TNF Inhibitor Therapy for Psoriatic Arthritis

As the beneficial effects of tumor necrosis factor (TNF) inhibitor therapy in rheumatoid arthritis (RA) became clear in the late 1990s, it was only a matter of time before interest grew in testing these agents in psoriatic arthritis (PsA), a disease widely perceived to have similar manifestations and pathophysiology. These agents proved to be effective in PsA as well and, despite increasing evidence of the differences between PsA and RA, they have become an important part of the treatment algorithm for both diseases. Differences remain, however, including apparent differences in the value of concomitant methotrexate therapy. There are five approved TNF inhibitors in the United States and all now have a labeled indication for PsA.

All agents in this class produce their effects by inhibiting extracellular signaling with the pro-inflammatory cytokine TNF-alpha. TNF-alpha functions early in the inflammatory cascade, so that inhibition of this protein may have important downstream effects on other pro-inflammatory cytokines, IL-1, IL-6 and IL-8. Inhibition of these cytokines also reduces surface expression of cellular adhesion molecules, reducing migration of activated lymphocytes into synovial tissue in PsA, and limiting the bone and cartilage damage by proteases released by these inflammatory cells.

TNF-alpha is produced by a variety of cells in the psoriatic synovium, but primarily by activated macrophages (Figure 7-1). Soluble TNF-alpha, in turn, binds to receptors on effector cells, such as synovial fibroblasts and osteoclasts; when activated, these cells drive the cartilage and bone destruction that can occur in PsA. By blocking TNF signaling, TNF inhibitors may block activation of target cells. All available TNF inhibitors bind soluble TNF-alpha and prevent it from binding to cell-surface TNF receptors, which exist in two variations, p55 and p75. The monoclonal antibody compounds may also bind TNF expressed on the cell surface, although the clinical importance of this distinction is unclear.

There are five TNF inhibitors currently approved by the FDA for use in the United States:

• Etanercept
• Infliximab
• Adalimumab
• Golimumab
• Certolizumab pegol.

All have a labeled indication for PsA.

TNF inhibitors are clearly more expensive than nonbiologic agents used to treat PsA. This fact, along with safety concerns, impacts any treatment decisions in the clinic. In most cases, the decision to treat with a TNF inhibitor is dependent both on response to prior therapies and the particular disease manifestations in the patient being managed. Group for the Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) and European League Against Rheumatism (EULAR) have both proposed recommendations for management of PsA that identify situations in which TNF inhibitor therapy is appropriate (Figure 6-2). For milder disease, including nondestructive arthritis, enthesitis and axial disease, treatment with an nonsteroidal anti-inflammatory drug (NSAID) may be sufficient. The presence of destructive or deforming arthritis warrants more aggressive therapy with either a nonbiologic disease-modifying antirheumatic drug (DMARD) or a TNF inhibitor, although there is far less evidence for the prevention of structural damage with conventional DMARDs. In patients with enthesitis or axial disease refractory to NSAID therapy, there is little evidence for benefit with nonbiologic DMARDs, so that TNF inhibitor therapy may be warranted in this situation.

The efficacy of the available TNF inhibitors in PsA appears to be comparable, although there are little comparative data to inform treatment decisions. In one prospective controlled trial, 100 patients at a single center in Italy who had experienced inadequate response to DMARD therapy were randomly assigned to open-label treatment with infliximab 5 mg/kg every 6 to 8 weeks, etanercept 25 mg twice-weekly or adalimumab 40 mg every other week. The response to all three agents was comparable, with 75%, 72% and 70% achieving an ACR20 response at 1 year with infliximab, adalimumab and etanercept, respectively. Patients treated with infliximab and adalimumab were noted to have the greatest improvement in psoriasis area and severity index (PASI) scores in this study.

When patients do not respond adequately to a TNF inhibitor, there is little evidence to suggest that increasing the dose may be beneficial. As previously noted, for joint symptoms at least, etanercept 50 mg twice weekly was no more effective than once weekly dosing in the PRESTA trial. In the ADEPT trial with adalimumab, patients in the open-label extension were able to increase their dose frequency to 40 mg weekly; 18.9% chose to do so, although the impact of this change on their disease activity is unclear from the publication. The addition of methotrexate or another DMARD may also be considered for patients with an inadequate response, although as previously noted, there is little evidence to suggest that methotrexate provides synergistic efficacy in PsA as opposed to RA. An unblinded trial suggested that cyclosporine could be safely used in combination with adalimumab and that the combination was more effective than either drug alone.

