusted change from baseline in mean affected BSA of the full study sample (n=1192) (descriptive, adjusted for baseline differences in mean BSA). All data presented as mean (95% CI). Adjusted for between-group comparison through the use of MMRM to account for baseline differences in mean BSA. SEC 150 (n=87; mean (95% CI)=−2.68% (−3.87% to −1.48%)) and SEC 300 (n=79; −2.69% (−3.97% to −1.42%)) pooled together for data visualisation purposes. BSA, body surface area; i, inhibitor; IL, interleukin; IXE, ixekizumab; JAKi, Janus kinase inhibitor; MMRM, mixed model for repeated measures; PsA, psoriatic arthritis; SEC, secukinumab; TNF, tumour necrosis factor.
Forest plots depicting comparative assessment (least squares mean difference (TJC, SJC, cDAPSA, BSA, PhysGA, PatGA, HAQ-DI, Pain VAS) or OR (MDA)) of early response to treatment (3 months after initiation) in the full sample of patients with real-world psoriatic arthritis (n=1192). All data presented as LSM difference (TJC, SJC, cDAPSA, BSA, PhysGA, PatGA, HAQ-DI and Pain VAS) or OR (MDA) and 95% CI. Adjusted for between-group comparison through the use of IPTW to account for all baseline differences. Secukinumab 150 mg: n=87; secukinumab 300 mg: n=79. BSA, body surface area; cDAPSA, clinical disease activity in psoriatic arthritis; HAQ-DI, Health Assessment Questionnaire-Disability Index; i, inhibitor; IPTW, inverse probability of treatment weighting; LSM, least squares mean; MDA, minimal disease activity; PatGA, patient global assessment; PhysGA, physician global assessment; SJC, swollen joint count; TJC, tender joint count; TNF, tumour necrosis factor; VAS, visual analogue scale.
Unadjusted and adjusted (descriptive, imputed) change from baseline to month 3 in clinical measures across treatment groups (n=1192)
IXE (n=341) vs TNFi (n=436)
At baseline, IXE-treated patients exhibited longer disease duration (IXE, 9.3±8.5 years; TNFi, 6.6±7.3 years) and larger proportions presented with b/tsDMARD experience (IXE, 70.4%; TNFi, 31.2%) and active PsO (IXE, 41.1%; TNFi, 34.2%), along with higher mean BSA (IXE, 5.5±10.8; TNFi: 4.4±9.5) but similar joint disease activity (mean cDAPSA: IXE, 27.8±0.9; TNFi, 26.7±0.8) (table 1).
At 3 months, the MMRM-adjusted descriptive analysis showed similar effectiveness on joint disease activity (TJC, SJC and mean cDAPSA) (figure 1), with similar proportions achieving cDAPSA remission (IXE, 0.9%; TNFi, 1.4%). As expected, in the skin domain, IXE exhibited markedly higher effects than TNFi in terms of adjusted CFB in BSA (IXE, −2.91; TNFi, −2.17) (figure 2) as well as the proportion of patients whose BSA shifted from ≥3% to <3% (IXE, 56.4%; TNF, 41.6%) (table 2). Lastly, adjusted for all baseline differences using IPTW, comparative analysis of IXE versus TNFi supported the observations from the descriptive analysis, that both drug types were similarly effective on joint disease activity at the 3-month timepoint (TJC, SJC and mean cDAPSA score; figure 3A–C) and that patients who had initiated treatment with IXE (p<0.05) exhibited significantly greater improvement in mean BSA than those who had initiated a TNFi (figure 3D). Finally, improvement in both PhysGA (p<0.05) and PatGA (p<0.05) was significantly greater for those who had initiated IXE rather than a TNFi (figure 3F,G), while the proportion of patients with MDA, HAQ-DI and Pain VAS improvements were similar between IXE and TNFi (figure 3E,F and I).
