Page 93 - 2019_03-Haematologica-web

RED score and hepcidin:ferritin in lower-risk MDS
(RCUD), refractory anemia with excess blasts-1 (RAEB-1), del 5q syndrome], or chronic myelomonocytic leukemia-1 (CMML-1) with a white blood cell count <13x109/L, (ii) low/intermediate (int)-1 risk according to the IPSS determined locally, (iii) hemoglo- bin <10 g/dL or RBC transfusion dependence, (iv) Eastern Cooperative Oncology Group performance status <2 and (v) sEPO level <500 IU/L. Exclusion criteria were: non-controlled hyperten- sion, or cardiovascular disease (uncontrolled angina pectoris, heart failure), renal insufficiency, sEPO level >500 IU/L, systemic infec- tion or chronic inflammatory disease, serum folate <2 ng/mL or vitamin B12 <200 pg/mL, and other non-MDS-related causes of anemia (e.g., hemolysis, hemorrhage, iron deficiency). All patients gave their written informed consent to biological investigations according to the recommendations of the local ethics committee (Comité de Protection des Personnes Paris V, CPP n. RCB 212- A01395-38, EUDRACT 2012-002990-7338) and the study was conducted in accordance with the Helsinki Declaration and regis- tered in ClinicalTrials.gov as NCT03598582.
Treatment
Patients received subcutaneous epoetin zeta 40 000 IU/week for 12 weeks. Response was evaluated after 12 weeks of treatment according to International Working Group 2006 criteria. Non- responders were excluded from the study while responsive patients continued on epoetin zeta for another 52 weeks. Patients still responding at week 52 could continue treatment, based on the physician’s decision. If hemoglobin levels exceeded 12 g/dL at any time before week 12, the dose of epoetin zeta was reduced to 20 000 IU/week. After week 12, the intervals between injections were increased by 1 week if hemoglobin levels exceeded 13 g/dL. The purpose of this dose adjustment was to reach epoetin zeta doses allowing hemoglobin levels to be maintained between 11 and 12 g/dL. During the dose adjustment period, weekly blood counts were performed. No prescription of iron was allowed in this trial, in order not to perturb iron metabolism markers. Each patient had a minimal follow-up of 52 weeks.
Biological endpoints
The primary endpoint of this study was to find new biomarkers capable of predicting the response to epoetin zeta.
Bone marrow aspirates were collected from all 70 patients at inclusion and then after 12 weeks. No samples were excluded based on clinical parameters. Fresh bone marrow aspirates were sent to Cochin Hospital, Paris, for centralized flow cytometry analysis of dyserythropoiesis, using the RED score, and gene sequencing. Ogata scores were also assessed locally in the hospi- tals of Mulhouse, Creteil, Tours, Grenoble and Cochin. Patients were re-evaluated at week 12 by flow cytometry using both the RED and Ogata scores, assessed centrally in Cochin Hospital.10
Blood plasma was also collected for quantitative analyses of hepcidin and GDF-15. Hepcidin levels from plasma samples col- lected in EDTA were measured by liquid chromatography coupled to tandem mass spectrometry in Louis Mourier Hospital using the method described by Lefebvre et al.17 The results are expressed as hepcidin:ferritin (x100) ratios which represent a measure of the adequacy of hepcidin level relative to iron body stores because hepcidin levels are known to be modulated by transfusion and inflammation. All patients, except one, had C-reactive protein val- ues <5 mg/L. GDF-15 was measured by enzyme-linked immunosorbent assay at Cochin hospital, using kits obtained from R&D Systems (Minneapolis, MN, USA).18
Genomic studies and bioinformatic analysis
Mononuclear cells from bone marrow aspirates were purified on a Ficoll gradient. Cell pellets were further processed for DNA
extraction using the DNA/RNA Kit (Qiagen, Hilden, Germany). All 70 samples were screened for mutations in a panel of 26 genes (ASXL1, CBL, DNMT3A, ETV6, EZH2, FLT3, IDH1, IDH2, JAK2, KIT, KRAS, NRAS, MPL, NPM1, PHF6, PTPN11, RIT1, RUNX1, SETBP1, SF3B1, SRSF2, TET2, TP53, U2AF1, WT1 and ZRSR2) by a next-generation sequencing assay using the Ion AmpliSeqTM library kit 2 384 n. 4480442 (Life Technologies, Chicago, IL, USA). All the samples were also screened for ASXL1 (including c.1934dupG; p.G646WfsX12) and SRSF2 mutations by Sanger sequencing. JAK2, NPM1 and FLT3-ITD mutations were analyzed by real-time polymerase chain reaction and fluorescent poly- merase chain reaction to confirm the next-generation sequencing data. Bioinformatic analysis was performed as previously described.19
Sample size justification and statistical analysis
The sample size was computed based on an expected treatment response rate of 50% to 60%, i.e. about 30 responders and 30 non- responders. Allowing for 10%-15% of the biological data being non-evaluable, 70 patients had to be included.
For continuous variables, values were expressed as medians and interquartile ranges (IQR) or means and compared using the Wilcoxon test. Categorical variables are reported as counts or per- centages with 95% confidence intervals (95% CI) and were com- pared using a Fisher exact or chi-squared test. Evaluation of the erythroid response (according to International Working Group 2006 criteria) after 12 weeks of treatment, i.e. the efficacy end- point, was considered as a binary variable (response versus non- response). The search for factors predictive of response was per- formed by logistic regression and results were presented as odds ratio (OR) with 95% CI. P values <0.05 were considered statisti- cally significant. Receiving operating characteristic (ROC) curves were used to determine the best thresholds of the prognostic fac- tors. All analyses were performed using JMP® software (version 10, SAS Institute Inc, Cary, NC, USA).
Results
Patients’ baseline characteristics
Seventy patients (31 males and 39 females) were recruit- ed in 16 French centers between January 2013 and March 2017. Their median age was 78 years (range, 57-93 years). At inclusion, the WHO classification of the patients was the following: 22 RCMD, 19 RCUD, 14 RARS, four RAEB- 1, six CMML, two del 5q-, and three MDS-unclassifiable. The IPSS classification was low in 43 (61.5%) and int-1 in 27 (38.5%) patients, whereas the IPSS-R classification was very low in 13 (18.5%), low in 47 (67%), intermediate in nine (13%) and high in one (1.4%) patient. Twenty (28%) patients were dependent on RBC transfusions (median >2 RBC transfusions/8 weeks) receiving a cumulated number of RBC concentrates ranging from 2 to 7 (median 3) (Table 1). Before treatment, next-generation sequencing was per- formed in 68 patients. The most frequent mutations involved SF3B1 (n=28), TET2 (n=20), ASXL1 (n=15), SRSF2 (n=9), DNMT3A (n=9), U2AF1 (n=7), IDH1/2 (n=6), and EZH2 (n=6) genes. One, two, three, four or five muta- tions were detected in 12, 12, 26, 10, seven and one patients, respectively, and 26% of patients had more than two mutations (Online Supplementary Figure S1).
Efficacy and safety of epoetin zeta
Overall the HI-E was 47.6% (33 patients). Among the 20 transfusion-dependent patients, eight obtained HI-E
haematologica | 2019; 104(3)
499