Over the past few years, several large-scale studies using next-generation sequencing (NGS) of whole-genomes (WGS) and whole-exomes (WES) have defined the mutational landscape of chronic lymphocytic leukemia (CLL) [1,2,3,4]. NGS studies have also revealed the clonal heterogeneity in CLL and showed that clonal evolution contributes to the variability in clinical course among CLL patients . Clonal evolution is considered a key condition in CLL progression and relapse after treatment. Most CLL cases are diagnosed during the inactive disease phase, genetic aberrations’ underlying progress in CLL activity leading to the need for therapy are poorly understood and should be explored. A large number of frequently mutated genes have been identified and several putative driver mutations likely to confer selective growth advantage to CLL tumor cells have been proposed [1,2,3]. In addition, clonal shifts between paired treatment-naïve and relapsed CLL samples have been reported due to pre-existing subclone expansion under therapeutic pressure, demonstrating that clonal evolution likely underlies CLL relapse [3, 5]. Nevertheless, there are still a limited amount of longitudinal WES studies analyzing consecutive CLL samples before treatment intervegntion . The acquisition of driver mutations accompanied by selectively neutral passenger changes during disease prior to therapy influence is therefore poorly documented. Here, WES was performed on consecutive treatment-naïve samples of CLL patients from three groups with different disease course: Active disease (AD) group: patients with an active disease before the second analyzed time-point (TP2); Stable disease (SD) group: cases with a period of stable phase after diagnosis followed by progression within 3 years after; and Indolent disease (ID) group: those with a long-term stable indolent disease. Moreover, we applied a novel integrative bioinformatics tool called “Cancer Genome Interpreter” to identify driver mutations .
Thirty-five CLL patients were included in the WES study. In total, 70 tumor samples – (two tumor time-points (TP) for each patient) - as well as 26 matched germline samples, were sequenced. Three groups of patients were characterized based on the disease activity at the second TP: (i) AD group (n = 20); (ii) SD group (n = 6); (iii) ID group (n = 9). Sampling points and group definition details are shown in Fig. 1. The disease activity was assessed according to iwCLL guidelines . Sample characteristics are summarized in Supplementary Table S1, sample processing and WES analysis are detailed in Supplemental Material. In order to distinguish driver from passenger mutations, the novel bioinformatics tool “Cancer Genome Interpreter” (CGI, https://www.cancergenomeinterpreter.org/home) was used;  defined driver mutations were consequently validated by deep-targeted sequencing (DTS), as described previously . Moreover, FISH data from testing of four recurrent cytogenetic aberrations (del13q, trisomy 12, del11q and del17p) were available for all samples.
WES analysis of samples from both TPs obtained from 26 CLL patients with available paired germline material showed presence of 25 somatic mutations. From WES analysis of 9 CLL patients with no available non-tumor control, 67 putatively somatic mutations were identified. Taken together, a total of 392 non-silent somatic or putatively somatic mutations (363 non-synonymous and 29 indels) were identified in 353 genes across the 35 CLL patients (Supplementary Table S2). Using CGI algorithm, 54 mutations were classified as “driver” and 338 mutations as “passenger” (Supplementary Table S2). The large majority of driver mutations (50/54, 92.6%) were further validated by deep-targeted sequencing (DTS) (Supplementary Table S3). Moreover, DTS of a 9-gene set recurrently mutated in CLL (TP53, SF3B1, NOTCH1, NFKBIE, BIRC3, POT1, MYD88, XPO1, and EGR2) revealed 7 mutations which were not detected by WES due to their low Variant Allele Frequency (VAF) (Supplementary Table S3). The 57 validated driver mutations were located in 35 different genes. The most frequently mutated genes were SF3B1 (8/35, 22.9%), NOTCH1 (4/35, 11.4%), NFKBIE (4/35, 11.4%), TP53 (3/35, 8.6%), BIRC3 (3/35, 8.6%), and RPS15 (3/35, 8.6%) (Fig. 2). Among the other genes with a driver mutation, 11 had previously been reported as drivers in CLL patients [2, 3]. Additionally, CGI analysis also predicted driver mutations in CDC73, DHX9, EGFR, ERCC6, FAT1, GATA3, G3BP1, HDAC2, IDH1, and PTCH1 genes that were unknown for CLL to date (Fig. 2). Among them, the tumor suppressor FAT1 has been related to chemo-refractoriness in CLL ; HDAC2 is known to be down-regulated in CLL ; and DHX9, GATA3, and IDH1 have been described to be recurrently mutated in other hematological malignancies .
