Ristocetin-Induced Platelet Aggregation for Monitoring of Bleeding Tendency in Ibrutinib-Treated Patients with Chronic Lymphocytic Leukemia

Background. Inhibition of Bruton´s tyrosine kinase (BTK) with the small molecule ibrutinib has significantly improved the survival of patients with chronic lymphocytic leukemia (CLL). BTK is also expressed in platelets. Collagen- and von Willebrand Factor (vWF)-dependent (ristocetin-induced) impairment of platelet function has recently been described (Levade M et al., Blood 2014, 124:3991-5;Kamel S et al., Leukemia 2015, 29:783-787) . Bleeding events were observed in 61% of patients in a recently published 3 year follow-up (Byrd JC et al., Blood 2015, 125:2497-2506).  Bleeding under ibrutinib is generally mild (CTC grade 1-2 corresponding to spontaneous bruising or petechiae), but grade 3 or 4 bleeding can be observed, particularly after trauma. We hypothesized that quantitative assessment of platelet aggregation in ibrutinib CLL patients could help (1) to predict bleeding tendency, and (2) to guide patients through invasive procedures.

Patients and Methods. Twenty-four adult patients with previously treated CLL (16 male/8 female, median age 67 years, range 55-84) received ibrutinib orally at a planned dose of 420mg/day and were regularly monitored and thoroughly investigated for bleeding tendency. The median time on ibrutinib was 7.5 months, (range 1-27). Bleeding events (any CTC grade) occurred in 13 (54%) and dose-reductions to 280 (N=12) or 140mg (N=3) (for bleeding, infections, or neutropenia) were made in 15 (63%) of patients during a median observation period of 5 months (range 1-12). Bleeding was observed in 4 of 6 patients with concomitant anticoagulation. Of note, only 1 of the 24 patients had a CTC grade 3 bleeding event, and no grade 4 or 5 events were observed. Ristocetin-induced platelet aggregation (RIPA, herein referred to as RCoF) was quantitatively measured in fresh hirudin-blood by whole blood aggregometry with a Multiplate® Analyzer (Roche Diagnostics). Platelet aggregation was expressed in AUC units (U) (normal range 98-180U). Controls included normal subjects (N=53). Consecutive samples before and during treatment were available in all patients. Statistical methods comprised t-Test and ANOVA using SAS.

Results. Ristocetin-induced platelet aggregation was already diminished before ibrutinib treatment (median 51 RCoF U) when compared to normal controls (Table 1). This is likely due to lower platelet counts in CLL patients influencing overall platelet aggregability (Hanke AA et al.,  Eur J Med Res 2010, 15:214-219). During ibrutinib treatment, platelet aggregation was substantially impaired (median of 22U). A direct comparison of available paired samples in 5 patients showed a significant decrease after ibrutinib initiation (51 to 14.5U; p=0.0028).  Of note, significantly lower values were measured at visits when bleeding events were documented (N=34) compared to patient visits without bleeding tendency (N=70) (median 13 vs. 42U; p<0.001). The median RCoF value was lower in patients with CTC grade > 2 (N=10) vs. <2 bleeding (11 vs. 14U). Similar results were obtained for collagen-dependent platelet function (bleeding vs. no bleeding: 17 vs. 19.5U; p=0.002). RCoF values were correlated with platelet count (r²=0.34; p<0.0001) at median values of 103 vs. 138 G/L in patients with or without bleeding, respectively. There was also a significant difference between the lowest RCoF values in individual patients with or without bleeding (7.5 vs. 16.5U; p=0.027) (Figure 1). No bleeding event was observed in patients whose lowest RCoF value was greater than 25U. Long-term kinetics of vWF-dependent platelet function was assessed in 7 patients and corresponded with ibrutinib dose. When ibrutinib was stopped, recovery of RCoF to greater than 70U was observed in as little as 48hours, suggesting a short time to normalization of platelet function.

Conclusion. These data indicate that quantitative assessment of vWF-dependent platelet function in ibrutinib treated patients may serve to monitor therapy particularly in the setting of bleeding tendency, anticoagulation, or planned invasive procedures. Further evaluation of platelet function as a pharmacodynamic marker seems warranted.

Table 1. VWF-dependent platelet function (RCoF) in normal subjects or CLL patients before and during ibrutinib treatment with or without bleeding.

Figure 1. Lowest RCoF values in individual patients with or without bleeding.