Evidence-Based Approach to Managing Non-Transfusion-Dependent Thalassemias (NTDT)

Transcription

1 Evidence-Based Approach to Managing Non-Transfusion-Dependent Thalassemias (NTDT) Ali T. Taher, MD, PhD, FRCP Professor of Medicine, Hematology & Oncology American University of Beirut Medical Center Beirut Lebanon Professor of Hematology & Medical Oncology (Adj.) Emory School of Medicine Atlanta USA

2 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

3 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

4 NTDT: a designation based on transfusion requirement NTDT patients do not require regular transfusions, although they may require occasional transfusions for growth failure, pregnancy, infections, and other specific situations 1 4 Non transfusion dependent thalassaemia (NTDT) β-thalassaemia intermedia Mild/moderate HbE/β-thalassaemia HbH disease (α-thalassaemia) HbS β-thalassaemia HbC thalassaemia Transfusions seldom required Occasional transfusions required (e.g. surgery, pregnancy, infection) Intermittent transfusions required (e.g. poor growth and development, specific morbidities) Regular, lifelong transfusions required for survival Transfusions not required α-thalassaemia trait β-thalassaemia minor 1. Taher AT, et al. Br J Haematol. 2011;152: Galanello R, Origa R. Orphanet J Rare Dis. 2010;5: Vichinsky E. Hematology Am Soc Hematol Educ Program. 2007; Muncie HL Jr, Campbell J. Am Fam Physician. 2009;80: Figure adapted from Musallam KM, et al. Haematologica. 2013;98: Transfusion dependent thalassaemia (TDT) β-thalassaemia major Severe HbE/β-thalassaemia Hb Barts hydrops (α-thalassaemia major) HbE, haemoglobin E; HbH, haemoglobin H.

5 The prevalence of NTDT is increasing worldwide due to migration Prevalence of Hb H disease in Europe and North America increasing due to migration 3 Hb H most common in Southeast Asia, Middle East and Mediterranean 3 Approx annual births 4 β thalassemia intermedia most common in Africa, Mediterranean, India and East Europe 1,2 World distribution of hemoglobinopathies 3 3 Hb E/β thalassemia most common in East India, Bangladesh and Southeast Asia 1 Approx. 19,000 annual births 4 The increased migration flow expands the reach of these diseases 1 Weatherall DJ. Blood Rev 2012;26 Suppl 1:S3 S6; 2 Available from: 3 Harteveld CL and Higgs DR. Orphanet J Rare Dis 2010;5:13; 4 Weatherall DJ. Blood 2010;115:

6 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

7 How it all started: realizing morbidity is highly prevalent and different from TDT in our clinics Complication (% of patients affected) Lebanon (n = 37) β-ti Italy (n = 63) Lebanon (n = 40) β-tm Italy (n = 60) Splenectomy Cholecystectomy Gallstones Extramedullary haemopoiesis Leg ulcers Thrombotic events Cardiopathy a PHT b Abnormal liver enzymes HCV infection Hypogonadism Diabetes mellitus Hypothyroidism a Fractional shortening < 35%. b Defined as pulmonary artery systolic pressure > 30 mmhg; a well-enveloped tricuspid regurgitant jet velocity could be detected in only 20 patients, so frequency was assessed in these patients only. HCV, hepatitis C virus. Taher A, et al. Blood Cells Mol Dis. 2006;37:12-20.

8 Observations were similar in reported literature β-thalassemia Major (Regularly transfused) Non-Transfusion-Dependent Thalassemias (NTDT) Facies Bone deformity (Hair on end) Hypothyroidism Hypoparathyroidism Cardiac siderosis Left-sided heart failure Hepatic failure Viral hepatitis Diabetes mellitus Hypogonadism Osteoporosis Silent cerebral ischemia Pulmonary hypertension Right-sided heart failure Extramedullary hematopoietic psuedotumors Hepatic fibrosis, cirrhosis and cancer Gall stones Splenomegaly Osteoporosis Venous thrombosis Leg ulcers Musallam KM, et al. Haematologica. 2013;98: Musallam KM et al. Haematologica. 2013;98:

9 The OPTIMAL CARE study Overview on Practices in β-thalassaemia Intermedia Management Aiming for Lowering Complication rates Across a Region of Endemicity: Cross-sectional study of 584 patients with β-ti from 6 comprehensive care centres in the Middle East and Italy n = 127 A.T. Taher K.M. Musallam n = 153 M.D. Cappellini n = 200 M. Karimi n = 51 A. El-Beshlawy n = 12 S. Daar n = 41 K. Belhoul M. Saned β-ti, β-thalassaemia intermedia. Taher AT, et al. Blood. 2010;115:

10 High morbidity rates were confirmed Parameter Frequency n (%) Age (years) < (29.5 ) (49.3) > (21.2) Male:female 291 (49.8) : 293 (50.2) Splenectomized 325 (55.7) Serum ferritin (µg/l) < 1, (64.4) 1,000 2, (30.6) > 2, (5) Complications Osteoporosis EMH Hypogonadism Cholelithiasis Thrombosis Pulmonary hypertension Abnormal liver function Leg ulcers Hypothyroidism Heart failure Diabetes mellitus 134 (22.9) 124 (21.2) 101 (17.3) 100 (17.1) 82 (14) 64 (11) 57 (9.8) 46 (7.9) 33 (5.7) 25 (4.3) 10 (1.7) Treatment Frequency n (%) Hydroxyurea 202 (34.6) Transfusion Never Occasional Regular Iron chelation None Deferoxamine Deferiprone Deferiprone + deferoxamine Deferasirox 139 (23.8) 143 (24.5) 302 (51.7) 248 (42.5) 300 (51.4) 12 (2.1) 3 (0.5) 21 (3.6) Taher AT, et al. Blood. 2010;115:

11 Frequency (%) Morbidities seemed to increase with advancing age * years years years > 32 years 120 treatment-naive patients *Statistically significant trend * 26.7 * * * * ALF, abnormal liver function; DM, diabetes mellitus; HF, heart failure. Taher AT, et al. Br J Haematol. 2010;150:486-9.