To identify somatic mutations which could be involved in clonal evolution, we analyzed the VAF dynamics between TP1 and TP2. Twenty-six out of 57 (46 %) driver mutations showed a significant change in allele frequency at the TP2: 4 were detected only at the TP2, 21 showed VAF increase at the TP2, and 1 mutation showed a decrease (Fig. 2). Additionally, FISH analysis of four recurrent cytogenetic aberrations at both TPs showed that 11/35 patients acquired one or more new cytogenetic alterations at TP2 (3/9 ID, 2/6SD, and 6/20 AD) (Fig. 2). The most often acquired aberration - deletion 13q, was detected in 7 cases (2/9 ID, 1/6SD, and 4/20 AD). Acquisition of deletion 11q was detected in 4 cases (2/9 ID, 1/6SD, and 1/20 AD). Two patients who acquired a 17p deletion were from the SD and AD group. Taking together the WES and FISH results, clonal evolution was observed in 5/9 ID patients, in 6/6SD patients and in 14/20 AD patients. Of note, 5/9 ID patients showed clonal evolution although they showed a long-term indolent disease (median follow-up = 158 months). Mutations in CLL drivers associated with aggressive clinical course such as TP53, BIRC3, RPS15, and NFKBIE [4, 13,14,15] were mostly detected within the AD/SD groups (Fig. 2). Nevertheless, there were well-known CLL driver mutations (NOTCH1, SF3B1) detected in two of eight ID patients, revealing the fact that the simple presence of such a mutation does not immediately lead to disease progression. Follow-up of these two patients with indolent disease already bearing a driver mutation at TP1 reached 99 (P5), and 213.1 (P46) months with no clinical evidence of disease activity to date as documented in Supplementary Table 1 (Fig. 1).
In summary, we performed a longitudinal study using whole-exome sequencing to characterize genetic alterations occurring during disease course before CLL-related therapy intervention in 35 CLL patients. We compared samples from indolent CLL to samples from a stable or active disease. To define potential driver mutations, we used novel integrative bioinformatics tool “Cancer Genome Interpreter”. We showed continual evolution with cytogenetic aberration and somatic mutation accumulation during the time prior to therapy intervention. Despite clonal evolution, including driver mutation presence in genes such as NOTCH1 or SF3B1, observed in indolent CLL cases, there was no clinical evidence of disease activity during long-term follow-up after sampling. We conclude that the acquisition of aberrations is not limited to the active disease phase or relapses after therapy [3, 5, 6]. Moreover, mutational profiles of indolent or outwardly stable CLL cases show that the presence of CLL clones bearing driver mutations do not have to correspond directly with disease progression. Therefore, simple mutation acquisition does not necessarily lead to immediate disease progression; nevertheless, accumulating changes precede the manifestation of disease activity. In addition, clonal evolution can occur in the absence of adverse prognostic factors such as the presence of high-risk cytogenetic alterations or unmutated IGHV. In fact, the acquisition of mutations can happen in the absence of any FISH alterations (P35 or P45) as well as in IGHV-mutated CLLs (P46). Unfortunately, analysis of genomic changes does not fully explain the transformation to a more aggressive stage in all CLL patients (P40). It was reported that epigenetic changes could also fuel CLL evolution during disease progression . Understanding CLL evolution from the time of diagnosis to therapy need may be essential to gain insight into the process of transformation from the initial inactive form to later more aggressive stages. Although white blood cells (WBC) count during disease course is more feasible than performing NGS studies, we have observed that the acquisition of genomic alterations does not have to simply correspond with an increase of WBC (P4 or 35). Then, genomic analysis should be made in larger longitudinal-based cohort studies in order to evaluate how to predict disease activation in CLL. On the other hand, to understand the genomic changes underlying CLL relapse, mutational analysis at the time of diagnosis may be irrelevant as additional aberrations may appear during time and clonal shifts are likely to happen. Such analysis should be done before therapy intervention to monitor tumoral clones that are responsible for CLL relapse.
