Expert recommendations for the use of apremilast in psoriatic arthritis (cid:2)

Introduction and objectives: Despite the evidence, there are doubts about the positioning of apremilast in the psoriatic arthritis (PsA) treatment algorithm. The objective of this project was to collect the scientiﬁc evidence and the experience of a group of rheumatologists who are experts in the management of PsA with apremilast in clinical practice in Spain


Introduction
Psoriatic arthritis (PsA) is a chronic inflammatory disease associated with a significant decline in function and quality of life 1 .Patients with PsA are at increased risk of obesity, insulin resistance, type 2 diabetes mellitus, hypertension, hyperlipidaemia, and cardiovascular disease 2 .Patterns of involvement in PsA have several features, including the presence of dactylitis and enthesitis 1 .
A diagnosis of PsA is an indication for treatment with systemic therapies 3 .Current therapeutic options for PsA and its complications include 4,5 non-steroidal anti-inflammatory drugs, corticosteroids, conventional disease-modifying antirheumatic drugs (DMARDs) (cDMARDs), and biologic drugs 6 .New drugs with new, more specific mechanisms of action against PsA have been developed in recent years, including synthetic disease-modifying antirheumatic drugs with a specific target, such as apremilast 7 .
Apremilast is an oral phosphodiesterase-4 inhibitor approved to treat PsA, alone or in combination with other DMARDs, in adult patients who have had an inadequate response or intolerance to previous treatment with a DMARD, and to treat moderate to severe chronic plaque psoriasis 8 .In contrast to non-specific immunosuppressive drugs, apremilast has an immunomodulatory action 9 , reducing the risk of severe opportunistic infections and reactivation of tuberculosis 10 .
Although the evidence from clinical studies is strong and consistent, there remain doubts about the positioning of apremilast within the therapeutic algorithm for PsA, as these studies are relatively recent and no direct comparative clinical trials are available 4,11 .Therefore, information from routine clinical practice is very important to draw strategic conclusions 12,13 .The aim of this project was to gather scientific evidence and the experience of a group of rheumatologists with expertise in the management of PsA on the use of apremilast in clinical practice in Spain using the Delphi methodology, generating a series of recommendations on the management of PsA with an eminently practical approach aimed at rheumatologists.

Materials and methods
The study was conducted in 4 phases using the Delphi technique 14 modified by Rand/UCLA recommendations 15 : 1) identification of clinical scenarios of interest by the Scientific Committee overseeing the management of the project; 2) development of the questionnaire by a panel of rheumatology experts; 3) e-mail survey in 2 rounds; and 4) collection and final analysis of the results.
1) The Scientific Committee comprised a national coordinator and 5 regional coordinators 11,[16][17][18][19] .Each regional member proposed several clinical scenarios where, due to the lack of consensus, it would be interesting to make therapeutic recommendations on the use of apremilast in PsA.After analysing and discussing the proposals, the following scenarios were chosen: 1) efficacy in peripheral PsA; 2) efficacy in enthesitis and dactylitis; 3) efficacy in PsA with skin involvement; 4) management of comorbidities in patients with PsA; and 5) safety implications in the use of apremilast.
2) The expert panel for this study was made up of a total of 17 rheumatologists with expertise in the treatment of PsA from different regions of Spain, including the 5 regional coordinators (Appendix A).This panel of experts attended a round table where different cases of apremilast use in routine clinical practice in the 5 proposed scenarios were discussed.This discussion, together with the previous literature review, was the basis for the questionnaire that had 50 questions and a total of 156 items to be assessed.All questions, in the form of statements (affirmative or negative), were designed to be answered on a Likert scale, where 1 meant strongly disagree with the statement and 9 meant strongly agree with the statement (Fig. 1).3) This questionnaire was sent to the expert panel in electronic format and answered anonymously.After analysing the results of the first round of voting, the items that had not reached consensus in the first round were submitted again to the expert panel for voting.The questionnaire of the second round included the medians of the scores of the first round for each question.4) Following the Rand/UCLA methodology 15 , each item's level of appropriateness was ranked according to the median and the mean of the absolute deviation from the median.Also, following the definitions of the BIOMED Concerted Action on Appropriateness for different panel sizes, the level of agreement for each item was classified as "agree", "neutral" or "disagree" (Fig. 1).Agreement on an item was determined when at least 77% of the panellists scored within the predefined 3-point interval (Likert score intervals: 1-3, 4-6, 7-9) in which the median score for that item fell.Disagreement on an item was determined when the scores of one third or more of the panellists were in the Likert interval 1-3, and of another third or more in the interval 7−9.Items where there was no agreement or disagreement were considered to have an "uncertain" level of consensus.
The study was conducted over 3 months, from 30 November 2017 to 14 February 2018.

