Evidence That Baseline Levels of Low-Density Lipoproteins Cholesterol Affect the Clinical Response of Graves' Ophthalmopathy to Parenteral Corticosteroids

Abstract

Background: High dose intravenous glucocorticoid (ivGC) therapy is the first line treatment in moderate to severe Graves' ophthalmopathy (GO) and is associated with a clinical response rate ranging from 50% to 80%. Recently, a positive correlation between total cholesterol and low-density lipoproteins cholesterol (LDLc) with GO presentation and activity has been described.

Objective: We aimed at evaluating whether, in patients with moderate to severe active GO treated with ivGC therapy, cholesterol, and LDLc could represent valuable predictive factors of medium-term GO outcome.

Methods: This single center retrospective study was conducted in a consecutive series of 87 patients undergone ivGC therapy because affected by moderate to severe active GO. Clinical outcome of GO was evaluated at week 6 (W6) and 12 (W12) in respect to baseline conditions (week 0) by the seven points CAS according to EUGOGO recommendations. Univariate analysis and binary logistic regression were performed for the outcome variable W12CAS.

Results: In patients with active GO, an early positive clinical response to ivGC therapy (as evaluated by CAS at 6W) was a strong determinant (OR=13) of the clinical outcome at week 12. Moreover, high levels of LDLc at baseline were positively associated with a reduction in the likelihood of being classified as improved at 12W. Patients with LDLc >193.6 mg/dl were very likely to respond negatively to ivGC therapy independently from the response at 6W. Based on these results, we propose a predictive decision-making model to be tested in future prospective studies.

Discussion: We found that, in patients with active GO, both an early clinical response to ivGC therapy and baseline LDLc levels are significant determinants of GO outcome (W12CAS). These data support the need of a cholesterol-lowering treatment before addressing these patients to ivGC therapy.

Keywords: Graves’ disease; Graves’ ophthalmopathy; cholesterol; low-density lipoprotein cholesterol; parenteral corticosteroids.

Figure 1

Stacked histogram of low-density lipoproteins cholesterol (LDLc) levels grouped for W12CAS categories.

Figure 2

Receiver operating characteristic (ROC) curve of the predicted probability of improvement at W12CAS calculated with binomial logistic regression.

Figure 3

Suggested decision-making algorithm. The graphs are constructed on the basis of binomial logistic regression prediction using the Youden’s best cut-off P(I_W12CAS) = 0.66. Each dot refers to a single patient; green dots refer to patients classified as Improved at W12CAS, while black dots refer to Not Improved Patients. The predicted probability of improvement at W12CAS [P(I_W12CAS)] provided by regression analysis is positioned on the x axis; cut-off = 0.664 (dotted line). The continuous predictive variable included in the model (LDLc) is positioned on the Y axis. The categories of the dichotomous predictive variable (W6CAS outcome) are highlighted in green (I_W6CAS) or red (NI_W6CAS) contour lines.

Figure 4

Steps of the decision-making algorithm.

Figure 5

Mechanistic hypothesis of relationship of hypercholesterolemia with chronic inflammation and resistance to corticosteroids effects in patients with Graves’ ophthalmopathy (GO).