Summary of findings
Description of the condition
Sickle cell disease (SCD) is an autosomal recessive genetic disorder caused by a single nucleotide mutation of the haemoglobin β-chain. The worldwide birth rate of people who are homozygous or compound heterozygous for symptomatic SCD is about 2.2 per 1000 births (Angastiniotis 1998; Modell 2008). However, the incidence of this disease varies widely between ethnic groups (Modell 2008). Populations originating from sub-Saharan Africa, the Middle East and parts of the Mediterranean are predominantly affected, but population movement has made SCD a worldwide problem.
The inheritance of one mutated gene (substitution of valine for glutamic acid in the sixth position of the beta chain of the haemoglobin molecule) together with the normal gene for adult haemoglobin (HbA) results in a mutant protein giving rise to a defective variant of haemoglobin. People with the heterozygous state (HbAS or sickle cell trait) are called sickle cell carriers. They are asymptomatic and need neither treatment nor occupational restrictions. The heterozygous state gives some advantage to carriers against malaria infection and has therefore persisted (Weatherall 1998; Richer 2005).
The homozygous state (HbSS) results from inheriting the mutation from both parents. Sickle cell disease can also occur when people are compound heterozygous by inheriting the sickle cell gene from one parent and another variant haemoglobin gene, such as haemoglobin C, D, OArab or E, or a β-thalassaemia gene from the second parent. This gives rise to HbSC, HbSD, HbSOArab, HbSE or HbSβ+ or HbSβ0. All of these genotypes cause clinically significant SCD with chronic haemolytic anaemia and a predisposition to blockage of small blood vessels resulting in painful crises, acute chest syndrome, cerebral infarction and other complications due to the sickling of the red blood cells. However, the overall severity and patterns of organ damage vary widely depending upon genotype (Platt 1991; Miller 2000; Powars 2005).
Unlike people with β-thalassaemia major, who require regular blood transfusions throughout life from soon after birth, the majority of people with SCD require red cell transfusions only occasionally and intermittently. These are required for indications such as the management of acute severe anaemia in children, e.g. due to splenic sequestration or transient Parvovirus B19 induced aplastic crisis, acute chest syndrome, acute stroke or as a prophylactic measure before operations. Red cell transfusions are not usually required for the management of the chronic anaemia, or acute painful episodes (Josephson 2007). By adulthood, the majority of people with SCD have received several red blood cell transfusions for various reasons.
A small but increasing number of people with SCD are on long-term transfusions, most commonly for secondary stroke prevention, but also for primary stroke prevention, or for recurrent pulmonary complications in people who have not responded to hydroxycarbamide. Over the last decade trials have evaluated the effect of regular prophylactic transfusions for the primary prevention of stroke in children with SCD (Adams 1998; Adams 2005). Trans-cranial doppler (TCD) screening is now recommended as routine care. High-risk children are identified by high-flow velocities on TCDs and should be offered long-term transfusion therapy.
Since the body has no physiological mechanism to actively excrete excess iron, repeated blood transfusions lead to an increased body iron burden including iron deposition into the liver, heart, pancreas and other endocrine organs. This mechanism is well known for people with thalassaemia on regular transfusion programs. As studies have shown, people with SCD with repeated blood transfusions can be affected by the same long-term problems due to iron overload (Vichinsky 2005a). Even if there is no solid evidence showing that iron chelation improves clinical outcome in SCD, iron chelation therapy is generally offered to iron-overloaded people with SCD.
Description of the intervention
Deferoxamine (DFO, Desferal
Oral preparations have been highly sought after for many years. In 1987 two studies showed that the orally active iron chelator deferiprone (1,2 dimethyl-3-hydroxypyrid-4-1, also known as L1, CP20, Ferriprox
How the intervention might work
Deferasirox (4-[3,5-bis(2-hydroxyphenyl)-1H-1,2,4-triazol-1-yl]-benzoic acid) also known as CGP 72670, ICL670 or Exjade
Adverse effects known from experiences in people with thalassaemia include gastrointestinal disturbances (nausea, stomach pain or diarrhoea) that are generally mild and a diffuse rash being more common at higher doses (Cappellini 2006). More rarely, fever, headache and cough are encountered. The main adverse effect with the use of deferasirox seems to be a mild to moderate elevation of the creatinine level in about a third of patients. Elevations of liver enzyme levels have also been described with a lower incidence (5.6%) (Cappellini 2006). As with standard therapy (deferoxamine), hearing loss and ocular disturbances including cataracts and retinal disorders have been reported with a very low incidence (< 1%).
Why it is important to do this review
Deferoxamine necessitates a serious commitment from the user and due to its adverse effects deferiprone is only approved as second line therapy in some countries. Thus, much hope is being placed in the newer oral chelator deferasirox which - as studies in people with thalassaemia have shown - apparently offers a promising line of treatment due to its iron chelation properties and safety and tolerability profile (Nisbet-Brown 2003; Cappellini 2006; Piga 2006; Cappellini 2007a). Although these results are presumably generalisable to people with SCD, a thorough evaluation of deferasirox in people with SCD seems to be warranted. To adequately manage the increasing number of people with SCD being regularly transfused due to TCD screening results, a systematic review of the effectiveness and safety of deferasirox looking at people with SCD according to Cochrane standards is needed.
To evaluate the effectiveness and safety of oral deferasirox for management of transfusional iron overload in people with SCD.
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) were considered for this review.
Types of participants
People with SCD (irrespective of genotype, age and from any setting worldwide), who have received repeated red blood cell transfusions in the past or who are receiving regular red blood cell transfusions currently which have resulted in iron overload (defined as ferritin levels of over 1000 ng/ml on at least two occasions).
Types of interventions
For oral deferasirox (all schedules and doses), the following comparisons are conceivable:
- deferasirox compared with no therapy or placebo;
- deferasirox compared with another iron chelating treatment schedule (i.e. deferoxamine or deferiprone or any combination thereof).
These comparisons constitute two separate groups and would have been analysed separately. However, the necessity of chelation therapy in iron-overloaded people is well-established and, if at all, only short-term, e.g. pharmacokinetic studies would be ethically justifiable. Longer-term studies with no therapy or placebo would not suffice the paradigm of equipoise and we did not expect to find, and in fact did not find, any studies comparing deferasirox to no therapy or placebo.
