High-Resolution Genomic Copy Number Analysis on Sequential Samples from the CLL8 Trial: Relation Between Clonal Evolution and Defects in DNA Damage Response?

Genomic abnormalities have strong prognostic impact in chronic lymphocytic leukemia (CLL). However, clonal evolution has been studied in a limited number of cases and not within the setting of current standard therapy. It was therefore our aim to study changes in the composition of copy number alterations (CNA) over time with and without the influence of chemo(immuno)therapy.

Sequential samples of 92 patients enrolled on the CLL8 trial of the GCLLSG were analyzed by Affymetrix® 6.0 single nucleotide polymorphism (SNP) arrays. Since procurement of a relapse sample was a prerequisite for this study, the cohort was not representative for the CLL8 trial [21% CR (N=19), 63% PR (N=58), 13% non response (N=12), 3% missing response (N=3)]. 48 patients received Fludarabine / Cyclophosphamide (FC), 44 patients FC plus Rituximab (FCR). Samples were taken at 3 time points: pre-treatment [N=27], time of first treatment in CLL8 [N=92] and post treatment at relapse / progressive disease [N=74]. The median observation period between samples was 35 months [range: 6-127] for the pre-treatment vs. first treatment and 41 months [range: 5-87] for the first treatment vs. post treatment (relapse) comparisons.

The majority of cases maintained genomic stability over time. This applied in particular for the comparison between pre-treatment and first treatment [N=21 of 27; 78%] but also for the comparison between first treatment and relapse [N=49 of 74; 66%].

The cohort was characterized by a high proportion of high-risk genomic abnormalities [30% del(11)(q22.3), 22% TP53 loss and/or mutation in the pre-treatment cohort; 36% del(11)(q22.3), 24% TP53 loss and/or mutation in the post treatment cohort]. The acquisition of novel clonal CNAs was associated with these: In the pre-treatment to first treatment comparison [N=27], only 6 [22%] cases had novel CNAs emerging over time and all of them carried an ATM loss by del(11)(q22.3). Eight cases with ATM and/or TP53 alteration and all cases without high-risk genomic aberrations [N=13] showed no evidence of clonal evolution. In the first treatment to relapse comparison [N=74], 25 [34%] cases had newly acquired CNAs at relapse and 20 of them [80%] carried high-risk genomic abnormalities [44% del(11)(q22.3), 36% TP53 loss and/or mutation]. Only 5 cases lacking ATM or TP53 alteration had newly acquired CNAs. In contrast, genomic stability was observed in 25 cases with high-risk genomic abnormalities [15 cases with del(11)(q22.3), 10 with TP53 loss and/or mutation) and 24 cases without high-risk abnormalities. No statistically significant increased incidence of clonal evolution was observed in IGHV unmutated cases [N=66 of 92] [24 (80%) cases with clonal evolution vs. 42 (70%) cases without clonal evolution, p=0.3].

Nine patients had samples available from all 3 time points. While mainly genomic stability could be observed prior to treatment, 3 cases acquired 3 novel CNAs each after therapy.

Comparing both treatment arms in the post treatment cohort [N=74] revealed a higher incidence of clonal evolution after treatment with FCR [FCR: N=16 of 35, 46%; FC: N=9 of 39, 23%; p=0.04] at a similar median observation period [36 and 35 months, respectively]. Also, the mean number of newly acquired CNAs at relapse was higher in the FCR treated group [3.3 vs. 2.6]. With regard to response no statistically significant differences were observed between cases with and without clonal evolution [cases with clonal evolution: 6 (24%) CR, 17 (64%) PR, 2 (8%) non-response; without: 8 (17%) CR, 31 (66%) PR, 8 (17%) non-response, p=0.5].

Loss of clonal lesions was rare, occurring only under the selection pressure of therapy: 16 CNAs in 11 cases were not observed anymore at relapse. 3 of these CNAs were subclonal at time of first treatment [10-20% allelic burden] and might not yet have re-emerged after relapse. 5 CNAs in 4 cases were lost at relapse. The remaining 8 CNAs were del(13q) [N=5] and del(11q) [N=3] that were lost for a probably more advantageous del(13q) / del(11q) clone. The appearance of a more advantageous del(13q) / del(11q) clone was linked to a larger deletion size [N=3] or a larger discontinuous deletion very likely resulting from chromothripsis [N=3].

The results of this study support previous data of a high genomic stability in CLL cases lacking alterations of TP53 and/or ATM. However, application of chemo(immuno)therapy did slightly increase the number of cases acquiring novel clonal CNAs.