Early death in acute promyelocytic leukemia (APL) in French centers: a multicenter study in 399 patients.

Leukemia (2014) 28, 2395–2424 © 2014 Macmillan Publishers Limited All rights reserved 0887-6924/14 www.nature.com/leu

LETTERS TO THE EDITOR

Monoclonal B-cell lymphocytosis is characterized by mutations in CLL putative driver genes and clonal heterogeneity many years before disease progression Leukemia (2014) 28, 2395–2398; doi:10.1038/leu.2014.226 Monoclonal B-cell lymphocytosis (MBL) is defined as an asymptomatic expansion of clonal B cells with less than 5 × 109/L cells in the peripheral blood and without other manifestations of chronic lymphocytic leukemia (CLL; for example, lymphadenopathy, cytopenias, constitutional symptoms).1 Approximately 1% of the MBL cohort develops CLL per year. Evidence suggests that nearly all CLL cases are preceded by an MBL state.2 Our understanding of the genetic basis, clonal architecture and evolution in CLL pathogenesis has undergone significant improvements in the last few years.3–8 In contrast to CLL, our knowledge of the genetics in MBL is still very fragmented, with few prior studies focused on the status of selected genetic abnormalities.1,9,10 In addition, previous reports have demonstrated the co-existence of two B-cell subclones in a small subset of MBL cases,11 and clonal changes in sequential MBL samples measured by IGHV mutational analysis.12 To better understand the genomic landscape of MBL and clonal evolution overtime, we performed a longitudinal analysis in a cohort of eight MBL cases, analyzed by whole-exome sequencing (WES) at two time points. The criteria for sample selection were that the individual with MBL had a sample collected at diagnosis of MBL and had one additional sample collected more than 30 months apart that had stored cells for bead selection of clonal B cells and normal non-B cells for DNA extraction. Samples for WES were collected on average 65 months apart (median 62.5 months, range 30–91). At the second time point analyzed, four cases were still MBL, whereas the remaining four cases had progressed to detectable lymphadenopathy by physical exam but had not yet required treatment. After the second sample analyzed, cases were followed up for an average of 26 months (median 24 months, range 3–72). Thus, the total median follow-up of the cohort from MBL diagnosis to the final visit was 88 months (average 91, range 75–120). Overall, one case remained as an MBL and the remaining seven cases had now progressed to detectable lymphadenopathy, and three of which required treatment by IWCLL criteria. The follow-up information and time points analyzed are summarized in Figure 1a. The MBL individuals provided written informed consent for the collection and use of samples for research purposes according to the Declaration of Helsinki. Clinical information of cases analyzed is provided in Supplementary Table S1. B lymphocytes were enriched from peripheral blood mononuclear cells using the EasySep Human CD19+ Cell Enrichment Kit without CD43 Depletion. T cells were enriched using the EasySep Human CD3 Positive Selection Kit (Stemcell Technologies, Vancouver, BC, Canada) and subsequently were used as germline samples in the sequencing genetic studies. After cell enrichment, all fractions were stained by four-color immunophenotypic analysis to assess sample purity. The following antibody conjugates were used: anti-CD5-fluorescein isothiocyanate, anti-CD16-phycoerythrin, anti-CD19-allophycocyanin and anti-CD3peridinin chlorophyll protein (Beckton Dickinson, Franklin Lakes, NJ, USA). Flow cytometry analysis was performed using the FACSCalibur

(Beckton Dickinson) and data analyzed using Cell Quest software. On the basis of FACS (fluorescence-activated cell sorting) analysis, we observed after enrichment an average of 91% of CD19+ cells (range 76–99%) and 91% of the CD19+ fraction were CD19+/CD5+ cells (range 66–99%). We used the values of the CD19+/CD5+ fraction to calculate the leukemic B-cell fraction and reduce any significant contamination of non-clonal B cells in each biopsy. DNA was extracted from the clonal B cells and non-clonal (that is, T cells) cells using the Gentra Puregene Cell Kit (Qiagen, Hilden, Germany). Extracted DNAs were fingerprinted to confirm the relationship between samples of the same MBL individual and to rule out sample cross-contamination between individuals. In addition, the molecular analysis of IGHV gene family was performed to confirm the existence of one or more B-cell clones (Supplementary Table S2). Mononuclear cells were used for this analysis. One microgram of RNA was converted to cDNA using the BioRad iScript cDNA Kit (Hercules, CA, USA). cDNA (2 μl) were amplified in separate PCR reactions for each of the seven IGHV family genes using a consensus sense primer and an IgM antisense primer. The reactions were carried out using the HotStarTaq PCR kit (Qiagen) in a total volume of 50 μl with 20 pmol of each primer. Cycling conditions were 94 °C for 15 min followed by 35 cycles of 30 s at 94 °C; 1 min at 60 °C; 1 min at 72 °C; and a final cycle of 72 °C for 10 min. PCR products were electrophoresed and visualized with ethidium bromide, purified using the Wizard PCR Prep kit (Promega, Madison, WI, USA) and sequenced using an automated sequencer (Applied Biosystems, Foster City, CA, USA). In those cases where gDNA was used (MBL04 and MBL09), the conditions were the same except that a consensus antisense primer to the IGHJ region was used instead of the IgM antisense primer. Resulting sequences were aligned to germline sequences using ImMunoGeneTics Information (IMGT) System reference sets and IMGT/V-Quest software (http://imgt.cines.fr). Genomic DNA from each sample was sheared and used for the construction of paired end sequencing library as described in the protocol provided by Illumina. The exome was captured using the Sure Select 50 Mb Exome Enrichment kit (Agilent, Santa Clara, CA, USA) following the manufacturer’s instructions. Samples were sequenced using Illumina HiSeq2000. 100 bp paired end reads were aligned to human genome hg19 using Novoalign (Novocraft Technologies, Selangor, Malaysia). Realignment and recalibration was done to take advantage of Best Practice Variant Detection v3 recommendations implemented in the GATK. Somatic singlenucleotide variations (SNVs) were genotyped using SomaticSniper, whereas insertions and deletions were called by GATK Somatic Indel Detector. Each variant in coding regions was analyzed with PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/) and MutationTaster (http://www.mutationtaster.org/) to predict biological effects. Pair-wise analyses were performed comparing each tumor time point with the germline sample. Variants with read depth less than 10X were excluded from further analysis. To validate the findings obtained from WES and increase the sensitivity of the screening, a panel of 24 genes relevant to CLL (Supplementary Table S3) was sequenced in all MBL samples at both time points. Targeted deep sequencing (TDS) was performed in a Ion Torrent

Accepted article preview online 18 July 2014; advance online publication, 12 August 2014

Letters to the Editor

2396

Figure 1. (a) Time points analyzed in each MBL case. Changes in Rai stage (Rai40) and start of treatment are indicated for each case. (b) Changes in allelic fractions of the mutations found in driver genes between time points analyzed. Y-axis shows the tumor fraction with each mutation. MBL8 and MBL10 changed Rai stage between time points (indicated with an arrowhead).

PGM (Life Technologies, Carlsbad, CA, USA) sequencer with an average coverage depth of 630 × (range 506 × –1180 × ). Finally, SNV, insertions and deletions (indels) were visually inspected using Integrative Genomics Viewer (IGV, Broad Institute). Sequencing findings were then integrated with FISH panel data previously performed in these cases (Table 1). Overall, an average of 122-fold depth coverage was obtained by WES (range 92 × –178 × ) with an average of 80% of the targeted regions being covered by at least 30 × . We found an average of 6.6 and 7.5 nonsynonymous SNVs and indels at the first and second time points, respectively. The complete list of mutations and the respective allelic fractions is shown in Supplementary Table S4. In four cases, we found mutations in driver genes Leukemia (2014) 2395 – 2424

previously identified in CLL, including ATM, DDX3X, EGR2, FBXW7, SAMHD1 and SF3B1. TDS not only validated the WES results, but also identified additional mutations in BIRC3, POT1 and NOTCH1 in regions that were initially missed due to low coverage obtained from WES in those specific regions. We found a fifth case, MBL01, with a damaging mutation affecting PRDM1, a gene frequently inactivated in diffuse large B-cell lymphomas.13 Of note, FBXW7 was the only gene recurrently mutated in our cohort. All mutations were found in highly evolutionarily conserved regions and considered damaging by PolyPhen and MutationTaster. In all four cases, we found the mutations in two or more driver genes (Table 1). Thus, MBL03 has mutations in ATM and POT1; MBL06 has mutations in BIRC3 and FBWX7, MBL08 has mutations © 2014 Macmillan Publishers Limited


2397 Table 1. Case no.

Summary of FISH and sequencing results in the MBL cases analyzed FISH

% Positive cells Time point 1

Time point 2

MBL01 MBL02 MBL03

Normal 13q − 13q −

— 41 31

— 94 95

MBL04 MBL06

13q − Trisomy 12

56 51

NA NA

MBL08

Trisomy 12

16

NA

MBL09 MBL10

Trisomy 12 11q −

46 11

NA NA

Gene

Mutation

Polyphen prediction

PRDM1 D344N Moderate — POT1 Q567X High ATM R925G Moderate — BIRC3 Q547Nfs21* High FBXW7 R505L Moderate NOTCH1 P2514Rfs4* High FBXW7 R465L Moderate EGR2 H384N Moderate SAMHD1 Q105X High SAMHD1 I136S Moderate — SF3B1 K666Q Moderate DDX3X D164G Moderate DDX3X I415V Moderate

Mutation taster prediction

Time point 1

Time point 2

P-value

Coverage depth

Allelic fractiona

Coverage depth

Allelic fractiona

Disease causing

342

0.467

410

0.303

0.0001

Disease causing Disease causing

377 641

0.18 0.140

359 837

0.04 0.020

0.0002 0.0001

causing causing causing causing causing causing causing

1457 154 244 368 263 736 702

1.000 0.60 0.941 0.306 0.038 0.003 0.260

2756 418 220 418 301 730 679

1.000 0.59 1.000 1.000 0.420 0.080 0.620

NS NS 0.007 0.0001 0.0001 0.0001 0.0001

Disease causing Disease causing Disease causing

624 530 927

0.000 0.09 0.51

528 382 933

0.220 0.23 0.30

0.0001 0.0009 0.0001

Disease Disease Disease Disease Disease Disease Disease

Abbreviations: FISH, fluorescence in situ hybridization; MBL, monoclonal B-cell lymphocytosis; NA, not applicable; NS, not significant. Only putative tumorigenic genes were included. All mutations are somatic. All mutations excepting PRDM1 (D344N) were validated using targeted sequencing. P-values indicate whether the allelic fraction differences between time points are significant. aThe values correspond to allelic fractions after tumor content correction.

in EGR2, FBXW7, NOTCH1 and two independent mutations in SAMHD1, whereas MBL10 has a mutation in SF3B1 and two independent mutations in DDX3X. We analyzed sequential time points to recreate the clonal architecture and evolution. The number of mutations found per case was insufficient to run comprehensive clustering analysis. However, we focused the analysis on the dynamics of mutations found in putative driver genes. In three of the four cases harboring mutations in putative driver genes (MBL03, MBL06 and MBL08), all the mutations were already detected at the first time point analyzed, but mostly found in subclones that either remained stable (MBL03 and MBL06) or become more prevalent overtime (MBL08; Figure 1b). In the remaining case (MBL10), the first time point analyzed was characterized by the presence of DDX3X (p.I415V) in nearly 50% allelic fraction and a second DDX3X mutation (p.D164G) in 10% of allelic fraction. In the second time point, an increased abundance of p.D164G and reduced abundance of p.I415V were observed. This opposite trend in prevalence suggests the presence of driver mutations in different subclones with alternated dominance between time points. Furthermore, at the second time point, a novel mutation in SF3B1 was found (p.K666Q). This mutation was not identified in the first time point even using deep sequencing. All four cases with putative CLL mutations eventually developed lymphadenopathy and/or required treatment on average 56 months after clinical recognition of MBL compared with an average of 61 months in the remaining cases. In all four samples, the existence of a single B-cell clone was confirmed by molecular analysis of IGHV gene (Supplementary Table S2). Previous studies have identified genetic abnormalities in a subset of MBL, including deletion 13q14 and NOTCH1 mutations.2,10 Furthermore, ATM mutations have been found in linkage studies of familial CLL.14 Our study brings new insights regarding the genomic landscape and clonal architecture of MBL, the precursor to CLL,1,15 by identifying mutations in one or more driver genes and confirming the existence of genetic subclones in a subset of MBL cases. Furthermore, our study brings new data providing insight into the temporal succession of genetic events in MBL and CLL pathogenesis. Thus, a recent study has shown the subclonal nature of FBXW7 in all four CLL cases harboring mutations in the gene, suggesting this mutation is a later event in CLL pathogenesis.5 However, our data demonstrate the © 2014 Macmillan Publishers Limited

existence of recurrent mutations of FBXW7 in the premalignant stages of the disease, and suggest that FBXW7 mutations can be an early genetic event in CLL pathogenesis. Remarkably, we identified tumorigenic mutations on average 60 months before clear disease progression occurred in MBL cases. These findings support the implementation of a targeted sequencing strategy including the subset of putative cancer genes now known to be recurrently mutated in CLL (~20 genes). Using available sequencing technologies, this subset of genes could be easily screened in very high depth coverage, thus providing early identification of oncogenic mutations in small subclones at the MBL phase. CONFLICT OF INTEREST RF has received a patent for the prognostication of MM based on genetic categorization of the disease. He has received consulting fees from Medtronic, Otsuka, Celgene, Genzyme, BMS, Lilly, Onyx, Binding Site, Millenium and AMGEN. He also has sponsored research from Cylene and Onyx. TS was supported by fundings from Hospira, Genentech, Glaxo-Smith-Kline, Jannsen, Celgene and Cephalon. The remaining authors declare no conflcit of interest.

ACKNOWLEDGEMENTS EB is a recipient of the Marriott Specialized Workforce Development Awards in Individualized Medicine and the Henry Predolin Foundation Career Development Award. This work was supported by the Henry Predolin Foundation and grant NIH-CA95241.

J Ojha1, C Secreto2, K Rabe2, J Ayres-Silva3, R Tschumper2, DV Dyke2, S Slager2, R Fonseca1, T Shanafelt2, N Kay2 and E Braggio1 1 Mayo Clinic, Scottsdale, AZ, USA; 2 Mayo Clinic, Rochester, MN, USA and 3 National Institute of Cancer, Rio de Janeiro, Brazil E-mail: [email protected] REFERENCES 1 Rawstron AC, Bennett FL, O'Connor SJ, Kwok M, Fenton JA, Plummer M et al. Monoclonal B-cell lymphocytosis and chronic lymphocytic leukemia. N Engl J Med 2008; 359: 575–583.

Leukemia (2014) 2395 – 2424


2398 2 Landgren O, Albitar M, Ma W, Abbasi F, Hayes RB, Ghia P et al. B-cell clones as early markers for chronic lymphocytic leukemia. N Engl J Med 2009; 360: 659–667. 3 Braggio E, Kay NE, VanWier S, Tschumper RC, Smoley S, Eckel-Passow JE et al. Longitudinal genome-wide analysis of patients with chronic lymphocytic leukemia reveals complex evolution of clonal architecture at disease progression and at the time of relapse. Leukemia 2012; 26: 1698–1701. 4 Knight SJ, Yau C, Clifford R, Timbs AT, Sadighi Akha E, Dreau HM et al. Quantification of subclonal distributions of recurrent genomic aberrations in paired pre-treatment and relapse samples from patients with B-cell chronic lymphocytic leukemia. Leukemia 2012; 26: 1564–1575. 5 Landau DA, Carter SL, Stojanov P, McKenna A, Stevenson K, Lawrence MS et al. Evolution and impact of subclonal mutations in chronic lymphocytic leukemia. Cell 2013; 152: 714–726. 6 Quesada V, Conde L, Villamor N, Ordonez GR, Jares P, Bassaganyas L et al. Exome sequencing identifies recurrent mutations of the splicing factor SF3B1 gene in chronic lymphocytic leukemia. Nat Genet 2012; 44: 47–52. 7 Schuh A, Becq J, Humphray S, Alexa A, Burns A, Clifford R et al. Monitoring chronic lymphocytic leukemia progression by whole genome sequencing reveals heterogeneous clonal evolution patterns. Blood 2012; 120: 4191–4196. 8 Wang L, Lawrence MS, Wan Y, Stojanov P, Sougnez C, Stevenson K et al. SF3B1 and other novel cancer genes in chronic lymphocytic leukemia. N Engl J Med 2011; 365: 2497–2506.

9 Greco M, Capello D, Bruscaggin A, Spina V, Rasi S, Monti S et al. Analysis of SF3B1 mutations in monoclonal B-cell lymphocytosis. Hemato Oncol 2013; 31: 54–55. 10 Rasi S, Monti S, Spina V, Foa R, Gaidano G, Rossi D. Analysis of NOTCH1 mutations in monoclonal B-cell lymphocytosis. Haematologica 2012; 97: 153–154. 11 Ghia P, Prato G, Scielzo C, Stella S, Geuna M, Guida G et al. Monoclonal CD5+ and CD5- B-lymphocyte expansions are frequent in the peripheral blood of the elderly. Blood 2004; 103: 2337–2342. 12 Marti GE, Shim YK, Albitar M, Middleton D, Abbasi F, Anderson A et al. Long-term follow-up of monoclonal B-cell lymphocytosis detected in environmental health studies. Cytometry B Clin Cytom 2010; 78: S83–S90. 13 Pasqualucci L, Compagno M, Houldsworth J, Monti S, Grunn A, Nandula SV et al. Inactivation of the PRDM1/BLIMP1 gene in diffuse large B cell lymphoma. J Exp Med 2006; 203: 311–317. 14 Bevan S, Catovsky D, Marossy A, Matutes E, Popat S, Antonovic P et al. Linkage analysis for ATM in familial B cell chronic lymphocytic leukaemia. Leukemia 1999; 13: 1497–1500. 15 Shanafelt TD, Ghia P, Lanasa MC, Landgren O, Rawstron AC. Monoclonal B-cell lymphocytosis (MBL): biology, natural history and clinical management. Leukemia 2010; 24: 512–520.

Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu)

CD34+CD38−CD58− cells are leukemia-propagating cells in Philadelphia chromosome-positive acute lymphoblastic leukemia Leukemia (2014) 28, 2398–2401; doi:10.1038/leu.2014.228 Prognosis of Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph+ALL) has improved with the use of tyrosine kinase inhibitors but most persons relapse. Some persons with Ph+ALL develop resistance to tyrosine kinase inhibitors but others relapse because of the persistence of quiescent leukemia stem cells (also termed leukemia-propagating cells (LPCs)). LPCs are defined by their ability to initiate human leukemia and to proliferate and self-renew in immune-deficient mice.1–4 LPCs in persons with acute myeloid leukemia have diverse phenotypes but most are CD34+CD38−.1,4 Recently, persons with acute myeloid leukemia and high LPCs frequencies in the bone marrow and persons whose bone marrow cells have a gene expression profile typical of LPCs are reported to have worse clinical outcomes following therapy with anti-leukemia drugs.5,6 In persons with Ph+ALL, CD34+CD38− cells were identified as LPCs in the non-obese diabetic/severe combined immunodeficient (NOD/SCID) xenograft assay,7 but the clinical relevance of this finding, if any, is unknown. CD58 is reported to be overexpressed in leukemia blasts and might be used as a marker of minimal residual disease in persons with B-cell ALL.8 Higher proportion of CD58+ cells is reported to correlate with better outcomes in B-ALL.9 In adults with Ph+ALL, there are no data on the potential prognostic importance of differences in CD58 expression patterns in CD34+ CD38− cells. We hypothesized candidate LPCs may be further enriched in the CD34+CD38−CD58− bone marrow fraction possibly translating to unfavorable prognosis. Sixty-three consecutive newly diagnosed adults with Ph+ALL were prospectively studied at Peking University Institute of Hematology from 1 January 2010 to 31 December 2012 (Supplementary Figure 1). Inclusion criteria included: (1) age

18–60 years; (2) diagnosis of ALL based on the 2008 World Health Organization criteria; (3) detection of the Ph-chromosome and/or BCR-ABL mRNA; (4) no contraindication to therapy with imatinib or an allotransplant. The study was approved by the Ethics Committee of Peking University People’s Hospital and written informed consent was obtained from all subjects before study entry in accordance with the Declaration of Helsinki. Induction chemotherapy included 1 cycle of a CODP regimen (cyclophosphamide, 750 mg/mE+2, day 1; vincristine 1.4 mg/mE +2, days 1, 8, 15, 22; daunorubicin, 40 mg/mE+2, days 1–3; prednisone, 1 mg/kg/day, days 1–21). Subjects achieving complete remission (CR) received 8 cycles of consolidation therapy including hyper-CVAD B (cycles 1, 3, 5 and 7; methotrexate, 1 g/mE+2, d 1; cytarabine, 1 g/mE+2, q12h, days 2–3) alternating with the hyperCVAD A (cycles 2, 4, 6 and 8; cyclophosphamide, 300 mg/mE+2, q12h, days 1–3; doxorubicin, 60 mg/mE+2, day 4; vincristine, 1.4 mg/mE+2, days 4, and 11; dexamethasone, 40 mg/day, days 1–4 and 11–14). Subjects also received imatinib (400 mg/day) during induction and consolidation therapy. After two cycles of consolidation, subjects with a suitable donor including an human leukocyte antigen-matched sibling, an human leukocyte antigen-matched unrelated donor or a human leukocyte antigenhaplotype identical-related donor were advised to receive an allotransplant.10–12 Subjects in the chemotherapy cohort received 6 more cycles of consolidation chemotherapy including the Hyper-CVAD B program alternating with the Hyper-CVAD A program followed by maintenance therapy (6-mercaptopurine, 60 mg/mE+2 daily; methotrexate, 20 mg/mE+2 weekly; monthly vincristine (4 mg/day, day 1)/prednisone (1 mg/kg/day, days 1–7) pulse) for 2 years. Multi-parameter flow cytometry analyses of CD58-FITC (Beckman-Coulter, Brea, CA, USA)/CD10-PE/CD19-APC-Cy7/ CD34-PerCP/ CD45-Vioblue/ CD38-APC (BD Biosciences, San Jose,


Leukemia (2014) 2395 – 2424

© 2014 Macmillan Publishers Limited


2399 Table 1.

+

+

−

−

Characteristics of Ph ALL subjects with CD34 CD38 CD58 or other phenotype at diagnosis CD34+CD38− CD58−

Other phenotype

P-value

13 36 (18–60) 6/7

50 38.5 (18–60) 30/20

0.5 0.5

Blood cell count at diagnosis Median WBC (×10E+9/l) (range)

32.5 (2.9–147)

30.1 (2.3–359)

0.5

Median BM blasts at diagnosis (%) (range) Median CD34+CD38−CD58− cells by BM blasts at diagnosis (%) (range) BCR-ABL (P190/P210) With/without additional cytogenetic abnormalities CR rate after 1 course of induction (%) Median time to achieve CR1 (days) (range)

85.3 (67–98) 82.3 (56–88.2) 8/5 2/11 61.5 56 (21–90)

79.0 (20.6–95.3) 6.5 (1.2–18.8) 30/20 16/34 90 31.5 (15–85)

0.09 o0.0001 1.00 0.32 0.03 0.04

5 8 (3/5)

12 38 (15/23)

BCR-ABL/ABL (%) during therapy After 1st cycle After 2nd cycle After 3rd cycle

19.3 (0–91.3) 16.3 (0–150) 11.9 (0–100.4)

5.7 (0–8.1) 5.18 (0–22.4) 1.2 (0–15)

0.05 0.02 0.02

Survival status BCR/ABL+at CR BCR/ABL+after the 3rd cycle CCR Relapse NRM

13 12 5 7 1

34 26 36 9 5

0.03 0.01 0.05 0.01 1.00

Characteristics Number of subjects Median age (range) Gender (male/female)

Post-consolidation therapy Chemo+IM Allo-HSCT (MSD/HID)+IM

0.3

Abbreviations: Allo-HSCT, allogeneic hematopoietic stem cell transplantation; BM, bone marrow; CR, complete remission; chemo, chemotherapy; CCR, continuous complete remission; HID, human leukocyte antigen-mismatched/haploidentical donors; IM, imatinib; MSD, human leukocyte antigen-matched sibling donors; NRM, non-relapse mortality; Ph+ALL, Philadelphia chromosome-positive acute lymphoblastic leukemia; WBC, white blood cell.

CA, USA) on gated leukemia blasts was performed using a multicolor MACSQuant Analyzer (Miltenyi Biotec, Bergisch Gladbach, Germany). Fluorescence-minus-one controls were used to determine positive events for CD34, CD38 and CD58. There was considerable heterogeneity in expression of CD38 and CD58 (Supplementary Figure 2). Samples with ⩾ 20% blasts expressing the relevant CD antigen were considered positive. CD34+ blasts with ⩾ 20% CD38 expression were defined as a CD34+CD38+ phenotype, whereas CD34+ blasts with o 20% CD38 expression were classified as the CD34+CD38− phenotype. CD58 expression was calculated as a percent in the CD34+CD38+ population or CD34+CD38− population. CD34+CD38+ blasts with ⩾ 20% CD58 expression were defined as the CD34+CD38+CD58+ phenotype, whereas CD34+CD38+ blasts with o 20% CD58 expression were classified as the CD34+CD38+CD58− phenotype. Similarly, CD34+ CD38− blasts with ⩾ 20% CD58 expression were determined to be the CD34+CD38−CD58+ phenotype, whereas CD34+CD38− blasts with o20% CD58 expression were classified as the CD34+ CD38−CD58− phenotype. Based on blast phenotypes at diagnosis, subjects were further divided into the CD34+CD38−CD58− cohort (N = 13) and other phenotype cohort (N = 50, including subjects with CD34+CD38−CD58+, CD34+CD38+CD58− or CD34+CD38+ CD58+ phenotypes and subjects with the above defined four fractions concurrently). The clinical characteristics of the two phenotype groups did not differ significantly (Table 1). Median follow-up was 24 months (range, 6–43 months) for all subjects and 30 months (range, 8–43 months) for survivors. The CD34+CD38−CD58− cohort had a lower proportion of CR after the first course of chemotherapy (62% vs 90%; P = 0.03). Median time to achieve CR in the CD34+ © 2014 Macmillan Publishers Limited

CD38−CD58− cohort was significantly longer compared with the other phenotype cohort (median, 56 days vs 32 days; P = 0.04). Significantly, higher levels of BCR/ABL mRNA were detected in subjects in remission in the CD34+CD38−CD58− cohort than persons in remission in the other phenotype cohort especially after the third cycle of therapy. Cumulative incidence of relapse at 3 year in the CD34+CD38−CD58− cohort was significantly higher compared with cumulative incidence of relapse in the other phenotype cohort (60% (54–65%) vs 19% (18–19%); P = 0.02). Three-year leukemia-free survival of subjects in the other phenotype cohort was significantly higher than in subjects in the CD34+CD38−CD58− cohort (69% (53–81%) vs 33% (9–60%); P = 0.04). The CD34+CD38−CD58− cohort also had worse 3-year survival than the other phenotype cohort (32% (6–62%) vs 71% (55–82%)), but this difference was not significant (P = 0.07) (Supplementary Figure 3). In multivariate analyses, the CD34+CD38−CD58− phenotype was an independent risk factor correlated with likelihood of achieving CR (P = 0.03, odds ratio (OR) = 0.4 (0.2–0.9)), relapse (P = 0.03, OR = 3.4 (1.1–10.5)), leukemia-free survival (P = 0.01, OR = 3.1 (1.3–7.4)) and survival (P = 0.03, OR = 2.8 (1.1–6.9)) (Supplementary Table 1). Because of differences in clinical outcomes between subjects with and without CD34+CD38−CD58− phenotype, we studied the ability of cells from Ph+ALL subjects with a CD34+CD38−CD58−, CD34+CD38−CD58+, CD34+CD38+CD58− and CD34+CD38+CD58+ phenotypes to initiate leukemia in a murine xenograft assay. The six subjects were classified into the other phenotype group because the CD34+CD38−CD58− fraction was detected in only a few blasts (Supplementary Table 2). Bone marrow mononuclear cells were stained with mouse anti-human CD58-FITC Leukemia (2014) 2395 – 2424


2400

Figure 1. Comparison of human Ph+ALL engraftment in the anti-CD122-conditioned NOD/SCID recipients transplanted with CD34+ CD38−CD58− vs other phenotype (CD34+CD38−CD58+, CD34+CD38+CD58− or CD34+CD38+CD58+) of Ph+ALL cells. (A) May–Giemsa staining and fluorescence in situ hybridization (FISH) analyses of leukemic blasts in the original Ph+ALL subjects. (B) The recipient transplanted with CD34+CD38−CD58− Ph+ALL cells at 12-week posttransplantation (+) exhibited splenomegaly and suppression of erythropoiesis in the femur. (C) Recipients transplanted with the other phenotype of Ph+ALL cells (black) exhibited higher overall survival (OS) compared with the primary recipients (blue) and the secondary recipients (red) transplanted with CD34+CD38−CD58− Ph+ALL cells, as estimated by the Kaplan–Meier method (Po0.0001 for comparison of recipients within a given graft dose and for all recipients combined). No significant difference was noted between the primary and secondary recipients transplanted with the CD34+CD38−CD58− Ph+ALL cells. (D) Low-magnification images of human Ph+ALL engraftment in bone sections (upper panels) of the CD34+CD38−CD58− cells transplanted recipients; engraftment was further confirmed by morphologic and cytogenetic analyses as well as hematoxylin and eosin (HE) and immunohistochemical staining with antihuman CD19 and CD34 antibodies at high magnification (lower panels). In contrast, no human engraftment was demonstrated in the recipients transplanted with other phenotype cells. (E) CD34+CD38−CD58− Ph+ALL cells infiltrated into recipient organs. HE staining and antihuman CD19 and CD34 antibody labeling of the brain, liver, spleen and kidney of a recipient transplanted with CD34+CD38−CD58− cells compared with a recipient transplanted with other phenotype cells.

(Beckman-Coulter) and CD34-PE/CD19-APC-Cy7/CD45-PerCP/ CD38-APC/CD3,CD4,CD8-PE-Cy7 monoclonal antibodies (BD Biosciences) and sorted using the FACS Aria II (Becton Dickinson, San Jose, CA, USA). In the viable CD3−CD4−CD8− bone marrow mononuclear cells, CD34+CD38−CD58−, CD34+CD38−CD58+, CD34+CD38+CD58− and CD34+CD38+CD58+ fractions were sorted (Supplementary Figure 4). Purity of each fraction was 497%. The anti-CD122 (interleukin-2 receptor β (IL-2Rβ))-conditioned NOD/ SCID xenograft assay was performed by intra-bone marrow injection.13,14 Doses were 1 × 10E+3, 1 × 10E+4 and 1 × 10E+5/ mouse. We found different engraftment kinetics in the blood of primary and secondary recipients when 1 × 10E+3, 1 × 10E+4 or Leukemia (2014) 2395 – 2424

1 × 10E+5 CD34+CD38−CD58− cells were transplanted. The efficiently engrafted human leukemia cells in all recipients transplanted with CD34+CD38−CD58− cells were phenotypically and clonally derived from the donor subjects analyzed by multiparameter flow cytometry and BCR/ABL mRNA. Human leukemia cells were also detected infiltrating into liver, kidney and brain of primary and secondary murine recipients transplanted with CD34+ CD38−CD58− cells by hematoxylin and eosin staining and immune histochemistry with rabbit anti-human CD34 and CD19 (Abcam, Cambridge, MA, USA) (Figure 1). In contrast, CD34+CD38−CD58+, CD34+CD38+CD58− and CD34+CD38+CD58+ cells transplanted at the same or higher doses of 1 × 10E+6 and 1 × 10E+7 cells failed to © 2014 Macmillan Publishers Limited


engraft. These data suggests Ph+ALL LPCs are derived from the CD34+CD38−CD58− cells. Self-renewal capacity of CD34+CD38−CD58− cells was studied by serial transplants in mice. High levels of human CD45+CD19+ engraftment were observed in all of the secondary recipients of CD45+CD34+CD38−CD58− cells. In contrast, when 1 × 10E+3, 1 × 10E+4, 1 × 10E+5 CD45+CD34+CD38−CD58+ or CD45+CD34+CD38+ fractions from the same CD34+CD38−CD58− primary recipients were transplanted, no human engraftment was detected in secondary recipients. These findings suggest that CD34+CD38−CD58− cells not only initiate human leukemia but also self-renewal. Limiting dilution analyses were performed to estimate LPCs frequencies in the above four cell fractions. We calculated a median frequency of 1 LIC in 128 CD34+CD38−CD58− cells (95% confidence interval, 11–626). No LPCs were found in the CD34+ CD38−CD58+, CD34+CD38+CD58− or CD34+CD38+CD58+ fractions even at higher injection doses. A previous report suggested Ph+ALL LPCs are found in the CD34+ CD38− population.7 Our data indicate these LPCs are found in the CD34+CD38−CD58− population. Subjects with the CD34+ CD38−CD58− phenotype had the worst clinical outcomes. We also found CD34+CD38− cells are heterogeneous. CD34+CD38−CD58− human Ph+ALL cells but not CD34+CD38−CD58+ can initiate Ph+ALL and self-renewal in anti-CD122-conditioned NOD/SCID mice. Archimbaud et al.9 reported a correlation between less CD58 expression on ALL blasts with worse survival. Similarly, we found worse outcomes in subjects with a CD34+CD38−CD58− phenotype. The conventional NOD/SCID mouse assay with intravenous injection is widely used to assay human hematopoietic stem cells and LPCs.1,4 Recent improvements include depletion of natural killer cells with anti-CD122 antibody and direct intra-medullary injection.13,14 Using this improved assay, we found candidate Ph+ALL LPCs in the CD34+CD38−CD58− fraction. Based on these data, we suggest the adverse clinical outcomes associated with the CD34+CD38−CD58− phenotype consistent with biological studies demonstrating that LPCs are quiescent and relatively resistant to chemotherapy.15 In conclusion, our study suggested that Ph+ALL LPCs are enriched in the CD34+CD38−CD58− phenotype which translates to adverse clinical outcomes. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS Professor Robert Peter Gale kindly reviewed the typescript. The work was supported by the National Natural Science Foundation of China (81370638 and 81230013), the National Clinical Priority Specialty, the Beijing Municipal Science and Technology Program (grant no. Z141100000214011), and Peking University People’s Hospital Research and Development Funds (RDB2012-23).

AUTHOR CONTRIBUTIONS X-JH designed the study and supervised the analyses and typescript preparation. YK performed the research, analyzed and interpreted the data, performed statistical analyses and wrote the typescript. All other authors participated in the collection of patient data. All of the authors agreed to submit the final typescript.

Y Kong1, Y-J Chang1, Y-R Liu1, Y-Z Wang1, Q Jiang1, H Jiang1, Y-Z Qin1, Y Hu1,2, Y-Y Lai1, C-W Duan3, D-L Hong3 and X-J Huang1,2 1 Peking University People’s Hospital, Peking University Institute of Hematology, Beijing Key Laboratory of Hematopoietic Stem Cell Transplantation, Beijing, China; 2 Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, China and 3 Department of Pathophysiology, Key Laboratory of Cell Differentiation and Apoptosis of Ministry of Education, Shanghai Jiao Tong University School of Medicine, Shanghai, China E-mail: [email protected] REFERENCES 1 Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med 1997; 3: 730–737. 2 Cobaleda C, Sanchez-Garcia I. B-cell acute lymphoblastic leukaemia: towards understanding its cellular origin. Bioessays 2009; 31: 600–609. 3 Kong Y, Yoshida S, Saito Y, Doi T, Nagatoshi Y, Fukata M et al. CD34+CD38+CD19+ as well as CD34+CD38-CD19+ cells are leukemia-initiating cells with self-renewal capacity in human B-precursor ALL. Leukemia 2008; 22: 1207–1213. 4 Lapidot T, Sirard C, Vormoor J, Murdoch B, Hoang T, Caceres-Cortes J et al. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice. Nature 1994; 367: 645–648. 5 Eppert K, Takenaka K, Lechman ER, Waldron L, Nilsson B, van Galen P et al. Stem cell gene expression programs influence clinical outcome in human leukemia. Nat Med 2011; 17: 1086–1093. 6 Gentles AJ, Plevritis SK, Majeti R, Alizadeh AA. Association of a leukemic stem cell gene expression signature with clinical outcomes in acute myeloid leukemia. JAMA 2010; 304: 2706–2715. 7 Cobaleda C, Gutierrez-Cianca N, Perez-Losada J, Flores T, Garcia-Sanz R, Gonzalez M et al. A primitive hematopoietic cell is the target for the leukemic transformation in human Philadelphia-positive acute lymphoblastic leukemia. Blood 2000; 95: 1007–1013. 8 Chen JS, Coustan-Smith E, Suzuki T, Neale GA, Mihara K, Pui CH et al. Identification of novel markers for monitoring minimal residual disease in acute lymphoblastic leukemia. Blood 2001; 97: 2115–2120. 9 Archimbaud E, Thomas X, Campos L, Magaud JP, Dore JF, Fiere D. Expression of surface adhesion molecules CD54 (ICAM-1) and CD58 (LFA-3) in adult acute leukemia: relationship with initial characteristics and prognosis. Leukemia 1992; 6: 265–271. 10 Huang XJ, Zhu HH, Chang YJ, Xu LP, Liu DH, Zhang XH et al. The superiority of haploidentical related stem cell transplantation over chemotherapy alone as postremission treatment for patients with intermediate- or high-risk acute myeloid leukemia in first complete remission. Blood 2012; 119: 5584–5590. 11 Yan CH, Liu DH, Liu KY, Xu LP, Liu YR, Chen H et al. Risk stratification-directed donor lymphocyte infusion could reduce relapse of standard-risk acute leukemia patients after allogeneic hematopoietic stem cell transplantation. Blood 2012; 119: 3256–3262. 12 Xiao-Jun H, Lan-Ping X, Kai-Yan L, Dai-Hong L, Yu W, Huan C et al. Partially matched related donor transplantation can achieve outcomes comparable with unrelated donor transplantation for patients with hematologic malignancies. Clin Cancer Res 2009; 15: 4777–4783. 13 Mazurier F, Doedens M, Gan OI, Dick JE. Rapid myeloerythroid repopulation after intrafemoral transplantation of NOD-SCID mice reveals a new class of human stem cells. Nat Med 2003; 9: 959–963. 14 Tanaka T, Tsudo M, Karasuyama H, Kitamura F, Kono T, Hatakeyama M et al. A novel monoclonal antibody against murine IL-2 receptor beta-chain. Characterization of receptor expression in normal lymphoid cells and EL-4 cells. J Immunol 1991; 147: 2222–2228. 15 Ishikawa F, Yoshida S, Saito Y, Hijikata A, Kitamura H, Tanaka S et al. Chemotherapy-resistant human AML stem cells home to and engraft within the bone-marrow endosteal region. Nat Biotechnol 2007; 25: 1315–1321.



