International, Practice-Based Recommendations for Prediction of Bleedings and Anti-Platelet/Anti-Coagulant (AP/AC) Co-Medications in CLL Patients Taking Ibrutinib

Introduction: bleeding events (grade 1-2, 50%, grade 3-4 <5%) have been repeatedly reported in patients taking BTK inhibitors, including ibrutinib (IBR) and acalabrutinib. We lack clear recommendations from real-life clinical practice experience, since all existing management decision support systems have been made out from clinical trials (many of which didn't allow the recruitment of patients taking AC). Furthermore in clinical trials, no thorough exploration of platelet aggregation test was available.

Methods: we reviewed the in vitro and in vivo evidence collected in 3 different centres from 3 countries (France, Russia, Austria) including: bleeding symptoms with grade, AP/AC co-medications, agonist-induced platelet aggregation tests (ristocetin RIPA, collagen CIPA), platelet counts. In the French series, complete PK evaluation (t0-0.5h-1h-2h-4h-6h) at one month was available for some patients (Cmax, Cmin, AUC).

Results: based on our first 121 pts, our aggregation tests monitoring point towards several common considerations:

1/ For patients taking IBR alone (without AP/AC):

- Half patients' platelets do not demonstrate any collagen-induced platelet aggregation (CIPA) defect with IBR, even sampled at various time points. This 50/50 segregation between IBR-sensitive and IBR-resistant pts is also observed among 47 healthy donors' platelets. Furthermore, acalabrutinib (sparing TEC kinase more than IBR) in vitro induced very similar (in time and size) aggregation defects only in the IBR-sensitive pts, reinforcing the view of a BTK-effect with the two inhibitors.

- Among the 50% early bleedings due to ibrutinib, 50% will recover over time (clinical and CIPA test recovery). As a consequence, in our global population 25-30% of patients still present bleeding after 6 months of therapy (100% grade 1-2). No PK parameter was found to explain the discrepancies of bleeding events among patients, suggesting dose reductions of IBR are not mandatory from the start to leverage bleeding risk. The late risk of severe haemorrhage (grade 3-4) seen in 5 patients seems to be linked with AC (not AP) intake.

- the CIPA test has a very good sensitivity (85.7%) and a good specificity (62.5%) to discriminate the patients with bleeding diathesis the first 3 months. The specificity (not sensitivity) was however significantly reduced (33.3%) after 3 months, compromising its utility in pts requiring AC therapy while on ibrutinib for >3 months. Yet, the large majority of patients with bleeding syndromes had a weak aggregation response to collagen 3.3 µg/ml. A lower dose of collagen in vitro can be used (2.2 instead of 3.3microg/ml) to increase specificity after 3-6 months of ibrutinib.

2/ for patients taking IBR + AP or AC

- the percentage of patients with bleeding syndromes before 6 months did not increase significantly compared to patient under IBR treatment alone (the bleeding grade did not increase neither). Conversely, after 6 months, with AP the risk increases a little bit to 40% (all grade 1-2 except 1 grade 3) as compared to 90% bleeding events in the 10 pts with AC (mainly VKA, including 2 grade 3 and 1 grade 4).

- the CIPA tests are no longer valid for patients under AP therapy before start of IBR, as their platelet aggregation response is very low, as expected. In vitro, addition of aspirin or clopidogrel to IBR-resistant platelets reduced aggregation to a significantly lesser extent than that observed in IBR-responsive platelets. This is of interest in the investigation of IBR pts requiring a dual AP therapy.

- Most patients under AC and IBR with bleedings had also weak CIPA results. This result would suggest that a weak platelet response to collagen could increase the risk of bleeding (inherent to the AC drug), and therefore heparin should be the preferred choice with ibrutinib.

- Take home message: our clinical and biological data collectively identify a strong concern with the use of a triplet therapy, IBR+AP+AC, which should be avoided.

Conclusion: the monitoring of CIPA testing is not recommended before starting IBR if no AP/AC is added. In clinical situations where AP/AC is mandated, it can predict an "IBR-related" additional risk factor the bleeding risk of the AP drug. Our results are currently being compared with those of RIPA testing (in the Austrian cohort). We will present updated recommendation charts (including direct anticoagulants use) during the meeting.