Safety and effectiveness of apremilast in Japanese patients with psoriatic disease: Results of a post‐marketing surveillance study

The safety and efficacy of apremilast in psoriatic disease has been demonstrated in clinical trials, including in Japanese patients. This post‐marketing surveillance study was conducted after approval of apremalast in Japan in 2016 to evaluate the safety and effectiveness of the drug in Japanese patients with plaque psoriasis (PsO) and psoriatic arthritis (PsA) in routine clinical practice. Patients (enrolled between September 1, 2017, and August 31, 2019), were observed for 12 months after apremilast treatment initiation or until discontinuation or withdrawal. Safety was assessed by evaluating adverse reactions (ARs) and serious ARs. Effectiveness measures in PsO included the proportion of patients who achieved global improvement and Physician's Global Assessment (PGA) scores of 0/1 and the change from baseline in the Dermatology Life Quality Index (DLQI) after 6 and 12 months treatment. The safety analysis set included 1063 patients (PsO, n = 992; PsA, n = 127). ARs and serious ARs were reported in 29.4% and 0.7% of patients, respectively; most occurred <1 month after apremilast initiation. There were no reports of fatal ARs, serious infections, hypersensitivity, or vasculitis. No new safety signals were identified. Among the key survey items, gastrointestinal disorders were the most common ARs (21.3%). In patients with PsO, after 6 and 12 months of treatment, effectiveness rates of achieving highly effective or effective global improvement of were 90.9% and 93.8%; PGA 0/1 was achieved by 42.7% and 58.1% of patients; mean decrease from baseline in total DLQI score was 4.2 (p < 0.0001) and 5.7 (p < 0.0001), respectively. Effectiveness was evaluated in a small number of patients with PsA for some measures; after 6 and 12 months of treatment, improvements were observed in global improvement effectiveness rates, Disease Activity Score in 28 Joints score, Visual Analog Scale score, and DLQI score. We conclude that orally administered apremilast was well tolerated and effective in Japanese patients with PsO and/or PsA enrolled in this post‐marketing surveillance study.


| INTRODUC TI ON
Plaque psoriasis (or psoriasis vulgaris; PsO) and psoriatic arthritis (PsA) are complex, multifactorial, chronic immune-mediated inflammatory diseases that primarily affect the skin and joints respectively and are associated with a high disease burden and significantly impaired health-related quality of life (HRQoL). 1,2e of the key inflammatory pathways in PsO pathogenesis involves the phosphodiesterase-4 (PDE4)-mediated degradation of cyclic adenosine monophosphate (cAMP), a key intracellular second messenger, which promotes an increase in pro-inflammatory mediators and a decrease in anti-inflammatory mediators. 3remilast (Otezla®; Amgen Inc.) is an orally administered, small molecule that selectively inhibits PDE4, resulting in increased intracellular cAMP levels and subsequent downstream effects on intracellular pathways involved in innate and adaptive immunity as well as in non-immune cells. 1 These effects suppress the production of a broad range of pro-inflammatory cytokines, such as tumor necrosis factorα, interleukin (IL)-17, and IL-23 and promote the expression of anti-inflammatory cytokines such as IL-10. 3As apremilast acts early in the inflammatory cascade, it is thought to lead to a broader and more balanced regulation of inflammatory mediators than biological agents that target a specific mediator. 1remilast has been studied extensively in clinical trials of patients with PsO and PsA including the phase 2b PSOR-005 study 4 and phase 3 ESTEEM 1 and 2 studies, 5,6 which showed that apremilast reduced the severity and extent of moderate to severe plaque psoriasis and was well tolerated.Furthermore, the phase 3 PALACE 1-4 studies 7-10 demonstrated the clinical benefits and favorable safety profile of apremilast among patients with PsA.In a phase 2b study conducted in Japan, 11 apremilast demonstrated similar efficacy and tolerability to the phase 2 and phase 3 studies conducted in multiple countries, [4][5][6]12 confirming that data from the pivotal phase 3 studies can be extrapolated to Japanese patients. Basd on those studies, apremilast was approved in Japan in 2016 for the treatment of PsO in patients with an inadequate response to topical therapies as well as for the treatment of PsA. 13 This post-marketing surveillance (PMS) study was conducted to investigate the safety and effectiveness of apremilast in Japanese patients with PsO and PsA treated in real-world clinical practice.

