A systematic review of the clinical effectiveness and cost-effectiveness of Pharmalgen® for the treatment of bee and wasp venom allergy

Background: Each year in the UK, there are between two and nine deaths from anaphylaxis caused by bee and wasp venom. Anaphylactic reactions can occur rapidly following a sting and can progress to a life-threatening condition within minutes. To avoid further reactions in people with a history of anaphylaxis to bee and wasp venom, the use of desensitisation, through a process known as venom immunotherapy (VIT), has been investigated and is in use in the UK. VIT consists of subcutaneous injections of increasing amounts of purified bee and/or wasp venom extract. Pharmalgen (R) products (ALK Abello) have had UK marketing authorisation for VIT (as well as diagnosis) of allergy to bee venom (using Pharmalgen Bee Venom) and wasp venom (using Pharmalgen Wasp Venom) since March 1995. Objective: This review assessed the clinical effectiveness and cost-effectiveness of Pharmalgen in providing immunotherapy to individuals with a history of type 1 [immunoglobulin E (IgE)-mediated] systemic allergic reaction to bee and wasp venom. Data sources: A comprehensive search strategy using a combination of index terms (e.g. Pharmalgen) and free-text words (e.g. allerg$) was developed and used to interrogate the following electronic databases: EMBASE, MEDLINE, The Cochrane Library. Review methods: Papers were included if they studied venom immunotherapy using Pharmalgen (PhVIT) in patients who had previously experienced a systemic reaction to a bee and/or a wasp sting. Comparators were any alternative treatment options available in the NHS without VIT. Included outcomes were systemic reactions, local reactions, mortality, anxiety related to the possibility of future allergic reactions, health-related quality of life (QoL) and adverse reactions (ARs) to treatment. Cost-effectiveness outcomes included cost per quality-adjusted life-years (QALYs) gained. Because of the small number of published randomised controlled trials (RCTs), no meta-analyses were conducted. A de novo economic model was developed to assess the cost-effectiveness of PhVIT plus high-dose antihistamine (HDA) plus adrenaline auto-injector (AAI) plus avoidance advice in relation to two comparators. Results: A total of 1065 citations were identified, of which 266 full-text papers were obtained. No studies were identified that compared PhVIT with any of the outlined comparators. When these criteria were widened to include different protocols and types of PhVIT administration, four RCTs and five quasi-experimental studies were identified for inclusion. The quality of included studies was poor, and none was conducted in the UK. Eight studies reported re-sting data (systemic reactions ranged from 0.0% to 36.4%) and ARs (systemic reactions ranged from 0.0% to 38.1% and none was fatal). No included studies reported quality of life. No published economic evidence relevant to the decision problem was identified. The manufacturer of PhVIT did not submit any clinical effectiveness or cost-effectiveness evidence to the National Institute for Health and Clinical Excellence in support of PhVIT. The results of the Assessment Group's (AG) base-case analysis show that the comparison of PhVIT+HDA+AAI versus AAI+HDA yields an incremental cost-effectiveness ratio (ICER) of 18,065,527 per QALY gained; PhVIT+HDA+AAI versus avoidance advice only yields an ICER of 7,627,835 per QALY gained. The results of the sensitivity analyses and scenario analyses showed that the results of the base-case economic evaluation were robust for every plausible change in parameter made. The results of the 'High Risk of Sting Patients' subgroup analysis show that PhVIT+HDA+AAI dominates both AAI+HDA and avoidance advice only (i.e. is less expensive and more effective). The 'VIT Anxiety QoL Improvement' subgroup analysis shows that PhVIT+HDA+AAI versus HDA+AAI has an ICER of 23,868 per QALY gained, and PhVIT+HDA+AAI versus avoidance advice only yields an ICER of 25,661 per QALY gained. Limitations: This review is limited to the use of Pharmalgen in the treatment of hymenoptera venom allergy and therefore does not assess the effectiveness of VIT in general. Conclusions: The current use of PhVIT in clinical practice in the NHS appears to be based on limited and poor-quality clinical effectiveness research. Available evidence indicates that sting reactions following the use of PhVIT are low and that the ARs related to treatment are minor and easily treatable. The results of the AG's de novo economic evaluation demonstrate that PhVIT+AAI+HDA compared with AAI+HDA and with avoidance advice only yields ICERs in the range of 8-20M pound per QALY gained. Two subgroups ('High Risk of Sting Patients' and 'VIT Anxiety QoL Improvement') were considered in the economic evaluation and the AG concludes that the use of PhVIT+AAI+HDA may be cost-effective in both groups. Future research should focus on clearly identifying groups of patients most likely to benefit from treatment and ensure that clinical practice is focussed on these groups. Furthermore, given the paucity of UK data in this area it would be informative if data could be collected routinely when VIT is administered in the NHS (e.g. rates of systemic adverse reactions to VIT, rates of systemic reactions to bee/wasp stings). Funding: The National Institute for Health Research Health Technology Assessment programme.