Persistence on TNF inhibitor therapy appears to be good, even outside of clinical trials. Data from the BSRBR found that over 75% of patients with PsA remained on their first TNF inhibitor over more than 2 years of follow-up. In patients who do not respond to a TNF inhibitor or in whom this response wanes, there are observational data to suggest that a switch to another agent in the class can be successful. In the largest of these observational studies, the clinical response to adalimumab over 12 weeks was similar in patients receiving adalimumab as their first TNF inhibitor and those switching from etanercept or infliximab. Treatment for patients with an inadequate response to multiple TNF inhibitors is much more challenging.

Enlarge 

Enlarge 

Etanercept

Etanercept, the first TNF inhibitor studied in psoriatic arthritis (PsA), is a fusion protein combining the extracellular portions of two recombinant p75 TNF receptors with the Fc portion of human IgG1. The resulting construct, which is produced commercially in a Chinese hamster ovary cell line, is fully human, although the hinge region is not naturally occurring and potentially antigenic.

The first evidence of the efficacy of etanercept in PsA came in a pilot study published in 2000. In this 12-week study, 60 patients with PsA were randomized to receive subcutaneous etanercept 25 mg twice weekly or placebo. Those receiving methotrexate or NSAIDs remained on stable doses of those drugs. The primary end point was the proportion of patients who met the psoriatic arthritis response criteria (PsARC), first developed for the PsA trials with sulfasalazine. Additional end points included the proportion of patients achieving ACR20, 50 and 70 responses. Skin response was assessed as the proportion of patients achieving 75% improvement in the psoriasis area and severity index (PASI75). At 12 weeks, 87% of the patients in the study had met a PsARC response vs. 23% of the placebo patients, a significant difference. An ACR20 was achieved by 73% of the etanercept patients vs. 13% of the placebo patients, also significant. In the patients with psoriasis involvement judged sufficient for assessment of response (≥ 3% body surface area), the difference between the etanercept-treated and placebo patients was also significant. Radiographic end points were not assessed in this study. Importantly, the side effect profile of etanercept in this trial (discussed in more detail below) was not different from that seen in rheumatoid arthritis (RA) trials.

One of the notable aspects of this first etanercept trial was the nature of the PsA patients enrolled, which has been mirrored in most other trials of TNF inhibitors in PsA since then. Patients were predominantly polyarticular, with a mean of approximately 20 tender and 14 swollen joints. In essence, this was a very “rheumatoid-like” population, which does not obviate the results of the study but does raise some questions about the generalizability of these results to patients with oligoarticular or purely axial disease. For the latter group at least, there are now published studies of etanercept and other TNF inhibitors in ankylosing spondylitis (AS) that confirm the efficacy of these agents in axial spondyloarthropathies.

Following this pilot trial, etanercept was studied in a much larger, multicenter trial with 205 patients. Again, patients were randomized to subcutaneous etanercept 25 mg twice weekly or placebo for 24 weeks. This trial, and all subsequent trials of TNF inhibitors in PsA, used the ACR response as the primary clinical end point rather than the PsARC. At the primary clinical end point of 12 weeks, 59% of the etanercept-treated patients achieved an ACR20 response vs. 15% of the placebo patients. Corresponding levels for ACR50 and ACR70 were 38% and 11% vs. 4% and 0% for placebo (Figure 7-2). These results were sustained through 24 weeks of therapy (Figure 7-3). As in the pilot study, patients taking methotrexate therapy at baseline were allowed to remain on this drug during the trial, and approximately 40% of the patients in both arms did so, at a mean dose of about 16 mg per week; the ACR20 response rate with etanercept for patients taking concomitant methotrexate was about the same as for those who were not (Figure 7-4).

This multicenter etanercept study was the first to examine the structural benefit of a TNF inhibitor in PsA. Using a modification of the Sharp radiographic scoring system that included both distal interphalangeal (DIP) joints and the first carpal metacarpal (CMC) joint (Figure 5-10), plain radiographs were scored at baseline, 24 weeks and 48 weeks. Patients remained on blinded therapy until the last patient enrolled had completed 24 weeks, at which point all were eligible to enroll in a 48-week open-label extension. Mean total Sharp score (mTSS) progression for the placebo group in this trial was small relative to what has been seen in RA (approximately 1 unit over 48 weeks), but this progression was clearly greater than in the etanercept group, where there was no mean total Sharp progression at both 24 and 48 weeks (Figure 7-5). As in RA, the impact of etanercept on progression was seen in both erosion and joint space narrowing scores.

