Search

FCR improves PFS and OS in CLL. Previous non-randomised Phase II trials suggest adding mitoxantrone to FCR (FCM-R) improves outcomes. The rituximab dose in CLL was inferred from lymphoma but in CLL even low doses of rituximab lead to the loss of CD20 expression (“shaving”) suggesting that low dose rituximab with chemotherapy may be effective.The primary objective was to assess whether the CR rate to FCM-miniR was not inferior to FCR. Secondary objectives were MRD eradication, overall response rates and safety. Late endpoints (not reported) were PFS, OS and MRD relapse.ARCTIC was a phase IIB, randomised, controlled, parallel group trial in previously untreated CLL performed in 34 UK centres. 206 patients were to be randomised to FCR or FCM-miniR with an 80% power to show that FCM-miniR did not lead to 10% worse CR rates compared to FCR. A formal interim analysis on the primary endpoint was planned after 103 patients. The schedule for oral FCR was equivalent to iv FCR with iv rituximab given on Day 1 (375mg/m2 Cycle 1; 500mg/m2 Cycles 2-6), oral fludarabine (24mg/m2/day for 5 days; Days 1-5) and oral cyclophosphamide (150mg/m2/day for 5 days; Days 1-5). FCM-miniR included intravenous mitoxantrone (6mg/m2/day) and 100mg rituximab on Day 1 of 6 cycles given 28 days apart. Patients with neutropenia delaying therapy received G-CSF (lenograstim 263mcg/day; Days 7-13) on all remaining cycles. Prophylaxis with co-trimoxazole and acyclovir was used.200 patients were recruited, 100 to each arm. The interim analysis after 103 patients indicated that although the CR rate to FCM-miniR was similar to predicted it was inferior to FCR. The independent DMC advised that the trial would not show non-inferiority for FCM-miniR and recruitment was stopped. A further 97 patients were recruited up to the interim analysis. 21 patients still receiving FCM-miniR were advised to cross-over to FCR. 100 patients received FCR, 79 FCM-miniR and 21 patients randomised to FCM-miniR crossed over to FCR.The treatment arms were well balanced for the following: median age 63 (36-80) and 20% (n=40) over 70; 67.5% male; 17% prog. stage A, 47% stage B and 35.5% stage C. 33.5% had a creatinine clearance under 70ml/min. 60% (87/146) had unmutated Ig genes; 4% (6/164) 17p deletion; and 15% (26/168) 11q deletion. More 11q del patients were randomised to FCM-miniR (15 vs 7). 29.5% (59/200) discontinued therapy early including 30% (n=30) FCR and 35.4% (n=28) FCM-miniR. 141 (70.5%) completed 6 cycles. 93 (49%) of 189 received G-CSF - more with FCM-miniR compared with FCR (54% vs 44%). 117 (58.5%) patients experienced a dose modification; 57% for FCR and 62% for FCM-miniR; the most common being dose delay in 47% of patients. 181 SAE's were reported from 104 patients; 79 events from 49% (49/100) receiving FCR, 80 from 58% (46/79) receiving FCM-miniR and 47% (9/22) cross-over patients. 62.5% of related SAE's were infectious. 3 patients died due to an SAE, 1 FCR and 2 FCM-miniR.We report only the 179 intention-to-treat patients who did not cross-over. To date 146 of 179 patients (82%) have undergone independent central review. 70% (102/146) achieved a CR or CRi; 78% (62/79) FCR and 60% (40/67) FCM-miniR. 13/22 (59%) patients with a baseline creatinine clearance of <60ml/min (received 50% fludarabine dose) achieved a CR or CRi. ORR was 94% (145/154) in evaluable patients, 96% (81/84) FCR and 91% (64/70) FCM-miniR. MRD by flow cytometry (sensitivity <10-4) was assessed in the marrow 3 months post-therapy; of assessable patients to date 51% (78/152) had undetectable MRD, 55% (47/85) FCR and 46% (31/67) FCM-miniR.ARCTIC was stopped prematurely as it was clear that the arm containing mitoxantrone and low dose rituximab would not be non-inferior to FCR. However both FCR and FCM-miniR performed well with CR rates of 78% and 60% respectively. This compares favourably to CR rates of 38% for FC in LRF CLL4 Trial and 44% for FCR in the GCLLSG CLL8 Trial. MRD was undetectable in 55% of patients receiving FCR and 46% for FCM-miniR. Oral FCR in front-line CLL results in extremely high CR rates and the eradication of detectable MRD in the majority. Low dose rituximab (100mg per cycle; 600mg total dose compared to 2875mg/m2 for standard FCR) in combination with FCM appears to be effective although FCM-miniR is inferior to full dose FCR. However there is more toxicity associated with the addition of mitoxantrone confounding the comparison of rituximab doses.Hillmen: Pharmacyclics: Research Funding; Janssen: Honoraria, Research Funding; Roche Pharmaceuticals: Honoraria, Research Funding; GlaxoSmithKline: Honoraria, Research Funding; Celgene: Honoraria. Off Label Use: The use of ofatumumab in combination with chlorambucil in previously untreated CLL. The reported trial will support the extension of the ofatumumab licence. Schuh:GSK: Honoraria; Celgene: Honoraria; Mundipharma: Honoraria.

