S151: combination of large spleen and low in vivo kinase inhibition is an early predictor of inferior molecular response and poor outcomes in chronic myeloid leukaemia

Background: Variability in the molecular response to frontline tyrosine kinase inhibitor (TKI) therapy in chronic phase chronic myeloid leukaemia may be driven by differences in the level of kinase inhibition induced. Crkl (Crk-like protein) is an immediate downstream substrate for the BCR::ABL1 protein. Comparing phosphorylated Crkl (pCrkl) to unphosphorylated Crkl provides a practical surrogate measure of kinase activity of BCR::ABL1 achieved on therapy. We have developed an in-house assay that measured in vivo BCR::ABL1 kinase inhibition (IVKI) after the first week of TKI treatment. Aims: To investigate the relationship between IVKI in patients (pts) measured in mononuclear cells after the first week of TKI treatment and subsequent molecular response and other response predictors. Methods: Blood at diagnosis (baseline, pre-TKI therapy) and after the first 7 days of treatment were collected for IVKI from 173 adult CP-CML pts enrolled in the Australian TIDEL-II study, who started treatment with 600mg imatinib/QD, and 73 ENESTxtnd patients, started on 300mg nilotinib BID. IVKI was defined as the percent reduction of pCrkl:Crkl in mononuclear cells from day 0 to day 7. Molecular response (MR) rates were calculated by cumulative incidence and compared using the Fine and Gray model. Two-group comparison was performed using the Mann-Whitney U test, Fisher’s exact test or two-sample Z test proportion. Results: Thirty-nine of 173 pts (23%) treated with imatinib 600mg/QD had low IVKI (<11% reduction from baseline level); this was associated with slower reduction of BCR::ABL1 over the first 3 months (BCR::ABL1 halving time: 24 vs 16 days, p<0.001), higher rates of failure to achieve early molecular response (EMR, <10% BCR::ABL1: 23% vs 9%, p<0.001, Fig 1A); lower rates of major molecular response (MMR:72% vs 85%, p=0.03), and MR4.5 by 36 months (26% vs 44%, p=0.037), compared to high IVKI pts. Low IVKI was significantly more common in imatinib recipients with the e13a2 transcript compared to e14a2 recipients (31% versus 18% p=0.04). Achieving IVKI levels much higher than 11% appeared to provide no additional benefit. Spleen size was significantly associated with IVKI (p=0.001). We divided spleen size into normal (0cm), palpable (1-9cm) and large (≥10cm). Spleen size was associated with EMR failure (Fig 1B). Notably 55% of pts with large spleen and low IVKI experienced EMR failure whereas the EMR rate in pts with large spleen and high IVKI was 12% (p=0.014, Fig 1C). Furthermore, pts with large spleen and low IVKI had higher incidence of blast crisis progression (18% vs 1% vs 5%, p=0.036), inferior MMR (45% vs 89% vs 78%, p=0.0009) and MR4.5 (0% vs 53% vs 30%, p=0.0002) compared to pts with normal spleen and high IVKI, and the remaining pts respectively. In nilotinib-treated pts (n=73), only 4% had low IVKI (n=3). This in vivo data further supports the notion that nilotinib is more effective at targeting BCR::ABL1 kinase activity than higher dose imatinib, consistent with the finding that nilotinib leads to faster and deeper molecular responses. It further suggests that IVKI is of limited relevance in the nilotinib setting because low IVKI is rarely observed. Summary/Conclusion: IVKI provides clinically relevant assessment of the adequacy of kinase inhibition achieved in the first week of imatinib therapy. This opens a unique opportunity to optimise the TKI dose or consider switching to a more potent TKI well before time-dependent molecular targets are assessable which may be particularly important for patients with large spleens.Keywords: Chronic myeloid leukemia, Tyrosine kinase inhibitor, BCR-ABL, Imatinib resistance