Interestingly, and importantly, when patients in this trial transitioned to the open-label extension, a similar impact on radiographic progression was seen during this portion of the trial as well. There was no further mean radiographic progression for those patients in the initial placebo group who received etanercept for an additional 48 weeks (Figure 7-6).

As a result of these trials, etanercept is approved for PsA in the United States. The 25-mg twice-weekly dose has been largely replaced by the 50-mg prefilled syringe and auto-injectors, which are equally effective. Higher doses of etanercept do not appear to offer any benefit in PsA, at least for joint and tendon manifestations. The Psoriasis Randomized Etanercept STudy in Subjects with Psoriatic Arthritis (PRESTA) was a blinded study of etanercept 50 mg weekly vs. 50 mg twice weekly in 752 patients with psoriasis and PsA. At 12 weeks, there was no difference between the two dose regimens in ACR response or enthesitis, although the improvement in skin disease was greater with the higher dose.

Etanercept may be used with or without concomitant methotrexate. In the SEAM-PsA trial, PsA patients naïve to DMARD therapy were randomized to receive blinded therapy with etanercept, methotrexate or the combination of both drugs. The primary endpoint was the ACR20 response at 24 weeks, achieved by 60.9% of the patients in the etanercept arm vs. 50.7% of those in the methotrexate arm, which was statistically different. The key secondary endpoint of minimal disease activity was achieved by 35.9% in the etanercept arm and 22.9% in the methotrexate arm, also a statistically significant difference. Response rates in the combination arm, however, were not different from those with etanercept monotherapy. This differs from RA, where combination therapy with methotrexate has been shown to be more effective than TNF inhibitor monotherapy. There are no controlled data on the benefit of concomitant methotrexate with the other available TNF inhibitors in PsA.

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Infliximab

Infliximab, the mouse-human chimeric anti-TNF monoclonal antibody, was approved by the FDA for the treatment of PsA based on two pivotal trials known as the Infliximab Multinational Psoriatic Arthritis Controlled Trial (IMPACT) and IMPACT 2. These trials were launched after a number of small series suggested benefit with infliximab in PsA. Both IMPACT and IMPACT 2 used a dose of infliximab of 5 mg/kg given every 8 weeks after initial loading doses at 0, 2 and 6 weeks. This dose was chosen rather than the 3 mg/kg starting dose used in rheumatoid arthritis (RA) as it was felt that PsA, a spondyloarthropathy, might respond better to the dosing used in irritable bowel disease (IBD), which shares some genetic and possible pathophysiologic characteristics.

The IMPACT trial enrolled 104 patients with psoriatic arthritis (PsA) who had previously had an inadequate response to one or more DMARDs. Disease activity and extent were similar to the etanercept trials, with a mean of 20 to 24 tender joints and 15 swollen joints. Patients were allowed, but not required, to continue concomitant therapy with one DMARD during the trial. All patients received blinded therapy with either infliximab or placebo at 0, 2, 6 and 14 weeks. At week 16, the primary end point of the trial, all patients were crossed over to open-label infliximab therapy through 50 weeks. At week 16, 65%, 46% and 29% of the infliximab-treated patients achieved an ACR20, 50 and 70, respectively. Only 10% of the placebo patients had achieved an ACR20, and none reached the higher levels of response. In patients with evaluable skin disease (psoriasis area and severity index [PASI] ≥ 2.5 at baseline), 68% had a 75% improvement in PASI score at 16 weeks with infliximab vs. none with placebo. Skin and joint responses were sustained through week 50, and the original placebo arm had caught up to the active treatment arm by that point. Radiographic benefit was observed with infliximab in this trial when compared with estimated rate of progression at study entry, although the assessment of benefit relative to placebo was limited by the short placebo treatment phase of the trial.

The second trial, IMPACT 2, had a similar design, except that crossover to active therapy for the placebo arm took place at 24 weeks, although patients with active disease could enter an early escape at 16 weeks. At week 24, the primary end point, the ACR20, 50 and 70 response rates for infliximab were 54%, 41% and 27%, compared with 16%, 4% and 2% for placebo (Table 7-1). Patients entering early escape at Week 16 were considered nonresponders at week 24. As in the IMPACT trial, there was a dramatic improvement in skin disease with infliximab, with 60.2% of evaluable patients achieving a PASI75 by week 24 compared with just 1.1% of the placebo patients. Also, as in the IMPACT trial, the skin and joint response was sustained through 50 weeks.