Leukemia (2014) 2395 – 2424

2401


2402

The prognostic importance of the presence of more than one focal lesion in spine MRI of patients with asymptomatic (smoldering) multiple myeloma Leukemia (2014) 28, 2402–2403; doi:10.1038/leu.2014.230 Patients with asymptomatic multiple myeloma (AMM) are at risk for development of symptomatic disease requiring therapy, at the rate of about 10% per year, in the first 5 years after diagnosis.1 However, while some patients with AMM have a relatively low risk approaching that of monoclonal gammopathy of undetermined significance, there is a subset of patients with AMM at very high risk for imminent development of disease complications (that is, within 1–2 years after diagnosis). Such complications may include bone fractures (commonly vertebral compression fractures) or acute renal injury and renal insufficiency, which increase the morbidity and mortality of MM patients. Currently, it is recommended that patients with AMM should be followed closely without therapy;2 however, a recent clinical trial indicated that for patients at high risk for progression, early intervention may be beneficial.3 Thus, for these ‘high-risk’ patients early intervention is now indicated by many experts,4 in order to avoid catastrophic complications of the disease. Nevertheless, the identification of high-risk patients is challenging and requires the consideration of several different disease characteristics. The Mayo Clinic group recently identified that patients with extensive BM infiltration (⩾60%) or with extremely abnormal free light chains (FLC) ratio (4100) are at very high risk for imminent progression within o2 years from the diagnosis of AMM.5,6 Our group has also confirmed that the above risk features identify a subset of patients with AMM at very high risk for progression, within 12–18 months from diagnosis.7 The Spanish group has proposed more sophisticated criteria, which require high-quality, flow cytometric analysis.8 In view of the need for widely available methods to identify high-risk patients, imaging studies may offer significant information. Moulopoulos et al. has published since the 90s the prognostic importance of spine magnetic resonance imaging (MRI) for patients with symptomatic9 and asymptomatic MM.10 Hillengass et al.11 published data indicating that patients with more than one focal lesion in whole-body MRI are at high risk for progression to symptomatic disease; however, whole-body MRI is not widely available. Our group also identified abnormal spine MRI (with either a focal or a diffuse pattern) as a factor associated with a high risk of progression to symptomatic MM.7 Recently the SWOG S0120 study also identified focal lesions in spine MRI as a predictor of disease progression.12 Thus, it is now increasingly recommended that spine MRI should be considered in the initial workup of patients with AMM with high-risk features and that patients with more than one focal lesion should be considered for early intervention.4 Herein, we present data from our database of patients with AMM with available spine MRI at diagnosis and with mature follow-up, in order to evaluate the prognostic importance of the presence of focal lesions in spine MRI for the identification of patients with AMM at high risk of progression. We analyzed the outcomes of 67 patients with AMM and a minimal follow-up of 2.5 years. An experienced radiologist in the field of MM (LM) reviewed all the MRIs using standard criteria.9,10,13 The median age of the patients was 63 years (range 35–88), 63% had immunoglobulin G and 37% immunoglobulin A MM. The median BM infiltration in trephine biopsy was 20% (range

11–90%) and 9% had ⩾ 60% plasma cells. All patients had more than 10% plasma cells probably due to the availability of BM trephine biopsy and immunohistochemistry in all of them. The median serum M-spike was 1.9 g/dl, with 48% having 42 g/dl and 11% having ⩾ 3 g/dl, a figure that is smaller than the one reported in the Mayo Clinic series.1 Median FLC ratio (involved to uninvolved FLC) was 7.6 (range 1–219) and 11% had FLC ratio ⩾ 100. After a median follow-up of 4 years (range 2.5–11 years), 21 (31%) patients have progressed to symptomatic MM and the 1-, 2- and 3-year rate of progression to symptomatic MM for the cohort is 13%, 21% and 29%, respectively, with an estimated median time to development of symptomatic MM of 70 months. At the time of diagnosis 13 (20%) patients had abnormal findings in their MRIs of the spine: 9 (14%) had more than one focal lesion (FL), while 2 had diffuse pattern without focal lesions; 2 patients had only one focal lesion and were excluded from further analysis. Median time to symptomatic MM for patients with more than one focal lesion was 15 months (95% confidence interval 6–26 months), while it exceeded 5 years for patients with no focal lesions (Po0.001), corresponding to a hazard ratio of 7.2 (95% confidence interval 3–18) (Figure 1). For patients with more than one FL the 1-year progression to symptomatic MM rate was 44%, the 2-year progression rate was 69% and the 3-year progression to MM rate was 85%. The respective 1-, 2- and 3-year progression rates for those with no FLs was 8%, 13% and 22%, respectively. We applied previously defined features associated with a high risk of progression (BM infiltration ⩾ 60%, FLC ratio 4100) along with the presence of more than one FL in spine MRI, and 85% of our patients had none of these risk factors and 15% had at least one. Among patients with more than one focal lesion, 6 (67%) had at least one of the other high-risk features, but in three it was the only high-risk feature. Median time to progression for those with none vs those with at least one high-risk feature was 45 years vs 9 months (Po0.001), 1-year progression rate was 4% and 60%, 2-year progression rate 11% and 71% and 3-year progression rate 21% and 89%, respectively. Our data support the prognostic importance of the MRI of the spine as a tool for the identification of patients at high risk for progression to symptomatic disease. Our results are in accordance with those reported by Hilengass et al.,11 according to which patients with more than one focal lesion had a median time to progression of 13 months while for those without focal lesion or with one focal lesion the time to progression exceeded 5 years.11 In the SWOG S0120 study the presence of more than one focal lesion was associated with increased risk of disease progression in the subset of patients with available MRIs of the spine.12 Our patients had spine MRI and not whole-body MRI that was used in the study by Hillengass et al.,11 which is associated with a wider field of view. Nevertheless, in the study of Hillengass et al.,11 15% of patients with AMM had focal lesions in the axial skeleton, a figure similar to what we found in our study, in which 16% of our patients had one or more focal lesions in the spine MRI. In the SWOG S0120 a similar proportion of patients (16%) had focal lesion in spine MRI. Thus, it is consistent across different cohorts that ~ 15% of patients with AMM present with focal lesions in spine MRI. MRI of the spine may miss extra-axial skeletal focal lesions, which in the study by Hillengass are present in 10% of


Leukemia (2014) 2395 – 2424



2403

2

1st Department of Radiology, School of Medicine, National and Kapodistrian University of Athens, Athens, Greece E-mail: [email protected]

REFERENCES

Figure 1. Time to progression to symptomatic MM for patients with AMM and more than one focal lesion in spine MRI vs those with no focal lesions or with one focal lesion in spine MRI.

patients with AMM and normal spinal MRI (vs 17% of patients with abnormal spinal MRI, with either focal lesion or diffuse pattern). However, MRI of the spine is less time consuming and causes less discomfort to the patient, has lower cost than whole-body MRI and is widely available, while it also identifies those patients with AMM at high risk for progression. In summary, our data support the use of spine MRI as the imaging modality of choice in order to help identify patients with AMM at high risk for progression to symptomatic disease, with an estimated probability of progression of ~ 70% within 2 years from diagnosis. In addition to other widely available prognostic tools (FLC ratio4100, BM infiltration ⩾ 60%) we are now in position to identify patients with AMM who should be offered immediate therapy, just like patients with symptomatic MM. CONFLICT OF INTEREST The authors declare no conflict of interest.

E Kastritis1, LA Moulopoulos2, E Terpos1, V Koutoulidis2 and MA Dimopoulos1 1 Department of Clinical Therapeutics, School of Medicine, National and Kapodistrian University of Athens, Athens, Greece and

1 Kyle RA, Remstein ED, Therneau TM, Dispenzieri A, Kurtin PJ, Hodnefield JM et al. Clinical course and prognosis of smoldering (asymptomatic) multiple myeloma. N Engl J Med 2007; 356: 2582–2590. 2 Kyle RA, Durie BG, Rajkumar SV, Landgren O, Blade J, Merlini G et al. Monoclonal gammopathy of undetermined significance (MGUS) and smoldering (asymptomatic) multiple myeloma: IMWG consensus perspectives risk factors for progression and guidelines for monitoring and management. Leukemia 2010; 24: 1121–1127. 3 Mateos MV, Hernandez MT, Giraldo P, de la Rubia J, de Arriba F, Lopez Corral L et al. Lenalidomide plus dexamethasone for high-risk smoldering multiple myeloma. N Engl J Med 2013; 369: 438–447. 4 Dispenzieri A, Stewart AK, Chanan-Khan A, Rajkumar SV, Kyle RA, Fonseca R et al. Smoldering multiple myeloma requiring treatment: time for a new definition? Blood 2013; 122: 4172–4181. 5 Larsen JT, Kumar SK, Dispenzieri A, Kyle RA, Katzmann JA, Rajkumar SV. Serum free light chain ratio as a biomarker for high-risk smoldering multiple myeloma. Leukemia 2013; 27: 941–6946. 6 Rajkumar SV, Larson D, Kyle RA. Diagnosis of smoldering multiple myeloma. N Engl J Med 2011; 365: 474–475. 7 Kastritis E, Terpos E, Moulopoulos L, Spyropoulou-Vlachou M, Kanellias N, Eleftherakis-Papaiakovou E et al. Extensive bone marrow infiltration and abnormal free light chain ratio identifies patients with asymptomatic myeloma at high risk for progression to symptomatic disease. Leukemia 2013; 27: 947–53. 8 Perez-Persona E, Vidriales MB, Mateo G, Garcia-Sanz R, Mateos MV, de Coca AG et al. New criteria to identify risk of progression in monoclonal gammopathy of uncertain significance and smoldering multiple myeloma based on multiparameter flow cytometry analysis of bone marrow plasma cells. Blood 2007; 110: 2586–2592. 9 Moulopoulos LA, Varma DG, Dimopoulos MA, Leeds NE, Kim EE, Johnston DA et al. Multiple myeloma: spinal MR imaging in patients with untreated newly diagnosed disease. Radiology 1992; 185: 833–840. 10 Moulopoulos LA, Dimopoulos MA, Smith TL, Weber DM, Delasalle KB, Libshitz HI et al. Prognostic significance of magnetic resonance imaging in patients with asymptomatic multiple myeloma. J Clin Oncol 1995; 13: 251–256. 11 Hillengass J, Fechtner K, Weber MA, Bauerle T, Ayyaz S, Heiss C et al. Prognostic significance of focal lesions in whole-body magnetic resonance imaging in patients with asymptomatic multiple myeloma. J Clin Oncol 2010; 28: 1606–1610. 12 Dhodapkar MV, Sexton R, Waheed S, Usmani S, Papanikolaou X, Nair B et al. Clinical, genomic, and imaging predictors of myeloma progression from asymptomatic monoclonal gammopathies (SWOG S0120). Blood 2014; 123: 78–85. 13 Moulopoulos LA, Gika D, Anagnostopoulos A, Delasalle K, Weber D, Alexanian R et al. Prognostic significance of magnetic resonance imaging of bone marrow in previously untreated patients with multiple myeloma. Ann Oncol 2005; 16: 1824–1828.

Primary acute myeloid leukemia cells with IDH1 or IDH2 mutations respond to a DOT1L inhibitor in vitro Leukemia (2014) 28, 2403–2406; doi:10.1038/leu.2014.235 Epigenetic dysregulation represents an emerging paradigm in the pathogenesis of myeloid malignancies, and the pharmacologic targeting of pathways involved in regulating epigenetic modifications is a promising therapeutic strategy.1 Potential targets include recurrently mutated genes encoding epigenetic modifiers (for example, DNMT3A, IDH1/2, TET2, EZH2), and altered epigenetic modifiers (like mixed lineage leukemia (MLL) fusion genes) that are known to initiate acute myeloid leukemia (AML).2 Some

leukemias with MLL translocations are dependent on the activity of DOT1L, a histone methyltransferase responsible for trimethylation of histone 3 on lysine 79 (H3K79); small-molecule inhibitors of DOT1L, like EPZ004777, can disrupt leukemic progression.3–5 DOT1L inhibitors are in various stages of clinical development for MLL-rearranged leukemias, but it is currently unclear whether these inhibitors may also be active in AML cases that lack MLL rearrangements. We recently developed an in vitro culture system for primary AML cells that captures the genetic complexity of primary AML samples, and that can be used to test the impact of novel

Accepted article preview online 5 August 2014; advance online publication, 29 August 2014


Leukemia (2014) 2395 – 2424


2404

pharmacologic agents on AML cells.6 In vitro testing of primary leukemic samples has been very challenging for DOT1L inhibitors, which requires exposure to cells for more than 1 week. However, with our system, primary human AML cells can be expanded for up to 2 weeks. We used this approach to test the effects of DOT1L inhibition against a set of cryopreserved, genomically characterized de novo AML patient samples, collected through a study approved by the Human Research Protection Office at Washington University School of Medicine after patients provided informed consent. We first verified the sensitivity of MLL-rearranged primary AML cells to the DOT1L inhibitor EPZ004777. Cryopreserved peripheral blood or bone marrow specimens collected at diagnosis were thawed and expanded for 2–4 days in vitro to allow for cells to recover, and were then incubated with increasing concentrations (0.1–10 μM) of EPZ004777 (or DMSO vehicle control) for a period of 10 days. We initially selected six adult AML samples (including four with MLL translocations and two without) to monitor the influence of EPZ004777 on cell growth. Vehicle-treated cells from each

patient expanded at variable rates; however, all samples achieved a minimum of twofold expansion by day 10. Dose-dependent growth inhibition was observed for two of the four MLL rearranged samples (one with MLL-AF6 and one with MLL-AF10) in the presence of EPZ004777. AML cell growth was minimally influenced at the maximum drug concentration in the non-MLL control samples (Figure 1a). Previous studies performed on cells expressing MLL fusions demonstrated that EPZ004777 reduced cell growth with delayed kinetics;4 indeed, reduced growth was not evident in the responsive samples for at least 6–7 days. Global levels of H3K79 methylation were reduced following exposure to EPZ004777, and expression of HOXA cluster genes was also decreased, as expected (Supplementary Figure S1). Each of the two MLL-rearranged samples that failed to exhibit a strong response to EPZ004777 had an MLL–ELL fusion (Figure 1b and Supplementary Table 1). The lack of response to EPZ004777 in these cases may relate to the reported absence of DOT1L in the MLL–ELL elongation complex.7,8

Figure 1. EPZ004777 alters growth and differentiation of primary AML cells with MLL rearrangements. (a) Impact of EPZ004777 treatment on the growth of primary AML cells with (top panels) and without MLL rearrangements (bottom panels). Absolute cell numbers on the y-axis; note the different scales due to different rates of growth. Cells were expanded using a previously described stromal co-culture technique6 (also see methods in Supplementary material) in the presence of DMSO or increasing concentrations of EPZ004777 (0.1, 1, 10 μM) over a 10-day period. Data represent mean values from two experiments assayed in duplicate ± s.d. (error bars). (b) Representative growth curve of MLL–ELL primary cells treated with increasing concentrations of EPZ004777 (left panel). Summary graph of cell growth results from two MLL–ELL samples treated with different doses of EPZ004777 (right panel); results are plotted as percent relative to DMSO control. Data represent mean values from two experiments assayed in duplicate ± s.d. (left) or s.e.m. (right). (c) Histograms depicting the cell surface expression of myeloid markers CD11b (left panels) and CD14 (right panels) in two representative patient samples (one with a MLL translocation and one without). Cells were incubated in the presence of 10 μM EPZ004777 for 10 days and analyzed by flow cytometry. (d) Wright–Giemsa-stained cytospins of representative MLL and non-MLL rearranged patient samples treated for 10 days with DMSO or 10 μM EPZ004777. Scale bars represent 20 μm. Leukemia (2014) 2395 – 2424



We also examined the effects of EPZ004777 on myeloid differentiation. Cell surface expression of CD11b and/or CD14 increases as myeloid cells differentiate toward neutrophils or mature monocytes, respectively. Expression of these markers increased in a dose-dependent manner after 10 days of drug treatment in the sensitive samples (Figure 1c, top). In contrast, the MLL-ELL and non-MLL control samples displayed only modest surface expression changes in the presence of drug (Figure 1c, bottom). Morphological examination of the cells corroborated the flow cytometric findings (Figure 1d). Collectively, these data suggested that this in vitro culture system would be suitable for testing the sensitivity of an expanded panel of primary AML samples to EPZ004777. We therefore evaluated AML cells with mutations in other epigenetic modifiers (for example, IDH1, IDH2, DNMT3A) or the MLL partial tandem duplication (MLL-PTD). All four cases with MLL-PTD alterations were sensitive to EPZ004777, consistent with

a recent report.9 Additionally, we also observed a potent response to EPZ004777 in an AML sample with a PICALM–MLLT10 fusion, which was predicted from a murine model.3 Surprisingly, many AML samples without MLL fusions or MLL-PTDs were also highly sensitive to treatment with EPZ004777. While EPZ004777 did not slow the growth of AML samples with isolated DNMT3A mutations (for example, UPN 721214 and 868442), 7/7 samples with canonical mutations in IDH1 (R132H/C) or IDH2 (R140Q) were sensitive. All cases demonstrated a dose-dependent decrease in viable cell numbers (compared to the vehicle control) after 10 days of drug treatment (Figures 2a and b). Of note, two of the seven samples also had MLL-PTD mutations (Supplementary Table 1). For the IDH-mutated samples, treatment with EPZ004777 for 10 days induced greater surface expression of CD11b and/or CD14 than vehicle alone (Figure 2c). Morphologic differentiation was also apparent (Figure 2d) in the sensitive samples. In total, 14/23 AML samples were sensitive to EPZ004777. All of the responders had an

Figure 2. EPZ004777 impairs the growth and differentiation of genetically defined, non-MLL rearranged primary AML cells. (a) Growth curves of primary cells with an IDH1 mutation (UPN 807970), a MLL-PTD mutation (UPN 817156), both mutations (UPN 914247) or neither mutation (UPN 868442) incubated in the presence of DMSO or increasing concentrations of EPZ004777 (0.1, 1, 10 μM) over a 10-day period. Refer to Supplementary Table S1 for more detailed information on individual primary samples. Absolute cell numbers on the y-axis; note the different scales due to different rates of growth. Cells were expanded using the stromal co-culture technique (see methods). Data represent mean values from two experiments assayed in duplicate ± s.d. (error bars). (b) Impact of EPZ004777 on the proliferation of IDH wild-type (WT; UPN 868442 and 721214) and IDH mutant (all others) primary cells; refer to Supplementary Table S1 for more details on individual samples. Results represent number of viable cells counted after 10 days, and are expressed as a percentage relative to DMSO controls, to adjust for differences in growth rates between different primary samples. Data represent mean values from two experiments assayed in duplicate ± s.e.m. (error bars). (c) Cell surface expression of CD11b (left panels) and CD14 (right panels) in two representative IDH mutant patient samples. Cells were incubated in the presence of 10 μM EPZ004777 for 10 days and analyzed by flow cytometry. (d) Wright–Giemsa-stained cytospins of representative IDH wild-type and mutant patient samples treated for 10 days with DMSO or 10 μM EPZ004777. Scale bars represent 20 μm. © 2014 Macmillan Publishers Limited

Leukemia (2014) 2395 – 2424

2405


2406 MLL rearrangement (that was not MLL-ELL), a PICALM–MLLT10 fusion, MLL-PTDs, and/or mutations in IDH1 or IDH2. In sum, our data suggest that a wider spectrum of AMLs may be responsive to DOT1L inhibitors than originally predicted. While the observed responses of AML samples with MLL-PTDs or the PICALMMLLT10 fusion were anticipated, the responses of IDH-mutant samples were not. The neomorphic enzymatic activity of mutated IDH1/2 produces 2-hydroxyglutarate (2-HG), an oncometabolite capable of inhibiting enzymes that are involved in a variety of epigenetic processes, including histone and DNA methylation.10 While IDH-mutated AMLs typically do not show dysregulation of HOXA cluster gene expression (which is common in MLL rearrangements and MLL-PTDs), global levels of H3K79 dimethylation (H3K79me2) have been reported to be modestly increased when either IDH1 or IDH2 are mutated;10–12 the precise locations of altered sites of H3K79 methylation are not yet known. Regardless, the observations presented here provide an impetus to expand studies of DOT1L inhibitors to AML samples with canonical IDH1 or IDH2 mutations, and to define the mechanisms by which it acts in these cases. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS Technical assistance was provided by the Alvin J. Siteman Cancer Center Tissue Procurement Core, the High Speed Cell Sorter Core, and the Molecular and Genomic Analysis Core at Washington University School of Medicine and Barnes-Jewish Hospital in St Louis, MO, which are all supported by the National Cancer Institute Cancer Center Support Grant P30CA91842. This work was supported by National Institutes of Health grant R01CA162086 (TJL), PO1CA101937 (TJL), K08HL116605 (JMK) and the Barnes Jewish Hospital Foundation (TJL), as well as the Doris Duke Charitable Foundation to Washington University (SMS, Clinical Research Fellow).

1

SM Sarkaria1, MJ Christopher1, JM Klco2 and TJ Ley1 Section of Stem Cell Biology, Division of Oncology, Department of Internal Medicine, Washington University in St Louis School of Medicine, St Louis, MO, USA and

2

Department of Pathology and Immunology, Washington University in St Louis School of Medicine, St Louis, MO, USA E-mail: [email protected]

REFERENCES 1 Shih AH, Abdel-Wahab O, Patel JP, Levine RL. The role of mutations in epigenetic regulators in myeloid malignancies. Nat Rev Cancer 2012; 12: 599–612. 2 Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med 2013; 368: 2059–2074. 3 Chen L, Deshpande AJ, Banka D, Bernt KM, Dias S, Buske C et al. Abrogation of MLL-AF10 and CALM-AF10-mediated transformation through genetic inactivation or pharmacological inhibition of the H3K79 methyltransferase Dot1l. Leukemia 2013; 27: 813–822. 4 Daigle SR, Olhava EJ, Therkelsen CA, Basavapathruni A, Jin L, Boriack-Sjodin PA et al. Potent inhibition of DOT1L as treatment of MLL-fusion leukemia. Blood 2013; 122: 1017–1025. 5 Bernt KM, Zhu N, Sinha AU, Vempati S, Faber J, Krivtsov AV et al. MLL-rearranged leukemia is dependent on aberrant H3K79 methylation by DOT1L. Cancer Cell 2011; 20: 66–78. 6 Klco JM, Spencer DH, Lamprecht TL, Sarkaria SM, Wylie T, Magrini V et al. Genomic impact of transient low-dose decitabine treatment on primary AML cells. Blood 2013; 121: 1633–1643. 7 Biswas D, Milne TA, Basrur V, Kim J, Elenitoba-Johnson KSJ, Allis CD et al. Function of leukemogenic mixed lineage leukemia 1 (MLL) fusion proteins through distinct partner protein complexes. Proc Natl Acad Sci USA 2011; 108: 15751–15756. 8 Lin C, Smith ER, Takahashi H, Lai KC, Martin-Brown S, Florens L et al. AFF4, a component of the ELL/P-TEFb elongation complex and a shared subunit of MLL chimeras, can link transcription elongation to leukemia. Mol Cell 2010; 37: 429–437. 9 Kuhn MWM, Hadler M, Daigle SR, Chen C, Sinha AU, Krivtsov AV et al. Myeloid leukemia cells With MLL partial tandem duplication are sensitive to pharmacological inhibition of the H3K79 methyltransferase DOT1L. Blood 2013; 122: 1256. 10 Lu C, Ward PS, Kapoor GS, Rohle D, Turcan S, Abdel-Wahab O et al. IDH mutation impairs histone demethylation and results in a block to cell differentiation. Nature 2012; 483: 474–478. 11 Wang P, Dong Q, Zhang C, Kuan P-F, Liu Y, Jeck WR et al. Mutations in isocitrate dehydrogenase 1 and 2 occur frequently in intrahepatic cholangiocarcinomas and share hypermethylation targets with glioblastomas. Oncogene 2013; 32: 3091–3100. 12 Xu W, Yang H, Liu Y, Yang Y, Wang P, Kim S-H et al. Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases. Cancer Cell 2011; 19: 17–30.