Types of outcome measures
- Time to death (any cause)
- Reduced end-organ damage due to iron deposition
- cardiac failure (necessitating medical treatment)
- endocrine disease (necessitating substitution hormone therapy or treatment of diabetes)
- histological evidence of hepatic fibrosis
- pathological surrogate markers of end-organ damage (i.e. elevated liver enzymes, elevated fasting glucose or pathological oral glucose tolerance test (OGTT), pathological measures (e.g. ejection fraction) in echocardiography)
- Measures of iron overload
- serum ferritin (ng/ml)
- iron levels in biopsies of liver and other tissue (mg/g liver dry weight)
- tissue iron assessment by SQUID (superconducting quantum interference device) (mg/g liver wet weight)
- tissue iron assessment by MRI (magnetic resonance imaging) (ms)
- Measures of iron excretion (urine and faeces) over 24 hours (mg/kg/d)
- Any adverse events
- raised levels of creatinine or kidney failure (above upper normal limit or rise of more than 20% above baseline level)
- skin rash
- gastrointestinal disturbances
- neutropenia or agranulocytosis (absolute neutrophil count (ANC) less than 1000/µl or less than 500/µl)
- raised levels of liver enzymes (above upper normal limit or raise of more than 20% above baseline level) or progression to liver fibrosis
- hearing loss
- eye problems (e.g. retinal toxicity)
- unanticipated adverse events as reported in the primary studies
- Participant satisfaction (measured e.g. by questionnaire) and compliance with chelation treatment (measured by the number of people in each arm that show adequate level of adherence to treatment (intake or application of iron chelator on five or more days per week).
- Cost of intervention per year.
For future updates, data from outcomes not defined a priori but which arise from the review will be collected, if the outcome is considered to be of clinical relevance.
Search methods for identification of studies
No language restriction was applied.
We identified relevant studies from the Cystic Fibrosis and Genetic Disorders Group's Haemoglobinopathies Trials Register using the terms: (sickle cell OR haemoglobinopathies general) AND ICL670(A).
The Haemoglobinopathies Trials Register is compiled from electronic searches of the Cochrane Central Register of Controlled Trials (CENTRAL) (updated each new issue of The Cochrane Library) and quarterly searches of MEDLINE. Unpublished work is identified by searching the abstract books of five major conferences: the European Haematology Association conference; the American Society of Haematology conference; the British Society for Haematology Annual Scientific Meeting; the Caribbean Health Research Council Meetings; and the National Sickle Cell Disease Program Annual Meeting. For full details of all searching activities for the register, please see the relevant section of the Cochrane Cystic Fibrosis and Genetic Disorders Group Module.
Date of the most recent search of the Group's Haemoglobinopathies Trials Register: 13 March 2014.
For this current, updated version, the following additional databases were searched in July 2012 and again in August 2013 (see Appendix 1; Appendix 2 for details): MEDLINE; MEDLINE In-Process; MEDLINE Daily Update; PubMed (limited to “Epub ahead of print”); Embase; Embase Alert; The Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials (Central), Other Reviews (DARE), Methods Studies, Technology Assessments, Economic Evaluations, Cochrane Groups Issue); Biosis Previews; ISI Web of Science; Derwent Drug File; XTOXLINE.
An RCT filter was used for searches in MEDLINE, Embase, Biosis Previews, ISI Web of Science, Derwent Drug File and XTOXLINE. Also, the search was limited to reports published in between 2009 and 2012 (search undertaken in July 2012) and to reports published between 2012 and 2013 (search undertaken in August 2013).
Since deferasirox treatment is an intervention where there is still ongoing research, the following four trial registries were searched on 03 June 2013 for all years available in all possible fields using the basic search function (using separately the following keyword terms; 'deferasirox', 'exjade', 'ICL670', 'ICL 760', 'CGP72670' and 'CGP 72670'):
- Current Controlled Trials Register - via www.controlled-trials.com; (all available registers were searched);
- ClinicalTrials.gov - via www.clinicaltrials.gov;
- ICTRP - via www.who.int./ictrp/en/;
- Deutsches Register klinischer Studien DRKS (German Clinical Trials Register) - via www.drks.de.
For the previous version of this review, in addition to the register search, several databases and ongoing trial registers were searched. See Appendix 3 for full details.
Searching other resources
Reference lists of all identified papers were screened additionally to identify other potentially relevant citations.
Contact was made with selected experts in the field as well as the manufacturer of deferasirox (Novartis) to request information on unpublished studies that involved deferasirox.
Data collection and analysis
Selection of studies
One author (JM) screened all titles and abstracts of papers identified by the search strategies for relevance. We only excluded citations which were clearly irrelevant at this stage. We obtained full copies of all potentially relevant papers. At this stage two review authors (JM and DB) independently screened the full papers, identified relevant studies and assessed eligibility of studies for inclusion. We resolved any disagreement on the eligibility of studies through discussion and consensus, or if necessary through a third party (GA). We excluded all irrelevant records and recorded details of the studies and the reasons for exclusion.
Data extraction and management
Aside from details relating to the risk of bias of the included studies, we extracted two groups of data.
- Study characteristics: place of publication; date of publication; population characteristics; setting; detailed nature of intervention; detailed nature of comparator; and detailed nature of outcomes. A key purpose of this data was to define unexpected clinical heterogeneity in included studies independently from the analysis of the results.
- Results of included studies with respect to each of the main outcomes indicated in the review question. We carefully recorded reasons why an included study did not contribute data on a particular outcome and considered the possibility of selective reporting of results on particular outcomes.
Two review authors (JM, DB) independently undertook data extraction using a data extraction form developed by the authors. The review authors resolved any disagreements by consensus or through discussion with a third author (GA). Once disagreements had been resolved, we recorded the extracted data on the final data extraction form. One review author (JM) transcribed these into RevMan 5.2 (Review Manager 2012). Another review author (DB, LS) verified all data entry for discrepancies.