IXE (n=341) vs IL-12/23i (n=35) and IL-23i (n=56)
At baseline, patients prescribed IXE were similar to the IL-12/23i and IL-23i groups with respect to disease duration, previous b/tsDMARD experience and skin involvement (regarding IL-12/23i); however, baseline joint disease activity was higher in the IXE group (mean cDAPSA: IXE, 27.8±0.9; IL-12/23i, 24.7±2.7; IL-23i, 26.7±2.0). Interestingly, compared with the IXE group, more patients prescribed an IL-12/23i or IL-23i already met cDAPSA criteria for remission at baseline (IXE, 0.9%; IL-12/23i, 5.7%; IL-23i, 3.6%). Regarding affected skin, patients prescribed an IL-23i had higher baseline skin involvement (mean BSA: 7.9±12.7; active PsO: 55.4%) than either IXE (5.5±10.8; 41.1%) or IL-12/23i (5.9±8.7; 48.6%) (table 1). The proportion of patients with axial involvement was higher in the IXE group (35.8%) than the IL-12/23i (28.6%) and IL-23i (19.6%) groups.
MMRM-adjusted response rates were higher with IXE than with IL-12/23i and IL-23i (pooled) in terms of joint disease activity (TJC, SJC, mean cDAPSA), skin (BSA), MDA and PROs (PatGA, PhysGA, HAQ-DI and Pain) (figures 1, 2, table 2). At the 3-month timepoint, a notably smaller proportion of IXE-treated patients (12.6%) were classified as HDA than those treated with an IL-12/23i (36.4%) or, to a lesser extent, an IL-23i (18.5%). The IPTW-adjusted comparative analysis of joint disease activity (TJC, SJC and cDAPSA) showed better results for IXE against the pooled group of IL-12/23i and IL-23i (figure 3A–C). Improvements in TJC, SJC and cDAPSA were significantly greater with IXE (LSM difference −3.6, p<0.01; LSM difference −2.4, p<0.05 and LSM difference −8.4, p<0.001) than with IL-12/23i and IL-23i (pooled). Similarly, those who had initiated IXE (p<0.01) exhibited significantly greater improvements in BSA than those who had initiated either an IL-12/23i or IL-23i (figure 3D). MDA at 3 months did not significantly differ (figure 3E). PROs reflected the pattern in skin and joints, with significantly more improvement in PhysGA (figure 3F), PatGA (figure 3G), HAQ-DI (figure 3H) and Pain VAS (figure 3I) after initiating IXE (PhysGA, p<0.05; PatGA, p<0.01; HAQ-DI, p<0.001; Pain VAS, p<0.001) than an IL-12/23i or IL-23i.
IXE (n=341) vs JAKi (n=124)
At baseline, patients initiating IXE were similar to those initiating a JAKi in terms of disease duration (IXE, 9.3±8.5; JAKi, 9.0±7.6) and prior b/tsDMARD experience (70.4%; 72.6%). Baseline joint disease activity was similar across groups (mean cDAPSA: IXE, 27.8±0.9; JAKi, 29.6±1.6) but compared with IXE (41.1%), fewer JAKi patients presented with active PsO (27.4%) while also exhibiting lower mean BSA (IXE, 5.5±10.8; JAKi, 3.2±8.2) (table 1).
At 3 months, the MMRM-adjusted descriptive analysis found that IXE and JAKi exhibited similar effectiveness on joint disease activity and PROs. Although the adjusted CFB in BSA was similar between IXE (−2.9) and JAKi (−3.0), the proportion of patients who shifted from a BSA ≥3% to <3% was higher with IXE (56.4%) than JAKi (50.0%) (table 2). Lastly, IPTW-adjusted comparative analysis of IXE versus JAKi revealed that these drug types were similarly effective at producing early improvements in joint disease activity (TJC, SJC and cDAPSA mean score, figure 3A–C) and PROs (figure 3F–I). However, somewhat unexpectedly, IXE was not significantly better than JAKi at improving skin involvement (figure 3D) or attainment of MDA (figure 3E), even though JAKi are not indicated to treat PsO. To better understand this, an ad hoc analysis using stratification based on baseline BSA is presented below.
IXE (n=341) and PDE4i (n=32)
Patients initiating IXE had more b/tsDMARD experience (IXE, 70.4%; PDE4i, 21.9%), as well as higher joint disease activity (mean cDAPSA: 27.8±0.9; 21.7±2.4) and skin involvement (mean BSA: 5.5±10.8; 4.4±5.4) at baseline (table 1). Comparative effectiveness analyses at the 3-month timepoint were not performed due to low PDE4i sample size.