Results
The 17 experts on the panel reached consensus in the first round on 58 of the 156 items (37.2%) (46 appropriate and 12 inappropriate).Of the remaining 98 items, consensus was reached on 35 more (21 appropriate and 14 inappropriate) in the second round.

Efficacy of apremilast in peripheral psoriatic arthritis
Of the 45 items included in this section of the questionnaire, the experts reached consensus on 24: appropriate in 21 (46.7%; 17 in the first round and 4 in the second round) and inappropriate in 3 items (6.7%; one in the first round and 2 in the second round) (Table 1).
The experts reached consensus that it would be necessary to establish (new) classifications of PsA, according to site, activity, or the presence of comorbidities, to define more specific therapeutic targets (Item 1 [I-1]).Moreover, the panel of experts agreed that, based on current therapeutic options, it would be necessary to explore and develop new pharmacological targets for the treatment  of PsA (I-3) and that, in this sense, apremilast, due to its efficacy and safety profile, is a new treatment paradigm for this condition (I-4).Regarding the tools to measure disease activity, the panel agreed that the ACR20 criterion (20% improvement in the American College of Rheumatology criteria) was not an adequate parameter (I-5).Regarding treatment response rates in PsA, the panel reached consensus that the ACR20 criterion, again, was not an appropriate parameter and that, of those suggested, the only criterion appropriate for this purpose would be that of minimal disease activity (MDA) (I-6).They agreed that patient-reported satisfaction would be a particularly relevant measure when assessing the therapeutic effect of apremilast (I-8).
The experts considered apremilast an appropriate option in patients with mildly disabling disease or with moderate disease activity and no radiographic progression factors (I-15), and in patients in whom DMARDs had failed or were contraindicated, including those not medically deemed eligible due to the efficacy/safety profile of biologic therapy (I-10, I-16).In this regard, the experts considered the use of apremilast to be particularly indicated in the following situations: presence of moderate inflammation, presence of nail, scalp or palmoplantar psoriasis, presence of comorbidities, and in polymedicated patients (I-11).In terms of the therapeutic goals for apremilast in PsA, the experts agreed that remission or at least low disease activity should be achieved (I-12).
However, in terms of the response to apremilast, the panel agreed that it would be appropriate, in patients with a slow response in the first 16 weeks, to maintain treatment until week 24 and then reassess, always in agreement with the patient (I-14).

Efficacy of apremilast in enthesitis and dactylitis
The experts reached consensus on 3 (60.0%;all in the second round) of the 5 items included in this section of the questionnaire (Table 2).
The experts agreed that the presence of enthesitis is a variable that would positively influence the choice of apremilast for the treatment of patients with PsA (I-9).They also agreed that in patients with moderate PsA in whom cDMARDs had failed or were contraindicated, the presence of dactylitis would positively influ-ence the choice of apremilast as an appropriate option before using biologic therapies (I-11, I-18).

Efficacy of apremilast in psoriatic arthritis with skin involvement
The experts reached consensus on one (50.0%,first round) of the 2 items included in this section of the questionnaire (Table 3).
The experts agreed that apremilast would be an appropriate therapeutic option, before using biologic drugs, to treat PsA patients with psoriasis in difficult sites such as nails, scalp or palms and soles of the feet (I-17).

