Peculiarities of psoriatic arthritis among Nigerian psoriasis patients: a comparative cross-sectional study

¹Rheumatology Unit, Obafemi Awolowo University Teaching Hospitals Complex, Ile-Ife, Osun state, Nigeria, ²Rheumatology Unit, The Royal Wolverhampton NHS Trust, United Kingdom, Corporate Services Centre, New Cross Hospital, Wolverhampton WV10 0QP, United Kingdom, ³Dermatology and Genitourinary Medicine unit, Faculty of Clinical Sciences, College of Medicine, University of Lagos, Lagos, Nigeria, ⁴Rheumatology Unit, Lagos State University College of Medicine, Lagos, Nigeria, ⁵Lagos State University Teaching Hospital, Ikeja, Lagos, Nigeria, ⁶Department of Preventive Medicine and Biostatistics, Uniformed Services University of the Health Sciences, Maryland, USA, ⁷Rheumatology Unit, Royal National Hospital for Rheumatic Diseases, Combe Park, Bath, United Kingdom, ⁸Faculty of Medicine, Dentistry and Health, University of Sheffield, United Kingdom, ⁹Department of Dermatology and Venereology, College of Health Sciences, Obafemi Awolowo University, Ile Ife, Osun State, Nigeria

^&Corresponding author
Adeola Ajibade, Rheumatology Unit, Obafemi Awolowo University Teaching Hospitals Complex, Ile-Ife, Osun state, Nigeria

Introduction    

Psoriatic arthritis (PsA) is a heterogeneous disease with multiple musculoskeletal and dermatological manifestations [1]. It is an inflammatory arthritis associated with psoriasis, which runs an indolent and progressive course [2]. Clinical diagnosis of PsA is mainly by characteristic signs and symptoms involving the skin and joints [2]. Psoriatic arthritis was initially considered to be a mild disease, but in the past decade, 40%-60% of patients with Psoriasis (PsO) were reported to develop erosive and deforming joint complications [3], making PsA a more severe condition than previously thought. Furthermore, Psoriatic arthritis-induced joint damaging complications lead to lower articular function, and higher mortality, and also affect patients´ ability to work, and influence their social relationships [3,4]. Prompt diagnosis and treatment can improve pain and inflammation, and prevent progressive joint involvement and damage [2].

The documented patterns of joint involvement in PsA include asymmetrical oligoarticular arthritis, symmetrical polyarthritis, distal interphalangeal arthropathy, arthritis mutilans, and spondylitis with or without sacroiliitis (axial) [5]. Classification of PsA was initially based on the Moll and Wright [6] classification criteria (1973), however, patients with cutaneous psoriasis who had musculoskeletal involvement, such as dactylitis, enthesitis, and tendinitis, were missed [7]. The Classification criteria for Psoriatic Arthritis (CASPAR) was developed in 2006, and published by the CASPAR study group [8]. The CASPAR criteria were statistically derived from a large prospective study [7]. The criteria are simple to use and highly specific, with sensitivity and specificity of 91.4% and 98.7% respectively [8,9], and a sensitivity of up to 99.1% in early PsA [10].

Psoriatic arthritis is a multifaceted disease, with a prevalence (among psoriasis patients) of 6 - 42% [11-13]. New developments in classification criteria, pathophysiology, evaluation tools, and therapeutic options are paving the way for a new look at the disease [14,15]. Despite these new developments, studies exploring the pattern of psoriatic arthritis in Sub-Saharan Africa, especially among Nigerians are still quite few. Psoriatic arthritis has been reported amongst rheumatology patients in Nigeria [16-18]. These studies place the prevalence of PsA in Nigeria at 0.5-0.8% [17,18]. These may however not reflect the true burden of the disease, as patients who do not get referred to rheumatology clinics, or who were not correctly diagnosed as PsA, would have missed out. Furthermore, there have been reports of a high prevalence of undiagnosed active PsA among PsO patients seen in dermatology clinics [19-21]. There is no systematic documentation of psoriatic arthritis among psoriasis patients, using CASPAR criteria, in Nigeria.

Objective: this study estimates the burden of psoriatic arthritis (using the CASPAR Criteria), and describes its peculiar pattern among Nigerian psoriasis patients from a regional tertiary hospital. It also compares characteristics of patients with and without psoriatic arthritis, to determine possible factors associated with the development of PsA. This will stimulate awareness, and aid in early diagnosis.