Ad hoc analysis: proportion of patients with MDA at month 3, stratified by baseline BSA
On the basis of lower baseline BSA in the JAKi group and the variability in baseline BSA across all groups, and in order to better understand the contribution of skin to the early achievement of MDA, we further stratified the patient population and re-analysed the proportions of patients with MDA at the 3-month timepoint between those with BSA ≥3% or <3% at baseline. This ad hoc analysis identified important numerical differences between groups: among patients with baseline BSA ≥3%, a moderately lower proportion attained MDA at the 3-month timepoint with JAKi and IL-12/23i than the other treatment types included. On the other hand, among the patients with baseline BSA <3%, a lower proportion reached MDA with SEC 300, IL-12/23i and IL-23i than other treatment types (table 2 and online supplemental figure S7A,B).
Discussion
In this 3-month interim analysis, we report baseline characteristics and early response data demonstrating the rapid, real-world effectiveness of available PsA treatments. At baseline, the overall patient profiles of the TNFi and SEC 150 groups were strikingly different from other treatment groups, exhibiting shorter disease duration, lower disease activity across domains and less previous b/tsDMARD exposure; this introduced additional complexity, particularly when pooling SEC 150–300. Our data also indicate that, among the patients initiating JAKi in a real-world setting, nearly one-third are receiving JAKi as a first-line treatment (ie, after NSAIDs and/or csDMARDs, naïve to other b/tsDMARDs), which is earlier in the course of treatment than the 20192 and 20223 EULAR recommendations. Interestingly, while clinical trials have not found significant differences between monotherapy and combination therapy,26 we further observed that nearly half of the patients in the JAKi group were concomitantly treated with csDMARDs, akin to TNFi clinical practice or possibly attributed to varying local regulatory requirements.
No striking differences in disease activity of the joints were observed between treatment groups at baseline, except for a higher percentage of baseline cDAPSA remission rates in IL-12/23i, IL-23i and PDE4i groups compared with other treatments. This suggests that treatment decisions might be influenced by other factors such as skin. We further observed that IXE, SEC 300 and IL-23i medications were primarily prescribed to patients with relevant skin involvement. In contrast, TNFi and JAKi were more commonly prescribed to patients with less skin involvement. Finally, patients with axial manifestations more often received treatment with an IL-17Ai, TNFi or JAKi, than either an IL-12/23i or IL-23i. Overall, our findings indicate that the predominant involvement of skin or axial at baseline may influence clinical treatment decisions towards different classes/drugs, largely in alignment with official recommendations.
In comparative analyses, we did not observe meaningful differences in the 3-month real-world effectiveness of IXE and SEC TOTAL (pooled dosage) with respect to either skin or joint disease activity. However, in the descriptive analysis, we observed different patient profiles within the IL-17Ai class. Patients prescribed SEC 300 had the highest rates of prior treatment failure, along with the highest joint disease activity and skin involvement. Consequently, the SEC 300 group exhibited less responsiveness to treatment compared with IXE and SEC 150, as evidenced by smaller decreases in TJC and baseline-adjusted cDAPSA mean scores, as well as a lower proportion of patients attaining MDA. The differences in baseline characteristics are consistent with the study from Joven et al,27 which subsequently found numerically lower persistence in the SEC 300 cohort that the investigators ascribed to a lack of effectiveness of SEC 300 compared with SEC 150 and IXE. These relevant baseline population differences make it challenging to interpret data from SEC-treated patients and the IL-17Ai class as a whole, and therefore further analyses are warranted, particularly by dose and line of treatment.
Regarding comparative effectiveness after 3 months of treatment, this study showed that after using IPTW to account for baseline differences, IXE was similarly effective as TNFi in improving joint disease activity, and significantly more effective than TNFi on skin as well as PatGA and PhysGA; this set of outcomes is noteworthy as the IXE group, in particular, had a higher rate of prior biological failures as well as longer mean disease duration than TNFi. Given that effective control of both skin and joints is essential for improving HRQoL in patients with PsA,28 the better performance in the skin domain may have led to the significantly greater improvements in PatGA and PhysGA observed following IXE treatment, relative to TNFi.
