Isoquercetin for thromboinflammation in sickle cell disease: a randomized double-blind placebo-controlled trial

Abstract

Data from a small trial in patients with cancer suggest that isoquercetin (IQ) treatment lowered thrombosis biomarkers and prevented clinical thrombosis, but, to our knowledge, no studies of IQ have been conducted to target thromboinflammation in adults with sickle cell disease (SCD). We conducted a randomized, double-blind, placebo-controlled trial in adults with steady-state SCD (hemoglobin SS [HbSS], HbSβ0thal, HbSβ+thal, or HbSC). The primary outcome was the change in plasma soluble P-selectin (sP-selectin) after treatment compared with baseline, analyzed in the intention-to-treat population. Between November 2019 and July 2022, 46 patients (aged 40 ± 11 years, 56% female, 75% under hydroxyurea treatment) were randomized to receive IQ (n = 23) or placebo (n = 23). IQ was well tolerated and all the adverse events (AEs; n = 21) or serious AEs (n = 14) recorded were not attributable to the study drug. The mean posttreatment change for sP-selectin showed no significant difference between the treatment groups (IQ, 0.10 ± 6.53 vs placebo, 0.74 ± 4.54; P = .64). In patients treated with IQ, whole-blood coagulation (P = .03) and collagen-induced platelet aggregation (P = .03) were significantly reduced from the baseline. Inducible mononuclear cell tissue factor gene expression and plasma protein disulfide isomerase reductase activity were also significantly inhibited (P = .003 and P = .02, respectively). Short-term fixed-dose IQ in patients with SCD was safe with no off-target bleeding and was associated with changes from the baseline in the appropriate direction for several biomarkers of thromboinflammation. The trial was registered at www.clinicaltrials.gov as #NCT04514510.

Graphical abstract

Figure 1.

Effect of IQ on plasma sP-selectin. Mean change from baseline in plasma sP-selectin level in participants from the IQ group and placebo group reveals no significant differences (IQ, n = 23; placebo, n = 23; ANCOVA, P = .64).

Figure 2.

Effects of IQ on whole-blood coagulation. Mean change from baseline in (A) clot initiation time (reaction-time or R-time) and (B) clot formation time (K time) showing significant prolongation from the baseline value after IQ treatment (ANCOVA, ∗P = .04 and ∗P = .02, respectively). (C) Fibrin crosslinking determined by the α-angle acuteness signified by the maximal amplitude showing significant reduction from the baseline value after IQ treatment (ANCOVA, ∗∗∗∗P = .0001). However, (D) clot strength (MA) was not affected by IQ treatment (P = .09). (E) Mean change from the baseline in whole-blood CI assessed by TEG showed a significant reduction in the IQ group compared with in the placebo group (ANCOVA, ∗P = .03; IQ, n = 22; placebo, n = 22). AU, arbitrary units; MA, maximum amplitude.

Figure 3.

Effects of IQ on platelet aggregation. Mean agonist-induced platelet aggregation shown as change from the baseline in impedance whole-blood aggregometry from patients from the IQ group (n = 22) and the placebo group (n = 22). IQ treatment only shows a significant effect on platelet aggregation after exposure to (A) low doses of collagen (ANCOVA, ∗P = .02; IQ, n = 21; placebo, n = 21); however, after standard platelet aggregation, agonists such as (B) thrombin, (C) adenosine 5′-diphosphate (ADP), (D) arachidonic acid, and (E) high dose of collagen (ANCOVA, P = .93, P = .72, P = .63, and P = .66; respectively) did not show any effect.

Figure 4.

Effects of IQ on TF antigen, activity, and gene expression. (A) Mean change from the baseline in the number of TF⁺ MVs showing reduction in both IQ and placebo groups but a significantly greater reduction in the placebo group (∗P = .02). (B) Mean change from the baseline in TF⁺ MV procoagulant activity (PCA), which is decreased in both groups but shows no significant difference (P = .51). (C) LPS-induced TF mRNA expression in PBMCs isolated from patients with SCD in the IQ group at baseline (BL) and after treatment (PT) showed a significant reduction in inducible TF mRNA after IQ treatment (∗∗P = .003; n = 20).

Figure 5.

Effect of IQ on plasma PDI antigen and plasma PDI activity. (A) Mean change from baseline in plasma PDI antigen was not different between the patients treated with IQ and those treated with placebo (ANCOVA, P = .52; IQ, n = 22; placebo, n = 22). (B) PDI reductase activity (mean change from the baseline value) is significantly decreased in the patients treated with IQ compared with those treated with placebo (ANCOVA, ∗P = .02; IQ, n = 22; placebo, n = 22).