Machine perfusion systems and cold static storage of kidneys from deceased donors

The assessment group identified 7 studies that compared at least 2 of: the RM3 renal preservation system, the LifePort kidney transporter, Belzer UW storage solution and Marshall's hypertonic citrate solution. Two of these studies, which were both retrospective record reviews, compared the 2 machines. Three of the studies compared the LifePort kidney transporter with Belzer UW storage solution (2 ongoing randomised controlled trials and 1 retrospective review) and 1 cohort study compared the LifePort kidney transporter with Marshall's storage solution. One study compared the 2 different storage solutions. No studies were identified that compared the RM3 renal preservation system with either of the storage solutions.

Two retrospective record reviews (reported as abstracts) compared the 2 machine preservation systems. One study (744 kidneys transplanted) was a review over a 5-year period that included a change in practice from the use of the RM3 renal preservation system to the LifePort kidney transporter. The kidneys included in this study were from extended criteria deceased heart-beating donors (78%) or non-heart-beating donors (22%). The second study (89 kidneys transplanted) reviewed transplant records over a 22-month period and included kidneys mainly from deceased heart-beating donors (98%). Reporting in both studies was insufficient for a thorough assessment of quality. The relative risks were calculated by the assessment group.

In the larger study, rates of primary non-function were reported as 3% and 2% in the RM3 and LifePort groups, respectively (relative risk [RR] 1.44; 95% confidence interval [CI] 0.59 to 3.54; p=0.42). Rates of delayed graft function were reported as 24% and 32% in the RM3 and LifePort groups, respectively (RR 0.76; 95% CI 0.62 to 0.94; p<0.01). Patient survival and graft survival were both reported as 97% and 93% in the RM3 and LifePort groups, respectively (RR 1.05; 95% CI 1.01 to 1.08; p<0.01). The smaller study reported graft survival at a different follow-up point. At 30 days this was 97% and 94% in the RM3 and LifePort groups, respectively (RR 0.97; 95% CI 0.89 to 1.06; p=0.49), and at 90 days 97% and 90% (RR 0.93; 95% CI 0.83 to 1.04; p=0.21).

Two ongoing randomised controlled trials and one retrospective record review compared Belzer UW storage solution with the LifePort kidney transporter. One study (the Machine Preservation Trial, 672 kidneys transplanted) included mainly kidneys from deceased heart-beating donors but also some from non-heart-beating donors. The other study (the PPART study, 90 kidneys transplanted) included only kidneys from non-heart-beating donors. The primary outcome in both studies was rate of delayed graft function. The retrospective record review (36 kidneys transplanted) included kidneys from non-heart-beating donors. The primary outcome for this study was immediate graft function.

The Machine Preservation Trial study reported a small statistically significant benefit in terms of graft survival favouring the use of machine perfusion. Further detailed results of the Machine Preservation Trial study were provided as academic-in-confidence and are not included in this document.

The PPART study reported no statistically significant differences between the LifePort kidney transporter and Belzer UW storage solution at 3-month follow-up. Rates of primary non-function were reported as 2% and 0% in the LifePort kidney transporter and Belzer UW storage solution groups, respectively (RR 3.00; 95% CI 0.13 to 71.74; p value not reported). Rates of delayed graft function were reported as 58% and 56% in the LifePort kidney transporter and Belzer UW storage solution groups, respectively (RR 1.04; 95% CI 0.73 to 1.49; p=0.99). Patient survival was reported as 98% and 100% in the LifePort kidney transporter and Belzer UW storage solution groups, respectively (RR 0.98; 95% CI 0.92 to 1.04; p=0.32). Rates of graft survival were reported as 96% and 100% in the LifePort kidney transporter and Belzer UW storage solution groups, respectively (RR 0.96; 95% CI 0.89 to 1.03; p=0.16).

The retrospective record review reported statistically significant results favouring the use of the LifePort kidney transporter compared with Belzer UW storage solution. Delayed graft function was reported as 28% and 89% in the LifePort kidney transporter and Belzer UW storage solution groups, respectively (RR 0.31; 95% CI 0.15 to 0.67; p<0.001).