Management of comorbidities in psoriatic arthritis patients
The experts reached consensus on 28 items: appropriate in 18 (36.7%;14 in the first round and 4 in the second round) and inappropriate in 10 (20.4%, 5 in the first round and 5 in the second round) of the 49 items included in this section of the questionnaire (Table 4).
The experts agreed that, in general, before starting treatment with apremilast in patients with PsA, a conventional blood test should be performed (I-22).They also considered it appropriate to undertake early screening and appropriate periodic monitoring of cardiovascular risks in patients with PsA at least once every 5 years, or after each change of treatment (I-32, I-33).They also agreed to define apremilast as an appropriate therapeutic option, before biologic treatments, in patients with cardiometabolic comorbidities (I-23).It would also be appropriate in patients with hepatic impairment versus other cDMARDs and in the treatment of patients with a history of congestive heart failure and dyslipidaemia versus biologic therapies (I-24, I-25).In terms of the comorbidities in patients with PsA requiring special attention before starting treatment with apremilast, there was agreement (consensus in disagreement) that it would not be necessary with arterial hypertension, type 2 diabetes mellitus, dyslipidaemia, demyelinating diseases, cardiovascular accidents, neoplasms, or heart failure.Special attention should be paid to patients showing symptoms of mood disorders, especially those with depressive symptoms with or without antidepressant treatment (I-26, I-47).
Finally, they agreed that the benefit/risk ratio of drugs with potentially negative effects on the common comorbidities of PsA  should be considered before starting treatment with apremilast (I-31).

Safety implications in the use of apremilast
The experts reached consensus on 37 items: appropriate in 24 (43.6%; 14 in the first round and 10 in the second round) and inappropriate in 13 (23.6%;6 in the first round and 7 in the second round) of the 55 items included in this section of the questionnaire (Table 5).
The experts agreed that pharmacological monitoring with apremilast not being necessary is an advantage over biologic therapies (I-27).In terms of patient follow-up, they agreed that it would be appropriate to perform a complete blood test once or twice a year, but inappropriate to repeat tuberculosis screening tests annually (I-29).
The experts agreed that it is a particularly suitable option to treat polymedicated patients, needle phobic patients, and those with surgery scheduled (I-35, I-36, I-38).
The experts agreed that one of the most relevant features of apremilast was its good safety profile and that it could be considered a safer option than biological therapies, especially in patients with recurrent, chronic or active infections (including human immunodeficiency virus infection, hepatitis B virus, hepatitis C virus and tuberculosis), a recent history of malignancy (<5 years), active malignancy or immunosuppressed states (I-41).
The expert panel agreed that the most appropriate management of patients experiencing gastrointestinal adverse effects or headaches with apremilast would include considering it inappropriate to reduce fluid intake (in the case of gastrointestinal adverse effects) and administering all patients antidiarrhoeals, antiemetics and/or analgesics prophylactically at the start of treatment.They considered a slower induction phase appropriate, and temporarily discontinuing treatment in cases where the adverse effect is clinically relevant and affects the continuity of treatment with apremilast, resuming treatment once the adverse effect has subsided, and administering antidiarrhoeals, antiemetics and/or analgesics only in cases where the adverse effect is clinically relevant (I-49, I-50).