| Study design
This was a multicenter, prospective, observational study (ClinicalTrials.govidentifier: NCT03284879; European Union electronic Register of Post-Authorization Safety Studies [EU PAS] identifier: EUPAS36684) conducted in Japan in patients with PsO and/or PsA who were treated with apremilast in routine clinical practice.
A total of 1086 patients were enrolled at 160 hospitals or clinics in Japan between September 1, 2017, and August 31, 2019, and were observed for 12 months after the start of apremilast treatment or until discontinuation or withdrawal.In cases of treatment discontinuation, data were recorded on concomitant medications and adverse events (AEs) for 2 months from the date of discontinuation.A centralized, electronic data capture system was used for patient registration (within 14 days of apremilast initiation) and data collection.
Participating physicians completed two case report forms (CRFs) during the observation period (one after 6 months [CRF1] and the other after 12 months [CRF2] of treatment) detailing patient demographic data and baseline characteristics, apremilast treatment data, concomitant medications, AEs, disease activity scores, and patientand physician-reported outcomes.
The study was conducted in compliance with the regulatory requirements stipulated in the Good Post-marketing Study Practice (GPSP) guidelines of Japan.Institutional review board and ethics committee approvals as well as informed consent were not required according to the GPSP ordinance.However, approval for the study was obtained if requested by a participating medical institution.

| Patient selection and treatment
Patients who received apremilast for the first time for PsO after an inadequate response to topical therapies and/or for PsA were enrolled in the study.Patients with a history of hypersensitivity to apremilast or to any formulation excipient(s) were excluded.
Patients initially received apremilast at a dose of 10, 20, or 30 mg, in accordance with Japanese prescribing information. 12

| Outcome measures
Safety was evaluated by recording all ARs and serious ARs.AEs reported by physicians other than those assessed as "not related" by physicians (including unknown/not specified) were tabulated as ARs.Serious ARs were defined as ARs assessed as "serious" by physicians.ARs were classified using preferred terms and system organ classes (SOC) in the Japanese version of the Medical Dictionary for Regulatory Activities (MedDRA/J) version 25.0.Listed/unlisted ARs were assessed based on the description in apremilast prescribing information (revised July 2020, 2nd edition).In this survey, safety information was collected with "gastrointestinal disorders," "serious infections," "serious hypersensitivity," "weight decreased," "vasculitis," "malignancies" and "depression and suicidal events" as the key survey items.The occurrence of ARs by time, number and the percentage of patients who experienced ARs, especially gastrointestinal disorders, in each of the patient background categories were recorded to identify factors potentially affecting the safety of apremilast.
Treatment effectiveness in patients with PsO was evaluated by the effectiveness rate, which was the proportion of patients who reported a global improvement of highly effective or effective, Physician's Global Assessment (PGA) scores of 0/1, DLQI score 0/1, or <5 and the change from baseline in the patient-reported DLQI, after 6 and 12 months of treatment. 14,15Effectiveness in patients with PsA was assessed by determining the proportion of patients who achieved global improvement and change from baseline in the Visual Analog Scale (VAS) scores for pain (0-100 mm, assessed by patients), Disease Activity Score in 28 Joints (DAS28-CRP), and DLQI after 6 and 12 months of treatment. 16

