APC/DTC Briefing Document

Transcription

1 London New Drugs Group Page 1 APC/DTC Briefing Document Review of Iron Chelators (deferasirox, deferiprone & desferrioxamine) for Iron Overload Contents Background 2 Iron Overload 4 Iron Chelating Agents 5 Deferasirox Clinical Evidence 8 Deferasirox Adverse Effects 13 Deferasirox Precautions & 16 Deferasirox Special Populations Economic Considerations 17 Deferasirox Place in Therapy 19 References 20 Produced for the London New Drugs Group Contact: Katie Smith & Victoria Gibson East Anglia Medicines Information Service Ipswich Hospital Heath Road Ipswich IP4 5PD Tel: Katie.smith@ipswichhospital.nhs.uk Further copies of this document are available from URL: Or Summary Iron overload is the main complication of regular blood transfusions which are used in the management of several conditions including the beta (β) thalassaemias, sickle cell disease, myelodysplastic syndrome and other rare anaemias. There are 3 iron chelating agents available in the UK; deferasirox and deferiprone which are administered orally and desferrioxamine which is administered via subcutaneous infusion. Desferrioxamine has the widest range of licensed indications. Deferiprone is not widely used in clinical practice because of its side effect profile. Guidance produced by the UK Thalassaemia Society in 2005 (before deferasirox was available) recommends that all children with transfusion dependent thalassaemia should be started on subcutaneous desferrioxamine infusions after receiving blood transfusions or when the serum ferritin level is consistently greater than 1000mcg/l. Patients who adhere to treatment usually complete their education, work and are expected to live at least until their mid forties. Data indicate that only about 50% of UK patients are able to adhere fully to subcutaneous desferrioxamine, because infusions are painful and poorly tolerated. Complications of untreated iron overload include iron related cardiomyopathy, hypothalamic and pituitary damage and reduced life expectancy. In practice, adherence to oral therapy is much improved in those who have difficulties with desferrioxamine infusions. Deferasirox is a new oral iron chelating agent licensed for the treatment of chronic iron overload due to frequent blood transfusions in patients aged 6 years and over with beta thalassaemia and as a suitable alternative for the treatment of iron overload in patients with other anaemias including sickle cell patients and patients aged 2-5 years, when desferrioxamine therapy is not tolerated contraindicated or inadequate. There has been one pivotal, non-inferiority, randomised, phase 3 trial in 586 beta thalassaemia patients aged over 2 years, who received either oral deferasirox 5-30mg/kg once daily or subcutaneous desferrioxamine infusions 20-50mg/kg daily. The success rate in reducing or maintaining liver iron concentration (LIC) levels was 52.9% for the deferasirox group (n=146) 66.4% for This is an NHS document not to be used for commercial and marketing purposes. Produced to inform local decision making using the best available evidence at the time of publication. The information in this document may be superseded in due course.

2 Page 2 The thalassaemias are a group of genetic disorders of haemoglobin synthesis, which result from production abnormalities in the globin chains of haemodesferrioxamine (n=184) group. The difference between the groups was 13.5% (95% CI , -5.4) i.e. 13.5% in favour of desferrioxamine. The confidence interval indicates that the response rate to deferasirox could be up to 21% lower than that achieved with desferrioxamine, therefore in the primary efficacy population, non-inferiority of deferasirox to desferrioxamine was not demonstrated. The authors explain this as being possibly due to proportionally lower doses of deferasirox relative to desferrioxamine, being administered to patients with LIC values <7mg Fe/g dry weight. The results of an open label subgroup analysis of 381 patients with liver iron concentration (LIC) 7mg Fe/g dry weight who received deferasirox doses above 20mg/kg, showed that the difference in success rate between the two groups was -0.3% (-10.2, 9.6). The effectiveness of deferasirox in treating transfusion dependent iron overload in patients with sickle cell disease has been studied in 1 published clinical trial and was found to be as effective as desferrioxamine. Use of deferasirox in patients with myelodysplastic syndrome has been studied but there are currently no published clinical trials. Since the aetiology of iron overload in patients with sickle cell disease and myelodysplastic syndrome is the same as beta thalassaemia (i.e. transfusion overload) it is widely agreed that deferasirox will be equally as efficacious. Gastro-intestinal disturbances were reported in about 26% of patients on deferasirox. Raised serum creatinine levels and deaths due to acute renal failure and cytopenias have been reported in the United States. The relationship of these events to treatment with deferasirox is uncertain. The Scottish Medicines Consortium (SMC) has accepted deferasirox for restricted use within NHS Scotland for the treatment of chronic iron overload associated with rare acquired or inherited anaemias requiring recurrent blood transfusions. The SMC does not recommend deferasirox for patients with myelodysplastic syndromes because the economic case has not been demonstrated. Approximate annual costs for treatment of beta thalassaemia with deferasirox are 12,220-19,859 compared to 2,309-12,937 for desferrioxamine and 4,160 for deferiprone. Additional costs that need to be considered with desferrioxamine are use of home care delivery or nurse services and the cost of an infuser device plus the associated equipment. The total annual cost of infused desferrioxamine in the UK has been estimated as 17,913. Despite its low cost, deferiprone is rarely used because of the side effect profile. Based on evidence from clinical trials, experience of clinicians and patients and the financial implications of the various treatment options, deferasirox is considered to be a first line treatment for iron overload. Background Iron overload is the main complication of regular blood transfusions which are used in the management of several conditions including the haemoglobinopathies; beta (β) thalassaemias and sickle cell disease, and myelodysplastic syndrome and other rare anaemias. 1, 2 Haemoglobinopathy refers to a range of genetically inherited disorders of red blood cell haemoglobin and includes sickle cell disease and the thalassaemias. Sickle cell disease and beta thalassaemia major are two of the commonest forms of this disorder. 3 They occur most commonly in populations whose ancestors come from Africa, Asia, Mediterranean, Middle and Far East and because of migration and intermarriage they are now seen in the Caribbean, South America and other parts of the world including Britain and North America. 3, 4

