0528: Apolipoprotein A1 and B Levels Indicate Specific Lipid Changes in Early Treatment Naïve Psoriatic Arthritis and Correlate with 1-year Disease Activity

Background/Purpose: Psoriatic arthritis (PsA) is associated with increased cardiovascular risk and a pro-atherogenic lipid profile. We hypothesize that alterations in lipid metabolism combined with systemic inflammation are present in the early disease stages of the disease and reflect pathophysiologic, metabolic changes characteristic for PsA. We investigated serum lipid profile in early PsA and examined whether changes in the profile or apolipoproteins are specific to early disease.

Methods: Adult patients with newly diagnosed treatment naïve PsA were included in the multicenter cohort study (N=75). Clinical, demographic characteristics and comorbidities were collected. The serum lipids were measured: total cholesterol (TC), non-HDL-cholesterol (non-HDL-c), LDL-cholesterol (LDL-c), HDL-cholesterol (HDL-c), triglycerides (TAG), apolipoprotein B (ApoB) and apolipoprotein A1 (ApoA1). The results were compared to sex- and age matched healthy controls (HC) (N=61) and rheumatoid arthritis (RA) patients (N=50)(Kruskall-Wallis). The lipid measurements were re-evaluated in PsA patients after 1 year follow-up.

Results: The three groups were matched by age; HC and PsA patients were matched by age and sex. PsA patients had mainly oligoarticular disease, low PASI scores and a short symptom duration (median: 0.6 years). The mean DAPSA score (SD) was 19.5(10.8), corresponding to moderate disease activity. Early RA patients (disease duration (median: 0.5 years) all had polyarticular disease and moderate disease activity, mean(SD) DAS-28-CRP of 4.61(2.0).


Among lipid measurements, HDL-c levels were lower in PsA than in HC and RA (df2, c²10, p=0.006, PsA vs HC p=0.013) (Fig 1A-E), the levels of TC, LDL-C, TAG were not different in the three groups. Significant differences in ApoA1 and ApoB levels were observed between PsA, RA and controls. ApoB was higher in PsA than in RA patients but lower than in controls (df2, c²43.8; p< 0.001). ApoA1 was markedly lower in PsA patients compared to both RA and controls (df2, c²118.9; p< 0.001)(Fig 1 F-H). In regression models, the levels of ApoA1, adjusted for age, sex, BMI, statin use, smoking status, CRP levels and interaction between them, were predictive of PsA diagnosis with 90.6% accuracy. In ROC analysis, ApoA1 was predictive of the diagnosis of PsA with a specificity of 82.4% and a sensitivity of 83.8% at an optimal cut-off value of 1403 µg/ml (AUC[95%CI] 0.886[0.83-0.941]) (Fig 2). At 1 year follow-up only ApoA1 levels, but not other lipid measurements, increased in PsA patients achieving DAPSA LDA, while ApoB levels remained comparable to baseline (Fig 3).

Conclusion: Decreased levels of ApoA1 and increased levels of ApoB are specific to early and treatment naïve patients with PsA and distinguish them from early RA patients and non-arthritic controls. In particular, circulating levels of ApoA1 are strongly predictive for the diagnosis of PsA. The ApoA1 levels increase after 1 year in PsA patients achieving DAPSA low disease activity suggesting link with active disease. These findings suggest that specific changes in lipid profile are present in early stages of PsA and that apolipoprotein measurements might be better markers of pro-atherogenic profile in PsA than classical serum lipids.