Evolving Understanding of the CLL Genome

Michaela Grubera,b,c and Catherine J. Wua,b,d

Over the past few years, massively parallel sequencing technologies have revealed with high resolution the

0037-1963/& 2014 Elshttp://dx.do

Funding: Dnational2013-624Family(SU2C IBlood IInstitute

Conflicts o

aDepartmenBoston, M

bBroad InscDepartmenHaemost

dDepartmeMedical

Address coInstitute,Dana 540

Seminars

tremendous genetic and epigenetic heterogeneity in chronic lymphocytic leukemia (CLL). We havelearned how the molecular architecture differs not only between affected individuals but also withinsamples and over time. These insights have catalyzed our understanding of the pathobiology of CLL andpoint to critical signaling pathways in the development and progression of the disease. Several key driveralterations have been identified, which serve to refine prognostic schemata but also to inspire thedevelopment of new therapeutic strategies. Ongoing advances in technology promise to further elucidatethe molecular basis of CLL, and this knowledge is anticipated to aid us in understanding and addressingthe clinical challenge presented by the vast variability in the clinical course of patients with CLL.Semin Hematol 51:177–187. C 2014 Elsevier Inc. All rights reserved.

A hallmark of chronic lymphocytic leukemia (CLL)is its tremendously variable clinical course. Asmany as 80% of CLL patients are asymptomatic

at diagnosis, but many progress to extensive lymphaden-opathy, hepatosplenomegaly, and life-threatening cytope-nias within only a few years. Others, however, remainasymptomatic over decades, with 20% to 30% having alife expectancy not significantly different from the generalpopulation.1,2

An enduring goal of CLL studies has been to betterunderstand the basis of this clinical variability. Of note,because of its high prevalence, relatively slow progression,and the ready availability of leukemia samples from patientperipheral blood, CLL has been continuously at theforefront of genomic research. Thus, while the firstprognostic schema, established in the 1970s,3,4 was basedon clinical features, newer studies have focused on the roleof somatic genomic alterations in the pathogenesis of CLL

$ - see front matterevier Inc. All rights reserved.i.org/10.1053/j.seminhematol.2014.05.004

r. Gruber acknowledges support by a Marie-Curie Inter-Outgoing Fellowship from the European Union (PIOF-924). Dr. Wu acknowledges support from the BlavatnikFoundation, American Association for Cancer Researchnnovative Research Grant), the National Heart, Lung, andnstitute (1RO1HL103532-01), and the National Cancer(1R01CA155010-01A1).f interest: none.

t of Medical Oncology, Dana-Farber Cancer Institute,A.

titute, Cambridge, MA.t of Internal Medicine I, Division of Haematology andaseology, Medical University of Vienna, Vienna, Austria.nt of Medicine, Brigham and Women’s Hospital, HarvardSchool, Boston, MA.

rrespondence to Catherine J. Wu, MD, Dana-Farber CancerDepartment of Medical Oncology, 450 Brookline Avenue,B, Boston, MA 02215. E-mail: cwu@partners.org

in Hematology, Vol 51, No 3, July 2014, pp 177–187

and, in turn, have examined their impact on clinicaloutcome. For example, mutational status of the immuno-globulin heavy chain variable (IGHV) gene divides CLLinto two genetically distinct groups, which likely reflectsdifferent cells of origin and has emerged as an importantdisease-associated prognostic factor.5,6 In a separate land-mark study, presence of four common CLL-associatedcytogenetic aberrations, deletions of chromosomes 11q,13q, and 17q as well as trisomy of chromosome 12, couldstratify CLL patients into prognostically distinct groupsand, for the first time, was linked with clinical course andsurvival.7

Remarkably, the introduction of next-generationsequencing has led to a breathtakingly exponential increasein the knowledge of the molecular underpinnings of CLLover the last 3 years (Figure 1). These recent studies haverevealed several notable, even surprising and paradigm-shifting insights such that our perception about this diseasehas greatly evolved in a relatively short span of time. First,whole-exome sequencing (WES) of large sample cohorts hasclearly shown the high degree of genetic variability amongCLL patients, with the discovery of novel common genemutations that likely play a role in the pathobiology of CLL.Furthermore, a startlingly high degree of intra-sample clonalheterogeneity was recently uncovered, based on the detec-tion and quantification of leukemia-specific alterations thatmark various cell subpopulations within a CLL sample. Atthe same time, genome-wide approaches to examine theDNA methylome have revealed epigenetic heterogeneityamong patients and within individual samples. Collectively,these observations demonstrate the complex interrelationshipbetween genetic and epigenetic features of each sample.Herein we review these novel discoveries, which are nowundergoing evaluation as features to incorporate into newprognostic and therapeutic schema that promise to improvethe clinical care of patients with CLL.

177

Figure 1. Evolution and growth in our understanding of chronic lymphocytic leukemia (CLL) heterogeneity over time.

M. Gruber and C.J. Wu178

PROFILING GENETIC HETEROGENEITY IN CLL

Massively parallel sequencing techniques have providedthe ability to rapidly sequence millions of DNA fragmentswith relatively low sample input. As an alternative to wholegenome sequencing, selective restriction of reads to thecoding regions of the genome by WES has drasticallyreduced the costs of sequencing per sample. This latterapproach has facilitated the rapid sequencing of large samplecohorts, which in turn has enabled the drawing of associ-ations between genetic alterations and clinical features. Overthe last few years, the results of up to a dozen whole CLLgenomes8–10 and 4300 whole CLL exomes11–14 acrossdifferent centers worldwide have been reported.

