Disease Characteristics and the Burden of Joint and Skin Involvement Amongst People With Psoriatic Arthritis: A Population Survey

The study was a non-interventional, cross-sectional survey that captured current and historic data from physicians on their consulting PsA patients and patient-reported outcomes. Physicians were identified via publicly available lists and invited to participate in the study. Those who agreed to participate were asked to complete a patient record form for a minimum of the next three adult patients they consulted with who had a physician-confirmed diagnosis of PsA, were aged ≥ 18 years, and for whom they were actively involved in drug management. The same patients were invited, by their physicians, to complete a voluntary questionnaire. Clinical diagnosis of PsA was based upon the physicians’ individual judgment with no restrictions on how the diagnosis was determined for inclusion in the study.

This study used data derived from physician-reported patient record forms (PRFs) and voluntary patient self-completed forms (PSCs). Physician-reported data included patient demographics, clinical characteristics including PsA symptoms, flaring patterns, and clinical outcome measures. Physicians also provided information on the number of joints involved in the patient’s PsA, and assessed overall, joint, and skin symptom severity by rating as either ‘None’, ‘Mild’, ‘Moderate’, or ‘Severe’, based on their own clinical judgement. Patient-reported data included information on general health, disease history and symptoms, medications, and patient-reported outcome measures (EuroQoL 5D) [18, 19], Work Productivity and Activity Impairment (WPAI) [20, 21], Health Assessment Questionnaire-Disability Index (HAQ-DI) [22, 23], and the Psoriatic Arthritis Impact of Disease (PsAID12) questionnaire [24]). The EQ-5D utility score is derived from answers to five different areas that include mobility, self-care, usual activities, pain/discomfort, and anxiety/depression and is converted to a score that is anchored at 0 for death and 1 for perfect health. The VAS allows respondents to report their perceived health status with a grade ranging from 0 (the worst possible health status) to 100 (the best possible health status). The HAQ-DI is an index measuring the quality of life related to health originally designed for assessment of patients with rheumatoid arthritis and has values from 0–3 where a value of 0 to 1 indicates mild difficulties to moderate disability, from 1 to 2: moderate to severe disability and from 2 to 3 severe to very severe disability. The PsAID is a PsA-specific instrument developed by the European League Against Rheumatism (EULAR) that provides a patient-reported measure of disease impact on life in general. The questionnaire uses a weighted scoring system, has a range of 0–10 with 4 being considered a patient acceptable symptom state [24, 25].

In order to assess the impact of joint and skin symptoms, PsA patients were also asked a short series of questions; ‘Given the choice of only one, would you rather be completely free of your…? [Skin symptoms; Joint symptoms]’, ‘Do you experience anxiety and/or depression as a direct result of your PsA?’, ‘Which of the following is the cause of your anxiety and/or depression? [Your skin symptoms; Your joint symptoms; Both your skin and joint symptoms]’.

To be eligible for inclusion in this analysis, physician-reported patient data had to include confirmation of the presence of skin psoriasis currently or historically, and/or have current BSA information available. Patients were grouped during the analysis according to whether they experienced joint symptoms alone, or joint symptoms plus active skin symptoms. Patients were defined as having ‘joint-only’ PsA if the treating physician reported body surface area affected by psoriasis was currently 0% or provided a response of ‘no diagnosis of psoriasis’. Similarly, they were defined as having ‘joint and skin’ PsA if their current BSA was > 0%. A sensitivity analysis was also conducted by grouping patients according to the number of affected joints (≤ 4 joints vs. > 4 joints) and for severity of skin symptoms (mild, moderate, severe). The number of affected joints was derived from asking the physician ‘How many joints are currently affected by this patient’s PsA?’ while the severity of skin symptoms stemmed from physician subjective assessment.

Physicians provided further estimates of the impact of PsA using a VAS and where this data was available the score on a scale of 0–100 was based on 0 being equivalent to the best possible health assessment and 100 the worst health assessment. This was collected for global, skin, and joint symptoms separately. Pain was assessed by a physician-reported numeric rating scale (NRS, range 1–10) with 1 representing no pain and 10 worst possible pain. Flaring was assessed by physicians reporting if the patient ever experiences flares (acute episodes), and their response based on their own subjective assessment—no formal definitions of flaring were provided.