The IMPACT 2 trial assessed radiographic progression in addition to clinical response. At week 24, there was a statistically significant difference in modified Total Sharp Score (mTSS) progression between the infliximab and placebo arms, with a magnitude that was similar to the difference in progression seen with etanercept (-0.70 units for infliximab vs. 0.82 units for placebo). Interestingly, the IMPACT 2 publication includes a probability plot of radiographic progression, which makes it clear that the differences between groups derive from just 20% of the patients who had much more progression in the placebo arm (Figure 7-7).

In addition to standard joint assessments, the infliximab trials also assessed the unique PsA clinical manifestations of dactylitis and enthesitis. In both IMPACT and IMPACT 2, the baseline prevalence of both dactylitis and enthesitis was similar in both treatment arms, yet there were significantly fewer infliximab-treated patients with each finding at the primary clinical end point (Figure 7-8 and Figure 7-9). IMPACT 2 looked as well at function and health-related quality of life, showing clinically meaningful improvement in Health Assessment Questionnaire (HAQ) and Short Form-36 (SF36) scores with infliximab relative to placebo. Finally, measurements of work productivity and employment in this trial demonstrated that infliximab treatment led to increased employment, improved productivity and fewer work days lost by week 14. Nail changes were not assessed systematically in the IMPACT studies, but an open-label trial of infliximab in 18 patients with psoriasis or PsA and nail involvement demonstrated marked improvement over 38 weeks of therapy.

Both IMPACT and IMPACT 2 enrolled patients with established PsA (11 years of disease in IMPACT and 8 years in IMPACT 2). A trial examined the effects of infliximab in a population of patients with earlier disease, naïve to methotrexate. In the open-label RESPOND trial, 115 patients with a mean duration of PsA of just 3 years were randomized to receive either methotrexate 15 mg weekly or methotrexate combined with infliximab 5 mg/kg given at 0, 2, 6 and 14 weeks. At the primary end point of 16 weeks, an ACR20, 50 and 70 response was achieved by 86%, 73% and 49% of the combination therapy patients, compared with 67%, 40% and 19% of those receiving methotrexate alone (Figure 7-10). A PASI75 was achieved with combination therapy by 97% of those with evaluable skin disease compared with 54% of those receiving methotrexate monotherapy. Rates of remission and improvement in dactylitis were also significantly greater in the combination arm. The open-label design of this trial presumably contributed to the very high response rates seen with infliximab therapy, rates that are much higher than those that have been seen in other trials of TNF inhibitors in PsA. Unfortunately, the design of this trial did not include a monotherapy infliximab arm, so there is no way to know whether the addition of methotrexate led to greater improvement than would have been achieved with infliximab alone.

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Adalimumab

Adalimumab is a fully human monoclonal antibody developed using phage display technology, which targets TNF. This technology uses a library of bacteriophages expressing DNA for a variety of heavy and light immunoglobulin chains. Chains with the highest affinity for a specific antigen, in this case TNF, are selected from this library, then recombined to produce a fully human, albeit non-native, antibody directed against TNF. Initially approved for the treatment of rheumatoid arthritis (RA), trials in psoriatic arthritis (PsA) and other spondyloarthropathies followed quickly.

The key pivotal trial in PsA, the Adalimumab Effectiveness in Psoriatic Arthritis Trial (ADEPT), was published in 2005. This trial enrolled 313 patients with active PsA, who were randomized to treatment with adalimumab 40 mg subcutaneously every other week or placebo for 24 weeks. The primary end point was the ACR20 response rate at 12 weeks. Results of this trial were remarkably similar to those of the pivotal etanercept and infliximab trials; at week 12, the ACR20, 50 and 70 response rates with adalimumab were 58%, 36% and 20% compared with 14%, 4% and 1% with placebo. This response was sustained through 24 weeks of blinded therapy (Figure 7-11). Fifty percent of the patients in this trial were taking methotrexate at entry. Methotrexate was continued through the course of the trial, and the clinical response was the same in patients with or without concomitant methotrexate use (Figure 7-12).