UNBIASED DISCOVERY OF KEY CLL DRIVERS

A major goal of the large-scale cancer sequencingstudies has been the identification of key alterations thatdrive malignancy. The development of massively parallelsequencing has led to the parallel development ofadvanced computational algorithms for analyzing thesebig datasets. In general, these algorithms detect cancer-specific alterations with a high probability of being cancerdrivers on the basis of whether they are present at asignificantly higher-than-expected rate given the knownbackground mutation rate of the cancer. In CLL, these

efforts have corroborated known CLL-associated altera-tions (ie, mutations in TP53 and ATM) but importantlyhave identified numerous previously unknown somaticchanges, the majority of which have been confirmed acrossindependent sample cohorts. The first studies of DNAsequencing in CLL found mutations in MYD88,NOTCH1, and XPO1, as well as in BIRC3.8,11 Subse-quently, results of WES of two well-powered cohorts of�100 patients each further detected several novel somaticalterations in CLL (in FBXW7, POT1, and CHD2).Strikingly, both studies identified the novel finding ofrecurrent mutations in the splicing machinery co-factorSF3B1 in 10% to 15% of patients.9,15 Most recently, thelargest single CLL sequencing cohort to date was reported,comprising 160 patients, in which numerous lowerfrequency mutations (in NRAS, KRAS, HIST1H1E,SAMHD1, and MED12) were identified.13

These somatic alterations are present in critical com-ponents of a number of cellular pathways and includeDNA damage and cell cycle control (TP53, ATM,POT1, and BIRC3), mRNA processing (XPO1 andSF3B1), NOTCH signaling (NOTCH1), inflammatorypathways (MYD88), and chromatin modification (CHD2)(Table 1).9,13,15 Several of the significantly mutated genesdisplay a clustering of mutations in hot spots within highlyevolutionarily conserved gene regions and strongly supportthe idea that they are positively selected gain-of-function

Table 1. Evidence for Co-segregation and Mutual Exclusivity of Genetic Alterations in CLL

Pathway Alteration Co-segregation Mutual Exclusivity Location

Alterations associated with M-IGHV8,9,15,22

Chromatin modification CHD2 15q26Inflammatory pathways MYD88 Del13q9,22,25 SF3B1,25 NOTCH125 3p22

Alterations associated with U-IGHV8,9,11,14,15,22,25,27,28,82,83

DNA damage response,cell cycle control

ATM Del(11q)9 11q22-q23*

BIRC3 Del(11q)2,25 TP53,2,24,30 NOTCH1,24

SF3B12411q22*

POT1 SF3B114 7q31.33TP53 Del(17p)2,9,25,49,80–82 SF3B1,28 Tris(12),49 BIRC324 17p13.1*

mRNA processing SF3B1 Del(11q)9,22 MYD8825, Tris(12)28

NOTCH122,25,27,79, TP53302q33.1

XPO1 NOTCH18 2p15NOTCH signaling NOTCH1 Tris(12),2,9,22,26,27,29

XPO1,8 TP5311,26SF3B1,22,25,27,79 MYD88,25

TP53,29 BIRC3,24

Del(13q)26

9q34.3

FBXW7 Tris(12)9,22 SF3B1,22 NOTCH19 4q31.3

*Involved in corresponding common chromosomal alterations.Abbreviations: CLL, chronic lymphocytic leukemia; M-IGHV, mutated immunoglobulin heavy chain variable (IGHV) gene; UM-IGHV,unmutated IGHV gene.

Understanding the CLL genome 179

alterations. For example, mutation in MYD88, a criticaladaptor molecule of the interleukin-1 receptor–Toll-likereceptor (TLR)-signaling pathway, has been found almostexclusively at position L265P, which is localized within theinterleukin-1 receptor–TLR domain. It is noteworthy thatthis particular mutation seems to occur selectively in B-cellmalignancies, most prominently in Waldenström’s macro-globulinemia and diffuse large B-cell lymphoma.16,17 Inanother example, NOTCH1, a transmembrane proteincentral to the Notch-signaling pathway, is recurrently affectedby a 2 base pair frameshift deletion in the C-terminal PESTdomain,11 leading to pathway activation, increased cellsurvival, and resistance against pro-apoptotic stimuli.18

SF3B1 encodes the core catalytic subunit of the spliceosomecomplex, and its mutations localize to 900 base pairs withinthe C-terminal region9,15,19 and have been noted to affectsplicing at 3' splice sites.20,21 Another recurrently mutatedgene affecting RNA processing is the nuclear transport geneXPO1, with mutations clustering at a highly conserved site atresidue E571K.8,9,22,23 Finally, the shelterin POT1 encodes aprotein essential for telomere function, of which recurrentmutations in CLL affect key residues required to bindtelomeric DNA and lead to substantial telomeric dysfunctionassociated with increased genomic instability and numerouschromosomal abnormalities.14

The significantly mutated CLL genes also include examplesof tumor suppressor genes (TP53, BIRC3, and ATM). TP53is furthermore involved in the region of chromosome 17p,and BIRC3 and ATM at 11q, which are often found deletedin CLL and which correspond to poor prognosis.9,13,24

Further clues on the functional role of alterations canbe inferred based on the patterns of co-segregation andmutual exclusivity (Table 1). Interestingly, the signifi-cantly mutated genes in CLL seem to be differentiallyrepresented between the IGHV mutated and unmutatedCLLs. Although the former seems to be associated with del(13q) and mutations in MYD88, the latter commonlydisplay mutations affecting NOTCH1, SF3B1, ATM, andTP53 and associations with trisomy 12, del(11q) anddel(17p), respectively.9,13,22,25–27 Lesions of BIRC3 andTP53 have been noted to occur in a mutually exclusivefashion. Likewise, mutations in SF3B1, NOTCH1, andMYD88 also seem to be exclusive of each other. Thesepatterns suggest possible distinct evolutionary paths,whereby certain subclonal alterations may confer advant-age when occurring in the genomic context of particularancestor lesions. Alternatively, mutual exclusivity couldindicate that alterations have highly similar downstreameffects; thus, functionally redundant secondary mutationsdo not provide any further advantage to the tumor cell.On the other hand, consistent co-occurrence suggestssynergistic effects between alterations that enhance fitnessof the malignant clone and promote selection of drivercombinations.

