IL-23 Inhibitors for the Treatment of Psoriatic Arthritis

IL12/23 and IL-23 inhibitors have demonstrated superiority over TNF inhibitors in certain PsA domains, although none have shown superiority for peripheral arthritis. Most clinical trials use the American College of Rheumatology (ACR) 20 or ACR 50 response, assessing peripheral joint disease, as the primary endpoint while other studies focus on different PsA domains, such as enthesitis. Nonresponse in the included studies was categorized as those who did not achieve the primary or secondary endpoint that is being discussed.

Efficacy of IL-12/23 Inhibitor

The PSUMMIT 1 and 2 trials examined the efficacy of ustekinumab (45 mg or 90 mg) versus placebo in patients with active PsA, demonstrating effectiveness in both patients never treated with biologics and patients previously treated with TNFi [6]. Both trials included patients with active disease despite prior treatment with non-steroidal anti-inflammatory drugs (NSAIDs) and conventional synthetic disease-modifying antirheumatic drugs (csDMARDS). Additionally, PSUMMIT 2 included patients with prior TNF inhibitor use [7]. In the PSUMMIT 1 trial, 4.9%(45 mg) and 2.0% (90 mg) discontinued due to lack of efficacy. At week 24, a higher proportion of patients receiving ustekinumab (45 mg and 90 mg) achieved ACR20 compared with placebo (42.4% and 49.5% vs 22.8%, p < 0.0001) [6]. In the PSUMMIT 2 trial, both ustekinumab arms (45 mg and 90 mg) demonstrated a superior ACR20 response at week 24 compared with placebo (43.7% and 43.8% versus 20.2%, p < 0.001). For patients with one prior TNF inhibitor, the proportion of subjects achieving an ACR20 response was 34.8% (45 mg) and 35.5% (90 mg) compared with 10.0% (placebo). Among those with more than one prior TNF inhibitor, the proportion of patients achieving an ACR20 responses was 37.8% (45 mg), 33.3% (90 mg), and 18.8% (placebo) at week 24 [7].

Efficacy of IL-12/23 Inhibitor versus TNF Inhibitor

The ECLIPSA prospective study assessed the efficacy of ustekinumab (45 mg or 90 mg) compared with TNFi [8]. The primary endpoint was the clearance of enthesitis, defined by a Spondyloarthritis Research Consortium of Canada (SPARCC) index of zero. In both treatment arms, all enthesitis measures improved between baseline and 24 weeks. Clearance of enthesitis was significantly higher in the ustekinumab arm compared with the TNFi arm (74% versus 42%, p = 0.018). Patients in both arms demonstrated significant improvements in median Disease Activity Score 28 (DAS28), with patients receiving ustekinumab improving from 4.0 to 2.7 (p < 0.001) and patients receiving TNFi from 4.4 to 2.8 (p < 0.001) [8]. No patients discontinued either trial medication due to lack of efficacy. These findings were limited by the open label design and small sample size.

Both treatment arms also showed statistically significant improvements in Psoriasis Area and Severity Index (PASI) scores. Complete remission of skin disease (PASI100) was observed in 59% of patients treated with ustekinumab and 29% of patients treated with TNFi (p = 0.039), while nearly complete resolution of skin disease (PASI90) occurred in 86% of patients treated with ustekinumab and 29% of patients treated with TNFi (p = 0.0001) [8].

Efficacy of IL-23 Inhibitor After TNF Inhibitor

IL-23 inhibitors have demonstrated significant efficacy after prior exposure to TNF inhibitors. The DISCOVER-1 trial evaluated guselkumab in patients with active PsA, including those previously treated with TNFi [9]. Patients were randomly assigned to receive guselkumab every 4 (q4wks) or 8 weeks (q8wks) with the placebo group transitioning to guselkumab at week 24. ACR20 at week 24 was achieved at a higher proportion in both guselkumab groups (q4wks and q8wks) compared with those receiving placebo (59.4% and 52.0% versus 22.2% with p < 0.001). Similar achievement rates of ACR20 responses were observed in those with and without prior TNFi use [9]. In DISCOVER-1, no patients discontinued guselkumab due to lack of efficacy, while in DISCOVER-2, 1.2% of patients in both guselkumab groups discontinued treatment for this reason. In DISCOVER-2, patients receiving guselkumab q4wk (64%) and q8wk (64%) showed superior achievement of ACR20 at week 24 compared with placebo (33%, p < 0.0001) [10].