Statistical analysis was performed using Stata (version 15.1). Patient demographics, clinical characteristics and PROs were described within each subgroup (joints & skin vs. joints only and ≤ 4 joints vs. > 4 joints for the sensitivity analysis) and overall, using n, mean and standard deviation (for numeric variables) or n and percent (for categorical variables). Clinical burden and PROs were also described (n, mean, SD, %) and compared between physician-reported skin severity (mild, moderate or severe).

When two groups were compared, numeric variables were compared using a t test; ordered categorical variables were compared using a Mann–Whitney test; binary categorical variables were compared using a Fisher’s exact test; other categorical variables were compared using a Chi-squared test. When more than two groups were compared, numeric variables were compared using an ANOVA; ordered categorical variables were compared using a Kruskal–Wallis test; categorical variables were compared using a Chi-squared test. p values < 0.05 were considered statistically significant.

The non-interventional, observational nature of the data collection does not result in patients being placed at risk from the study. Patients provided informed consent to participate in the study and did not provide any personally identifiable information. All responses were anonymized to preserve respondent (physician and patient) confidentiality and all participating physicians and patients were assigned a study number to aid anonymous data collection and to allow linkage of data during data collection and analysis.

The research was conducted in accordance with national market research and privacy regulations (EphMRA, U.S. Department of Health and Human Services National Institutes of Health, HIPAA). Ethical approvals were sought and granted through Western-IRB (Study ID number 1183030) for the US & Canada, and through Freiburg Ethics Commission (FEKI) for all other countries.

Across the 2703 PsA patients, 54.8% were male, with a mean (SD) age of 49.2 (13.4) years, and a mean (SD) BMI of 26.9 (6.7). ‘Joint only’ and ‘joint and skin’ patients were comparable demographically, except in the distribution of employment status (Table 1).

In the overall population, most patients had mild severity disease reported by physicians (Table 2) with a mean of 1.83 current symptoms, 3.19 affected joints, and a pain score of 2.75. Physician-determined overall disease severity was worse in patients with ‘joint and skin’ involvement than those with ‘joint only’ PsA, with a significantly higher proportion reported as moderate or severe (30.7 vs. 14.2%, p < 0.001) (Table 2). Additionally, the mean number of affected joints was 4.0 in patients with ‘joint and skin’ involvement and was significantly higher than in patients with joint only PsA where a value of 1.8 was seen (p < 0.001). This picture of detrimental clinical characteristics in patients with ‘joint and skin’ symptoms extended to physician assessed disease activity including global, joint, and skin VAS scores (Table 2).

Both groups of patients received csDMARDs in approximately comparable percentages (34.3 and 30.6%, respectively), however, there were significant differences in the percentage of patients receiving targeted synthetic DMARDs or biologic DMARDs in the two categories. Patients with ‘joint only’ PsA were significantly less likely to be currently receiving tsDMARDs (apremilast or tofacitinib) compared with patients with ‘joint and skin’ symptoms (6.9 vs. 11.5%, p = 0.001). Conversely, patients with joint only symptoms were more likely to receive biologic DMARDs compared with patients with skin and joint symptoms (67.4 vs. 54.0%, p < 0.001) (Table 1).

Physician-reported flaring was available for 2665 patients and suggested that 58.9% of PsA patients experienced flares at any time (Table 3) with a greater percentage of patients with active ‘joint and skin’ symptoms experiencing a flare currently, within the last 12 months or at periods longer than 12 months compared with ‘joint-only’ patients (62.3 vs. 52.7%). In the overall population, the mean number of flares was 0.65 in the last 12 months and did not vary significantly in patients with ‘joint only’ (0.60) and joint and skin symptoms (0.68). The mean duration of a flare was 17.4 days in the overall population and again did not vary significantly with symptoms being 17.8 days in joint only patients and 17.3 days in joint and skin patients.

Physicians reported that in the overall population 197 patients (7.4%) currently had a flare with the current flare severity assessed as mild (18.3%), moderate (62.4%), and severe (19.3%). When assessed in terms of ‘joint only’ and ‘joint and skin’ status there were no significant differences between groups with the current flare assessed as mild in 27.7% of joint only patients versus 15.3% of joint and skin patients. Corresponding values for moderate and severe flares were 57.4 vs. 64.0% and 14.9 vs. 20.7%.

Of patients currently flaring, 47.7% of patients indicated that both their skin and joints were currently worse than normal; 40% reported only their joints were worse than normal while only 8.4% reported only their skin was worse than normal.

A total of 573 patients provided an assessment of mental health burden associated with PsA. Over 41% of patients indicated that they currently experienced anxiety/depression as a result of their PsA (Fig. 1).

