Presentation: Can Pain Sensitivity SNPs be Replicated in Rheumatoid Arthritis (RA) Patients? (2007)

1064 Can Pain Sensitivity SNPs be Replicated in Rheumatoid Arthritis (RA) Patients?

Purpose: A recent report of a genetic component to inflammatory pain sensitivity (Tegeder, 2006) holds promise in advancing the treatment of RA pain. This study seeks to confirm the role of several SNPs reported to influence pain perception in a sample of RA patients.
Methods: The sample included Caucasian CCP+ patients who met ACR criteria for rheumatoid arthritis enrolled in a large single center RA longitudinal cohort study. At baseline and at up to three yearly follow-up visits anti-CCP antibodies and C-reactive protein level were measured; an interview asked about pain sensitivity and whether the patient had experienced a flare in the last 6 months; and a rheumatologist performed a joint exam. The question “How much arthritis pain do you feel today?” is the primary outcome and is measured on a scale of 0 (no pain) to 10 (severe pain).
Samples were genotyped using the Affymetrix 100K chip. Of the 152 SNPs purported to regulate inflammatory and neuropathic pain sensitivity, 3 were genotyped on the 100K platform (rs7142517, rs10483639, rs998259). Using Haploview software, we found one additional SNP, rs10498472, in linkage disequilibrium at r-squared > .8 with the remaining 12 SNPs.
We used four generalized estimating equation (GEE) models adjusted for repeated measures to test whether each of the 4 SNPs predict pain sensitivity, adjusting for potential confounders.
Results: Four hundred twenty patients (mean baseline disease duration 17.5 yrs, sd 12.5, 82% female, mean age 58.9 sd 13.3) contributed from 1 to 4 visits for a total of 1187 observations. The mean baseline pain score was 3.3, sd 2.8. In the unadjusted GEE models none of the SNPs met a Bonferroni adjusted significance criteria of 0.006: rs7142517 p=0.14, rs10483639 p=0.27, rs10498472 p=0.79, and rs998259 p=0.94. To test whether the SNPs influenced pain perception given the same level of pain, we adjusted for objective measures of disease activity (DAS28-CRP assessed yearly and any flares in the last 6 months), age, and gender. In the adjusted model none of the SNPs exhibited a significant association with self-reported pain (p= 0.64, 0.23, 0.52, and 0.81 respectively).
Conclusions: We were unable to replicate the report that these SNPs are related to pain sensitivity in a population of RA patients. Since the Affymetrix chip only included 3 of the15 purported pain-regulating SNPs, further research is needed to see whether the other SNPs can be replicated in RA patients.

R.J. Glass, Millennium Pharmaceutical, 2 Research grants; Biogen IDEC MA, Inc., 2 Research grants; Bristol Myers Squibb Foundation, 2 Research grants.