EHA 2026 preview – BeOne versus Nurix in the front line

If at last year’s ASH conference BeOne’s BTK degrader BGB-16673 opened up the slightest of leads over its rival, Nurix’s bexobrutideg, in chronic lymphoblastic leukaemia patients relapsed on a BTK inhibitor, a new battle is now brewing.This concerns CLL patients naive to BTK inhibition, and here abstracts revealed for the upcoming European Hematology Association meeting appear to give a slight advantage to Nurix. A separate question is whether the two molecules’ efficacy in first-line CLL is better than that of a typical BTK inhibitor like Brukinsa or Imbruvica.Certainly, it’s far too early to conclude that bexobrutideg is better than BGB-16673; the numbers of patients involved are tiny, the response rate data are nearly identical, and not enough has been disclosed in the abstracts to gauge whether BeOne might again be troubled by a toxicity profile that previously seemed a little worse than Nurix’s.The data being presented at EHA concern BTK inhibitor-naive cuts of the same two studies that earlier yielded results in BTK-pretreated patients: BeOne’s Cadance-101 and Nurix’s NX-5948-301.The headline numbers are best response rates of 86% for BGB-16673 and 84% for bexobrutideg – effectively identical and in similar numbers of patients (22 and 19 respectively). However, what makes Nurix look good is that BGB-16673 has had nearly double the median follow-up for its responses to develop, and that bexobrutideg is said to be well tolerated at the same 600mg dose highlighted in the relapsed setting.Meanwhile, BeOne is focusing on 200mg, perhaps mindful of earlier toxicity concerns. BGB-16673’s EHA abstract does spell out some adverse events, but to be fair not enough is being disclosed at present about bexobrutideg’s toxicity profile to make an accurate comparison.Mizuho analysts, who cover Nurix, said the data continued to suggest an “opportunity for bexobrutideg to differentiate on safety”. All bexobrutideg and BGB-16673 responders are said to be in ongoing response at the data cutoffs.Treatment-naives?So far so good, but what might have flattered Nurix’s dataset is the inclusion of patients naive to any systemic therapy, who make up half the 19-strong dataset, and would naturally be expected to do better than those who had been pretreated. Still, it’s not clear that BeOne’s BTK-naive dataset doesn’t include treatment-naives – just that none are spelled out.Either way, both BTK degraders in these early CLL settings look to be in line with BTK inhibitors: in first-line CLL BeOne’s Brukinsa scored a 93% ORR in the Sequoia trial, while AbbVie/J&J’s Imbruvica yielded 82% in Resonate-2.BGB-16673 and bexobrutideg are both in phase 3 studies in patients relapsed on a covalent BTK inhibitor, but no pivotal plans in the BTK-naive setting have yet emerged. Cross-trial comparison in CLL patients naive to BTK inhibition BGB-16673 (BeOne)Bexobrutideg (Nurix)StudyCadance-101NX-5948-301Dose200mg600mgPts naive to prior BTKi...2219...of which naive to any systemic treatmentNone disclosed10Best response rate in BTKi-naive pts86%84%Median follow-up8.3 mth4.9 mthSafetyGr≥3 TEAEs 33%; gr≥3 neutropenia/neutrophil count decrease 15%; no cases of major haemorrhage, opportunistic infections or febrile neutropenia“Continues to be well tolerated in pts treated across all doses”Discontinuations owing to TEAE6% (1 death deemed not treatment-related)Not disclosedData cutoff15 Dec 20251 Jan 2026Source: EHA.