Yao G, Albon E, Adi Y, Milford D, Bayliss S, Ready A, Raftery J, Taylor R S
Health Economics Facility, Health Services Management Centre, University of Birmingham, UK.
Health Technol Assess. 2006 Dec;10(49):iii-iv, ix-xi, 1-157. doi: 10.3310/hta10490.
To review the clinical and cost-effectiveness of basiliximab, daclizumab, tacrolimus, mycophenolate mofetil (MMF), mycophenolate sodium (MPS) and sirolimus as possible immunosuppressive therapies for renal transplantation in children.
Electronic databases were searched up to November 2004.
Data from selected studies were extracted and quality assessed. An economic model [Birmingham Sensitivity Analysis paediatrics (BSAp)] was produced based on an adaptation of a model previously developed for the assessment of the cost-effectiveness of immunosuppressants in adults following renal transplant.
For the addition of basiliximab, one unpublished paediatric randomised control trial (RCT), reported that the addition of basiliximab to tacrolimus-based triple therapy (BTAS) failed to significantly improve 6-month biopsy-proven acute rejection (BPAR), graft function, graft loss and all-cause mortality. No significant difference between groups was seen in 6-month or 1-year or longer graft loss, all-cause mortality and side-effects. In a meta-analysis of adult RCTs, the addition of basiliximab to a ciclosporin, azathioprine and steroid regimen (CAS) significantly reduced short-term BPAR. There was no significant difference in short- or long-term graft loss, all-cause mortality or side-effects. One adult RCT was included for the addition of daclizumab to CAS, which reported reduced 1-year BPAR, although no difference between groups was seen in either 1- or 3-year graft loss, all-cause mortality and side-effects. For tacrolimus versus ciclosporin, one unpublished paediatric RCT found that a regimen of tacrolimus, azathioprine and a steroid (TAS) reduced 6-month BPAR and improved graft function [glomerular filtration rate (GFR)] compared with CAS. This improvement in BPAR with tacrolimus was as shown in the meta-analysis of adult RCTs. There was evidence, particularly in children, that in comparison with ciclosporin, tacrolimus may reduce long-term graft loss, although there is no benefit on total mortality. The total level of withdrawal in children was reduced in children receiving tacrolimus. Adult RCTs showed an increase in post-transplant diabetes mellitus with tacrolimus. For MMF versus azathioprine, a meta-analysis of adult RCTs showed MMF [regimen of ciclosporin, MMF and a steroid (CMS)] to reduce 1-year BPAR compared with azathioprine (CAS). There was evidence, particularly in children, that in comparison with azathioprine, tacrolimus may reduce long-term graft loss, although there is no benefit on total mortality. There was an increase in the level of cytomegalovirus infection with MMF, although the overall level of withdrawal due to adverse events was not different to that of azathioprine-treated adults. No study comparing MPS with azathioprine (CAS) was identified. In an adult RCT comparing MMF with MPS, there was no significant difference between groups in 1-year efficacy or side-effects. One unpublished paediatric RCT assessed the addition of sirolimus to CAS. BPAR, graft loss and all-cause mortality were not reported. In two adult RCTs, compared with azathioprine, sirolimus reduced 1-year BPAR, reduced graft function (as assessed by an increased serum creatinine) and increased the level of hyperlipidaemia. No significant differences were seen in other efficacy and side-effect outcomes. On an adult RCT comparing sirolimus with ciclosporin, there were no significant differences between groups in 1-year efficacy or side-effects with the exception of an increased level of hyperlipidaemia with sirolimus substitution. Both the assessment group and drug companies assessed the cost-effectiveness of the newer renal immunosuppressants currently licensed in children using an adaptation (BSAp) of the Birmingham Sensitivity Analysis (BSA) model. This model is based on a 10-year extrapolation of 1-year BPAR results sourced from paediatric RCTs or adult RCTs (where paediatric RCTs were not available). The addition of basiliximab and that of daclizumab to CAS was found to increase quality-adjusted life-years (QALYs) and decreased overall costs, a finding that was robust to sensitivity analyses. The incremental cost-effectiveness ratio (ICER) of replacing ciclosporin with tacrolimus was highly sensitive to the selection of the hazard ratio for graft loss from acute rejection, dialysis costs and the incorporation (or not) of side-effects. The ICERs for tacrolimus versus ciclosporin ranged from about 46,000 pounds/QALY to about 146,000 pounds/QALY. Although sensitive to varying the hazard ratio for graft loss with acute rejection, the ICER for replacing azathioprine with MMF remained in excess of 55,000 pounds/QALY.
In general, compared with a regimen of ciclosporin, azathioprine and steroid, the newer immunosuppressive agents consistently reduced the incidence of short-term biopsy-proven acute rejection. However, evidence of the impact on side-effects, long-term graft loss, compliance and overall health-related quality of life is limited. Cost-effectiveness was estimated based on the relationship between short-term acute rejection levels from RCTs and long-term graft loss. Both the addition of daclizumab and that of basiliximab were found to be dominant strategies, that is, regarding cost savings and increased QALYs. The incremental cost-effectiveness of tacrolimus relative to ciclosporin was highly sensitive to key model parameter values and therefore may well be a cost-effective strategy. The incremental cost-effectiveness of MMF compared with azathioprine, although also sensitive to model parameter, was unattractive. There is a particular need for RCTs to assess the use of MMF, MPS and daclizumab for renal transplantation in children where no such evidence currently exists. Future comparative studies need to report not only on the impact of the newer immunosuppressants on short- and long-term clinical outcomes but also on side-effects, compliance, healthcare resource, costs and health-related quality of life.