3 Page 3 globin. 1,2 They are divided into the α, β, δβ, or εγδβ thalassaemia, according to which globin chain is produced in reduced amounts. 1,2 The clinical features of the thalassaemias are similar but the beta thalassaemias are the most clinically significant and common type of thalassaemia. An estimated 700 people with thalassaemia major live in the UK with 20 babies born each year with beta thalassaemia. 4 Beta thalassaemia major results from absent or reduced β chain production. The α-chain synthesis proceeds at a normal rate, which causes an imbalanced globin chain synthesis. The excess α-chains are unstable and as a result the red blood cells do not form correctly and are eventually destroyed prematurely. 1, 2 Beta thalassaemia major usually presents by 2 years of age, with a haemoglobin level of 6g/dl and symptoms of severe anaemia. 1, 2 This results in increased erythropoietin production, which can cause bone marrow hyperactivity leading to serious deformities of the skull and long bones and pathologic fractures. It can also impair growth and delay or prevent puberty. Splenomegaly results from increased abnormal red blood cells in the circulation. 1, 2 Life expectancy is decreased in people with untreated beta thalassaemia major. People who don t receive adequate treatment often do not survive beyond puberty. 1, 2 All patients require standard treatment consisting of regular blood transfusions given every 3-4 weeks, to correct the anaemia. Bone marrow transplant may be an alternative treatment option but this is confined to 25% of patients aged 17 years or less with an HLA matched sibling. 5 There are three types of sickle cell disease: sickle cell anaemia, haemoglobin sickle cell disease and sickle beta thalassaemia which are caused by inherited abnormal haemoglobin formation due to the presence of haemoglobin S. 5-7 Sickle shaped red blood cells clog capillaries, causing organ ischaemia. Acute pain (crises) may develop frequently. Infection, bone marrow aplasia, or lung involvement can develop acutely and be fatal. Crises are treated with analgesics and other supportive measures including blood transfusions when there is a cycle of closely spaced painful crises. 7 Transfusions are used to prevent long term recurrent cerebral thrombosis in children <18 years of age who have suffered at least one stroke. 5 Around 10% of patients develop a stroke and these patients will all be on lifelong 4 weekly transfusions. 5 A further 5-10%, who have not suffered a stroke, have high risk factors for stroke and these children also receive 4 weekly infusions for an indefinite period, but is likely to last throughout childhood. 5,8 In 1973, the average life span of sickle cell patients was 14 years, now it is 50 years if they are managed adequately. 6 An estimated 10,500 adults and children with sickle cell disease live in the UK with 445 new births occurring each year 4 Sickle cell disease is now more common than other familiar genetic disorders such as cystic fibrosis and haemophilia. 4 Myelodysplastic syndrome (MDS) affects the bone marrow and results in ineffective and/or inappropriate haematopoiesis. 9 This can lead to anaemia, which is treated with regular blood transfusions; neutropenia and/or thrombocytopenia. Symptoms tend to reflect the most affected cell line and may include pallor, weakness and fatigue, fever and infections; and increased bruising, petechiae, epistaxis and mucosal bleeding. Splenomegaly and hepatomegaly are common. MDS may convert to acute myeloid leukaemia. 9 Over 90% of cases of MDS occur in individuals over 60 years of age. 10 Most patients will de-

4 Page 4 velop transfusion dependency, of which over 50% are likely to benefit from desferrioxamine. Deferiprone is not recommended for routine use due to lack of published data and concerns over safety and efficacy Iron Overload Each unit of blood contains iron which can not be excreted from the body. 15 A typical thalassaemia patient will accumulate mg/kg of iron per day. 15 Excessive iron is deposited in body tissue as haemosiderin and is very toxic. Free non-transferrin bound iron also has the potential to form free radicals which can cause oxidative damage. 16 Excess iron accumulates in nearly all tissues and severe damage can occur to the liver, heart, thyroid, pituitary, hypothalamus, pancreas and joints. Once the body has accumulated 12-24g of iron, significant clinical manifestations of iron toxicity will occur. 15 Total body iron content can reach as high as 50g, compared with normal levels of 2.5g in women and 3.5g in men. 1 Symptoms of iron overload do not usually occur until irreversible organ damage has occurred. 1 They include fatigue, hepatomegaly, bronze skin pigmentation, loss of libido, arthralgias, diabetes or cardiomyopathy. 1 Liver involvement can lead to elevated aminotransferase (ALT and AST) levels, bridging fibrosis or cirrhosis. 1 All patients who require regular blood transfusions will also require iron chelation therapy. 5 Children who have been both adequately transfused and chelated may grow and develop normally, pass through a normal puberty and survive to adult life. 15 Iron overload is identified by elevations in plasma iron concentration, transferrin saturation and plasma ferritin concentration. 1 Table 1 shows the normal therapeutic ranges for these, as well as the levels seen in iron overload. Liver iron concentrations above 7mg Fe/ g dry weight are associated with increased morbidity and mortality. 15 Serum ferritin is also raised in several inflammatory conditions. Table 1: Haematological and histological markers of iron overload Normal Therapeutic Range Iron overload patients Plasma iron Children: mcg/ml >100mg/ml concentration Adults: mcg/ml Transferrin saturation 15-50% % Plasma ferritin Child 2-15yrs: 7-142ng/ml Men: ng/ml Women: ng/ml >1000ng/ml Total iron binding 4.5-8mmol/l ( Reduced capacity (TIBC) mcg/ml) Liver iron concentration (milligrams of iron per dry weight of liver tissue; Fe/g dry weight) Men: 0.2-2mg Fe/g dry weight Women: mg Fe/g dry weight >2mg/g dry weight

