Open access journal of the Ferrata-Storti Foundation, a non-profit organization
Open access journal of the Ferrata-Storti Foundation, a non-profit organization
Trisomy 12 represents the third most frequent chromosomal aberration in chronic lymphocytic leukemia (CLL) (15–20% of cases) and often (approx. 60% of cases) occurs as the sole cytogenetic lesion.1,2 Within the hierarchical model of genetic subgroups commonly used in clinical practice, +12 as single aberration confers an intermediate prognostic risk, with a median time to progression of 33 months and a median overall survival (OS) of 114 months.1
TheNOTCH1 gene has been shown to have an essential biological role in hematopoiesis.3 Following the pivotal study that identifiedNOTCH1 mutations in CLL and provided initial evidence on the unfavorable clinical outcome associated withNOTCH1 alterations,4 two independent studies of the CLL coding genome have recently identified activating mutations of theNOTCH1 gene in approximately 10% of CLL at diagnosis.5,6 The prevalence ofNOTCH1 mutations increases with disease aggressiveness.5 At diagnosis,NOTCH1 mutations show an adverse impact on outcome, confirmed in at least four series,4–7 and act independently of other clinico-biological features, includingTP53 disruption.7 Among CLL cytogenetic subgroups,NOTCH1 mutations are distributed in a mutually exclusive fashion withTP53 disruption and are enriched in CLL carrying +12, where they recur in approximately 25% of patients.7
Based on the emerging association betweenNOTCH1 alterations and +12, we investigatedNOTCH1 mutations in a series of untreated +12 CLL. We observed that in these patients,NOTCH1 mutations: i) cluster within cases with no additional cytogenetic abnormalities; ii) induce a particular transcriptional profile; and iii) refine outcome prediction.
This multicenter study evaluated 104 patients carrying +12: 54 were males and 50 females, with a median age of 65 years (interquartile range 56–72). All cases satisfied the IWCLL diagnostic criteria for CLL8 and were selected on the basis of: i) untreated disease; ii) availability of biological material; and iii) presence of +12, independent of additional chromosomal abnormalities.
Patients gave their informed consent to blood collection and biological analyses, in agreement with the Declaration of Helsinki. The study was approved by the Ethical Committee of Policlinico Umberto I, “La Sapienza” University of Rome (n. 2182/16.06.2011) and of the Ospedale Maggiore della Carità, Novara, northern Italy, associated with the Amedeo Avogadro University of Eastern Piedmont (protocol code 59/CE; study n. CE 8/11).
Lymphocyte morphology, immunophenotype, FISH analysis,IGHV andTP53 sequencing were performed as previously described.9
TheNOTCH1 (exon 34; RefSeq NM_017617.2) mutation hotspot previously identified in CLL7 was analyzed by direct sequencing of genomic DNA extracted from blood mononuclear cells. Purified amplicons were subjected to conventional DNA Sanger sequencing using the ABI PRISM 3100 Genetic Analyzer (Applied Biosystems, CA, USA). The presence ofNOTCH1 c.7544_7545delCT alleles was also investigated by ARMS PCR. Further details are reported in theOnline Supplementary Design and Methods.
Overall survival (OS) was measured from the date of initial presentation to the date of death (event) or last follow up (censoring). Survival analysis was performed by the Kaplan-Meier method. Further details are reported in theOnline Supplementary Design and Methods.
For oligonucleotide array analysis, the HGU133 Plus 2.0 gene chips (Affymetrix, Santa Clara, CA, USA) were used. Sample preparation and microarray processing were performed as previously described.10 Further details are reported in theOnline Supplementary Design and Methods.
Table 1A.Characteristics of CLL patients harboring +12 according to theNOTCH1 mutation status.
Table 1B.Univariate and multivariate analysis for overall survival in CLL patients harboring +12a.
NOTCH1 mutations occurred in 25 of the 104 untreated CLL with +12 investigated (24%) (Table 1A), were represented in all cases by frameshift deletions, including the c.7544_7545delCT in 22 of 25 (88%) cases, and preferentially associated with use of unmutatedIGHV genes (84%,P=0.003).NOTCH1 mutations occurred independent of gender, thus suggesting thatNOTCH1 mutations might be an important marker of unfavorable prognosis in both male and female CLL patients.
Figure 1.(A) Distribution ofNOTCH1 mutations among genetic subgroups of +12 CLL. (B) Overall survival according toNOTCH1 mutation status in +12 CLL. (C) Hierarchical stratification of overall survival according to genetic lesions in +12 CLL. (D) Comparison betweenNOTCH1 mutated andNOTCH1 wt CLL samples. Relative levels of gene expression are depicted with a color scale: red represents the highest level of expression and blue represents the lowest level. The table reports the functional annotation analysis, performed using the DAVID database, of differentially expressed genes betweenNOTCH1 mutated andNOTCH1 wt CLL. The biological processes reported are ordered according to theirP value.
Trisomy 12 occurred as an isolated chromosomal abnormality in 73 of 104 (70%) cases, while it was associated to other cytogenetic abnormalities in 31 of 104 (30%) cases (Figure 1A). Mutual information analysis revealed a clustering ofNOTCH1 mutations among CLL harboring +12 as a sole abnormality (22 of 73, 30%) compared to patients harboring +12 in addition to other cytogenetic lesions (2 of 31, 6%) (I=0.027;P=0.008) (Figure 1A). Consistently, +12 CLL harboringNOTCH1 mutations carried deletion 13q14 only exceptionally (0 of 19). Consistent with pivotal observations,5,7 also in +12 CLL,NOTCH1 mutations distributed in a mutually exclusive fashion with deletions of 17p13 and 11q22-q23 (Figure 1A).
