Gene therapy for haemophilia

  • Protocol
  • Intervention



This is the protocol for a review and there is no abstract. The objectives are as follows:

To evaluate the safety and efficacy of gene therapy in the treatment of people affected with haemophilia A and haemophilia B.


Description of the condition

Haemophilia is an X-linked bleeding disorder characterized by spontaneous or provoked, often uncontrolled, bleeding into joints, muscles and other soft tissues, causing significant pain, swelling and permanent damage if left untreated. Haemophilias are caused by a genetic deficiency of specific proteins in the blood called clotting factors. Haemophilia A is due to an inherited deficiency of factor VIII and haemophilia B is due to a deficiency of factor IX. The incidence of haemophilia A is around 1 per 5000 to 10,000 male births, and about 1 per 60,000 births for haemophilia B. Approximately 80% of affected children are born deficient in factor VIII (haemophilia A) and 20% are deficient in factor IX (haemophilia B) (Bi 1995; Wong 2011). Haemophilias are inherited in an X-linked recessive manner. Hence, haemophilias occur almost exclusively in males, but also in some carrier females due to variances in inactivation of X chromosomes.

The goal, when managing these patients, is to control both the frequency and severity of bleeding episodes and ultimately to prevent permanent joint damage and death in severe cases. Replacement of the missing clotting factor from exogenous sources is the mainstay of therapy in haemophilia. Before World War II, treatment of haemophilia was limited to the transfusion of whole blood or fresh plasma. The first clotting factor concentrate was discovered in the 1960s in the form of cryoprecipitate. In the 1970s, plasma-derived clotting factors were isolated which created a paradigm shift in the management of haemophilia (Brinkhous 1968; Webster 1965). By 1980, the commercially available freeze-dried factor VIII and IX concentrates increased the average lifespan of people with haemophilia to 60 years from a mere 27 years in the 1940s (Evatt 2006; Wong 2011). Early factor VIII and IX concentrates were derived from pooled human plasma from up to 20,000 donors.

Though hepatitis B and C were known risks of pooled human serum derived factor concentrates, they were considered acceptable due to a drastic improvement in the quality of life of people with haemophilia (Kasper 1972; Makris 1990; Mannucci 1977). In 1982 came the first reports of patients with haemophilia succumbing to acquired immune deficiency syndrome (AIDS) and only later it was realized that plasma-derived factor concentrates were responsible for transmission of the still unidentified infectious agent (Aronstam 1993; Gill 1983; Goedert 1989). Nearly 90% of Americans with severe haemophilia have been reported to be infected with HIV in the 1980s via contaminated pooled serum products (NHF 2013). HIV was isolated in early 1984 and by early 1985 heating of factor concentrates became standard practice to kill the virus before transfusion (Wong 2011). Subsequently, several safeguards were employed to prevent donor derived factor transfusion induced infections, such as: donor screening; chromatographic purification; and viral inactivation (Wong 2011). Fortunately, since 1986, there have been no reported cases of HIV transmission through factor concentrates at least in the USA (NHF 2013). The human factor IX gene was cloned in 1982, followed by the production of human rFIX in Chinese hamster ovary cells (Anson 1984; Choo 1982). Factor VIII was cloned and sequenced in 1984 (Gitschier 1984; Toole 1984) and the first recombinant factor VIII (rFVIII) product was approved for clinical use in 1992 (Wong 2011).

The initial treatment modality was episodic or 'on demand' replacement of the missing factor as needed, after the onset of bleeding. However, several studies have shown that prophylactic administration of clotting factor concentrates to prevent bleeding episodes preemptively is more beneficial than episodic administration ( Iorio 2011; Manco-Johnson 2007) The goal of clotting replacement is to keep factor activity greater than one per cent. The currently used adjuvant therapies include the use of antifibrinolytic agents (in both haemophilia A and B) and desmopressin (in mild and moderate haemophilia A only). The antifibrinolytic agents such as e-aminocaproic acid and tranexamic acid exert their effect by inhibiting the proteolytic activity of plasmin and, therefore, inhibiting fibrinolysis (Wong 2011). Desmopressin works mainly by releasing von Willebrand's factor from its storage sites and stabilizing factor VIII in the plasma (Castaman 2008; Leissinger 2001).

Description of the intervention

In gene therapy, a change is made in the genetic material in certain target cells by means of a vector (usually viruses) in such a way that the change is functional and produces an adequate end product so as to correct the defect underlying the disease. In haemophilia, recombinant genetic material is introduced into the target cells which cause production of the clotting factor which is deficient in the disease.