5 Page 5 Persistent serum ferritin levels of >2500mcg/l are associated with an increased risk of cardiac disease and death. A level of mcg/l is generally recommended in thalassaemia patients. 15 It has been shown in MDS patients that serum ferritin levels in excess of 1000mcg/l are related to reduced survival. Overall survival is decreased by 30% for every 500mcg/l over the 1000mcg/l threshold. 12 Direct measurement of liver iron, by liver biopsy, provides an accurate measure of body iron stores, however this method is invasive, iron deposition can be patchy and the results 15, 17 show poor reproducibility. Direct non-invasive measurement of hepatic iron stores is possible by magnetic susceptometry (with the use of a super conducting quantum interference device or SQUID) and is either equivalent to or more accurate than the measurement of hepatic iron by liver biopsy. 17 Only four centres worldwide can currently use this technique. 17 A more recently developed alternative is R2 magnetic resonance imaging (MRI) which has been approved by the FDA and EMEA for measurement of hepatic iron. Hepatic iron may not accurately represent iron deposition in other vital organs such as the heart. A parameter, determined by MRI, known as T2, can be used to measure liver and myocardial iron accumulation. T2 has 15, 17 not been validated by the FDA or EMEA although it is the method of choice to measure iron load in the heart. In clinical practice iron load is monitored by changes in plasma ferritin concentrations. 1 Iron Chelating Agents There are three iron chelating agents licensed in the UK. 19 The licensed indications for each are shown in table Desferrioxamine is administered by subcutaneous infusion and there is over 40 years experience of its use in clinical practice. 15 Deferiprone was the first oral iron chelating agent licensed in the UK and has 15, 21 been available since It is licensed for use in beta thalassaemia patients over the age of One randomised controlled trial has shown deferiprone to be comparable with desferrioxamine as a single chelator agent 15, 23 over a period of one year. Long term studies show that liver iron is not adequately controlled by deferiprone monotherapy in a significant proportion of patients. 15 Deferiprone may be more effective than desferrioxamine at removing cardiac iron and consequently patients taking it may have a survival benefit over desferrioxamine. 15, There is some anecdotal evidence that adherence to therapy is improved in patients treated with deferiprone because patients prefer the oral route of administration. 15 Adverse effects of deferiprone include agranulocytosis (0.5-1% of patients), arthropathy (4-50% of patients), nausea, neutropenia and intermittent increases in transaminase levels. 15 Agranulocytosis has been fatal especially when deferiprone has been used in non-licensed indications. 27 The patients neutrophil count must be monitored every week in an attempt to identify vases of neutropenia and agranulocytosis. 21 The reported findings during clinical trials that deferiprone was associated with liver 15, 17 fibrosis has not been confirmed. A recent Cochrane review found that both deferiprone and desferrioxamine produce a significant reduction in iron stores in transfusion dependent, iron overloaded people and that there was no evidence to suggest that either was more clinically efficacious, although there was little consistency between outcomes and little information to fully assess the methodological quality of most of the included trials. 28 The review concluded that there was insufficient

6 Page 6 Table 2: Licensed indications for iron chelating agents Drug Deferasirox Deferiprone Desferrioxamine Indications 1. Treatment of chronic iron overload due to frequent blood transfusions ( 7ml/kg/month of packed red blood cells) in patients with beta thalassaemia major aged 6 years and older. 2. Treatment of chronic iron overload due to blood transfusions when desferrioxamine therapy is contraindicated or inadequate in the following patient groups: in patients with other anaemia s, in patients aged 2 to 5 years, in patients with beta thalassaemia major with iron overload due to infrequent blood transfusions (<7ml/kg/month of packed red blood cells). The treatment of iron overload in patients with thalassaemia major when desferrioxamine therapy is contra-indicated or inadequate. 1. Treatment for chronic iron overload, e.g. transfusional haemosiderosis in patients receiving regular transfusions (e.g. thalassaemia major). primary and secondary haemochromatosis in patients in whom concomitant disorders (e.g. severe anaemia, hypoproteinaemia, renal or cardiac failure) preclude phlebotomy. 2. Treatment for acute iron poisoning. 3. For the diagnosis of iron storage disease and certain anaemias. 4. Aluminium overload - In patients on maintenance dialysis for end stage renal failure where preventative measures (e.g. reverse osmosis) have failed and with proven aluminium-related bone disease and/or anaemia, dialysis encephalopathy; and for diagnosis of aluminium overload. evidence to change current treatment recommendations. The review also stated that no major differences in compliance between the two treatments was observed as almost all those in the included trials achieved good to excellent compliance with both deferiprone and desferrioxamine. 28 Guidance produced by the UK Thalassaemia Society in 2005 (before deferasirox was available) recommends that all children with transfusion dependent thalassaemia should be started on subcutaneous desferrioxamine infusions after receiving blood transfusions or when the serum ferritin level is consistently greater than 1000mcg/l. 15 The starting frequency should be 1-2 nights per week and increased over the first year of therapy to the standard dose in childhood of 20-40mg/kg over hours on 5-6 nights per week. 15 Deferiprone therapy and combination therapy with desferrioxamine and deferiprone, should be restricted to patients who have evidence of high iron stores and who are judged to be at high risk of iron related tissue damage, after attempts have been made to optimise adherence with des-

