Repurposing established drugs for new indications

Leflunomide and hydroxychloroquine combination therapy for Sjögren’s syndrome

Eefje Otten - van der Heijden

Prof.dr T.R.D.J. (Tim) Radstake
Dr J.A.G. (Joel) van Roon & dr A.A. (Aike) Kruize
October 27, 2020
16:15 h


The immunopathology of pSS is characterized by broad immune-activation involving a wide variety of immune cells, ultimately resulting in B cell hyperactivity and the formation of ectopic lymphoid structures in a part of the pSS patients. Since effective therapeutic options are still lacking, symptoms may only be (partly) relieved by symptomatic measures. Considering the complex interplay of all involved immune cells and their effector molecules, the central hypothesis to this thesis is that successful inhibition of immune-activation in pSS requires a combination of drugs targeting both overlapping and distinct immunopathological pathways, jointly resulting in broad immune inhibition.

Following our hypothesis, the combination of the established cDMARDs leflunomide and hydroxychloroquine may hold promise of clinical efficacy, based on their well characterized properties. Thus, in this thesis, the efficacy (in vitro and in vivo) and safety of leflunomide and hydroxychloroquine combination therapy in pSS patients was investigated.

A detailed overview of the current knowledge on pSS immunopathology is provided in Chapter 2, and the place of cDMARDs in its treatment is described. A great deal of information on immunopathology of pSS remains to be elucidated, however significant progress has been made in recent years. Central to the process is focal infiltration with mononuclear cells of target organs, interacting in a complex manner and jointly inducing B cell hyperactivity and formation of auto-antibodies. Great potency was expected from bDMARDs. However, clinical studies have failed to show clinical efficacy up to now. Also cDMARDs have been investigated in pSS, but all together knowledge on cDMARDs in pSS is limited due to the lack of standardized inclusion criteria and outcome measures and lack of knowledge on safety profiles. In addition, there is little information on optimal dosing in pSS. In recent years, knowledge on molecular pathways underlying pSS immune pathology has emerged, creating new opportunities for ‘old’ drugs to be repurposed. Given the upregulation of potential additive pathways in pSS, synergism between the different cell-types involved is likely to occur. Consequently, combining drugs targeting several pathways underlying
T-cell, B-cell and pDC (and other cell types such as NK and ILC cells) activation seems reasonable. In this respect, a combination therapy with two complementary cDMARDs holds promise. Leflunomide and hydroxychloroquine may be a good combination, with leflunomide targeting mainly activated T-cells and to a lesser extent B-cells and, with HCQ inhibiting mainly B-cells and pDCs that can be activated by TLR7 and 9-induced immune activation.

In Chapter 3, we showed in an in vitro system that TCR/TLR9 activation of PBMCs induced strong proliferation of T and B-cells and production of CXCL13, IFN-α, IFN-γ, IgG and IgM. Leflunomide dose-dependently inhibited all measured parameters, where HCQ potently and dose-dependently decreased B cell proliferation, CXCL13, IFN-α, IgG and IgM production. At different concentration combinations leflunomide and hydroxychloroquine inhibited several immune hallmark features more potently than each single compound. Clear additive inhibition of T- and B-cell proliferation and CXCL13 production was seen using suboptimal dosages of leflunomide and hydroxychloroquine. IFN-α and B-cell activity, reflected by immunoglobulin production, were already potently inhibited by this concentration of hydroxychloroquine alone, therefore the assumed additive effect could not be seen. Additive inhibition of IFN-γ required higher dosages of leflunomide and hydroxychloroquine, dosages that might be achieved less easy in vivo. Thus, a combination of leflunomide and hydroxychloroquine at clinically applicable concentrations additively inhibits immune activation in vitro/ex vivo, supporting a potential implementation of this drug combination in pSS treatment.

Given the well-described mechanisms of action and anti-inflammatory activities of leflunomide and hydroxychloroquine in literature and the promising results of leflunomide and hydroxychloroquine combination therapy in vitro, a placebocontrolled, double-blinded randomized trial investigating this therapy in pSS patients was performed (Chapter 4). The study included 29 pSS patients with active disease, reflected by an ESSDAI score of 5 or greater. 21 patients were treated with leflunomide/hydroxychloroquine combination therapy, 8 patients were assigned to placebo treatment. Disease activity reflected by ESSDAI score, the primary endpoint, significantly decreased in the leflunomide/hydroxychloroquine arm compared to the placebo arm. Also other clinical parameters such as the patient reported outcome ESSPRI score including its constituents ESSDAI pain and ESSDAI fatigue, and saliva output improved. Histological assessment by parotid biopsies supported these positive findings. Leflunomide/hydroxychloroquine combination therapy proved to be safe and well-tolerated. In addition to these results, response to therapy could be predicted with high accuracy by a model comprising ten circulating proteins. 

In Chapter 5, five different IFN- associated biomarkers (IFN-score in whole blood and in PBMCs, MxA in whole blood, CXCL10 and Galectin-9 in serum) in patients treated with either leflunomide/hydroxychloroquine or placebo were assessed and their potential for treatment monitoring was investigated. At baseline, these biomarkers correlated with each other and also modestly with the ESSDAI scores. At the end of the study, all biomarkers with the exception of IFN-score in PBMC were significantlydecreased compared to the baseline value in the leflunomide/hydroxychloroquine treated group. Only for protein biomarkers MxA, CXCL10 and Galectin-9 a correlation with changes (delta) in ESSDAI values at 24 weeks was seen. The presence of an IFNsignature at baseline, based on the IFN-score, was not predictive of clinical response. In contrast, patients responding to the therapy showed strong early decreases of MxA and Galectin-9 levels from week 8 onwards, whereas non-responding patients showed changes comparable to those of placebo-treated patients. Subsequent ROC analysis revealed good predictive values for early changes in (delta) MxA and Galactin-9 levels after 8 weeks, the latter outperforming the former.

In Chapter 6 we focused on Galectin-9 as a biomarker that reflects the IFN signature in patients with SLE, APS and pSS. The findings of this chapter are in line with previous findings from our group showing that Galectin-9 is an easy to measure and accurate biomarker for the IFN signature in SLE and APS patients and correlates with disease activity measured by the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI). Usage of Galectin-9 as a biomarker for the IFN signature needs further confirmation, testing generalizability. Data from this thesis indicate that at the least Galectin-9 might play an additional role in monitoring of pSS immunopathology and disease activity. Higher levels of Galectin-9 were found in pSS patients compared to non-Sjögren sicca patients, correlating with ESSDAI scores and serum IgG levels. Leflunomide / hydroxychloroquine treatment induced a decline in Galectin-9 levels, parallel with changes in B cell hyperactivity, disease activity measured by ESSDAI and ESSPRI, and markers of IFN activity. This warrants further study of Galectin-9 as a biomarker for disease monitoring in pSS.