Methods     

Study design and setting: this was a hospital-based cross-sectional study carried out at the dermatology outpatient unit of the Lagos University Teaching Hospital (LUTH) over two years (January 2017 - January 2019). Lagos University Teaching Hospital is a 700-bed Federal government-owned tertiary institution that serves as a major referral centre for Southwestern Nigeria. Lagos, the main geographical location that the hospital serves, is multi-ethnic, with an estimated population of 17.5 million.

Participants: all patients with biopsy-established psoriasis, seen within the study period, who were at least 16 years of age, and gave written informed consent, were serially recruited. Patients with other established inflammatory arthropathies were excluded. Ethical approval was given by the LUTH Health Research Ethics Committee (ADM/DCST/HREC/APP/1350).

Sample size determination: this was done using Fischer´s formula for sample size determination for cross-sectional study. Estimated disease prevalence (P) was determined based on a hospital-based study done in Nigeria [22], which found a prevalence of arthritis of 3.2% among patients with psoriasis. With a p-value of 3.2% (0.032), n was calculated to be 47.6. Adding a 10% attrition rate of 4.76, the calculated minimum sample size was 52.36. To obtain a more rounded estimate, which will allow normal distribution for analysis, 60 participants were recruited.

Data source/collection: an interviewer-administered proforma was used to collect demographic data, relevant clinical information, and laboratory/radiological details. Clinical diagnoses of psoriasis and severity score were made by a dermatologist, and histological confirmation was made by a dermatology pathologist. The participants were interviewed and examined by a rheumatologist. Medical history collected from participants included present or past psoriasis lesions, duration of psoriasis, type of psoriasis, joint pain and/or swelling, joint stiffness, dactylitis, tenosynovitis, enthesitis, enthesitis site, number and type of joint involvement, nail involvement, eye involvement, co-morbidities (hypertension, diabetes mellitus, dyslipidaemia, gout, HIV infection), family history of psoriasis or arthritis, medication history and cigarette/alcohol intake. Physical examination including inter-triginous areas was conducted to determine the presence of psoriatic lesions. A detailed musculoskeletal examination was conducted to assess tender and/or swollen joints, tenosynovitis, and dactylitis. Schober´s test (determined thoracolumbar flexibility), Occiput-to-wall measurement, Sacro-iliac/axial joint tenderness, and the FABER test were also carried out. Enthesitis points were assessed using the Spondyloarthritis Research Consortium of Canada (SPARCC) enthesitis index [15,23].

Variables

Main outcome: participants who had inflammatory articular disease (inflammatory joint/back pain, synovitis, tenosynovitis, entheseal inflammation) were considered to have met the entry point for CASPAR criteria (Annex 1). These participants were then assigned points based on the five categories of the CASPAR criteria. Any participant who scored ≥3 points was categorized as psoriatic arthritis.

Co-variates: blood samples were collected for relevant investigations, including Fasting Lipids Profile (FLP), erythrocyte sedimentation rate (ESR), high sensitivity C-reactive protein (HS-CRP), serum uric acid, fasting blood glucose (FBG) and rheumatoid factor (RF). For subjects with physical findings suggestive of inflammatory articular disease, a digital X-ray was done on at least one involved joint.

Contextual variables: the following socio-demographic variables were also explored in the study - age (young/middle age/elderly), gender (male/female), occupation (students/artisans/civil servants/none), formal education (secondary education or less/tertiary education)

Bias and confounders: to reduce selection bias, ALL patients who satisfied the inclusion and exclusion criteria were recruited. The use of logistics regression analysis suppressed cofounders that may emerge as predictors of PsA among PsO patients, patients with other established inflammatory arthropathies were excluded to suppress confounders.