Discussion
Consensus recommendations based on expert opinion and developed using validated and rigorous methodologies produce useful guidelines, providing experts with valuable specific information.In this study, the Delphi methodology 14 was used to achieve consensus and generate recommendations for rheumatologists in the management of PsA.The expert participation was good, with a 100% response rate in the 2 rounds.
According to the experts, the appropriate patient for treatment with apremilast would have moderately active, mildly disabling PsA, including those with skin involvement in difficult sites and other comorbidities commonly associated with the disease, with an inadequate response or contraindication to cDMARDS and not yet candidates for biologic therapies.
The presence of enthesitis and dactylitis, as well as psoriasis in difficult sites, would be factors that would positively influence the use of apremilast before that of biologic therapies.In the PALACE 1-3 studies 18,20,21 , which assessed the safety and efficacy of apremilast in adult patients with active PsA despite prior treatment with cDMARDs or biologics, improvements in peripheral activity parameters characteristic of PsA were observed at weeks 16 and 24.Apremilast also demonstrated greater ACR20 responses in patients who had not previously been exposed to biologic therapy 21 .
Regarding the items relating to the use of apremilast versus cDMARDs, the consensus among experts is in line with the approved indication for the drug, which recommends the use of apremilast in patients who have had an inadequate response or in whom there are contraindications to the cDMARDs 8 .In the PALACE study, the responses observed in the group treated with apremilast were similar between the patients who were receiving and those not receiving concomitant cDMARDs, including methotrexate.In addition, a greater ACR20 response was observed at week 16 in patients previously treated with cDMARDs or biologics receiving apremilast compared to patients receiving placebo 8 .
Regarding the use of apremilast versus biologic therapies, there was consensus among experts that apremilast is an appropriate choice in patients naïve to biologic therapies.Biologic therapies have an activity ceiling of around 60% [22][23][24][25][26] , which means that up to 40% of patients with PsA, at best, do not respond to initial treatment.Furthermore, these drugs have immunosuppressive activity, which can generate problems of immunogenicity or toxicity, depending on the patient [22][23][24][25][26] .In this regard, and due to their immunomodulatory action, the risk of serious opportunistic infections and reactivation of tuberculosis in patients treated with apremilast is lower than with other immunosuppressive drugs 10 and does not require screening for tuberculosis 8 .Another advantage of apremilast over biologic therapies is that it does not require extensive drug monitoring 8 , which is consistent with the results of previous studies that concluded that there is no need for prior laboratory analysis or continuous monitoring of laboratory parameters 27 .
Regarding the management of comorbidities, the experts considered that apremilast would be an appropriate therapeutic option in patients with PsA and other comorbidities commonly associated with the disease.Furthermore, no prior screening or follow-up for most of the comorbidities commonly associated with PsA would be required, which is an advantage over cDMARDs and biologics.
Regarding the items relating to the safety of apremilast, in general, the experts' responses are in line with the results obtained in previous studies 18,20,21 , which demonstrated the efficacy of apremilast in PsA, with a very favourable safety and tolerability profile, and a very convenient dosage and form of administration 27,28 .The experts agreed that one of the most relevant features of apremilast was its good safety profile and that it could be considered a safer therapeutic option than conventional systemic treatments.The experts also agreed on the management of gastrointestinal adverse effects and headaches, 2 of the most common treatmentrelated adverse effects 8 .
In conclusion, apremilast, due to its efficacy, safety, and tolerability profile, is a new therapeutic option for patients with psoriasis and PsA, especially patients with comorbidities, including infections and/or neoplasms, and covers some hitherto unmet needs 10,29,30 .

Fig. 1 .
Fig. 1.Summary outline of the conditions used to assess the Delphi questionnaire.

Table 1
Efficacy of apremilast in peripheral psoriatic arthritis.
ACR20: 20% improvement in the criteria of the American College of Rheumatology; AMDF: Arithmetic Mean of the Desirability Function; CPDAI: Composite Psoriatic Disease Activity Index; DAPSA: Disease Activity for Psoriatic Arthritis; DAS-28: Disease Activity Score; DMARDs: Disease-modifying Anti-rheumatic Drugs; HAQ: Health Assessment Questionnaire; MAD-M: Mean absolute deviation from the median; MDA: Minimum Disease Activity criterion; PASDAS: Psoriatic Arthritis Disease Activity Score; PsA: Psoriatic Arthritis; PsAJAI: Psoriatic Arthritis Joint Activity Index; PsAQol: Psoriatic Arthritis Quality of Life; PsARC: Psoriatic Arthritis Response Criteria; RA: Rheumatoid Arthritis.

Table 2
Efficacy of apremilast in enthesitis and dactylitis.

Table 3
Efficacy of apremilast in psoriatic arthritis with skin involvement.

Table 4
Management of comorbidities in patients with psoriatic arthritis.

Table 5
Safety implications in the use of apremilast.