In addition, the COSMOS trial investigated the use of guselkumab compared with placebo in patients with prior TNF inhibitor inadequate response [11]. Patients were randomized by baseline csDMARD use and prior number of TNFi used. In this trial, 2.6% of patients receiving guselkumab (n = 189) discontinued due to inadequate efficacy. At week 24, a larger proportion of patients receiving guselkumab compared with placebo achieved ACR20 (44.4% and 19.8%, respectively, p < 0.001). The subgroups of patients receiving guselkumab who had prior treatment with one TNFi agent (47.3%) or two agents (22.7%), also had a higher proportion of patients achieving an ACR20 response compared with those receiving placebo (21.2% and 9.1%). Additionally, guselkumab was superior to placebo in Health Assessment Questionnaire-Disability Index (HAQ-DI) (p = 0.003), ACR50 (p = 0.001), Short Form 36 Physician Component Summary (SF-36 PCS) score (p < 0.001), and PASI100 (p < 0.001) (Table 1) [11].

Table 1 Summary of Phase III clinical trials evaluating the efficacy of IL-23 inhibitors compared with placebo in those with active psoriatic arthritis

The KEEPsAKE 1 and 2 trials evaluated the efficacy of risankizumab compared with placebo in patients with previous nonresponse to a biologic or csDMARD [12]. More patients initially receiving risankizumab achieved ACR20, the primary endpoint, compared with placebo (51.3% versus 26.5%; p < 0.001) with sustained efficacy through week 52 (58.5%). Patients who switched from placebo to risankizumab at week 24 also demonstrated improved ACR 20 response at week 52 (26.5% at week 24 versus 55.7% at week 52). Those who initially received risankizumab met all secondary endpoints at week 24 as well including the change in HAQ-DI, PASI 90, Minimal Disease Activity (MDA), change in SF-36 PCS score, and change in Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-Fatigue) score compared with placebo (p < 0.05) [13].

Real World Data of IL-23 Inhibitors

Several real-world studies have demonstrated the long-term persistence of IL-23 inhibitors compared with TNFi. A retrospective study by Vaiopoulos et al. evaluated the efficacy of guselkumab and risankizumab, stratifying patients based on prior biologic exposure, including prior TNFi, IL-17, IL-23, and biologic-naïve status [14]. IL-23 inhibitors significantly improved peripheral arthritis exhibiting a decrease in mean DAPSA score from 43.6 ± 20.9 to 20.5 ± 18.5 after 24 weeks (p < 0.001). PASI100 was also met in 100% of biologic naïve patients and 81.8% of patients with prior TNF inhibitor exposure (p = 0.09). Patients exposed to IL-17 or IL-23 inhibitors prior to guselkumab or risankizumab also achieved PASI 100 (64.1%, p = 0.04). In a study by Walsh et al., more patients continued guselkumab at 12 months (71.5%) compared with TNF inhibitors (43.7%) [15].

Persistence may vary for IL-12/23 inhibitors than IL-17 inhibitors compared with TNF inhibitors. Another study by Gossec et al. demonstrated a similar persistence rate at 12 months for ustekinumab and TNFi (72.4% and 70.5%) [16]. The mean time on both classes of biologics was also similar with a duration of 13.1 months for ustekinumab and 12.7 months for TNF inhibitors.

IL-17 Inhibitors for the Treatment of Psoriatic ArthritisEfficacy of IL-17 Inhibitors versus TNF Inhibitor

Several trials suggest that IL-17 inhibitors may not be superior to TNF inhibitors when treating peripheral PsA. The EXCEED trial compared secukinumab with adalimumab in patients with PsA, and demonstrated that secukinumab was not superior to adalimumab, as 67% in the secukinumab and 62% in the adalimumab group achieved ACR20 response at week 52 (p = 0.072) [17]. In this trial, 2.6% of patients receiving secukinumab (n = 426) and 5.4% receiving adalimumab (n = 427) discontinued the trial agent owing to lack of efficacy (Table 2).

Table 2 Summary of phase III clinical trials evaluating the efficacy of IL-17 inhibitors in those with active psoriatic arthritis

Another trial, Spirit H2H, compared ixekizumab and adalimumab in biologic naïve patients with PsA and PsO with a history of an inadequate response to csDMARDs [18]. Patients were randomized on the basis of severity of PsO, mild and moderate to severe, and received different doses based on which arm they qualified for. Of the adalimumab arm (n = 283), 1.1% discontinued owing to lack of efficacy, compared with 2.1% in the ixekizumab arm (n = 283). The primary endpoint of achieving ACR50 and PASI100 was assessed at week 24. The proportion of patients achieving ACR50 and PASI100 simultaneously was higher for those receiving ixekizumab (36%) compared with those receiving adalimumab (28%, p = 0.036) [18]. However, there was no statistically significant difference in the treatment arms receiving ixekizumab and adalimumab when only ACR50 was assessed (50.5% versus 46.4%, respectively, p = 0.338), although the PASI100 response was notably greater in patients receiving ixekizumab compared with patients receiving adalimumab (60% vs 47%, p = 0.001).

