Genetic risk of prediabetes and diabetes development in chronic myeloid leukemia patients treated with nilotinib

Genetic risk of prediabetes and diabetes development in chronic myeloid leukemia patients treated with nilotinib

Accepted Manuscript Title: Genetic risk of prediabetes and diabetes development in chronic myeloid leukemia patients treated with nilotinib Author: Br...

781KB Sizes 0 Downloads 49 Views

Accepted Manuscript Title: Genetic risk of prediabetes and diabetes development in chronic myeloid leukemia patients treated with nilotinib Author: Bruno Martino, Corrado Mammì, Claudia Labate, Silvia Rodi, Domenica Ielo, Manuela Priolo, Maurizio Postorino, Giovanni Tripepi, Francesca Ronco, Carmelo Laganà, Caterina Musolino, Marianna Greco, Giorgio La Nasa, Giovanni Caocci PII: DOI: Reference:

S0301-472X(17)30663-X http://dx.doi.org/doi: 10.1016/j.exphem.2017.07.007 EXPHEM 3563

To appear in:

Experimental Hematology

Received date: Revised date: Accepted date:

29-3-2017 19-7-2017 20-7-2017

Please cite this article as: Bruno Martino, Corrado Mammì, Claudia Labate, Silvia Rodi, Domenica Ielo, Manuela Priolo, Maurizio Postorino, Giovanni Tripepi, Francesca Ronco, Carmelo Laganà, Caterina Musolino, Marianna Greco, Giorgio La Nasa, Giovanni Caocci, Genetic risk of prediabetes and diabetes development in chronic myeloid leukemia patients treated with nilotinib, Experimental Hematology (2017), http://dx.doi.org/doi: 10.1016/j.exphem.2017.07.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Brief Communications

Genetic risk of prediabetes and diabetes development in chronic myeloid leukemia patients treated with nilotinib Bruno Martino1, Corrado Mammì2, Claudia Labate2, Silvia Rodi2, Domenica Ielo1, Manuela Priolo2, Maurizio Postorino3, Giovanni Tripepi4, Francesca Ronco1, Carmelo Laganà2, Caterina Musolino5, Marianna Greco6, Giorgio La Nasa6, Giovanni Caocci6 1

Operative Unit of Hematology, Grande Ospedale Metropolitano “Bianchi-Melacrino-

Morelli”, Reggio Calabria, Italy; 2

Operative Unit of Medical Genetics, Grande Ospedale Metropolitano “Bianchi-Melacrino-

Morelli”, Reggio Calabria, Italy; 3

Operative Unit of Nephrology, Grande Ospedale Metropolitano “Bianchi-Melacrino-

Morelli”, Reggio Calabria, Italy; 4

CNR-IBIM, Epidemiology Research Unit and Clinical Pathophysiology of Renal Disease

and Hypertension, Reggio Calabria, Italy 5

Operative Unit of Hematology, Azienda Ospedaliero Universitaria, University of Messina,

Messina, Italy 6

Department of Medical Sciences and Public Health, University of Cagliari, Cagliari, Italy

Correspondence to: Giovanni Caocci Department of Medical Sciences and Public Health, University of Cagliari, “R. Binaghi” Hospital Via Is Guadazzonis, 3 09126 Cagliari, Italy Tel. ++390-70-6092800 Fax. ++390-70-6092936 E-mail: [email protected] Running Title: Insulin resistance in nilotinib-treated CML patients Keywords: Genetic Risk Score, Chronic Myeloid Leukemia, Diabetes, Nilotinib Abstract word count: 229 Text word count: 1588 Tables: 1 Figures: 2 References: 29

1 Page 1 of 15

Graphical Abstract

Highlights 

Prediabetes and diabetes represent adverse events in CML treated with nilotinib.



