Long-term effectiveness and drug survival of golimumab in patients affected by psoriatic arthritis with cutaneous involvement

Psoriatic arthritis (PsA) is a chronic inflammatory disease, belonging to the group of spondyloarthritis (SpA), typically associated with psoriasis (PsO) and characterized by the presence of both articular and periarticular structures involvement [1]. The prevalence of PsO ranges from 2 to 3% in the general population, and PsA may affect 10–30% of PsO patients [2]. Joint involvement in PsA is a potentially debilitating disease: patients may present arthritis to both small and large peripheral joints, enthesitis, and the axial disease, leading to progressive erosive arthritis and severe functional impairment in more than half of the patients [3]. PsA pathogenesis is only partially understood and still remains to be completely clarified. The presence and the overexpression of cytokines such as tumor necrosis factor (TNF)-α, interleukin (IL)-17, and IL-23, plays a role in the pathogenic ways linking PsA and PsO [4]. The use of TNF inhibitors (TNFi) approximately two decades ago has dramatically improved PsA treatment. However, several unmet needs concerning the management of both joint and skin manifestations still remained [5]. Randomized-placebo controlled clinical trials involving TNFi have shown excellent results for PsA and PsO treatment. Etanercept, infliximab, adalimumab, golimumab (GLM), and certolizumab have been demonstrated to be effective in PsA and in distinctive aspects of the disease, as skin disease, peripheral arthritis, enthesitis, and dactylitis. Moreover, their effectiveness was demonstrated in improving quality of life and work ability and in reducing radiographic progression [6]. GLM is a TNFi approved by the European Medicines Agency (EMA) and the US Food and Drug Administration (FDA) for the treatment of active rheumatoid arthritis, active PsA, active ankylosing spondylitis, severe non-radiographic axial SpA, polyarticular juvenile idiopathic arthritis, and moderate to severe active ulcerative colitis [7]. GLM is a fully human monoclonal antibody IgG1k-neutralizing TNF-α. Clinical trials as GO-REVEAL studies demonstrated a satisfactory efficacy of GLM in improving PsA signs and symptoms and in treating the structural damage caused by the disease [8]. Treatment recommendations underline the relevance of the burden of skin PsO in patients affected by PsA and skin manifestations need to be a relevant part of the decision-making process for PsA therapeutic approach [9]. However, the effectiveness evaluation of GLM in a real-life setting is a crucial issue, in particular for skin manifestation. To date, only a few unsponsored long-term studies reported the efficacy of GLM treatment in PsA patients and their treatment adherence and, to our knowledge, the GLM effectiveness according to patients characteristics and lines of treatment (i.e., 1st, 2nd, and 3rd or more lines) was rarely considered [10, 11]. Moreover, limited data are available on the effectiveness and safety of GLM in patients affected by both PsO and PsA in bio-naïve and TNF-insufficient responders (TNF-IR), in particular concerning skin sustained efficacy.

To address these knowledge gaps, the aims of this multicenter observational study were to evaluate the efficacy and the long-term treatment retention rate in patients affected by PsA with associated PsO treated with GLM in a real-life setting. Moreover, we aimed at evaluating the treatment retention rate based on patient characteristics, such as gender, body mass index (BMI), presence of comorbidities, smoking habit, and line of GLM treatment.

In this observational retrospective, multicenter study in patients with moderate to severe PsA with cutaneous manifestations of PsO, GLM was effective for both articular and skin involvement. Indeed, GLM was effective in reducing all clinical parameters analyzed at all the time points of evaluation until 48 months of observation including inflammatory indexes and pain. Data at 36 months were not shown because stackable with those at 48 months. Clinical phenotype of patients was characterized by cutaneous involvement represented in all patients while, concerning articular and peri-articular manifestations, a high prevalence of polyarticular involvement and enthesitis was found. Long-standing disease, high prevalence of erosions, and presence of axial disease demonstrated the presence of a challenging study population of PsA patients. Moreover, the high disease activity at baseline measured by specific composite scores supported the severity of PsA. The statistically significant reduction of clinimetrics indexes, inflammatory markers, and the gain of a high percentage of patients in DAPSA low disease activity support the efficacy of GLM in our “difficult-to-treat” PsA patients. Remission according to DASPA score, an ambitious endpoint, was obtained in 20% of patients at T12 and 36.8% of patients were in clinical remission at T48. In this context, we assume that TNF inhibition represents one of the key treatments for PsA and PsO, as confirmed by clinical trials demonstrating the efficacy of TNFi in both PsA and PsO [5]. The reduction of inflammatory markers, such as CRP, considered a biomarker of radiographic progression in SpA, may be considered as an indirect outcome of the reduction of radiographic damage [17]. Based on ASDAS-CPR and BASDAI scores, significant improvement was reached and maintained within the first 12 months of treatment and sustained during the following 48 months of follow-up for axial involvement. Most patients evaluated in our study had been previously treated with other biologic agents (TNFi) and only 41.9% of patients were bio-naïve with 57% in concomitant csDMARD treatment. The drug survival of GLM was independent of the line of treatment used and the previous use of bDMARDs and can be considered effective both as first-line therapy and in patients with failure for one or more bDMARDs. Data from international and national registries, such as the GISEA registry or the BioRx.si, have compared the characteristics of patients starting treatment with GLM [11, 18, 19]. bDMARD-naïve and bDMARD-experienced patients were compared and no difference in the persistence of GLM was found. The switch among the same group of drugs, as the TNFi family, is still a debate in PsA management and patients failing a first TNFi agent usually use a different target [11]. However, several data from the literature seem to support the efficacy of switching among TNFi in PsA [10, 19]. The evaluation of drug persistence in our real-life experience demonstrated a good percentage (54%) of patients still on therapy after 4 years of GLM treatment. Other studies, evaluating the same outcome, estimated approximately 47% of PsA patients, from the UK, remained on their initial TNFi therapy after 5 years and low persistence of 32% after 5 years of follow-up on PsA patients in Germany, while no differences among TNFi were observed [20‐22].