| Statistical analysis
The target population size was 1000 patients.Based on the results of a Japanese clinical trial 11 in which three of 241 (1.2%) patients with PsO experienced serious infections, the threshold of serious infections in this study was set at 1.2%.It was estimated that 814 patients would need to be included in the safety analysis set to have ≥80% power for detecting serious infections at the specified threshold (1.2%), even if the true risk was more than twice this threshold (i.e., >2.4%).Therefore, the planned population size was set at 1000 patients to account for dropouts during the 12-month observation period.
The safety analysis set included all patients with completed CRFs who did not meet safety exclusion criteria (i.e., registration outside the contract period, enrollment after 14 days of apremilast administration, patient duplication, no apremilast administration, treatment initiation prior to 1 September 2017, and no hospital visit after apremilast initiation).The effectiveness analysis set included all patients in the safety analysis set who did not meet the following effectiveness exclusion criteria: treatment of diseases other than approved indications, previous apremilast treatment, no effectiveness assessments, and patients not fulfilling "inadequate response to topical therapies" criteria.Data were summarized using descriptive statistics, with means and standard deviations (SD) and medians and ranges (minimum and maximum) used for continuous variables, and the number and proportion of patients used for categorical variables.The relationship between patient baseline characteristics and safety was evaluated using the Fisher's exact or Pearson's chi-squared tests.A logistic regression Cox analysis was performed with specific items as explanatory variables (i.e., sex, age, diagnosis, disease duration, and comorbidities) to investigate risk factors for the occurrence of gastrointestinal disorders; odds ratios (ORs) and their 95% confidence intervals (CIs) were calculated.
A p-value of <0.05 was considered statistically significant.No imputation was made for missing data.All statistical analyses were performed using the SAS software version 9.4 (TS1M5) (SAS Institute Inc).
Patients were counted once for each diagnosis (PsO or PsA) if they had both PsO and PsA.The safety analysis set included 1063 patients, with 992 patients having PsO and 127 having PsA, meaning that 56 patients were diagnosed as having both PsO and PsA.
Seventeen patients were excluded from the safety analysis set for the following reasons: enrollment after 14 days of apremilast administration (n = 4), no apremilast administration (n = 2), and no hospital visit after apremilast initiation (n = 11).The effectiveness analysis set included 1013 patients (954 patients with PsO and 115 with PsA); 50 patients were excluded because of no effectiveness assessments.
In the safety analysis set, 693 of 1063 patients were followed for 6 months and completed CRF1; 370 patients had discontinued treatment, mainly due to AEs (n = 124), no hospital visit (n = 92), or inadequate response (n = 91; Figure 1).After 12 months' follow-up, CRF2 was completed by 546 patients; 145 patients discontinued treatment, mostly because of an inadequate response (n = 39), no hospital visit (n = 36), or AEs (n = 27).The reason for discontinuation of two patients was unknown.Baseline characteristics and treatment patterns of the safety analysis set are summarized in Table 1.
Most ARs were reported within 1 month of initiating apremilast treatment and decreased in number gradually over time (Table 4).
Diarrhea was the only common AR (≥5% of patients) occurring within 1 month of treatment.
Among the key survey items, gastrointestinal disorders were reported in 226 (21.3%) patients, weight decreased in three (0.3%) patients, malignancies occurred in two (0.2%) patients, and depression and suicidal events in three (0.3%) patients (two cases of depression and one suicidal event but no death).Serious infections, serious hypersensitivity, or vasculitis were not reported.Baseline patient characteristics associated with a significant increase in the incidence of gastrointestinal disorders were smoking history (p = 0.0253), comorbidities (p = 0.0003), comorbidities except for liver, kidney, or psychiatric disorders, or infections (p = 0.0052), and prior and concomitant medications (p = 0.0168 and p = 0.0007, respectively; Table 5).In multivariate logistic regression analyses, no significant association was found between any of the explanatory variables (i.e., sex, age, diagnosis, disease duration, or comorbidities) and the occurrence of gastrointestinal disorders (Figure 2).Among the 226 patients with gastrointestinal disorders, 63 patients discontinued apremilast treatment.
Improvements in quality of life were also observed; the total DLQI score decreased significantly from baseline (mean [SD] decrease of 4.2 [5.4] at 6 months; 5.7 [6.8] at 12 months; p < 0.0001 at both time-points) (Tables 6 and 7).

| DISCUSS ION
Results of this large-scale, real-world, PMS study in Japanese patients with PsO and/or PsA suggests that apremilast has an acceptable safety and tolerability profile and has positive clinical assessments of effectiveness and physician/patient-reported outcomes.7][8][9]17 Our results are also in line with those of three single-center, real-world studies in Japanese patients with PsO, conducted at Jichi Medical University, [18][19][20]  a Severity of PsA is included in Table S2.
b Disease or symptom that was cured before apremilast initiation.
c Disease or symptom present at apremilast initiation.