uGRS < 10 predicts higher prediabetes/diabetes free survival after nilotinib



In clinical practice uGRS could help tailor the best TKI therapy

Abstract Impaired fasting glucose (IFG) and type 2 diabetes (T2D) represents adverse events in Chronic Myeloid Leukemia (CML) patients treated with the second-generation tyrosine kinase inhibitor (TKI) nilotinib. A genetic risk score (uGRS) for the prediction of insulin 2 Page 2 of 15

resistance, consisting of 10 multiple single-nucleotide polymorphisms (SNPs), has been proposed. We evaluated the uGRS predictivity in 61 CML patients treated with nilotinib. Patients were genotyped for IRS1, GRB14, ARL15, PPARG, PEPD, ANKRD55/MAP3K1, PDGFC, LYPLAL1, RSPO3, and FAM13A1 genes. The uGRS was based on the sum of the risk alleles within the set of selected SNPs. Molecular response (MR)3.0 and MR4.0 were achieved in 90% and 79% of the patients, respectively. Before treatment, none of the patients had abnormal blood glucose. During treatment and subsequently follow-up of 80.2 months (range 1-298), 7 patients (11.5%) developed diabetes requiring oral treatment, after a median of 14 months (range 3-98) since nilotinib. Twelve patients (19.7%) developed prediabetes. Prediabetes/diabetes-free survival was significantly higher in patients with an uGRS below 10 compared to higher scores (100% vs 22.8±12.4%, p<0.001). Each increment of 1 unit on the uGRS caused a 42% increase in the prediabetes/diabetes risk (HR=1.42; CI: 1.04-1.94; p=0.026). The presence of more than 10 allelic variants associated to insulin secretion, processing, sensitivity and clearance is predictive of prediabetes/diabetes developing in CML patients treated with nilotinib. In clinical practice uGRS could help tailor the best TKI therapy.

3 Page 3 of 15

Introduction Nilotinib is a second-generation tyrosine kinase inhibitor (TKI) effective in the treatment of chronic myeloid leukemia (CML), with a toxicity profile characterized by several metabolic alterations, including dyslipidemia, metabolic syndrome and insulin resistance, the latter mainly represented by impaired fasting glucose (IFG) and diabetes mellitus (DM).1,2 The pathogenesis of glucose metabolic alterations is poorly understood, considering that other TKIs as imatinib and dasatinib seem to improve glucose levels and have been investigated for treatment of Type 2 DM (T2D).3,4 The higher fasting insulin levels and inadequate endogenous insulin (C-peptide) secretory capacity observed in nilotinibtreated CML patients would seem associated with insulin resistance.5-7 Type 2 diabetes is characterized by impaired insulin secretion by pancreatic beta cells and peripheral insulin resistance. Genome-wide association (GWA) studies have investigated the relationship between T2D risk and susceptibility loci variants involved in insulin secretion, proinsulin processing, insulin sensitivity and clearance.8-10 Scott et al. identified 10 multiple single-nucleotide polymorphisms (SNPs) associated with insulin resistance and validated an unweighted genetic risk score (uGRS) for prediction of insulin resistance and T2D development.11 According to European LeukemiaNet recommendations, patients with T2D or prediabetes may receive nilotinib, but with strict monitoring of fasting glucose and glycosylated haemoglobin (HbA1c).12 In clinical practice, it would be useful to identify CML patients at risk of developing treatment-related glycometabolic side effects. We evaluated the ability of the uGRS to predict insulin resistance in CML patients treated with nilotinib.

4 Page 4 of 15

Methods From July 2015 to June 2016, 61 CML patients attending the Operative Unit of Hematology of the “Bianchi-Melacrino-Morelli” Hospital in Reggio Calabria and the “Binaghi” Hospital in Cagliari were enrolled in our study. All patients had received first- or second-line nilotinib treatment. Written informed consent was obtained from all patients. Major Molecular response (MR3.0) was assessed according to the International Scale (IS). Complete molecular response (CMR) was defined as undetectable BCR-ABL by Real-time PCR (RT-PCR) with an assay sensitivity of MR4.0. At baseline and during treatment all patients were evaluated for fasting plasma glucose (FPG), fasting insulin, HbA1c serum levels, serum C-peptide, HDL and LDL cholesterol, triglycerides and Homeostasis Model Assessment – Insulin Resistance (HOMA-IR) values.13,14 Oral glucose tolerance tests (OGTT) were performed in patients with FPG values between 100 and 110 mg/dl (5.56 mmol/L and 6.11 mmol/L).