As the numbers of studies examining the incidence ofthese key mutations in CLL have grown, it has alsobecome clear that their frequency in patient groups largelydepends on the composition of the sequenced cohort.Thus, although mutation frequency in SF3B1 rangesbetween 4% and 12% in early CLL, it rises to 17% to

Figure 2. Frequency of genetic alterations in chronic lymphocytic leukemia (CLL) depending on the cohort. The largestcohort per center has been taken into account, and the number of mutated cases over total size of the cohort is noted.Unselected cohorts have been included from: DFCI (Dana-Farber Cancer Institute)/Broad Institute,9,13 Italy (AmedeoAvogadro University of Eastern Piedmont, Novara; Sapienza University, Rome)/Columbia University,2,11,24,28–30 MLL(Munich Leukemia Laboratory),22,26 and SCALE (Scandinavian Lymphoma Etiology).25,79 Included are also cohorts fromthe clinical trials UK LRF (UK Lymphoma Research Foundation) CLL4,27,80 GCLLSG (German CLL Study Group) CLL4,49 andGCLLSG CLL2H.81 “Early” ¼ newly diagnosed and untreated patients without explicit evidence of progressive disease;“progressive” ¼ patients with symptomatic CLL requiring treatment or with relapse after treatment.

M. Gruber and C.J. Wu180

24% of patients by the time of disease progression.9,24,25

Similarly, mutations in TP53, ATM, and NOTCH1 havea higher incidence in cohorts with advanced disease acrossindependent studies.9,27–29 The mutation rate ofNOTCH1 is further markedly increased in patients withlymphomatous transformation.12 By contrast, mutationrates of MYD88 appear stable throughout the course ofCLL (Figure 2).9,30

Altogether, these patterns of association between distinctgenetic alterations suggest the presence of different potentialtrajectories in the development of the CLL genome.Moreover, the patterns of genetic alterations are distinctin each individual CLL, and they demonstrate the tremen-dous interindividual heterogeneity in CLL (Figure 3A).

INTRA-LEUKEMIC GENETIC HETEROGENEITYAND CLONAL EVOLUTION

Even as we have gained greater understanding ofvariation in the spectrum of genetic alterations amongpatients with CLL, analysis of WES data has revealed theextensive degree of genetic heterogeneity within individ-ual samples. The existence of genetically distinct sub-populations was already suggested by studies usingfluorescence in situ hybridization, in which manychromosomal alterations were observed only in subfrac-tions of cells.7 Moreover, the number of genomicalterations was shown to increase throughout the course,from newly diagnosed to progressive and further torelapsed CLL.31–33 Application of WES and whole-genome sequencing methods, however, have providedmore precise quantitation and more global assessment ofthis phenomenon.

Using these technologies, a series of recent studies haveindicated that a single time point genetic profile of a CLLsample represents a snapshot of multiple different tumorcell populations that are related to each other and changingover time. By integrating information on the allelicfrequencies of mutations together with local copy numberand purity information, Landau et al13 recently demon-strated the possibility of inferring proportions of cellsubpopulations harboring a genetic alteration (Figure 3B).Interestingly, certain somatic mutations were detectedpreferentially in clonal or subclonal fashion, suggesting anorder of alterations corresponding to earlier and laterdrivers. Presence of a subclonal driver by itself also indicatedpoor prognosis and more rapid disease progression. Fur-thermore, of the 149 patients in the study, a subset ofsamples assessed longitudinally showed clear patterns ofclonal evolution commonly after therapy, which was alsoassociated with worse overall outcome. Importantly, aggres-sive subclones representing a majority of the tumor cellpopulation at relapse were often already detectable inpretreatment samples. Complex and highly heterogeneousevolutionary dynamics with linear and branching subclones,as well as marked clonal shifts over time and multiple cyclesof therapies, were also seen in smaller series of 22 patientsstudied by using DNA arrays as well as in three patientsrepeatedly assessed by next-generation sequencing in adetailed multi time-point analysis over up to 7 years.10,34

PROFILING EPIGENETIC HETEROGENEITYIN CLL

The recent explosion in the number of mutations inknown epigenetic regulatory genes identified across human

Figure 3. (A, B) Inter- and intra-leukemic genetic and (C, D) epigenetic heterogeneity in chronic lymphocytic leukemia(CLL) revealed by next-generation sequencing. Panels A, B, and C were adapted with permission from Wang et al,9

Landau et al,13 and Kulis et al,45 respectively. Panel D was provided by Landau et al.46

Understanding the CLL genome 181

cancers has underscored the importance of epigenetic controlin tumor suppression.35 It is highly likely that theseepigenetic changes cooperate with genetic mutations to moldthe evolutionary tumor landscape. The best-studied epige-netic modification to date is CpG methylation, which occursby conversion of cytosine in DNA to 5-methylcytosine byaddition of methyl groups to CpG sites regulating geneexpression. In cancer, global genome-wide hypomethylationis accompanied by localized hypermethylation and anincrease in expression of DNA methyltransferase.36

Epigenetic Differences among Patients

CpG methylation profiles clearly differ among prog-nostic subcategories of CLL.37,38 These aberrantly methy-lated loci have been shown to include genes involved inCLL pathobiology, such as BCL2,39 TCL1,40 death-associated protein kinase 1 (DAPK1),41 LPL, ZAP70,and NOTCH1, as well as gene regulators and pathways

involved in B-cell signaling.42 Also aberrantly methylatedin CLL are several microRNA (miRNA) promoters, andthis change can affect the expression of those miRNAswith known roles in CLL.43 Finally, two lincRNA iso-forms were found deregulated in CLL through changes intheir promoter methylation and degree of histone mod-ification. They map to the frequently deleted region ofchromosome 13q14 and seem to play a role in controllingtranscription of multiple other genes at this locus.44

Highly comprehensive genome-wide methylation pro-filing has been performed by using the Illumina 450 Karrays. These DNA hybridization chips evaluate 485,000methylation sites per sample, covering 495% of CpGislands but also miRNA promoters and CpG sites outsideof CpG islands. In a study analyzing samples from 139CLL patients gathered by using this technology, significantmethylation differences were found in association withCLL genotypes (ie, mutations in SF3B1 and NOTCH1;trisomy 12 or del[11q]). Furthermore, methylation