综述巴利昔单抗、达利珠单抗、他克莫司、霉酚酸酯（MMF）、麦考酚钠（MPS）和西罗莫司作为儿童肾移植可能的免疫抑制疗法的临床疗效和成本效益。
检索截至2004年11月的电子数据库。
提取所选研究的数据并进行质量评估。基于先前开发的用于评估肾移植后成人免疫抑制剂成本效益的模型改编，建立了一个经济模型[伯明翰敏感性分析儿科模型（BSAp）]。
对于添加巴利昔单抗，一项未发表的儿科随机对照试验（RCT）报告称，在基于他克莫司的三联疗法（BTAS）中添加巴利昔单抗未能显著改善6个月活检证实的急性排斥反应（BPAR）、移植肾功能、移植肾丢失和全因死亡率。在6个月、1年或更长时间的移植肾丢失、全因死亡率和副作用方面，各治疗组之间无显著差异。在一项成人RCT的荟萃分析中，在环孢素、硫唑嘌呤和类固醇方案（CAS）中添加巴利昔单抗可显著降低短期BPAR。在短期或长期移植肾丢失、全因死亡率或副作用方面无显著差异。纳入一项将达利珠单抗添加到CAS中的成人RCT，该研究报告1年BPAR降低，尽管在1年或3年的移植肾丢失、全因死亡率和副作用方面，各治疗组之间无差异。对于他克莫司与环孢素的比较，一项未发表的儿科RCT发现，与CAS相比，他克莫司、硫唑嘌呤和类固醇方案（TAS）可降低6个月BPAR并改善移植肾功能[肾小球滤过率（GFR）]。他克莫司在BPAR方面的改善与成人RCT的荟萃分析结果一致。有证据表明，尤其是在儿童中，与环孢素相比，他克莫司可能会降低长期移植肾丢失，尽管对总死亡率无益处。接受他克莫司治疗的儿童停药总水平有所降低。成人RCT显示，使用他克莫司后移植后糖尿病的发生率增加。对于MMF与硫唑嘌呤的比较，一项成人RCT的荟萃分析显示，与硫唑嘌呤（CAS）相比，MMF[环孢素、MMF和类固醇方案（CMS）]可降低1年BPAR。有证据表明，尤其是在儿童中，与硫唑嘌呤相比，他克莫司可能会降低长期移植肾丢失，尽管对总死亡率无益处。MMF治疗的巨细胞病毒感染水平有所增加，尽管因不良事件导致的总体停药水平与硫唑嘌呤治疗的成人无差异。未找到比较MPS与硫唑嘌呤（CAS）的研究。在一项比较MMF与MPS的成人RCT中，各治疗组在1年疗效或副作用方面无显著差异。一项未发表的儿科RCT评估了在CAS中添加西罗莫司的情况。未报告BPAR、移植肾丢失和全因死亡率。在两项成人RCT中，与硫唑嘌呤相比，西罗莫司降低了1年BPAR，降低了移植肾功能（通过血清肌酐升高评估）并增加了高脂血症水平。在其他疗效和副作用结果方面无显著差异。在一项比较西罗莫司与环孢素的成人RCT中，除西罗莫司替代后高脂血症水平增加外，各治疗组在1年疗效或副作用方面无显著差异。评估组和制药公司均使用伯明翰敏感性分析（BSA）模型的改编版（BSAp）评估了目前在儿童中获批的新型肾免疫抑制剂的成本效益。该模型基于从儿科RCT或成人RCT（如无儿科RCT）中获取的1年BPAR结果进行10年的外推。发现将巴利昔单抗和达利珠单抗添加到CAS中可增加质量调整生命年（QALY）并降低总体成本，这一结果在敏感性分析中具有稳健性。用他克莫司替代环孢素的增量成本效益比（ICER）对急性排斥反应导致的移植肾丢失的风险比、透析成本以及副作用的纳入（或不纳入）高度敏感。他克莫司与环孢素的ICER范围约为46,000英镑/QALY至约146,000英镑/QALY。尽管对急性排斥反应导致的移植肾丢失的风险比变化敏感，但用MMF替代硫唑嘌呤的ICER仍超过55,000英镑/QALY。
总体而言，与环孢素、硫唑嘌呤和类固醇方案相比，新型免疫抑制剂持续降低了短期活检证实的急性排斥反应的发生率。然而，关于对副作用、长期移植肾丢失、依从性和总体健康相关生活质量的影响的证据有限。成本效益是根据RCT的短期急性排斥反应水平与长期移植肾丢失之间的关系进行估计的。发现添加达利珠单抗和巴利昔单抗均为优势策略，即在成本节约和增加QALY方面。他克莫司相对于环孢素的增量成本效益对关键模型参数值高度敏感，因此很可能是一种具有成本效益的策略。MMF与硫唑嘌呤相比的增量成本效益，尽管也对模型参数敏感，但并不具有吸引力。特别需要进行RCT来评估MMF、MPS和达利珠单抗在儿童肾移植中的应用，目前尚无此类证据。未来的比较研究不仅需要报告新型免疫抑制剂对短期和长期临床结果的影响，还需要报告对副作用、依从性医疗资源、成本和健康相关生活质量的影响。