Glucose abnormalities and TD2 diagnosis were evaluated

according to American Diabetes Association (ADA) criteria: prediabetes (IFG) was diagnosed by FPG between 100-125 mg/dl (5.6-6.9 mmol/L) or 2 hour-OGTT between 140199 mg/dl (7.8-11.0 mmol/L); diabetes was diagnosed by FPG ≥ 126 mg/dl (7.0 mmol/L) or 2 hour-OGTT ≥ 200 mg/dl (11.1 mmol/L).15 To evaluate the combined effects of 10 SNPs associated with insulin resistance in GWA studies, we used the additive model described by Scott and calculated the uGRS11. Variants included in the score were in or near the IRS1, GRB14, ARL15, PPARG, PEPD, ANKRD55/MAP3K1, PDGFC, LYPLAL1, RSPO3, and FAM13A1 genes. Individual genetic risk scores were based on the number of accumulated risk alleles that each individual had within the set of selected genetic variants. Each individual might have 0, 1, or 2 risk alleles in each of the variants and was assigned a total score from 0 to 20. The effect of the uGRS on the outcome variable, was investigated by Kaplan-Meier survival analysis,

5 Page 5 of 15

univariate and multivariate (age-adjusted) Cox regression. The prognostic power of the uGRS was established using receiver operating characteristic (ROC) curve analysis.

Results and discussion The characteristics of the 61 chronic phase CML patients are shown in Table 1. Thirty-six patients were treated with frontline nilotinib 300 mg twice daily. Twenty-five patients underwent second-line treatment with nilotinib 300 mg twice daily because of imatinib failure (13 patients) or intolerance (12 patients). The median follow-up was 80.2 months (range 1-298). MR3.0 and CMR were achieved in 90% and 79% of the patients, respectively. None of the patients had a glycemic value above the cut-off of 110 mg/dl before starting treatment with nilotinib. Seven patients (11.5%) developed overt diabetes requiring oral treatment: 3 after treatment with imatinib and 4 during frontline treatment with nilotinib. Onset of diabetes occurred at a median of 14 months (range 3-98) after starting treatment. Fourteen patients (23%) - 8 receiving frontline nilotinib and 6 receiving second-line nilotinib - developed IFG according to ADA criteria. Only one of these patients developed overt diabetes requiring treatment. The other 14 patients remained border line throughout the follow-up. When using the OGTT, 12 patients (19.7%) developed prediabetes in a period ranging from 6 to 95 months (median 10 months): 5 patients during frontline treatment and 2 during second-line treatment. None of these patients developed overt diabetes during the follow-up. No difference in prediabetes/diabetes risk was found following frontline or second-line nilotinib treatment. The uGRS ranged from a minimum value of 5/20 to a maximum value of 14/20. FPG, fasting insulin, HBA1c, C-peptide, HDL and LDL cholesterol, triglycerides, HOMA-IR, familiarity for diabetes were not associated with high values of uGRS or the development of diabetes. None of the patients with an uGRS score below 10 presented OGTT results

