Poor Tolerance of MTX

Although the data are mostly for MTX used in rheumatoid arthritis (RA), the situation is similar for PsA; indeed, as for RA, the first conventional synthetic DMARD (csDMARD) recommended is MTX, even more so with associated psoriasis [1]. Thus, in RA, compliance with MTX is estimated at 60–80% [19,20,21,22], and the various international registries for follow-up of biological treatments report a monotherapy rate of about 30% of patients [23,24,25,26,27,28]. Approximately 30% of patients cannot tolerate MTX and 30% cannot tolerate a dosage > 15 mg per week [29, 30]. Overall, 10–30% of patients stop MTX because of intolerance, most often of digestive origin (nausea, abdominal pain) or alopecia, hepatic cytolysis, etc. [31, 32]. Nikiphorou et al. showed in 762 patients with RA and 193 with PsA that MTX had been stopped in 260 (34%) patients with RA and 71 (36%) patients with PsA most commonly as a result of gastrointestinal intolerability. The median duration of MTX treatment was 10 months in both groups, and overall, one-third of patients with RA and PsA stop MTX most commonly as a result of poor tolerability (this analysis did not include patients who suffer from side effects but continue therapy) [31]. Because of this poor tolerability, the possibility to stop the treatment when initiating targeted therapy would represent real added value.

The Efficacy of Targeted Therapies Combined with MTX is Not Superior to That of Monotherapy in PsA

Contrary to what is observed in RA treatment (i.e., greater efficacy of targeted therapies, especially anti-TNFα agents combined with MTX versus monotherapy), none of the targeted therapies available for PsA have shown superior efficacy when combined with MTX versus monotherapy [33,34,35,36,37,38,39,40,41,42]. A systematic review of the literature showed that the anti-TNFα–MTX combination was no more effective than anti-TNFα monotherapy in PsA [33], and this was demonstrated for each of the available products (Table 1) [4, 5, 8, 13, 34,35,36,37,38,39,40,41,42].

Table 1 Efficacy of targeted therapies combined with methotrexate (MTX) or in monotherapy in psoriatic arthritis (PsA)Anti-TNFα Monoclonal Antibodies (mAbs) are Immunogenic in PsA

Despite less data for PsA than RA, studies show that anti-TNFα mAbs can be responsible for immunization and the development of anti-drug antibodies (ADAs) in 20–30% of cases during PsA treatment. Also, as in RA, this immunization can be significantly reduced when combined with MTX [43,44,45,46,47] (Table 2). Three studies in patients with PsA treated with adalimumab showed the development of ADAs in 18–29% of cases, a lower serum adalimumab concentration, and poorer clinical outcome in patients with ADAs; MTX use was significantly correlated with low prevalence of ADAs [43,44,45] (Table 2). In the GO-REVEAL trial, the incidence of ADAs in response to golimumab at week 14 was low (4.6%), and no patient taking MTX at baseline had ADAs [46]. By week 104, 5.4% of golimumab-treated patients had ADAs, and this was less common in patients taking versus not taking MTX at baseline (1.6% vs 9.1%) [46]. In the IMPACT 2 study, only 3.6% of patients taking MTX at baseline were positive for ADAs to infliximab through week 66 as compared with 26.1% not taking MTX [47]. ACR improvement at week 54 was lower for patients with than without ADAs (22% vs 33%), and mild to moderate infusion reactions were 3.5-fold more common in patients with than without ADAs [47].

Table 2 Immunogenicity of anti-TNF mAbs in psoriatic arthritis

Thus, during the treatment of PsA with anti-TNFα mAbs, ADAs develop in 10–30% of patients and are associated with low serum concentrations of the drug and a secondary loss of efficacy and therefore less therapeutic maintenance. In addition, the development of ADAs is reduced when treatment is combined with MTX.

Anti-IL-17, Anti-IL-12/23, and Anti-IL-23 Antibodies are Poorly Immunogenic and JAKi Not at all Immunogenic

Unlike anti-TNFα mAbs, the other therapeutic antibodies (anti-IL17, anti-IL-12/23, and anti-IL-23) used for treating PsA are poorly immunogenic (Table 3).