It was first shown in 1989 that human factor IX could be synthesized and secreted into the circulation of laboratory animals after transplantation of genetically modified human fibroblasts using a human factor IX cDNA containing retrovirus (Nathwani 2004; Palmer 1989). Since then, several translational approaches have been developed for the clinical application of gene therapy in haemophilia, but most of these have only resulted in transient benefits (Manno 2003; Manno 2006; Powell 2003; Roth 2001; VandenDriessche 1999; Xu 2003). Finally, in 2011, a new adeno-associated virus (AAV) vector for factor IX gene transduction that could directly be given to an adult by a peripheral vein infusion was developed (Nathwani 2006; Nathwani 2011; Tuddenham 2012).

Gene therapy is not without risk. Introduction of new genetic material via vector viruses may cause unpredictable outcomes due to the alteration of the genetic material (insertional mutagenesis) (Hacein-Bey-Abina 2003; Miller 2005; Nakai 2005), there is a theoretical risk that the viral vector may regain its potential to produce new viral particles, or may cause harmful immune reactions (High 2011; Manno 2003; Mingozzi 2011).

How the intervention might work

The currently available treatment for haemophilia is lifelong clotting factor replacement, a regimen which is both expensive and is not easily available (Iorio 2011; Ponder 2008). Worldwide, about 75% of haemophilia patients do not have access to adequate care (Ponder 2008). Due to the short half-life of the clotting factors in the blood, replacement is required usually every other day for haemophilia A and every two to three days for haemophilia B. The frequency and mode of administration (intravenous (IV)) of these treatment poses a huge range of challenges to people affected with haemophilia and their families (need for IV catheters and related complications such as infections and thrombosis) (Santagostino 2010).

At USD 1 per unit of recombinant factor VIII, the annual cost of on-demand factor replacement therapy for a single 50 kg adult patient of haemophilia is USD 150,000 in the United States of America (Manco-Johnson 2007; Ponder 2011). This cost increases to about USD 300,000 for prophylactic therapy. This amounts to an individual lifetime cost of over USD 20 million in factor replacement costs alone (Ponder 2011). Such an exorbitant therapy is neither available frequently, nor affordable to almost 75% of the people with haemophilia in the world who live in developing countries (Ponder 2008; Ponder 2011). These patients continue to have significant morbidities and die young.

In comparison to the above treatment modalities, gene therapy holds substantial promise. The key advantage being that it is considered to be a one time intervention and permanently curative. This intervention will lead to complete avoidance of the need for IV infusions, reduced hospital visits, a decrease in the use of other interventions and their side effects and ultimately reduced costs (Nathwani 2011). The commercial cost of gene therapy using the viral vector is projected at USD 30,000 per patient (Ponder 2011). Such a therapy would be life changing for the patients.

Why it is important to do this review

Gene therapy holds significant promise of a substantially better treatment modality. Apart from that, gene therapy offers the potential of a lifetime cure, a better quality of life and freedom from various related morbidities. Theoretically, once gene therapy is administered, the affected individual will be asymptomatic with respect to haemophilia. There are still doubts with regards to long-term sustenance of the effects, the unintended consequences or adverse effects and the costs of therapy. We aim to conduct this review in the hopes of answering some of the above questions, specifically in terms of the benefits and safety of gene therapy in comparison to standard treatment in people affected with haemophilia.


To evaluate the safety and efficacy of gene therapy in the treatment of people affected with haemophilia A and haemophilia B.


Criteria for considering studies for this review

Types of studies

Randomised or quasi-randomised clinical trials including controlled clinical trials.

Types of participants

Individuals with haemophilia A or B of all ages who do not have inhibitors to factor VIII or IX.

Types of interventions

Gene therapy (with or without standard treatment) compared with standard treatment (factor replacement) or other 'curative' treatment such as stem cell transplantation.

