Skin disease

Psoriasis and psoriaform skin lesions are a feature of SpA. Many measures have been used to assess disease activity in skin psoriasis. The earliest measure to assess the extent of skin involvement was the body surface area (BSA) based on the ‘rule of nines’ used to assess the extent of burns. This allows estimation of the extent of skin surface involved, assuming that the head represents approximately 10% of the surface area, upper limbs 20%, trunk 30% and lower limbs 40% . This estimation of skin surface has been validated using the volume of ointment required to cover an area although the head represented a slightly lower percentage of body surface area, as no ointment was used on the scalp .

In practical assessment of the BSA, an estimation of area is commonly performed using a handprint (the patients’ own) to represent 1% of body surface area. This handprint or palm includes the dorsal aspect of the palm, thumb and all four fingers and has been shown to equate to between 0.51% and 0.91% of the BSA . In the majority of patients, the use of a handprint or palm to assess BSA should therefore overestimate the extent of psoriasis, particularly in mild psoriasis. Assessing skin psoriasis using BSA has reasonable intra-observer reliability, but the inter-observer reliability is variable. Any bias can be minimised by ensuring that the same assessor measures BSA longitudinally.

The psoriasis area and severity index (PASI) was developed within a clinical trial and is the most commonly used outcome measure for psoriasis . This outcome measure combines the assessment of BSA and disease severity. The head, upper limbs, trunk and legs are assessed separately and then combined using weightings based on the BSA (head = 0.1, upper limbs = 0.2, trunk = 0.3, lower limbs = 0.4). The degree of erythema, induration and desquamation in each area are graded on a semi-quantitative scale from 0 (no involvement) to 4 (very marked change) and these scores are summed to measure disease severity. The use of these subjective scores has been validated compared with the objective outcome measures showing significant correlations . The area of involvement of psoriasis in each body area is graded from 0 to 6 based on percentage coverage of skin disease (0% = 0, <10% = 1, 10–29% = 2, 30–49% = 3, 50–69% = 4, 70–89% = 5, 90–100% = 6) and these body area scores are multiplied by the disease severity score and the weighting for each body area. This gives a total score of between 0 and 72 .

The PASI relies heavily on the BSA for components of the score and is therefore subject to the limitations of the BSA, particularly overestimation. The PASI performs better in the assessment of psoriasis, as it includes a score for disease severity as well as area. However, it does not include assessment of the most important features of severity as nominated by both dermatologists (systemic symptoms and pustular change) and patients (embarrassment and quality of life) .

The other major problem is the responsiveness of the scale. There is evidence of general poor responsiveness and problems at either end of the scale. Although the theoretical range of PASI score is 0–72, in practice, it is rare to find a PASI score over 40, meaning that nearly half of the score is unused. At the other end of the scale, the PASI has poor sensitivity to change and responsiveness when skin psoriasis is mild.

Several modifications to the PASI have been proposed to try to improve its performance. The psoriasis log-based area and severity index (PLASI) still scores area in six categories, but the boundaries for each category are based on a log scale (0%, 1–2%, 3–5%, 6–10%, 11–21%, 22–46%, 47–100%) . The psoriasis exact area and severity index (PEASI) was developed by the same team, but uses actual BSA instead of an area score . Both increase the weighting of the BSA thus improving sensitivity. The self-administered PASI (SAPASI) uses an anatomical sketch which the patient is asked to shade in to mark their psoriasis. The patient is also asked to score the colour, thickness and scaling of one typical lesion using three visual analogue scales (VAS). The SAPASI is calculated using the area and severity scores . It has good intra- and inter-rater reliability and correlates well with the PASI suggesting that it would be valid for use in studies where an investigator is not available .

Due to an ongoing dissatisfaction with the PASI, other scoring systems have been developed, using similar parameters. The ‘lattice system physician’s global assessment’ (LS-PGA) requires an assessment of BSA to decide on a category of BSA involved (0%, 1–3%, 4–9%, 10–20%, 21–29%, 30–50% and 51–100%) and then a factoring in of thickness, erythema and scale. This results in a global assessment from ‘clear’ to ‘very severe’ psoriasis . The LS-PGA has reasonable inter- and intra-rater reliability and has shown good correlation with PASI and BSA.

One of the latest instruments developed for use in psoriasis is the ‘Copenhagen psoriasis severity index’ (CoPSI). This instrument scores erythema, thickness and scaling, each on a scale of 0 to 3 at 10 sites of the body, creating a possible score of 0–90 . Testing found a substantial intra- and inter-rater reliability (>80%) for the CoPSI and the PASI, as well as good correlation . The exclusion of BSA removes the inherent inaccuracies with this evaluation, while the number of sites involved provides an estimation of disease extent. The CoPSI should also provide better differentiation at the milder end of psoriasis cases .