Table 3 Immunogenicity of anti-IL-17, IL-12/23, and IL-23 mAbs in psoriatic arthritis

The low immunogenicity potential of secukinumab was first confirmed in psoriasis studies [48, 49]; across six phase III clinical trials, treatment-emergent ADAs (TE-ADAs) developed in 11/2842 patients (0.4%), with no association with secukinumab dose, frequency, or mode of administration [48]. In two phase III extension studies up to 5 years, TE-ADAs developed in 32/1636 (< 2%) patients, with no effect on the efficacy, safety, or pharmacokinetics of secukinumab [49]. In a systematic review of 443 study publications of chronic inflammatory diseases, secukinumab was responsible for the development of less than 1% of ADAs and ustekinumab for 1–11% [50]. TE-ADAs developed in 5 (0.35%) of 1414 patients with PsA treated with secukinumab, with no association with secukinumab dose, frequency, administration mode, secukinumab pharmacokinetics, or loss of efficacy [51].

In a phase III study of treatment with ixekizumab for moderate to severe plaque psoriasis, TE-ADAs were never detected in most patients (82.6%; 308/373) [52]. Among TE-ADA-positive patients (n = 65, 17.4%), 56 had low or moderate maximum titers, with serum concentrations and efficacy comparable to those of TE-ADA-negative patients. The median serum concentration in the high-titer TE-ADA group was in general comparable to that in the low-titer group, and for most patients, TE-ADAs had a negligible impact on ixekizumab serum concentration and efficacy [52]. In patients with PsA receiving ixekizumab (SPIRIT-P1 and -P2), among 107 patients taking MTX, 11 (10.3%) were TE-ADA-positive, 10 with low titers, and among 119 patients without MTX, 14 (12%) were TE-ADA-positive, 12 with low titers. ADAs did not affect the ixekizumab ACR20 response at week 52 [53].

In one phase IIA study of psoriasis, the incidence of anti-bimekizumab (BKZ) antibodies was 34% in the BKZ + placebo group and 47% in the BKZ group (probably with a very sensitive detection method), with no impact on drug exposure, clinical response, or safety [54]. The Summary of Product Characteristics for BKZ showed that ADAs developed in approximately 31–57% of patients, 33–44% of whom had neutralizing antibodies across indications, with no clinically meaningful impact on clinical response, and no association between immunogenicity and treatment-emergent adverse events [55].

A post hoc analysis of 112 PsA serum samples of individuals receiving open-label ustekinumab along with MTX (ustekinumab/MTX, n = 58) or placebo (ustekinumab/pbo, n = 54) revealed ADAs in 11 ustekinumab/pbo- and 19 ustekinumab/MTX-treated patients over 52 weeks (p > 0.05). At week 52, ADA-confirmed patients did not differ significantly from ADA-negative patients (p > 0.05) in safety or clinical outcomes. Thus, concomitant MTX had no significant impact on ustekinumab immunogenicity and ADA formation was not associated with impairments in ustekinumab safety, efficacy, or trough levels [56].

In 5-year data from two phase III randomized studies with guselkumab in patients with psoriasis, 111/770 (14.4%) and 146/943 (15.5%) were ADA-positive at some point through week 264 [57]. Only 5 (4.5%) and 8 (5.5%) ADA-positive patients had neutralizing antibodies equivalent (0.76% of patients receiving guselkumab with ADA-evaluable samples). Through week 264, serum guselkumab concentrations were comparable between ADA-positive and -negative patients, and the development of guselkumab ADAs did not seem to affect the clinical response [57]. In phase III pooled assays of PsA (DISCOVER 1 and 2 trials), at 1 year (DISCOVER 1), ADAs developed in 4.5% (49/1094) of patients, and at 2 years (DISCOVER 2) in 7.3% of patients [58].

For individuals receiving risankizumab for up to 52 weeks in psoriasis clinical trials, TE-ADAs and neutralizing antibodies were detected in 24% (263/1079) and 14% (150/1079), respectively. For most individuals, ADAs to risankizumab were not associated with changes in clinical response or safety. For patients receiving risankizumab for up to 28 weeks in PsA clinical trials, TE-ADAs and neutralizing antibodies were detected in 12.1% (79/652) and 0% (0/652), respectively, and were not associated with changes in clinical response or safety [59].

JAKi, which are not therapeutic antibodies but rather small chemical molecules, by definition are not immunogenic.

Thus, contrary to what is observed with anti-TNFα mAbs, the other therapeutic classes used for treating PsA (anti-IL-17, -12/23, -23) are poorly immunogenic or not at all immunogenic (JAKi). In addition, when neutralizing ADAs are produced, there is no impact on the serum concentration or clinical efficacy and therefore therapeutic maintenance. Finally, MTX does not seem to affect this immunization phenomenon.