Assessment of risk of bias in included studies
Two review authors (JM, DB) assessed every study using a simple form and followed the domain-based evaluation as described in the Cochrane Handbook for Systematic Reviews of Interventions 5.0 (Higgins 2008a).
We assessed the following domains as having either a low, unclear or high risk of bias:
- concealment of allocation;
- blinding (of participants, personnel and outcome assessors);
- incomplete outcome data;
- selective outcome reporting;
- other sources of bias.
We reviewed the assessments and discussed any inconsistencies between the review authors in the interpretation of inclusion criteria and their significance to the selected studies. We resolved any disagreements through discussion with a third author (GA). We did not automatically exclude any study as a result of a rating of an 'unclear' or 'high' risk of bias. We present the evaluation of the risk of bias of the included studies in tabular form in the Results section of the review.
Measures of treatment effect
We analysed extracted data using the most up-to-date version of RevMan available at the time of analysis (Review Manager 2012).
We planned to extract hazard ratios with their 95% confidence intervals (CI) for the time-to-event outcomes mortality and end-organ damage. If hazard ratios were not given, we planned to use indirect estimation methods described by Parmar (Parmar 1998) and Williamson (Williamson 2002) to calculate them.
Since we were unable to either extract these data from the study reports or receive the necessary information from the primary investigators, as an alternative we used the proportions of participants with the respective outcomes measured at certain time points (i.e. three months, six months, then six-monthly intervals) to be able to calculate risk ratios (RR).
We expressed any results for binary outcomes as RR with 95% CIs as measures of uncertainty. Continuous outcomes were expressed as mean differences (MD) with 95% CIs as measures of uncertainty.
Unit of analysis issues
We did not include any cross-over studies nor non-inferiority studies. Therefore, the following methods of analysis were not used for this current version of the review. However, for future updates, when conducting a meta-analysis combining results from cross-over studies, we plan to use the methods recommended by Elbourne (Elbourne 2002). For combining parallel and cross-over studies, we will use the methods described by Curtin (Curtin 2002a; Curtin 2002b; Curtin 2002c).
For some outcomes, a possible perception of the comparison might be whether deferasirox is not inferior to standard treatment with deferoxamine. Therefore, a per-protocol analysis would be of interest, as is often used for non-inferiority studies, for our primary outcome as well as for the groups one to five of our secondary outcomes.
Studies included in future versions of this review may report information concerning the intention-to-treat (ITT) or per protocol (PP) population or both which we would analyse according to Witte who lists several proposals for analysing these, depending on the data available (Witte 2004).
- If all studies report only an ITT analysis (or all studies report only a PP analysis), we will perform a non-inferiority meta-analysis based on Witte's `perfect case' proposal.
- If some studies report only an ITT analysis and others only a PP analysis (exclusively), we will perform meta-regression with analysis type as a covariate.
- If some studies report only an ITT analysis and others only a PP analysis, whilst others report both, we will undertake a sensitivity analysis.
- If all studies give enough information to do both analyses, we will analyse a bivariate model.
For time-to-event data, we would state non-inferiority if the relative difference in hazard ratios is less than 10%. For RRs, we would define non-inferiority as a RR difference of less than 10% in treatment failures compared to standard therapy. For the continuous outcomes of 'measures of iron overload and iron excretion' as well as 'costs' we would also consider a relative difference of 10% as equivalent.
Dealing with missing data
We requested any missing data from the original investigators.
Assessment of heterogeneity
For outcomes for which both studies provided data, we assessed heterogeneity using Chi
In future updates of this review, if moderate or high heterogeneity is detected and far more than the present two studies are included (i.e. 10 or more studies), we intend to explore clinical heterogeneity by examining differences between groups as detailed below (Subgroup analysis and investigation of heterogeneity).
Assessment of reporting biases
We made a great effort to minimise the likelihood of publication bias by the use of a comprehensive search strategy including the search of abstracts and contacting the manufacturer of deferasirox. We did not use funnel plots to assess publication bias, since asymmetry is difficult to detect with a small number of studies (i.e. less than 10) and we could only include two studies in this updated version of the review. If in future we will be able to include more than 10 studies in this review, we will use funnel plots to graphically assess the likelihood of publication bias. We took care in translating the results of the included studies into recommendations for action by involving all review authors in drawing conclusions.
We conducted meta-analyses using a fixed-effect model as the primary analysis. If we found marked clinical, methodological or statistical heterogeneity (I
Subgroup analysis and investigation of heterogeneity
We did not perform subgroup analyses as outlined in our protocol since only two studies could be included. For the same reason investigation of clinical or methodological heterogeneity was not done. However, for future updates of this review, we plan to assess clinical heterogeneity by examining differences due to:
- age of participants (i.e., 0 to 2 years; 3 to 5 years; 6 to 11 years; 12 to 17 years; 18 years or older);
- age at commencement of the intervention (i.e., 0 to 2 years; 3 to 5 years; 6 to 11 years; 12 to 17 years; 18 years or older);
- baseline measures of iron overload (i.e. LIC (mg Fe/g dw): ≤3; >3 to ≤ 7; < 7 to ≤ 14; < 14 or serum ferritin (ng/mL (or both)): < 1000; > 1000 to ≤ 2500; > 2500 to ≤ 4000; > 4000).
Subgroup analyses are planned for different:
- doses of intervention (≤ 10 mg/kg; > 10 to ≤ 20 mg/kg; > 20 to ≤ 30 mg/kg; > 30 mg/kg);
- genotypes of SCD (e.g. HbSS, HbS/β-Thal or HbSC).
Sensitivity analyses based on assessment of risk of bias and publication status (unpublished and published studies) were not performed since only two studies could be included in this review. However, for future updates of this review we plan to investigate the robustness of our results through a sensitivity analysis on the basis of the risk of bias of the included studies by defining the following groups: low risk of bias (successful blinding of patients, people involved in treatment and care and outcome assessors; adequate allocation concealment; loss to follow-up of less than 20%); high risk of bias (no blinding, inadequate allocation concealment and loss to follow-up of more than 20%); unclear risk of bias (rating of unclear risk of bias in at least one of these three categories).