6 Page 6 of 15

indicative of pre-diabetes or diabetes (Figure 1). Prediabetes/diabetes-free survival was significantly higher in CML patients with an uGRS below 10 compared to patients with higher scores (100% vs 22.8±12.4%, p<0.001) (Figure 2). Cox regression analysis showed that each increment of 1 unit on the uGRS caused a 42% increase in the risk of developing prediabetes/diabetes (HR=1.42; CI: 1.04-1.94; p=0.026). In a multivariate Cox model including age and the uGRS, only the genetic risk score maintained a significant association with the onset of prediabetes/diabetes (p=0.025). A ROC curve was generated to measure the prognostic accuracy of the uGRS. The area under the curve (AUC) resulted to be 0.80 (CI: 0.65-0.92), demonstrating a high discriminatory power of the test. For exploratory purpose we calculated the uGRS in an additional small cohort of 12 CML patients treated with imatinib (median follow up 86 months, range 44-112), without metabolic alterations. The median uGRS was 9±2. While nilotinib and dasatinib are generally associated with improved glycemic control, nilotinib has been shown to variably impair glucose metabolism.3,4 In phase II second-line trials of patients treated with nilotinib 400mg twice daily, following resistance or intolerance to imatinib, the incidence of all grade hyperglycemia was 70% and of grade 3-4 hyperglycemia 12%.16 In the Enact trial of 1422 CML patients treated with second-line nilotinib, 0.1% of them showed grade 3-4 hyperglycemia suspected to be drug-related.17 In phase III studies of frontline treatment, hyperglycemia was observed in almost 50% of the patients treated with nilotinib 300 mg twice daily; 20% developed diabetes within 3 years but relatively few patients required anti-diabetic treatment.5,18

A high rate (11%) of grade 3-4

hyperglycemia was observed in the recent GIMEMA trial,19 which overlaps well with the results of real life data sets.2,20,21 Few studies have evaluated the safety of nilotinib therapy in CML patients with pre-existing type 2 diabetes (T2D).6,19,22 Within the ENESTnd trial, 41 nilotinib-treated patients reported T2D at baseline; 68% of these patients were on anti-

7 Page 7 of 15

diabetic medication at study entry. Overall, hyperglycemia occurring during treatment was mild, transient and manageable. Only 5 patients (12.2%) had to interrupt nilotinib because of adverse events related to diabetes.22 Multiple mechanisms are involved in the pathogenesis of the glycometabolic changes observed after TKI treatment. In animal models of T2D, imatinib has shown to reduce insulin resistance through reduction in endoplasmic reticulum stress markers and c-jun N-terminal kinase (JNK) activity, that increase insulin resistance in peripheral tissues.23-25 Stimulation of glucose uptake following the expression of BCR-ABL tyrosine kinase appears to have an important role in the suppression of apoptosis and inappropriate survival of CML cells. 26 Downregulation of BCR-ABL-stimulated glucose transport by imatinib is mediated by relocation of glucose transporter proteins from the cell surface to interior compartments, suggesting a possible therapeutic effect on inappropriate cancer cell survival.27 The mechanism by which nilotinib induces hyperglycemia remains unknown. Given its superior selectivity for BCR-ABL with respect to imatinib and dasatinib, it may not inhibit PDGF, JNK and c-KIT pathways. Moreover, it causes inadequate secretion of endogen insulin which is reversible after discontinuation of therapy6. Whether nilotinib-induced glycometabolic alterations are to be considered a true clinical problem remains a matter of debate. The incidence of diabetes requiring anti-diabetic drugs in nilotinib clinical trials is relatively small, ranging from 1.8% to 5.3%. Hyperglycemia occurring during nilotinib treatment in patients with pre-existing TD2 is usually mild, transient and manageable.22 In our study, diabetes developed in 11.5% of the patients after a median of 14 months (range 3-98) from starting treatment, and was easily managed. When we compared the onset of diabetes with a general population of the same region and age group, no statistically significant differences emerged.28 Nevertheless, the increasing number of older patients affected by CML worldwide points to an urgent need for