Recently, none of the TNFi agents was found to be more persistent than others as first-line therapy, while secukinumab was found to be superior to other biologics when indicated as second-line therapy [23]. However, data from Slovenia showed a better persistence of GLM compared to other TNFis in bDMARD-experienced AS and PsA patients [19].

The use of combination therapy as bDMARDs with MTX or other csDMARDs is still controversial in the management of PsA since its rational use is based on indirect evidence or expert opinion [24]. However, csDMARDs may be warranted in peripheral joint manifestations when bDMARDs are not effective [25]. We have not observed significant differences in survival rates among patients using MTX or not, confirming that GLM was effective in both combination treatment and monotherapy. This result suggests to using the GLM also in those patients that are intolerant to csDMARDs, making MTX treatment not necessary when a bDMARD is started. Furthermore, no difference in survival rates was observed when patients were divided according to the presence of comorbidities, BMI, or gender.

MetS was strongly associated with a lower probability to achieve MDA in TNFi-treated PsA patients, and different studies reported a higher prevalence of MetS in PsA patients compared to other rheumatic diseases [26, 27]. In our cohort of patients, MetS was not a negative predictor factor for GLM efficacy, in contrast with other findings supporting the hypothesis that MetS or overweight and obesity may affect TNFi therapy, as adalimumab or etanercept. GLM efficacy might not be affected by the presence of MetS because the dosing regimen is adequate to the weight of the patients being MetS related to chronic inflammatory state and adipokines. Lifestyle such as smoking habit is associated with high disease activity in PsA patients and may exert a negative role during the treatment with bDMARDs [28]. We have not observed the influence of smoking in our 4 years of follow-up, making GLM effective also in smokers.

Among other factors affecting treatment response, gender may influence bDMARDs clinical response remaining a challenge in PsA. In particular, female gender has been associated with a prevalent polyarticular phenotype, high swollen joint count, and peripheral erosive joint involvement compared to males, while female gender correlated with lower rates of extraarticular manifestations and axial involvement [28, 29]. Moreover, the female gender generally associates with poor rates of response to TNFi and a lower probability of achieving remission compared to males [30], especially when comorbidities, such as MetS, occur [30‐32]. In our cohort, we did not observe gender differences in the efficacy of GLM, since gender was not a relevant factor in patient outcomes. The high prevalence of IBD in our cohort is due to the early referral in the combined gastro-rheuma outpatient clinic by gastroenterologists from the involved hospitals [33, 34]. Thyroid dysfunction may be associated with the worst clinical outcome in several inflammatory conditions and in PsA [26, 35, 36]. According to our data, thyroid disorders may represent a negative prognostic factor in the PsA study population for treatment persistence even if this result needs to be confirmed in a larger study cohort. Thyroid disorders may have a relevant impact on disease perception and activity and so may have an influence on treatment efficacy.

A very low rate of primary and secondary failure was observed with a good safety profile with no serious adverse events reported consistent with previous reports of patients treated with GLM [37‐39].

This study encompasses limitations such as the retrospective nature of the study, the presence of low cutaneous disease activity at baseline, the relatively low number of PsA patients enrolled due to the inclusion of patients with both joint and skin manifestations, and the absence of imaging in the follow-up (radiographic evaluation of efficacy), even if ultrasound, MRI, and/or X-ray was performed in all patients to confirm the diagnosis at baseline.

In this multicentric study, we described the clinical efficacy in joint and periarticular structures as well as on PsO manifestations on moderate to severe PsA patients treated with GLM in a real-life setting. Treatment efficacy was supported by a good drug persistence of GLM regardless of the presence of comorbidities and patients’ characteristics. Our results were consistent with no differences in terms of clinical response and efficacy between males and females, smokers and no-smokers, obese and normal-weight patients, and lines of treatment, confirming GLM efficacy and safety in a long-term real-life setting.