FCR improves PFS and OS in CLL. Previous non-randomised Phase II trials suggest adding mitoxantrone to FCR (FCM-R) improves outcomes. The rituximab dose in CLL was inferred from lymphoma but in CLL even low doses of rituximab lead to the loss of CD20 expression (“shaving”) suggesting that low dose rituximab with chemotherapy may be effective.

No relevant conflicts of interest to declare.

Chronic lymphocytic leukemia (CLL) is a B-cell lymphoproliferative disease which follows a heterogeneous clinical course. Although the two staging systems, based on clinical parameters, have been effective in stratifying patients into different risk groups, they both fail to identify the early stage patients who will progress rapidly and therefore benefit from an early or more intensive therapy. This is important as most patients present at an early stage of disease.A number of biological markers have been identified to date to aid prediction. These include the IgVH mutational status, CD38 and Zap-70 expression. Although these parameters provide fairly accurate prognostic information, neither marker alone or combined can identify either Binet stage A patients who are going to progress, or the stage B or C patients who may require alternative treatment at the onset. It is therefore important to identify new predictive markers which may provide additional or better prognostic information.The microRNAs are endogenous, non-coding RNAs that play key regulatory roles in a diverse range of pathways, including development, cell proliferation, differentiation and apoptosis. These 18–24 nucleotide single-stranded RNAs are involved in post-transcriptional gene regulation, by binding to complimentary sites in the 3' UTR of messenger RNAs (mRNA), usually resulting in gene silencing. A number of findings early on in the history of microRNAs suggested their potential role in human cancer, and in 2006 Calin et al directly associated de-regulated expression of miR-15 and -16 in the development of CLL. Since this time, a number of microRNAs have been implicated in CLL lymphomagenesis, including miR-92, which is a mature member of the miR-17-92 cluster. The miR-17-92 is located on chromosome 13q31 and shown to act in an oncogenic capacity (oncomiR-1). With respect to CLL, although the expression levels of mature microRNAs of the miR-17-92 cluster have been assessed, the role of such expression in predicting disease outcome has never been examined.We used qRT-PCR to assess the expression levels of mature microRNAs of the miR-17-92 cluster, relative to normal CD19+ B-cells. We report that, despite being transcribed from the same parental cluster, the expression levels of all mature microRNAs vary, with miR-17-5p and -18 showing significantly higher levels than the other members of the cluster (p = <0.001 – 0.001 and p = <0.001 – 0.0256, respectively). This variation is not the result of known microRNA polymorphisms. In all Binet stage patients, high expression of miR-18 (p = 0.021), but low miR-17-5p (p = 0.039) significantly predicted for shorter treatment free survival (TFS), while overexpression of miR-19a (p = 0.084) and -19b (p = 0.088) but a low level of miR-92 (p = 0.188) showed a trend for prediction. Within the Binet stage A group, miRs -19a (p = 0.036), -19b (p = 0.09), -17-5p (p = 0.01) and -92 (p = 0.076) retained prognostic significance. Combined expression of either high, or low risk miRs significantly improved the prediction based on each factor alone in both all Binet stage patients (p = 0.004 and p = 0.025, respectively) and also in the Binet stage A group (p = 0.012 and p = 0.013, respectively). The combined expression of predictive microRNAs with traditional prognostic makers (IgVH, CD38 and Zap-70) improved prediction based on either factor alone. Expression of the high risk factor miR-18 identified the patients with low CD38 who were more likely to progress (p<0.001) and similarly, the low risk factor miR-17-5p predicted those patients with low Zap-70 or mutated IgVH who were more likely to require earlier treatment (p = 0.021 and p = 0.01, respectively). Multivariate analysis showed that expression of miR-18 (H.R 9.871) and Zap-70 (H.R 4.443) or miR-19a (H.R 6.999) and Zap-70 (H.R 22.523) retained independent prognostic significance within all Binet stage and stage A patients, respectively.In summary, high levels of miRs -18, -19a and -19b but low expression of miRs -17-5p and -92 predict for shorter TFS in CLL. This is the first study to implicate members of the miR-17-92 cluster as risk factors in CLL. Given the recent identification of the involvement of miR-17-92 oncomiR-1 in the activation of anti-apoptotic and proliferative pathways, the role of this cluster in CLL warrants further investigation.No relevant conflicts of interest to declare.

