Background/Purpose: A positive MRI at baseline is a strong predictor for a positive follow-up MRI¹. But many questions about the volatility of the lesions over short follow-up periods remain unsolved. Is it possible for lesions to disappear, occur or move localisation? Or is their location rather consistent over time?

Objectives: To describe if and how the location of active inflammatory lesions change and if lesions can occur or disappear after 3 months without changing treatment.

Methods: 90 patients with chronic back pain (≥3 months, ≤2 years, onset ≤45 years) who were included in the SPondyloArthritis Caught Early (SPACE)-cohort underwent a MRI of the SI joints (MRI-SIJ) at baseline and after 3 months follow-up.

Results: Out of all the patients 66/90 (73.3%) did not show any lesions at any of the time points. In 12 patients no difference was found between baseline and follow-up MRI with respect to the SI joint (only left side affected in 7 patients, only right side in 2 and in 3 patients both SIJ were affected). In 5 patients without lesions at baseline new inflammation occurred (4 times only in the right SIJ and once in both SIJs). In 4 patients with inflammation at baseline the inflammation was no longer present at follow-up. In 3 out of 6 patients with inflammation in both SIJ at baseline, the inflammation was only present in the left SIJ after 3 months. The other 3 patients presented inflammation in both SIJ at baseline as well as follow-up. 24/90 (26.7%) patients had a positive MRI at baseline or follow-up, or both. The quadrant in which the lesion was present did not change over time in 6 patients. Lesions disappeared from quadrants in 9 patients (lesion disappeared in only 1 quadrant in 7 patients and in 3 quadrants in 2 patients) and occurred in 7 patients (MRI changed from negative to positive in 5 and remained positive in 2 patients). The lesions moved between quadrants in 2 patients. In these 2 patients lesions disappeared from one quadrant and occurred in another. In 20/24 patients the medication between baseline MRI and follow-up did not change (18 patients used NSAIDs and 2 of them did not use any medication) and 4 patients changed medication (2 patients switched NSAID type and the other 2 started NSAID treatment in that period). Out of the patients that changed from a negative to positive MRI (n=5) or visa versa (n=4), only 1 patient also changed medication, by switching to another NSAID.

Conclusion: BME lesions on MRI occur or disappear at SIJ level in 9% of the patients with chronic back pain after 3 months. From all patients with a positive MRI at baseline, lesions did not change location in 50% of them at SIJ level while, at quadrant level, less than 30% of the patients showed stability in the location of lesions. So, there is quite some volatility of lesions over a short follow-up period of 3 months only.