Landau DA, Wu CJ. Chronic lymphocytic leukemia: molecular heterogeneity revealed by high-throughput genomics. Genome Med. 2013;5:47.
Puente XS, Bea S, Valdes-Mas R, Villamor N, Gutierrez-Abril J, Martin-Subero JI, et al. Non-coding recurrent mutations in chronic lymphocytic leukaemia. Nature. 2015;526:519–24.
Landau DA, Tausch E, Taylor-Weiner AN, Stewart C, Reiter JG, Bahlo J, et al. Mutations driving CLL and their evolution in progression and relapse. Nature. 2015;526:525–30.
Ljungstrom V, Cortese D, Young E, Pandzic T, Mansouri L, Plevova K, et al. Whole-exome sequencing in relapsing chronic lymphocytic leukemia: clinical impact of recurrent RPS15 mutations. Blood. 2016;127:1007–16.
Burger JA, Landau DA, Taylor-Weiner A, Bozic I, Zhang H, Sarosiek K, et al. Clonal evolution in patients with chronic lymphocytic leukaemia developing resistance to BTK inhibition. Nat Commun. 2016;7:11589.
Smith EN, Ghia EM, DeBoever CM, Rassenti LZ, Jepsen K, Yoon KA, et al. Genetic and epigenetic profiling of CLL disease progression reveals limited somatic evolution and suggests a relationship to memory-cell development. Blood Cancer J. 2015;5:e303.
Tamborero D, Rubio-Perez C, Deu-Pons J, Schroeder MP, Vivancos A, Rovira A, et al. Cancer Genome Interpreter annotates the biological and clinical relevance of tumor alterations. Genome Med. 2018;10:25.
Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero G, Dohner H, et al. iwCLL guidelines for diagnosis, indications for treatment, response assessment, and supportive management of CLL. Blood. 2018;131:2745–60.
Kubesova B, Pavlova S, Malcikova J, Kabathova J, Radova L, Tom N, et al. Low-burden TP53 mutations in chronic phase of myeloproliferative neoplasms: association with age, hydroxyurea administration, disease type and JAK2 mutational status. Leukemia. 2018;32:450–61.
Messina M, Del Giudice I, Khiabanian H, Rossi D, Chiaretti S, Rasi S, et al. Genetic lesions associated with chronic lymphocytic leukemia chemo-refractoriness. Blood. 2014;123:2378–88.
Van Damme V, Govaerts E, Nackaerts K, Dooms C, Wauters I, Vansteenkiste J. Clinical factors predictive of long-term survival in advanced non-small cell lung cancer. Lung Cancer. 2013;79:73–6.
Cazzola M. IDH1 and IDH2 mutations in myeloid neoplasms--novel paradigms and clinical implications. Haematologica. 2010;95:1623–7.
Mansouri L, Sutton LA, Ljungstrom V, Bondza S, Arngarden L, Bhoi S, et al. Functional loss of IkappaBepsilon leads to NF-kappaB deregulation in aggressive chronic lymphocytic leukemia. J Exp Med. 2015;212:833–43.
Rossi D, Fangazio M, Rasi S, Vaisitti T, Monti S, Cresta S, et al. Disruption of BIRC3 associates with fludarabine chemorefractoriness in TP53 wild-type chronic lymphocytic leukemia. Blood. 2012;119:2854–62.