Abstract 2727

Abstract 1363

Clinically useful prognostic markers are well described in chronic lymphocytic leukaemia (CLL) and include cell surface expression of CD38 and levels of the intracellular kinase ZAP70. The variable hinge region of the immunoglobulin heavy chain gene (VH) mutation status is well described and highly prognostic but at present remains primarily a research tool. The presence or absence of activation of intracellular signalling cascades by phosphorylation in response to cross linking of the B cell receptor (BCR) in vitrohas also been shown to carry prognostic merit in CLL in a number of studies. These studies have used western blotting based approaches making screening of large sample numbers time consuming.

Clinically useful prognostic markers are well described in chronic lymphocytic leukaemia (CLL) and include cell surface expression of CD38 and levels of the intracellular kinase ZAP70. The variable hinge region of the immunoglobulin heavy chain gene (VH) mutation status is well described and highly prognostic but at present remains primarily a research tool. The presence or absence of activation of intracellular signalling cascades by phosphorylation in response to cross linking of the B cell receptor (BCR) in vitro has also been shown to carry prognostic merit in CLL in a number of studies. These studies have used western blotting based approaches making screening of large sample numbers time consuming. We investigated the correlation between intracellular kinase signalling and CD38, ZAP70 and VH status using a novel ELISA assay for phospho Extracellular signal Regulated Kinase (pERK). CLL cells from 101 patients were either stimulated by BCR cross-linking with goat anti-human IgM Fc5μ fragment or treated with Goat IgG F(ab’) fragment as an isotype control or treated with Phorbol-12-Myristate-13-Acetate (PMA) as a positive control. All treatments were given for a 15 minute incubation prior to cell lysis and protein extraction. We validated our sample pool with analysis of cumulative survival versus VH mutation status. Using a 98% deviation from germline as the unmutated : mutated cut off this confirmed a significant survival advantage in the mutated cases p=0.011 (Mantel-Cox log rank, unmutated mean survival 3209 days vs. mutated mean survival 7569 days). For the analysis of signalling status samples demonstrating a greater than 2 fold increase in absolute pERK levels over isotype were designated as signallers. Of 101 samples analysed 33 (33%) were signallers. Mean survival for signallers was 2826 days whereas for non-signallers it was 7059 days. However this difference was not significant p=0.404 (Mantel-Cox log rank). A significant positive correlation between signaller status and VH concordance to germline was demonstrated (p=0.014, Mann Whitney test; Pearson correlation coefficient 0.226, p=0.023). CD38 positivity and pERK signalling status also showed a significant relationship (p=0.003, Chi Squared test; Pearson correlation coefficient 0.364, p=0.002). The relationship between pERK and ZAP70 did not achieve significance (p=0.054, Chi Squared test; Pearson correlation coefficient 0.211, p=0.039). Taken as a whole these data provide further strong evidence that retention of the ability to signal across the BCR is important in CLL prognosis. We have demonstrated that ELISA is an highly efficient and fully quantitative tool for such analysis. ELISA also facilitates the study of larger sample numbers in comparison to more traditional techniques. We are continuing to use ELISA on different targets as an effective way of progressing the study of intracellular signalling. This study of intracellular signalling remains a valid and important area of translational research in CLL.