Statistical methods: data was analysed using the commercially available statistical package for social sciences (SPSS) IBM version 22.0 analytical software. Shapiro-Wilks test was used to determine the normality of the distribution of variables. Continuous variables were summarized and presented as means and standard deviations (or median and interquartile range where the variable distributions are skewed). Categorical variables were summarized as frequencies and percentages. Associations between psoriatic arthritis and the tested determinants were evaluated using the Chi-square test for categorical variables and the independent sample T-test (or Mann-Whitney U-test) for continuous variables. Firth logistic regression test was carried out to identify possible predictors of PsA. Missing data were assigned a universal value, and labeled as missing data, for analysis. Statistical significance was set at a p-value of <0.05. Participants who fulfilled the CASPAR criteria (Annex 1) were categorized as Psoriatic Arthritis. For this study, simultaneous occurrence of psoriasis was defined as psoriasis lesion(s) and psoriatic arthritis starting within 12 months of each other [24].

Results     

Description of study participants: a total of 65 potentially eligible participants were approached, out of which 5 were found ineligible (2 refused consent, while 3 had other established inflammatory arthropathies). Sixty (60) study participants were recruited and analyzed, with 32 (53.3%) males and 28 (46.7%) females. The mean age was 47.5 ± 16.9 years. Most patients had chronic plaque psoriasis (73.3%). Of the location variants, the majority (41.7%) had scalp psoriasis. About two-thirds (61.7%) of patients had psoriasis in multiple sites while 25.0% were already on immunosuppressives (methotrexate± sulfasalazine/apremilast), of which 40% fell into the PsA group. Less than a tenth (9.4%) of patients had a positive schober´s test, all of whom fell into the category of PsA. The clinical parameters of study participants are summarized in Table 1.

Outcome data-psoriatic arthritis based on CASPAR criteria: fifteen (25%) participants fulfilled the classification criteria for psoriatic arthritis. Of note, there were no significant differences in sociodemographic characteristics of PsA and non-PsA patients as shown in Table 2. Nine (60.0%) of the participants classified as PsA were females, while 6 (40.0%) were males; (F:M ratio = 1.5:1). The mean age of PsA patients based on CASPAR criteria was 53.0±16.5yrs. The median duration of PsA was 18.5 (12, 48) months. Some participants had joint complaints that did not fulfill CASPAR criteria - 8 (13.3%) patients had non-specific arthralgia, 9 (15.0%) had osteoarthritis, and 1 (1.7%) had isolated plantar fasciitis (Figure 1).

Clinical characteristics of psoriatic arthritis: psoriatic arthritis was further classified based on the Moll and Wright articular classification (Figure 2). None of the 15 patients who fulfilled the CASPAR criteria had arthritis mutilans. Oligo/mono-articular pattern occurred most frequently in 10 (66.7%) cases, a poly-articular, and purely axial pattern was found in 2 (13.3%) patients each, while the Distal Inter-Phalangeal (DIP) joint pattern was found in only 1 (6.7%) patient. Eight (53.3%) patients had mixed articular patterns (peripheral and axial involvement). Psoriasis preceded the onset of arthritis in 11 (73.4%) patients, while arthritis preceded psoriasis onset in 2 (13.3%). Two (13.3%) patients had simultaneous onset of arthritis and psoriasis. Chronic plaque Ps (53.3%) and Scalp Ps (40%) were still the most frequently seen morphological and location variants respectively, amongst this subgroup. Twelve PsA patients had peripheral radiographs done; only one patient had marginal erosions, 4 (33.3%) had joint space narrowing, while the majority (7) had normal findings. The majority (8) also had normal findings on axial radiographs. Five (38.5%) had features suggestive of spasm of the back muscle, evidenced by straightening of the normal lumbar lordosis. Demonstrable sacro-ilitis was not found in any subject. An axial radiograph was done on 13 subjects.

Comparative analysis of patients with and without psoriatic arthritis: inflammatory eye disease (IED) was significantly more prevalent in patients with PsA (26.7% vs 2.2%, p = 0.01). Nail involvement occurred significantly more frequently among patients with PsA (66.7%) compared to those without PsA (31.1%), with a p-value of 0.02. Dactylitis occurred in 3 (20.0%) of patients with psoriatic arthritis, while none of those without PsA had dactylitis (p = 0.01). All (100%) of the patients with PsA had enthesitis, while only 1 (2.2%) of patients without PsA had, with a level of significance of <0.001 (Table 3). The most common entheseal sites involved were the tibial tubercle, plantar fascia, and iliac crest. There were no statistically significant differences among the laboratory parameters of both groups.

