Psoriatic Arthritis Spondylitis Radiology Index: A Modified Index for Radiologic Assessment of Axial Involvement in Psoriatic Arthritis

DISCUSSION

Although differences between PsA and AS have been described, clinicians can readily appreciate that some patients with PsA have axial involvement that is indistinguishable from AS, while other patients appear to show characteristic features: asymmetry, non-marginal syndesmophytes, paravertebral ossification, and fusion of the cervical posterior elements. Clearly, any instrument designed to assess radiologic severity and progression has to take into account both the classical and differential features. The new index for assessing psoriatic spondylitis additionally and separately considers the posterior elements of the cervical spine, and thus is a simple modification of one of the existing instruments. The PASRI therefore extends and improves the radiologic assessment of axial PsA. However, ours is a preliminary study and further validation of this index is clearly required.

Incorporation of other distinctive and distinguishing features of psoriatic spondylitis was considered, but was rejected on the following grounds. There is no clear radiologic definition of non-marginal syndesmophytes: although these have been described as more bulky and distinct from the corner of the vertebral body¹, including a scoring system for these was considered to introduce too much interobserver bias. Should a valid and reliable definition be introduced it may be possible to incorporate this feature into a new radiologic score. Paravertebral ossification was also considered, but rejected on the grounds that this feature is too infrequent to be of use in such a radiologic score¹⁷. A method of scoring asymmetry (based on that suggested for peripheral joints¹⁸) was also considered but not included, as it was felt that this was a quantitative rather than a qualitative difference, as suggested for the peripheral joint involvement of PsA.

Visualization of the spinal facet joints on plain AP and lateral radiographs is sometimes difficult. This is particularly so for the thoraco-lumbar spine. However, we found little difficulty in visualizing and interpreting the cervical facet joints, probably because of the orientation of these joints to the incident beam of the radiation. The cervical facet joints are more horizontal and face slightly inferior and posterior in the transverse plane, making them easier to see in the standard lateral view. For the lumbar spine, the more vertical orientation with overlapping anterior and posterior edges makes them hard to interpret in the standard lateral radiograph. However, it is possible to visualize some of the lumbar facet joints on the AP view, but degenerative changes and the necessary oblique angle of the incident radiation (the beam is usually focused on L4) in many of the joints make complete assessment difficult. For this reason it is usual to conduct plain radiography of the lumbar facet joints with standard oblique views. Although taking oblique views of the lumbar spine would have enabled us to visualize the lumbar facet joints, we did not consider the additional radiation exposure to be justified. For the same reasons, additional views of the cervical spine were not carried out.

Although the BASRI was quicker to perform than both the PASRI and mSASSS, the scoring rules for the BASRI do not allow detailed assessment of the posterior elements. Further, if the BASRI were used it would need further modification, as it was designed to have a minimum score of 2 (at least bilateral sacroiliitis grade 2), but in PsA it is possible to have axial involvement without radiologic sacroiliitis^5,19, thus creating the chance of zero scores. In our study, nearly 10% of patients had radiologic changes in the cervical and lumbar spine in the absence of radiologic sacroiliitis.

The ASessment in Ankylosing Spondylitis (ASAS) group has endorsed the mSASSS as the preferred method to assess structural change and to assess progression of structural damage in clinical trials. This decision was based on such factors as floor and ceiling effects, validity, and sensitivity to change⁹. It is appropriate, therefore, to have a modified index for PsA based on the mSASSS. The addition of the posterior elements is a relatively minor modification, but it clearly enhances the correlations with anthropometric and, in particular, the patient reported outcomes compared to the mSASSS (Table 4). In fact, correlations between mSASSS and anthropometric measures were much worse than those reported by Chandran, et al⁸, where mean scores for both clinical and radiographic measures were comparable. However, the correlations between mSASSS and clinical measures were largely comparable to those reported by Wanders, et al⁹, where patients had both a longer duration of disease and more severe impairment. In our study the correlations between the PASRI and clinical measures were not so different from the BASRI, which is of interest, as the BASRI scoring system can incorporate an evaluation of fusion in the posterior elements, albeit in an ordinal manner.

Our study has provided information on a new radiologic score for assessing spinal structural damage in PsA that extends and incorporates features of both the BASRI and mSASSS. Both the PASRI and BASRI appear valid for the radiologic assessment, but the PASRI offers more scope for measuring change and has greater face validity. Further information is required on progression with time as it is not anticipated that improvement, due to efficacious treatment, for example, will occur in these scores.