TA B L E 1 (Continued)
][9][10][11]17,[21][22][23] The incidence of gastrointestinal ARs was higher in patients with a history of smoking, which represents one of the important findings of this study.It was also higher in patients with comorbidities as well as in those who had received prior and concomitant medications with no significant association between explanatory variables (such as sex, age, diagnosis, disease duration, or comorbidities) and the occurrence of gastrointestinal disorders as determined by multivariate logistic regression analyses.
Upper respiratory tract infections and nasopharyngitis were also frequently observed in clinical and observational studies of apremilast but were not commonly observed as an AR in this PMS study, suggesting a low possible causal connection with apremilast.
Moreover, no serious infections were reported with apremilast in this study.In contrast, some biologics and Janus kinase inhibitors used in managing psoriatic disease in Japan may be associated with a higher risk of infections, which requires appropriate screening and monitoring of patients according to recommendations from Japanese guidance. 24,25erall, the rates of serious ARs remained low in this study and occurred only in 0.7% (7/1063) of patients, whereas the phase 2b study in Japanese patients reported an AR incidence of 2.4% (2/85 patients), 26 and PROMINENT, a phase 3b study in Japanese patients with mild to moderate PsO, 17 reported a similar incidence (2.6%; 4/152 patients) of serious treatment emergent adverse events (TEAEs).
PsO and PsA are chronic, systemic diseases associated with increased risk of cardiovascular and thrombotic events. 27In this b Chi-squared test.
c Disease or symptom that was cured before apremilast initiation.
d Disease or symptom present at apremilast initiation.

TA B L E 3 (Continued)
PMS study, the occurrence of cardiovascular ARs with apremilast was low.This is consistent with the results of the pooled analysis of 15 clinical studies of apremilast that demonstrated a low incidence of major adverse cardiovascular events; they were nearly identical between apremilast and placebo groups in the placebo-controlled period (0.1%). 23e incidences of ARs with apremilast related to the key survey items listed in this PMS study were low for weight decrease, malignancies, depression and suicidal events, and none of these events showed an increasing trend over time.Importantly, no instances of serious hypersensitivity, vasculitis or fatal ARs were observed in this PMS, further confirming that no new safety signals were identified.
In terms of onset, most ARs reported in this PMS, such as gastrointestinal disorders, were reported within 1 month of starting apremilast and gradually decreased over time.][7][8][9][10] Resolution within 1 month was also typical for the gastrointestinal disorders of nausea and diarrhea, although these continued in some patients, highlighting the need for management.
When ARs do emerge, it is important for physicians to manage these reactions by using shared decision making with patients, 28 including expectations regarding management, likely resolution, and the importance of continuing apremilast as a systemic treatment, given the nature of PsO and PsA.
][7][8][9][10][11] Apremilast showed high effectiveness in this study, with global improvement seen in >90% of patients with PsO as well as those with PsA.Patients with PsO achieved a significant improvement in the severity of lesions and QoL.Patients with PsA achieved significant improvements in global improvement, joint pain VAS, DAS28, and DLQI; however, the effectiveness of apremilast in PsA in this study should be interpreted cautiously as the evaluable population for VAS, DAS28, and DLQI assessments was small.
Effectiveness results in terms of PGA 0/1 achievement was comparable between patients with PsO in this study and the efficacy results in the phase 3 PROMINENT study in Japan. 17Specifically, PGA 0/1 was achieved in 42.7% of patients in this study at 6 months compared with 43.7% of patients at 16 weeks in PROMINENT. 17In terms of quality of life DLQI 0/1 was achieved in 33.0% of patients at 6 months in this study and 41.2% of patients at 16 weeks in PROMINENT. 17The b Chi-squared test.
c Disease or symptom that was cured before apremilast initiation.
d Disease or symptom present at apremilast initiation.