Lymphodepletion followed by infusion of suicide gene-transduced donor lymphocytes to safely enhance their antitumor effect: a phase I/II study Leukemia (2014) 28, 2406–2410; doi:10.1038/leu.2014.237 Donor lymphocyte infusions (DLIs) can produce lasting remissions in patients with relapsed chronic myeloid leukemia (CML) after allogeneic stem cell transplantation (SCT), but are less effective in non-CML diseases.1,2 Chemotherapy-induced lymphodepletion with cyclophosphamide (Cy) and fludarabine (Flu) before DLI has been shown to not only enhance activation of donor lymphocytes but also cause significantly more severe graft-versus-host disease (GVHD) than DLI alone,3 thus limiting its application. To safely balance the toxic versus beneficial effects of activated donor lymphocytes, we combined prior lymphodepletion with the

infusion of donor T cells engineered to carry a suicide gene for treating patients with aggressive hematologic malignancies. Donor T cells were transduced with gibbon ape leukemia viruspseudotyped retroviral particles harboring a bicistronic Thy1– IRES–TK vector coding for the herpes simplex virus type 1 thymidine kinase (TK) suicide gene and the human CD90 (Thy1) GPI-anchored membrane molecule as a selection marker. Transduced cells were then expanded during 15–19 days and enriched through magnetic selection of CD90+ cells before their infusion, as previously described.4 This process of transduction/expansion/ selection was associated with a significant increase in the proportion of CD3+ and CD8+ cells, specifically of memory/ activated phenotype (Supplementary Figure 1). In this phase I/II

Accepted article preview online 8 August 2014; advance online publication, 2 September 2014

Leukemia (2014) 2395 – 2424



2407

Figure 1. Immunomonitoring in PB after Cy–Flu/DLI–TK. (a) Absolute counts of CD3+, CD4+, CD8+, CD19+ and CD3+CD56+ cells. Follow-up is given in days following Cy–Flu/DLI–TK in all graphs. (b) Relative counts of activated HLA-DR+ (left panels) or CD69+ (right panels) among CD4+ (upper panels) and CD8+ (lower panels) cells. (c) Relative counts of naive (CD45RA+CD62+, left panels), central-memory (CD45RA−CD62L+, middle panels) and effector-memory (CD45RA−CD62L−, right panels) cells among CD4+ (upper panels) and CD8+ (lower panels) cells. (d) Relative and (e) absolute counts of CD90+ cells among CD3+ cells. (f) Relative counts of TK+ cells among peripheral blood mononuclear cells as determined by quantitative PCR. Histograms represent the mean ± s.e.m. of patient values (n = 9 at day 15, n = 8 at days 30 and 90, n = 7 at day 180, n = 5 at day 270 and n = 3 at day 360). TK transgene was undetectable in patient 3 all along his follow-up. For some patients at some time points, values were not available because of profound lymphopenia or when the follow-up was stopped. Mean values for cell counts were compared between different time points using the two-tailed paired Student's t-test. The degree of significance is indicated as follows: *Po0.05; **Po0.01; ***Po0.001. © 2014 Macmillan Publishers Limited

Leukemia (2014) 2395 – 2424


2408

n=10 Cy-Flu/DLI-TK GVHD n=2

GVHD grade O-I n=8 Patient 8

Patient 2

GCV Flow cytometry + % CD90 cells/CD3+

PBMC

GCV

GCV

Liver

PCR % TK copies/cells

Leukemia residual disease (NPM1)

After salvage chemotherapy

0

50

At SCT

100 150 days after HSCT

200

250

90 80 70 60 50 40 30 20 10 0

Cy-Flu/DLI-TK

pre-DLI day 15 M1

M3

M6

M9

months after Cy-Flu/DLI-TK

Cy-Flu/DLI-TK

1.0

1.0

0.8

0.8

Overall Survival

Cumulative incidence of progression

Free light chain (mg/L)

Relapse post-HSCT

100 10 1 0.1 0.01 0.001 0.0001 0.00001

0.6 0.4 0.2

0.6 0.4 0.2 0.0

0.0 0

12 24 Months after DLI-TK

0

12 24 Months after DLI-TK

Figure 2. Graft-versus-host and GVT effects after Cy–Flu/DLI–TK. (a) Two patients (no. 2 and 8) developed grade ⩾ II acute-like GVHD and thus received GCV according to protocol guidelines. Patient 2 had a grade III liver GVHD and thus received intravenous GCV (5 mg/kg b.i.d.) from day 43 following Cy–Flu/DLI–TK for 2 weeks. As shown by flow cytometry analysis and PCR quantification of TK transgene copies, this led to the rapid disappearance of TK+ cells in PB and liver. Patient 8 had a grade III cutaneous and digestive GVHD that was treated with intravenous GCV from day 7, leading to rapid clearance of CD90+ cells in PB as attested by flow cytometry analysis. PCR quantification of TK transgene copies was not available at early time points before GCV introduction for this patient owing to very profound lymphopenia. (b) The Cy–Flu/DLI–TK regimen mediates a GVT effect in patient 2 with acute myeloid leukemia harboring a nucleophosmine 1 (NPM1) mutation. The leukemia minimal residual disease follow-up is quantified since SCT as a ratio (%) of mutated NPM1 mRNA copy number on reference beta-glucuronidase mRNA sequence. (c) Antitumor effect following Cy–Flu/DLI–TKin patient 1 with free light chain myeloma, as attested by the free kappa light chain dosage upon time. (d) Cumulative incidence of progression and (e) Kaplan–Meier overall survival in 10 patients following Cy–Flu/DLI–TK.

trial, the safety and efficacy of the combination treatment (Cy–Flu/ DLI–TK) were studied between February and December 2011 in 10 adults with relapsed myeloma (n = 5) or myelodysplasia/acute leukemia (n = 5) in three French centers (Supplementary Table 1). Leukemia (2014) 2395 – 2424

Mean age at inclusion was 56 years (range: 44–68). Patients had received SCT from a sibling (n = 8) or a matched unrelated (n = 2) donor since a mean interval of 24 months (range: 6–56) and were free of immunosuppressive therapy at inclusion. © 2014 Macmillan Publishers Limited


2409 Patient 10 (Supplementary Table 1) previously failed to respond to one standard DLI, whereas others had not received any DLI before inclusion. After receiving Cy (50 mg/kg at day − 6) and Flu (25 mg/m2/day from day − 6 to day − 2) intravenously, patients were infused at day 1 with a mean cell dose of 5.2 (range: 1.7–10.1)x106 CD3+ cells/kg, of which 98% were CD90+ cells (range: 97–99%). DLI products contained a mean ratio of 0.6% (range: 0.1–1.6) CD4+ FoxP3+Helios+ regulatory T cells. This trial was registered at www. clinicaltrials.gov as NCT01086735. The Cy–Flu regimen was rapidly lymphodepleting so that CD4+, CD8+, CD19+ and CD56+ cells were quasi-undetectable in peripheral blood (PB) at the time of DLI infusion (Figure 1a). Thereafter, the kinetics of immune reconstitution differed among subpopulations. CD8+ and CD56+ cells had reached pre-depletion values within 2 weeks from lymphodepletion, and had doubled at 8 weeks. It took 8 weeks for B cells to reach pre-depletion values and this was even much slower for CD4+ cells, which reached their baseline level after 1 year (Figure 1a). T-cell activation and naive/ memory markers were monitored during the 3 months following Cy–Flu/DLI–TK. During that period, we observed a significant increase in the proportion of HLA-DR-expressing CD4+ and CD8+ T cells (Figure 1b), which was associated with an increase in the proportion of cells with a memory phenotype and a decrease in those with a naive phenotype (Figure 1c). By flow cytometry, transduced cells could be detected in PB immediately after DLI–TK. From baseline values for natural CD90+ cells,5 the percentage of CD90+ cells increased during the first week to reach around 5% of circulating T cells and decreased thereafter (Figure 1d). The absolute number of CD90+ T cells peaked at day 30 after infusion (Figure 1e). In patient 3, no increase in the percentage of CD90+ cells was observed at any time point after DLI–TK. Using PCR amplification of the TK gene, all patients had detectable transduced cells in PB during their whole follow-up except patient 3 (Figure 1f). No acute infusion-related toxicity was observed. Three patients developed acute-like GVHD following TK+ cell infusion (Supplementary Table 1). At day 14 after DLI–TK, patient 9 had a grade I cutaneous GVHD that did not need ganciclovir (GCV) treatment according to protocol guidelines and resolved with local steroids. Patient 8 had a grade III cutaneous and digestive GVHD at day 7 after DLI–TK that was treated with intravenous administration of GCV (5 mg/kg b.i.d.). This led to complete and lasting resolution of symptoms, correlated with clearance of TK+ cells in PB (Figure 2a). Patient 2 had a grade III liver GVHD histologically documented at day 43 after DLI–TK. Treatment with GCV led to the rapid disappearance of TK+ cells in PB and liver (Figure 2a) but without clinical improvement after 2 weeks, so that an additional immunosuppressive therapy with steroids and, secondarily, anti-CD25 monoclonal antibody had to be instituted. Liver tests improved partially with persisting transaminase elevation. To adjust treatment, a second liver biopsy was thus performed to evaluate still active GVHD, but was immediately complicated by an intra-abdominal hemorrhage that was rapidly fatal. This complication was the sole fatal adverse event during follow-up (Supplementary Table 2). In three patients with detectable disease at the time of treatment (myeloma in patients 1 and 5 or acute leukemia in patient 2), we observed a graft-versus-tumor (GVT) effect attributable to the Cy–Flu/DLI–TK (Figures 2b–d). This effect was durable in patient 5 and patient 2 remained disease-free until her death from liver biopsy complication (Figure 2b). Two additional patients (no. 3 and 4) who had undetectable disease at inclusion remained disease-free on the long term. Seven out of 10 patients progressed in 10 ± 3 months following Cy–Flu/DLI–TK (Figure 2d), disease progression being at least transiently sensitive to additional therapy in all but one. With a median follow-up of © 2014 Macmillan Publishers Limited

22 months after Cy–Flu/DLI–TK, six patients are alive (Figure 2e), four of them being disease-free. The present trial represents the first investigation of suicide gene-transduced DLI following a lymphodepletive regimen for malignancy relapse after allogeneic SCT. As in previous phase I/II clinical trials where donor TK+ cells have been tested,6–10 TK-transduced T cells proved to be safe, with no documented adverse event related to the gene modification. After their infusion, transduced cells expanded in PB in all but one patient. In this patient, as we never detected any transduced cells even at early time points, it is unlikely that this could have been caused by an immunization against the exogenous TK protein, as described by others.10 In other patients, the total number of CD90+ cells in the compartment of PB was around 107 at day 1, thus representing ~ 2% of the total number of injected cells. This repartition is close to the one of PB lymphocytes, which is known to represent about 2% of the total number of body lymphocytes. At day 30, the peak expansion time, CD90+ cells had increased over 15-fold. Interestingly, (i) the recovering T cells had an activated phenotype and (ii) the recovery and then expansion above pre-lymphodepletion values of the CD8+ and NK cells was much faster than that of CD4+ cells. This should create an effectortuned milieu favorable for a GVT effect. GCV was used in two patients to control GVHD. In these two patients, GCV led to the rapid elimination of the TK+ cells. This was sufficient to control GVHD in one of the patient. In the other, there was only a partial control of disease after 2 weeks of treatment. As GCV treatment led to the disappearance of TK+ cells from PB and liver, persisting GVHD was thus related to non-transduced effector cells rather than insensitive TK+ cells. TK− cells could have arisen from either non-transduced cells from the DLI, although they represented only 1.4% of infused cells or from T cells surviving to the lymphodepletive treatment. In both cases, lymphodepletion may have favored their activation and proliferation upon alloreactive stimulation. In this patient, a related GVT effect was documented but the long-term evolution of GVHD could not be evaluated because of sudden death following a liver biopsy. Miller et al.,3 who tested a similar Cy–Flu/DLI approach using the same Cy–Flu doses and schedule, but unmanipulated cells in quite comparable patients, observed a 60% incidence of grade II to IV acute GVHD, higher than the 20% observed in our trial. This discrepancy might be related to the cell dose used as in their study patients received a 20 times higher T-cell dose than in ours (108 vs a mean of 5.2x106 CD3+ cells/kg). This and the safety profile of the treatment should now prompt to infuse higher doses of transduced cells. Modifications of the transduced cell manufacturing process through the use of additional stimulating cytokines10,11 and/or the concomitant depletion of regulatory T cells from the DLI12 represent promising strategies to improve clinical results. The precise proper effect of lymphodepletion before cell infusion will also need to be validated in controlled trials for future developments. Combination of antitumor agents with DLI13 or the use of donor T cells encoding a chimeric antigen receptor (CAR) targeting a tumor antigen14 might be also mandatory to further increase the GVT effect of DLI. In the recent development of CAR T cells, the use of lymphodepletion before cell infusion may help in improving the antitumor effect,15 whereas the combined use of suicide genes will permit securing the approach in prevision of unexpected adverse events following CAR T-cell infusion.11 CONFLICT OF INTEREST SM, JLC, FML and DK are inventors of a patent application claiming the association of lymphodepletion with suicide gene transduction and Treg depletion of DLI. JLC, FML and DK are shareholders of LTKfarma, a company developing suicide gene-modified T cells for cancer and autoimmune diseases. The remaining authors declare no conflict of interest.

Leukemia (2014) 2395 – 2424


2410

ACKNOWLEDGEMENTS We thank P Jouany, S Katsahian and V Millul for study management, C Pautas and Y Hicheri for clinical care and E Atti and T Berthy for technical assistance (all in Assistance Publique-Hôpitaux de Paris, France). The study was sponsored and monitored by the Regional Clinical Research Office, Paris and supported by a grant from the Programme Hospitalier de Recherche Clinique (PHRC ID, P 010506) and recurrent funding from the Centers for Clinical Investigation in Biotherapy from the Henri Mondor and Pitié-Salpêtrière hospitals. The sponsors had no role in study design, data analysis, data interpretation or writing of the report.

S Maury1,2,3, M Rosenzwajg4,5,6,7,8, R Redjoul1, A Marcais9, A Xhaard10, M Cherai4,5, L Cabanne1, G Churlaud4, F Suarez9, G Socié10, L Gregoire11, K Debbache1, C Bernard4,5,6, J-L Beaumont12, N Azar4, O Boyer13, F Roudot-Thoraval2,11, JL Cohen2,3, C Cordonnier1,2, FM Lemoine4,5,13 and D Klatzmann4,5,6,7,8 1 AP-HP, Hôpital Henri Mondor, Department of Hematology, DHU Virus-Immunity-Cancer, Créteil, France; 2 IMRB, University Paris Est Créteil (UPEC), INSERM U955 team 21, Créteil, France; 3 Center for Clinical Investigation in Biotherapy, Créteil, France; 4 AP-HP, Hôpital Pitié-Salpêtrière, Biotherapy Department, Paris, France; 5 AP-HP, Hôpital Pitié-Salpêtrière, Clinical Investigation Center in Biotherapy (CIC-BTi), Paris, France; 6 AP-HP, Hôpital Pitié-Salpêtrière, Inflammation-ImmunopathologyBiotherapy Department (i2B), Paris, France; 7 Sorbonne Université, UPMC Univ Paris 06, UMRS 959, Immunology-Immunopathology-Immunotherapy (I3), Paris, France; 8 INSERM, UMR_S 959, Immunology-ImmunopathologyImmunotherapy (I3), Paris, France; 9 Department of Hematology, Hôpital Necker, Paris, France; 10 Department of Hematology–Transplantation, Hôpital St Louis, Paris, France; 11 Department of Public Health, Créteil, France; 12 Department of Hemapheresis, Etablissement Français du Sang, Créteil, France and 13 Sorbonne Universités, UPMC Univ Paris 06, UMR-S INSERM 1135, CIMI-Paris, Paris, France E-mail: [email protected] or [email protected] REFERENCES 1 Schmid C, Labopin M, Nagler A, Bornhauser M, Finke J, Fassas A et al. Donor lymphocyte infusion in the treatment of first hematological relapse after allogeneic stem-cell transplantation in adults with acute myeloid leukemia: a retrospective risk factors analysis and comparison with other strategies by the EBMT Acute Leukemia Working Party. J Clin Oncol 2007; 25: 4938–4945.

2 Porter DL, Alyea EP, Antin JH, DeLima M, Estey E, Falkenburg JH et al. NCI First International Workshop on the Biology, Prevention, and Treatment of Relapse after Allogeneic Hematopoietic Stem Cell Transplantation: Report from the Committee on Treatment of Relapse after Allogeneic Hematopoietic Stem Cell Transplantation. Biol Blood Marrow Transplant 2010; 16: 1467–1503. 3 Miller JS, Weisdorf DJ, Burns LJ, Slungaard A, Wagner JE, Verneris MR et al. Lymphodepletion followed by donor lymphocyte infusion (DLI) causes significantly more acute graft-versus-host disease than DLI alone. Blood 2007; 110: 2761–2763. 4 Lemoine FM, Mesel-Lemoine M, Cherai M, Gallot G, Vie H, Leclercq V et al. Efficient transduction and selection of human T-lymphocytes with bicistronic Thy1/HSV1-TK retroviral vector produced by a human packaging cell line. J Gene Med 2004; 6: 374–386. 5 Blair A, Hogge DE, Ailles LE, Lansdorp PM, Sutherland HJ. Lack of expression of Thy-1 (CD90) on acute myeloid leukemia cells with long-term proliferative ability in vitro and in vivo. Blood 1997; 89: 3104–3112. 6 Tiberghien P, Ferrand C, Lioure B, Milpied N, Angonin R, Deconinck E et al. Administration of herpes simplex-thymidine kinase-expressing donor T cells with a T-cell-depleted allogeneic marrow graft. Blood 2001; 97: 63–72. 7 Burt RK, Drobyski WR, Seregina T, Traynor A, Oyama Y, Keever-Taylor C et al. Herpes simplex thymidine kinase gene-transduced donor lymphocyte infusions. Exp Hematol 2003; 31: 903–910. 8 Ciceri F, Bonini C, Stanghellini MT, Bondanza A, Traversari C, Salomoni M et al. Infusion of suicide-gene-engineered donor lymphocytes after family haploidentical haemopoietic stem-cell transplantation for leukaemia (the TK007 trial): a non-randomised phase I-II study. Lancet Oncol 2009; 10: 489–500. 9 Borchers S, Provasi E, Silvani A, Radrizzani M, Benati C, Dammann E et al. Genetically modified donor leukocyte transfusion and graft-versus-leukemia effect after allogeneic stem cell transplantation. Hum Gene Ther 2011; 22: 829–841. 10 Cieri N, Mastaglio S, Oliveira G, Casucci M, Bondanza A, Bonini C. Adoptive immunotherapy with genetically modified lymphocytes in allogeneic stem cell transplantation. Immunol Rev 2014; 257: 165–180. 11 Hoyos V, Savoldo B, Quintarelli C, Mahendravada A, Zhang M, Vera J et al. Engineering CD19-specific T lymphocytes with interleukin-15 and a suicide gene to enhance their anti-lymphoma/leukemia effects and safety. Leukemia 2010; 24: 1160–1170. 12 Maury S, Lemoine FM, Hicheri Y, Rosenzwajg M, Badoual C, Cherai M et al. CD4 +CD25+ regulatory T cell depletion improves the graft-versus-tumor effect of donor lymphocytes after allogeneic hematopoietic stem cell transplantation. Sci Transl Med 2010; 2: 41ra52. 13 Shimoni A, Kroger N, Zander AR, Rowe JM, Hardan I, Avigdor A et al. Imatinib mesylate (STI571) in preparation for allogeneic hematopoietic stem cell transplantation and donor lymphocyte infusions in patients with Philadelphia-positive acute leukemias. Leukemia 2003; 17: 290–297. 14 Kochenderfer JN, Dudley ME, Carpenter RO, Kassim SH, Rose JJ, Telford WG et al. Donor-derived CD19-targeted T cells cause regression of malignancy persisting after allogeneic hematopoietic stem cell transplantation. Blood 2013; 122: 4129–4139. 15 Grupp SA, Kalos M, Barrett D, Aplenc R, Porter DL, Rheingold SR et al. Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N Engl J Med 2013; 368: 1509–1518.


