Background Cryotherapy is widely and empirically used in an adjuvant and symptomatic setting in inflammatory rheumatic diseases, with a low level of evidence.

Objectives The aim of this work was to evaluate the effects of two local cryotherapy (LC) techniques (ice or hyperbaric cold CO2) applied twice on pain and pro-inflammatory cytokine levels in arthritic knees.

Methods Thirty-five patients suffering from non-septic knee arthritis were included (16 gouts, 8 calcium pyrophosphate dehydrate crystal deposition diseases, 5 spondyloarthritides, 6 rheumatoid arthritides). Patients didn't receive any concurrent anti-inflammatory drug (corticosteroids, NSAIDs, colchicine, DMARDs, biologics) throughout the study period (24 hours). They were randomized to receive local ice (for 30 minutes) or cold pulsed CO2 (for 2 minutes - Cryo+® Cryonic) at 9 A.M and 5 P.M. Synovial fluid was collected just before the first cold application then 24 hours later (also at 9 A.M). Synovial fluid cell count and cytokine levels (IL-6, IL-1β, TNF-α, IL-17A, VEGF by Multiplex flow cytometry – CBA® BD biosciences) were measured and compared before and after the two LC applications. Paired Wilcoxon tests were performed. The factors associated with cytokine level evolution were assessed (uni- then multivariate analysis by logistic regression).

Results Pain VAS significantly decreased after 24 hours in both treatment groups (Ice: 5.3±1.3 versus 2.7±1; n=20;p=0.0003 – CO2: 5.6±1.4 versus 3.1±1.2; n=15;p=0.002). Synovial fluid leukocyte count non-significantly decreased (23580±10015 versus 19120±8981; p=0.49). IL-6 levels significantly decreased after 2 LC applications 456100±163570pg/mL versus 281500±142006pg/mL; n=35; p=0.0001). This decrease was also significant in ice-treated subgroup (n=20; p=0.0002) but not in CO2-treated patients (n=15; p=0.07). In multivariate analysis, IL-6 decrease was associated with a higher CRP rate at baseline (r=0.35; p=0.004) and a greater skin temperature drop during LC (r=0.33; p=0.03). IL-1β levels also decreased after treatment (26.75±8.95pg/mL versus 23.10±10.11pg/mL; n=35; p=0.02) but the difference didn't reach significance in treatment subgroups (Ice: n=19; p=0.07 – CO2: n=15; p=0.15). In multivariate analysis, IL-1β decrease was associated with “ice” modality (r=3.25; p=0.03) and negatively associated with BMI (r=-4.02; p=0.04). VEGF levels also decreased globally (1391±295pg/mL versus 1263±281pg/mL; n=34; p=0.03) but not significantly in subgroups (Ice: n=19; p=0.2 – CO2: n=15; p=0.07). IL-17A rates also tended to decrease (16.3±15.75pg/mL versus 14.95±15.2pg/mL; n=35; p=0.065). TNF-α levels tended to decrease in CO2 group (n=15; p=0.08).

Conclusions LC (ice or pulsed CO2) applied twice during one day showed anti-inflammatory effects by reducing synovial IL-6, IL-1β and VEGF levels in arthritic knees. Both techniques equally reduced pain but sample sizes were still insufficient to show any difference in terms of local cytokine level reduction.

Disclosure of Interest None declared