Types of outcome measures

Primary outcomes
  1. Bleeding episode(s) (bleeding frequency, number of bleeding episodes per year, or as reported in the trial)

  2. Factor VIII or IX supplementation requirement (frequency as reported in the trials)

  3. Serious adverse events (resulting in death or life threatening complications, inpatient hospitalisation, significant or permanent disability, or one that requires additional intervention to prevent permanent damage or disability)

    1. evidence of clotting factor antibody development

    2. evidence of organ toxicity

    3. evidence of tumour development

Secondary outcomes
  1. Measures of haemostasis

    1. clotting factor plasma levels

    2. activated partial thromboplastin time (aPPT)

    3. international normalised ratio (INR)

    4. prothrombin time (PT)

  2. Non-serious adverse events (any adverse event other than those mentioned above)

  3. Joint damage (changes in: clinical joint function; orthopedic joint score; or radiologic joint score)

  4. Quality of life (as measured by standardised instruments)

Search methods for identification of studies

Electronic searches

We will identify relevant trials from the Cystic Fibrosis & Genetic Disorders Group's Coagulopathies Trials Register using the term: gene therapy.

The Coagulopathies 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 and the handsearching of the journal - Haemophilia. Unpublished work is identified by searching the abstract books of major conferences: the European Haematology Association conference; the American Society of Hematology conference; the British Society for Haematology Annual Scientific Meeting; and the Congress of the World Federation of Haemophilia. 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.

Additionally, we will directly search CENTRAL, PubMed and Embase using a specialised sensitive search strategy.

We shall search and consider including trials irrespective of publication status or language.

Searching other resources

We will search the reference lists and contact authors of included trials, to identify any additional, potentially relevant trials for inclusion. We shall search available clinical trial registers or meta registers for any ongoing trials, specifically in the following registers.

Data collection and analysis

Selection of studies

Search results from all sources will be put together in a reference manager software and duplicates will be excluded. Two review authors (JN and SK) will then independently screen the search results using titles and abstracts to exclude irrelevant trials. We shall classify the remaining trials into included, excluded and those awaiting classification, using the predefined criteria for selecting trials. For those trials awaiting classification, we will evaluate the report(s) and then make the judgement regarding inclusion or exclusion. If we are not able to make a judgement from the report(s), we shall attempt to contact the trial authors and make the classification once we have obtained sufficient information. We will resolve any disagreements by a consensus meeting with the third author (AS). We shall clearly document the reasons for exclusion. We will also scrutinise each trial report to ensure that multiple publications from the same trial are included only once and all such reports will be linked to the original trial report in the reference list of included trials.

Data extraction and management

Two review authors (SK and JN) will independently extract the data from the trials considered for inclusion using a structured data extraction form. We will resolve any disagreements by discussion and, in any cases where there are persisting disagreements, we will seek arbitration by another author (MEM).

We will extract the following data.

  1. General information: title; authors; source; country; year of publication; setting.

  2. Trials characteristics: design; method of randomisation; allocation concealment; blinding of outcome assessors.

  3. Patients: inclusion and exclusion criteria; underlying disease; sample size; baseline characteristics of patients; losses to follow up.

  4. Interventions: type of vector; dose.

  5. Additional treatment: immunosuppressive therapy.

  6. Outcomes: as mentioned above (primary and secondary outcomes).

If the full text versions do not provide sufficient information, we will contact the corresponding author for further details.

Assessment of risk of bias in included studies

Two review authors (SK and JN) will independently assess the risk of bias of the included trials on the basis of generation of the randomisation sequence, allocation concealment, blinding (of the participants, personnel and outcome assessors), attrition bias, and selective outcome reporting. We will assign judgements on the above domains of risk of bias into high, low and unclear, based on table 8.5d of the Cochrane Handbook for Systematic Reviews of interventions (Higgins 2011a). We aim to resolve any disagreements by referring to the trial report and, if necessary, through arbitration with two other authors (AS and MEM).

Measures of treatment effect

Different types of data will be analysed as below.

Rate data

When an outcome is measured in terms of rates we shall perform the analysis using rate ratios (RR) with their 95% confidence intervals (95% CI).

Continuous data

When outcome data are continuous, we shall extract means and standard deviations (SD). If the outcomes are reported using different continuous scales we shall use the standardized mean difference (SMD) and their associated 95% CIs to perform the analysis.

Dichotomous data

We shall extract the number of events and the total number of people reporting outcomes and analyse them using risk ratios (RR) and their associated 95% CI.

Unit of analysis issues

Trials with multiple treatment groups

In cases of trials with multiple treatment groups we shall ensure that a participant group is not represented more than once in the same analysis. We shall either combine groups to make pair wise comparisons if possible, or exclude groups by making a decision on which group is most clinically relevant and extract the data only for that group.

Cluster-randomised trials

We do not expect to find cluster randomised trials for this intervention. However, if we do find such trials which report data adjusted for design effect, we shall use the information as such. If the effect estimates are not adjusted, we shall extract the intracluster correlation coefficient (ICC), if provided in the trial report, or contact the authors for the same and adjust the effect estimate for the design effect.