OPEN

Five gene probes carry most of the discriminatory power of the 70-gene risk model in multiple myeloma Leukemia (2014) 28, 2410–2413; doi:10.1038/leu.2014.232 The prognostic value of gene expression profiling (GEP) in multiple myeloma (MM) has been reported by several groups.1–4 We have previously published a 70-gene classifier (GEP70) that identifies patients with high risk for short progression-free

survival (PFS) and overall survival (OS).1 The GEP70 model was developed from data on patients enrolled in Total Therapy 2 (TT2).1 Its discriminatory power has been validated in several published data sets in the transplant, non-transplant and relapse settings (reviewed in Johnson et al.5). We applied the GEP70 model to 56 previously treated patients with available baseline GEP information who were enrolled in Total Therapy 6 (TT6), a

Accepted article preview online 31 July 2014; advance online publication, 5 September 2014

Leukemia (2014) 2395 – 2424



2411 tandem transplant trial the details of which are provided in Supplementary Methods. The gene expression profiles have been deposited at the NCBI GEO data repository (http://www. ncbi.nlm.nih.gov/geo/) under GEO accession number GSE57317. Sample procurement and processing for GEP, as well as calculations of the GEP70 risk score, have been reported previously.1 The estimated 1-year survival was 62% for the high-risk group and 97% for the low-risk group by GEP70 (Supplementary Figure S1A, P o 0.0001). To investigate whether this striking difference in outcomes was driven by a few genes, all 70 probe sets of the GEP70 risk model were ranked by their P-values, based on univariate Cox regression analysis for OS in TT6 (Supplementary Table S1). The five probe sets with the smallest P-values (ENO1, FABP5, TRIP13, TAGLN2 and RFC4) were combined to create a continuous score, using methodology similar to that used to develop the GEP70 model.1 Because each of the five probe sets had a positive association with short OS in TT6, the GEP5 score was simply the mean of log2 transformed expression levels of the five probe sets. An optimal cutoff for the new risk score (hereafter referred to as GEP5) was then established with the running log-rank test, so that patients with scores higher than the cutoff were deemed to have high-risk MM and others to have low-risk (Figure 1a), with an estimated OS at 1 year of 60% and 95%, respectively (1-year PFS 50% and 91%, respectively). All five genes identified in this study were previously reported to be involved in cell proliferation and have been associated with development and survival in different cancers. ENO1 encodes alpha-enolase. Initiation of translation at an alternative translation start site results in a shorter isoform that produces MYC binding protein 1, which acts as a transcriptional repressor and possibly as a tumor suppressor.6 Overexpression of FABP5, a member of the family of fatty acid-binding proteins, was associated with poor survival in triple-negative breast cancer and with resistance to all-trans retinoic acid in a preclinical model of pancreatic ductal adenocarcinoma.7,8 TRIP13 encodes a hormone-dependent transcription factor that interacts with the ligand-binding domain of thyroid hormone receptors and may play a role in early-stage non-small-cell lung cancer.9 Association of TAGLN2 overexpression and short survival, metastasis and disease progression has been shown for several cancers.10,11 RFC4 encodes the 37-kDa subunit of the replication factor C protein complex, which, together with the proliferating cell nuclear antigen, is required for DNA elongation.12 Because the number of patients treated on TT6 was relatively small and follow-up short (median follow-up 26.5 months), a larger data set of 275 uniformly treated patients on TT3a with a longer follow-up was then used to investigate the new GEP5 score’s applicability to previously untreated myeloma. We validated the new GEP5 cutoff for patients enrolled in TT3b (n = 166).13 Gene expression data for TT3a and TT3b have previously been published and are deposited in the ArrayExpress archive (http://www.ebi.ac.uk/arrayexpress) under the accession number E-TABM-1138. A new optimal cutoff for the GEP5 model of 10.68 was identified from TT3a using the running log-rank statistics, which identified significant differences in OS and PFS for the groups with high- and low-risk disease. Importantly, these differences are comparable to those obtained by the GEP70 risk model with its established cutoff1 (Figure 1b and Supplementary Figure S1B). In the validation cohort (TT3b), risk distinction using GEP5 was very similar to GEP70 (Figure 1c and Supplementary Figure S1C) and both were comparable to results in the TT3a training set. We also applied GEP5 to a publicly available external data set of previously untreated patients (HOVON65/GMMG-HD4, n = 288)4 as a second validation set, where GEP5 also differentiated between a high-risk and © 2014 Macmillan Publishers Limited

a low-risk population with significantly different survival (Figure 1d). In order to address the question whether the five probe sets in the GEP5 were truly the best choice, we randomly selected 10 000 quintuplets from all the probe sets within the 70 gene model to create 10 000 continuous scores using the same methodology as for the GEP5 score. Among the 10 000 random scores tested, only 40 performed better in TT6. Of these 40 only 1 performed better in the TT3 test set and none was superior to GEP5 in the TT3b validation set (Supplementary Figure S2 and Supplementary Table S2). We also examined randomly selected continuous scores in TT6 with probe sets ranging between 1 and 10. Of a total of 42 485 models considered, only 1236 had a smaller P-value than GEP5 in TT6. Among those 1236 scores, 68 had a smaller P-value when tested in the TT3a test set and none performed better than GEP5 in the TT3b validation set (Supplementary Figure S3 and Supplementary Table S3). Although some of these random scores showed a better correlation with survival in single data sets, none were consistently better than the GEP5 score across different data sets. The GEP5 always ranked among the top 2% of all scores in all data sets analyzed (data not shown). On multivariate stepwise analysis, the GEP5-defined high-risk designation was selected as the most adverse variable linked to inferior PFS, with an estimated hazard ratio of 3.44 (95% CI: 2.02–5.86), whereas the GEP70 model was selected for OS (Supplementary Table S4). Table 1 summarizes the univariate survival analysis of the GEP5 and GEP70 models. Cross-tabulation of GEP70 and GEP5 risk (low vs high) for TT3A, and TT3B showed an agreement rate between the two models of 0.89, and 0.87, respectively (Supplementary Table S5). GEP70 and GEP5 currently require the use of microarray technology that interrogates the expression levels of more than 47 000 transcripts and variants simultaneously. To assess whether a more targeted approach, only measuring the expression of a small number of genes, could reliably predict risk in MM, we analyzed 48 RNA samples of previously untreated patients on TT3a and TT3b with available GEP data using the nanoString nCounter, with a code set consisting of all five genes (ENO1, FABP5, TAGLN3, TRIP13 and RFC4) of the GEP5 signature and the housekeeping genes RPL27, RPL30, RPS13, RPS29 and SRP14 (code set sequences are provided in Supplementary Table S6). Technical and biological normalization were performed using the nSolver software provided by nanoString. The correlation between microarray and nanoString-based gene expression for all five genes was between r = 0.64 and r = 0.87. Using the normalized nanoString data, we computed a nanoString-based GEP5 score (nsGEP5) applying the same methodology as for the microarray-based GEP5. nsGEP5 and GEP5 correlated very well with r = 0.852 (Supplementary Figure S4A). The receiver operator curve revealed an area under the curve of 0.897, suggesting that GEP5 high/low risk can be predicted using nsGEP5 (Supplementary Figure S4B). In summary, high-risk myeloma remains one of the greatest therapeutic challenges. The striking difference in survival of previously treated patients among GEP70 low- and high-risk groups motivated our search for fewer responsible genes. We indeed identified a set of five genes that are highly predictive of survival in multiple independent data sets. The nsGEP5 based on targeted evaluation of the expression levels of these five genes using the nanoString technology showed a very good correlation with GEP5 (based on microarray data). This new technology could reduce cost and sample requirements and has the great potential of making gene expression-driven risk assessment available to a broader patient population. However, the nsGEP5 will have to be evaluated in an independent homogeneous set of clinical samples before it can be utilized in the routine clinical setting. Recently a large-scale proteomics experiment involving 85 patients with MM identified ENO1, Leukemia (2014) 2395 – 2424


2412

Figure 1. GEP5 distinguishes a high- and a low-risk group with significantly different OS and PFS in the TT6 discovery set, the TT3A training set, and the TT3B and HOVON65/GMMG-HD4 validation sets. Left panels show overall survival, right panels progression-free survival. (a) TT6 discovery set; (b) TT3A training set; (c) TT3B validation set; (d) HOVON65/GMMG-HD4 validation set (Po 0.0001, all panels). Leukemia (2014) 2395 – 2424



Table 1. Summary of the GEP70 and GEP5 models by P values from univariate analysis when GEP5 and GEP70 were considered as both binary and continuous variables Protocol TT3a

Outcome variable

Gene predictor

P in continuous Cox analysis

P in binary log-rank analysis

OS

GEP5 GEP70 GEP5 GEP70

1.76E − 10 1.65E − 11 1.14E − 05 7.75E − 10

3.87E − 05 2.98E − 10 0.001044 2.75E − 08

GEP5 GEP70 GEP5 GEP70

4.17E − 10 2.11E − 08 1.72E − 10 3.50E − 08

5.67E − 06 1.09E − 06 3.35E − 07 1.85E − 05

PFS TT3b

OS PFS

FABP5 and TAGLN2 among a set of 24 proteins that are associated with short OS.14 This set of 85 patients included 47 who were enrolled in TT3b. The correlation of expression at both mRNA (via our GEP analyses) and protein levels supports the biological relevance of the genes included in the GEP5 model. Work is in progress to identify agents that can effectively target these prognostic genes. CONFLICT OF INTEREST BB has received research funding from Celgene and Millennium, is a consultant to Celgene and Millennium, and is a co-inventor on patents and patent applications related to use of GEP in cancer medicine that have been licensed to Myeloma Health, LLC. SZU is a consultant to Celgene, Millennium and Onyx. He has received research funding from Onyx and Celgene, and speaking honoraria from Celgene. The remaining authors declare no conflict of interest.

ACKNOWLEDGEMENTS Peggy Brenner, a science editor supported by the University of Arkansas for Medical Sciences, provided editorial assistance to the authors during preparation of this manuscript. This manuscript was supported by a grant from the National Institutes of Health (P01CA055819).

AUTHOR CONTRIBUTIONS BB, JE, JC and CJH conceived and planned the project; PQ, QZ, AH and JC conducted statistical analyses; patients were accrued by BB, FvR, SW and SZU; all authors wrote and approved the manuscript. 1

2

1

1

1

2413

REFERENCES

1

CJ Heuck , P Qu , F van Rhee , S Waheed , SZ Usmani , J Epstein , Q Zhang1, R Edmondson1, A Hoering2, J Crowley2 and B Barlogie1 1 Myeloma Institute for Research and Therapy, Little Rock, AR, USA and 2 Cancer Research and Biostatistics, Seattle, WA, USA E-mail: [email protected]

1 Shaughnessy JDJr, Zhan F, Burington BE, Huang Y, Colla S, Hanamura I et al. A validated gene expression model of high-risk multiple myeloma is defined by deregulated expression of genes mapping to chromosome 1. Blood 2007; 109: 2276–2284. 2 Decaux O, Lode L, Magrangeas F, Charbonnel C, Gouraud W, Jezequel P et al. Prediction of survival in multiple myeloma based on gene expression profiles reveals cell cycle and chromosomal instability signatures in high-risk patients and hyperdiploid signatures in low-risk patients: a study of the Intergroupe Francophone du Myelome. J Clin Oncol 2008; 26: 4798–4805. 3 Wu P, Walker BA, Brewer D, Gregory WM, Ashcroft J, Ross FM et al. A gene expression-based predictor for myeloma patients at high risk of developing bone disease on bisphosphonate treatment. Clin Cancer Res 2011; 17: 6347–6355. 4 Kuiper R, Broyl A, de Knegt Y, van Vliet MH, van Beers EH, van der Holt B et al. A gene expression signature for high-risk multiple myeloma. Leukemia 2012; 26: 2406–2413. 5 Johnson SK, Heuck CJ, Albino AP, Qu P, Zhang Q, Barlogie B et al. The use of molecular-based risk stratification and pharmacogenomics for outcome prediction and personalized therapeutic management of multiple myeloma. Int J Hematol 2011; 94: 321–333. 6 Subramanian A, Miller DM. Structural analysis of alpha-enolase. Mapping the functional domains involved in down-regulation of the c-myc protooncogene. J Biol Chem 2000; 275: 5958–5965. 7 Liu RZ, Graham K, Glubrecht DD, Germain DR, Mackey JR, Godbout R. Association of FABP5 expression with poor survival in triple-negative breast cancer: implication for retinoic acid therapy. Am J Pathol 2011; 178: 997–1008. 8 Gupta S, Pramanik D, Mukherjee R, Campbell NR, Elumalai S, de Wilde RF et al. Molecular determinants of retinoic acid sensitivity in pancreatic cancer. Clin Cancer Res 2012; 18: 280–289. 9 Kang JU, Koo SH, Kwon KC, Park JW, Kim JM. Gain at chromosomal region 5p15.33, containing TERT, is the most frequent genetic event in early stages of non-small cell lung cancer. Cancer Genet Cytogenet 2008; 182: 1–11. 10 Zhang Y, Ye Y, Shen D, Jiang K, Zhang H, Sun W et al. Identification of transgelin-2 as a biomarker of colorectal cancer by laser capture microdissection and quantitative proteome analysis. Cancer Sci 2010; 101: 523–529. 11 Nohata N, Sone Y, Hanazawa T, Fuse M, Kikkawa N, Yoshino H et al. miR-1 as a tumor suppressive microRNA targeting TAGLN2 in head and neck squamous cell carcinoma. Oncotarget 2011; 2: 29–42. 12 Loor G, Zhang SJ, Zhang P, Toomey NL, Lee MY. Identification of DNA replication and cell cycle proteins that interact with PCNA. Nucleic Acids Res 1997; 25: 5041–5046. 13 Nair B, van Rhee F, Shaughnessy JDJr, Anaissie E, Szymonifka J, Hoering A et al. Superior results of Total Therapy 3 (2003-33) in gene expression profiling-defined low-risk multiple myeloma confirmed in subsequent trial 2006-66 with VRD maintenance. Blood 2010; 115: 4168–4173. 14 Edmondson R, Chavan S, Heuck C, Epstein J, Barlogie B. Combining proteomics and gene expression profiling identifies proteins/genes associated with short overall survival in multiple myeloma. Blood (ASH Annu Meet Abstr) 2012; 120: Abstract 197.

This work is licensed under a Creative Commons AttributionNonCommercial-ShareAlike 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http:// creativecommons.org/licenses/by-nc-sa/4.0/


Phase II study of pomalidomide in high-risk relapsed and refractory multiple myeloma Leukemia (2014) 28, 2413–2415; doi:10.1038/leu.2014.248 Pomalidomide (Pom) is an IMiD immunomodulatory agent that is now FDA approved for treatment of patients who have

received ⩾ 2 prior therapies, including lenalidomide (Len) and bortezomib (Bor), and have demonstrated disease progression on or within 60 days of completion of the last line of therapy.1 There are limited data evaluating efficacy of Pom in high-risk RRMM with prior exposure/refractoriness to Len and how best to



Leukemia (2014) 2395 – 2424


2414 combine it with other available multiple myeloma (MM) therapies for this patient population. Herein, we are reporting a novel, response-adaptive, phase II clinical trial design that used Pom in high-risk RRMM with sequential addition of specific MM therapies. The study was approved by the UAMS IRB. All subjects were informed of the investigational nature of this study and gave written voluntary consent in accordance with institutional and federal guidelines. Patients with MM and high-risk features that has relapsed or resistant to ⩾ 1 prior MM therapy were eligible for enrollment. High-risk features were defined by having high risk on gene expression profiling (GEP) (GEP 702 or GEP803 signatures), elevated LDH or having abnormal metaphase cytogenetics at the time of disease relapse or progression. Patients with RRMM over 18 years of age and ECOG performance status 0–2 were eligible. All subjects were required to take aspirin or alternate prophylactic anticoagulation while on study. Cycle #1 Pom was given at 4 mg orally on days 1–21 q 4 weeks; in the absence of ⩾ PR by the IMWG criteria4 after two cycles, dexamethasone (Dex, 4–20 mg orally q week) and/or Bor (1.3 mg/m2 IV days 1,4,8,11) and/or Vorinostat (Vor, 100–200 mg po days 1–21 q 4 weeks) could be added; in the absence of ⩾ PR by IMWG criteria after four cycles, cyclophosphamide (Cy, 300 mg/m2 q week) could be added. The primary end point was progression free survival (PFS) at 1 year after enrollment; secondary end points included overall survival (OS) and overall response rates (ORR). Univariate and multivariate Cox proportional hazards regression models were used to identify associations between covariates and survival outcomes. OS and PFS were depicted using Kaplan–Meier curves. The overall response rate was defined as the percentage of patients achieving partial response (PR) or better, prior to off-study date. Best response was defined as the best response achieved while enrolled by using IMWG criteria.4 Toxicity was estimated using the National Cancer Institute Common Toxicity Criteria version 3.0. For Pom pharmacogenomic study, GEP analyses were performed on 18 patients with paired baseline and 48 h post Pom samples. In comparison with other agents, we also investigated expression changes of patients’ samples obtained prior to and 48 h after single-agent therapy with thalidomide (Thal), Len, Dex or Bor (see Supplementary file for methodology). Seventy-one patients with HRMM were enrolled between October 2011 and August 2012 (Table 1). Baseline patient characteristics included age ⩾ 65 years in 45%, cytogenetic abnormalities in 86%, GEP70 high risk in 14/34 (41%), GEP80 high risk in 28/34 (82%), extramedullary disease in 7/71 (10%) and secondary plasma cell leukemia in 3/71 (4%). In total, 68/71 patients (96%) had a prior autologous stem cell transplant (ASCT), 30/71 (42%) had ⩾ 2 ASCT. Patients had received a median of five prior lines of therapy (range 1–10). All the 71 patients (100%) were Len exposed, 58 (82%) with Len-refractory (Len-R) MM, 64 (90%) with Bor-refractory (Bor-R) MM, and 53 (75%) with dual-refractory (dual-R) MM prior to enrollment. Thirty-four (48%) patients received ⩾ 6 cycles of treatment, 9 (13%) received ⩾ 12 cycles. Patients who received ⩾ 6 cycles, had addition of Dex (67%), Bor+Dex (18%), Bor+Dex+Cy (12%), Bor+ Dex+Cy+Vor (3%) (Supplementary Figure S1). As of July 2013, 45 patients (63%) discontinued therapy primarily due to progression or death. Twenty (28%) patients achieved ⩾ PR as best response (1 Len-R, 2 Bor-R, 6 in dual-R), with median DoR for patients ⩾ PR of 3 months (range 0.5–21 months). Regimen(s) used for patients ⩾ PR in each cycle until response were listed in Supplementary Table S1 and other characteristics of responders were presented in Supplementary Table S2. Fifty-eight patients (5 Len-R, 8 Bor-R, 43 dual-R) had ⩾ stable disease, giving an overall disease control rate of 82%. The largest changes in serum M protein from baseline were displayed in Supplementary Figure S2. The OS and PFS for the whole group (n = 71) at 12 months were 63% and Leukemia (2014) 2395 – 2424

Table 1.

Patient Characteristics n/N (%)

Age ⩾ 65 yr Albumino3.5 g/dl B2M ⩾ 3.5 mg/l B2M45.5 mg/l CRP ⩾ 8 mg/l Hbo10 g/dl LDH ⩾ 190 U/l Cytogenetic abnormalities Hypodiploidy Deletion 17p t(4;14) t(14;16) t(14;20) GEP 70-gene high risk GEP 80-gene high risk GEP CD-1 subgroup GEP CD-2 subgroup GEP HY subgroup GEP MS subgroup GEP PR subgroup Lenalidomide refractory prior to enrollment Bortezomib refractory prior to enrollment

Figure 1.