This study enrolled patients with established (mean 9.5 years) and active (mean 25 tender and 14 swollen joints) PsA. Improvement in psoriatic skin disease was assessed in patients with > 3% body surface area (BSA) of psoriasis at baseline. In this group, 59% achieved a psoriasis area and severity index (PASI)75 at 24 weeks with adalimumab compared with 1% with placebo. Enthesitis and dactylitis were improved with adalimumab relative to placebo in this trial, although the differences did not reach statistical significance. Function, measured by the Health Assessment Questionnaire (HAQ), and quality of life, measured by the Short Form-36 (SF36) and the dermatology life quality index (DLQI), were also significantly improved with adalimumab relative to placebo. Fatigue was also assessed in this trial and was significantly improved with adalimumab compared with placebo at 24 weeks.

The impact of adalimumab on radiographic progression in this trial was quite similar to that of other TNF inhibitors in PsA. At 24 weeks, the modified Total Sharp Score (mTSS) progression was -0.2 units with adalimumab compared with 1.0 unit with placebo. A probability plot showed that similar to the data from the IMPACT 2 trial, this difference was driven by the 20% of placebo patients who had progressive radiographic damage (Figure 7-13). The impact of adalimumab on structural progression was similar for patients taking concomitant methotrexate and those who were not. Multiple sub-analyses did not identify any particular subgroup for which adalimumab provided greater structural benefit.

Patients in the ADEPT trial were eligible for open-label therapy with adalimumab after 24 weeks. Clinical and radiographic benefits, including quality of life measures, were sustained through 2 years of therapy. Persistence was good, with only 16% (44 of 285) of patients who entered the open-label study discontinuing therapy over 2 years.

Adalimumab has been studied in a second large trial in PsA, specifically in patients with concurrent or prior DMARD use. This trial enrolled 100 patients who were randomized to 12 weeks of adalimumab 40 mg every other week or placebo. DMARDs being taken at study entry were continued, except for cyclosporine or tacrolimus. The ACR20 response rate was 39% at 12 weeks compared with 16% in the placebo arm (Figure 7-14). Psoriatic skin disease was also improved with adalimumab compared with placebo. Fatigue, quality of life (SF36), enthesitis, and dactylitis were all numerically, but not statistically, improved with adalimumab at 12 weeks.

The arthritis response to adalimumab in this trial was lower than has been seen with other trials of TNF inhibitors in PsA, although the ACR20 response rate improved to 65% with an additional 12 weeks of open-label therapy. Additional analysis of the demographic and clinical data from this study did not provide an explanation for the apparent delayed response. An analysis of data on joint disease from three trials of adalimumab in psoriatic skin disease also confirmed the efficacy of this agent in PsA.

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Golimumab

Golimumab is a fully human monoclonal antibody that targets TNF. It has been approved for use in psoriatic arthritis (PsA) in the United States based on the Golimumab-Randomized Evaluation of Safety and Efficacy in Subjects With Psoriatic Arthritis Using a Human Anti-TNF Monoclonal Antibody (GO-REVEAL) trial. This trial included 405 patients with active psoriasis and PsA randomized to receive blinded subcutaneous injections with placebo, golimumab 50 mg, or golimumab 100 mg every 4 weeks for 6 months. The primary end point for this trial was at 14 weeks, by which point 51% of the patients receiving golimumab 50 mg and 45% of those receiving golimumab 100 mg had achieved an ACR20 response compared with 9% in the placebo arm (Figure 7-15). At 24 weeks, the ACR50 and 70 response rates for golimumab 50 mg, the commercially approved dose, were 32% and 19%, compared with 4% and 1% for placebo ± methotrexate.

As with other trials of TNF inhibitors in PsA, concurrent use of methotrexate (48% of patients) did not affect response rates in this study. Neutralizing antibodies to golimumab were observed in approximately 5% of the patients in this study; only 1 of 19 patients with antibodies was taking concomitant methotrexate.

Dactylitis was numerically, but not statistically, improved with golimumab relative to placebo. Improvement in enthesitis, however, was statistically greater for both doses of golimumab at 14 and 24 weeks when assessed with the Maastricht Ankylosing Spondylitis Entheses Score (MASES) enthesitis index. Physical function, measured by Health Assessment Questionnaire (HAQ), was also significantly improved for both doses of golimumab at 24 weeks.

Golimumab was effective for skin disease in this trial. At week 14, 40% of the patients with > 3% body surface area (BSA) psoriasis at baseline who received 50 mg had achieved a PASI75 compared with 3% of the placebo patients. As with joints, this response was independent of methotrexate use. This trial was the first large trial with a TNF inhibitor to systematically measure improvements in nail disease. PGA of nail disease and assessment of a target nail using the nail psoriasis severity index (NAPSI) were significantly improved with both doses of golimumab at 14 and 24 weeks.