12 The pathophysiology and risk factors were predicted but not yet evaluated Impaired α:β globin ratio Red cell pathology Ineffective erythropoiesis haemolysis Gall stones Iron overload Anaemia Tissue oxygenation Hypercoagulable state Erythroid marrow expansion Diabetes mellitus Growth deficiency Hypothyroidism Hypoparathyroidism Hypogonadism Hepatic cancer Renal disease Leg ulcers Thrombotic events Pulmonary hypertension Bone deformities Osteoporosis Hepatosplenomegaly Extramedullary haematopoietic pseudotumors Musallam KM, et al. Haematologica. 2013;98:833-4.

13 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

14 Ineffective erythropoiesis leads to iron overload in NTDT Ineffective erythropoiesis Chronic anemia/hypoxia?gdf-15 TWGF-1 HIFs Tmprss6 Erythropoietin Hepcidin Ferroportin Intestinal iron absorption Release of recycled iron from reticuloendothelial system Transfusions Minor role Iron overload Musallam KM, et al. Curr Opin Hematol. 2013; 20:

15 Serum ferritin (ng/ml) Chronic anemia is also independently associated with clinical morbidity and iron overload in NTDT A hemoglobin of <7 g/dl was the level below which all patients developed a morbidity while a hemoglobin of >10 g/dlwas the level after which none of the patients had a morbidity (area under the curve = 0.84, 95% CI: 0.70 to 0.97, p b 0.001) Hemoglobin level (g/dl) Pearson's r= , p b Taher AT, et al. Blood Cell Mol Dis. 2015;55:108-9.

16 Extramedullary hematopoietic psuedotumors in NTDT as a result of ineffective erythropoiesis and anemia/hypoxia Chronic anemia Defective Hb F unloading Hypoxia Thoracic and lumbar regions are most commonly involved Symptoms develop as a result of pressure on surrounding structures Spinal cord compression and possible irreversible neurological damage is most significant and debilitating Increased bone marrow hematopoietic activity and expansion + Formation of erythropoietic tissue masses Liver EXTRAMEDULLARY hematopoiesis Glands Pleura Spleen Spine Haidar R et al. Eur Spine J 2010;19: ; Castelli R et al. Am J Med Sci 2004;328: ; Chehal A et al. Spine 2003;28:E245 E249; Borgna-Pignatti C. Br J Haematol 2007;138:

17 Leg ulcers in NTDT can develop in the context of anemia and tissue hypoxia Leg ulcers are more common in older than in younger patients with β-ti Leg ulcers 2 The skin at the extremities of elderly β-ti patients can be thin due to reduced tissue oxygenation; this makes the subcutaneous tissue fragile and increases the risk of lesions Ulcers are very painful and difficult to cure 1 Risk factors: severe anaemia, ineffective erythropoiesis, splenectomy, and hypercoagulability levels 3 5 A role for local IOL is suggested; IOL may play a role in the pathophysiology of leg ulcers by causing oxidative stress and not just by local accumulation 6 1. Taher AT, et al. Blood Cells Mol Dis. 2006;37: Levin C, Koren A. Isr Med Assoc J. 2011;13: Taher AT, et al. Blood. 2010;115: Musallam KM, et al. Blood Cells Mol Dis. 2011;47: Musallam KM, et al. Cold Spring Harb Perspect Med. 2012;a Matta BN, et al. J Eur Acad Dermatol Venereol. 2014;28(9):

18 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

19 Since it was apparent that NTDT patients develop iron overload in the absence of transfusions, the question was should we be concerned with this iron overload?

20 Mean serum ferritin level and 95% CI (ng/ml) SF levels increased over 11 years in 52 non-chelated β-ti patients: The ORIENT study p< Musallam KM, et al. Haematologica 2014;99:e

21 Cardiac T2* (ms) LIC (mg Fe/g dry wt) Iron overload and the heart Normal cardiac R2* Normal LIC (< 3 mg Fe/g dry wt) Mild LIC (3 7 mg Fe/g dry wt) Moderate LIC (7 15 mg Fe/g dry wt) Severe LIC (> 15 mg Fe/g dry wt) n=20 n=49 Serum ferritin (ng/ml) Cardiac R2* (Hz) No evidence of cardiac siderosis even in NTDT patients with considerable iron overload Origa R, et al. Haematologica. 2008;93: Roghi A, et al. Ann Hematol. 2010;89: Taher AT, et al. Am J Hematol. 2010;85: Mavrogeni S, et al. Int J Cardiovasc Imaging. 2008;24:

22 Borgna-Pignatti C et al. Br J Haematol 2004;124: ; Maakaron JE et al. Ann Hepatol 2013;12: ; Maakaron JE et al. Br J Haematol 2013;161:1; Mancuso A. World J Hepatol 2010;2: ; Restivo Pantalone G et al. Br J Haematol 2010;150: HCC is one of the most clinically severe complications of NTDT & may be related to iron overload Several case reports and case series suggest an association between iron overload and HCC in hepatitis C-negative patients with NTDT HCV Plus iron Possible reasons for increased risk of HCC in NTDT Iron overload is undiagnosed (serum ferritin underestimates iron burden compared with TDT) Chelation is started late or not at all NTDT patients survive longer than TDT, often an older population Hepatocellular > macrophage distribution

23 HCC found in non-viral, non-cirrhotic livers, highlighting the potential severity of iron overload and its complications Case reports and case series in the literature reviewed for consistency between patient characteristics HCC observed in patients with non-viral (hepatitis), non-cirrhotic livers Of 36 cases, 22 were β TI patients, of whom six were negative for hepatitis B or C Survey for HCC using an abdominal ultrasound every 6 months is recommended in at-risk patients Sex Age* Survival HCV Ab HCV RNA HBV Ab HBs Ag AFP (ku/l) Serum ferritin (ng/ml) Serum ferritin peak (ng/ml) LIC M 48 Alive at 26 months NA F 61 5 months NA F months + NA NA NA 990 NA NA M 73 7 months NA NA NA 574 NA NA M 54 1 months M 51 4 years Vaccinated HCC is a serious yet often overlooked complication in NTDT with appropriate screening, early detection and adequate iron chelation, complications can be minimized *At diagnosis Ab, antibody; AFP, alpha-fetoprotein; HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen Maakaron JE et al. Ann Hepatol 2013;12:

24 LIC (mg Fe/g dry wt) Iron overload and other morbidities p = p = p = p < p = p < p = p = p < p < On multivariate analysis, a 1 mg Fe/g dw increase in LIC was significantly associated with higher odds of thrombosis, pulmonary hypertension, hypothyroidism, osteoporosis, and hypogonadism Adjusted for age, gender, splenectomy status, transfusion history, total hemoglobin level, fetal hemoglobin level, platelet count, NRBC count, and serum ferritin level 168 non-chelated β-ti Morbidity absent Morbidity present 0 Musallam KM, et al. Haematologica. 2011;96:

25 Renal dysfunction in thalassemia is generally caused by persistent hypoxia, anemia, and severe IOL Anemia Glomerular dysfunction vascular resistance results in GFR Hypoxia Tubulointerstitial dysfunction epithelial mesenchymal transdifferentiation accumulation of extracellular matrix activation of fibroblasts diffusion capacity of oxygen Iron overload Tubulointerstitial dysfunction direct iron cytotoxicity on tubules and glomerular permeability causes enzymuria + proteinuria moderate tubular atrophy and interstitial fibrosis Glomerular dysfunction mild to moderate glomerulosclerosis GFR GFR, glomerular filtration rate. Adapted from Mallat NS, et al. J Nephrol 2013;26: Adapted from Musallam KM, Taher AT. J Am Soc Nephrol. 2012;23: Iron chelation Tubular dysfunction vacuolization of proximal tubular epithelium apoptosis of tubule cell tubular absorption of solutes Glomerular dysfunction reversible haemodynamic changes

26 Renal dysfunction in β-ti is associated with iron overload 48% had evidence of glomerular hyperfiltration 14% patients had proteinuria* patients had elevated LIC>7 mg Fe/g dw, NTBI and nucleated RBC counts A considerable proportion of patients showed evidence of abnormally elevated egfr, with a declining trend towards advancing age r s =0.357 P=0.011 r s =0.444 P=0.001 Occurrence of proteinuria associated with anemia, hemolysis and iron toxicity *UPr/UCr ratio >500 mg/g N=50 Ziyadeh FN, et al. Nephron Clin Prac. 2012;121:c136-c43.

27 Long-term follow-up of NTDT patients is crucial to identify TI patients at risk of end-stage renal disease 127 TI patients were observed for 10 years 6 (4.7%) developed ESRD that required regular haemodialysis 4 showed no abnormality in urine sediment or proteinuria 2 showed glomerular-range proteinuria 1 reached nephrotic range (biopsy showed FSGS) Tubular atrophy and interstitial fibrosis Segmental sclerosis in a glomerulus Iron deposits (Prussian blue stain) ESRD, end-stage renal disease; FSGS, focal segmental glomerulosclerosis. Mallat NS, et al. Blood Cells Mol Dis. 2013;51:146-8.

28 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

29 Hypercoagulability and vascular disease in NTDT Transfusion Splenectomy Iron overload Pathologic RBCs Thrombin generation (phosphatidyl serine exposure) Rigidity, deformability, and aggregation Endothelial injury Expression of adhesion molecules and tissue factor Platelet abnormalities Thrombocytosis Chronic activation Adhesion and aggregation Circulating microparticles Antithrombin III Protein C Protein S Endocrine & hepatic dysfunction? Atherosclerosis THROMBOSIS Musallam KM, et al. Haematologica. 2013;98:833-4.