There was a significant difference in the patient-reported cause of their anxiety/depression with ‘both joint and skin symptoms’ being mentioned by 56.1% of ‘joint only’ patients and 65.0% of ‘joint and skin’ patients. Skin symptoms were the cause of anxiety/depression in 3.0% of joint only patients and 11.7% of joint and skin patients, while joint symptoms were the cause of anxiety/depression in 40.9% of joint only patients and 23.3% of joint and skin patients.

Based on the question “Given the choice of only one, would you rather be completely free of your skin symptoms or joint symptoms?”, 61.6% of patients prioritized joint symptoms, while 38.4% prioritized skin symptoms. The symptom prioritization did not differ significantly (p = 0.1513) between the ‘joint only’ and ‘joint and skin’ patient groups with 64.7 and 60.2%, respectively, prioritizing joint symptoms. Even in patients with a BSA = 0 or ‘no diagnosis of psoriasis’, 35.3% of patients in the ‘joint only’ group prioritized skin symptoms as most important compared with 39.8% of patients in the ‘joint and skin’ group (Fig. 1). Joint and skin patients managed by rheumatologists prioritize the symptoms in a similar way (Supplementary Table 1), and results are consistent when analyzed on a regional basis (Supplementary Fig. 1).

Across a range of PROs, patients with ‘joint and skin’ involvement showed significantly worse outcomes compared with patients with ‘joint-only’ symptoms. Physical function measured by the HAQ-DI was statistically significantly worse in the ‘joint and skin’ group compared to the ‘joint-only’ group (0.57 vs. 0.33, p < 0.001). Likewise, quality of life and health status as assessed by the EQ5D and PsAID questionnaires were also statistically significantly worse in the ‘joint and skin’ group compared to the ‘joint-only’ group. The mean (SD) EQ-5D index was 0.85 (0.17) in the ‘joint only’ group and 0.79 (0.17) in patients with ‘joint and skin’ involvement (Table 4). A similar pattern was seen with the EQ-5D visual analogue scale where the respective values were 77.68 (17.51) and 71.98 (19.01) and in both cases differences between the two groups were significant (p < 0.001).

The mean PsAID12 score (SD) was significantly higher at 2.91 (2.08) in patients with ‘joint and skin’ involvement compared with 1.66 (1.86) in patients with ‘joint-only’ involvement. For each of the 12 questions making up this validated score, patients with ‘joint and skin’ involvement showed significantly higher mean scores compared with patients with ‘joint-only’ symptoms (p < 0.0001 across all scores) (Fig. 2).

Significantly more work impairment was seen in three out of the four WPAI domains in patients with ‘joint and skin’ involvement compared with those with ‘joint-only’ involvement (Table 4).

PRO results are presented regionally in Supplementary Tables 4, 5, 6, and follow the same directional trend to the global results.

The relative impact on patient-reported outcomes in patients with a physician-assessed joint count of ≤ 4 vs. > 4 in patients with ‘joint-only’ symptoms is shown in Table 5. Across a range of outcomes patients with four or more affected joints showed significantly worse clinical outcomes ranging from increased disease progression, more current flares, an increase in the mean number of current symptoms, and increased physician-reported pain (Table 5). Alongside clinical outcomes, patient-reported outcomes showed a consistent and significant increase in disability, health status, and work impairment with increased number of affected joints.

Outcomes were also assessed in the 1743 patients with joint and skin PsA stratified according to physician-reported skin severity. Table 6 shows the clinical characteristics of patients with joint and skin involvement stratified according to physician subjective assessment of skin severity. Patients with joint and skin PsA and severe skin disease showed significantly higher levels of physician-reported joint disease, disease progression and pain levels compared with patients with less severe skin involvement (all p < 0.001). Severe joint disease increased from 1.9% of patients with mild skin disease severity to 21.9% of patients with severe skin disease while more patients were either unstable (4.3% of patients with mild skin disease to 18.5% and 31.3% of patients with moderate or severe skin disease) or deteriorating (2.6%, 9.1% and 34.3% of patients with mild, moderate and severe skin disease). Differences in patients managed by rheumatologists and dermatologists are shown in Supplementary Tables 2 and 3.

Current flaring (any flaring of the skin or joint symptoms) in patients with joint and skin involvement was more likely in patients with severe skin symptoms with 45.9% compared with 4.3% of patients with mild skin symptoms. Patients with joint and skin symptoms were more likely to have a current flare than joint-only patients but adverse outcomes increased in patients with joint and skin symptoms and with increasing skin disease.