Acquired Bruton's tyrosine kinase mutations in patients treated on the ibrutinib and rituximab, ibrutinib alone, and ibrutinib and venetoclax arms of the national multi-centre Phase III FLAIR study in previously untreated CLL patients

Abstract Background: The FLAIR trial confirmed that ibrutinib alone (I), ibrutinib and rituximab (IR), and MRD-directed ibrutinib and venetoclax (I+V) have resulted in improved progression free survival (PFS) compared to FCR for front line treatment in CLL. In particular, I+V was found to have a 5-years estimated PFS of 93.9% compared to 58.1% with FCR. Covalent Bruton's tyrosine kinase (BTK) inhibitors work by binding to the C481 residue of the BTK protein, preventing the autophosphorylation of the protein and therefore the proteins' catalytic properties. Mutations in the BTK gene at C481 and T474, have been reported in relapsing patients treated with ibrutinib. To understand and compare the mechanism of resistance in patients who relapse on ibrutinib-containing treatment regimens, patients' DNA was examined for presence of treatment related mutations in BTK. Here we present detailed patterns of clonal progression in patients with progressive disease (PD) treated with I, IR and I+V, using digital droplet PCR to track mutations through serial samples at 6 monthly time points throughout the trial. Methods: Peripheral blood DNA samples from patients with PD (96 IR, 33 I; 8 I+V) were analysed for the presence of BTK mutations at the C481 and T474 hotspots. ddPCR was used to detect C481S (both c.1442G>C and c.1441T>A), C481R, C481F and T474I mutations. Allelic burden was expressed as fractional abundance (adj FA), adjusted for disease burden as determined by flow cytometry. Where mutations were found at disease progression, serial samples from initiation of the trial and at 6 monthly intervals until disease progression, were then analysed to establish when the mutation first arose and to track its evolution over time. Due to the high sensitivity of the ddPCR technique it was possible to detect mutations at <0.01%. Results: Among patients with PD, 23/96 (24%) IR and 6/33 (18%) I patients were found to have one or more BTKmut at the time of disease progression; no I+V patient was found to have a BTKmut at either hotspot. The majority (18/29) of BTKmut patients had more than one BTKmut. Of the total 52 BTKmuts detected within the 29 patients, there were 19 C481S (c.1442G>C), 22 C481S (c.1441T>A), 4 C481R, 2 C481F and 2 T474I. Neither C481F nor T474I were ever found to be the sole BTKmut in PD patient, unlike the other C481 mutations. No BTKmut was found at baseline demonstrating that these mutations were treatment related. The median length of time on trial before the detection of a BTKmut was 67 months. Of the BTKmut patients 19/29 were found to have a detectable mutation prior to PD. The median length of time from 1st detected BTK mutation to PD was 7 months. The majority of BTKmut patients were late progressors with 22/29 progressing at ≥72 months. In the other 7/29 patients, PD occurred between 33-71 months with a median time of 67 months. Evolutionary dynamics show mutations rising as treatment continues and then dropping upon cessation of treatment. In certain cases however, as the number of CLL cells start to expand at an increasing rate, the adj FA within the cancer cell population starts to decrease. The possible explanation would be the dominance of other clones not containing the specific BTK mutation. BTKmut patients can be broadly split into two groups, one where the mutation is present in a large percentage of the cell population ≥5% (12/29 pts), and one where the mutation is instead only present in a small percentage of the population <5% (17/29). Of the 29 PD patients with BTKmut, 24 were IGHVUNMUT (83%) which represent a higher % compared to the prevalence of IGHVUNMUT in the population of PD patients in general (63%), indicating that the IGHV mutational status is not only a prognostic marker for PFS and OS, but is associated with an increased risk of developing a BTKmut during prolonged treatment with ibrutinib. TP53mut were also found in 4/29 of the PD patients with BTKmut which is a known risk of acquiring BTKmut. Conclusion: 21% of patients progressing on ibrutinib-containing regimens had a BTKmutat the C481 or T474 residues, leading to the conclusion that there are other factors contributing to relapse. PD with BTKmut in ibrutinib-containing arms occurred late during treatment. cBTKmut at either BTK hotspot were not detected in I+V patients, indicating that the combination of ibrutinib and venetoclax mitigated the development of the BTK mutation related resistance mechanism.