Desmopressin acetate (DDAVP) for preventing and treating bleeding in people with mild or moderate haemophilia A
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To establish whether DDAVP is safe, effective and affordable for preventing and treating bleeding episodes, as a result of minor trauma or surgery, in people with mild or moderate haemophilia A.
Background
Description of the condition
Haemophilia A is a rare congenital bleeding disorder (Srivastava 2012). It is an X‐linked recessive bleeding disorder that results from the deficiency of coagulation factor VIII. A linear relationship exists between plasma levels of the deficient protein and the expected clinical phenotype, with severe forms of the condition defined as having plasma levels of the deficient protein below 1 international unit (IU)/mL, moderate forms between 1 and 5 IU/mL, and mild forms between more than 5 IU/mL and the lower limit of the local normal values (or 40 IU/mL where the range is not defined) (Srivastava 2012). The prevalence of moderate and mild haemophilia is not known, but estimated at about half of the overall prevalence of haemophilia, which is approximately 10 to 15 per 100,000 males (Stonebraker 2010a). People with haemophilia A present with a tendency to bleed, either spontaneously (usually severe forms, but potentially moderate or mild as well) or as a consequence of minor trauma or surgery (all severities). All bleeding episodes can be managed with replacement treatment, i.e. the therapeutic intravenous administration of factor VIII concentrate, available as either plasma derived or recombinant products. Factor VIII concentrates can be used to raise the plasma level concentration to the levels needed to prevent bleeding before surgery (Srivastava 2012). However, there are drawbacks of factor replacement therapy. Firstly, the administration of exogenous factor concentrates (either plasma derived or recombinant) is sometimes complicated by the occurrence of inhibitory antibodies, making subsequent administration of factor concentrates ineffective (Gouw 2013). The development of inhibitors is more common in people with severe disease, but it has been recently shown to also occur in a significant proportion of those with mild disease (Giuffrida 2008; Hay 1998; Mauser‐Bunschoten 2012). In people with mild disease, the complication is regarded as serious, as the individual is then considered to have a severe form of the condition. Secondly, factor concentrates have to be administered intravenously and require a cumbersome reconstitution process (Feldman 2007; Miners 2009). Thirdly, factor replacement therapy comes at a significant cost, and its availability is limited to residents of higher income countries (Stonebraker 2010b).
Description of the intervention
Desmopressin acetate (1‐deamino‐8‐D‐arginine vasopressin, also known as DDAVP) is a synthetic analogue of the nonapeptide arginine, derived from vasopressin. It may be administered as a slow intravenous infusion (using a 4 µg/mL preparation diluted in saline), or as a subcutaneous injection (using a 15 µg/mL preparation). An intranasal spray preparation is also available, although some people have difficulty mastering the administration technique. The spray is particularly suitable for home treatment, although evidence from a randomised placebo‐controlled cross‐over trial showed only a non‐significant trend to reduce blood loss in women with an inherited bleeding disorder and menorrhagia (Kadir 2002). In a recent Cochrane review of this treatment for preventing and treating acute bleeds during pregnancy in women with congenital bleeding disorders, the authors were unable to draw any conclusions as no trials were eligible for inclusion (Karanth 2015). Alternatively, subcutaneous treatment at home is an option with excellent or good responses recorded in 94% of treatments of bleeding episodes (excluding menorrhagia) and 86% of treatments of menorrhagia (Rodeghiero 1996). As a result of its antidiuretic effects, there is a risk with DDAVP of fluid overload and hyponatraemia (Mannucci 1997), which can lead to seizures in susceptible individuals, particularly young children and pregnant women (Mannucci 2005). Hence, during therapy, fluid intake should be limited and plasma sodium levels frequently monitored. Treatment with DDAVP is not indicated for haemophilia B since it does not affect factor IX levels.
How the intervention might work
Desmopressin acetate induces factor VIII and von Willebrand factor (VWF) release from endothelial cells and boosts plasma levels of factor VIII and VWF (Mannucci 1981), which in turn positively affects platelet function (Colucci 2014). In those with inherited bleeding disorders, DDAVP infusion results in a two‐ to six‐fold increase from baseline in factor VIII and VWF plasma levels, although individuals have differing responses (Lethagen 1987), which have been associated with specific gene mutations (Castaman 2009). The initial response is reproducible in any particular individual if two or three days elapse between administrations. In people with mild or moderate haemophilia A, when factor VIII can be raised to an appropriate therapeutic level, DDAVP is often considered the treatment of choice because it avoids the expense and potential hazards of using a clotting factor concentrate (Franchini 2010; Mannucci 1997). In carriers of haemophilia, DDAVP is particularly useful in the treatment or prevention of bleeding (Leissinger 2001).