M. Gruber and C.J. Wu182

profiling was able to categorize patients into three distinctclinicopathologic groups (Figure 3C). Specifically, CLLswith unmutated IGHV were strongly related to naive Bcells, whereas CLLs with predominantly mutated IGHValigned with mature B cells. These results suggest amethylation imprint corresponding to the putative cell oforigin. Unexpectedly, a third group was uncovered withmainly mutated IGHV but with a methylation signaturemore similar to naive B cells. Compared with the othertwo groups, patients in the third group had an intermedi-ate clinical outcome. It was noted that methylation ofCpGs in the gene bodies outside CpG islands showedstrongest correlation with gene expression.45

Intra-leukemic Epigenetic Heterogeneity

Given the existence of genetically distinct subcloneswithin individual patient cancer cells, it might be expectedthat epigenetic alterations could also show complex intra-tumoral differences with changes over time. Indeed, recentanalysis of 450 K array data revealed high heterogeneitywithin individual CLL samples compared with healthy Bcells (Figure 3D).46 Also using this platform, Cahill et al42

examined methylation in paired diagnostic and follow-upsamples. Although they found 42000 sites differentiallymethylated between IGHV mutated and unmutated CLL,they observed no significant differences in methylationpatterns over time or between peripheral blood and lymphnodes. These data support the idea that altered methyl-ation is an earlier, rather than later, leukemogenic event. Arecent 450 K array analysis on 28 longitudinally followedup CLL cases, however, showed that the majority ofgenetically evolved cases also showed epigenetic changesbut that epigenetic evolution was not observed in theabsence of genetic changes.47 These results suggest atemporal hierarchy in which genetic alterations precedemarked epigenetic changes and yet co-operate together. Inthe same study, Oakes et al additionally applied next-generation targeted bisulfite sequencing on 28 selectedregions. In contrast to array-based detection, massivelyparallel sequencing technologies provide the opportunityto study methylation at base pair resolution and withsequence context. Thus, allele-specific methylation in CLLcould be identified, exhibiting a stochastic pattern withrandom distribution between neighboring CpGs, whichcontrasted starkly with physiologically imprinted regionsthat showed methylation consistently occurring on thesame allele. Reduced-representation bisulfite sequencing isa recently introduced NGS technology that enablesgenome-scale methylation examination restricted to CG-rich sites, and hence is cost-effective. A recent reduced-representation bisulfite sequencing study of 104 primaryCLL samples uncovered the high degree of intra-tumoralmethylation heterogeneity in CLL compared with normalB cells. This heterogeneity was evaluated to stem primarilyfrom stochastic variation in DNA methylation, termed“locally disordered methylation.” Thus, the high

heterogeneity in methylation status within patient samplesseems to arise not from an admixture of cell subpopulationswith distinct ordered methylation patterns but rather from anincreased proportion of cells with disordered methylation oftheir genome. This high degree of “noise” in the CLLmethylome was also linked with altered transcriptionalregulation, clonal evolution, and adverse clinical outcome.46

CLINICAL IMPLICATIONS OF GENOMICDISCOVERIES IN CLL

Many of the recent insights on CLL gained by thevarious large-scale sequencing studies have substantialclinical implications. Overall, they provide new markersthat serve to refine prognostic information but also pointto promising novel therapeutic targets and new strategiesfor rationally approaching CLL treatment.

Refining CLL Prognostic Schemata

Multiple studies on unselected cohorts comprisingmainly untreated patients have demonstrated the prognosticvalue of four driver alterations: mutations in SF3B1,NOTCH1, BIRC3, and TP53.9,11,19,22,24,48 Other studiesassessed the impact of these lesions in more selected cohorts,such as patients enrolled in therapeutic clinical trials. Forexample, analysis of the German CLL4 trial cohort, whichcompared fludarabine with or without cyclophosphamide,found that the poor prognostic impact of TP53 mutationswas equivalent to chromosomal del(17p), and that theselesions together were associated with extraordinary poorresponse to chemotherapy.49 Similarly, in the UK LRFCLL4 trial, comparing different fludarabine-containingregimens, integration of mutations in TP53 together withthose in SF3B1 and NOTCH1 confirmed that patients withTP53 alterations had the shortest survival and pooresttherapeutic response. However, SF3B1 and NOTCH1mutations, albeit not associated with fludarabine response,both had an independent negative impact on overallsurvival.27 Finally, in the German CLL2H cohort,who received alemtuzumab in the setting of fludarabinerefractoriness, patients with NOTCH1 mutations also hadlonger progression-free survival, suggesting that this sub-group might particularly benefit from the anti-CD52 anti-body treatment. In contrast, SF3B1 mutation had noimpact on response rates or overall and progression-freesurvival in this population.50 These contrasting resultsindicate that different drivers may have a variable role,depending on the context.

Given the explosion of studies that have genotypedpatients across unselected cohorts and a limited set ofclinical trial cohorts, Rossi et al2 sought to integrate thismutational information (on TP53, NOTCH1, SF3B1, andBIRC3) together with conventional FISH cytogenetic datain an effort to improve the predictive ability of clinicalprognostic schema. Indeed, when this was done, predictiveaccuracy was significantly enhanced in a large cohort of

Understanding the CLL genome 183

637 newly diagnosed CLL cases and was validated in anindependent group of 370 patients. Patients with TP53 orBIRC3 lesions consistently had the worst prognosis,followed by patients with mutations in SF3B1 andNOTCH1 and del(11q). Notably, these high-risk lesionsdeveloped in �20% of initially low-risk subjects duringthe course of the disease, among whom del(11q) andabnormalities of TP53, NOTCH1, and SF3B1 as well asBIRC3 again represented a major fraction.2 Another largestudy assessing mutations of TP53, NOTCH1, SF3B1,FBXW7, MYD88, and XPO1 together with cytogeneticsand IGHV mutational status in 1160 patients confirmedthe strong prognostic value of NOTCH1, TP53, andSF3B1 alterations with independent impact on overallsurvival of the latter two.22