This extended cohort corroborates the high prevalence ofNOTCH1 mutations in +12 CLL, where the overall frequency ofNOTCH1 mutations in +12 patients consistently ranges from 24.5 to 28.6%.7,11 Furthermore, +12 patients harboringNOTCH1 mutations prevalently belong to aggressive cases, i.e. cases with an unmutatedIGHV gene status, in line with recent findings,7,11 and expression of CD38 (NOTCH1 mutated/CD38 positive, n=20 of 25 (80.0%)vs. NOTCH1 wild-type/CD38 positive, n=39 of 77 (50.6%) (P=0.010).
At variance, no difference emerged in the distribution of ZAP-70 positivity (NOTCH1 mutated/ZAP-70 positive, n=15 of 24 (62.5%)vs. NOTCH1 wild-type/ZAP-70 positive, n=34 of 77 (44.2%) (P=0.116).
Finally, the analysis of this specific cohort of patients showed an enrichment ofNOTCH1 mutations in CLL harboring+12 as sole cytogenetic abnormality.
After a median follow up of seven years, 36 of 103 evaluable patients had died, for a median OS of 11.6 years (95% CI: 10.4–12.7). By univariate analysis, the crude impact ofNOTCH1 mutations on survival in +12 CLL was an approximately 2.4 fold increase in the hazard of death (HR: 2.46; 95% CI: 1.26–4.81) and a significant shortening of OS (P<0.01) (Table 1B;Figure 1B). Other variables associated with shorter OS were age, Binet stage,IGHV mutation status andTP53 disruption.
Multivariate analysis selectedNOTCH1 mutations as an independent risk factor of OS (HR: 3.04; 95% CI: 1.39–6.64;P<0.01), after adjusting for age (>vs. ≤65 years), Binet stage (B-Cvs. A),IGHV mutation status andTP53 disruption by mutation and/or deletion.NOTCH1 mutations, 13q14 deletion, 11q22-q23 deletion andTP53 status were used to build a hierarchical model of genetic subgroups to predict OS in CLL with +12. The outcome of +12 CLL withNOTCH1 mutations was poor, similar to cases with +12 andTP53 disruption or 11q22-q23 deletion and significantly worse than patients with +12 as an isolated abnormality or plus 13q14 deletion (Figure 1C).
NOTCH1 mutations represent, therefore, an independent adverse prognostic factor of OS among +12 CLL, providing a new genetic prognostic stratification of patients with this intermediate risk marker.
To understand whetherNOTCH1 mutations induced a distinctive transcriptional profile in +12 CLL, we compared 7NOTCH1 mutatedvs. 7NOTCH1 wild-type cases in a cohort of patients carrying +12 (Online Supplementary Table S1).
This analysis showed thatNOTCH1 mutated cases formed a tight clustering (Figure 1D), with 2 patients incorrectly placed. Of these, one later developed aTP53 mutation and a myelodysplastic syndrome (CLL_00248,Online Supplementary Table S1).
Sixty-five differentially expressed genes (Online Supplementary Table S2) were selected, the majority being upmodulated inNOTCH1 mutated samples. DAVID functional annotation analysis highlighted an overrepresentation of cell cycle related genes, indicating thatNOTCH1 mutations induce a proliferative advantage that might explain the clinically aggressive behavior (Figure 1D). We also observed significantly higher levels of IgM expression inNOTCH1 mutated cases. It is known that IgM expression is higher in cells with increased ability to respond to external stimuli,12,13 indicating that theNOTCH1 mutated clone might survive and expand also thanks to these interactions. Intriguingly, approximately 30% of the upregulated transcripts were located on chromosome 12. This might be due to the fact that +12 was present as a single alteration in allNOTCH1 mutated cases, whereas in theNOTCH1 wild-type subgroup the scenario was more complex, with 3 cases displaying only +12 and 4 cases one or more additional chromosomal aberrations.
Among the transcripts previously reported to be associated with a +12 signature, we confirmed the upregulation ofANAPC5,GLIPR1,TIMELESS andSLC2A6.14,15
In summary, this study highlights thatNOTCH1 mutations in +12 CLL: i) preferentially cluster with cases harboring +12 as sole genetic abnormality; ii) account for 24% of cases and are more frequently detected in cases with unfavorable biological markers; iii) associate with a particular gene expression profile; iv) predict a poor outcome, stratifying +12 CLL in two distinct subgroups; and v) are not associated withTP53 disruption or 11q deletion, thus making them even more useful in a genetic hierarchical prognostic model, given their mutual exclusivity. This observation sheds light on the heterogeneous clinical course of +12 CLL patients and allows us to refine the intermediate prognostic risk of this chromosomal lesion. The transcriptional profile characterized by an overrepresentation of genes involved in cell cycle and proliferation suggests the potential biological contribution ofNOTCH1 mutations in determining the aggressive behavior of the disease.
This work was supported by AIRC (to RF) and AIRC, Special Program Molecular Clinical Oncology, 5 × 1000, n. 10007, Milan, Italy (to GG and to RF), Compagnia di San Paolo, Torino, Italy, Fondazione per le Biotecnologie, Torino, Italy, Fondazione Buzzati-Traverso, Pavia, Italy, Progetto FIRB-Programa “Futuro in Ricerca” 2008 (to DR), PRIN 2008 (to GG) and 2009 (to DR), MIUR, Rome, Italy, Progetto Giovani Ricercatori 2008 (to DR), Ministero della Salute, Rome, Italy, and Novara-AIL Onlus (to GG), Novara, Italy.