Abstract
Genomic studies have revealed that individual tumours are genetically heterogeneous containing genetic alterations which can evolve over time. The recently discovered novel recurrent mutations in CLL, i.e. SF3B1, NOTCH1, but also TP53 and ATM, were more often subclonal, indicating that they tend to arise later in leukemic development and contribute to disease progression. Although, it has been suggested that the presence of a subclonal driver mutation is an independent risk factor for disease progression in CLL, this has not yet been tested in a prospective trial with treatment naive patients.
To study the incidence and clinical impact of clonality of novel recurrent mutations in a prospective trial in high risk treatment naive CLL patients treated with fludarabine and cyclophosphamide with (FCA) or without (FC) alemtuzumab.
In the HOVON68 trial, treatment naïve CLL patients with a high risk profile defined as either 17p deletion, 11q deletion, trisomy 12, unmutated IGHV and/or VH3-21, were randomized between treatment with 6 cycles of FC and FCA. We evaluated 119/272 (43.8%) trial patients based on availability of DNA. This cohort did not differ from the total trial population as to clinical characteristics and prognostic factors. Mutational analysis of SF3B1, NOTCH1, TP53, BRAF, KRAS, NRAS, EZH2, MYD88 and PIK3CA was performed by next generation sequencing. Mutations were classified as subclonal in case of an allelic frequency of either 2-40% or 60-90%. Extensive analysis of ATM using Sanger sequencing (ex 4-65) and ATM functional analysis by assessing IR induced phosphorylation of ATM targets was performed in 82 patients.
There were 75 (63%) patients with a subclonal mutation and/or chromosomal abnormality in this cohort of 119 high risk treatment naïve patients. In total 19 (16%) SF3B1, 12 (10%) NOTCH1, 16 (13%) TP53, 5 (4%) BRAF, 5 (4%) KRAS, 2 (2%) NRAS and 1 (1%) MYD88 mutations were found. ATM mutations were identified in 11 out of 82 patients. Mutations in TP53 correlated with 17p deletion (p<0.0001), ATM and SF3B1 mutations with 11q deletion (p=0.009 and 0.008 respectively) and NOTCH1 mutations with trisomy 12 (p=0.002). RAS, TP53 and NOTCH1 mutations were mainly subclonal (90-67%), whereas the other aberrations were subclonal in about 50%. Patients with subclonal mutations had a significantly lower overall response rate (ORR; p=0.04) as compared to patients with clonal mutations. At the individual gene level ORR and CR were decreased in patients with TP53 mutation (p=0.001 and p=0.01 respectively) but not in those with the other mutations. FCA improved ORR in patients with TP53 mutations (14% vs 67%; p=0.06). However in patients with the other mutations there were no differences in ORR/CR between the two treatment arms.
With a median follow-up of 42.5 months progression free survival (PFS) tends to be shorter in patients with subclonal mutations (median 29 vs 35 months; p=0.09). PFS was only significantly decreased in patients with TP53 mutation (p<0.001). FCA improved PFS in patients with mutated TP53 (median 3 vs 16 months; p=0.02), and in those with SF3B1 mutation (median 33 vs 44 months; p=0.05). Overall survival (OS) was only decreased in TP53 mutated (p=0.02) patients and there was a trend for improved OS due to FCA (median 24 vs 67 months; p=0.09).
The majority of novel recurrent mutations and chromosomal abnormalities were found to be subclonal, and subclonality was associated with inferior outcome in this high risk CLL treatment naïve patient group. As expected, TP53 mutations had an adverse prognostic impact, which was overcome by alemtuzumab. In addition, we found that patients with an SF3B1 mutation might benefit from alemtuzumab.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