The National Psoriasis Foundation (NPF) developed an entirely new responder index called the NPF psoriasis score (NPF-PS). This combines induration at two target sites, ratio of current BSA to baseline, physician and patient global assessment and patient assessment of itch, all scored 0–5. These components are then summed to create a scale from 0 to 30 . An initial study found that the NPF score identified a significant responsiveness after treatment and also identified this at an earlier time point than the PASI, perhaps suggesting that it is more discriminative .

Finally, it must be noted that all of the measures discussed above to evaluate psoriasis were designed for use and validated in plaque psoriasis and have not been assessed in other forms of psoriasis.

Nail disease

Nail involvement is a common feature of both skin psoriasis and psoriatic arthritis. It has been estimated to occur in up to 50% of patients with uncomplicated skin psoriasis and is even more prevalent amongst patients with PsA, between 63% and 83% .

Psoriatic nail disease can be broadly divided into psoriasis affecting the nail matrix and the nail bed. Involvement in the nail matrix results in changes to the nail plate. The most common change is pitting, depressions in the surface of the nail plate resulting from sloughing of abnormal cells laid down when there is involvement in the proximal matrix. If the involvement is in the distal matrix, the abnormality is seen deeper in the nail causing leuconychia. More extensive involvement can result in crumbling of the nail plate. Psoriasis affecting the matrix can be viewed directly in the lunula resulting in red spots. Involvement in the nail bed results in distinct patterns of nail disease. Onycholysis and subungual hyperkeratosis are caused by psoriasis affecting the hyponychium, the transitional area between nail bed and normal epidermis. Involvement more proximal in the nail bed can result in ‘oil drop change’, salmon spots or splinter haemorrhages.

Clearly, assessment of nail disease in PsA must take all of these features into account. Prior to the development of any systematic scoring system, many therapeutic studies of nail disease used subjective measures of nail involvement graded by the investigator as mild/moderate/severe or using a visual analogue scale (VAS).

One of the earliest published nail assessments was developed by the Bath group. Jones et al. used a predefined scoring system where each fingernail was scored for the presence of pitting, onycholysis, hyperkeratosis and dystrophy. Each of these features scored one point with a maximum score of 40 for the 10 fingernails . This scoring system was then used and adapted by the Oxford group who used the same scoring system for finger and toenails with a maximum score of 80 (the psoriasis nail severity score or PNSS) .

In 2003, the nail psoriasis severity index (NAPSI) was devised by dermatologists. This scored nails by dividing each nail into four quadrants and scoring one point for each quadrant, which has any of the features of nail bed psoriasis (onycholysis, oil drop, splinter haemorrhages and nail bed hyperkeratosis) and one point for any features of nail matrix psoriasis (pitting, leuconychia, red spots in lunula and nail plate crumbling) . This results in a score of 0–8 per nail, 0–80 for all fingernails or 0–160 if toenails are included. The advantage of this scoring system is that it does not require scoring for each feature separately but does allow differentiation between nail matrix and nail bed involvement. Initial assessment revealed reasonable reproducibility and a subsequent study showed moderate-to-good interobserver reliability . Following on from this, the NAPSI has been used as an outcome measure in trials of biological agents.

Subsequent to this, a modification, the mNAPSI was proposed by rheumatologists working in PsA. They began a study to validate the NAPSI score, but with the aim of modifying the scoring system as required to “enhance its performance characteristics.” Digital photographs of fingernails were scored by six physicians using the original NAPSI, then two focus sessions were held to discuss problems with the scoring instrument. The biggest obstacle found to scoring accurately was the use of quadrants, which were difficult to quantify. It was also felt that the individual features of nail disease should be considered separately to improve sensitivity. The only exception to this was oil drop dyschromia which was considered along with onycholysis as they represent the same pathology. To avoid the use of quadrants, a more quantitative scale for different features of nail disease was introduced, scoring pitting, onycholysis and crumbling as 0–3, and the other features as present or absent in each nail. Therefore, the possible range of scores is 0–14 per nail, or 0–140 for all fingernails, the toenails being excluded because of the difficulties ascribing the changes to psoriasis . Comparison of mNAPSI scores showed excellent inter-rater reliability and internal consistency as well as strong correlation with physician VAS for nail severity . This was confirmed by an independent study where 10 rheumatologists and nine dermatologists, regardless of speciality, showed an excellent agreement for the mNAPSI score. Despite this, an obvious disadvantage of this score can be the time taken to complete the proforma, which may be several minutes if there is extensive nail involvement.

Nail disease

Nail involvement is a common feature of both skin psoriasis and psoriatic arthritis. It has been estimated to occur in up to 50% of patients with uncomplicated skin psoriasis and is even more prevalent amongst patients with PsA, between 63% and 83% .