7 Page 7 ferrioxamine. 15 The main problem with iron chelation therapy is compliance to regular subcutaneous infusions of desferrioxamine. They are unpopular and are often resisted by patients. 5, 15, 29 The infusions are time consuming to set up, they are painful, requiring the introduction of a subcutaneous needle on each occasion which can be distressing, followed by continuous attachment to an infuser device for , 15 hours. As beta thalassaemia and sickle cell disease are inherited conditions, many families have several affected children. Complications of iron toxicity are iron related cardiomyopathy, hypothalamic and pituitary damage which can result in cardiac arrhythmias, short stature, delayed or absent puberty and infertility respectively. Other endocrine problems including glucose intolerance and hypothyroidism have also been directly linked to transfusion iron overload, and the ability of the patient to adhere to iron chelation therapy. 15 If untreated, iron overload in thalassaemia major is fatal by the 2 nd or 3 rd decade of life. 15, 30 The consequences of iron overload are the same for both sickle cell and beta thalassaemia patients. Iron overload may be responsible for as many as 20% of sickle cell deaths. 5, 31, 32 Life expectancy in thalassaemia patients, in western countries such as the United Kingdom, is linked to the effectiveness of iron chelation therapy, which in turn is linked to compliance. Patients who adhere to treatment usually complete their education, work and find a partner and are expected to live at least until their mid forties. Data indicate that only about 50% of UK patients are able to adhere fully to current iron chelation therapy and that less adherent patients gain on average only 10 years 30, 33, 34 of life. Full prescribing instructions can be found in the summary of product characteristics (SPC) for deferasirox (Exjade). 20 It is recommended that treatment with deferasirox be started after the transfusion of approximately 20 units (about 100ml/kg) of packed red blood cells or when there is evidence from clinical monitoring that chronic iron overload is present (e.g. serum ferritin >1,000mcg/ l). Doses (in mg/kg) must be calculated and rounded to the nearest whole tablet size. Deferasirox is available in three tablet strengths (125mg, 250mg and 500mg). 20 The recommended initial daily dose of deferasirox is 20mg/kg body weight. It is recommended that serum ferritin be monitored every month and that the dose of deferasirox is adjusted, if necessary, every 3 to 6 months based on the trends in serum ferritin. Dose adjustments may be made in steps of 5 to 10mg/kg and are tailored to the individual patient's response and therapeutic goals (maintenance or reduction of iron burden). Doses above 30mg/kg are not recommended because there is only limited experience with doses above this level. If serum ferritin falls consistently below 500mcg/l, an interruption of treatment should be considered. 20 Deferasirox must be taken once daily on an empty stomach at least 30 minutes before food, preferably at the same time each day. The tablets are dispersed by stirring in a glass of water or orange or apple juice (100 to 200ml) until a fine suspension is obtained. After the suspension has been swallowed, any residue must be re-suspended in a small volume of water or juice and swallowed. The tablets must not be chewed or swallowed whole. 20 The dosing recommendations for paediatric patients are the same as for adult patients. Changes in weight of paediatric patients over time must be taken

8 Page 8 into account, when calculating the dose. In children aged between 2 and 5 years, exposure is lower than in adults. This age group may therefore require higher doses than are necessary in adults. However, the initial dose should be the same as in adults, followed by individual titration. 20 Clinical Evidence for Deferasirox There has been one pivotal phase III study and five phase II studies assessing the efficacy and safety of deferasirox The only adequate and well controlled randomised trial was study The reason for this is that the EMEA agreed that the safety and efficacy of deferasirox could be extrapolated from beta thalassaemia as a model disease. 18 The other trials were either exploratory or uncontrolled. 35, 36 For randomised trials, blinding was not performed as it was proposed that subcutaneous administration of placebo for 48 weeks to patients randomised to deferasirox was unacceptable. 35, 36 Phase III trial In study 0107, a phase III, randomised, active controlled, open label, trial, 586 patients, older than 2-years with beta thalassaemia and transfusional iron overload were randomised to receive either oral deferasirox (n=296) or subcutaneous infusions of desferrioxamine (n=290) for one year. 37 The initial dose of deferasirox or desferrioxamine was dependent on the liver iron concentration (LIC) at screening as shown in table 3. Deferasirox was taken once daily every morning, 30 minutes before breakfast. Desferrioxamine was administered as a subcutaneous infusion for 8 hours on five consecutive days per week. The initial dose of deferasirox and desferrioxamine were to remain unchanged during the 1-year study period unless the evaluation of safety and efficacy markers indicated that dose adjustment was necessary. 37 Patients with a baseline LIC of 2-7mg Fe/g dry weight were allowed to continue their previous doses of desferrioxamine, even if doses were higher than specified by the study protocol as it was considered unethical to reduce the dose of well managed patients to a sub-optimal level for the purpose of a trial. 18 Blood transfusions were regularly administered during the study period according to the patient s requirements; the amount of blood and iron received by each patient was carefully assessed. 37 The main objective of the trial was to demonstrate non-inferiority of deferasirox compared to desferrioxamine regarding its effects on liver iron concentration (LIC) across all groups. 37 The primary endpoint was success rate of deferasirox at reducing or maintaining LIC levels as demonstrated by the change from baseline in LIC levels. The success criteria for the primary endpoint are shown in table 4. Table 3: Dose of deferasirox or desferrioxamine to be used based on screening LIC Screening LIC Dose (mg Fe/g dw) Deferasirox Desferrioxamine 2-3 5mg/kg/day 20-30mg/kg/day >3 7 10mg/kg/day 25-35mg/kg/day > mg/kg/day 35-50mg/kg/day >14 30mg/kg/day 50mg/kg/day