8 Page 8 of 15

parameters capable of individuating patients at risk for developing nilotinib-related prediabetes or diabetes. Over the past few years, GWA studies have been widely used to identify genetic risk factors for complex diseases. Scott et al. developed a genetic score for the prediction of insulin resistance independent of body size.8-10, 29 Our results show that the presence of more than 10 allelic variants associated to insulin secretion, processing, sensitivity and clearance is strongly predictive of diabetes. Patients with an uGRS score below 10 remained in a state of prediabetes/diabetes-free (figure 2) and each increment of 1 unit on the uGRS correlated with a 42% increased risk of developing prediabetes/diabetes. In a small cohort of imatinib long term treated patients, without metabolic alterations, we found a median value of uGRS = 9±2; imatinib might improve glycemic control in those patients with a genetic predisposition to prediabetes/diabetes development (uGRS≥10).3,4 CML patients with prediabetes or diabetes must be stringently monitored during nilotinib treatment.12

Moreover, the uGRS could provide additional information in those

patients undergoing nilotinib treatment who present low or moderate cardiovascular risk. Further perspective studies of larger populations are needed to confirm the predictive power of the uGRS, which in clinical practice can help tailor the best anti-TKI therapy.

DECLARATIONS Competing Statement: The authors have no conflicts of interest to disclose Acknowledgements We are deeply grateful to the patients who participated in this study. We also wish to thank Anna Maria Koopmans for professional writing assistance. Funding: none

9 Page 9 of 15

Authors contributions: Conception and design: MB Collection and assembly of data: MB, CM, CL, SR, DI, MP, MP, FR, CL, CM, MG, GLN; Statistical analysis: GT; Manuscript writing: GC, BM, Final approval of manuscript: MB, CM, CL, SR, DI, MP, MP, GT, FR, CL, CM, MG, GLN, GC

References 1) Golemovic M, Verstovsek S, Giles F, Cortes J, Manshouri T, Manley PW, et al. AMN107, a novel aminopyrimidine inhibitor of Bcr-Abl, has in vitro activity against imatinib-resistant chronic myeloid leukemia. Clin Cancer Res. 2005;11(13):4941-7. 2) Iurlo A, Orsi E, Cattaneo D, Resi V, Bucelli C, Orofino N, et al. Effects of first- and second-generation tyrosine kinase inhibitor therapy on glucose and lipid metabolism in chronic myeloid leukemia patients: a real clinical problem? Oncotarget. 2015;6(32):3394451. 3) Malek R, Davis SN. Tyrosine kinase inhibitors under investigation for the treatment of type II diabetes. Expert Opin Investig Drugs. 2016;25(3):287-96. 4) Prada PO, Saad MJ. Tyrosine kinase inhibitors as novel drugs for the treatment of diabetes. Expert Opin Investig Drugs. 2013;22(6):751-63. 5) Rea D, Gautier JF, Breccia M, Saglio G, Hughes TP, Kantarjian HM et al. Incidence of Hyperglycemia by 3 Years in Patients (Pts) with Newly Diagnosed Chronic Myeloid Leukemia in Chronic Phase (CML-CP) Treated with Nilotinib (NIL) or Imatinib (IM) in ENESTnd. Blood 2012; 120: 168. 6) Ito Y, Miyamoto T, Chong Y, Maki T, Akashi K, Kamimura T. Nilotinib exacerbates diabetes mellitus by decreasing secretion of endogenous insulin. Int J Hematol. 2013;97(1):135-8.