Psoriatic nail disease can be broadly divided into psoriasis affecting the nail matrix and the nail bed. Involvement in the nail matrix results in changes to the nail plate. The most common change is pitting, depressions in the surface of the nail plate resulting from sloughing of abnormal cells laid down when there is involvement in the proximal matrix. If the involvement is in the distal matrix, the abnormality is seen deeper in the nail causing leuconychia. More extensive involvement can result in crumbling of the nail plate. Psoriasis affecting the matrix can be viewed directly in the lunula resulting in red spots. Involvement in the nail bed results in distinct patterns of nail disease. Onycholysis and subungual hyperkeratosis are caused by psoriasis affecting the hyponychium, the transitional area between nail bed and normal epidermis. Involvement more proximal in the nail bed can result in ‘oil drop change’, salmon spots or splinter haemorrhages.

Clearly, assessment of nail disease in PsA must take all of these features into account. Prior to the development of any systematic scoring system, many therapeutic studies of nail disease used subjective measures of nail involvement graded by the investigator as mild/moderate/severe or using a visual analogue scale (VAS).

One of the earliest published nail assessments was developed by the Bath group. Jones et al. used a predefined scoring system where each fingernail was scored for the presence of pitting, onycholysis, hyperkeratosis and dystrophy. Each of these features scored one point with a maximum score of 40 for the 10 fingernails . This scoring system was then used and adapted by the Oxford group who used the same scoring system for finger and toenails with a maximum score of 80 (the psoriasis nail severity score or PNSS) .

In 2003, the nail psoriasis severity index (NAPSI) was devised by dermatologists. This scored nails by dividing each nail into four quadrants and scoring one point for each quadrant, which has any of the features of nail bed psoriasis (onycholysis, oil drop, splinter haemorrhages and nail bed hyperkeratosis) and one point for any features of nail matrix psoriasis (pitting, leuconychia, red spots in lunula and nail plate crumbling) . This results in a score of 0–8 per nail, 0–80 for all fingernails or 0–160 if toenails are included. The advantage of this scoring system is that it does not require scoring for each feature separately but does allow differentiation between nail matrix and nail bed involvement. Initial assessment revealed reasonable reproducibility and a subsequent study showed moderate-to-good interobserver reliability . Following on from this, the NAPSI has been used as an outcome measure in trials of biological agents.

Subsequent to this, a modification, the mNAPSI was proposed by rheumatologists working in PsA. They began a study to validate the NAPSI score, but with the aim of modifying the scoring system as required to “enhance its performance characteristics.” Digital photographs of fingernails were scored by six physicians using the original NAPSI, then two focus sessions were held to discuss problems with the scoring instrument. The biggest obstacle found to scoring accurately was the use of quadrants, which were difficult to quantify. It was also felt that the individual features of nail disease should be considered separately to improve sensitivity. The only exception to this was oil drop dyschromia which was considered along with onycholysis as they represent the same pathology. To avoid the use of quadrants, a more quantitative scale for different features of nail disease was introduced, scoring pitting, onycholysis and crumbling as 0–3, and the other features as present or absent in each nail. Therefore, the possible range of scores is 0–14 per nail, or 0–140 for all fingernails, the toenails being excluded because of the difficulties ascribing the changes to psoriasis . Comparison of mNAPSI scores showed excellent inter-rater reliability and internal consistency as well as strong correlation with physician VAS for nail severity . This was confirmed by an independent study where 10 rheumatologists and nine dermatologists, regardless of speciality, showed an excellent agreement for the mNAPSI score. Despite this, an obvious disadvantage of this score can be the time taken to complete the proforma, which may be several minutes if there is extensive nail involvement.

Enthesitis

An enthesis is the site of attachment to bone of tendon, ligament or joint capsule. Enthesitis can cause pain, tenderness and swelling at entheseal sites, sometimes with resultant bony proliferation, such as bone spurs at the calcaneum. Enthesitis is recognised as a manifestation of all the spondyloarthritides, but the majority of outcome measures for enthesitis have been developed and validated in patients with AS.

The first enthesitis index was developed by Mander et al. A list of all entheses easily accessible to clinical examination was created and this was tested on outpatients with AS. Sites which scored zero on this pilot study were eliminated, leaving 66 entheseal sites (see Fig. 1 ) to be examined for tenderness and graded on a semi-quantitative score from 0 to 3 . A further study with 19 AS patients showed correlation of the Mander enthesitis index (MEI) with pain and stiffness VAS scales and a reduction in the score with non-steroidal anti-inflammatory drug (NSAID) treatment. There was some variability between different examiners performing the MEI, but intra-observer variability was not formally tested .

Enthesitis sites recorded by the principle enthesitis indices (Mander Enthesitis Index (MEI), Maastricht Ankylosing Spondylitis Enthesitis Score (MASES), Spondyloarthritis Research Consortium of Canada Enthesitis Index (SPARCC), Leeds Enthesitis Index (LEI)).