The GO-REVEAL trial demonstrated structural benefit with golimumab. At 24 weeks, the modified Total Sharp Score (mTSS) progression was -0.16 for golimumab 50 mg and 0.27 for placebo. Interestingly, unlike other trials with TNF inhibitors, concomitant use of methotrexate appeared to produce greater protection from structural progression in this study. Both radiographic and clinical benefit were sustained through 1 year as patients transitioned to open-label therapy after 24 weeks.

The intravenously administered preparation of golimumab, marketed as Simponi Aria by Janssen, has also been approved for the treatment of PsA, based on the 24-week GO-VIBRANT study. In this study, 480 PsA patients were randomized to golimumab 2 mg/kg, administered at 0, 2, 4, 12 and 20 weeks, or placebo. The primary endpoint of the study was an ACR20 response at week 14, achieved by 75% of the golimumab group compared with 22% of the placebo group. Corresponding ACR50 and ACR70 responses also favored golimumab (44% and 25%, compared with 6% and 2%). This trial also looked at the more stringent composite response of minimal disease activity (MDA), met by 27% of the golimumab group at 14 weeks, compared with 4% of the placebo group. PASI75 was achieved by 59% of the golimumab treated patients at 14 weeks, compared with 14% of placebo-treated patients. Golimumab also demonstrated reduced radiographic progression relative to placebo in this study.

Enlarge 

Certolizumab pegol is a TNF inhibitor composed of an Fab fragment from a humanized anti-TNF monoclonal antibody fused to a polyethylene glycol (PEG) moiety. The smaller Fab fragment can be produced using an E coli system rather than a more complex mammalian cell culture, while the PEG confers greater stability and a longer half-life. The results from the RAPID-PsA trial were published, and certolizumab pegol is now approved by the FDA for this indication.

The RAPID-PsA trial enrolled 409 patients with active psoriatic arthritis (PsA) who had failed at least one prior DMARD; prior use of another TNF inhibitor was not an exclusion, and approximately 19.5% of the patients in the trial had previously been treated with another such agent. Participants were randomized to receive placebo or certolizumab pegol (400 mg at weeks 0, 2 and 4, followed by one of two dose schedules: 200 mg every 2 weeks or 400 mg every 4 weeks). The primary end point was the ACR20 response at 12 weeks. Blinded therapy was continued through 24 weeks, although nonresponders at 14 and 16 weeks were able to escape to active treatment.

Response to certolizumab pegol in the RAPID-PsA trial was comparable to other TNF inhibitors in PsA, with 58% of the certolizumab pegol 200 mg every-2-weeks group and 51.9% of the certolizumab pegol 400 mg every-4-weeks group achieving an ACR20 at 12 weeks compared with 24.3% in the placebo arm (Figure 7-16). Investigators noted that ACR20 response was statistically greater than placebo as early as 1 week after the first dose and was sustained through 24 weeks. In patients with baseline dactylitis or enthesitis, treatment with certolizumab pegol resulted in statistically greater improvement compared with the placebo group. No sub-analysis of clinical response according to background methotrexate use is yet available.

Certolizumab pegol was very effective for skin disease in this trial. At 24 weeks, the psoriasis area and severity index 75 (PASI75) response rate in subjects with > 3% body surface area (BSA) psoriasis averaged 61.4% for the two certolizumab pegol treatment groups compared with 15.1% in the placebo group. Function was also improved with certolizumab pegol; the reduction in Health Assessment Questionnaire Disability Index (HAQ-DI) was 0.50 at 24 weeks for the combined certolizumab pegol groups compared with 0.19 for placebo. Radiographic data were captured in this study but have not yet been reported.

The safety profile in the RAPID-PsA trial was reported to be similar to that seen with certolizumab pegol in rheumatoid arthritis (RA), the approved indication for this agent. Serious adverse events occurred in 4% of the placebo group vs. 7% of the combined certolizumab pegol groups. There was one death in each of the certolizumab pegol treatment arms, one due to a myocardial infarction, and one due to sudden death of unknown cause; no deaths were reported in the placebo arm.

The results of the RAPID-PsA study show that treatment with certolizumab pegol is effective in patients with PsA with rapid improvement in clinical signs and symptoms of PsA, including arthritis, enthesitis, dactylitis and skin involvement.

