Standardized Targeted Sequencing of 29 Genes with Recurrent Mutations in Chronic Lymphocytic Leukemia (CLL) Helps to Better Define Clinical Correlates and Genomic Hotspots

Background: CLL is a genetically heterogeneous disease with a variety of chromosomal abnormalities and gene mutations present at both the onset of the disease and acquired after treatment. Many of these genomic alterations have clinical implications as they may portend worse survival outcomes or resistance to chemo-immunotherapy, and therefore contribute to the comprehensive risk stratification in CLL pts. Since 2010, 654 CLL pts seen at MDACC had a standardized gene sequencing panel performed to help with prognostication. Here, we summarize the associations of these mutations with clinical characteristics and define the significance of recurrent mutational hotspots.

Methods: From the 680 total samples (654 distinct pts with 26 pts having two samples at various time points), DNA from either peripheral blood or bone marrow that contained >10% CLL cells underwent targeted sequencing with a next generation sequencing-based panel of 29 genes (EndCLL Assay V1), including ATM, BIRC3, BTK, CALR, CARD11, CD79A, CD79B, CHD2, CSMD3, CXCR4, DDX3X, EZH2, FAT1, FBXW7, KLHL6, LRP1B, MAPK1, MUC2, MYD88, NOTCH1, PLCG2, PLEKHG5, POT1, SF3B1, SPEN, TGM7, TP53, XPO1 and ZMYM3. Clinical characteristics such as FISH abnormalities, chromosome karyotype, previous treatments, ZAP70 and somatic hypermutation (IGHV) status were assembled from when this assay was performed.

Results: Overall, 400 (59%) pts were treatment-naïve and 280 (41%) were previously treated. Those with mutated IGHV accounted for 236 (35%) of pts while 321 (47%) had unmutated IGHV with 123 (18%) unknown. Stratifying pts by hierarchical FISH model, 99 (15%) had del(17p), 102 (15%) had del(11q), 93 (14%) had trisomy 12, 204 (30%) had del(13q) , and 91 (13%) had no abnormalities. The majority of patients (70%) had ≥1 discovered mutation with TP53 (22%), SF3B1 (18%), NOTCH1 (13%) and ATM (13%) being the most commonly mutated genes. Previously treated pts were significantly enriched for del(11q) (p=<0.001), del(17p) (p=<0.001), complex karyotype (p=<0.001) and mutations in BTK (p=<0.001), SF3B1 (p=0.01) and TP53 (p=<0.001) while treatment-naïve pts were enriched for mutations in MYD88 (p=0.01). Separating pts into IGHV -mutated versus -unmutated, mutated pts were significantly enriched for del(13q) (p=<0.001) and mutations in CD79B (p=0.03), KLHL6 (p=0.001)and MYD88 (p=<0.001) while IGHV -unmutated patients had increased proportion of del(11q) (p=<0.001), del(17p) (p=<0.001), complex karyotype (p=<0.001) and mutations in ATM (p=0.001), NOTCH1 (p=<0.001) , SF3B1 (p=<0.001) , TP53 (p=<0.001) and XPO1 (p=<0.001). Shared pairwise associations discovered in both treated and untreated cohorts were found in trisomy 12 with BIRC3 (p=<0.001), CHD2 (p=0.01) and SPEN (p=0.01), BIRC3 / FBXW7 (p=0.03) and NOTCH1 / POT1 (p=0.02) while exclusions occurred between trisomy 12/del(13q) (p=<0.001). Comparing mutations within the DNA-binding domain of TP53 versus ones in all other domains showed no clinical correlations. However, multiple mutations within TP53 were associated with prior treatment (p=0.03) and trisomy 12 (p=0.04) while a single TP53 mutation was more commonly associated with ATM mutation (p=0.04). NOTCH1 mutation at P2514fs is a known hotspot with 91 (81%) such mutations noted in our cohort. However, mutations outside of this hotspot were correlated with del(13q) (p=0.03) and FAT1 mutation (p=0.01).

Conclusion: In the largest cohort of targeted gene sequencing to date, expected abnormalities and mutations were found within previously treated and untreated pts in del(11q), del(17p), complex karyotype and mutations in ATM, TP53, SF3B1 and MYD88 . New associations between IGHV -mutated and del(13q) along with mutations in CD79B and KLHL6 were discovered along with new shared pairwise associations in treatment-naïve and previously treated pts between trisomy 12 with CHD2 and SPEN, BIRC3 / FBXW7 and NOTCH1 / POT1, which may suggest common activated pathways and pathogenesis of disease. Recurrent regional mutations such as the DNA-binding domain in TP53 and P2514fs in NOTCH1 had no associated clinical correlates but having either single or multiple TP53 mutations did have clinical correlates. Future direction will include associating other genomic hotspots with clinical characteristics, survival outcomes and clonality.