Anti-TNFα mAbs Seem to Have Better Therapeutic Maintenance When Combined with MTX

Because of the immunogenicity of anti-TNFα mAbs and its reduction by MTX, logically, combined treatment with MTX may allow for better therapeutic maintenance during PsA treatment; this hypothesis was strongly suggested by a systematic review of the literature and data from several registries or observational studies [23, 33, 60,61,62,63,64,65].

A systematic review of the literature published in 2015 and therefore taking into account only anti-TNFα agents included therapeutic, controlled, randomized, and observational studies comparing anti-TNFα agents as monotherapy or combined with MTX, particularly regarding therapeutic maintenance, in PsA [33]. Drug survival data were reported for four registries. In the NOR-DMARD registry, concomitant MTX was associated with better 1-year TNFα antibody survival in patients with PsA, and no combination-treatment courses were discontinued as a result of lack of efficacy, whereas lack of efficacy was the reason for approximately 40% of monotherapy discontinuations [23].

In another analysis of NOR-DMARD data, patients receiving TNF inhibitors as monotherapy (n = 170) or combined with MTX (n = 270) had similar baseline characteristics, except for higher swollen joint count in the MTX than non-MTX group [60]. Drug survival analyses favored patients receiving MTX, although not significantly (p = 0.07), most prominently those receiving infliximab (p = 0.01). A similar trend was seen for adalimumab (p = 0.12, not significant, probably because of lack of power), and the difference in the etanercept group was negligible (p = 0.79). In a Cox regression analysis, lack of concomitant MTX and current smoking were independent predictors of TNF inhibitor discontinuation. Discontinuation rates were markedly higher for patients receiving infliximab as monotherapy because of adverse events (p < 0.001) [60].

The South Swedish Arthritis Treatment Group noted no baseline differences between 100 patients receiving TNF inhibitor monotherapy and 161 patients receiving concomitant MTX, other than greater non-steroidal anti-inflammatory drug use in the MTX than no-MTX group [61]. Etanercept treatment and high C-reactive protein level at treatment initiation were associated with better drug survival over 7 years. The improved drug survival with concomitant MTX seemed related to significantly fewer dropouts because of adverse events [61].

In the Danish biologics registry, 54% of 410 patients with PsA were receiving concomitant MTX at baseline [62]. Baseline MTX use was more prevalent among patients receiving infliximab (70%) than adalimumab (49%) or etanercept (39%). In a Cox regression analysis, male sex, C-reactive protein level > 10 mg/l, concomitant MTX use and low visual analogue scale score for patient health at baseline were associated with longer drug survival. The adjusted hazard ratio (HR) for discontinuing the TNF inhibitor associated with lack of MTX use was 1.37 (95% CI 1.07–1.75). Lack of MTX use was associated with discontinuation due to adverse events but not lack of efficacy [62]. In contrast to the above findings, a study of the Canadian Rhumadata clinical database and registry found that concomitant MTX did not improve 5-year retention with adalimumab or etanercept (52% for combination therapy vs 67% monotherapy; p = 0.74) [63]. One study compared treatment outcomes for 15,332 patients with PsA treated with TNF inhibitor and csDMARD co-treatment (62%) versus TNF inhibitor monotherapy (38%) in patients who initiated a first TNF inhibitor in 13 European countries in 2006–2017. MTX co-medication was associated with improved remission for adalimumab and infliximab and improved retention for infliximab. No effect of co-medication was demonstrated for etanercept [64]. Finally, a prospective observational study of 82 patients with PsA receiving etanercept with or without concomitant MTX in one Italian center found no significant difference in the main demographic and clinical features, including rates of withdrawal due to inefficacy or toxicity, between patients receiving etanercept alone and those concomitant MTX therapy; concomitant MTX treatment did not seem a positive predictor of drug survival [65].

Anti-IL-17 Antibodies May Have Better Therapeutic Maintenance than Anti-TNFα Antibodies as First-Line Targeted Therapy in PsA

There are limited data on the maintenance of different targeted therapies and even less on how they compare over time in PsA. In addition, the data on persistence observed during extensions of phase III, registry, or observational studies are sometimes “optimistic”, are quite heterogeneous, and do not always reflect real life.

Maintenance of Anti-IL-17 Antibodies

In a retrospective, multicenter study of 178 patients with PsA receiving secukinumab as a first-, second-, or third-line or greater bDMARD therapy (for 37%, 21%, and 42% of patients, respectively), the overall 24-month persistence rate was 67% (95% CI 60–74). Patients receiving first-line secukinumab had the highest 24-month persistence (83% [95% CI 73–92], p = 0.024) [66].