10 Page 10 of 15

7) Zdenek R, Belohlavkova P, Cetkovsky P, Faber E., Klamova H, Ludmila M, et al. Comparison of Glucose and Lipid Metabolism Abnormality during Nilotinib, Imatinib and Dasatinib Therapy - Results of Enigma 2 Study. Blood. 2014;124:1813. 8) Strawbridge RJ, Dupuis J, Prokopenko I, Barker A, Ahlqvist E, Rybin D, et al. Genomewide association identifies nine common variants associated with fasting proinsulin levels and provides new insights into the pathophysiology of type 2 diabetes. Diabetes. 2011;60(10):2624-34. 9) Manning AK, Hivert MF, Scott RA, Grimsby JL, Bouatia-Naji N, Chen H, et al. A genome-wide approach accounting for body mass index identifies genetic variants influencing fasting glycemic traits and insulin resistance. Nat Genet. 2012;44(6):659-69. 10) Dimas AS, Lagou V, Barker A, Knowles JW, Mägi R, Hivert MF, et al. Impact of type 2 diabetes susceptibility variants on quantitative glycemic traits reveals mechanistic heterogeneity. Diabetes. 2014;63(6):2158-71. 11) Scott RA, Fall T, Pasko D, Barker A, Sharp SJ, Arriola L, et al. Common genetic variants highlight the role of insulin resistance and body fat distribution in type 2 diabetes, independent of obesity Diabetes. 2014;63(12):4378-87. 12) Steegmann JL, Baccarani M, Breccia M, Casado LF, García-Gutiérrez V, Hochhaus A, et al. European LeukemiaNet recommendations for the management and avoidance of adverse events of treatment in chronic myeloid leukaemia. Leukemia. 2016;30(8):1648-71. 13) Levy JC, Matthews DR, Hermans MP. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care. 1998; 21: 2191-2192. 14) Gutt M, Davis CL, Spitzer SB, Llabre MM, Kumar M, Czarnecki EM, et al. Validation of the insulin sensitivity index (ISI(0,120)): comparison with other measures. Diabetes Res Clin Pract. 2000; 47: 177-184.

11 Page 11 of 15

15) American Diabetes Association. Classification and diagnosis of diabetes. Diabetes Care. 2015;38 Suppl:S8-S16. 16) Kantarjian HM, Giles FJ, Bhalla KN, Pinilla-Ibarz J, Larson RA, Gattermann N, et al. Nilotinib is effective in patients with chronic myeloid leukemia in chronic phase after imatinib resistance or intolerance: 24-month follow-up results. Blood. 2011;117(4):1141-5. 17) Nicolini FE, Turkina A, Shen ZX, Gallagher N, Jootar S, Powell BL, et al. Expanding Nilotinib Access in Clinical Trials (ENACT): an open-label, multicenter study of oral nilotinib in adult patients with imatinib-resistant or imatinib-intolerant Philadelphia chromosome-positive

chronic

myeloid

leukemia

in

the

chronic

phase.

Cancer.

2012;118(1):118-26. 18) Larson RA, Hochhaus A, Hughes TP, Clark RE, Etienne G, Kim DW, et al. Nilotinib vs imatinib in patients with newly diagnosed Philadelphia chromosome-positive chronic myeloid

leukemia

in

chronic

phase:

ENESTnd

3-year

follow-up.

Leukemia.

2012;26(10):2197-203. 19) Castagnetti F, Breccia M, Gugliotta G, Martino B, D'Adda M, Stagno F, et al; GIMEMA CML Working Party. Nilotinib 300 mg twice daily: an academic single-arm study of newly diagnosed

chronic

phase

chronic

myeloid

leukemia

patients.

Haematologica.

2016;101(10):1200-1207. 20) Breccia M, Loglisci G, Salaroli A, Serrao A, Alimena G. Nilotinib-mediated increase in fasting glucose level is reversible, does not convert to type 2 diabetes and is likely correlated with increased body mass index. Leuk Res. 2012;36(4):e66-7. 21) Breccia M, Muscaritoli M, Gentilini F, Latagliata R, Carmosino I, Rossi Fanelli F, et al. Impaired fasting glucose level as metabolic side effect of nilotinib in non-diabetic chronic myeloid leukemia patients resistant to imatinib. Leuk Res. 2007;31(12):1770-2.