Efficacy of IL-17 Inhibitor After TNF Inhibitor

The SPIRIT-P1 and SPIRIT- P2 trials suggest that IL-17 A inhibitors and TNF inhibitors may provide similar efficacy for treating PsA, although IL-17 A inhibitors still maintain efficacy following TNF inhibitor therapy. These trials assessed the efficacy of ixekizumab and adalimumab versus placebo in biologic-naïve patients (SPIRIT-P1) or in patients who had an inadequate response to TNF inhibitors (SPIRIT-P2). In SPIRIT-P1, patients who initially received adalimumab switched to ixekizumab at week 24. ACR 20 at week 24, the primary endpoint, was achieved in 57.4% of those receiving adalimumab and 60% in those receiving ixekizumab every 2 weeks or 4 weeks. [19] For those who switched from adalimumab to ixekizumab, ACR 20 was achieved at week 52 in 64.9% compared with 68.9% of those who received ixekizumab throughout the duration of the trial [20]. At week 24, PASI 90 was also met in 61.9% of those receiving ixekizumab and 36.8% of those receiving adalimumab. An increased proportion of patients achieved PASI 90 after switching from adalimumab to ixekizumab (50.7%) [20]. SPIRIT-P2 showed similar efficacy of ixekizumab after TNF failure with an increased proportion of patients achieving ACR 20 compared with placebo with a sustained response through week 52 (Table 2) [21, 22].

Efficacy of IL-17 A/F Inhibitor After TNF Inhibitor

IL-17 A/F inhibition is efficacious following TNF inhibitor therapy. The BE COMPLETE study evaluated bimekizumab compared with placebo in patients with active PsA and inadequate response or intolerance to TNF inhibitors [23]. Patients were stratified by previous nonresponse to TNF inhibitors (1 or 2 prior) or by inadequate response to TNF inhibitors. Of the bimekizumab arm (n = 267), 0.4% discontinued due to lack of efficacy. The primary endpoint, ACR 50, was achieved in 42.7% of participants receiving bimekizumab and 6.8% receiving placebo (p < 0.0001). Bimekizumab was also superior in multiple secondary endpoints including HAQ-DI score change from baseline (−0.38 versus −0.07, p < 0.0001), PASI 90 (68.8% versus 6.8%, p < 0.0001), SF-36 PCS score change from baseline (7.3 versus 1.4, p < 0.0001), and MDA response (44.2% versus 6%, p < 0.0001) [23].

A network meta-analysis of efficacy outcomes for 22 biologics including bimekizumab in the treatment of PsA and PsO was conducted separating participants into biologic naïve and TNF inhibitor experienced groups for analysis [24]. Bimekizumab and comparators were ranked by efficacy using the surface under the cumulative rank curve (SUCRA). For patients who were TNF inhibitor experienced, bimekizumab ranked first in participants achieving ACR20, ACR 70, and PASI 90 and second for ACR 50. For the biologic naïve arm, bimekizumab ranked first for participants achieving PASI 100 and second for achieving PASI 90; however, bimekizumab ranked third for ACR70, fifth for ACR50, and sixth for ACR 20 [24].

Real World Data of IL-17 Inhibitors

Two real world retrospective studies evaluated the persistence of IL-17 inhibitors compared with IL-12/23 inhibitors. Persistence was defined as how long the patient continued the medication. For patients with PsA, IL-17 inhibitors had higher persistence compared with IL-12/23 inhibitors and no difference was noted when comparing persistence between IL-12/23 and TNF inhibitors (Table 3) [25]. Letarouilly et al. supported the superior persistence of IL-17 inhibitors compared with IL-12/23 [26]. The median persistence of secukinumab was notably higher than ustekinumab with sustained persistence at 1 and 2 years (Table 3).

Table 3 Summary of real-world studies evaluating long-term use of biologics in patients with psoriatic arthritisBiologics for the Treatment of Axial Manifestations of Psoriatic Arthritis

Axial manifestations of psoriatic arthritis have not been fully assessed in most clinical trials due to the lack of axial specific clinical measures. TNF inhibitors have been effective for axial manifestations, and recent clinical trials also support the use of IL-17 inhibitors for clinically significant improvement in axial disease.

Efficacy of IL-17 Inhibitors

The MAXIMISE trial compared secukinumab with placebo in patients who were biologic naïve with axial disease [27]. Patients who initially received placebo were rerandomized to secukinumab at week 12. Magnetic resonance imaging (MRI) was performed at week 12 and 52 to assess active inflammation of the spine and sacroiliac joints with approximately 60% of patients in the trial having active inflammation observed on MRI at baseline. At week 12, the primary endpoint, Assessment of Spondylarthritis International Society (ASAS) 20, was met in similar proportions of patients receiving secukinumab 300 mg and 150 mg and was superior to placebo (63% and 66% versus 31%, p < 0.0001). This efficacy was continued through week 52 with 81% of those receiving secukinumab throughout the trial achieving ASAS 20 and 77% of those who switched from placebo to secukinumab achieving ASAS 20. ASAS 20 response was achieved in 68% of patients with positive MRI findings receiving secukinumab compared with 27% of those with positive MRI findings receiving placebo [27].