IL-17Ai and IL-23i are two major classes of PsA treatment, yet a H2H comparison has never been conducted. This study provides another meaningful finding from the adjusted comparative effectiveness analysis, in that IXE—across nearly all clinical parameters, except MDA—exhibited greater real-world effectiveness than IL-12/23i and IL-23i (pooled). While there are no H2H trials comparing IL-17Ai versus IL-23i PsA treatment, our results support the data from H2H PsO trials comparing the attainment of PASI100 by week 12 of treatment with IL-17Ai versus IL-23i, IXORA-R29 (41% (IXE) vs 25% (guselkumab)) and IXORA-S30 (36% (IXE) vs 15% (ustekinumab)), complementing the paucity of data in PsA. Moreover, improvements with IXE specifically exceeded those with IL-23i at the 3-month timepoint in both joint disease activity (CFB in TJC, SJC, cDAPSA) and skin involvement (adjusted CFB in BSA and BSA shift >3 to <3%; table 2). The difference between IXE and IL-23i could be due to the slower onset of action of IL-23i relative to IL-17Ai treatments, which has been demonstrated in the skin31 and indicated here in the joints. Analysis of later timepoints in PRO-SPIRIT will enable us to understand the effectiveness of this treatment class in the longer term.
Finally, it is worth noting that the comparative analysis found JAKi treatment to be similarly effective as IXE in terms of disease activity at the 3-month timepoint. However, the JAKi group had more naïve patients at baseline than recommended by EULAR and GRAPPA guidelines, along with less initial skin involvement and higher concomitant use of csDMARDs for combination therapy. Furthermore, the ad hoc analysis indicated that in patients with baseline BSA ≥3%, the effectiveness (as measured by MDA) was higher numerically with IXE than JAKi.
As an observational study, PRO-SPIRIT has several limitations. These include patient selection and allocation bias, as the choice of initiating or switching treatment with IXE, a bDMARD, or a tsDMARD was solely based on the judgement of the treating physicians, who were reimbursed for their participation by the study sponsor (Eli Lilly and Company) regardless of the medication selected. To account for the resulting (and anticipated) differences in baseline characteristics, we used a statistical weighting approach, but we cannot ensure that all confounding variables were identified and accounted for. Moreover, despite applying statistical weighting for the number of previous b/ts DMARDs, future analyses could benefit from adjusting for whether patients are b/tsDMARDs-naïve or experienced. Additionally, we are not able to comment on differences in safety, because adverse event data were only collected following standard pharmacovigilance programmes at the country level and only as reasons for discontinuation for the purpose of this study. A further limitation is that, while certain groups such as IXE and TNFi had large sample sizes, other groups were substantially smaller; we pooled groups where possible to ameliorate this and enable evaluation, but this pooling is itself an additional limitation. Finally, in order to avoid an increased type 1 error rate, we were limited to a planned set of comparisons, which prohibited comparing all groups across every metric. These between-group differences at baseline, likely stemming from selection/allocation biases in assigning certain patient types to specific medication classes, required statistical adjustments before comparing subsequent treatment responses. These adjustments could have also influenced data interpretation.
In summation, although several bDMARDs and tsDMARDs with different MoAs have been approved for the treatment of PsA, there remains a paucity of RWE exploring the relative effectiveness of these therapies in a heterogeneous population. For instance, TNFi, SEC and ustekinumab were only recently shown to be effective in a real-world setting.32 Furthermore, to our knowledge, only two studies have ever reported comparative analysis data—one comparing SEC and ustekinumab33 and the other examining TNFi versus ustekinumab.34 One additional registry-based study compared multiple PsA drugs but only reported persistence,35 while another observational study reported RWE of IXE and SEC persistence.27 In conclusion, PRO-SPIRIT is one of the first large-scale, multinational studies generating RWE on the effective use of IXE and other b/tsDMARDs with a wide variety of MoAs in clinical practice. This RWE study confirms the efficacy of IXE as reported in RCTs, which is consistent regardless of patient baseline characteristics or clinical domain involvement, and additionally provides valuable comparative analyses among different treatment groups and treatment populations in a real-world setting. This RWE brings rheumatologists one step closer to holistic care of the patient and individualising treatment options for patients with PsA that focus on both early responses in the joints and skin as well as baseline disease severity.