30 Type of event Thromboembolic events in a large cohort of β-ti patients Patients (N = 8,860) 6,670 with β-tm 2,190 with β-ti 146 (1.65%) thrombotic events 61 (0.9%) with β-tm 85 (3.9%) with β-ti Risk factors for developing thrombosis in β-ti were age (> 20 years) previous thromboembolic event family history splenectomy Venous Stroke DVT PE PVT STP Others TM (n = 61) TI (n = 85) Thromboembolic events (%) DVT = deep vein thrombosis PVT = portal vein thrombosis; STP = superficial thrombophlebitis Taher A, et al. Thromb Haemost. 2006;96:

31 OPTIMAL CARE study: multivariate analysis on risk factors for thrombosis in splenectomised patients Parameter Group OR 95% CI p value NRBC count 300 x 10 6 /L Group III 1.00 Referent Group II Group I < Platelet count 500 x 10 9 /L Group III 1.00 Referent Group I patients had significantly higher Group II NRBC, platelets, and PHT occurrence, and Group I were mostly non-transfused PHT Group III 1.00 Referent Group II Group I Transfusion naivety Group III 1.00 Referent Group II Group I < NRBC = nucleated red blood cell; PHT = pulmonary hypertension; OR = adjusted odds ratio; CI = confidence interval. Taher AT, et al. J Thromb Haemost. 2010;8:

32 Silent cerebral infarction in splenectomized patients Study Number of patients Age (years) MRI technique Definition of infarction Prevalence of SCI (95% CI) Manfre et al Mean: Tesla Abnormally high signal 37.5% ( ) Range: 9-48 T2- spin-echo weighted intensity on long TRweighted images imaging Two blinded neuroradiologists Taher et al Mean: 32.1 Range: Tesla 60.0% ( ) Karimi et al Mean: 24.3 Range: Teli et. al Mean: 12 Range: T1-, T2- gradient-echo-, FLAIR-, diffusionweighted imaging a Other diagnoses were ruled out based on the radiological appearance of lesions (e.g. viral encephalopathy or multiple sclerosis) or absence of associated risk factors and clinical symptoms (e.g. vasculitis or Binswanger s disease). FLAIR, fluid attenuation inversion recovery; N/A, data not available; CI, confidence interval. Abnormally high signal intensity on the T2- and FLAIR-weighted images a Two blinded neuroradiologists 1.5 Tesla Abnormally high signal 26.7% ( ) T1-, T2-, FLAIRweighted intensity on the T2- and FLAIR- weighted images imaging One neuroradiologist N/A N/A 0% Musallam KM, et al. Thromb Res. 2012;130:

33 Probability of abnormality Patients (n) Risk factors for silent cerebral infarction Age (years) Increasing age and transfusion naivety are associated with a higher incidence and multiplicity of white matter lesions Occasionally transfused Non-transfused No. of abnormalities > 50 Age (years) 0 1 > 1 Taher AT, et al. J Thrombosis Haemost. 2010;8:54-9.

34 Risk factors for PHT: OPTIMAL CARE β-ti patients (N = 64) Parameters PHT+* PHT Splenectomized % AOR (95% CI) History of thromboembolic events % AOR (95% CI) Nucleated RBC count Mean (SD; x 10 6 /L) AOR (95% CI) Transfusion naivety % AOR (95% CI) Iron chelation naivety % AOR (95% CI) Hydroxyurea naivety % AOR (95% CI) ( ) ( ) (199.2) 2.59 ( ) ( ) ( ) ( ) (94.5) * p < for all parameters for patients with PHT versus those without PHT AOR = adjusted odds ratio Karimi M, et al. Eur J Intern Med. 2011;22:

35 LIC (mg Fe/g dry wt) Probability of PHT (%) Risk of Pulmonary Hypertension in NTDT increases with high LIC, advancing age, and splenectomy SF LIC TCG and SF TCG and LIC r = p = 0.01 r = p = Tricuspid gradient (mmhg) 5,000 4,000 3,000 2,000 1,000 0 SF (µg/l) All patients Non-splenectomised Splenectomised Age (years) PHT (defined as PASP 30 mmhg) present in 38.5% Significantly correlated with LIC Not correlated with age, Hb level, and SF level PHT prevalence in thalassaemia was 2.1% (TI 4.8%, TM 1.1%) PASP, pulmonary artery systolic pressure; TCG, tricuspid gradient. Derchi G, et al. Circulation. 2014;129: Isma eel H, et al. Am J Cardiol. 2008;102:363-7.

36 Mean scale score With all this morbidity it was not surprising that β-ti patients had poor HR-QoL * * * β-ti β-tm *P < 0.05 Musallam KM, et al. Eur J Haematol. 2011;87:73-9.

37 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

38 Splenectomy for NTDT patients Splenomegaly and hypersplenism are inevitable in untransfused/rarely transfused patients Due to extramedullary hematopoiesis, splenic pooling and chronic passive congestion of the spleen Leads to worsening of anemia, neutropenia, thrombocytopenia Enlargement can be associated with abdominal pain and risk of rupture Thus, splenectomy is a common practice in NTDT patients especially that it can lead to an increase the total hemoglobin level by 1-2 g/dl and avoid blood transfusions Cholecystectomy during splenectomy is commmon in NTDT patients who are at increased risk of gall stone formation from increased hemolysis Taher AT, et al. Br j Haematol. 2011;152: Abi Saad GS, et al. Br J Surg. 2011;98:

39 Splenectomy for NTDT patients: Adverse events Infections Increased susceptibility to infections especially in patients <5 years Overwhelming post-splenectomy sepsis, which can be rapidly fatal Appropriate vaccinations and prophylactic antibiotics are necessary Vascular events Venous thromboembolism Pulmonary hypertension Silent cerebral infarction Taher AT, et al. Br j Haematol. 2011;152: Abi Saad GS, et al. Br J Surg. 2011;98:

40 The OPTIMAL CARE study splenectomized patients: 325/584 Complication Parameter RR 95% CI p value EMH Splenectomy Transfusion < Hydroxyurea Pulmonary hypertension Age > 35 years Splenectomy < Transfusion < Hydroxyurea Iron chelation Heart failure Transfusion < Thrombosis Age > 35 years Hb 9 g/dl Serum ferritin 1,000 µg/l Splenectomy < Transfusion < Cholelithiasis Age > 35 years < Female Splenectomy < Transfusion < Iron chelation < Abnormal liver function Serum ferritin 1,000 µg/l Taher AT, et al. Blood. 2010;115:

41 The OPTIMAL CARE study splenectomized patients: 325/584 (cont.) Complication Parameter RR 95% CI p value Leg ulcers Age > 35 years Splenectomy Transfusion Hydroxyurea Hypothyroidism Splenectomy Hydroxyurea Osteoporosis Age > 35 years < Female of most disease-related complications Splenectomy < Transfusion < Hydroxyurea < Iron chelation Hypogonadism Female < Serum ferritin 1,000 µg/l < Transfusion < Hydroxyurea < Iron chelation Splenectomy was independently associated with an increased risk Taher AT, et al. Blood. 2010;115:

42 Splenectomy: best practice Splenectomy is generally avoided in NTDT patients younger than 5 years Splenectomy is reserved for cases of: Worsening anemia leading to poor growth and development When transfusion therapy is not possible or iron chelation therapy is unavailable Hypersplensim Leucopenia or thrombocytopenia causing clinical problems such as recurrent bacterial infection or bleeding Splenomegaly Accompanied by symptoms such as left upper quadrant pain or early satiety Massive splenomegaly (largest dimension >20 cm) with concern about possible splenic rupture Taher AT, et al. TIF Guidelines 2013.

43 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

44 Progressive anaemia or enlargement of the spleen Use of blood transfusions in NTDT Occasional blood transfusions Infections Pregnancy Surgery More frequent transfusions * Childhood: poor growth and development Adulthood: prevention or management of specific complications Prevention** Thrombotic or cerebrovascular disease PHT with / without secondary heart failure EMH Leg ulcers *longer intervals than in β-tm and for defined durations ** in high-risk populations Taher AT et al. Blood Rev 2012;26S:S24, Oliviery NF Blood Rev 2012;26S:S28, Vichinsky E Blood Rev 2012;26S:31; Taher AT et al. TIF Guidelines 2013.

45 The OPTIMAL CARE study occasionally/regularly transfused patients: 445/584 Complication Parameter RR 95% CI p value EMH Splenectomy Transfusion < Hydroxyurea Pulmonary hypertension Age > 35 years Splenectomy < Transfusion < Hydroxyurea Iron chelation Heart failure Transfusion < Thrombosis Age > 35 years Hb 9 g/dl Serum ferritin 1,000 µg/l Splenectomy < Transfusion < Cholelithiasis Age > 35 years < Female Splenectomy < Transfusion < Iron chelation < Abnormal liver function Serum ferritin 1,000 µg/l Taher AT, et al. Blood. 2010;115:

46 The OPTIMAL CARE study occasionally/regularly transfused patients: 445/584 (cont.) Complication Parameter RR 95% CI p value Leg ulcers Age > 35 years Splenectomy Transfusion Hydroxyurea Hypothyroidism Splenectomy Hydroxyurea cholelithiasis, and leg ulcers Osteoporosis Age > 35 years < Female Transfusion Splenectomy therapy was associated 4.73 with an increased < risk Transfusion < of endocrinopathy Hydroxyurea < Iron chelation Hypogonadism Female < Serum ferritin 1,000 µg/l < Transfusion < Hydroxyurea < Iron chelation Transfusion therapy was protective for thrombosis, EMH, PHT, HF, Taher AT, et al. Blood. 2010;115:

47 Alloimmunization Risk factors Minimally transfused and newly transfused patients Old age at first transfusion Ethnic and racial diverse population Splenectomy Allogeneic WBC within the transfusate Management Fully phenotyped matched blood should be given Erythropoietin in combination with iron and folic acid Chou ST et al. Br J Haematol. 2012;159:

48 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

49 Extramedullary hematopoiesis Pulmonary hypertension Venous thromboembolism Heart failure Leg ulcers Abnormal liver function Diabetes mellitus Hypothyroidism Hypogonadism Osteoporosis Median HbF (%) Elevations in HbF ameliorate morbidity in NTDT * * * 63 untransfused β-ti * * 40 No 30 Yes *P<0.001; P=0.003; P=0.002 Musallam KM, et al. Blood. 2012;119:364-7.