9 Page 9 Table 4: Success criteria for primary endpoint in study 0107 Baseline LIC (Fe/g dw) LIC after 1 year (Fe/g dw) Success Failure 2-<7mg 1-<7 mg < 1mg or 7mg 7-<10mg 1-<7mg < 1mg or 7mg 10mg Decrease in LIC 3mg Decrease in LIC < 3mg In most patients (84%), LIC was determined by liver biopsy. In patients with contra-indications to liver biopsy, mainly children, SQUID analysis was performed. 37 Non-inferiority was demonstrated if the lower limit of the 95% confidence interval for the difference in success rate between deferasirox and desferrioxamine was above -15% in the primary efficacy population. 37 Secondary endpoints included Changes in mean serum ferritin levels. Changes in net body iron balance. Comparison of results by dose. Safety assessments. The results of study 0107 are shown in tables 5 and 6. The mean age of the patients was 17 years with a median of 15 years. Many of the patients were under 16 years of age (51%). Approximately two thirds of each group had a baseline LIC value of 7mg Fe/g dry weight or more at baseline; 69% in the deferasirox group and 68% in the desferrioxamine group. Most patients completed 1 year of therapy: 541 (92.3%) of 586 underwent both baseline and 1 year LIC assessments. Discontinuations were relatively similar in the groups receiving deferasirox (n=17) and desferrioxamine (n=12). 37 In the primary efficacy population, non-inferiority of deferasirox to desferrioxamine was not demonstrated. The success rate for deferasirox group (n=146) was 52.9% and for desferrioxamine (n=184) was 66.4%. The difference between the groups was -13.5% (95% CI -21.6, -5.4) i.e. 13.5% in favour of desferrioxamine. The confidence interval indicates that the response rate to deferasirox could be up to 21% lower than that achieved with desferrioxamine. Therefore in the primary efficacy population, non-inferiority of deferasirox to desferrioxamine was not demonstrated. The authors explain this as being possibly due to proportionally lower doses of deferasirox relative to desferrioxamine being administered to patients with LIC values <7mg Fe/g dw. 37 Non-inferiority was demonstrated in the group of patients who were allocated to the higher dose groups. In a subgroup analysis of 381 patients with LIC 7mg Fe/g dry weight who received deferasirox doses above 20mg/kg and desferrioxamine 35mg/kg, the difference in success rate between the two groups was -0.35% [-10.2, 9.6] in favour of desferrioxamine. If only patients who had LIC determined by biopsy are included, the success rate for deferasirox (n=176) was 59.7% and with desferrioxamine (n=179) was 58.7% with a difference and 95% CI of 1.0% (-9.2, 11.2). In both these analyses, the lower limit of the CI is above -15%. 37 In patients with baseline LIC values less than 7mg Fe/g dry weight, who received either deferasirox at doses of 5mg or 10mg/kg per day (n=34), the success rate was 40%. The success rate in the

10 Page 10 Table 5: Primary efficacy results, study 0107: Success rates based on change in LIC levels Deferasirox Desferrioxamine Dose 5 mg/ All All kg mg/kg mg/kg mg/kg Patients with Biopsy & SQUID results No. of patients who met the success criteria Success rate (95% CI) 40% (16.3, 67.7) 40% (28.5, 51.5) 36.7% (26.1, 47.3) 74.1% (66.0, 82.2) 52.9% (47.0, 58.8) 66.4% (60.9, 72.0) Difference Not applicable -13.5% (-21.6, -5.4) No. of patients with baseline LIC levels < 7mg Fe/g dry weight n/a n/a Success rate (95% CI) 40% (16.3, 67.7) 40% (28.5, 51.5) n/a n/a 40% (29.6, 50.4) 82% (74.8, 90.7) Difference Not applicable -42.8% (-55.9, -29.7) No. of patients with baseline LIC levels 7mg Fe/g dry weight n/a n/a Success rate (95% CI) n/a n/a 36.7% (26.1, 47.3) 74.1% (66.0, 82.2) 58.6% (51.7, 65.6) 58.9% (52.0, 65.9) Difference Not applicable -0.3% (-10.2, 9.6) comparable desferrioxamine group (n=72) was 82.8%. The between group difference was -42.8% (55.9, -29.7%). 37 A mean reduction in LIC of -5.3 ± 8.0mg Fe/g dry weight (p<0.001) occurred in the deferasirox group and of -4.3 ± 5.8mg Fe/dry weight in the comparable desferrioxamine group. The difference between the two groups was not statistically significant (p=0.367). 37 Table 6 shows the changes from baseline in serum ferritin, liver iron concentration and iron excretionintake ratio in relation to dose. The mean values for these markers at the start and the end of the study are not included in the study report so it is unclear if any of the reductions brought the patients back into the normal range. For patients receiving 5mg/kg/day and 10mg/kg/day, all parameters indicated an increase in iron burden; for those receiving 20mg/kg/day deferasirox, the parameters indicate that iron burden was unchanged and for those receiving 30mg/kg/day they indicated that iron burden was reduced. These results seem to suggest that the higher the baseline liver iron concentration, the higher the dose used, and the greater the effects seen. This study does not show that deferasirox is noninferior to desferrioxamine. Doses of 20mg/kg/day and 30mg/kg/day both reduce body iron levels however; this