Obvious advantages of DDAVP over plasma products are the much lower cost and the absence of any risk of transmission of viral infections. The decision to use DDAVP must be based on both the baseline concentration of factor VIII, the increment achieved, and the duration of treatment required. Each individual's response should be tested prior to therapeutic use, as there are significant differences between them. The response to intranasal DDAVP is more variable and therefore less predictable (Franchini 2010; Mannucci 1997).
Dosage and administration
Desmopressin acetate can be given by intravenous infusion, subcutaneously, or as a nasal spray. It is important to choose the correct preparation of DDAVP because some lower dose preparations are used for other medical purposes. The recommended dose of DDAVP administered intravenously or subcutaneously is 0.3 µg/kg per dose.
The peak response is seen approximately 60 minutes after administration (therapeutic peak level at 30 to 90 minutes) either intravenously or subcutaneously, but it varies between individuals and depends on the route of administration (Mannucci 2012). Closely spaced repetitive use of DDAVP over several days may result in decreased response (tachyphylaxis). Factor concentrates may be needed when higher factor levels are required for a prolonged period (Srivastava 2012). Rapid infusion may result in tachycardia, flushing, tremor, and abdominal discomfort. A single metered intranasal spray of 1.5 mg/mL in each nostril is appropriate for adults. For an individual with a body weight of less than 40 kg, a single dose in one nostril is sufficient (Leissinger 2001; Mannucci 1997). Although the intranasal preparation is available, some people find it difficult to use, especially in cases of nasal congestion, and it may be less efficacious than when given subcutaneously. As a result of its antidiuretic activity, water retention and hyponatraemia can be a problem. When repeated doses are given, the plasma osmolality or sodium concentration should be measured (Franchini 2010; Mannucci 1997). In most adults, hyponatraemia is uncommon. Due to water retention, DDVAP should be used with caution in young children and is contraindicated in children under two years of age who are at particular risk of seizures secondary to cerebral oedema (due to water retention) (Franchini 2010; Srivastava 2012). There are case reports of thrombosis (including myocardial infarction) after the infusion of DDAVP and it should be used with caution in those with a history of, or who are at risk of, cardiovascular disease (Mannucci 1997).
Why it is important to do this review
Since its first clinical use in 1977, DDAVP has become the treatment of choice for people with haemophilia A with factor VIII levels between 5% and 50%. An average a two‐ to six‐fold factor VIII increase is observed in most people and a return to baseline occurs usually within eight hours. This review is important given that DDAVP may be used to treat bleeding complications in those with mild or moderate haemophilia A, instead of the more expensive and risky factor concentrate. In fact, DDAVP is devoid of any risk of blood‐borne infection transmission (as compared to plasma‐derived factor concentrates) and devoid of any risk of inducing inhibitors to factor VIII (as compared to recombinant factor VIII). There is currently no available systematic appraisal of the evidence of the use of DDAVP to prevent and treat bleeding complications in people with mild or moderate haemophilia A. It is necessary to establish whether its use in this population is safe, effective and affordable for preventing and treating bleeding episodes as a result of minor trauma or surgery.
Objectives
To establish whether DDAVP is safe, effective and affordable for preventing and treating bleeding episodes, as a result of minor trauma or surgery, in people with mild or moderate haemophilia A.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) and quasi‐RCTs, also studies that have all of the following design features will be included:
-
prospective recruitment;
-
participants allocated to different interventions using methods that are not random;
-
control group (including historical controls);
We will include published and unpublished studies.
Eligible studies will recruit people with mild or moderate haemophilia A requiring treatment for surgery or trauma. While RCTs and quasi‐RCTs are eligible for inclusion, we do not anticipate identifying studies of these designs given the rarity of the disease and the concomitant occurrence of surgery or trauma (where treatment is not planned a priori). Indeed, several different barriers to planning and executing RCTs in this field can be identified. Firstly, the multiplicative interaction of a rare condition (the coagulation disorder) with a relatively rare occurrence (surgery or trauma requiring treatment) restricts the eligible population even further. Secondly, RCTs need an active comparator (most untreated individuals do bleed, and most of those receiving any form of treatment do not), which significantly inflates the required sample size. A further reason is that RCTs in a surgical setting are generally more difficult, both due to the reluctance of individuals and doctors to agree to randomisation, and for the high cost of being ready to enrol participants from several centres, if the surgeries are emergencies rather than elective in reference centres. Finally, RCTs by design, will usually not recruit enough participants to identify rare adverse effects and the duration (limited by cost restrictions) will not be sufficient to identify long‐term adverse effects.