The aforementioned studies have addressed the ques-tion of whether the presence or absence of a drivermutation in a CLL sample is prognostic. However, ithas become evident that addressing the fraction of CLLcells bearing a driver mutation within a clonally heteroge-neous population and which may participate in clonalevolution can present daunting challenges to physicianstreating patients with this disease. Clearly, therapy imposesa strong selective pressure that affects the relative propor-tions of subclones harboring driver mutations within themalignant population. Thus, the dominant driver muta-tions in relapse samples are frequently already present inminor subclones before treatment.13 These observationssupport the well-established “watch and wait” strategy ofwithholding treatment for asymptomatic CLL. However, amore thorough understanding of these interclonal dynam-ics might anticipate which subpopulations could becomeproblematic in the future and thereby suggest approachesto personalize antecedent treatment.51 Early detection ofdriver lesions in small subclones may potentially requiresurveillance using ultra-deep sequencing approaches,although validation of this strategy is lacking. In supportof this idea, however, a recent study demonstrated thatpresence of minor subclones harboring TP53 mutationscomposed of as little as 2% of the cancer cells withina population in early CLL predated its emergence as adominant subclone at relapse and the development of achemo-refractory phenotype.52

Novel Therapeutic Targets

To date, the only genetic features that have driventherapeutic decisions are deletions of 17p and TP53alterations.48 However, the relative high frequency ofsomatic mutations in specific genes indicates that certainpathways are likely essential to CLL and that targetingthese gene pathways could provide effective therapy.Indeed, a number of agents targeting these genes andpathways are under development, and they includeinhibitors of Notch signaling (γ-secretase inhibitors),53

the spliceosome,54 nuclear RNA export,23 and telomer-ase.55 Another potential new target in CLL includes

activation-induced cytidine deaminase (AID), which drivessomatic hypermutation in normal B-cell development bythe generation of point mutations and initiation ofdouble-strand breaks.56,57 More than 40% of CLLs havemarked expression of AID; in these cases, inhibition ofhomologous recombination in vitro preferentially inducedapoptosis in leukemia cells.58

Besides pointing to potentially vulnerable pathways inCLL, recurrently mutated genes in CLL have also recentlybeen used to define the subclonal architecture of CLL.These studies have strongly suggested a specific temporalhierarchy in the acquisition of genetic and epigeneticchanges, with clonal events (ie, “earlier”) enriched foralterations, which more selectively affect B cells.13 Theseobservations suggest that elimination of CLL early indisease course may be particularly effective, before itsgenetic diversification. Consistent with the idea thattargeting the common “trunk”59 might be an effectivetherapeutic strategy, clinical studies targeting B cell–specific pathways, such as by using anti-CD20 antibodiesor novel kinase inhibitors of B-cell receptor signalingpathways, have demonstrated striking efficacy in CLL.60

FUTURE DIRECTIONS

Within the same time frame that studies using mas-sively parallel sequencing in CLL have been reported,analysis of other cancers using these technologies havebeen also completed. These datasets together with CLLdata have yielded information regarding the fundamentalmechanisms underlying somatic alteration in cancer andhave also identified processes that are common or uniqueto CLL compared with other tumors.56,57 Recent NGSefforts have examined the mutation spectrum of othermature B-cell lymphomas as well; not surprisingly, similaralterations appear across the malignancies, but CLL alsohas its own unique constellations of abnormalities(Figure 4).8,11,13,15,17,61–69

Overall, these studies across cancers have yielded anumber of further important insights. First, overall muta-tion rates of different malignancies have been established,and it is clear that CLL is a low mutation-rate tumor,similar to other leukemias, and is mutated up to 10-foldless than the carcinogen-induced cancers.56,57,70 Second,examination of the spectrum of mutations and theorganization of affected gene regions have revealedevidence that cancers occasionally undergo catastrophicshattering events (ie, so-called “chromothripsis”) as well ascoordinated genomic rearrangements across different chro-mosomes.71,72 Finally, comparison of CLL in relation toother cancers has revealed the spectrum of mutations inCLL to be consistent with a footprint of somatic hyper-mutation that is conventionally catalyzed by enzymes ofthe AID/APOBEC family of cytidine deaminases.56,57

At this juncture, the continued advances in technologyand analytic tools promise the ability to gain answers to anumber of critical questions in the near term. First, by

Figure 4. Recurrent putative driver alterations in mature B-cell non-Hodgkin lymphomas. Abbreviations: BL, Burkitt’slymphoma; CLL, chronic lymphocytic leukemia; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; MCL,mantle cell lymphoma; MM, multiple myeloma assessed by NGS (TS, targeted sequencing [NGS or Sanger Sequencing],RNAseq, or SNP array; WES, whole-exome sequencing; WGS, whole-genome sequencing); SMZL, splenic marginal zonelymphoma; WM, Waldenström’s macroglobulinemia. Shown are all alterations that were significant in CLL in at least onestudy. For alterations in other entities, selected alterations have been validated and occurred in a significant and highproportion (410% of cases) in at least one study or were identified across independent studies and were significant in atleast one study.