Other analyses: a firth logistic regression was carried out to evaluate the independent effects of the variables that had a significant association with psoriatic arthritis. These were - inflammatory eye disease (IED), nail involvement, enthesitis, dactylitis, peripheral joint pain and swelling, and back pain. Enthesitis and dactylitis had null cells in the bivariate analysis and were therefore excluded. Back pain, peripheral joint pain, and nail involvement are the only variables that remained significant following the logistic regression (Table 4).

Discussion     

The prevalence of PsA among psoriasis patients has been reported to vary between 6 and 42% [11] worldwide, however, the diagnosis is still often missed [25]. The prevalence in our study (25%) falls within this range and is similar to what was reported by Çinar et al. [24] in Turkey (25.4%) and Reich et al. [21] in Germany (20.6%). About one-third of participants had chronic joint complaints that did not meet the CASPAR criteria for PsA - similar to findings in other studies [26-29]. This is not unexpected in our study, considering that osteoarthritis (OA) is common in the mean age group obtained. There have been controversies about the designation of isolated enthesitis as PsA: while some experts believe that rheumatoid factor-negative patients with isolated enthesitis are considered to have PsA, others do not agree with this [30]. For this study, isolated plantar fasciitis was not classified as having PsA.

Asymmetric oligo/mono-articular pattern was the most common articular pattern found, occurring in 66.7% of cases, while no arthritis mutilans were found. A higher occurrence of the oligo-articular pattern was also reported by Akpabio et al. [17] (in a previous Nigerian study) and Cinar et al. [24]. Kumar and colleagues [31], however, reported polyarthritis as the most common articular pattern in India. This may be because the average duration of PsA was higher in the Indian study population (4.19 years). In most studies and series, the frequency of oligo/mono-articular involvement has been put at 11 - 70%, with a frequency of arthritis mutilans at <5% [6,32]. Akpabio et al. [17] and Kumar et al. [31] found 1 (one) arthritis mutilans each in Lagos (Nigeria) and India respectively. These results (including ours) substantiate the previous reports that arthritis mutilans are rare. Spondylitis is reported to be uncommon as a predominant feature of PsA [6], as seen in this study.

The majority of participants had normal findings on radiographs. None of the patients had detectable sacro-ilitis on radiography, while only one had marginal erosion on the peripheral x-ray. Akpabio et al. [17] demonstrated radiographic sacro-ilitis in only one patient. Reports from other African sub-regions, and outside Africa, such as Turkey [24] demonstrated sacro-ilitis in 37.5% of patients on imaging (50% by radiograph). In South Africa, Green et al. [33] found only about 5% of cases with axial PsA without radiographic sacro-ilitis. While in India, Kumar et al. [31] showed X-ray abnormalities in 29% (14% had erosions, while 9% had joint space narrowing). It is worthy of note, however, that 25% of the participants in our study had been on immunosuppressive therapy for their psoriasis, and only 2 (3.3%) had a prior diagnosis of PsA. It is possible that the immunosuppressives had a modifying effect on the development and/or progression of joint complications. It could also be that articular damage progresses slowly in Nigerian patients with inflammatory arthritis. This is considering that radiographic changes were mostly mild in Nigerian patients with rheumatoid arthritis in a report by Adelowo et al. [34]. Akpabio et al. also found sacro-ilitis in only one patient in his cohort [17]. These observations expose areas that need to be further explored and understood in PsA care in Nigeria. Non radiographic sacroilitis may have however been detected on MRI which was not done in this study [35].

Documented reports have shown that in the majority of cases, cutaneous manifestations develop before arthritis, a subset may have concurrent skin and joint manifestation, while more rarely, joint manifestation may precede cutaneous disease - a condition known as “psoriatic arthritis sine psoriasis” [2,36]. In this study, 73.4% had psoriasis before joint manifestations, similar to what has been documented in the literature (60 - 80%)[25]. Patients who had initial arthritis before skin manifestation had been managed for seronegative R.A. and osteoarthritis, despite atypical symptoms. A high index of suspicion, together with the knowledge of extra-articular PsA manifestations, in the setting of atypical arthritis presentations may reduce the missed diagnosis of PsA.