12 Page 12 of 15

22) Saglio G, Larson RA, Hughes TP, Issaragrisil S, Turkina AG, Marin D et al. Efficacy and safety of nilotinib in chronic phase (CP) chronic myeloid leukemia (CML) patients (Pts) with type 2 diabetes in the ENESTnd trial. Blood 2010;116:3430. 23) Han MS, Chung KW, Cheon HG, Rhee SD, Yoon CH, Lee MK, et al. Imatinib mesylate reduces endoplasmic reticulum stress and induces remission of diabetes in db/db mice. Diabetes. 2009;58:329-36. 24) Hägerkvist R, Jansson L, Welsh N. Imatinib mesylate improves insulin sensitivity and glucose disposal rates in rats fed a high-fat diet. Clin Sci (Lond). 2008;114:65-71. 25) Hirosumi J, Tuncman G, Chang L, Görgün CZ, Uysal KT, Maeda K, et al. A central role for JNK in obesity and insulin resistance. Nature. 2002;420:333-6. 26) Barnes K, McIntosh E, Whetton AD, Daley GQ, Bentley J, Baldwin SA. Chronic myeloid leukaemia: an investigation into the role of Bcr-Abl-induced abnormalities in glucose transport regulation. Oncogene. 2005;24:3257-3267. 27) Bentley J, Walker I, McIntosh E, Whetton AD, Owen-Lynch PJ, Baldwin SA. Glucose transport regulation by p210 Bcr-Abl in a chronic myeloid leukaemia model. Br J Haematol. 2001;112:212-215. 28) http://www.ibdo.it/pdf/DiabetesMonitor-FactandFigure2014.pdf accessed on February 10th, 2017 29) Langenberg C, Sharp SJ, Schulze MB, Rolandsson O, Overvad K, Forouhi NG et al. Long-term risk of incident type 2 diabetes and measures of overall and regional obesity: the EPIC-InterAct case-cohort study. InterAct Consortium. PLoS Med. 2012;9(6):e1001230

13 Page 13 of 15

Figure 1. Distribution of risk alleles in CML patients according to the results of oral glucose tolerance tests (OGTT) and the unweighted genetic risk score (uGRS). No patient with uGRS values below 10 had OGTT results indicative of pre-diabetes or diabetes.

Figure 2. Prediabetes/diabetes-free survival according to genetic risk score (uGRS) score in CML patients following nilotinib treatment.

14 Page 14 of 15

Table 1: Characteristics and treatment of 61 patients with chronic myeloid leukemia in the chronic phase. Age at diagnosis, mean years (range) Sex, N° (%) Male Female First line Nilotinib 300 mg BID, N° (%) Second line Nilotinib 300 mg BID, N° (%) Imatinib failure Imatinib intollerance Overall follow-up from diagnosis, mean months (range) Follow-up Nilotinib front line, mean months (range) Follow-up Nilotinib second line, mean months (range) Sokal Risk, N° (%) low intermediate high Metabolic features FPG (mg/dl), mean (range) Insulin (umol/L), mean (range) HbA1c (%), mean (range) C-peptide (umol/L), mean (range) HDL cholesterol (mg/dl), mean (range) LDL cholesterol (mg/dl), mean (range) Triglycerides (mg/dl), mean (range) HOMA-IR, mean (range) Patients developing diabetes during treatment, N° (%) Time in appearance of diabetes from Nilotinib start, median months (range) Patients developing prediabetes during treatment, N° (%) Time in appearance of pre-diabetes from Nilotinib start, median months (range)

54

(18 - 84)

30 31 36 25 13 12 80.2 42.4 66.3

(49.2) (50.8) (59) (41) (21.3) (19.7) (1-298) (1-104) (11-131)

24 16 21 Baseline 86.4 (62-109) 7.7 (1.1-69) 5.3 (4.2-5.9) 2.4 (0.60-7.5) 54.3 (20-110) 97.4 (50-155) 142 (58-404) 2.8 (0.4-24.8) 7 14

(39) (26) (35) Follow up 95.4 (66-186) 7.6 (1.2-77) 5.6 (4.6-7.3) 2.6 (0.5-7.7) 58.4 (27-99) 72.5 (50.4-199) 101.6 (42-475) 3.2 (0.7-25.2) (11.5) (3-98)

12

(19.7)

10

(6-95)

FPG = fasting plasma glucose; HbA1c = serum glycosylated hemoglobin, HOMA-IR= Homeostasis Model Assessment – Insulin Resistance.

15 Page 15 of 15