M. Gruber and C.J. Wu184

Understanding the CLL genome 185

integration of genetic and transcriptional data, we canunderstand genotype and phenotype relationships. Dis-tinct RNA expression modules have already been linked toCLL genotypes, and 460% of these associations werecorroborated in an independent validation cohort yieldingdefined hypotheses for experimental studies.73 Second, thelow mutation rate of CLL also suggests that we can likelygain comprehensive analysis of the mutation landscape in thenear future with a saturating number of exomes.70 Third,single-cell sequencing technologies promise to yield oppor-tunities to dissect genetic heterogeneity in CLL, further tracethe evolutionary tree of individual cases, and strengthenunderstanding of genotype and phenotype relationships inthis disease.74 Lastly, investigating CLL development bycomparison between monoclonal B-cell lymphocytosis andCLL, as well as by learning how germline variants predisposeto CLL, might help us to further understand evolutionarytrajectories in the disease. Recent meta-analyses of genome-wide population studies to examine CLL susceptibility haveassessed on the order of magnitude of thousands of patientsand control subjects. These efforts have corroborated 13previously known risk loci for CLL and have newly foundanother 13 loci associated with inherited disease suscepti-bility. Interestingly, some of the identified genes suggest anoverlap with regions of somatic driver alterations, involvingBCL2 or the 3' UTR of POT1.75,76 Striking reports on CLLcases with germline mutations in putative drivers exist forDAPK141; for SAMHD1, a nuclease involved in DNAdamage response77; and for genes of the microRNAprecursor miR-16-1-miR-15a.78

CONCLUSIONS

Recent technologic developments have yielded a breath-taking increase in our knowledge about the CLL genome.As we have reviewed herein, the insights gained fromdissecting the nature and basis of clonal evolution in CLLfuel a myriad of further lines of investigation that can haveconcrete impact on the clinical practice and treatment ofCLL. From a more basic biologic standpoint, future studieswill likely address whether different genomic alterationsprovide distinct roles at defined stages in development ofthe disease, from predisposition to initial transformation ofa clonal B-cell population and to progression with relapseddisease after treatment and even transformation.

REFERENCES1. Oscier D, Dearden C, Erem E, et al. Guidelines on the

diagnosis, investigation and management of chronic lym-phocytic leukaemia. Br J Haematol. 2012;159:541-64.

2. Rossi D, Rasi S, Spina V, et al. Integrated mutational andcytogenetic analysis identifies new prognostic subgroups inchronic lymphocytic leukemia. Blood. 2013;121:1403-12.

3. Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN,Pasternack BS. Clinical staging of chronic lymphocyticleukemia. Blood. 1975;46:219-34.

4. Binet JL, Auquier A, Dighiero G, et al. A new prognosticclassification of chronic lymphocytic leukemia derived froma multivariate survival analysis. Cancer. 1981;48:198-206.

5. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation statusand CD38 expression as novel prognostic indicators inchronic lymphocytic leukemia. Blood. 1999;94:1840-7.

6. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, StevensonFK. Unmutated Ig V(H) genes are associated with a moreaggressive form of chronic lymphocytic leukemia. Blood.1999;94:1848-54.

7. Dohner H, Stilgenbauer S, Benner A, et al. Genomicaberrations and survival in chronic lymphocytic leukemia.N Engl J Med. 2000;343:1910-6.

8. Puente XS, Pinyol M, Quesada V, et al. Whole-genomesequencing identifies recurrent mutations in chronic lym-phocytic leukaemia. Nature. 2011;475:101-5.

9. Wang L, Lawrence MS, Wan Y, et al. SF3B1 and othernovel cancer genes in chronic lymphocytic leukemia. N EnglJ Med. 2011;365:2497-506.

10. Schuh A, Becq J, Humphray S, et al. Monitoring chroniclymphocytic leukemia progression by whole genomesequencing reveals heterogeneous clonal evolution patterns.Blood. 2012;120:4191-6.

11. Fabbri G, Rasi S, Rossi D, et al. Analysis of the chroniclymphocytic leukemia coding genome: role of NOTCH1mutational activation. J Exp Med. 2011;208:1389-401.

12. Fabbri G, Khiabanian H, Holmes AB, et al. Genetic lesionsassociated with chronic lymphocytic leukemia transforma-tion to Richter syndrome. J Exp Med. 2013;210:2273-88.

13. Landau DA, Carter SL, Stojanov P, et al. Evolution andimpact of subclonal mutations in chronic lymphocyticleukemia. Cell. 2013;152:714-26.

14. Ramsay AJ, Quesada V, Foronda M, et al. POT1 mutationscause telomere dysfunction in chronic lymphocytic leuke-mia. Nat Genet. 2013;45:526-30.

15. Quesada V, Conde L, Villamor N, et al. Exome sequencingidentifies recurrent mutations of the splicing factor SF3B1gene in chronic lymphocytic leukemia. Nat Genet. 2012;44:47-52.

16. Ngo VN, Young RM, Schmitz R, et al. Oncogenically activeMYD88 mutations in human lymphoma. Nature. 2011;470:115-9.

17. Treon SP, Xu L, Yang G, et al. MYD88 L265P somaticmutation in Waldenstrom's macroglobulinemia. N Engl JMed. 2012;367:826-33.

18. Arruga F, Gizdic B, Serra S, et al. Functional impact ofNOTCH1 mutations in chronic lymphocytic leukemia.Leukemia. 2014;28:1060-70.

19. Wan Y, Wu CJ. SF3B1 mutations in chronic lymphocyticleukemia. Blood. 2013;121:4627-34.

20. Wan Y, Brooks A, Wang L, et al. SF3B1 mutation alters theselection of 3' RNA splice sites in chronic lymphocyticleukemia. Blood. 2013;122:117.

21. Ferreira PG, Jares P, Rico D, et al. Transcriptome character-ization by RNA sequencing identifies a major molecular andclinical subdivision in chronic lymphocytic leukemia.Genome Res. 2014;24:212-26.

22. Jeromin S, Weissmann S, Haferlach C, et al. SF3B1mutations correlated to cytogenetics and mutations inNOTCH1, FBXW7, MYD88, XPO1 and TP53 in 1160untreated CLL patients. Leukemia. 2014;28:108-17.

M. Gruber and C.J. Wu186

23. Lapalombella R, Sun Q, Williams K, et al. Selectiveinhibitors of nuclear export show that CRM1/XPO1 is atarget in chronic lymphocytic leukemia. Blood. 2012;120:4621-34.