Cross-over trials

We do not expect to find cross-over trials as successful gene therapy is irreversible.

Dealing with missing data

We will contact the primary investigators if information is missing or if there are unclear outcome data, summary data or individual data. We shall extract data for the number of people randomised and the number analysed for each outcome. If there is a difference, we shall calculate the percentage change and report it as loss to follow up.

For dichotomous outcomes, we will perform a per protocol analysis on the data.

For outcomes with continuous data that are missing SDs, we will either calculate these from other available data such as standard errors (SE), or will impute them, if possible. We shall impute SDs on the basis of SDs for the same outcome using the same scale, from within our review or from other similar studies, reviews or meta-analyses (Higgins 2011b).

Assessment of heterogeneity

We shall visually examine a forest plots for overlapping confidence intervals and use the Chi2 test for homogeneity and the I2 test for heterogeneity, to estimate the contribution of true inter trial variability. The Chi2 statistic will be considered to represent significant heterogeneity at a level of 10% significance (P value < 0.1). We will use an I2 statistic value of less than 40% to represent mild heterogeneity that can be ignored; 30% to 60% to represent moderate and 50% to 90% significant heterogeneity, the reasons for which we shall try to explore by subgroups. We shall consider an I2 value of 75% or greater as substantial heterogeneity, in which case we shall not perform meta-analysis.

Assessment of reporting biases

We shall assess for publication bias using funnel plots (provided that there are at least 10 included trials in an analysis). When funnel plots show asymmetry, we shall consider publication bias as a possible reason after excluding other reasons, such as heterogeneity and outcome reporting biases (Sterne 2011).

Data synthesis

We shall attempt to pool results only when there is no clinical heterogeneity and when there is no substantial statistical heterogeneity (I2 < 75%). When heterogeneity is present we shall analyse it using the random-effects model. If we have trials with varying designs in the same comparison we shall attempt to pool the effect estimates adjusted for design effect by using the generic inverse variance method. We shall attempt to pool data of outcomes reported at the same point of follow up across trials. If this is not possible we shall pool trial data at one, two and five years.

If meta-analysis is not possible, we shall perform a descriptive, qualitative critical appraisal of the outcome information from the included trials.

Subgroup analysis and investigation of heterogeneity

In case we observe moderate or substantial heterogeneity in any of the analyses, we plan to explore the possible causes by undertaking the following subgroup analyses:

  1. age;

  2. type of haemophilia (haemophilia A versus haemophilia B);

  3. gene therapy technique (ex vivo versus in vivo transduction);

  4. vector used.

We would like to compare, if possible, all outcomes following gene therapy in younger versus older individuals. This is because older individuals undergoing gene therapy may already have significant co-morbidities from ongoing disease process. Method and site of cellular transduction can also vary the adverse event profile and if there are significant trials available then we will do a subgroup analysis comparing the different methods. If one particular approach seems to have more adverse effects than the other then we will conduct the meta-analysis both with and without the given approach and report the result separately.

Sensitivity analysis

We shall perform appropriate sensitivity analysis for the primary outcomes to examine the robustness of any assumptions made during the review process, specifically with regard to loss to follow up and high risk of bias in allocation of participants.

With regard to loss to follow up and for trials with missing outcome data, which is greater than 10% of the number randomised, we shall examine the change in effect estimates by assuming worse outcomes to those lost to follow up. We shall assign high risk of attrition bias to trials that have a significant change in results.


We would like to thank the South Asian Cochrane Network & Centre for providing logistical and technical support.

Contributions of authors

Roles and responsibilities
Task Who will undertake this task?
Protocol stage: draft the protocolAS, JN, MEM, SK, VS
Review stage: select which trials to include (2 + 1)AS, SK, MEM
Review stage: extract data from trials (2 + 1)SK, JN, AS
Review stage: enter data into RevManSK, JN
Review stage: carry out the analysisAS, MEM
Review stage: interpret the analysisAS, MEM, JN, SK
Review stage: draft the final reviewAS, JN, MEM, SK, VS
Update stage: update the reviewAS, JN, MEM, SK, VS

Declarations of interest

The authors declare that they have no conflicts of interest.

Sources of support

Internal sources

  • South Asian Cochrane Network & Centre, India.

External sources

  • Cochrane Cystic Fibrosis & Genetic Disorders Group, UK.