50 Experience with HbF inducers in β- thalassemia Agent Main positive findings Limitations DNA methylation inhibitiors 5-azacytidine Marked hematological responses achieved Few studies Small sample sizes Safety concerns Decitabine Hematological responses achieved Few studies Favorable effects on red cell indices Small sample sizes Well-tolerated Hydroxyurea Hematological responses achieved Heterogonous phenotypes studied together Favorable effects on red cell, hemolysis, and hypercoagulability indices Heterogeneous study endpoints evaluated together Favorable effects on clinical morbidities Ideal dose and duration of therapy still Well-tolerated controversial Lack of efficacy on long-term therapy Data on predictors of response remain inconsistent Short-chain fatty acids Hematological responses achieved Favorable effects on red cell and hemolysis indices Well-tolerated Erythropoietic stimulating agents Hematological responses achieved Favorable effects on combination with hydroxyurea Well-tolerated Thalidomide and derivatives Hematological responses achieved Well-tolerated Small sample sizes Lack of efficacy on long-term therapy Few studies Small sample sizes High doses required No additive effects with short-chain fatty acids Few studies Small sample sizes Musallam KM, et al. Blood 2013;121:

51 The OPTIMAL CARE study: patients on hydroxyurea: 202/584 Complication Parameter RR 95% CI p value EMH Splenectomy Transfusion < Hydroxyurea Pulmonary hypertension Age > 35 years Splenectomy < Transfusion < Hydroxyurea Iron chelation Heart failure Transfusion < Thrombosis Age > 35 years Hb 9 g/dl Serum ferritin 1,000 μg/l Splenectomy < Transfusion < Cholelithiasis Age > 35 years < Female Splenectomy < Transfusion < Iron chelation < Abnormal liver function Serum ferritin 1,000 μg/l Taher AT, et al. Blood. 2010;115:

52 The OPTIMAL CARE study: patients on hydroxyurea: 202/584 (cont.) Complication Parameter RR 95% CI p value Leg ulcers Age > 35 years Splenectomy Transfusion Hydroxyurea Hypothyroidism Splenectomy Hydroxyurea Osteoporosis Age > 35 years < Female hypothyroidism, and 1.97 osteoporosis Splenectomy < Transfusion < Hydroxyurea < Iron chelation Hypogonadism Female < Serum ferritin 1,000 μg/l < Transfusion < Hydroxyurea < Iron chelation Hydroxyurea treatment was protective for EMH, PHT, leg ulcers, Taher AT, et al. Blood. 2010;115:

53 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

54 1. Musallam KM, et al. Blood Rev. 2012;26S:S Taher A, et al. Am J Hematol. 2010;85: Taher AT, et al. Br J Haematol 2009;146: SQUID = superconducting quantum interfering device NTBI = Non-transferrin-bound iron Assessment of iron overload in NTDT There are several methods to assess iron overload in NTDT patients; each carrying their own advantages and disadvantages 1 : Liver iron concentration (LIC) MRI Biopsy SQUID Serum ferritin Cardiac MRI No evidence of cardiac siderosis in NTDT 2 May be reserved for older patients with high LIC Other markers (NTBI, Transferrin Sat) 3 Few clinical studies

55 Serum ferritin level (ng/ml) 10, Serum ferritin underestimates iron burden in NTDT TI Linear (TI) TM Linear (TM) LIC (mg Fe/g dw) Taher A, et al. Haematologica. 2008;93: Musallam KM, Taher AT. N Engl J Med. 2011;364:1476.

56 Which NTDT patients should receive iron chelation therapy and at what thresholds Patients 10 years of age Increased iron-related morbidity risk beyond this age 1 Patients with a LIC 5 mg Fe/g dw Increased risk of morbidity beyond this threshold 2,3 Iron chelator efficacy beyond this threshold is established (THALASSA) 4,5 Patients with a serum ferritin 800 ng/ml (LIC not available) Best reflects a LIC of 5 (specificity/sensitivity analysis) 6 Increased morbidity risk beyond this threshold 7 1. Taher AT, et al. Br J Haematol. 2010;150: Musallam KM, et al. Haematologica. 2011;96: Musallam KM, et al. Blood Cells Mol Dis. 2013;51: Taher AT, et al. Blood. 2012;120: Taher AT, et al. Ann Hematol. 2013;92(11): Taher A, et al. Br J Haematol. 2015;168(2): Musallam KM, et al. Haematologica 2013;98(S1):S1172.

57 Prevalence (%) Patients with a LIC 5 have a higher prevalence of morbidities (n=168, β-ti, non-chelated) P< P= LIC < 5 mg/g dw 50.0 P= P=0.008 P=0.005 The prevalence of any morbidity was 84.7% vs. 50% and multiple morbidity 30.0 was 60.2% vs. 16.1%, in the 5 and 28.6<5 groups, respectively (p<0.001) Thrombosis PHT Hypothyroidism Osteoporosis Hypogonadism Musallam KM, et al. Blood Cells Mol Dis 2013;51:35-8.

58 LIC (mg/g dw) Morbidity rate difference around the LIC of 5 threshold Thrombosis vs. < vs. < Pulmonary hypertesnion vs. < vs. < Hypothyroidism vs. < vs. < Within Osteoporosis a clinically relevant 29.6 LIC range 9 (3-15), vs. <9 the LIC of 5 threshold 29.3had 8 vs. < the highest absolute risk difference for development of a morbidity Hypogonadism vs. <7 6 vs. < Any morbidity vs. < vs. < Multiple morbidity vs. < Absolute risk difference Between LIC 5 vs. <5 mg/g (%) Absolute risk difference (%) Musallam KM, et al. Blood Cells Mol Dis 2013;51:35-8.