11 Page 11 Table 6: Comparison of changes in ferritin, LIC and the iron excretion-intake ratio by dose (mg/kg/day) Baseline LIC Mean serum ferritin (mcg/l) Change from baseline Mean LIC, (mg Fe/g dry weight) Mean iron excretionintake ratio 3mg Fe/g dry weight or less Deferasirox 5mg ± ± ± Desferrioxamine 30mg ± ± ± >3 to 7mg Fe/g dry weight Deferasirox 10mg +833 ± ± ± Desferrioxamine 35mg +32 ± ± ± >7 to 14mg Fe/g dry weight Deferasirox 20mg -36 ± ± ± Desferrioxamine 40mg -364 ± ± ± >14mg Fe/g dry weight Deferasirox 30mg -926 ± ± ± Desferrioxamine 50mg ± ± ± was only similar to reductions seen with desferrioxamine, in the 30mg/ kg/day subgroup. Doses of 5mg/kg/ day and 10mg/kg/day were not effective. Patient satisfaction was assessed at 4 and 24 weeks and at the end of the study. Significantly more patients on deferasirox previously treated with desferrioxamine indicated they would be willing to continue deferasirox compared with patients on desferrioxamine who would be willing to continue desferrioxamine (85.8% vs. 13.8%, p<0.001). 38 Phase II studies Study 0104 was a randomised, double blind, placebo controlled, dose escalation study involving 23 patients with beta thalassaemia and transfusional iron overload who received either placebo (n=5) or deferasirox at doses of 10mg/kg/day (n=5), 20mg/ kg/day (n=6) and 40mg/kg/day (n=7) for 12 days. 39 All patients had previously received desferrioxamine 20mg/kg/day for at least 4 weeks prior to screening. And had serum ferritin values between ng/ml and LIC 3.5mg Fe/g dry weight. Net iron excretion and change from baseline in serum ferritin were used to as efficacy measures. All three doses increased net iron excretion. 39 Study 0105 was a 48 week, randomised, open label, safety study involving 71 beta thalassaemia patients with transfusional iron overload who received 10mg/kg/day or 20mg/kg/day of deferasirox or 40mg/kg/day of desferrioxamine. 40 Efficacy was assessed as secondary objective by determination of LIC levels. Decreases in liver iron levels were comparable in the deferasirox 20mg/kg and desferrioxamine groups; baseline values of 8.5 and 7.9mg Fe/g dry weight decreased to 6.6 and 5.9 Fe/ g dry weight respectively by week 48. Deferasirox 10mg/kg/day resulted in only a minimal fall in LIC of -0.4mg Fe/g dry weight. Over the study duration, serum ferritin remained stable in the deferasirox 20mg/kg group, but increased modestly in the 10mg/kg group. 40

12 Page 12 Study 0106 was a 48 week, open label, non comparative study, involving 40 paediatric beta thalassaemia patients with chronic iron overload. All patients commenced deferasirox, at a dose of 10mg/kg/day regardless of their baseline LIC. 41 The primary endpoint was safety and tolerability of deferasirox. From study week 12, doses could be increased by 5mg/kg/ day, up to a maximum of 30mg/kg/ day on a case by case basis. The mean deferasirox dose over the study period was 11.3mg/kg/day (median 10.3, range ). Overall LIC increased gradually from week 12 as mean daily iron intake was higher than excretion. The safety profile was as expected from previous trials and there was no difference between children and adolescents. The study authors concluded that the dose used in 18, 41 the trials was too low. Study 0108 was a multicentre, open label, non comparative trial which aimed to evaluate the efficacy and safety of deferasirox in paediatric and adult patients with congenital and acquired anaemia complicated by chronic iron overload who could not be appropriately or adequately treated with desferrioxamine. 35, 42, 43 Deferasirox treatment and success rate criteria were identical to study Effective chelation with deferasirox in patients was demonstrated if the success rate was above 50%. The main secondary endpoint was change from baseline in LIC. The results of study 0108 are detailed in tables 7 and 8. The success rate was higher than 50% but this was not statistically significant. The reduction in LIC was statistically significant. 35, 42, 43 The results of a subset of MDS patients in study 0108 with MDS have been presented as abstracts and posters but not fully published papers, so they can not be critically appraised. Table 7: Study 0108; Success rates based on change in LIC Per protocol population Beta thalassaemia Rare anaemias All patients Intent to treat population Patients with Biopsy & SQUID results Success rate 56.3% 56.5% 56.4% 50.5% (95% CI) (45.4, 67.1) (45.9, 67.0) (48.8, 63.9) (43.3, 57.8) p-value - - p=0.051 p=0.441 Patients with baseline LIC <7mg Fe/g dry weight Success rate 20% 61.5% 43.5% 40.0% (95% CI) (2.5, 55.6) (31.6, 86.1) (23.2, 65.6) (21.1, 61.3) p-value Not stated Patients with LIC 7mg Fe/g dry weight Success rate 61.4% 55.6% 58.5% 52.2% (95% CI) (50.0, 72.8) (44.1, 67.0) (50.3, 66.6) (44.4, 60.0) p-value - - p=0.022 p=0.289

13 Page 13 Table 8: Change from baseline in LIC (mean ± SD) Beta thalassaemia Other anaemias All patients Total number of patients Change in mean LIC from -4.7± ± ± 7.7 baseline P value p<0.001 p<0.001 p<0.001 Patients with baseline LIC <7mg Fe/g dry weight Change in mean LIC in patients with baseline LIC <7mg Fe/g dry 6.0 ± ± ± 4.1 weight Number of patients with baseline LIC >7mg Fe/g dry weight Change in mean LIC in patients with baseline LIC >7mg Fe/g dry -6.1 ± ± ± 7.4 weight P value p<0.001 p<0.001 p<0.001 There were 47 patients with MDS in the study of which 29 completed. 78.7% of patients had a baseline LIC 7mg Fe/g dw. Deferasirox decreased mean LIC by 5.7mg Fe/g dw (SD ± 6.3). Reasons for noncompletion included death considered due to underlying disease (4), consent withdrawal (6), drug no longer required (1) and adverse events 45, 46 (7). In study 0109, a randomised, multicentre, open label study, 195 patients with iron overload related to sickle cell disease; aged 2 years and above, received either deferasirox 5-30mg/ kg/day (n=32) or desferrioxamine 20-60mg/kg/day (n=63). 44 The primary objective of the trial was to assess the safety and tolerability of deferasirox. Efficacy was assessed as a secondary objective by estimating absolute and relative LIC change. The change from baseline in LIC for deferasirox was -3.0 ± 6.2mg Fe/g dry weight and for desferrioxamine was -2.8 ± 10.4mg Fe/g dry weight. The difference in the response to deferasirox and desferrioxamine was not statistically significant (p=0.669) in patients with sickle cell disease and baseline LIC <7mg Fe/g dry weight, who received the lower doses of deferasirox and desferrioxamine. Mean LIC was also maintained. The results of study 0109 indicate comparative efficacy of desferasirox and desferrioxamine. 44 Deferasirox: adverse effects In study 0107, 254 (85.8%) patients developed adverse events in the deferasirox group compared with 246 (84.8%) in the desferrioxamine group. Fifty-two patients experienced serious adverse events in the deferasirox group and 25 patients in the desferrioxamine group. Four deaths were reported in the study; 1 in the deferasirox group and 3 in the desferrioxamine group these were all related to underlying disease rather 35, 37 than the drug. During clinical trials, increases in serum creatinine >33% on 2 consecutive occasions occurred in about 36% of patients. Increases above the upper limit of the normal range on 2 occasions occurred in 2.4% of patients in study 0107, 2.3% of patients in study 0109 and 10.3% of patients in study 0108.