Types of participants
People with mild or moderate haemophilia A requiring treatment for surgery or trauma. We will define moderate haemophilia as plasma levels of factor VIII between 1 and 5 IU/mL and mild haemophilia as plasma levels of factor VIII between more than 5 IU/mL and the lower limit of the local normal values (or 40 IU/mL where the range is not defined).
Types of interventions
Treatment with DDAVP administration, any route, any dosage, any regimen as follows.
Intervention
-
DDVAP alone
-
DDAVP and antifibrinolytics
-
DDVAP and factor concentrates
Comparator
-
No treatment
-
Antifybrinolytics (with or without DDAVP)
-
Factor concentrates (with or without DDAVP)
Types of outcome measures
Primary outcomes
-
Bleeding (as defined in the included studies and may include bleeding scale, amount transfused in mL, spontaneous or post‐surgical bleeds)
-
Serious adverse effects (as defined by the study authors (for example, hypertension, allergic reaction, tachycardia, decreased sodium levels, seizure))
Secondary outcomes
-
Effective haemostasis during surgery (as defined in the included studies)
-
Bleeding resolution (as reported by the participant or physician)
-
Need for transfusion of blood components (red blood cells (RBC), platelets)
-
Post‐infusion plasma factor level (absolute value or change from baseline)
-
Any mild or moderate adverse effects (as defined by the study authors)
-
Direct cost of treatment (as described in the included studies)
-
Need for re‐intervention (surgical haemostasis)
Search methods for identification of studies
Resources will be searched using the search strategies given in the appendices (Appendix 1; Appendix 2). No restrictions will be made for publication year or language.
Electronic searches
The following electronic databases will be searched:
-
Ovid MEDLINE (1946 to present);
-
SCOPUS (1960 to present);
-
ClinicalTrials.gov www.clinicaltrials.gov (coverage unknown).
Searching other resources
In an effort to identify further studies, we will contact authors of any included studies and the manufacturers of DDAVP. We will also check reference lists of eligible studies.
Data collection and analysis
Particular attention will be given to explicitly stating any deviation from the pre‐defined methods of analysis described below which is imposed by the nature of retrieved data. If there are a large number of deviations, we will consider adding a specific paragraph titled: 'Deviation from the protocols' in the discussion section of the final review.
Selection of studies
Two authors will screen titles and abstracts to identify potentially relevant studies and then screen the full reports. If disagreements arise on the suitability of a study for inclusion in the review, a consensus will be reached by discussion with a third author.
Data extraction and management
Two authors will independently extract data using a structured form including:
-
type of study design;
-
what researchers did (using the Cochrane Non‐randomised Studies Methods Group checklists) (Reeves 2011).
-
participant characteristics;
-
setting;
-
detailed nature of intervention and control or, comparator;
-
detailed nature of outcomes;
-
confounding factors considered, the comparability of the groups base don the confounding factors, methods to control for the confounding factors and adjusted and unadjusted effect estimates.
We plan to group outcome data at immediately post‐treatment and at one week post‐treatment. If outcome data are recorded at other time points, we will consider examining these as well.
Assessment of risk of bias in included studies
We will use the Cochrane risk of bias tool to assess the risk of bias of eligible RCTs (Higgins 2011). This includes the assessment of six domains: sequence generation; allocation sequence concealment; blinding; incomplete outcome data; selective outcome reporting; and ‘other’ potential sources of bias. Items are assessed by: (a) providing a description of what happened in the study; (b) providing a judgement on the adequacy of the study with regard to the item.
We will assess risk of bias in non‐randomised studies using the ACROBAT‐NRSI tool recently developed by the Cochrane Methods Group (Sterne 2014). The tool is an extension of the existing tool for assessing risk of bias in RCTs (Higgins 2011). Seven domains of bias are assessed as being at low, moderate, serious or critical risk of bias, each assessment informed by answers to a series of pre‐defined questions. The domains are bias due to: confounding; selection of participants; classification of interventions; departures from intended interventions; missing data; measurement of outcomes; and the selection of the reported result.