There are conflicting reports about the duration of psoriasis and the development of PsA. Some authors have found PsA to be associated with a long duration of psoriasis [37], while others have not [32]. The median duration of psoriasis was higher for patients with PsA in this study compared with non-PsA, though not statistically significant. Nail involvement occurs in 20% to 40% of patients with uncomplicated psoriasis, whereas 60% to 80% of patients with psoriatic arthritis have nail involvement; mostly associated with the DIP pattern [6]. This study demonstrated a significantly higher nail involvement among PsA patients compared to those without. This highlights the importance of being familiar with the nail features of Psoriasis (pitting, ridging, onycholysis, hyperkeratosis) [6,35], and the benefits of joint management/evaluation of patients with dermatologists. The only participant with a DIP pattern in this study did not have nail involvement. Inflammatory eye disease has been particularly associated with PsA [32,38-40]. Uveitis has consistently been associated with Spondyloarthropathies; while in PsA in particular [38], the most frequent inflammatory eye disease were conjunctivitis and uveitis [39-41]. Inflammatory eye disease was found in our PsA study population, and this suggests that ophthalmology assessment will be a beneficial requirement in the evaluation of PsO and PsA patients (as this is not yet an established practice in Nigeria).

Other co-morbidities that have been implicated in PsA in literature [38,39] like cardiovascular diseases, obesity, diabetes mellitus, liver disease, gout, and uric acid were not significant in this study. Though ESR was higher in the PsA group, the difference was not statistically significant.

Generalizability: the study is hospital-based, so generalization of the findings to the public may require some caution. It however is a reflection of cases seen in this clime, moreso, in a multi-ethnic location such as Lagos.

Limitations: admittedly, the population for this study is small. It is, however, a reflection of the PsO burden of the chosen centre, as a previous estimate at the same centre yielded a total of 124 patients for 4.5 years [22]. A multi-center approach in the future will be optimum. Investigations that could have been carried out to exhaustively investigate and/or assess co-morbidities and associations (such as genetic studies, HLA B27, dual energy X-ray absorptiometry, magnetic resonance imaging, and colonoscopy) could not be done on account of exorbitant cost and/or unavailability (and were not the objectives of this study).

Conclusion     

This study showed that psoriatic arthritis may not be rare among Nigerian psoriasis patients as previously suggested. One out of every four psoriatic patients had PsA using the CASPAR criteria and the most consistent feature was enthesitis found in 100% of the participants. The mean age of presentation of PsA was 53.0 ± 16.5 years and the duration of PsO onset was significantly longer in participants with PsA. In this study, scalp psoriasis, nail involvement, and inflammatory eye disease (uveitis and conjunctivitis) were associated with PsA. Articular damage/progression in this study was mild/slow. The study highlights the importance of a multi-disciplinary approach to the holistic care of psoriatic arthritis patients. There is a need for a high index of suspicion for early diagnosis and management of PsA. Early involvement of specialties to cater for the articular, peri-articular, and extra-articular manifestations, will benefit the patients with regards to symptom relief, better quality of life, and prevention of complications. These teams may include rheumatologists, dermatologists, ophthalmologists, physical/occupational therapists, orthotics service and radiologists.

Disclaimer: the opinions and assertions expressed herein are those of the authors, and do not reflect the official policy or position of the Uniformed Services University of the Health Sciences or the Department of Defense.

Competing interests     

All authors declare no competing interests.

Authors' contributions     

All authors contributed to the study's conception and design. Material preparation, data collection, and analysis were performed by Adeola Ajibade, Olusola Ayanlowo, Hakeem Olaosebikan, and Bolanle Olapeju. The study was supervised from beginning to end by Olufemi Adelowo. The first draft of the manuscript was written by Adeola Ajibade, and all authors commented on previous versions of the manuscript. All authors have read and approved the final manuscript.

Acknowledgments     

The authors acknowledge the Central Research Laboratory of the College of Medicine, University of Lagos.

Tables and figures     

Table 1: clinical characteristics of participants

Table 2: comparison of demographic and clinical characteristics of patients with and without psoriatic arthritis

Table 3: factors in psoriatic arthritis

Table 4: possible predictors of psoriatic arthritis among patients with psoriasis

Figure 1: arthropathies in psoriasis

Figure 2: articular patterns of psoriatic arthritis; A) Moll and Wright articular pattern; B) other articular patterns

Annex     

Annex 1: the caspar (classification of psoriatic arthritis) criteria (PDF KB 66)

References