High frequency low penetrance risk alleles for chronic lymphocytic leukaemia (CLL) have been identified at 6p25.3 (rs872071, IRF4/MUM1), 11q24.1 (rs735665, GRAMD1B), 15q23 (rs7176508), 2q37.1 (rs13397985, SP140), 2q13 (rs17483466, ACOXL) and 19q13.32 (rs11083846, PRKD2). Given a role in determining risk of disease it is plausible that these allelic variants also play a role in disease progression and overall survival. In order to test this hypothesis, polymorphic status was determined in a case series of 403 patients diagnosed with CLL recruited via 5 clinical centres in the United Kingdom. Mean follow-up time was 4447 days and 169 patients were deceased at the time of last follow-up. Kaplan-Meier survival analysis and the log-rank test were used to investigate the prognostic significance of constitutional genetic markers. Polymorphic variation at rs735665, rs7176508, rs13397985, rs17483466 and rs11083846 was not significantly association with time from diagnosis to first treatment (p>0.05, log-rank test). However, a genetic variant in the 3′UTR of the interferon regulatory factor 4 (IRF4)/multiple myeloma oncogene 1 (MUM1) gene was significantly associated with poor prognosis. Specifically, carriers of the disease-associated variant at rs872071 had a significantly shorter mean time from diagnosis to first treatment (3983 days), compared to non-carriers (7340 days)(p=0.001), although there was no significant difference in overall survival (p=0.29). The association with time to first treatment remained significant in multivariate Cox regression analysis that included age, gender, immunoglobulin heavy chain variable region mutational status and stage of disease at diagnosis in the model (p=0.021). IRF4 is expressed in germinal centre (GC) and post-GC B-cells, and is important for B-cell maturation and homeostasis. Taken together with the aetiological evidence, these data suggest that common allelic variation in IRF4 not only affects risk of CLL, but also predicts poor prognosis.

High frequency low penetrance risk alleles for chronic lymphocytic leukaemia (CLL) have been identified at 6p25.3 (rs872071, IRF4/MUM1), 11q24.1 (rs735665, GRAMD1B), 15q23 (rs7176508), 2q37.1 (rs13397985, SP140), 2q13 (rs17483466, ACOXL) and 19q13.32 (rs11083846, PRKD2). Given a role in determining risk of disease it is plausible that these allelic variants also play a role in disease progression and overall survival. In order to test this hypothesis, polymorphic status was determined in a case series of 403 patients diagnosed with CLL recruited via 5 clinical centres in the United Kingdom. Mean follow-up time was 4447 days and 169 patients were deceased at the time of last follow-up. Kaplan-Meier survival analysis and the log-rank test were used to investigate the prognostic significance of constitutional genetic markers. Polymorphic variation at rs735665, rs7176508, rs13397985, rs17483466 and rs11083846 was not significantly association with time from diagnosis to first treatment (p>0.05, log-rank test). However, a genetic variant in the 3′UTR of the interferon regulatory factor 4(IRF4)/multiple myeloma oncogene 1(MUM1) gene was significantly associated with poor prognosis. Specifically, carriers of the disease-associated variant at rs872071had a significantly shorter mean time from diagnosis to first treatment (3983 days), compared to non-carriers (7340 days)(p=0.001), although there was no significant difference in overall survival (p=0.29). The association with time to first treatment remained significant in multivariate Cox regression analysis that included age, gender, immunoglobulin heavy chain variable region mutational status and stage of disease at diagnosis in the model (p=0.021). IRF4 is expressed in germinal centre (GC) and post-GC B-cells, and is important for B-cell maturation and homeostasis. Taken together with the aetiological evidence, these data suggest that common allelic variation in IRF4not only affects risk of CLL, but also predicts poor prognosis.