Enlarge 

Safety

The overarching theme of the safety data from clinical trials of TNF inhibitors in psoriatic arthritis (PsA) is that the safety concerns are the same as those that have been well described in rheumatoid arthritis (RA). An analysis of data from 596 PsA patients in the British Society for Rheumatology Biologics Register (BSRBR) found that the safety profile with these agents was not substantially different from that with nonbiologic DMARDs, and that the safety in PsA patients was similar to that seen in serologically negative RA patients.

Infections

The primary concern is infection, including both bacterial and opportunistic or atypical infections. Bacterial infections are seen in patients taking TNF inhibitors, although it is not entirely clear that they are seen more commonly than in patients taking nonbiologic DMARDs or with active, persistent inflammatory disease. Serious bacterial infections requiring hospitalization or intravenous antibiotics may occur and appear to be more prevalent during the first 6 to 12 months of TNF inhibitor therapy, likely because patients at greatest risk may develop this complication early, then discontinue therapy. Patients with active or chronic infections should not be treated with TNF inhibitors; there are little data on the safety of continued use of these agents in patients who develop an infection while on therapy, although anecdotal evidence suggests that they can safely be restarted in many cases after the infection has resolved.

Opportunistic infections, including fungal and mycobacterial infections, have been seen with all TNF inhibitors and all agents carry a black box warning on their labeling about these infections. There is some evidence that these infections are more common in patients treated with infliximab, perhaps because of the unique pharmacokinetics of this intravenously administered agent.

Tuberculosis

Prescreening for mycobacterium tuberculosis (mTB) exposure with a tuberculin skin test or an interferon gamma release assay blood test is recommended for all agents in the class, and has been shown to reduce the risk of reactivation of latent mTB. In the United States, atypical mycobacterial infections, while rare, appear to have become a more common complication of TNF inhibitor therapy than mTB infection, presumably because there are no screening tests for such infections. The risk of infection should be discussed with all patients initiating therapy with a TNF inhibitor, and they should be cautioned to alert other health care providers about their use of these agents so that unusual infections can be considered if they become ill.

The value of repeat testing for TB exposure in patients being treated with TNF inhibitors remains unclear. There are no consensus recommendations for annual screening, although some clinicians choose to do so, particularly in situations where there is a high risk of de novo exposure.

Hepatitis B

Hepatitis B reactivation has been reported with TNF inhibitor therapy. Prescreening is recommended, and these agents should not be used in patients with chronic hepatitis B. There is evidence that they may be used with caution in patients with hepatitis C.

Herpes Zoster

Herpes zoster has been shown to occur more frequently in patients receiving TNF inhibitors. Unfortunately, current recommendations preclude the administration of the herpes zoster vaccine, a live vaccine, to patients receiving biologic therapy, so it would be reasonable to consider giving this vaccine prior to initiating treatment. In general, there is no evidence that TNF inhibitors impact either the safety or the efficacy of other vaccines, although live vaccines should be avoided.

Lymphoma

TNF inhibitors carry a labeled warning concerning the risk of lymphoma, based on cases of lymphoma seen with therapy in early rheumatoid arthritis (RA) trials. Subsequent data has suggested that this risk is more closely related to disease activity in RA than to the treatment. TNF inhibitors also carry labeled warnings about the risk of hematologic and solid malignancies in children and young adults. The level of increased risk is clouded by the uncertain risk of malignancy in patients with juvenile inflammatory arthritis likely to receive these therapies; some data have suggested that the risk of malignancy is increased in this population. Skin cancers, including squamous cell and basal cell carcinomas, appear to be increased in patients treated with TNF inhibitors. Patients with a history of such cancers should be advised to undergo regular skin checks while on therapy. This risk may be even greater in psoriasis patients who have a received phototherapy or psoralen plus UVA (PUVA) therapy. There is some evidence that the risk of melanoma is increased with TNF inhibitor therapy, although this has not been consistently shown in all trials.

Demyelinating Disease and Congestive Heart Failure

Two other serious adverse events of note are demyelinating disease and congestive heart failure (CHF). Both appear to be exacerbated by TNF inhibition and these agents are contraindicated in patients with a history of multiple sclerosis (MS), optic neuritis or uncontrolled CHF. Some reports have suggested that MS may develop in patients treated with TNF inhibitors without a prior history of this disease, although this is rare.