Zenz T, Eichhorst B, Busch R, Denzel T, Habe S, Winkler D, et al. TP53 mutation and survival in chronic lymphocytic leukemia. J Clin Oncol. 2010;28:4473–9.
The research leading to these results has mainly received funding from the European Union Seventh Framework Programme [FP7/2007–2013] under Grant Agreement no 306242-NGS-PTL. In addition, this work was supported by grants from the Spanish Fondo de Investigaciones Sanitarias PI15/01471, PI18/01500, Instituto de Salud Carlos III (ISCIII), European Regional Development Fund (ERDF) “Una manera de hacer Europa”, “Consejería de Educación, Junta de Castilla y León” (SA085U16), “Proyectos de Investigación del SACYL”, Spain: GRS 994/A/14, BIO/SA10/14, BIO/SA31/13, GRS 1172/A/15,“Fundación Memoria Don Samuel Solórzano Barruso”, by grants (RD12/0036/0069) from Red Temática de Investigación Cooperativa en Cáncer (RTICC), Centro de Investigación Biomédica en Red de Cáncer (CIBERONC) CB16/12/00233 and USAL "Programa XIII". M. Hernández-Sánchez is supported by FEHH-Janssen (“Sociedad Española de Hematología y Hemoterapia”). M Quijada-Álamo is supported by an “Ayuda Predoctoral de la Junta de Castilla y León” (JCYL-EDU/529/2017). We are grateful to I. Rodríguez, S. González, T.Prieto, M. Á. Ramos, A. Martín, A. Díaz, A. Simón, M.del Pozo, V. Gutiérrez and S. Pujante from Centro de Investigación del Cáncer, Salamanca, for their technical assistance. D. Tamborero is supported by project SAF2015–74072-JUN, which is funded by the Agencia Estatal de Investigación (AEI) and Fondo Europeo de Dearrollo Regional (FEDER). This work was supported by Seventh Framework Programme (NGS-PTL/2012–2015/no.306242) and Ministry of Education, Youth and Sports (2013–2015, no. 7E13008); by the Ministry of Education, Youth and Sports of the Czech Republic under the CEITEC 2020 project (LQ1601); by the Ministry of Health, Czech Republic - conceptual development of research organization (FNBr, 65269705); by the Specific University Research (nr. MUNI/A/0968/2017) provided by MEYS; and by the project CZ.02.1.01/0.0/0.0/16_013/0001634 National Center for Medical Genomic - modernization of infrastructure and research of genetic variation in the population, funded by OP RDE. We acknowledge the CF Genomics CEITEC MU supported by the NCMG research infrastructure (LM2015091 funded by MEYS CR) for their support with obtaining the scientific data presented in this paper. We acknowledge S. Takacova from CEITEC MU for her help with the sample selection and processing.
MHS designed the experiment, analyzed data, and wrote the manuscript; JK designed the experiment, performed validation analysis and wrote the manuscript; AER performed sample selection and contributed to the interpretation of the results; LR and KP performed biostatistical analysis; DT performed bioinformatics analysis; MA, KPl, and RB performed sample selection; NT and MQ contributed to the interpretation of the results, VB performed validation analysis; AAM, MD, and AGC provided clinical data; NLB performed bioinformatics analysis; JMHR and SP designed the experiment and wrote the manuscript. All authors revised the manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
About this article
Cite this article
Hernández-Sánchez, M., Kotaskova, J., Rodríguez, A.E. et al. CLL cells cumulate genetic aberrations prior to the first therapy even in outwardly inactive disease phase. Leukemia 33, 518–558 (2019). https://doi.org/10.1038/s41375-018-0255-1
Blood Advances (2019)
Genetic dynamics in untreated CLL patients with either stable or progressive disease: a longitudinal study
Journal of Hematology & Oncology (2019)