59 Proportion of patients surviving without morbidity Patients with SF 800 μg/l have a higher incidence of morbidities over 11 years Time to first morbidity (years) Number at risk SF 300 μg/l SF > 300 to < 800 μg/l SF 800 μg/l SF 300 µg/l SF > 300 to < 800 µg/l SF 800 µg/l Chi-square = p value = ORIENT study Musallam KM, et al. Haematologica. 2014;99:e

60 Sensitivity (%) SF >800 ng/ml is predictive of LIC >5 mg/g When MRI is unavailable, SF thresholds are useful to indicate initiation of ICT, dose escalation, and/or dose interruption Patients with SF > 800 μg/l had a high probability of LIC 5 mg Fe/g dry wt SF > 2,000 μg/l was highly predictive of LIC 7 mg Fe/g dry wt SF 300 μg/l appears adequate to safely interrupt DFX therapy SF = 800 µg/l Specificity (%) However, around 50% of patients with SF<800 could still have LIC >5 ROC, receiver operating characteristic. Taher AT, et al. Br J Haematol. 2015;168:

61 DFO and DFP in NTDT DFO 1 Deferiprone 2 Significant decline in serum ferritin after 6 months of DFO treatment Significant UIE after 12 hours of continuous DFO (except in patients aged < 1 year) In some patients, substantial UIE despite modest SF levels SF levels of no value in predicting UIE No sigificant differences in excretion across doses 1. Cossu P. Eur J Pediatr. 1981;137: Pootrakul P, et al. Br J Haematol. 2003;122: Porter JB, et al. Int J Hematol. 2001;94: Trachtenberg F, et al. Am J Hematol. 2011;85: Delea TE, et al. Transfusion. 2007;47:1919. Significant reductions seen in mean SF, hepatic iron, red cell membrane iron, and serum NTBI levels SF ± SD: initial 2168 ± 1142 μg/l; final 418 ± 247 μg/l Significant mean increase in serum EPO also observed Increase in Hb values in 3 patients; reduction in transfusion requirements in 4 patients NTBI, non-transferrin-bound iron; UIE, urinary iron excretion.

62 Deferasirox in NTDT: THALASSA study design THALASSA study evaluated the efficacy of two deferasirox regimens (5 and 10 mg/kg/day) in NTDT patients, based on the change in LIC from baseline compared with placebo Randomize NTDT patients (2:1 DFX/placebo) CORE EXTENSION Screening (28 days) Deferasirox 5 mg/kg/day Placebo 5 mg/kg/day Deferasirox 10 mg/kg/day DFX LIC < 3 mg Fe/g dry wt: interrupt LIC 3 15 mg Fe/g dry wt: 10 mg/kg/day Placebo 10 mg/kg/day LIC > 15 mg Fe/g dry wt: 20 mg/kg/day 24 weeks 52 weeks 104 weeks LIC was measured by MRI after every 6 months SF was measured monthly The THALASSA study demonstrated a significant reduction in LIC in both deferasirox arms with a greater reduction in the 10 mg/kg/day group Taher AT, et al. Blood. 2012;120:970-7; Taher AT, et al. Ann Hematol. 2013;92:

63 Median ± 25th/75th percentile absolute change in SF (μg/l) Mean absolute change in LIC (mg Fe/g dry wt) Deferasirox continues to reduce iron burden over 2 years ,000 Deferasirox core starting dose 5 mg/kg/day Deferasirox core starting dose 10 mg/kg/day Placebo/deferasirox BL Time (months) Deferasirox core starting dose 5 mg/kg/day Deferasirox core starting dose 10 mg/kg/day Placebo/deferasirox BL Time (months) Deferasirox core + extension: median dose = 9.5 mg/kg/day Deferasirox extension: median dose = 10.8 mg/kg/day Placebo/deferasirox: median dose = 14.0 mg/kg/day BL, baseline. Taher AT, et al. Ann Hematol. 2013; 92:

64 Deferasirox consistently reduces iron burden across all evaluated patient subgroups Age Age <18 (n=21) Age 18 (n=145) Gender Male (n=89) Female (n=77) Race Caucasian (n=94) Asian (n=69) Black (n=2) Other (n=1) Baseline LIC 7 (n=31) >7 15 (n=77) >15 (n=57) Baseline SF (ng/ml) > (n=17) > (n=67) >1000 (n=82) Splenectomy Yes (n=88) No (n=78) n=20 Underlying disease β TI (n=95) α thalassemia (n=22) Hb E/β thalassemia (n=49) Deferasirox 10 mg/kg/day better n=7 n=47 n=28 n=26 n=29 n=4 n=23 n=27 n=29 n=30 n=24 n=16 n=24 n= Least squares mean difference in absolute change in LIC from baseline (95% CI) n=9 n=8 CI, confidence interval Taher AT et al. Am J Hematol 2013;88: n=32 n=22 n=32 n=46 n=30 n=24 n=20 n=27 n=28 n=26 n=8 n=21 n=12 n=20 n=15 n=7 Placebo better n=7

65 THALASSA: most common ( 3 patients overall) drug-related adverse events AE, n (%) Deferasirox 5 mg/kg/day (n = 55) Deferasirox 10 mg/kg/day (n = 55) Placebo 5 mg/kg/day (n = 28) Placebo 10 mg/kg/day (n = 28) Total (N = 166) Nausea 3 (5.5) 4 (7.3) 1 (3.6) 3 (10.7) 11 (6.6) Skin rash 2 (3.6) 5 (9.1) 0 1 (3.6) 8 (4.8) Diarrhoea 0 5 (9.1) 0 1 (3.6) 6 (3.6) Headache 2 (3.6) 1 (1.8) 0 2 (7.2) 5 (3.0) Upper abdominal pain 2 (3.6) 1 (1.8) (1.8) Abdominal pain 1 (1.8) 1 (1.8) 1 (3.6) 0 3 (1.8) Overall incidence of AEs was comparable between deferasirox and placebo Most drug-related AEs were of mild-to-moderate severity and resolved without discontinuation of treatment Taher AT, et al. Blood. 2012;120: Taher AT, et al. Ann Hematol. 2013; 92: AE, adverse event.