Further validation was attempted using the longitudinal OASIS cohort study designed to investigate the long-term outcome of AS. The authors provided evidence of a correlation between the MEI and other disease activity measures in AS including the Bath AS disease activity index (BASDAI). They also tried to simplify the MEI. The first step was to remove the tenderness scoring from 0 to 3 and replace it with a dichotomous 0/1 score for tenderness. The modified MEI showed similar correlation with other disease activity measures. A process of data reduction was performed using the data from patients who had an MEI > 0 to reduce the number of sites to be assessed. The most commonly involved enthesis was identified and noted, with exclusion of this enthesis in subsequent rounds. This step-wise reduction was performed until 80% of all patients with a positive MEI score were identified. Data were analysed from three time points and then, all potential sites were combined. After exclusion of entheses difficult to localise or near to other sites, there were 13 sites (see Fig. 1 ) left for assessment in the Maastricht AS entheses score (MASES) . Correlations between the enthesitis scores and disease activity measures were similar, with no marked reduction in correlation seen with the shortened MASES . Other new modifications of the MEI and distinct scoring systems have been proposed as enthesitis indices but few have been validated.

In particular, establishing criterion validity for these clinical outcome measures has been difficult because of a lack of a gold standard. Ideally, a gold standard would have associated evidence of tissue abnormality from histopathological studies. However, biopsy of tendons is neither safe nor easy, and there are limited research data available. Magnetic resonance imaging (MRI) has been shown to identify bone-marrow oedema at tendon insertions and abnormal signal around the enthesis . Ultrasound scanning using grey-scale and power Doppler to identify increased vascularisation in and around the enthesis can identify abnormal findings in symptomatic and asymptomatic entheses . Given that ultrasound identified power Doppler signal, and MRI bone-marrow oedema have been correlated with tissue evidence of inflammation in other rheumatic manifestations , it seems likely that imaging is the best gold standard available at present.

More recently, the Spondyloarthritis Research Consortium of Canada (SPARCC) created a new outcome measure for enthesitis in SpA using information from ultrasound and MRI studies to decide which entheseal sites should be included. They picked 16 sites (see Fig. 1 ) which had been identified as being commonly involved in imaging studies and could be clinically identified. Inter-observer reliability was good and a substantial correlation was seen between enthesitis score and other disease activity measures. However, the reduction in enthesitis score in patients treated with anti-tumour necrosis factor (TNF) therapy was not significant after 12 weeks of therapy . Reduced versions of the SPARCC enthesitis index using more commonly involved sites (SPARCC 8/16 score) and only sites which discriminated between treatment and placebo (SPARCC 6/16 score) showed larger effect sizes and standardised response means .

All of the entheseal outcome measures discussed previously were developed and validated principally for patients with AS. The SPARCC score was developed using a full spectrum of SpA patients, but the validation was only in patients with AS . The Leeds enthesitis index (LEI) is the only measure developed specifically for PsA. The methodology for this study was identical to that of the development of the MASES using dichotomous scoring for pain and then step-wise data reduction to identify and retain only those sites with most common involvement. This resulted in an index with six entheseal sites (bilateral lateral epicondyle, bilateral medial femoral condyle and bilateral Achilles tendon insertion – see Fig. 1 ) . This index was then compared with other entheseal indices including the MEI, MASES and others, in an open-label longitudinal study. The LEI showed closest correlation with other disease activity measures, a large effect size and the smallest floor effect when compared with the MEI. This low-floor effect means that it can identify the majority of patients with enthesitis using just six sites, making it far more feasible .

The reliability of the LEI, MASES, SPARCC and other enthesitis scoring systems was investigated in a study using AS and PsA patients. Nineteen patients were examined by 10 experienced rheumatologists. There was excellent agreement for all patients in all of the indices. In patients with AS, the MASES and SPARCC indices performed better, but in patients with PsA, the LEI and SPARCC showed excellent agreement .

The use of clinical assessment tools for enthesitis has now become widespread in clinical trials despite the debate about which particular scoring system is optimal. However, as mentioned above, there are some crucial problems with the criterion validity of these clinical techniques. Imaging studies have shown changes within tendons and around tendon sheaths at entheses but the correlation with clinically appreciable tenderness or swelling is less convincing . In addition to the soft tissue changes visualised, MRI scanning has identified the involvement of bone adjacent to the enthesis. No studies have addressed whether this can be clinically identified and whether it correlates with clinical enthesitis counts.

The other limitation of clinical enthesitis counts is the specificity of the finding of tenderness in these areas. Many of the entheseal points are relatively near joints and accepted fibromyalgia points, raising the possibility that misclassification could occur. The key to the reliability of these tools in clinical practice is the training provided to assessors in localising the correct points.