Efficacy of IL-23 Inhibitors

Although the efficacy of IL-23 inhibitors for axial disease has not been established, trials for ustekinumab in axial PsA have demonstrated poor efficacy. Ustekinumab demonstrated similar efficacy to TNF and IL-17 inhibitors for ankylosing spondylitis in prior studies and was hypothesized to be an efficacious treatment option for PsA. However, the primary outcome, ASAS 40, was not met in three previous clinical trials and the efficacy of ustekinumab was similar in those who had previous TNF nonresponse (23%) compared with those who were TNF naïve (26%) [28, 29]. Although future research is needed to define the role of IL-23 inhibitors in axial disease, a current phase 4 clinical trial, STAR, is being conducted to assess the efficacy of guselkumab for treatment of axial PsA [30].

Efficacy of Biologics in the Treatment of PsOIL-23 Inhibitor After TNF Inhibitor

The presence of concomitant PsO might influence the recommended sequence of biologic therapy, depending on the extent of the disease. IL-23 and IL-17 inhibitors have shown greater efficacy for the skin compared with TNF inhibitors in several phase III clinical trials. The VOYAGE 1 and 2 trials compared guselkumab to adalimumab in both biologic naïve and experienced patients. In both trials, patients who initially received placebo transitioned to guselkumab at week 16. In addition, patients who were nonresponders to adalimumab in VOYAGE 2 initiated guselkumab at week 28. In VOYAGE 1, 0.9% of patients receiving guselkumab (n = 329) and 3.6% of patients receiving adalimumab discontinued the study agent owing to lack of efficacy [31]. Guselkumab was superior to adalimumab in both trials with 73.3% and 70% achieving PASI 90 in VOYAGE 1 and VOYAGE 2 at week 16 compared with 49.7% and 46.8%, respectively (p < 0.001) [31, 32]. Guselkumab also demonstrated a decreased rate of nonresponse, defined as not achieving PASI 90, at week 28 compared with adalimumab (19% versus 45%). In addition, those who transitioned from adalimumab to guselkumab had an improvement in skin clearance from week 24 (PASI90, 54.8%) to week 48 (PASI90, 66.1%).

IL-17 A/F Inhibitor versus TNF Inhibitor

The BE SURE trial compared bimekizumab and adalimumab in patients with moderate to severe PsO [33]. Patients transitioned from adalimumab to bimekizumab at week 24. Bimekizumab was superior to adalimumab with 86.2% achieving PASI 90 at week 16 compared with 47.2% (p < 0.001). At week 24, 29.6% of those receiving adalimumab achieved PASI 100 compared with 66.8% of those receiving bimekizumab. Further, 66.7% of those who switched from adalimumab to bimekizumab achieved PASI 100 at week 56 compared with 70.2% of those who initially received and continued bimekizumab, exhibiting similar efficacy [33].

The BE OPTIMAL trial also compared bimekizumab and adalimumab but included patients with concomitant PsA [34]. The primary endpoint, ACR 20 at week 16, was met in 62.2% of patients receiving bimekizumab and 68.6% of patients receiving adalimumab. PASI 100 at week 16 was met in 47.5% of patients receiving bimekizumab and in 20.6% of patients receiving adalimumab. Bimekizumab also demonstrated an increased sustained response in achieving PASI 100 compared with adalimumab at week 52 (60.8% versus 48.5%, respectively) [34]. Of the bimekizumab arm (n = 431), 0.2% discontinued due to lack of efficacy, compared with the adalimumab arm (n = 140) with no discontinuation due to this reason.

IL-23 Inhibitor versus IL-17 Inhibitors

The Phase 3, randomized, controlled trial, ECLIPSE, was a direct comparator trial of guselkumab and secukinumab for the treatment of moderate to severe PsO [35]. Overall, 17% of patients in the trial had concomitant PsA although no rheumatologic endpoints were collected. Of the patients receiving guselkumab (n = 534), 0.4% discontinued study treatment due to inadequate efficacy, compared with patients receiving secukinumab (n = 514) with 1.4% discontinuation due to lack of efficacy. The primary endpoint, PASI 90 at week 48, was met in 84% of patients receiving guselkumab and 70% of patients receiving secukinumab (p < 0.0001) [35]. Secukinumab also had an increased number of patients discontinue the trial due to lack of efficacy compared with guselkumab (1.4% versus 0.4%).