Medical writing, statistical analysis, editorial and other assistance
Cedric Laedermann of Eli Lilly and Company contributed extensively to the design, analysis and writing of the manuscript. Statistical analysis was supported by C P Shyam and Akash Bagchi of Eli Lilly and Company. Medical writing assistance in the preparation of this article was provided by Nicola Roe, PhD, and Lynn Buckley, PhD, of Eli Lilly and Company. Quality review and support was provided by Certara Synchrogenix. The support for this assistance was funded by Eli Lilly and Company.
Data availability statement
Data are available on reasonable request. Eli Lilly and Company provides access to all individual participant data collected during the trial, after anonymisation, with the exception of pharmacokinetic or genetic data. Data are available to request 6 months after the indication studied has been approved in the USA and EU and after primary publication acceptance, whichever is later. No expiration date of data requests is currently set once data are made available. Access is provided after a proposal has been approved by an independent review committee identified for this purpose and after receipt of a signed data sharing agreement. Data and documents, including the study protocol, statistical analysis plan, clinical study report, blank or annotated case report forms, will be provided in a secure data sharing environment. For details on submitting a request, see the instructions provided at www.vivli.org.
Ethics statements
Patient consent for publication
Ethics approval
This is a large-scale observational (non-interventional) study taking place across 175 sites, each of which have their own form of an Institutional Board who reviewed and approved the study protocol. If it is absolutely necessary, we will gather these documents and provide them; however, we do not have them readily available. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
Eli Lilly and Company would like to thank the participants and their caregivers, without whom this work would not be possible. A full list of trial investigators can be located in the online supplemental appendix.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors This trial was designed by Eli Lilly and Company. All authors provided critical feedback and contributed to the final manuscript, which was led by KJN (guarantor) and supervised by IdlT. Statistical analysis was led and supervised by MN and drafting was led by DK.
Funding This study is sponsored by Eli Lilly and Company, Indianapolis, Indiana, USA.
Competing interests LEK has received honoraria or fees for serving as a speaker or consultant from AbbVie, Amgen, Biogen, Bristol Myers Squibb, Eli Lilly and Company, Gilead, GSK, Janssen, Merck, Novartis, Pfizer and UCB. He has received investigator-initiated study grants from AbbVie, Biogen, Eli Lilly and Company, Janssen, Novartis, Pfizer and UCB. KJN and MN are employees and minor shareholders of Eli Lilly and Company. JM has received grants, speaker honoraria or travel support, or participated on an advisory board for AbbVie, Biogen, Boehringer Ingelheim, Bristol Myers Squibb, Eli Lilly and Company, Fresenius Kabi, Galapagos, Medac, Mylan, Novartis, Pfizer, Roche Chugaï, Sanofi and Viatris. EL has received speaker honoraria from AbbVie, Eli Lilly and Company, Janssen, Novartis and UCB. WT has received grants, consulting fees, speaker honoraria and/or travel support from AbbVie, Eli Lilly and Company, GSK, Janssen, Novartis, Ono-Pharma, Pfizer and UCB. RA has received consulting fees, speaker honoraria and/or travel support from AbbVie, Amgen, Bristol Myers Squibb, Celltrion, Eli Lilly and Company, Galapagos, Gilead, Janssen, Mylan, Novartis, Pfizer, Roche Chugaï, Viatris and UCB. VC has received grants, royalties or consulting fees, or had a leadership role in, AbbVie, Amgen, Bristol Myers Squibb, Canadian Psoriasis Network, Eli Lilly and Company, Fresenius Kabi, Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA), Janssen, Novartis, UCB and University Health Network. AMF has received speaker honoraria, payment for expert testimony and/or travel support from AbbVie, Eli Lilly and Company, Janssen, Pfizer, Novartis and UCB. BZ, DK and TH are employees and minor shareholders of Eli Lilly and Company. NG has received grants, consulting fees, speaker honoraria, travel support or equipment/services, or participated on an advisory board for AbbVie, AstraZeneca, Eli Lilly and Company, Galapagos, Janssen, Novartis and UCB. AK, WF and IdlT are employees and minor shareholders of Eli Lilly and Company. DGMcG has received consulting fees, speaker honoraria, research grant support and/or travel support from AbbVie, Almiral, Bristol Myers Squibb, Eli Lilly and Company, Janssen, Novartis, Pfizer and UCB.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.