14 Page 14 These increases were dose dependent and occurred mainly with daily doses of mg/kg. Lower doses of 5-10mg/kg were better tolerated, therefore the increases may be linked to excessively rapid removal of iron. About two thirds of the patients showing serum creatinine increases returned below the 33% level without dose adjustment. 68 patients had a dose reduction or adjustment due to this. In 41.2% (28 out of 68) of these patients, a dose reduction led to a decrease in the serum creatinine. However, in all but 2 of these cases the values remained above baseline. In a third of these patients, the serum creatinine increase did not always respond to a dose reduction or interruption. 14, 24, 26, The preventability and long term consequences of the modification of the renal function have been questioned. No demonstration of risk prevention by progressively and slowly increasing the dose has been shown. The mechanism of these renal alterations can not be explained. No effective measure has been identified to reverse impaired renal function. Interruption of treatment is not always completely efficient. 20, 35 Renal colic/nephroliathiasis was reported less frequently in patients treated with deferasirox compared to desferrioxamine (0% vs. 3% in study 0107 and 0% vs. 0% in studies 0108 and 0109). Proteinuria, dysuria and nephropathy were also observed. The reversibility of renal adverse events 35, 37, has not been established. The number of patients in study 0107 who reported jaundice and hyperbilirubinaemia was higher in the group receiving 30mg/kg of deferasirox. Almost 2/3 patients in the deferasirox group had increases in transaminases during the study. A similar number of patients receiving 5, 10, 20 and 30mg/kg of deferasirox experienced hepato-biliary disorders, suggesting the absence of a dose effect. However, some of these events could be due to underlying disease or to insufficient iron chelation. The incidence of hepato-biliary adverse events such as cholecystitis, biliary colic, and cholelithiasis was higher in the deferasirox group than in the desferrioxamine group (2% vs. 0.8%), although overall the adverse 20, 35 events were uncommon. The percentage of patients who experienced ECG abnormalities and rhythm disorders in study 0107 (during 1 year of therapy) was 4.7% in the deferasirox group vs. 4.1% in the desferrioxamine group. 35 Cardiac dysfunction is a known complication of severe iron overload. 20 Skin rashes are considered to be a common side effect. Rashes are generally of mild severity and resolve, on average, spontaneously in less than one week. When interruption of treatment may be necessary, treatment may be reintroduced after resolution of the rash, at a lower dose followed by gradual dose escalation. 20, 35 Gastro-intestinal disturbances are common and were reported in about 26% of patients on deferasirox in the clinical trials. Nausea, vomiting, diarrhoea and abdominal pain were mainly mild to moderate, transient and generally resolve if treatment continues. Diarrhoea was reported more commonly in paediatric patients. A few cases of ulcer, pancreatitis, GI haemorrhage, haematemesis, dyspepsia, gastritis were reported in 20, 35 patients receiving deferasirox. As with other iron chelators, high frequency hearing loss and lenticular opacities have been uncommonly observed in patients treated with deferasirox. 20 Lens opacities and signs of deafness have been observed at a similar frequency in clinical trials in the deferasirox and desferrioxamine groups. 35 Thirteen (13) patients permanently discontinued therapy in study 0107 due to

15 Page 15 adverse events - 9 in the deferasirox group and 4 in the desferrioxamine group. Therapy discontinuations were mainly due to skin and liver disorders. 35 In post marketing surveillance, two fatal cases involving renal failure have been spontaneously reported in 47, 48 the United States. Both patients had underlying disease: the first case of fatal renal disease occurred in a patient with acute lymphocytic leukaemia but had no history of renal disease and no concomitant nephrotoxic drugs. The second one developed sepsis with multi-organ dysfunction. Both cases occurred shortly after the initiation of deferasirox and the fatal outcome occurred rapidly after the symptoms began. A possible causal relationship with deferasirox could not be excluded. The frequency of acute renal failure can not be calculated based on these spontaneous reports. 20 Additionally, there were post marketing reports of cytopenias, including agranulocytosis, neutropenia and thrombocytopenia in patients treated with deferasirox where some of the patients died. The relationship of these episodes to treatment with deferasirox is uncertain. The frequency of these events can not be calculated and they are not listed as adverse effects in the SPC. 20 Most of the affected patients had preexisting haematologic disorders that are frequently associated with bone marrow failure. Further, cases of leukocytoclastic vasculitis, urticaria, and hypersensitivity reactions (including anaphylaxis and angioedema) were reported. A pooled interim analysis of deferasirox treatment from the pivotal study and ongoing extension phases of four multicentre, open label trials has recently 49, 50 been presented. Data on patient characteristics and deferasirox dosing are presented in table 9. These results have not been fully published and consequently the results cannot be fully evaluated. The crossover cohort were those patients who received desferrioxamine and were crossed over to deferasirox at the end of the core phase (crossover cohort). To date, 833 patients (80.6%) remain on treatment with deferasirox, of which 543 (77.2%) are in the deferasirox cohort and 290 (87.9%) are in the crossover cohort. The primary reasons for discontinuation were adverse effects/abnormal laboratory parameters/ death (n=92), consent withdrawal (n=62) and unsatisfactory therapeutic effect (n=29). Adverse effects led to discontinuation in 48 patients. The most common drug related adverse effects causing discontinuation were increased ALT, diarrhoea, pruritic rash, proteinuria and/or glycosuria. Drug related adverse effects occurred early in treatment and were mainly Table 9 patient characteristics and deferasirox dosing Deferasirox Crossover All patients cohort cohort Total number of patients Age 2-16 years Age 16 years Median deferasirox exposure, (2.5) 78.3 (1.5) 91.0 (1.8) weeks (years) Mean deferasirox dose ± SD, mg/day 20.5 ± ± ± 6.1