Description of studies
Results of the search
The updated searches for this current version of the review (run in July 2012 and again in August 2013) identified 715 citations, including 253 duplicates (Figure 1). The title and abstract screening of the remaining 462 citations identified 25 as potentially eligible for this review. The full text screening excluded 24 citations for the following reasons:
|Figure 1. Study flow diagram combining first search & update searches (most recent search of databases: August 2013; most recent search of the Cochrane CFGD Group's Trials Register: 13 March 2014)|
- review or editorial/comment or other form of published article (N = 7);
- other intervention not including deferasirox (N = 1);
- observational data on people with SCD (N = 14);
- cost-effectiveness analyses on people with SCD (N = 1);
- RCT (SWiTCH trial) comparing hydroxyurea and phelobotomy to transfusions and chelation (chelation group included three patients on deferoxamine) (N = 1).
The identified abstract described a study which was labelled as ongoing in the previous version of our review (Vichinsky 2011). By ascertaining that this study had been completed by May 2011, further data listed on ClinicalTrials.gov, could be included in our analysis.
In addition, search of the Cystic Fibrosis and Genetic Disorders Group's Haemoglobinopathies Trials Register identified a full-text journal article to this study published after our electronic search was run (Vichinsky 2011). We added the additional data given in the published paper to our review (in addition to the data given in the ClinicalTrials.gov entry) and ascertained that no inconsistencies were to be found.
In August 2009, a search of eligible trials was undertaken for the previous version of the review. A total of 420 citations were identified, including 221 duplicates. The title and abstract screening of the 199 unique citations identified 62 as potentially eligible for this review. However, after full text screening no additional reports could be included. Reasons for exclusions were:
- included people with other disease than SCD (thalassaemia (N = 27), myelodysplastic syndrome (N = 14), or other conditions (N = 4))
- review or editorial/comment or other form of published article (N = 5)
- observational data on people with SCD (N = 12)
The initial search for the previous version of this review, was run on 15 August 2008 and identified 1177 citations, including 648 duplicates. The title and abstract screening of the 529 unique citations identified 153 as potentially eligible for this review. The full text screening excluded 147 citations for any of the following reasons:
- included people with other disease than SCD (thalassaemia (N = 72), myelodysplastic syndrome (N = 24) or other conditions (N = 5))
- review or editorial/comment or other form of published article (N = 31)
- other intervention not including deferasirox (N = 1)
- observational data on people with SCD (N = 12)
- cost-effectiveness analyses on people with SCD (N = 2)
For the previous version of this review, we identified six references reporting on a single study (Vichinsky 2007) (Figure 2).This RCT compared deferasirox with deferoxamine. These six reports specifically describe the RCT phase of the study (Vichinsky 2005b; Vichinsky 2005c; Vichinsky 2006a; Vichinsky 2006b; Vichinsky 2007a; Vichinsky 2008a). Of these six reports, two were published as full text articles (Vichinsky 2007a; Vichinsky 2008a) and four as abstracts (Vichinsky 2005b; Vichinsky 2005c; Vichinsky 2006a; Vichinsky 2006b). The remaining five reports listed under the Vichinsky study ID (Vichinsky 2007) report on the extension phases of four clinical trials including patients with SCD, thalassaemia, myelodysplastic syndrome (MDS) and other transfusion-dependent anemias (Cappellini 2007b; Cappellini 2007c; Piga 2007; Vichinsky 2007b; Vichinsky 2008b). The described total groups (consisting of the deferasirox groups of two observational studies and/or the primary deferasirox group(s) and the cross-over groups, i.e. patients switching to deferasirox after finishing the two primary randomised studies) only contain patients treated with deferasirox. There are no data presented of the initial control groups receiving deferoxamine. Therefore, the data presented are only observational data of a deferasirox group without a control group.
|Figure 2. Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.|
This study randomised 203 eligible people in a 2:1 ratio to receive either deferasirox or deferoxamine. The randomisation was performed using an interactive voice response system and was stratified according to the following age groups: 2 years to under 6 years; 6 years to under 12 years; 12 years to under 16 years and 16 years and older. The randomisation sequence included permuted block groups of six patients for each of the four age strata.
Only 195 of the 203 randomised patients received treatment and were included in the analysis. Patient characteristics were similar in both groups ( Table 1). Drug dosing was stratified according to baseline liver iron concentration (LIC) into four groups, with baseline LIC:
- ≤ 3 mg Fe/g dry weight (dw);
- > 3 to 7 mg Fe/g dw;
- > 7 to14 mg Fe/g dw;
- > 14 mg Fe/g dw.
For the reported LIC values a correction factor of 3.33 and for the adjusted LIC values a correction factor of 6.66 was used to convert the wet weight to dry weight values (Brittenham 1982; Olivieri 1997).
Deferasirox was given once daily each morning as a dispersed solution in water, half-an-hour before breakfast. The dosing algorithm according to baseline LIC and average daily doses administered are described in an additional table ( Table 2).
The second study enrolled 212 patients who were randomised 2:1 to receive either deferasirox (20 mg/kg/day) or deferoxamine (175 mg/kg/week) for 24 weeks. No information regarding randomisation procedure was given in the abstract nor on ClinicalTrials.gov. After 24 weeks, the patients receiving deferoxamine were crossed over to receive deferasirox for the remaining 80 weeks; we therefore limited our analysis to the data available at 24 weeks.
Of the 212 included patients, 9 patients were excluded due to GCP violations and a further 12 patients did not receive the study drug deferoxamine. Therefore, only data on a group of 191 patients were provided in the available abstract and on clinicaltrials.gov. Patient characteristics were similar in both groups ( Table 3).
Our search identified one RCT comparing alternative treatment (hydroxyurea and phlebotomy) to standard treatment (transfusions and chelation) (Ware 2011). This trial did not fulfil our stated inclusion criteria and we therefore decided to exclude this study.
Risk of bias in included studies
We assessed the risk of bias of the included studies according to the six categories outlined in Assessment of risk of bias in included studies. Our interpretation of the published reports of the two studies is as follows (Vichinsky 2007; Vichinsky 2011).
|Figure 3. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
Both studies do not describe the method of randomisation. However, the allocation is characterized as 'randomised'. Therefore, the risk of bias rating for both studies is 'unclear'.
