Session Type: ACR Poster Session A
Session Time: 9:00AM-11:00AM
Background/Purpose: Disease activity of polymyalgia rheumatica (PMR) (1, 2) may be evaluated using the PMR activity score (PMR-AS). For patients without measure of CRP, or patients having a treatment modifying its measure (i.e. anti-IL6), the PMR-AS could be unmeasurable or biased. Three options should be suggested: 1-to use ESR instead CRP but anti IL6 may also modify ESR; 2- to use PMR-AS without CRP (clin-PMR-AS) as an item but its validation is mandatory; 3- to replace CRP by a virtual value imputed on the basis of the other parameters (CRP-imputed-PMR-AS). Our goal was to build a PMR-AS using imputed value of CRP.
Methods: We used two independent cohorts of patients: the CRI cohort (137 visits of 89 PMR patients without any treatment or treated by corticosteroids) (2) and the TENOR cohort (20 patients, 20 visit at inclusion without any treatment, 20 visit at W4, 8, 12 during tocilizumab infusions, and 20 visit at 16, 20 and 24 weeks with treatment by steroids) (2). In the CRI cohort, we evaluated the correlation (Spearman) between the items of the PMR-AS to define which of them may be the best to perform an imputation of CRP. Then we built a scatter plot representing PMR-AS and clin-PMR-AS. Their correlation could be represented by an equation y=ax+b. We verified that the CRP-imputed-PMR-AS in patients of the TENOR cohort without treatment by tocilizumab gave a good correlation with PMR-AS with CRP. Finally, we evaluated the difference of PMR-AS without CRP and the CRP-Imputed-PMR-AS in the TENOR cohort during the visit 4, 8 and 12.
Results: On the CRI cohort, we observed a good correlation between the items of the PMR-AS, the clin-PMR-AS and PMT-AS. Agreement between PMR-AS with and without CRP was excellent (using cut off 0-7-17-∞; only 5/137 were discordant, kappa: 0.93) but as anticipated, the clin-PMR-AS was lower than the PMR-AS. On a scatterplot representing the PMR-AS (y) according to the clin-PMR-AS (x), a straight line y=1.12 x clin-PMR-AS + 0.26 represented their association, and with use it for CRP imputation. The mean +/- SD was very close between the PMR-AS, clin-PMR-AS, and CRP-imputed-PMR-AS, respectively. Using cut off 0-7-17-∞, we obtained a slightly higher concordance for the CRP-imputed-PMR-AS (kappa=0.95). This suggested that it was not really necessary to do an imputation to separate patient in groups of PMR-AS except for high clin-PMR-AS. The replication in the TENOR cohort before and after treatment by tocilizumab confirmed that CRP-imputed-PMR-AS did not modify the results using the original PMR-AS (during treatment by tocilizumab, means for PMR-AS, clinPMR-AS, and CRP-Imputed-PMR-AS were 11.3+/-8,1, 11+/-8.1 and 12.0+/-9, respectively).
Conclusion: This study supplies evidence that a CRP-imputed-PMR-AS may be used to monitor PMR activity for patients without available CRP or treated by anti IL6. Nevertheless, use of PMR-AS (or Clin-PMR-AS) in patients treated by tocilizumab lead to a very small proportion of false classification of patient splited in low-disease activity or high-disease activity. References: 1- Leeb et al. Ann rheum Dis 2004;63:1279-83. 2- Binard A et al, Arthritis Rheum, 2008, 59:263-269 3-Devauchelle-Pensec V, 2016, feb 29
To cite this abstract in AMA style:Devauchelle V, Saraux L, Berthelot JM, Cornec D, Marhadour T, Jousse-Joulin S, De Bandt M, Gouillou M, Saraux A. Polymyalgia Rheumatica Activity Score without C-Reactive Protein [abstract]. Arthritis Rheumatol. 2016; 68 (suppl 10). https://acrabstracts.org/abstract/polymyalgia-rheumatica-activity-score-without-c-reactive-protein/. Accessed December 13, 2019.
« Back to 2016 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/polymyalgia-rheumatica-activity-score-without-c-reactive-protein/