Paradoxical development of flares of skin psoriasis has been reported with TNF inhibitor therapy, both in patients with a history of psoriasis and in others, such as those with rheumatoid arthritis (RA), without this history. Flares of psoriasis may sometimes be managed by the addition of methotrexate, but they frequently require discontinuation of the inciting agent. Switching to an alternate TNF inhibitor does not always lead to a recurrence of the psoriasis.

Injection Site Reactions

In terms of frequency, the most common adverse events specifically attributable to TNF inhibitors are likely to be injection site and infusion reactions. Injection site reactions are relatively common, although serious reactions have not been reported, and these reactions generally fade with continued therapy. Stinging or burning sensations with injections are common. There is some evidence that this is less of a problem with golimumab and certolizumab pegol than with etanercept and adalimumab.

Lab Abnormalities

Laboratory abnormalities, including cytopenias and elevated liver enzymes, have been reported with TNF inhibitor therapy, but these events are not frequent enough to warrant specific recommendations for regular laboratory monitoring.

Pregnancy

All TNF inhibitors are labeled as pregnancy Class B, meaning that animal studies have not shown any risk to the fetus but there are no well-controlled human studies to confirm this. They may be continued in pregnancy if the benefit in terms of controlling disease activity is concerned to outweigh any potential risk to the fetus. The Organization of Teratology Information Specialists (OTIS) maintains a registry of women treated with biologic or immunosuppressive medication during therapy. Women continuing TNF inhibitors should be encouraged to participate in this registry.

General Management

TNF inhibitors are clearly more expensive than nonbiologic agents used to treat PsA. This fact, along with safety concerns, impacts any treatment decisions in the clinic. In most cases, the decision to treat with a TNF inhibitor is dependent both on response to prior therapies and the particular disease manifestations in the patient being managed. The current (2021) Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) recommendations for management of PsA identify situations in which TNF inhibitor therapy is appropriate, and are presented in and summarized in Figure 6-11. For milder disease, including nondestructive arthritis, enthesitis, and axial disease, treatment with an NSAID may be sufficient. The presence of destructive or deforming arthritis warrants more aggressive therapy with either a nonbiologic DMARD or a TNF inhibitor, although there is far less evidence for the prevention of structural damage with conventional DMARDs. In patients with enthesitis or axial disease refractory to NSAID therapy, there is little evidence for benefit with nonbiologic DMARDs, so that TNF inhibitor therapy may be warranted in this situation.

The efficacy of the available TNF inhibitors in PsA appears to be comparable, although there are little comparative data to inform treatment decisions. In one prospective controlled trial, 100 patients at a single center in Italy who had experienced inadequate response to DMARD therapy were randomly assigned to open-label treatment with infliximab 5 mg/kg every 6 to 8 weeks, etanercept 25 mg twice-weekly, or adalimumab 40 mg every other week. The response to all three agents was comparable, with 75%, 72%, and 70% achieving an ACR20 response at 1 year with infliximab, adalimumab, and etanercept, respectively. Patients treated with infliximab and adalimumab were noted to have the greatest improvement in PASI scores in this study.

When patients do not respond adequately to a TNF inhibitor, there is little evidence to suggest that increasing the dose may be beneficial. As previously noted, for joint symptoms at least, etanercept 50 mg twice weekly was no more effective than once weekly dosing in the PRESTA trial. In the ADEPT trial with adalimumab, patients in the open-label extension were able to increase their dose frequency to 40 mg weekly; 18.9% chose to do so, although the impact of this change on their disease activity is unclear from the publication.¹⁵ The addition of methotrexate or another DMARD may also be considered for patients with an inadequate response, although as previously noted, there is little evidence to suggest that methotrexate provides synergistic efficacy in PsA as opposed to RA. An unblinded trial suggested that cyclosporine could be safely used in combination with adalimumab and that the combination was more effective than either drug alone.

Persistence on TNF inhibitor therapy appears to be good, even outside of clinical trials. Data from the BSRBR found that over 75% of patients with PsA remained on their first TNF inhibitor over more than 2 years of follow-up.²⁶ In patients who do not respond to a TNF inhibitor or in whom this response wanes, there are observational data to suggest that a switch to another agent in the class can be successful. In the largest of these observational studies, the clinical response to adalimumab over 12 weeks was similar in patients receiving adalimumab as their first TNF inhibitor and those switching from etanercept or infliximab. Treatment for patients with an inadequate response to multiple TNF inhibitors is much more challenging.

Enlarge