Concealment of allocation
One study used an interactive voice response system; we therefore assumed that adequate allocation concealment was achieved and rated its risk of bias as 'low' (Vichinsky 2007). The risk of bias for the second study was concluded to be 'unclear' as this domain was not discussed (Vichinsky 2011).
Incomplete outcome data
The first Vichinsky study included 203 patients, but only 195 received treatment and were included in the safety population (Vichinsky 2007). Reasons for this are not stated. Additionally, a similar percentage of patients of both groups (11.4% of the deferasirox group and 11.1% of the deferoxamine group) either discontinued treatment due to adverse events (5.3% on deferasirox and 3.2% on deferoxamine) or were lost to follow up (two patients from the deferasirox group, one patient from the deferoxamine group), withdrew consent (six patients from the deferasirox group, one patient from the deferoxamine group) or were excluded due to protocol violations (three patients from the deferoxamine group). No intention-to-treat analysis with regard to efficacy was performed. Data are only presented as per protocol although the actual numbers of patients analysed is not very clearly stated. Therefore, we assessed this study as having a high risk of bias for this domain.
In the second Vichinsky study, 212 patients were originally included. Nine participants from one site were excluded due to severe 'Good Clinical Practice' (GCP) violations (Vichinsky 2011). Only 191 of the remaining 203 patients received treatment and were considered for analysis, i.e. 12 out of 68 are missing in the deferoxamine group at 24 weeks. For efficacy analysis (serum ferritin) 18 patients in deferasirox and 6 patients in deferoxamine were not considered. Therefore, we assessed this study as having a high risk of bias for this domain.
According to the methods section of the first Vichinsky study, ferritin values were monitored monthly, clinical and laboratory assessments were performed for safety reasons at 12, 24 and 36 weeks; LIC was also determined by SQUID at 24 weeks (Vichinsky 2007). These data are not reported. Also, it is unclear whether other possible outcomes were measured but not reported. We did not have any access to the original trial protocols to evaluate this. We contacted the corresponding author, but did not receive a response to our inquiry. Therefore, the risk of bias rating for this study is high risk.
The second study provided detailed outcome data on clinicaltrials.gov (Vichinsky 2011). Therefore, we assessed this study as having a low risk of bias for this domain.
Other potential sources of bias
Both studies were sponsored by Novartis. A potential influence of the manufacturer of deferasirox on the reporting of the results can not be excluded due to the cooperation between Novartis staff and the investigators. Therefore the risk of bias rating is unclear for both studies.
Effects of interventions
Only two studies are included in this review. Thus, sensitivity analyses, subgroup analyses or assessments of heterogeneity were not undertaken for this version of the review.
1. Overall mortality
There were only limited data available on the primary outcome: the second Vichinsky trial reported one death in the deferasirox group (assessed as unrelated to the study drug) and no deaths in the deferoxamine group, resulting in a relative risk of 1.26 (95% CI 0.05 to 30.41) (Vichinsky 2011). The interpretation of this analysis is hampered by this study's short follow up. The first Vichinsky trial did not report whether any deaths had occurred (Vichinsky 2007).
1. Reduced end-organ damage due to iron deposition
There were no data available on cardiac failure or histological evidence of hepatic fibrosis. One study reported on incidence of type 2 diabetes mellitus: one patient in the deferasirox group and no patient in the deferoxamine group contracted diabetes (Vichinsky 2011). This difference is not significant, RR 1.26 (95% CI 0.05 to 30.41) ( Analysis 1.2). No other data were available on endocrine disease.
Elevated liver function tests (LFTs) (alanine transaminase (ALT) ≥ 5x upper normal limit (UNL) on two consecutive visits), reported in two studies, were observed in five and two patients, respectively, treated with deferasirox and no patients on deferoxamine, although the difference was not significant, RR 3.66 (95% CI 0.47 to 28.65) ( Analysis 1.3) (Vichinsky 2007; Vichinsky 2011). The second study reported on the number of patients in each group with abnormal liver function tests: 2 patients and 0 patients in the deferasirox group and in the deferoxamine group, respectively (no significant difference), RR 2.10 (95% CI 0.10 to 42.97) ( Analysis 1.4) (Vichinsky 2011). However, since no definition of the event 'liver function test abnormal' was given, we did not pool these data with the first study.
No information was available on the surrogate markers elevated fasting glucose; pathological oral glucose tolerance test; or pathological measures in echocardiography.
2. Measures of iron overload
Ferritin reduction was greater in people treated with deferoxamine at the end of the two studies (12 months and 24 weeks, respectively). These results were statistically significant with a MD of change of 440.69 µg/l in favour of deferoxamine (95% CI 11.73 to 869.64) ( Analysis 1.5) (Vichinsky 2007; Vichinsky 2011). However, only data on 83 out of 132 (62.9%) and 117 out of 135 (86.7%) patients, respectively, treated with deferasirox and 33 out of 63 (52.4%) and 50 out of 56 (89.3%), respectively, treated with deferoxamine were available according to the study authors.
Liver tissue iron assessment by SQUID, reported in one study (Vichinsky 2007), showed no statistically significant difference for the overall group of patients, MD -0.20 mg Fe/g dw (95% CI -3.15 to 2.75) ( Analysis 1.6), nor for the subgroup of patients receiving simple transfusions, MD -0.20 mg Fe/g dw (95% CI -1.97 to 1.57) ( Analysis 1.7). For the subgroup of patients receiving exchange transfusions, the deferasirox-treated patients showed a higher, statistically significant reduction in LIC, MD -5.20 mg Fe/g dw (95% CI -8.56 to -1.84) ( Analysis 1.7). Neither results of LIC from liver biopsies nor from MRI assessment were available.
3. Measures of iron excretion (urine and/or faeces) over 24 hours
There were no data available on iron excretion in urine or faeces.
4. Adverse events
Except for occurrence of serum creatinine values exceeding the upper limit of normal (ULN) and alanine aminotransferase levels > 5x ULN, the first Vichinsky study did not report on rare adverse events (frequency of less than 10%) (Vichinsky 2007); according to the ClinicalTrials.gov entry, the second study reported only on adverse events occurring in more than 5% of participants, while no such threshold was chosen for serious adverse events (Vichinsky 2011).