24. Rossi D, Fangazio M, Rasi S, et al. Disruption of BIRC3associates with fludarabine chemorefractoriness in TP53wild-type chronic lymphocytic leukemia. Blood. 2012;119:2854-62.

25. Cortese D, Sutton LA, Cahill N, et al. On the way towards a'CLL prognostic index': focus on TP53, BIRC3, SF3B1,NOTCH1 and MYD88 in a population-based cohort.Leukemia. 2014;28:710-3.

26. Weissmann S, Roller A, Jeromin S, et al. Prognostic impactand landscape of NOTCH1 mutations in chronic lympho-cytic leukemia (CLL): a study on 852 patients. Leukemia.2013;27:2393-6.

27. Oscier DG, Rose-Zerilli MJ, Winkelmann N, et al. Theclinical significance of NOTCH1 and SF3B1 mutations inthe UK LRF CLL4 trial. Blood. 2013;121:468-75.

28. Rossi D, Bruscaggin A, Spina V, et al. Mutations of theSF3B1 splicing factor in chronic lymphocytic leukemia:association with progression and fludarabine-refractoriness.Blood. 2011;118:6904-8.

29. Rossi D, Rasi S, Fabbri G, et al. Mutations of NOTCH1 arean independent predictor of survival in chronic lymphocyticleukemia. Blood. 2012;119:521-9.

30. Messina M, Del Giudice I, Khiabanian H, et al. Geneticlesions associated with chronic lymphocytic leukemiachemo-refractoriness. Blood. 2014;123:2378-88.

31. Bea S, Lopez-Guillermo A, Ribas M, et al. Geneticimbalances in progressed B-cell chronic lymphocytic leuke-mia and transformed large-cell lymphoma (Richter's syn-drome). Am J Pathol. 2002;161:957-68.

32. Shanafelt TD, Hanson C, Dewald GW, et al. Karyotypeevolution on fluorescent in situ hybridization analysis isassociated with short survival in patients with chroniclymphocytic leukemia and is related to CD49d expression.J Clin Oncol. 2008;26:e5-6.

33. Shanafelt TD, Witzig TE, Fink SR, et al. Prospectiveevaluation of clonal evolution during long-term follow-upof patients with untreated early-stage chronic lymphocyticleukemia. J Clin Oncol. 2006;24:4634-41.

34. Braggio E, Kay NE, VanWier S, et al. Longitudinal genome-wide analysis of patients with chronic lymphocytic leukemiareveals complex evolution of clonal architecture at diseaseprogression and at the time of relapse. Leukemia. 2012;26:1698-701.

35. Shen H, Laird PW. Interplay between the cancer genomeand epigenome. Cell. 2013;153:38-55.

36. Ziller MJ, Gu H, Muller F, et al. Charting a dynamic DNAmethylation landscape of the human genome. Nature. 2013;500:477-81.

37. Rahmatpanah FB, Carstens S, Hooshmand SI, et al. Large-scale analysis of DNA methylation in chronic lymphocyticleukemia. Epigenomics. 2009;1:39-61.

38. Kanduri M, Cahill N, Goransson H, et al. Differentialgenome-wide array-based methylation profiles in prognosticsubsets of chronic lymphocytic leukemia. Blood. 2010;115:296-305.

39. Hanada M, Delia D, Aiello A, Stadtmauer E, Reed JC. Bcl-2gene hypomethylation and high-level expression in B-cellchronic lymphocytic leukemia. Blood. 1993;82:1820-8.

40. Yuille MR, Condie A, Stone EM, et al. TCL1 is activated bychromosomal rearrangement or by hypomethylation. GenesChromosomes Cancer. 2001;30:336-41.

41. Raval A, Tanner SM, Byrd JC, et al. Downregulation ofdeath-associated protein kinase 1 (DAPK1) in chroniclymphocytic leukemia. Cell. 2007;129:879-90.

42. Cahill N, Bergh AC, Kanduri M, et al. 450K-array analysisof chronic lymphocytic leukemia cells reveals global DNAmethylation to be relatively stable over time and similar inresting and proliferative compartments. Leukemia. 2013;27:150-8.

43. Baer C, Claus R, Frenzel LP, et al. Extensive promoter DNAhypermethylation and hypomethylation is associated withaberrant microRNA expression in chronic lymphocyticleukemia. Cancer Res. 2012;72:3775-85.

44. Garding A, Bhattacharya N, Claus R, et al. Epigeneticupregulation of lncRNAs at 13q14.3 in leukemia is linked tothe In Cis downregulation of a gene cluster that targetsNF-kB. PLoS Genet. 2013;9:e1003373.

45. Kulis M, Heath S, Bibikova M, et al. Epigenomic analysisdetects widespread gene-body DNA hypomethylation inchronic lymphocytic leukemia. Nat Genet. 2012;44:1236-42.

46. Landau DA, Clement K, Boyle P, et al. Increased localdisorder of DNA methylation forms the basis of high intra-leukemic epigenetic heterogeneity and enhances CLL evo-lution. Blood. 2013;122:596.

47. Oakes CC, Claus R, Gu L, et al. Evolution of DNAmethylation is linked to genetic aberrations in chroniclymphocytic leukemia. Cancer Discov. 2014;4:348-61.

48. Pospisilova S, Gonzalez D, Malcikova J, et al. ERICrecommendations on TP53 mutation analysis in chroniclymphocytic leukemia. Leukemia. 2012;26:1458-61.

49. Zenz T, Eichhorst B, Busch R, et al. TP53 mutation andsurvival in chronic lymphocytic leukemia. J Clin Oncol.2010;28:4473-9.

50. Schnaiter A, Paschka P, Rossi M, et al. NOTCH1, SF3B1,and TP53 mutations in fludarabine-refractory CLL patientstreated with alemtuzumab: results from the CLL2H trial ofthe GCLLSG. Blood. 2013;122:1266-70.

51. Puente XS, Lopez-Otin C. The evolutionary biography ofchronic lymphocytic leukemia. Nat Genet. 2013;45:229-31.

52. Rossi D, Khiabanian H, Spina V, et al. Clinical impact ofsmall TP53 mutated subclones in chronic lymphocyticleukemia. Blood. 2014;123:2139-47.