TA B L E 5 (Continued)
proportion of patients achieving DLQI 0/1 in our study at 12 months was higher than in a single-center clinic study in Japan (54.5% vs 28.6%). 21In this study, the 12-month survival rate was 46.8% whereas data from a retrospective study showed that the 12-month survival rate of apremilast was 53.4% in Japanese patients with PsO or PsA. 19 increase in prescription of apremilast among oral PsO drugs (mainly because of its improved safety profile) was shown in a single-center study 29 and a similar trend was also reported in an epidemiological survey of PsO patients in the Japanese Society for Psoriasis Research. 30Apremilast, with its convenient oral administration and anti-inflammatory effect but not immunosuppressive mode of action, favorable safety profile, as well as no requirement for laboratory monitoring, offers a unique therapeutic option for patients with PsO and/or PsA.Note: Physician Global Assessment (PGA) and Dermatology Life Quality Index (DLQI) scores were calculated in patients in whom these scores could be calculated at the start and 12 months after the start of apremilast treatment.
TA B L E 7 Effectiveness of apremilast 12 months after treatment initiation in patients with psoriasis in the effectiveness analysis set.

F I G U R E 3
Response rate for Physician's Global Assessment (PGA) 0/1, Dermatology Life Quality Index (DLQI) 0/1 and DLQI <5 at 6 and 12 months in patients with plaque psoriasis (PsO).n indicates the number of patients assessed by PGA using 6-or 12-month data from patients in whom each score could be calculated at two time-points, at the start of apremilast treatment and 6 or 12 months after the start of apremilast treatment and whose PGA score at the start of treatment was not "None," "Minor," or "Non-judgeable."N indicates the number of patients assessed by total DLQI scores using 6-or 12-month data from patients in whom each score could be calculated at two timepoints, at the start of apremilast treatment and 6 or 12 months after the start of apremilast treatment.SD, standard deviation.

3. 3 . 2 |
Patients with PsA At the 6-month time-point, 86 patients with PsA were included in the analysis for global improvement.At the 12-month time-point, F I G U R E 1 Completion and discontinuation status after 6 and 12 months of treatment.a Patients were counted once for each reason if they had more than one reason for discontinuation.CRF, case report form.
Baseline characteristics and treatment patterns of patients treated with apremilast (safety analysis set).
mission was achieved by 52.6% and 77.8% of patients respectively, and a low disease activity was achieved by 21.1% and 22.2% of patients, respectively.The total DLQI score decreased significantly from baseline (mean [SD] decrease of 4.7 [4.1] at 6 months and 4.8[4.5]at 12 months).
Adverse reactions (ARs) in patients in the safety analysis set (N = 1063).
TA B L E 2 a Causality with the drug could not be ruled out for three patients but all five patients eventually recovered or achieved remission.b Individual weight decreases were 5.4 kg, 5.5 kg, and value unknown.a Fisher's exact test.
Occurrence of adverse reactions (AR) by time-period.Occurrence of adverse drug reactions (ARs) of gastro intestinal disorders by patient baseline characteristics.
a Fisher's exact test.
Multivariate logistic regression analyses of patient factors associated with gastrointestinal (GI) disorders.CI, confidence interval.Physician Global Assessment (PGA) and Dermatology Life Quality Index (DLQI) scores were calculated in patients in whom these scores could be calculated at the start and 6 months after the start of apremilast treatment.
2F I G U R E 2 TA B L E 6 Effectiveness of apremilast 6 months after treatment initiation in patients with psoriasis in the effectiveness analysis set.Our study has several limitations including the non-comparative open-label, non-randomized study design.This could result in selection bias, information bias, and attrition bias, all of which could affect the study results.It should also be noted that previous clinical and observational studies typically reported TEAEs rather than ARs (events reported by physicians other than those assessed as "not related" by physicians [including unknown/not specified]), as assessed in the current study.This makes direct comparison of results with previous studies difficult.Finally, the number of patients with PsA was relatively small, making the results in this population more difficult to generalize compared with results associated with patients with PsO.In conclusion, this large, observational, PMS study demonstrated that orally administered apremilast was not associated with new