Creatinine levels were reported by the first Vichinsky study and partly also by the second Vichinsky study (Vichinsky 2007; Vichinsky 2011). Mild stable increases in creatinine were observed more often (though not statistically significant) in patients treated with deferasirox, RR 1.64 (95% CI 0.98 to 2.74) ( Analysis 1.8); the mean increase in creatinine over the course of the study was 3.24 µmol/l, significantly higher in the deferasirox-treated group (95% CI 0.45 to 6.03) ( Analysis 1.9). However, no statistically significant difference was seen in frequency of increases > UNL of creatinine (reported by two studies), RR 0.72 (95% CI 0.12 to 4.18) ( Analysis 1.8).
Adverse events of any kind were reported significantly more often in the deferoxamine group, RR 0.88 (95% CI 1.03 to 5.55) ( Analysis 1.10) (Vichinsky 2011). However, no difference was observed regarding the incidence of serious adverse events. They occurred with similar frequency in both groups ( Analysis 1.10). Also, the difference in serious adverse events suspected to be related to the study drug did not reach significance ( Analysis 1.10).
Rash occurred significantly more often in the deferasirox group, RR 2.39 (95% CI 1.03 to 5.55) ( Analysis 1.10). Gastrointestinal adverse effects such as abdominal pain, nausea, vomiting or diarrhoea were collected separately, so the overall proportion of patients experiencing gastrointestinal side effects can not be given. Statistically significant differences in favour of deferoxamine were observed regarding the adverse events: diarrhoea, RR 3.09 (95% CI 1.53 to 6.26); and nausea, RR 2.06 (95% CI 1.11 to 3.80) ( Analysis 1.10). There were no data available regarding the frequency of neutropenia or agranulocytosis, nor regarding the occurrence of hearing problems or eye problems. Further adverse events are described in another graph; no additional statistically significant differences were observed ( Analysis 1.10). The mean growth velocity in cm per year, reported in one study, was about 1 cm higher in the deferasirox group; however, these differences were not statistically significant for any of the analysed age groups ( Analysis 1.12).
Available data for five adverse events showed marked heterogeneity (I
5. Participant satisfaction
Participant satisfaction was reported by the first Vichinsky study (Vichinsky 2007). Both satisfaction with treatment, RR 3.13 (95% CI 1.99 to 4.93) and convenience of this treatment, RR 3.85 (95% CI 2.28 to 6.47) were significantly better with deferasirox compared to deferoxamine ( Analysis 1.13; Analysis 1.14). Also, the likelihood of continuing chelation therapy estimated by the patients themselves was significantly higher with deferasirox, RR 6.86 (95% CI 3.38 to 13.91) ( Analysis 1.15). These effects were also reflected by the overall rate of discontinuations, RR 0.53 (95% CI 0.31 to 0.92), but not by the rate of discontinuations due to adverse effects, RR 1.03 (95% CI 0.29 to 3.63)) ( Analysis 1.16).
6. Cost of intervention per year
This review is aimed at determining the effectiveness and safety of deferasirox for the management of iron overload in people with SCD.
Firstly, we planned to compare deferasirox to either no treatment or placebo. However, no study addressing this question could be identified, which is not surprising considering that the necessity of chelation therapy in iron-overloaded people is well-established and, if at all, only short-term, e.g. pharmacokinetic studies, would be ethically justifiable. Longer-term studies comparing deferasirox with no therapy or placebo would not suffice the paradigm of equipoise and would therefore be unethical.
Summary of main results
Two trials, involving 203 and 212 people, respectively, compared the efficacy and safety of deferasirox and deferoxamine after 12 months and 24 weeks, respectively. The primary outcome of the included studies was safety and tolerability; efficacy was only addressed as a secondary endpoint. Thus, only limited data were available for most of our pre-specified patient-important long-term efficacy outcomes. One study reported on our primary outcome (mortality). However, follow up was too short to allow valid conclusions on mortality to be drawn. Also, only limited data were available for reduced end-organ damage (incidence of diabetes mellitus). In a five-year extension phase, three deaths assessed as unrelated to the study drug occurred; surrogate markers of end-organ damage were also reported. However, it has to be kept in mind that observational data are less reliable than randomised controlled trial data and thus should be interpreted with caution (Vichinsky 2007).
Two additional efficacy outcomes we were planning to analyse were addressed in the included studies, namely the surrogate marker serum ferritin (addressed in both included studies) and LIC measured by SQUID (addressed in one study). Serum ferritin reduction was greater with deferoxamine with a mean difference of change of 440.69 µg/l (95% CI 11.73 to 869.64). Deferasirox was effective in reducing LIC; however, there was no statistically significant difference in LIC reduction between deferasirox and deferoxamine for the overall group of patients adjusted for transfusion category, MD -0.20 mg Fe/g dw (95% CI -3.15 to 2.75) (Vichinsky 2007). However, based on the results from one large phase III trial including 586 people with thalassaemia, deferasirox presumably offers a similar efficacy as deferoxamine (Cappellini 2006), in particular if higher doses of deferasirox are used.
Adverse events were observed frequently in both groups. Statistically significant differences in the frequency between the deferasirox- and the deferoxamine-treated groups were encountered for rash, diarrhoea and nausea in favour of deferoxamine, and for adverse events of any kind in favour of deferasirox. The mean increase of creatinine was also significantly higher with deferasirox, although the clinical relevance is debatable. Further differences in the potential for adverse effects between deferasirox and deferoxamine could not be detected, although the data suggest differences, e.g. with regard to growth velocity in favour of deferasirox, probably due to the sample size of the reporting study. Long-term adverse events were not recorded in either of the included studies. Also, it should be stressed that the investigators of the 2007 trial reported only adverse events that occurred in at least one treatment group with a frequency of more than 10% (Vichinsky 2007). Therefore, in this particular study, no information regarding potentially rarer adverse events such as kidney failure, agranulocytosis or eye and hearing problems in people with SCD was reported (Vichinsky 2007). The later trial reported only on adverse events occurring in more than 5% of participants, while no such threshold was established for serious adverse events (Vichinsky 2011). Since both cataracts or lenticular opacities and hypoacusis or neurosensory deafness have been reported in less than 1% of patients in the phase III trial of thalassaemia patients, regular screening at 12-month intervals is advised. Also, periodic monitoring of serum creatinine and liver function test on a monthly basis is recommended.