Thirteen quality registries in rheumatology participating in the European Spondyloarthritis Research Collaboration Network provided longitudinal, observational data to assess real-life 6- and 12-month effectiveness (i.e., retention, remission, low disease activity, and response rates) of secukinumab in 2017 patients with PsA. Secukinumab retention rates were 86% and 76% after 6 and 12 months, respectively, significantly higher in biologic-naïve patients than patients who previously received at least two b/tsDMARDs (90% and 82% vs 85% and 72%) [67].

A French retrospective study collected data from 842 patients with PsA of whom 15.8%, 24.2%, and 60.0% received secukinumab as a first-, second-, or third-line or greater bDMARD, respectively. The 1-year retention rate for secukinumab was 64% (95% CI 61–68) and was numerically greater with first- versus second- and third-line or greater therapy: 77% (95% CI 69–84) vs 65% (59–72) and 61% (57–66) [68].

Studies Comparing the Maintenance of Anti-IL-12/23 and Anti-TNFα Antibodies

The prospective, observational study PsABio followed patients with PsA who were prescribed first- to third-line treatment with ustekinumab or a TNF inhibitor. At 3 years, among 895 patients, the proportion still on their initial treatment was similar with ustekinumab and TNFi treatment (49.9% and 47.8%). Drug persistence was better for patients with first- or second-line bDMARD treatment than those with third-line treatment [69]. In the Italian part of the PsABio study (222 patients, 101 on ustekinumab and 121 on TNF inhibitor), at 1 year, 74.3% of the ustekinumab group continued treatment up to a mean of 12 ± 3 months versus 63.6% in the TNF inhibitor group. Ustekinumab had better persistence than a TNF inhibitor, overall and in specific subgroups (female patients, monotherapy without MTX, body mass index < 25 or > 30 kg/m2, patients receiving ustekinumab as second-line treatment instead of a second TNF inhibitor) [70].

Studies Comparing the Maintenance of Secukinumab and Anti-TNFα Agents

EXCEED, a double-blind, active-controlled, phase IIIb, multicenter study, evaluated the efficacy and safety of secukinumab (300 mg) versus adalimumab as first-line biological monotherapy for 52 weeks in patients with active PsA; 61/426 (14%) patients in the secukinumab group discontinued treatment by week 52 versus 101/427 (24%) in the adalimumab group, which suggested that the treatment retention rate was higher with secukinumab than adalimumab [71].

The biologic registers of the Nordic countries were used to identify all patients with PsA starting secukinumab or a TNF inhibitor in 2015–2018. All analyses were stratified by line of biologic treatment. The study included 6143 patients with 8307 treatment courses (secukinumab, n = 1227; adalimumab, n = 1367) and found no clinically significant difference in treatment retention or response rates between the two agents. The adjusted HRs for discontinuation with the first, second, and third line of treatment were 0.98 (95% CI 0.68–1.41), 0.94 (0.70–1.26), and 1.07 (0.84–1.36), respectively [72].

Studies Comparing the Maintenance of b/tsDMARDs

A nationwide cohort study involved the administrative healthcare database of the French health insurance system [73]. All adults with psoriasis (PsO) and PsA who were new users of biologics (not in the year before the index date) from January 1, 2015 to May 31, 2019 were included and followed up through December 31, 2019; 16,892 patients with PsO were included. Of these patients, 10,199 (60.4%) started a TNF inhibitor, 3982 (23.6%) an IL-12/23 inhibitor, and 2711 (16.0%) an IL-17 inhibitor. An additional 6531 patients with PsA were included: 4974 (76.2%) started a TNF inhibitor, 803 (12.3%) an IL-12/23 inhibitor, and 754 (11.5%) an IL-17 inhibitor. Overall 3-year persistence rates were 40.9% and 36.2% for PsO and PsA, respectively. After adjustment, the IL-17 inhibitor was associated with higher persistence as compared with the TNF inhibitor for PsO (weighted HR 0.78 [95% CI 0.73–0.83]) and PsA (weighted HR 0.70 [95% CI 0.58–0.85]) and as compared with the IL-12/23 inhibitor for PsA (weighted HR 0.69 [95% CI 0.55–0.87]). The overall persistence rate for PsA was relatively low, 73%, 49%, and 36% at 1, 2, and 3 years, respectively: 72%, 49%, and 35% for anti-TNFα antibodies; 72%, 48%, and 38% for anti-IL-12/23 antibodies; and 77%, 51%, and 42% for anti-IL-17 antibodies. The findings of this cohort study suggested that treatment persistence is better for an IL-17 inhibitor than TNF inhibitor as first-line targeted therapies for PsO and PsA [