Portal de Periódicos da CAPES

Different impact of NOTCH 1 and SF 3B1 mutations on the risk of chronic lymphocytic leukemia transformation to Richter syndrome

2012; Wiley; Volume: 158; Issue: 3 Linguagem: Inglês

10.1111/j.1365-2141.2012.09155.x

1365-2141

Davide Rossi, Silvia Rasi, Valeria Spina, Marco Fangazio, Sara Monti, Mariangela Greco, Carmela Ciardullo, Rosella Famà, Stefania Cresta, Alessio Bruscaggin, Luca Laurenti, Maurizio Martini, Pellegrino Musto, Francesco Forconi, Roberto Marasca, Luigi Maria Larocca, Robin Foà, Gianluca Gaïdano,

PI3K/AKT/mTOR signaling in cancer

Richter syndrome (RS) represents the development of an aggressive lymphoma, most commonly diffuse large B-cell lymphoma (DLBCL), in the context of chronic lymphocytic leukaemia (CLL). At least two types of RS exist: (i) transformation of CLL into a clonally related DLBCL, that accounts for ~80% of cases; and (ii) development of a DLBCL unrelated to the CLL clone. Clonally related RS and clonally unrelated RS are distinct disorders (Rossi et al, 2011a). Clinically, transformation into a clonally related RS is frequently lethal with an expected survival of a few months, while CLL patients developing a clonally unrelated RS display a survival probability in the range of de novo DLBCL (Rossi et al, 2011a). Biologically, clonally related RS frequently acquire genetic lesions of TP53, MYC and NOTCH1, which are otherwise absent or exceptional in clonally unrelated RS (Rossi et al, 2011a). NOTCH1 and SF3B1 mutations are predictors of poor outcome in CLL (Puente et al, 2011; Rossi et al, 2011b, 2012; Wang et al, 2011; Quesada et al, 2012). Here, we investigated a large multi-institutional series of 605 CLL patients to verify whether NOTCH1 and SF3B1 mutations at diagnosis may help in the identification of CLL at high risk of RS transformation. The study was approved by the Ethical Committee of the Ospedale Maggiore della Carità di Novara associated with the Amedeo Avogadro University of Eastern Piedmont (Protocol Code 59/CE; Study Number CE 8/11). Patients provided informed consent in accordance with local Institutional Review Board requirements and the Declaration of Helsinki. CLL was diagnosed according to National Cancer Institutes-International Workshop on CLL criteria (Hallek et al, 2008). RS diagnosis was histologically-proven (Müller-Hermelink et al, 2008). NOTCH1, SF3B1, TP53 and IGHV mutations were analysed by Sanger sequencing (Rossi et al, 2012). Probes (Abbott) used for fluorescent in situ hybridization analysis were: LSI13 and LSID13S319, CEP12, LSIp53 and LSIATM (Rossi et al, 2011a). CD38 expression was analysed by flow cytometry, utilizing a cut-off of 30% to define positivity (Rossi et al, 2009). VH CDR3 subset analysis and nomenclature were according to Murray et al (2008). Clonal relationship between CLL/DLBCL paired samples was established by comparing IGHV-D-J sequences. Time-to-event analysis utilized the Kaplan–Meier method. The cumulative incidence of RS development was calculated accounting for death as a competing risk, and was compared across groups with the Gray's test. The association between exposure variables and outcome was estimated by univariate and multivariate Cox regression analysis. Further details are reported in Supplementary Methods. Features of the CLL cohort are reported in Table SI. At CLL presentation, NOTCH1 mutations occurred in 12·2% patients, being mostly represented (82·4%) by a recurrent two bp frameshift deletion (c.7544_7545delCT). SF3B1 mutations occurred in 5·8% cases and were mainly represented (50·0%) by the K700E missense substitution. After a median follow-up of 6·4 years, 30/605 (4·9%) CLL had transformed into a clonally related RS, while 10/605 (1·6%) had developed a clonally unrelated DLBCL. NOTCH1 mutations at disease presentation recurred in 12/30 (40·0%) CLL that subsequently transformed into a clonally related RS, whereas they were consistently absent in the 10 CLL that developed a clonally unrelated RS. SF3B1 mutations were absent in all CLL that subsequently transformed to a RS, both clonally related or unrelated. The close relationship between NOTCH1 mutations and clonally related RS was reinforced by the following observations: (i) NOTCH1 mutations were the second most frequent genetic lesion (40·0%) at transformation