As one would expect, satisfaction with, and convenience of, deferasirox treatment was significantly higher; the same advantage of deferasirox was observed in an ancillary study to the phase III trial with thalassaemia patients (Cappellini 2007a). These effects were reflected by the overall rate of discontinuations, which were significantly higher in the deferoxamine group.
However, it remains unclear whether compliance with deferasirox will exceed compliance achieved with deferoxamine in the long term, which is of paramount importance to prevent morbidity and mortality in people with secondary iron overload (De Sanctis 2006; Wolfe 1985).
Overall completeness and applicability of evidence
The aim of this review, namely to assess the efficacy and safety of deferasirox in the treatment of people with SCD and secondary iron overload, could only be partly accomplished. Due to limitations in data from only two studies with both a relatively short follow up and a small number of participants (12 months, 195 participants and 24 weeks, 203 participants, respectively), it is difficult to arrive at unambiguous conclusions. The available data, as well as data from studies in people with thalassaemia, show that deferasirox is effective in reducing iron overload. However, whether deferasirox is more effective than deferoxamine could not be shown; our data actually suggest that deferoxamine is more effective (as measured by serum ferritin) based on the appropriate ratio of doses of deferasirox and deferoxamine used in these two studies. It could be argued that there is no need for deferasirox to be more efficacious since equivalent effectiveness in conjunction with better compliance and a more favourable safety profile could well be a huge improvement for affected patients. For this reason, the only completed phase III trial in people with thalassaemia was designed and analysed as a non-inferiority trial (Cappellini 2006).
This review, based on the two relatively small included studies, was not able to address some patient-important outcomes such as mortality or end-organ damage in sufficient detail to allow clear conclusions to be drawn, since these events are very rare and therefore the detection of statistically significant differences in studies of this size is unlikely. Given RCTs, large enough to detect these events, are unlikely to be conducted, further large observational studies will need to provide the main evidence for decision-making. Therefore, recommendations regarding the choice of iron chelating treatment in people with SCD can only be based on limited evidence. While there is indirect evidence from studies of patients with thalassaemia, it can not necessarily be assumed that for example, the safety profile will be exactly the same in people with SCD compared to people with thalassaemia.
Quality of the evidence
There is only limited evidence on the effects of deferasirox in people with SCD. Only two studies were available for inclusion in this review. Only 195 participants of the initially 203 eligible patients and 191 of initially 212 eligible patients, respectively, were included in the safety analysis. The proportion of patients included in the per protocol efficacy analyses ranged from 57.1% for serum ferritin to 85.2% for LIC measured by SQUID (Vichinsky 2007). The reasons for this are not stated. The overall risk of bias of the included studies was moderate and is described in detail in the risk of bias tables (Characteristics of included studies). Participants were not blinded to treatment due to the different modes of application of deferasirox and deferoxamine. However, blinding of assessors would have been feasible and could have improved the quality of some rather subjective outcomes, e.g. some adverse events such as nausea or abdominal pain. Finally, it is noted that both studies included in this review were sponsored by Novartis, the manufacturer of deferasirox. In the past, studies conducted with pharmaceutical company involvement in other therapeutic areas have been shown to contain a bias towards the drugs of the sponsor (Bhandari 2004).
Potential biases in the review process
A very comprehensive search strategy was applied to identify all potential studies and their reports. Despite identifying 12 reports (including 9 abstracts) of a single RCT (Vichinsky 2007) and also a full paper, an abstract, and data reported on clinicaltrials.gov for a further RCT (Vichinsky 2011), data for several relevant outcomes pre-specified in our protocol were not available. Several of these outcome measures are, however, important when making an informed and balanced decision on choice of chelator. While it is likely that some of these outcomes were not measured during the trial, others might have been collected but not reported. Unfortunately, even after contacting the primary investigators, we have not been able to obtain any additional data to date.
To minimize bias in the process of doing this review, we followed the rigid methodology for systematic reviews as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008b).
Agreements and disagreements with other studies or reviews
A review based on a systematic literature search focusing on deferasirox was published in 2006 (VanOrden 2006). In addition to data from a phase III trial, the 2006 review included evidence from phase I and phase II studies, as well as from pharmacokinetic studies, in both humans and non-humans. However, since the studies on patients with SCD included in this review was not yet published, their review did not include any data from a RCT of patients with SCD. Most of the patients included in the 2006 review had thalassaemia and approximately three-quarters of these patients are from a single phase III RCT. The authors made no attempt to pool the data; so findings are presented narratively (including observational data). The authors concluded that their findings suggest that deferasirox is as safe and effective as deferoxamine. However, further studies would be needed to assess its efficacy in other transfusion-requiring diseases such as SCD.
A further systematic review published in 2007 summarized the available data from five phase I and II and one phase III study (Lindsey 2007). All six studies are critically discussed, but no data were pooled and only synthesized qualitatively. Based on the only trial looking at efficacy as a primary endpoint (Cappellini 2006), the authors came to the conclusion that the two agents had similar efficacy although, overall the non-inferiority of deferasirox could not be shown by the primary phase III study investigators. Tolerability was assessed as good, even though deferasirox was associated with a higher incidence of adverse effects. The authors concluded that long-term efficacy and safety remained to be established.
In 2009 a comprehensive health technology assessment on deferasirox for secondary iron overload in patients with chronic anaemia, such as thalassemia and SCD was published (McLeod 2009). It identified 14 RCTs looking at various iron chelation regimens with a high degree of heterogeneity between trials in terms of trial design and outcome reporting. Only three of these compared deferasirox to deferoxamine, none of them contained data that could be included in a meta-analysis. Furthermore, eight economic evaluations were included in their report. The authors concluded that there appeared to be little difference between agents in terms of reducing serum ferritin. The economic evaluations appeared to demonstrate the cost-effectiveness of deferasirox compared to deferoxamine. However, the authors stated that both their clinical and economic analyses were restricted by the available evidence and should only be considered as exploratory.