16 Page 16 transient and mild to moderate severity. The pattern and incidence rates of adverse effects in paediatric and adult populations were similar in the deferasirox and crossover cohorts. Increased ALT levels that exceeded 5 times the ULN at 2 consecutive visits were noted in 80 (7.7%) patients. There is currently no published long term safety data beyond 2.6 years. Precautions Serum creatinine should be assessed twice prior to initiating treatment. 20 Serum creatinine, creatinine clearance and/or plasma cystatin C levels should be monitored weekly in the first month of treatment or after modification of therapy and monthly thereafter. Dose reduction of deferasirox should be considered if there is a rise in serum creatinine by >33% above baseline on 2 consecutive occasions. Full details regarding monitoring of renal function and adjustment of deferasirox dose can be found in the SPC. 20 Extra care is required in patients also receiving other medicinal products which can depress renal function, elderly and patients with pre-existing renal disease. Tests for proteinuria should be performed monthly. 15 Blood counts should also be monitored in line with standard clinical management for patients with haematological disorders that are frequently associated with bone marrow failure and treatment interruption should be considered in patients who develop unexplained cytopenia. Monthly monitoring of serum ferritin is recommended to assess the patient's response to treatment. Liver function test elevations have been observed in patients treated with deferasirox. It is recommended that liver function tests should be performed monthly. If a persistent and progressive increase in serum transaminases occurs that cannot be attributed to other causes, treatment should be interrupted. 20 Auditory and ophthalmic testing is recommended before starting treatment and every 12 months thereafter. 20 Body weight, height and sexual development should be monitored every 12 months in all paediatric patients treated with deferasirox. 20 Cardiac function should be monitored in patients with severe iron overload during long term treatment with deferasirox. 20 The bioavailability of deferasirox is affected by food; it should therefore be taken on an empty stomach at least 30 minutes before food, preferably at the same time each day. 20 The concomitant use of deferasirox with aluminium containing antacid preparations is not recommended. 20 Special populations Deferasirox is not recommended in pregnant or lactating women or in patients with severe hepatic impairment. 20 Deferasirox has not been studied in patients with renal impairment and is contra-indicated in patients with estimated creatinine clearance of <60mls/ minute. 20

17 Page 17 Economic considerations Drug Dose Cost per year Desferrioxamine 20-60mg/kg daily by subcutaneous infusion 2,309-12,937 for 5-7 days every week [ 1,455-8,721] Deferiprone 25mg/kg three times daily 4,160 Deferasirox 20-30mg/kg once daily 12,220-19,859 NHS prices from MIMS September Costs are approximate and are based on an average body weight of 54kg, which has been suggested as the mean patient weight for patients needing iron chelation therapy. 51 The cost in brackets is an example of the treatment costs if medicines were purchased through an agreed NHS contract. Doses are shown for general comparison and do not imply therapeutic equivalence. There are approximately 700 patients with beta thalassaemia major in the UK of which it is estimated that 97% would be eligible for iron chelation. 4 The prevalence of sickle cell disease is estimated to be approximately 10,500 patients the majority of which are in London. 4, 52 It is estimated that 23% of sickle cell patients are eligible for iron chelation. 52 The prevalence of myelodysplastic syndrome is around 3,800 and it is estimated that 13% are eligible for iron chelation. 52 The overall estimate for the number of eligible patients iron chelation in the UK is 7 per 100, While the drug cost of desferrioxamine is relatively low additional costs to the NHS may be incurred e.g. home care delivery or nurse services. In addition the infuser device used may significantly affect cost effectiveness; (e.g. use of a Graseby pump plus associated equipment will add a further 1,500 for the first year rising to an additional 14,000 per annum should an elastomer balloon delivery system be used). The total annual cost per patient of infused iron chelation therapy in the UK has been estimated as 17, A cost utility analysis study conducted by Karnon et al has estimated the resource use and costs for equipment for desferrioxamine treatment to be 7,552 annually per patient. 51 This study was funded by Novartis Pharmaceuticals Limited. The analysis assumed that deferasirox has equivalent efficacy to desferrioxamine and that patients receiving deferasirox are similarly compliant with patients receiving desferrioxamine infusions. Costs relating to monitoring and adverse events were also assumed to be similar. The analysis calculated that deferasirox is expected to gain quality adjusted life years (QALY) at an additional cost of 891 per QALY gained. The additional costs associated with desferrioxamine and used in the analysis are shown in table 10. Feedback obtained during the Scottish Medicines Consortium submission highlighted that the estimate for the number of patients who use a balloon infuser was too high for Scotland and that fewer patients actually use a balloon infuser in practice. Consequently, Novartis recalculated the figures and the SMC recommendation was based on these. In comparison with desferrioxamine, the following was calculated for the Scottish Medicines Consortium. In patients with beta thalassaemia the cost per QALY was around 20,000, in patients with sickle cell disease it was around 30,000 per QALY and in patients with myelodysplastic syndrome it is over 38,000 per QALY. From this, the SMC concluded that the cost effectiveness of deferasirox was acceptable in patients with β thalassaemia and sickle cell disease but the case had not been demonstrated in patients with MDS. 54