66 Mean creatinine clearance ± SD (ml/min) Mean UPCR ± SD (mg/mg) Mean ALT ± SD (U/L) Mean serum creatinine ± SD (µmol/l) There were no progressive changes in ALT, creatinine or UPCR in patients treated with deferasirox or placebo Deferasirox 5 mg/kg/day Deferasirox 10 mg/kg/day Placebo 5 mg/kg/day Placebo 10 mg/kg/day Time (months) Time (months) Time (months) Time (months) 3 patients had 2 consecutive serum creatinine increases >33% above baseline and > ULN ALT, alanine aminotransferase; UPCR, urine protein/creatinine ratio Taher A, et al. Blood. 2012;120:970-7.

67 THALASSA: safety profile of deferasirox remains consistent as NTDT patients approach target LIC < 3 for interrupting chelation Laboratory parameters at baseline and before reaching LIC < 3 mg Fe/g dry wt Parameter (mean ± SD) Baseline End of Period 1 a End of Period 2 a Creatinine, μmol/l 51.8 ± ± ± 19.9 Creatinine clearance, ml/min ± ± ± 52.3 Alanine aminotransferase, U/L 31.4 ± ± ± 6.8 Urinary protein/creatinine ratio, mg/mg a Last available assessment. 0.2 ± ± ± 0.1 Taher AT, et al. Eur J Haematol. 2014;92:521-6.

68 THETIS was an open-label, single-arm, multicenter Phase IV study to evaluate the efficacy and safety of deferasirox in NTDT patients with iron overload Screening (4 weeks) to determine patient eligibility Open-label treatment (52 weeks) with deferasirox: Starting dose 10 mg/kg/day for 4 weeks At Week 4, dose adjustments according to baseline LIC Results from THALASSA have demonstrated that more aggressive chelation should be initiated earlier in the course of treatment 15 mg/kg/day for patients with baseline LIC >7 but 15 mg Fe/g dw 20 mg/kg/day for patients with baseline LIC >15 mg Fe/g dw 10 mg/kg/day for patients with baseline LIC 5 but 7 mg Fe/g dw Primary objective of THETIS is to assess the efficacy of deferasirox in patients with NTDT based on change in LIC from baseline after 52 weeks of treatment At Week 24, dose adjustments according to Week 24 LIC Increase dose by 5 10 mg/kg/day if Week 24 LIC >15 mg Fe/g dw Increase dose by 5 mg/kg/day if Week 24 LIC >7 but 15 mg Fe/g dw and Week 24 LIC reduction from baseline <15% Increase dose by 5 10 mg/kg/day if Week 24 LIC increased from baseline to the next category Same dose if Week 24 LIC remained in the same baseline LIC category Decrease dose by 5 10 mg/kg/day if Week 24 LIC decreased from baseline to the next category Taher AT, et al. Blood Cells Mol Dis. 2016;57:23-9.

69 THETIS change in LIC and SF With significant and clinically relevant reductions in iron burden alongside a safety profile similar to that in THALASSA, these data support earlier escalation with higher deferasirox doses in ironoverloaded non-transfusion-dependent anaemia patients. Taher AT, et al. Blood Cells Mol Dis. 2016;57:23-9.

70 Algorithm for Iron Overload Assessment and Chelation Therapy in NTDT NTDT 10 years ( 15 years in deletional HbH disease) LIC Q 1 2 years SF Q 3 months LIC 5 mg Fe/g dry wt (SF 800 μg/l) Yes No DFX 10 mg/kg/day LIC Q 6 12 months (SF Q 3 months) LIC after 6 months > 7 mg Fe/g dry wt (SF > ,000 μg/l) and < 15% decrease from baseline DFX 20 mg/kg/day LIC 3 mg Fe/g dry wt (SF 300 μg/l) Discontinue DFX Taher AT, et al. TIF Guidelines TIF guidelines. Available from:

71 Agenda Non-transfusion-dependent thalassemia (NTDT) Morbidity in NTDT Clinical complications profile Ineffective erythropoiesis and anemia Iron overload and target organ damage Hypercoagulability and vascular disease Management of NTDT Splenectomy Transfusions Fetal hemoglobin induction Iron chelation Novel agents

72 Takeaways Ineffective erythropoiesis alongside anemia are the hallmarks of disease process in NTDT These in turn lead to increased intestinal iron absorption (iron overload) and a hypercoagulable state Iron overload and hypercoagulability are associated with a multitude of serious clinical complications involving almost every organ system Multiple morbidity in NTDT has significant impact on quality of life Splenectomy should only be performed in indicated situations as it is associated with a variety of adverse events Beneficial roles of tailored transfusion therapy, iron chelation therapy, and fetal hemoglobin induction in this patient population are established Data from novel therapies in NTDT patients are awaited