Genetic Evolution Promotes Progression and Drives Relapse in CLL

Chronic lymphocytic leukemia (CLL) is generally an indolent process, but the natural history is highly variable with some patients succumbing within one to two years of diagnosis as a result of aggressive, treatment-refractory disease. Like many other types of cancer, CLL is genetically heterogeneous, and specific chromosomal abnormalities have been shown to correlate with survival. For example, loss of the TP53 locus as a consequence of deletion affecting chromosome 17p has negative prognostic implications whereas loss of the region on chromosome 13q that encodes microRNAs 15 and 16 is prognostically favorable. CLL patients often relapse after their initial treatment, and relapsed patients are more resistant to subsequent treatment. Clonal evolution has been thought to account for disease relapse and chemotherapy resistance;¹  however, little was known about the dynamics of this process or the nature of the combination of genetic abnormalities that determine clinical outcomes in CLL.

In the current study, Landau and colleagues have addressed these issues by investigating intratumoral genetic heterogeneity using whole-exome sequencing and somatic copy number analysis in paired wild-type germ line material and mutant B cells from 149 patients with CLL. By applying two recently developed analytical tools (MeTect and ABSOLUTE), the authors identified five recurrent chromosomal abnormalities [del (8p), del (13q), del (11q), del (17p), and trisomy 12] and 20 mutated CLL driver genes (the term driver is used to indicate that the mutant gene contributes significantly to malignant transformation) and classified these mutations as either clonal or subclonal. Clonal mutations including MYD88, trisomy 12, and del (13q) occurred in the majority of cancer cells and represented early events in CLL development; subclonal mutations (e.g., SF3B1, TP53, ATM, and RAS) were present in only a subset of leukemic cells and were acquired after tumor initiation. The clonal mutations are primarily involved in B-cell malignancies, while the subclonal mutations are often found in other cancers.

Landau and colleagues further studied the association between clonal and subclonal mutations and CLL development, treatment, and clinical outcomes. They reported that age and mutated IGVH status, but not ZAP70 expression, were associated with a greater number of clonal but not subclonal mutations. Notably, subclonal, but not clonal, mutations were increased in 29 chemotherapy-treated patients compared with 120 chemotherapy-naïve patients. This finding was explained by ablation of dominant clones by treatment and the expansion of pre-existing subclones that gained a competitive advantage when treatment was applied. To challenge this hypothesis, evolution of somatic mutations was directly assessed in 18 samples collected at two time points (interval range between time points: 3.1-4.5 years). Expansion of subclones with driver mutations was detected in 10 of 12 chemotherapy-treated patients while clonal equilibrium was maintained in five of six untreated patients. Furthermore, the presence of driver mutations in subclones prior to therapy impacted both prognosis and clinical outcome. The 10 patients whose post-treatment samples showed evidence of clonal evolution had shortened failure-free survival from time of next therapy (FFS_Rx), and in a separate analysis, patients (n=39) with pretreatment subclonal drivers had shorter FFS_Rx than those (n=29) without subclonal drivers, independent of established prognostic markers.