Confounders
We expect potential confounders to be: indication (more severely affected people or those with previous bleeding on DDAVP will receive factor concentrate); dose administered; type of surgery; surgeon; centre or setting; concomitant treatment (e.g. tranexamic acid); and treatment duration. However, if we identify any further confounders in the included studies, we will also take these into account. We will create a table listing the confounders as columns, the studies as rows, and indicating whether each study has:
-
restricted participant selection, so that all groups had the same value for the confounder;
-
demonstrated baseline balance between groups for the confounder;
-
matched on the confounder; or
-
adjusted for the confounder using appropriate statistical analyses to quantify the effect size.
Measures of treatment effect
We will analyse data as per guidance from chapter 9 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011).
For dichotomous outcome measures we will collect data on the number of participants experiencing an event and the total number in the allocated treatment group. We plan to calculate an odds ratio (OR) and 95% confidence interval (CI).
For continuous outcomes we will record either the mean change from baseline for each group or mean post‐treatment values and standard deviation (SD) for each group. We plan to calculate the mean difference (MD) and 95% CIs . If the included studies report the same outcome using different scales, then we will analyse data using the standardised mean difference (SMD).
For time‐to‐event data we will calculate hazard ratios (HR) and 95% CIs.
Where appropriate, we will extract adjusted estimates and their standard error (SE).
Unit of analysis issues
We do not anticipate that there will be any unit of analysis issues; specifically, cluster‐randomized and cross‐over trials are not appropriate for this intervention. However, we are aware that there may be potential for issues to arise due to the effect of practitioner or organisational units on the methods of allocation to intervention or comparator.
Dealing with missing data
We will contact the authors of included studies for clarification if we believe any collected data or information are missing from publications. We will provide a narrative report of any instances where data or information are missing and this will inform our risk of bias assessment.
Assessment of heterogeneity
We plan to assess heterogeneity by visual inspection of the forest plots, the Chi2 test and the I2 statistic. The I2 statistic describes the percentage of total variation across studies that is due to heterogeneity rather than chance (Higgins 2003). The values of I2 lie between 0% and 100%, and a simplified categorization of heterogeneity that we plan to use is:
-
0% to 40%: might not be important;
-
30% to 60%: may represent moderate heterogeneity*;
-
50% to 90%: may represent substantial heterogeneity*;
-
75% to 100%: considerable heterogeneity*.
*The importance of the observed value of I2 depends on: (a) magnitude and direction of effects; and (b) strength of evidence for heterogeneity (e.g. P value from the Chi2 test, or a CI for I2).
We will consider a P value of 0.1 for Chi2 as significant due to the low power of the test.
Assessment of reporting biases
In order to assess the bias introduced from small or unpublished studies, we plan to construct a funnel plot if we are able to include at least 10 studies in a meta‐analysis. If we observe any asymmetry, we will consider other reasons for this as well as heterogeneity and selective outcome reporting.
Data synthesis
If we are able to include sufficient studies, we plan to use a fixed‐effect model in the first instance and a random‐effects model as part of a proposed sensitivity analysis. Studies with different designs (or which have different design features) will not be combined in a meta‐analysis.
If included studies are not sufficiently homogeneous to combine in a meta‐analysis, we will display the results of included studies in a forest plot but will suppress the pooled estimate.
We will analyse adjusted, rather than unadjusted, effect estimates, i.e. analyses that attempt to ‘control for confounding’. If several adjusted estimates are reported, we will use the estimate from the model that includes the largest number of confounders considered in the list above.
We will undertake meta‐analyses of adjusted estimates using the generic inverse‐variance method. It should be noted that the effect size estimate is still at risk of bias due to residual confounding as all methods to control for confounding are imperfect.
For unadjusted estimates, we will undertake meta‐analyses using the Mantel Haenszel method for dichotomous outcomes and the inverse variance for continuous outcomes.
Subgroup analysis and investigation of heterogeneity
We do not anticipate undertaking any subgroup analyses for this review.
Sensitivity analysis
In addition, if there are sufficient studies (at least 10) included in the review, we plan to undertake the following sensitivity analyses:
-
a comparison of analyses using a random‐effects model to a fixed‐effect model when combining studies;
-
analysis based on the overall risk of bias of the studies (we will decide on a level for overall risk of bias after visual inspection of the plot of the distribution of review authors' judgements across studies for each risk of bias item);
-
a comparison of primary outcomes by different identified study design features.