53. Groth C, Fortini ME. Therapeutic approaches to modulat-ing Notch signaling: current challenges and future prospects.Semin Cell Dev Biol. 2012;23:465-72.

54. Bonnal S, Vigevani L, Valcarcel J. The spliceosome as atarget of novel antitumour drugs. Nat Rev Drug Discov.2012;11:847-59.

55. Ruden M, Puri N. Novel anticancer therapeutics targetingtelomerase. Cancer Treat Rev. 2013;39:444-56.

56. Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signaturesof mutational processes in human cancer. Nature. 2013;500:415-21.

57. Lawrence MS, Stojanov P, Polak P, et al. Mutationalheterogeneity in cancer and the search for new cancer-associated genes. Nature. 2013;499:214-8.

58. Lamont KR, Hasham MG, Donghia NM, et al. Attenuatinghomologous recombination stimulates an AID-inducedantileukemic effect. J Exp Med. 2013;210:1021-33.

Understanding the CLL genome 187

59. Greaves M, Maley CC. Clonal evolution in cancer. Nature.2012;481:306-13.

60. Niemann CU, Jones J, Wiestner A. Towards targetedtherapy of chronic lymphocytic leukemia. Adv Exp MedBiol. 2013;792:259-91.

61. Bea S, Valdes-Mas R, Navarro A, et al. Landscape of somaticmutations and clonal evolution in mantle cell lymphoma.Proc Natl Acad Sci U S A. 2013;110:18250-5.

62. Okosun J, Bodor C, Wang J, et al. Integrated genomicanalysis identifies recurrent mutations and evolution patternsdriving the initiation and progression of follicular lym-phoma. Nat Genet. 2014;46:176-81.

63. Love C, Sun Z, Jima D, et al. The genetic landscape ofmutations in Burkitt lymphoma. Nat Genet. 2012;44:1321-5.

64. Richter J, Schlesner M, Hoffmann S, et al. Recurrentmutation of the ID3 gene in Burkitt lymphoma identifiedby integrated genome, exome and transcriptome sequencing.Nat Genet. 2012;44:1316-20.

65. Schmitz R, Young RM, Ceribelli M, et al. Burkitt lym-phoma pathogenesis and therapeutic targets from structuraland functional genomics. Nature. 2012;490:116-20.

66. Pasqualucci L, Trifonov V, Fabbri G, et al. Analysis of thecoding genome of diffuse large B-cell lymphoma. NatGenet. 2011;43:830-7.

67. Rossi D, Trifonov V, Fangazio M, et al. The coding genomeof splenic marginal zone lymphoma: activation ofNOTCH2 and other pathways regulating marginal zonedevelopment. J Exp Med. 2012;209:1537-51.

68. Chapman MA, Lawrence MS, Keats JJ, et al. Initial genomesequencing and analysis of multiple myeloma. Nature.2011;471:467-72.

69. Lohr JG, Stojanov P, Carter SL, et al. Widespread geneticheterogeneity in multiple myeloma: implications for targetedtherapy. Cancer Cell. 2014;25:91-101.

70. Lawrence MS, Stojanov P, Mermel CH, et al. Discovery andsaturation analysis of cancer genes across 21 tumour types.Nature. 2014;505:495-501.

71. Stephens PJ, Greenman CD, Fu B, et al. Massive genomicrearrangement acquired in a single catastrophic event duringcancer development. Cell. 2011;144:27-40.

72. Baca SC, Prandi D, Lawrence MS, et al. Punctuatedevolution of prostate cancer genomes. Cell. 2013;153:666-77.

73. Kotliar D, Landau DA, Litvin O, et al. Reconstructing agenotype-phenotype map in chronic lymphocytic leukemia.Blood. 2013;122:2857.

74. Method of the year 2013. Nat Methods. 2014;11:1.75. Berndt SI, Skibola CF, Joseph V, et al. Genome-wide

association study identifies multiple risk loci for chroniclymphocytic leukemia. Nat Genet. 2013;45:868-76.

76. Speedy HE, Di Bernardo MC, Sava GP, et al. A genome-wide association study identifies multiple susceptibility locifor chronic lymphocytic leukemia. Nat Genet. 2014;46:56-60.

77. Clifford R, Louis T, Robbe P, et al. SAMHD1 is mutatedrecurrently in chronic lymphocytic leukemia and isinvolved in response to DNA damage. Blood. 2014;123:1021-31.

78. Calin GA, Ferracin M, Cimmino A, et al. A MicroRNAsignature associated with prognosis and progression in chroniclymphocytic leukemia. N Engl J Med. 2005;353:1793-801.

79. Mansouri L, Cahill N, Gunnarsson R, et al. NOTCH1 andSF3B1 mutations can be added to the hierarchical prognos-tic classification in chronic lymphocytic leukemia. Leuke-mia. 2013;27:512-4.

80. Gonzalez D, Martinez P, Wade R, et al. Mutational status ofthe TP53 gene as a predictor of response and survival inpatients with chronic lymphocytic leukemia: results from theLRF CLL4 trial. J Clin Oncol. 2011;29:2223-9.

81. Zenz T, Habe S, Denzel T, et al. Detailed analysis of p53pathway defects in fludarabine-refractory chronic lympho-cytic leukemia (CLL): dissecting the contribution of 17pdeletion, TP53 mutation, p53-p21 dysfunction, and miR34ain a prospective clinical trial. Blood. 2009;114:2589-97.

82. Balatti V, Bottoni A, Palamarchuk A, et al. NOTCH1mutations in CLL associated with trisomy 12. Blood. 2012;119:329-31.

83. Zenz T, Vollmer D, Trbusek M, et al. TP53 mutationprofile in chronic lymphocytic leukemia: evidence for adisease specific profile from a comprehensive analysis of 268mutations. Leukemia. 2010;24:2072-9.