32/71 33/71 41/65 20/65 31/70 30/71 23/71 61/71 35/71 6/71 1/71 2/71 0/71 14/34 28/34 2/34 8/34 3/34 5/34 14/34 58/71 64/71

(45%) (46%) (63%) (31%) (44%) (42%) (32%) (86%) (49%) (8%) (1%) (3%) (0%) (41%) (82%) (6%) (24%) (9%) (15%) (41%) (82%) (90%)

Kaplan–Meier Estimates of OS and PFS from enrollment.

13%, respectively (Figure 1). An increased risk for OS was observed among patients with LDH ⩾ 190 U/l (HR = 4.62, Po 0.001), C-reactive protein ⩾ 8 mg/l (HR = 2.13, P = 0.041), hemogloblino 10 g/dl (HR = 3.47, P o 0.001) and albumin o3.5 g/dl (HR = 2.83, P = 0.005) (Supplementary Table S3). The most common toxicities were listed in Supplementary Table S4. At the false discovery rate of 0.01 by paired significance analysis of microarrays, expression levels of 55 probesets (48 genes) significantly changed in response to Pom treatment (Supplementary Table S5). Pathway analysis indicated that these 48 genes were mostly related to cell adhesion, inflammatory and hypersensitivity response, such as ITGB1, ITGB7, LGALS3BP (Lectin), GAS6, LYN, RAC1, PYCARD, CASP1, CORO1B, F11R and DKK1, all with higher expression values observed post Pom. To address the question whether these Pom-induced gene expression alterations were unique to Pom, we also investigated expression changes of purified plasma cells obtained prior to and 48 h after single-agent therapy with Thal, Len, Dex or Bor. We found that all three IMiD class of drugs (Pom, Thal and Len) induce gene expression changes in similar directions which were not observed after Dex or Bor (Supplementary Figure S3). In addition, we checked the cereblon5 (CRBN) expression level after 48 h Pom © 2014 Macmillan Publishers Limited


but no significant change was found. However, interferon regulatory factor 4 (IRF4), one of the downstream targets of CRBN previously reported to be critical for myeloma cell survival,5,6 was downregulated post Pom treatment (P = 0.001), suggesting a CRBN-dependent anti-myeloma activity for POM. Interestingly, IRF4 expression was also decreased after Dex but not Bor. The zinc finger-containing transcription factors, Ikaros family zinc finger proteins 1 and 3 (IKZF1 and IKZF3), can be selectively bound by the CRBN-CRL4 ubiquitin ligase and resulted in ubiquitination and degradation.7,8 In myeloma cell lines, it has been reported that Len downregulates IKZF1 and IKZF3 protein levels but not their mRNA levels.7,8 In our study, we found that the mRNA level of IKZF1 was significantly upregulated in MM patients after Pom treatment (P = 0.001), which could be a response of MM cells to counter the IKZF1 protein degradation effect of Pom. But no significant change was found for the IKZF3 expression after Pom. Furthermore, we found that the expression-based proliferation index9 was lower (P = 0.05) after 48 h Pom, which implies the killing effect of Pom towards the malignant plasma cells (more discussion of Pom pharmacogenomic study can be found in Supplementary file). Pom has shown to be active in RRMM in several singleagent10,11 early phase clinical trials and perhaps more pronounced response in combination with steroids.12–15 The initial phase I study demonstrated activity with two dosing schedules10 (MTD = 2 mg QD and 5 mg QOD). The efficacy and safety data from the single-agent phase II study showed an ORR of 63%, including CR in 3 patients (5%), VGPR in 17 patients (28%), and PR in 18 patients (30%).12 More recently, Pom has been combined with other FDA approved and novel agents for RRMM with response rates varying from 48.5–72% (see Supplementary file for references). There are limited data on the differential efficacy of Pom in RRMM patients based on biologic features of the disease at the time of relapse. In the present phase II trial, a response-adaptive strategy was used to treat RRMM with high-risk features. Pom demonstrates good anti-myeloma activity in a heavily pretreated, high-risk RRMM in combination with other anti-myeloma drugs. We also performed pharmacogenomic analyses utilizing global GEP in a subset of subjects, showing that Pom has a unique CRBNmediated mechanism of action. Given the activity in patients with poor prognostic features, further studies are needed to evaluate Pom in combination with other novel agents in front line setting for high-risk MM, perhaps in a response-adaptive schema. CONFLICT OF INTEREST SZU is a consultant to Celgene, Millennium, Onyx and Sanofi. He has received research funding from ArrayBioPharma, Celgene, Onyx, Janssen, Pharmacyclics, and speaking honoraria from Celgene, Millennium and Onyx. BB has received research funding from Celgene and Novartis. He is a consultant to Celgene and Genzyme and has received speaking honoraria from Celgene and Millennium. BB is a co-inventor on patents and patent applications related to use of GEP in cancer medicine.

ACKNOWLEDGEMENTS Clinical trial and translational GEP studies were supported by Celgene Corporation.

AUTHOR CONTRIBUTIONS Designed study and wrote the manuscript: SZU, QZ; analyzed the data: SZU, BB, KS, EH, AH, JC, QZ, PQ, SY, NP; contributed patients: SZU, BB, FVR; reviewed manuscript: SZU, BB, KS, QZ, EH, AH, SW, NP, DS, SP, FVR, JC, BB.

SZ Usmani1,2, Q Zhang1, K Stratton3, P Qu3, S Yaccoby1, E Hansen3, D Steward1, S Panozzo1, N Petty1, A Hoering3, S Waheed1, F Van Rhee1, J Crowley3 and B Barlogie1 1 Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, AR, USA; 2 Levine Cancer Institute/Carolinas Healthcare System, Charlotte, NC, USA and 3 Cancer Research and Biostatistics, Seattle, WA, USA E-mail: [email protected]

REFERENCES 1 Lacy MQ, McCurdy AR.. Pomalidomide. Blood 2013; 122: 2305–2309. 2 Shaughnessy JD Jr., Zhan F, Burington BE, Huang Y, Colla S, Hanamura I et al. A validated gene expression model of high-risk multiple myeloma is defined by deregulated expression of genes mapping to chromosome 1. Blood 2007; 109: 2276–2284. 3 Shaughnessy JD Jr, Qu P, Usmani S, Heuck CJ, Zhang Q, Zhou Y et al. Pharmacogenomics of bortezomib test-dosing identifies hyperexpression of proteasome genes, especially PSMD4, as novel high-risk feature in myeloma treated with Total Therapy 3. Blood 2011; 118: 3512–3524. 4 Durie BG, Harousseau JL, Miguel JS, Blade J, Barlogie B, Anderson K et al. International uniform response criteria for multiple myeloma. Leukemia 2006; 20: 1467–1473. 5 Zhu YX, Braggio E, Shi CX, Bruins LA, Schmidt JE, Van Wier S et al. Cereblon expression is required for the antimyeloma activity of lenalidomide and pomalidomide. Blood 2011; 118: 4771–4779. 6 Shaffer AL, Emre NC, Lamy L, Ngo VN, Wright G, Xiao W et al. IRF4 addiction in multiple myeloma. Nature 2008; 454: 226–231. 7 Kronke J, Udeshi ND, Narla A, Grauman P, Hurst SN, McConkey M et al. Lenalidomide causes selective degradation of IKZF1 and IKZF3 in multiple myeloma cells. Science 2014; 343: 301–305. 8 Lu G, Middleton RE, Sun H, Naniong M, Ott CJ, Mitsiades CS et al. The myeloma drug lenalidomide promotes the cereblon-dependent destruction of Ikaros proteins. Science 2014; 343: 305–309. 9 Zhan F, Huang Y, Colla S, Stewart JP, Hanamura I, Gupta S et al. The molecular classification of multiple myeloma. Blood 2006; 108: 2020–2028. 10 Schey SA, Fields P, Bartlett JB, Clarke IA, Ashan G, Knight RD et al. Phase I study of an immunomodulatory thalidomide analog, CC-4047, in relapsed or refractory multiple myeloma. J Clin Oncol 2004; 22: 3269–3276. 11 Richardson PG, Siegel D, Baz R, Kelley SL, Munshi NC, Laubach J et al. Phase 1 study of pomalidomide MTD, safety, and efficacy in patients with refractory multiple myeloma who have received lenalidomide and bortezomib. Blood 2013; 121: 1961–1967. 12 Lacy MQ, Hayman SR, Gertz MA, Short KD, Dispenzieri A, Kumar S et al. Pomalidomide (CC4047) plus low dose dexamethasone (Pom/dex) is active and well tolerated in lenalidomide refractory multiple myeloma (MM). Leukemia 2010; 24: 1934–1939. 13 San Miguel J, Weisel K, Moreau P, Lacy M, Song K, Delforge M et al. Pomalidomide plus low-dose dexamethasone versus high-dose dexamethasone alone for patients with relapsed and refractory multiple myeloma (MM-003): a randomised, open-label, phase 3 trial. Lancet Oncol 2013; 14: 1055–1066. 14 Richardson PG, Siegel DS, Vij R, Hofmeister CC, Baz R, Jagannath S et al. Pomalidomide alone or in combination with low-dose dexamethasone in relapsed and refractory multiple myeloma: a randomized phase 2 study. Blood 2014; 123: 1826–1832. 15 Leleu X, Attal M, Arnulf B, Moreau P, Traulle C, Marit G et al. Pomalidomide plus low-dose dexamethasone is active and well tolerated in bortezomib and lenalidomide-refractory multiple myeloma: Intergroupe Francophone du Myelome 2009-02. Blood 2013; 121: 1968–1975.



Leukemia (2014) 2395 – 2424

2415


2416

AKT-induced reactive oxygen species generate imatinibresistant clones emerging from chronic myeloid leukemia progenitor cells Leukemia (2014) 28, 2416–2418; doi:10.1038/leu.2014.249 BCR-ABL1 fusion tyrosine kinase transforms hematopoietic stem cells and cause chronic myeloid leukemia in chronic phase (CML-CP), which is a stem cell (leukemia stem cell = LSC)-derived but progenitor (leukemia progenitor cell = LPC)-driven disease.1 BCR-ABL1 kinase is leukemogenic only when expressed in a HSC with self-renewal capacity, thereby inducing LSC. LSCs are capable of generating large numbers of LPCs: leukemia common myeloid (LCMPs) and leukemia granulocyte/macrophage

(LGMPs) progenitors, which cannot self-renew and eventually differentiate to mature elements. In CML blast phase leukemia progenitors (LCMP and LGMP) also display the ability to selfrenew and to sustain leukemogenesis; therefore they are LSC candidates. CML cells display the signature of genomic instability.2 The resistance of CML cells to tyrosine kinase inhibitors (TKIs) such as imatinib is often caused by point mutations in the BCR-ABL1 kinase domain, which may emerge under TKI therapy. Overall, point mutations in the kinase domain of BCR-ABL1 have been detected in 50–90% of patients with


Leukemia (2014) 2395 – 2424



acquired resistance to imatinib; TKI-resistant (TKIR) clones have been detected in Lin−CD34+ cells, including Lin−CD34+ CD38− LSCs and Lin−CD34+CD38+ LPCs.3 Generation of the second and third generation of TKIs (dasatinib, nilotinib and ponatinib) to overcome TKI resistance may result in appearance of novel mutations, including compound mutations (polymutants).4 We reported that TKI-naive and TKI-treated LSCs and LPCs accumulate high levels of ROS and oxidative DNA damage, producing mutations in BCR-ABL1 kinase causing TKI resistance.5,6 There are several possible explanations for persistent elevated levels of ROS and oxidative DNA damage in CML-CP cells surviving TKI treatment. For example, the effect of TKIs on BCR-ABL1 kinase-induced signaling pathways stimulating ROS production may be obscured by growth factors, usually resulting in incomplete inhibition or even stimulation of STAT5, AKT, RAC2 and MAPK.7,8 Therefore, instead of developing novel inhibitors to target the elusive BCR-ABL1 kinase, drugs downregulating ROS production should be combined with existing TKIs to prevent mutations and to cause more radical elimination of CML-CP cells. Phosphatidylinositol 3-kinase (PI3K) kinase–mTOR signaling has been implicated in production of ROS in BCR-ABL1-positive cell lines.9 Recently we showed that RAC2, a putative downstream effector of PI3K, can alter the electron flow through mitochondrial respiratory chain complex III to elevate ROS in LSCs and LPCs.8 Here we evaluated the role of AKT serine/threonine kinase, another PI3K downstream effector, in generation of ROS-induced oxidative DNA damage and TKI resistance in LSCs and LPCs. As described before, AKT and RAC2 were inhibited in BCR-ABL1positive 32Dcl3 cells by expression of specific dominant-negative mutants AKT(K179M) and RAC(T17N),8,10 respectively, and in Lin−

CD34+ CML-CP cells by AKT activation inhibitor perifosine11 and RAC inhibitor NSC23766, respectively (Figure 1a). Inhibition of AKT did not affect the activity of RAC and inhibition of RAC did not affect AKT activity, clearly indicating that their activation status does not depend on each other. In the presence of growth factors, AKT(K179M) mutant and perifosine diminished ROS levels in annexin V-negative living BCR-ABL1-positive 32Dcl3 cells and Lin−CD34+ cells, respectively (Figures 1b and c). Perifosine effectively downregulated ROS in Lin−CD34+ CML-CP cells in the G0/G1, S and G2/M cell cycle phases (Figure 1d), which was associated with reduction of oxidative DNA lesions, 8-oxoG and DSBs (Figures 1e and f). Finally, inhibition of AKT either by AKT (K179M) mutant or by perifosine resulted in reduction of accumulation of TKIR clones in BCR-ABL1-positive 32Dcl3 cells (Figures 1g and h). To determine if AKT is responsible for overproduction of ROS in imatinib-treated LSCs and LPCs, Lin−CD34+ CML-CP cells were incubated with imatinib in the presence of growth factors, and ABL1 and AKT activation and ROS levels were measured. Despite inhibition of ABL1 kinase activity, AKT activation was mostly preserved in imatinib-treated Lin−CD34+ cells (Figure 2a), in concordance with other report.7 As expected from previous studies with PI3K inhibitors LY294002 or wortmannin and BCRABL1 kinase inhibitor imatinib,13 targeting AKT, a downstream effector of PI3K, with perifosine enhanced the inhibitory effect of imatinib on clonogenic growth of Lin−CD34+ CML-CP cells (Supplementary Figure 1). This effect may depend on perifosine-mediated elevation of ROS (including mitochondrial ROS) in a subpopulation of annexin V-positive leukemia cells committed to apoptosis (Supplementary Figure 2), in concordance with another report.14

Figure 2. AKT elevated ROS in imatinib-treated LPCs but not in LSCs. (a) Lin−CD34+cells from a normal donor (N) and from a CML-CP patient were untreated (CML) or treated with 1 μM imatinib (CML+IM) in the presence of growth factors. Total tyrosine phosphorylated proteins (P-Tyr), AKT phosphorylated on serine 473 (pAKT) and total AKT protein were detected by western analysis. (b) Lin−CD34+ cells from 3–6 CML-CP patients were untreated (black bars) and treated with 1 μM imatinib (light grey bars), 10 μM perifosine (dark grey bars) or 1 μM imatinib+10 μM perifosine (white bars) in the presence of growth factors. Phospho-ABL1 (pABL1; phospho-Y245), phospho-AKT (pAKT; phospho-T308) and ROS (DCFDA) were detected in annexin V-negative Lin−CD34+CD38− LSCs and Lin−CD34+CD38+ LPCs as described before.5,8 *P o0.05 in comparison with untreated cells. Figure 1. RAC-independent AKT-induced ROS caused oxidative DNA damage, resulting in accumulation of imatinib-resistant clones. (a) BCR-ABL1transformed 32Dcl3 cells transfected with AKT(K179M) and Rac(T17N) dominant-negative mutants or empty plasmids (E),8,10 and Lin−CD34+ CMLCP cells treated with 10 μM AKT inhibitor perifosine, 25 μM NSC23766 or diluent (C)8,11 were tested for activation of AKT and RAC. Western analyses detect AKT phosphorylated on serine 473 (AKT-pS473) and Rac bound to GTP, as described before;8,12 total levels of AKT and RAC were also determined as loading controls. (b) ROS were measured with DCFDA in BCR-ABL1-positive 32Dcl3 cells transfected with empty plasmid (black bar) and AKT(K179M) mutant (grey bar). (c–f) Lin−CD34+ CML-CP cells were left untreated (black bars) or incubated with 10 μM perifosine (grey bars) in the presence of growth factors.8 (c) ROS were measured with DCFDA in annexin V-negative cells as described before.5,8 (d) ROS were detected by DCFDA in G1, S and G2/M phases, determined by Vybrant DyeCycle Orange live cell staining (Invitrogen/Molecular Probes) as described before.8 ROS measurements are shown on the left side, and percentages of cells in cell cycle phases are indicated at the bottom. (e) 8-oxoG and (f) γ-H2AX detected by specific immunofluorescence as described before.8 (g,h) BCR-ABL1-positive 32Dcl3 cells transfected with AKT(K179M) mutant or empty plasmid (g) and untreated (Control) or treated with 1 μM perifosine (h) were cultured for 10 weeks. The frequency of TKI-resistant (TKIR) clones was determined as described before.8 *Po0.05. © 2014 Macmillan Publishers Limited

Leukemia (2014) 2395 – 2424

2417


2418 Sustained AKT activation was also detected in imatinibtreated Lin−CD34+CD38− LSCs and Lin−CD34+CD38+ LPCs when compared to imatinib-naive counterparts (Figure 2b). However, perifosine reduced ROS levels in imatinib-naive and imatinib-treated Lin−CD34+CD38+ LPCs but not in Lin−CD34+ CD38− LSCs (Figure 2b). Therefore it appears that AKT kinase plays an important role in generation of ROS in imatinib-naive and imatinib-treated Lin−CD34+CD38+ LPCs, but it is expendable in Lin−CD34+CD38− LSCs. Since accumulation of DNA lesions such as 8-oxoG and DSBs directly depends on ROS levels in Lin− CD34+CD38− LSCs and Lin−CD34+CD38+ LPCs,5,8 we postulate that AKT kinase regulates oxidative DNA damage in LPCs, but not in LSCs. In conclusion, we hypothesize that in imatinib-treated CML-CP patients AKT serine/threonine kinase plays a prominent role in accumulation of TKIR clones emerging from Lin−CD34+CD38+ LPCs, but probably not from Lin−CD34+CD38− LSCs. The mechanisms responsible for this cell compartment-specific AKT-mediated effect on genomic instability in CML-CP are unknown. Although AKT remained active in imatinib-treated Lin−CD34+CD38+ LPCs and Lin−CD34+CD38− LSCs, intrinsic differences between leukemic progenitor and stem cells may contribute to the selective AKT effect in LPCs.15 Moreover, it remains to be determined if AKT and RAC employ overlapping or different downstream signaling pathways. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS This work was supported by NIH/NCI grant CA134458 to T Skorski.

M Nieborowska-Skorska, S Flis and T Skorski Department of Microbiology and Immunology, and Fels Institute for Cancer Research and Molecular Biology, Temple University School of Medicine, Philadelphia, PA, USA E-mail: [email protected] REFERENCES 1 Marley SB, Gordon MY. Chronic myeloid leukaemia: stem cell derived but progenitor cell driven. Clin Sci (Lond) 2005; 109: 13–25.