after TP53 disruption (60·0%); and (ii) NOTCH1 mutations observed in the RS phase were already present at CLL presentation in all patients, either at the clonal (n = 5) or subclonal (n = 7) level, suggesting a preferential selection of NOTCH1 mutated clones during disease progression and transformation. Notably, none of the clonally related RS acquired SF3B1 mutations at transformation. NOTCH1 mutations were associated with a ~5·8-fold increase in the crude hazard of transformation into a clonally related RS (HR: 5·81; P < 0·001) (Table 1). Consistently, CLL harbouring NOTCH1 mutations had a significantly higher cumulative probability of transforming into a clonally related RS (45·0% at 15 years) compared to CLL without NOTCH1 mutations (4·6% at 15 years) (P < 0·001) (Fig. 1A). The association between NOTCH1 mutations and transformation into a clonally related RS also remained significant after correcting for death as a competing risk (P < 0·001). Conversely, NOTCH1 mutations had no impact on the development of a clonally unrelated DLBCL (P = 0·494). NOTCH1 mutations were significantly enriched among cases harbouring other factors known to affect RS risk, including CD38 expression (27·6%, P < 0·001) and IGHV4-39 gene usage (36·8%, P = 0·005), especially if rearranged in a stereotyped fashion in subset 8 (66·6%; P < 0·001) (Rossi et al, 2009). By multivariate analysis, NOTCH1 mutations were selected as an independent risk factor of transformation into a clonally related RS (HR: 3·71; P = 0·002), after adjusting for trisomy 12 and/or 11q22-q23 deletion and/or TP53 disruption, IGHV4-39 usage, and CD38 expression (Table 1), which, at CLL diagnosis, are well established risk factors for subsequent RS development (Rossi et al, 2009). The observation of a high frequency of NOTCH1 mutations in IGHV4-39 positive CLL prompted investigation of the interaction between IGHV4-39 usage and NOTCH1 mutations in the model for clonally RS prediction. Accordingly, bivariate analysis segregated three risk categories (Fig. 1B): (i) CLL harbouring both NOTCH1 mutations and IGHV4-39, which constituted the disease category with the highest risk of transformation; (ii) CLL harbouring NOTCH1 mutations but lacking IGHV4-39, which showed an intermediate risk of transformation; and (iii) CLL lacking NOTCH1 mutations, independent of IGHV4-39 usage, which showed a very low risk of transformation. SF3B1 mutations, a new biomarker of CLL progression (Rossi et al, 2011b; Wang et al, 2011; Quesada et al, 2012), showed no impact on RS development (P = 0·616) (Table 1). This observation further reinforces the notion that predictors of CLL transformation are different, at least in part, from predictors of other CLL clinical outcomes (Rossi et al, 2009). This study documents that NOTCH1 mutations are an independent prognosticator of CLL transformation into clonally related RS, whereas they have no impact on the development of clonally unrelated RS. Importantly, up to 40–50% of NOTCH1 mutated patients are ultimately projected to develop and die of clonally related RS. On these bases, the poor prognosis associated with NOTCH1 mutations in CLL may be explained, at least in part, by a substantial risk of developing RS. This study was supported by AIRC, Special Program Molecular Clinical Oncology, 5 × 1000, No. 10007, Milan, Italy (to G.G. and to R.Foà.); Progetto FIRB-Programma "Futuro in Ricerca" 2008 (to D.R.), PRIN 2008 (to G.G.) and 2009 (to D.R.), MIUR, Rome, Italy; Progetto Giovani Ricercatori 2008 (to D.R.) and Ricerca Sanitaria Finalizzata 2008 (to G.G. and P.M.), Ministero della Salute, Rome, Italy; Novara-AIL Onlus, Novara, Italy (to G.G.); Compagnia di San Paolo, Turin, Italy (to R.Foà.). S.M. is supported by fellowships from Novara-AIL Onlus, Novara. D.R. and G.G. designed the study, interpreted data and wrote the manuscript; S.R., V.S., M.G., C.C., R.Famà, S.C, and A.B performed molecular studies and interpreted data; S.M. performed FISH analysis and interpreted data; D.R. and M.F. performed statistical analysis L.L., M.M, F.F, and R.M. provided well characterized biological samples and clinical data; P.M., L.M.L., and R.Foà. contributed to data analysis and interpretation. The authors have no conflict of interest to disclose. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.