A review published in 2011 evaluated different chelation treatments in sickle cell anaemia, including deferasirox, regarding their effectiveness, safety and costs (Lucania 2011). Based on one RCT (Vichinsky 2007), two prospective cohort studies (Cappellini 2010; Raphael 2009) and one case report (Yusuf 2008) in people treated with deferasirox, the authors concluded that deferasirox was effective and seemed to be as safe as deferoxamine. However, the authors stated that follow up was too short to judge treatment with deferasirox as completely safe; and concerns about patients with possible pre-existing renal impairment should be emphasized. The overall conclusion by the authors was that chelation treatment in SCD, even though widely recommended and practised, had been based on little efficacy and safety evidence. According to this review, the cost-benefit ratio had also not been fully explored.
An economic evaluation in 2009 reported that deferasirox resulted in US$3197 savings compared to deferoxamine with a gain of 2.63 quality-adjusted life-years per patient (Kim 2009). We were unable to assess the costs of treatment with deferasirox since no data were reported in the included studies.
Finally, several reports have recently raised concerns about the safety of deferasirox and possible toxic side effects. Kontoghiorghes summarized several reports on mortality caused by deferasirox, with one study giving a 11.7% mortality rate (1935 of 16514 patients) (Kontoghiorghes 2013). An observational study reported that nine out of 10 observed patients presented at least one sign of proximal tubular dysfunction at a mean (SD) of 17.2 (8.9) months after initiation of deferasirox therapy (Dubourg 2012). Therefore, a systematic safety review including observational data seemed to be warranted.
Implications for practice
Deferasirox offers a new interesting option to the treatments available for secondary iron overload in people with sickle cell disease. It appears to be of similar efficacy as deferoxamine depending on the appropriate ratio of doses of deferoxamine and deferasirox being compared as chosen in the two randomised studies, or even higher, with regard to surrogate endpoints after 12 months and 24 weeks, respectively. Accordingly, it is now widely used due to its easier mode of application. The short-term safety of deferasirox seems to be acceptable; however, follow up in the included studies was too short to evaluate comparative efficacy and long-term side effects. Currently, only limited evidence on long-term efficacy and safety is available from the extension phase of one trial following patients treated with deferasirox only.
Even though there are no randomised trials looking at long-term patient-important outcomes in people with sickle cell disease on deferoxamine, there is long-standing experience with this drug. Thus, for people faced with initiating iron chelating therapy, both treatment options should be discussed in detail. Advantages and disadvantages should be pointed out, taking account of the limited evidence on deferasirox, this should be followed by fully informed shared decision-making. For people being well-managed with deferoxamine, it does not seem to be necessary to change the regimen to deferasirox. For people experiencing intolerance to deferoxamine or with low compliance due to the laborious mode of application, deferasirox should be offered as an alternative. However, effort should be made to enrol these patients in clinical trials to further gain insight into the effects of deferasirox; if no randomised controlled trials enrolling patients are being conducted, observational data should be collected in surveillance studies to better monitor adverse events.
Implications for research
Further studies on the effects of deferasirox in people with sickle cell disease are needed for several reasons. Firstly, they need to overcome the limitations of the available two randomised controlled trials, namely short duration and limited data on patient-relevant outcomes. Also, focusing on efficacy measures is of high importance. Furthermore, the producer of deferasirox sponsored not only the included studies in sickle cell disease, but also the aforementioned studies in other conditions, such as thalassaemia. In our opinion, investigator-initiated studies designed and conducted independently from Novartis are desirable. Finally, further studies should be undertaken looking at, e.g. different dosing schemes or a combination of deferasirox with other iron chelators (or both) to establish the optimal treatment algorithm for individual patients based on their iron overload status.
We thank the peer reviewers for their valuable comments which helped us to improve protocol and review. We would also like to thank the editorial team, namely Tracey Remnington and Nikki Jahnke, for their great support in preparing the protocol and this review. Christina Reese helped with the literature search and retrieval of full articles for the first version of this review. Claire McLeod gave valuable input at the protocol stage.
Data and analyses
- Top of page
- Summary of findings [Explanations]
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Index terms
Appendix 1. Search strategies August 2013
Appendix 2. Search strategies July 2012
Appendix 3. Search strategies June 2009
Last assessed as up-to-date: 6 May 2014.
Protocol first published: Issue 4, 2008
Review first published: Issue 8, 2010
Contributions of authors
Joerg Meerpohl: Conception, design and coordination of the review. Data collection and data management as well as analysis and interpretation of the data. Writing of the review and approval of the final version.
Lisa Schell: Data collection and data management. Involvement in writing the review. Verification of data entries for discrepancies.
Gerta Ruecker: Statistical advice and methodological support. General advice on the review and approval of the final version.
Edith Motschall: Advice on the initial search strategy and literature searches for the current version of this review .
Nigel Fleeman: Co-author of the HTA report by McLeod (McLeod 2009). General advice on the review and approval of the final version.
Charlotte Niemeyer: Interpretation of the data, clinical expertise, providing general advice on the review and approval of the final version.
Dirk Bassler: Data collection and data management. Analysis and interpretation of data. Involvement in writing the review and approval of the final version.
Declarations of interest
Joerg Meerpohl enrolled two adolescents with thalassaemia and one with Diamond-Blackfan anaemia in a post marketing surveillance study on deferasirox and participated once in a Novartis advisory board meeting on paediatric iron overload. None declared for other authors.
Medical Subject Headings (MeSH)
Anemia, Sickle Cell [blood; *therapy]; Benzoates [adverse effects; *therapeutic use]; Chelation Therapy [adverse effects; *methods]; Deferoxamine [adverse effects; therapeutic use]; Erythrocyte Transfusion [adverse effects]; Ferritins [blood]; Iron Chelating Agents [adverse effects; *therapeutic use]; Iron Overload [*drug therapy; etiology]; Randomized Controlled Trials as Topic; Triazoles [adverse effects; *therapeutic use]
MeSH check words
* Indicates the major publication for the study