2 Perrotti D, Jamieson C, Goldman J, Skorski T. Chronic myeloid leukemia: mechanisms of blastic transformation. J Clin Invest 2010; 120: 2254–2264. 3 Sorel N, Bonnet ML, Guillier M, Guilhot F, Brizard A, Turhan AG. Evidence of ABL-kinase domain mutations in highly purified primitive stem cell populations of patients with chronic myelogenous leukemia. Biochem Biophys Res Commun 2004; 323: 728–730. 4 Gibbons DL, Pricl S, Posocco P, Laurini E, Fermeglia M, Sun H et al. Molecular dynamics reveal BCR-ABL1 polymutants as a unique mechanism of resistance to PAN-BCR-ABL1 kinase inhibitor therapy. Proc Natl Acad Sci USA 2014; 111: 3550–3555. 5 Bolton-Gillespie E, Schemionek M, Klein HU, Flis S, Hoser G, Lange T et al. Genomic instability may originate from imatinib-refractory chronic myeloid leukemia stem cells. Blood 2013; 121: 4175–4183. 6 Nieborowska-Skorska M, Hoser G, Hochhaus A, Stoklosa T, Skorski T. Anti-oxidant vitamin E prevents accumulation of imatinib-resistant BCR-ABL1 kinase mutations in CML-CP xenografts in NSG mice. Leukemia 2013; 27: 2253–2254. 7 Konig H, Holtz M, Modi H, Manley P, Holyoake TL, Forman SJ et al. Enhanced BCR-ABL kinase inhibition does not result in increased inhibition of downstream signaling pathways or increased growth suppression in CML progenitors. Leukemia 2008; 22: 748–755. 8 Nieborowska-Skorska M, Kopinski PK, Ray R, Hoser G, Ngaba D, Flis S et al. Rac2MRC-cIII-generated ROS cause genomic instability in chronic myeloid leukemia stem cells and primitive progenitors. Blood 2012; 119: 4253–4263. 9 Kim JH, Chu SC, Gramlich JL, Pride YB, Babendreier E, Chauhan D et al. Activation of the PI3K/mTOR pathway by BCR-ABL contributes to increased production of reactive oxygen species. Blood 2005; 105: 1717–1723. 10 Skorski T, Bellacosa A, Nieborowska-Skorska M, Majewski M, Martinez R, Choi JK et al. Transformation of hematopoietic cells by BCR/ABL requires activation of a PI-3k/Akt-dependent pathway. EMBO J 1997; 16: 6151–6161. 11 Kondapaka SB, Singh SS, Dasmahapatra GP, Sausville EA, Roy KK. Perifosine, a novel alkylphospholipid, inhibits protein kinase B activation. Mol Cancer Ther 2003; 2: 1093–1103. 12 Ren SY, Bolton E, Mohi MG, Morrione A, Neel BG, Skorski T. Phosphatidylinositol 3kinase p85{alpha} subunit-dependent interaction with BCR/ABL-related fusion tyrosine kinases: molecular mechanisms and biological consequences. Mol Cell Biol 2005; 25: 8001–8008. 13 Klejman A, Rushen L, Morrione A, Slupianek A, Skorski T. Phosphatidylinositol-3 kinase inhibitors enhance the anti-leukemia effect of STI571. Oncogene 2002; 21: 5868–5876. 14 Rahmani M, Reese E, Dai Y, Bauer C, Payne SG, Dent P et al. Coadministration of histone deacetylase inhibitors and perifosine synergistically induces apoptosis in human leukemia cells through Akt and ERK1/2 inactivation and the generation of ceramide and reactive oxygen species. Cancer Res 2005; 65: 2422–2432. 15 Kornblau SM, Qutub A, Yao H, York H, Qiu YH, Graber D et al. Proteomic profiling identifies distinct protein patterns in acute myelogenous leukemia CD34+CD38stem-like cells. PLoS One 2013; 8: e78453.


OPEN

Relationship of different platelet response criteria and patient outcomes in a romiplostim myelodysplastic syndromes trial Leukemia (2014) 28, 2418–2421; doi:10.1038/leu.2014.253 Thrombocytopenia in lower-risk myelodysplastic syndrome (MDS) contributes to an increased risk of bleeding and is associated with shortened survival.1–3 Empiric platelet response criteria have been used in MDS clinical trials mostly with disease-modifying drugs, often as surrogates for clinical outcomes (Table 1a). The value of these criteria has not been rigorously evaluated, most importantly

not in trials of agents specifically targeting platelet production. Romiplostim is currently approved in the United States for the treatment of thrombocytopenia in patients with chronic immune thrombocytopenia (ITP) who have had an insufficient response to corticosteroids, immunoglobulins or splenectomy, and is marketed under the name Nplate. Results of trials of romiplostim in MDS suggest that romiplostim treatment improves platelet counts as monotherapy and when combined with azacitidine, decitabine or lenalidomide.4–7

Accepted article preview online 2 September 2014; advance online publication, 23 September 2014

Leukemia (2014) 2395 – 2424



2419 Table 1a.

Platelet response criteria evaluated

Platelet response criteria

Duration Definition 8 weeks For patients with baseline platelets o100 × 109/l: Baseline platelets 420 × 109/l, absolute increase of ⩾ 30 × 109/l Baseline platelets o20 × 109/l, increase to 420 × 109/l and by ⩾ 100% 8 weeksa For patients with baseline platelets o100 × 109/l: An absolute increase of ⩾ 30 × 109/l For platelet transfusion-dependent patients, stabilization of platelet counts and platelet transfusion independence 8 weeksa For patients with baseline platelets o100 × 109/l: A ⩾ 50% increase in platelet count with an absolute increase 410 × 109/l but o30 × 109/l

17

IWG 2006 HI-P

IWG 2000 Major11

IWG 2000 Minor11

Italian MDS group10 Complete (same as IWG AML 200318) None Any None

Platelet count 4100 × 109/l and no bleeding Baseline platelets 420 × 109/l, no bleeding and absolute increase of ⩾ 30 × 109/l Baseline platelets o20 × 109/l, increase to 420 × 109/l and by at least 100% 4 weeks Continuous platelet response

Durable CALGB19 ITP20

None None

⩾ 50% restitution of the initial deficit (to 140 × 109/l) Platelet count450 × 109/l

Abbreviations: AML, acute myeloid leukemia; CALGB, Cancer and Leukemia Group B; HI-P, hematologic improvement—platelet; ITP, immune thrombocytopenia; IWG, International Working Group; MDS, myelodysplastic syndrome. aAs platelets were measured every week in this trial, for IWG 2000 response measures, an 8-week duration was used for these analyses, rather than 2 months.

Table 1b.

Association of platelet response rates and transfusion needs with various platelet response criteria

Platelet response criteria

IWG 2006 HI-P IWG 2000 Major or Minor

P-value

Bleedingb

Romiplostim N = 167, n (%)

3 (3.6) 2 (2.4)

61 (36.5) 56 (33.5)

o0.001 15.6* (4.7–51.8) o0.001 21.2* (5.0–90.3)

9.8* (7.8–12.3) 13.0* (9.8–17.2)

2.3* (2.1–2.6) 1.4* (1.0–2.0) 2.6* (1.4–4.7) 2.5* (2.2–2.8) 1.2 (0.8–1.6) 2.2* (1.2–4.1)

63 (37.7) 115 (68.9) 82 (49.1)

o0.001 15.4* (5.0–47.2) o0.001 5.1* (2.8–9.1) o0.001 19.7* (6.9–56.5)

5.8* (4.7–7.2) 2.6* (2.3–2.9) 9.5* (7.8–11.4)

2.6* (2.3–2.9) 1.8* (1.2–2.7) 1.9* (1.0–3.4) 2.1* (1.9–2.3) 1.2 (0.9–1.7) 1.5 (0.9–2.6) 2.5* (2.3–2.7) 1.8* (1.3–2.4) 2.2* (1.2–3.8)

75 (44.9) 95 (56.9)

o0.001 12.5* (4.9–31.5) o0.001 6.3* (3.2–12.4)

1.8* (1.6–2.0) 2.8* (2.4–3.2)

2.8* (2.5–3.1) 0.8 (0.6–1.1) 2.1* (1.9–2.3) 0.9 (0.7–1.2)

6 (7.2) 17 (20.5)

All RR (95% CI)

OR (95% CI)b

CSBE RR (95% CI)

OSc

Placebo N = 83, n (%)

Italian MDS group Complete (IWG AML 2003) 4 (4.8) Any 26 (31.3) Durable 4 (4.8) CALGB ITP

Platelet transfuse RRa (95% CI)

Response rate

HR (95% CI)

1.6 (0.9–2.9) 1.6 (0.9–2.9)

Abbreviations: AML, acute myeloid leukemia; CALGB, Cancer and Leukemia Group B; CI, confidence interval; CSBE, clinically significant bleeding event; HI-P, hematologic improvement—platelet; HR, hazard ratio; IPSS, International Prognostic Scoring System; ITP, immune thrombocytopenia; IWG, International Working Group; MDS, myelodysplastic syndrome; OR, odds ratio; OS, overall survival; RR, rate ratio. aRomiplostim-treated patients only, responders vs nonresponders; adjusted for baseline platelet count and IPSS status. bRomiplostim vs placebo; adjusted for baseline platelet count and IPSS status. c Romiplostim-treated patients, nonresponders vs responders; adjusted for baseline age (65 years or older) and IPSS status. Statistically significant OR and RR are bolded with asterisks.

Data from a 58-week placebo-controlled trial of romiplostim in patients with lower-risk MDS and thrombocytopenia7 were used to evaluate how the platelet response measures associate with clinical outcomes. This trial was discontinued early owing to concerns that the potential small benefit of bleeding reduction did not outweigh the potential risk for diagnosis of acute myeloid leukemia (AML) as defined by pathology or initiation of AML-type treatment; the final analysis of the 58-week data set showed comparable AML rates in both arms.7 We describe here to what extent the currently clinically available platelet response criteria were applicable in this study and whether they were associated with platelet transfusions, bleeding and overall survival. Patients with International Prognostic Scoring System (IPSS) low/int-1 risk MDS (N = 250) were randomized 1:2 to 26 weeks of weekly placebo (N = 83) or romiplostim (N = 167), with dose adjusted for platelet count, followed by a 4-week washout period and bone marrow biopsy, another 24 weeks as randomized (extended treatment period), and a second 4-week © 2014 Macmillan Publishers Limited

washout period and bone marrow biopsy (ClinicalTrials.gov NCT00614523).7 The 4-week washouts occurred before bone marrow biopsies so that the study drug would not affect results, as romiplostim can be associated with transient blast cell count elevations. Patients then entered long-term follow-up. Eligible patients were receiving supportive care only (that is, not disease-modifying therapy), with platelet counts (1) ⩽ 20 × 109/l or (2) ⩽ 50 × 109/l and a history of bleeding. Patients were stratified by baseline IPSS status (low, int-1) and platelet count (o , ⩾ 20 × 109/l). Platelet response was assessed using various platelet response criteria (Table 1a). Clinically significant bleeding events (CSBEs) were defined as grade ⩾ 2 on the modified World Health Organization (WHO) bleeding scale.8,9 All analyses were performed post hoc. The Fisher’s exact test was used to test the association of romiplostim and platelet response. The Mantel–Haenszel method was used to calculate a pooled odds ratio of romiplostim and placebo across the randomization stratification factors of baseline platelet count Leukemia (2014) 2395 – 2424


2420

Figure 1. Survival curves. Shown are the survival probability for those who had achieved an IWG 2006 HI-P response (red dashed line) and those who did not (solid blue line), all in romiplostimtreated patients. Number of patients at risk is shown on the graph immediately above the x axis. P = 0.0070 by log-rank test.

and IPSS. A Cox regression model including prognostic factors and platelet response as covariates was used to predict overall survival in romiplostim-treated patients. Poisson regression models, including baseline platelet count, IPSS status and platelet response as covariates, were used to predict bleeding events and platelet transfusions in romiplostim-treated patients. As this was an analysis based on response, a landmark sensitivity analysis was performed using platelet counts from the first 26 weeks to determine platelet response, excluding patients who discontinued the trial within 26 weeks, and analyzing outcomes that occurred after 26 weeks; survival curves by response status were plotted per landmark at week 26. CSBEs were too infrequent for the landmark sensitivity analysis to be meaningful. Overall survival was recorded up to the last observation in long-term follow-up, and bleeding events and platelet transfusions were measured in the extended treatment period only. The six platelet response measures (Table 1a) were used to evaluate changes in platelet counts in the 58-week placebocontrolled trial in post hoc analyses. Patients (placebo, N = 83; romiplostim, N = 167) were mostly male (59.2%) and Caucasian (94.0%). Median (Q1, Q3) age was 70.0 (61.0, 77.0) years and median (Q1, Q3) baseline platelet count was 19.3 (12.5, 30.3) × 109/l. Median (Q1, Q3) MDS duration was 0.44 (0.13, 1.74) years. Most patients were MDS WHO classification refractory cytopenia with multilineage dysplasia (67.6%). Romiplostim treatment was significantly associated with platelet response by all criteria studied (Table 1b). For example, romiplostim-treated subjects were 15.6 times more likely to have hematologic improvement—platelet (HI-P) than placebo-treated subjects. All platelet response criteria also reflected whether patients required platelet transfusions, with nonresponders having more platelet transfusions than responders. The association between platelet response criteria and clinical outcomes such as bleeding (all and CSBE) were evaluated, as in Table 1b. All response criteria showed significant association between response status and overall bleeding events, with nonresponders being more likely than responders to have bleeding events. Only HI-P, complete response as presented by the Italian MDS group,10 and durable response were significantly associated with less CSBE. These same measures, and International Working Group (IWG) 2000 criteria, were significantly associated with improved overall survival. Survival curves for HI-P, the platelet response measure most significantly associated with survival, are shown in Figure 1 for romiplostim-treated patients. AML rates for Leukemia (2014) 2395 – 2424

romiplostim-treated patients with HI-P as compared with those without HI-P were similar, 8.5% vs 8.3%, with an odds ratio (95% confidence interval) of 1.02 (0.31, 3.38). Landmark sensitivity analyses were performed for all measures described above to determine whether this being an analysis based on response and discontinuation of patients affected trial results. For the overall survival end point, after excluding patients who discontinued in the first 26 weeks, the sample size decreased from 167 to 143. Among the 24 subjects who were excluded, 12 died. Smaller sample size and fewer events contributed to slightly larger P-values, although results were generally consistent with the original analyses. Differences included that complete response10 and IWG 2000 major response11 were marginally significantly associated with overall survival (P = 0.077 and 0.053, respectively). All platelet response measures remained significantly associated with all bleeding and platelet transfusions (data not shown). Data from this large placebo-controlled romiplostim trial indicate that platelet response criteria, developed empirically from clinical experience and trials using disease-modifying agents, are heterogeneous, result in a wide range of response rates for the same patient population and are predictive of clinical outcomes in thrombocytopenic MDS patients treated with a thrombopoietin mimetic. This is in keeping with the finding that thrombocytopenia per se has been associated with worse prognosis in MDS, including an increased risk of disease progression.1–3 For the first time, we show that platelet response to a thrombopoietin mimetic is positively associated with overall survival. Possibly, romiplostim has a beneficial effect through reducing potentially life-threatening thrombocytopenia or other as-yet-unrecognized broader effects. Thrombopoietin has previously been shown to stimulate other hematopoietic lineages.12 It is unclear whether the improved outcomes are associated with response to romiplostim or that, inherently, patients that respond have better outcomes. While no difference in survival was seen with romiplostim vs placebo,7 better selection of patients could lead to improved survival outcomes with romiplostim. A positive association between treatment and survival is also seen for the disease-modifying therapy azacitidine in higher-risk MDS.13 For studies examining survival and treatment with erythropoiesis-stimulating agents (ESAs) in MDS, results have been mixed. A retrospective multivariate analysis of patients treated with ESAs with or without granulocyte colony-stimulating factors (G-CSFs) reported that survival, but not disease progression, was improved in the ESA-responsive cohort compared with an untreated IPSS/IMRAW (International MDS Risk Analysis/Workshop) cohort.14 Another multivariate analysis, comparing patients treated with ESAs plus G-CSFs with a control cohort of untreated MDS patients, found better survival with ESA treatment, particularly for those requiring fewer than two red blood cell units transfused per month.15 However, a small randomized phase 3 Eastern Cooperative Oncology Group (ECOG) study of patients receiving supportive care alone or supportive care plus ESAs with or without G-CSFs found no difference in survival, and that survival was increased for those who responded to ESA treatment.16 Whether ESA treatment improves survival in MDS may become clearer as ongoing studies report results. In summary, these data indicate that platelet response criteria, specifically those that incorporate durable response, such as IWG 2006, correlate with overall survival and have the potential to be used as interim markers for clinically significant outcomes. However, a limitation of this data set is that the study drug treatment was ended early owing to concerns regarding transient increases in peripheral blast cell counts with romiplostim that put patients at risk for the diagnosis of and treatment for AML. Therefore, the data set was incomplete, and it is possible that different results regarding the association of platelet response measures and clinical outcomes would have been obtained with a © 2014 Macmillan Publishers Limited


2421 fuller data set. Evaluation is needed of these associations in either past MDS clinical trials of interventions to raise platelets or future ones to confirm that these findings occur in the broader context. CONFLICT OF INTEREST UP has received honoraria from Amgen, GlaxoSmithKline (GSK), Celgene and Novartis. MAS is a member of advisory boards for Amgen and Celgene, and a speaker for Celgene. HK has research grants from Amgen. AG is a member of advisory boards and a speaker for Amgen and GSK. GJM has been a member of advisory boards and speakers bureau for Amgen, Novartis, GSK and Celgene, received research funding from Celgene and has been a consultant for GSK. CJ and ASY are employees of and stockholders in Amgen. PF has received honoraria and research grants from Amgen, Celgene, Janssen Roche, GSK and Novartis.

ACKNOWLEDGEMENTS Medical writing assistance was provided by Susanna Mac of Amgen. Funding for these analyses was provided by Amgen.

AUTHOR CONTRIBUTIONS All authors were involved in analysis and interpretation of the data, drafting the publication and/or revising it critically for important intellectual content, and approving the final draft.

DISCLAIMER We had full access to the data and are fully responsible for content and editorial decisions for this manuscript.

U Platzbecker1, MA Sekeres2, H Kantarjian3, A Giagounidis4, GJ Mufti5, C Jia6, AS Yang7 and P Fenaux8 1 University Hospital Carl Gustav Carus Dresden, Medizinische Klinik und Poliklinik I, Dresden, Germany; 2 Leukemia Program, Cleveland Clinic Taussig Cancer Institute, Cleveland, OH, USA; 3 Department of Leukemia, The University of Texas MD Anderson Cancer Center, Houston, TX, USA; 4 Clinic for Oncology, Hematology and Palliative Medicine, Marien Hospital Düsseldorf, Düsseldorf, Germany; 5 King’s College London, London, UK; 6 Amgen Inc., South San Francisco, CA, USA; 7 Amgen Inc., Thousand Oaks, CA, USA and 8 Service d'hématologie clinique, Hopital Avicenne Universite Paris XIII, Bobigny, France E-mail: [email protected] REFERENCES 1 Kantarjian H, Giles F, List A, Lyons R, Sekeres MA, Pierce S et al. The incidence and impact of thrombocytopenia in myelodysplastic syndromes. Cancer 2007; 109: 1705–1714. 2 Kantarjian H, O'Brien S, Ravandi F, Cortes J, Shan J, Bennett JM et al. Proposal for a new risk model in myelodysplastic syndrome that accounts for events not considered in the original International Prognostic Scoring System. Cancer 2008; 113: 1351–1361. 3 Sekeres MA, Schoonen WM, Kantarjian H, List A, Fryzek J, Paquette R et al. Characteristics of US patients with myelodysplastic syndromes: results of six cross-sectional physician surveys. J Natl Cancer Inst 2008; 100: 1542–1551. 4 Greenberg PL, Garcia-Manero G, Moore M, Damon L, Roboz G, Hu K et al. A randomized controlled trial of romiplostim in patients with low- or intermediate-risk myelodysplastic syndrome (MDS) receiving decitabine. Leuk Lymph 2012; 54: 321–328. 5 Kantarjian HM, Giles FJ, Greenberg PL, Paquette RL, Wang ES, Gabrilove JL et al. Phase 2 study of romiplostim in patients with low- or intermediate-risk myelodysplastic syndrome receiving azacitidine therapy. Blood 2010; 116: 3163–3170.


6 Wang ES, Lyons RM, Larson RA, Gandhi S, Liu D, Matei C et al. A randomized, double-blind, placebo-controlled phase 2 study evaluating the efficacy and safety of romiplostim treatment of patients with low or intermediate-1 risk myelodysplastic syndrome receiving lenalidomide. J Hematol Oncol 2012; 5: 1–13. 7 Giagounidis A, Mufti GJ, Kantarjian HM, Fenaux P, Sekeres MA, Szer J et al. Results of a randomized, double-blind study of romiplostim versus placebo in patients with low/intermediate-1-risk myelodysplastic syndrome and thrombocytopenia. Cancer 2014; 120: 1838–1846. 8 Heddle NM, Cook RJ, Webert KE, Sigouin C, Rebulla P. Methodologic issues in the use of bleeding as an outcome in transfusion medicine studies. Transfusion 2003; 43: 742–752. 9 Rebulla P, Finazzi G, Marangoni F, Avvisati G, Gugliotta L, Tognoni G et al. The threshold for prophylactic platelet transfusions in adults with acute myeloid leukemia. Gruppo Italiano Malattie Ematologiche Maligne dell'Adulto. N Engl J Med 1997; 337: 1870–1875. 10 Oliva EN, Santini V, Zini G, Palumbo GA, Poloni A, Cortelezzi A et al. Efficacy and Safety of Eltrombopag for the Treatment of Thrombocytopenia of Low and Intermediate-1 IPSS Risk Myelodysplastic Syndromes: Interim Analysis of a Prospective, Randomized, Single-Blind, Placebo-Controlled Trial (EQoL-MDS). Blood (ASH Annual Meeting Abstracts) 2012; 120: Abstract 923. 11 Cheson BD, Bennett JM, Kantarjian H, Pinto A, Schiffer CA, Nimer SD et al. Report of an international working group to standardize response criteria for myelodysplastic syndromes. Blood 2000; 96: 3671–3674. 12 Kaushansky K, Broudy VC, Grossmann A, Humes J, Lin N, Ren HP et al. Thrombopoietin expands erythroid progenitors, increases red cell production, and enhances erythroid recovery after myelosuppressive therapy. J Clin Invest 1995; 96: 1683–1687. 13 Fenaux P, Mufti GJ, Hellstrom-Lindberg E, Santini V, Finelli C, Giagounidis A et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet Oncol 2009; 10: 223–232. 14 Park S, Grabar S, Kelaidi C, Beyne-Rauzy O, Picard F, Bardet V et al. Predictive factors of response and survival in myelodysplastic syndrome treated with erythropoietin and G-CSF: the GFM experience. Blood 2008; 111: 574–582. 15 Jadersten M, Malcovati L, Dybedal I, Della Porta MG, Invernizzi R, Montgomery SM et al. Erythropoietin and granulocyte-colony stimulating factor treatment associated with improved survival in myelodysplastic syndrome. J Clin Oncol 2008; 26: 3607–3613. 16 Greenberg PL, Sun Z, Miller KB, Bennett JM, Tallman MS, Dewald G et al. Treatment of myelodysplastic syndrome patients with erythropoietin with or without granulocyte colony-stimulating factor: results of a prospective randomized phase 3 trial by the Eastern Cooperative Oncology Group (E1996). Blood 2009; 114: 2393–2400. 17 Cheson BD, Greenberg PL, Bennett JM, Lowenberg B, Wijermans PW, Nimer SD et al. Clinical application and proposal for modification of the International Working Group (IWG) response criteria in myelodysplasia. Blood 2006; 108: 419–425. 18 Cheson BD, Bennett JM, Kopecky KJ, Buchner T, Willman CL, Estey EH et al. Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol 2003; 21: 4642–4649. 19 Silverman LR, Demakos EP, Peterson BL, Kornblith AB, Holland JC, Odchimar-Reissig R et al. Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: a study of the cancer and leukemia group B. J Clin Oncol 2002; 20: 2429–2440. 20 Kuter DJ, Rummel M, Boccia R, Macik BG, Pabinger I, Selleslag D et al. Romiplostim or standard of care in patients with immune thrombocytopenia. N Engl J Med 2010; 363: 1889–1899.

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/

Leukemia (2014) 2395 – 2424


2422

Early death in acute promyelocytic leukemia (APL) in French centers: a multicenter study in 399 patients Leukemia (2014) 28, 2422–2424; doi:10.1038/leu.2014.240 In the past two decades, the combination of all-trans retinoic acid (ATRA) and anthracycline-based chemotherapy (CT) has improved the outcome of acute promyelocytic leukemia (APL) compared with CT alone. This has been achieved by reducing the incidence of relapse and increasing the complete remission (CR) rate, the latter through almost disappearance of cases of resistant leukemia and, in most studies, by reducing the frequency of early deaths.1–4 In recent APL clinical trials, CR ranging from 90 to 95% have been reported, almost all failures resulting from early death (ED), mainly caused by bleeding, sepsis or differentiation syndrome (DS).5,6 Some recent studies have, however, reported ED rates as high as 15–30% in newly diagnosed APL when patients not included in clinical trials were analyzed together with protocol patients. In addition, when assessed in the United States SEER database,7 ED did not seem to have much improved between 1977 and 2007, in spite of the advent of ATRA treatment. We retrospectively analyzed the incidence and causes of ED in APL in 17 French centers over a 5-year period (December 2006 to December 2011) included or not in APL 2006 trial, the only active French APL trial during that period. Diagnosis of APL was made on a morphological basis, and had to be confirmed by presence of t(15;17) translocation and/or promyelocytic leukemia gene-retinoic acid recepter alpha gene rearrangement. Every effort was made to collect all APL cases that presented during the 5-year period of analysis but was not included in APL 2006 by checking all APL diagnoses made in the hematology, cytogenetic and molecular laboratories of the institution, and by including patients admitted directly to the hospital intensive care unit (ICU) or to clinical departments of the hospital other than that of hematology to avoid any bias. In APL 2006 trial, patients aged o 70 years received an ATRA plus CT induction course consisting of ATRA 45 mg/m2/d until hematological CR and idarubicin 12 mg/m2/d for 3 days plus aracytin (AraC) 200 mg/m2/d for 7 days. Patients achieving CR were then randomized to consolidation cycles with or without arsenic trioxide (ATO), and to maintenance treatment. Patients aged 470 years received a regimen combining ATRA, ATO and attenuated chemotherapy cycles. Patients not included in APL 2006 generally received ATRA combined with an induction CT regimen similar to that of APL 2006 trial. For each patient, intervals between first blood count, hospital admission and ATRA onset were particularly analyzed. ED was defined as death within 30 days of admission, irrespective of its cause. As shown in Table 1, 399 cases of newly diagnosed APL were diagnosed between December 2006 and December 2011 in the 17 participating centers. Females accounted for 48% of the patients (193 F/206 M) and median age was 51 years (16–87). Thirty-six (9%) patients were older than 75 years and 105 (26%) older than 65 years. White blood cell (WBC) count o 10 000/mm3 (standardrisk APL) was observed in 73% of patients and 27% had WBC 410 000/mm3 (high-risk APL). In all, 274 patients (68.7%) were included in APL 2006 trial and 125 (31.3%) were not included. Reasons for non-inclusion in APL 2006 trial were: 29 (7.3%) patients refusing or unable to give consent, 20 (5%) initial admissions in ICU, 20 (5%) older age (median age in those patients was 82.5 years, range 71–87) and/or comorbidities, 9 (2.3%)

previous cancers, 11 (2.7%) contraindications to anthracyclines, 5 (1.4%) other exclusion criteria (pregnancy, HIV and poor socioeconomic conditions), 6 (1%) rapid deaths before ATRA onset and 25 (6.3%) unknown reasons. Patients not included in APL 2006 trial were characterized by older age (median 62 vs 47 years, P = 0.0001) but a similar proportion of high-risk patients (30% vs 25%, P = 0.39), whereas there were no significant differences for other baseline characteristics. Twenty of the 36 (55.5%) patients aged ⩾ 75 years and 20 of the 32 (62.5%) patients initially admitted in ICU were not included in APL 2006 trial. Median interval from first abnormal blood count to ATRA onset was 1 day (Q1–Q3: 0–3), and o1 day, 1 day and 41 day in 28%, 26% and 46% of the patients, respectively. Similarly, median interval from hospitalization to ATRA onset was o 1 day (Q1–Q3: 0–1) and o 1 day, 1 day and 41 day in 51%, 24% and 25% of the patients, respectively. Finally, median interval from bone marrow aspirate to ATRA onset was also o 1 day (Q1–Q3: 0–1) and was o1 day, 1 day and 41 day in 68%, 17% and 15% of the patients, respectively. In patients not included in APL trial, 90 (72%) received ATRA plus anthracycline-based CT; 26 (24.2%) received ATRA plus nonanthracycline-based CT: 4 (3.2%) received ATRA plus ATO, 3 (2.6%) received ATRA plus AraC and ATO, and 15 (12%) received ATRA alone. The remaining nine non-included patients (7%) died without having received any treatment. Median age of those nine patients was 72 years and five had baseline WBC 410 000/mm3. In two of them, aged 71 and 84 years, attending physicians decided not to start treatment owing to major comorbidities, whereas one, aged 87, refused treatment. Of the remaining six patients, three were initially admitted in ICU, and all six died within 24 h of their admission, mainly due to central nervous system (CNS) bleeding. CR was achieved in 362/399 (90.4%) of the patients, 9 patients died before treatment onset, 1 patient had resistant leukemia and 27 patients (6.9%) had ED after treatment onset. Causes of ED included DS (n = 4), CNS bleeding (n = 11, including 5 patients who died before treatment), sepsis (n = 7), myocardial infarction (n = 2), multiple organ failure (n = 5) and uncertain (n = 7). After exclusion of patients who died before receiving ATRA, ED was seen in 2.5% of patients included in APL 2006 trial and 17% of non-protocol patients (P o 0.0001). Other prognostic factors of ED were older age (P = 0.005) and initial admission in ICU (P o 10−4), whereas high WBC (P = 0.155) count had no significant impact. None of the three intervals analyzed here (from first abnormal blood count, from hospitalization and from bone marrow aspirate to ATRA onset) had any influence on the ED rate. Moreover, 20 (64.5%) of the 31 patients admitted directly in ICU, 30 (83%) of the 36 patients aged 75 or greater, and 93 (86%) of the 108 high-risk patients achieved CR. Twenty-nine patients had prolonged intervals (⩾5 days) from hospitalization to ATRA onset, including three high-risk patients. Three of them, all with standard-risk APL at baseline, had ED. Thus, in our series, the ED rate was 9.6%, and appeared lower than the 17–29% ED rates reported in several recent studies where protocol and non-protocol APL patients were included7–12 (Table 2). Our results were closer to the 11% ED rate reported in another similar study.10 Reasons to explain those differences are uncertain. One would tend to attribute our relatively low ED rate


Leukemia (2014) 2395 – 2424



2423 Table 1.

Baseline patient characteristics n (%)

Patient characteristics

399 51 (37–66) 105 (26%) 36 (9%) 206 (51.6%) 27%

n Median age, years (Q1–Q3)a Age 465 years Age 475 years Male gender WBC 410 G/l (high-risk APL) Hospital department of first admission Hematology Medicine Surgery ICU

77% 13% 2% 8%

Abbreviations: APL, acute promyelocytic leukemia; ICU, intensive care unit; Q, quartile; WBC, white blood cell. aQ1–Q3, 25–75%.

Table 2. Early death rates recently reported in newly diagnosed APL (irrespective of inclusion or not in a clinical trial) Number of patients

Period

Early death rate (%)

105 1400 70 204 399

1997–2006 1992–2007 1997–2009 1992–2009 1993–2007

29 17.3 26 11 21.8

Swedish Registry8 SEER (US database)7 Stanford University9 Chicago, New York and Haifa10 a Canadian Cancer Registry12 a

Haifa: Rambam Medical Center (Northern Israel’s largest hospital and a tertiary referral center); Chicago: Northwestern University (University medical center) and John J Stroger Hospital of Cook County (Public Hospital); New York: Memorial Sloan-Kettering Cancer Center (Freestanding cancer center).

to the fact that a large majority of patients were very rapidly admitted, and that ATRA was rapidly started, as soon as APL was suspected by bone marrow morphology: 51% of our patients received ATRA on the day of admission, 68% received ATRA on the day bone marrow aspirate was done and only 25% received ATRA 41 day after admission. Altman et al.10 reported that the percentage of ED secondary to bleeding increased significantly when the onset of ATRA was delayed beyond the day APL was suspected. In our study, we found no significant difference between patients who experienced ED and those who achieved CR when analyzing intervals from hospital admission to ATRA onset, but the number of patients with delay in treatment onset was quite small. Also of note was the fact that, in our report, all patients were diagnosed and treated after 2006, whereas the five previous reports analyzing EDs in APL included patients often diagnosed in the 1990s, when the importance of rapid onset of ATRA treatment was possibly less appreciated. Moreover, in a recent Canadian study, ED rate improved over time, suggesting that the importance of rapid treatment was less appreciated in the 1990s.12 In the present report, 31.3% of the patients were not included in the current French national trial, mainly owing to older age, comorbidities, direct admission in ICU and rapid death. Although the proportion of patients included in clinical trials was probably lower in the other reports, this number was available only in the Stanford University experience (10 of the 70 patients included in a clinical trial).9 Not unexpectedly, the outcome of non-protocol patients was significantly poorer than that of patients included in the trial (82% vs 97.5% CR). This was largely attributable to the higher incidence of ED seen in patients older than 75 and patients © 2014 Macmillan Publishers Limited

with direct admission to ICU. Older age was found to be a prognostic factor of ED in all previously published APL series.7–10 Nevertheless, the CR rate of 83% in elderly patients remains quite high. As APL is not associated with a higher relapse risk in elderly patients, avoiding ED is a priority in that age group. Likewise, the CR rate of 64.5% observed in APL patients initially admitted to the ICU is rather high. Although we probably captured all APL cases admitted in the 17 centers analyzed in this report, which were the largest centers treating APL patients in France, we cannot be certain that the proportion of patients included in APL 2006 trial was similar in other smaller French centers. In those centers, a potential higher proportion of non-protocol patients could potentially contribute to a higher ED rate. One could also argue that we overlooked some possible APL deaths occurring before reaching a hematology department. However, evaluating the number of those potential cases will always remain difficult, as those patients precisely do not reach specialized centers able to make a definite diagnosis of APL. Thus, we found that, in large French centers, APL patients are generally rapidly admitted to the hospital where ATRA is rapidly started on the basis of morphological diagnosis. This may have contributed to the relatively low rate of ED (9.6%) observed. Avoiding ED in APL is particularly crucial given the low relapse rate now observed, even in high-risk patients.11 Whether substitution of ATRA–ATO regimens for ATRA–chemotherapy regimens, at least in standard-risk APL, will further reduce the ED rate, will have to be confirmed.

CONFLICT OF INTEREST The authors declare no conflict of interest.

AUTHOR CONTRIBUTIONS XT, CR, NV, JD, ED, PH, DB, J-BM, AS, CM, CP, PB, J-PM, MH-B, NF, ER and HD provided the data. RR collected the data. LD, PF and LA designed the research. RR, PF and LA wrote the paper. LA analyzed the data.

R Rahmé1, X Thomas2, C Recher3, N Vey4, J Delaunay5, E Deconinck6, P Hirsch7, D Bordessoule8, J-B Micol9, A Stamatoullas10, C Mariette11, C Pautas12, P Bories13, J-P Marolleau14, M Hunault-Berger15, N Fegueux16, E Raffoux17, H Dombret17, L Degos17, P Fenaux1 and L Adès1 1 Service d'Hématologie Senior, Hôpital Saint Louis, Université Paris 7, Paris, France; 2 Department of Hematology, Centre Hospitalier Lyon Sud, Lyon, France; 3 Hematology, CHU Purpan, University of Toulouse, Toulouse, France; 4 Hematology Department, Institut Paoli Calmettes, Marseille, France; 5 Service d'Hématologie Clinique, Centre Hospitalier Universitaire de Nantes, Nantes, France; 6 Hematology, CHU Besancon, Besancon, France; 7 Service d’Hématologie Clinique, Hôpital Saint-Antoine, AP-HP, Paris, France; 8 CHU Limoges, Service d'Hématologie Clinique et Thérapie Cellulaire, Limoges, France; 9 Hematology, Institut Gustave Roussy, Villejuif, France; 10 Hematology Department, Centre Henri Becquerel, Rouen, France; 11 Onco-Hematologie, CHU Grenoble, Grenoble, France; 12 Hôpital Henri Mondor, AP-HP, Université Paris 12, Créteil, France; 13 CHU Hautepierre, Université Strasbourg, Strasbourg, France; 14 Hematologie, Hopital Sud, CHU Amiens, Amiens, France; 15 CHU Angers, Angers, France; 16 CHU Montpellier, Montpellier, France and 17 Hôpital Saint-Louis, AP-HP, Université Paris 7, Paris, France E-Mail: [email protected] Leukemia (2014) 2395 – 2424


2424

REFERENCES 1 Tallman MS, Andersen JW, Schiffer CA, Appelbaum FR, Feusner JH, Ogden A et al. All-trans-retinoic acid in acute promyelocytic leukemia. N Engl J Med 1997; 337: 1021–1028. 2 Ades L, Chevret S, Raffoux E, de Botton S, Guerci A, Pigneux A et al. Is cytarabine useful in the treatment of acute promyelocytic leukemia? Results of a randomized trial from the European Acute Promyelocytic Leukemia Group. J Clin Oncol 2006; 24: 5703–5710. 3 Sanz MA, Montesinos P, Rayon C, Holowiecka A, de la Serna J, Milone G et al. Risk-adapted treatment of acute promyelocytic leukemia based on all-trans retinoic acid and anthracycline with addition of cytarabine in consolidation therapy for high-risk patients: further improvements in treatment outcome. Blood 2010; 115: 5137–5146. 4 Lo Coco F, Avvisati G, Vignetti M, Fioritoni G, Liso V, Ferrara F et al. Front-line treatment of acute promyelocytic leukemia with AIDA induction followed by risk-adapted consolidation: results of the AIDA-2000 Trial of the Italian GIMEMA Group. blood 2004; 104: 392. 5 Montesinos P, Bergua JM, Vellenga E, Rayon C, Parody R, de la Serna J et al. Differentiation syndrome in patients with acute promyelocytic leukemia treated with all-transretinoic acid and anthracycline chemotherapy: characteristics, outcome and prognostic factors. Blood 2008; 113: 775–783. 6 De Botton S, Dombret H, Sanz M, Miguel JS, Caillot D, Zittoun R et al. Incidence, clinical features, and outcome of all trans-retinoic acid syndrome in 413 cases of

Leukemia (2014) 2395 – 2424

7

8

9

10

11

12

newly diagnosed acute promyelocytic leukemia. The European APL Group. Blood 1998; 92: 2712–2718. Park JH, Qiao B, Panageas KS, Schymura MJ, Jurcic JG, Rosenblat TL et al. Early death rate in acute promyelocytic leukemia remains high despite all-trans retinoic acid. Blood 2011; 118: 1248–1254. Lehmann S, Ravn A, Carlsson L, Antunovic P, Deneberg S, Mollgard L et al. Continuing high early death rate in acute promyelocytic leukemia: a populationbased report from the Swedish Adult Acute Leukemia Registry. Leukemia 2011; 25: 1128–1134. McClellan JS, Kohrt HE, Coutre S, Gotlib JR, Majeti R, Alizadeh AA et al. Treatment advances have not improved the early death rate in acute promyelocytic leukemia. Haematologica 2012; 97: 133–136. Altman JK, Rademaker A, Cull E, Weitner BB, Ofran Y, Rosenblat TL et al. Administration of ATRA to newly diagnosed patients with acute promyelocytic leukemia is delayed contributing to early hemorrhagic death. Leuk Res 2013; 37: 1004–1009. Kelaidi C, Chevret S, De Botton S, Raffoux E, Guerci A, Thomas X et al. Improved outcome of acute promyelocytic leukemia with high WBC counts over the last 15 years: the European APL Group experience. J Clin Oncol 2009; 27: 2668–2676. Paulson K, Serebrin A, Lambert P, Bergeron J, Seftel Met al. Acute promyelocytic leukaemia is characterized by stable incidence and improved survival that is restricted to patients managed in leukaemia referral centres: a pan-Canadian epidemiological study. Br J Haematol 2014; 166: 660–666.


Evolving characteristics and outcome of secondary acute promyelocytic leukemia (APL): A prospective analysis by the French-Belgian-Swiss APL group.

Acute Promyelocytic Leukemia (APL): Comparison Between Children and Adults.

Clinical characteristics of acute promyelocytic leukemia manifesting as early death.

JSH guideline for tumors of hematopoietic and lymphoid tissues: leukemia: 2. Acute promyelocytic leukemia (APL).

RARα bcr Isoforms in Indian Acute Promyelocytic Leukemia (APL) Patients and Clinicohematological Correlation.

ATRA syndrome within the scope of therapy of acute promyelocytic leukemia (APL)].

ATRA) for Acute Promyelocytic Leukemia (APL).

Disparities in early death and survival in children, adolescents, and young adults with acute promyelocytic leukemia in California.

CD11b expression in acute promyelocytic leukemia.

Thrombo-hemorrhagic deaths in acute promyelocytic leukemia.

Acute promyelocytic leukemia in children and adolescents.

Resistance to therapy in acute promyelocytic leukemia.

Nephrotic syndrome in acute promyelocytic leukemia.

Thrombotic events in acute promyelocytic leukemia.

Prompt all-trans retinoic acid administration and early death in acute promyelocytic leukemia: what does the evidence say?

Targeting of leukemia-initiating cells in acute promyelocytic leukemia.

Methodological aspects of minimal residual disease assessment by flow cytometry in acute lymphoblastic leukemia: A French multicenter study.

Faggot cells in a non-promyelocytic acute myeloid leukemia.

Pneumatosis Intestinalis in a Patient with Acute Promyelocytic Leukemia.

Trisomy 13 in a case of acute promyelocytic leukemia.

Late differentiation syndrome in acute promyelocytic leukemia: a challenging diagnosis.

RARA fusion genes in acute promyelocytic leukemia: a review.

Acute appendicitis secondary to acute promyelocytic leukemia.

Activation of a promyelocytic leukemia-tumor protein 53 axis underlies acute promyelocytic leukemia cure.