One sequential cohort study (60 kidneys transplanted) compared Marshall's hypertonic citrate solution with the LifePort kidney transporter. This study included kidneys from non-heart-beating donors, where death was controlled. For the first 2 years of the study all kidneys were stored using the solution, after this point they were stored using the perfusion machine. The significance tests reported were calculated by the assessment group.

This study reported that no kidneys suffered from primary non-function. Rates of delayed graft function were reported as 53% and 87% in the LifePort kidney transporter and Marshall's hypertonic citrate solution groups, respectively (RR 0.64; 95% CI 0.43 to 0.93; p=0.012). The rates of patient survival and graft survival were reported as the same. After 1 year of follow-up survival rates were reported as 100% and 93% in the LifePort kidney transporter and Marshall's hypertonic citrate solution groups, respectively (RR 1.07; 95% CI 0.96 to 1.20; p=0.24); 2-year survival rates were 97% and 90%, respectively (RR 1.07; 95% CI 0.94 to 1.23; p=0.30).

One retrospective record review (58,607 kidneys transplanted) of kidneys from deceased donors included in the Collaborative Transplant Study database included data for kidneys stored using either Belzer UW storage solution (53,560 kidneys) or Marshall's hypertonic citrate solution (5,047 kidneys). This study specifically considered differences in graft survival of kidneys that had undergone different durations of cold ischaemia.

The assessment group's analysis of the data from the study suggests no statistically significant differences between the 2 solutions. The 3-year graft survival rates in kidneys that had had up to 18 hours of cold ischaemic time were 81% and 80% in the Belzer UW storage solution and Marshall's storage solution groups, respectively (RR 1.02; 95% CI 0.99 to 1.04; p=0.13). The 3-year graft survival rates in kidneys that had had more than 36 hours of cold ischaemic time were 75% and 73% in the Belzer UW storage solution and Marshall's storage solution groups, respectively (RR 1.03; 95% CI 0.96 to 1.11; p=0.45). Comparing different durations of cold ischaemic time, the study suggests that the incidence of graft failure increases as cold ischaemic time increases.

The manufacturers of the technologies did not submit economic analyses. The assessment group identified 2 published economic analyses, one from the UK and another from Canada, both using a healthcare system perspective. The UK study reported cost per quality-adjusted life year (QALY), while the Canadian study reported cost per delayed-graft-function event avoided. Both studies reported that machine perfusion was associated with lower costs and greater benefits than cold static storage. Both economic analyses were completed before the most recent data from the PPART and Machine Preservation Trial studies became available.

The assessment group developed an economic model that made 3 comparisons. First, LifePort machine perfusion was compared with Belzer UW storage solution. This comparison was completed in 2 different populations: kidneys from non-heart-beating donors using data from the PPART study and kidneys mainly from deceased heart-beating donors using data from the Machine Preservation Trial study. Second, LifePort machine perfusion was compared with Marshall's hypertonic citrate solution using data from a cohort study. Third, Belzer UW storage solution was compared with Marshall's hypertonic citrate using data from a retrospective record review.

The assessment group was unable to do any cost-effectiveness analyses that included the RM3 machine perfusion system because cost data, although requested, were not made available.

The model was a Markov state transition model that included the health states immediate graft function, delayed graft function, transplant failure, explantation and a return to dialysis, and subsequent transplantation. The characteristics of the cohort modelled were chosen to be consistent with data obtained from UK Transplant and The Renal Registry. The cohort was followed up until almost all patients (97%) had died. The assessment group developed a standard data set for use in the model which was modified to reflect the comparisons described above.

Cost data for machine perfusion were annualised and it was assumed that perfusion machines were used for 10 years with no resale value afterwards. The estimated number of kidneys stored by each machine per year was calculated based on the total number of transplantations per year divided by the number of transplant centres. This estimate was 61 kidneys per year for analyses using data from the Machine Preservation Trial and 16 kidneys per year for analyses using data from the PPART study. The costs for machine perfusion also included an annual maintenance contract and the costs of the perfusion kit and solution used in the machine. This resulted in a cost per kidney stored of £544 for the analyses using data from Machine Preservation Trial and £737 for the analyses using data from PPART. The costs of storing a kidney using cold static storage included the costs of the solution and the box required to store the kidney. This was calculated to be £262.53 per kidney with Belzer UW storage solution and £49.73 per kidney with Marshall's solution.

Utility data were derived from published literature. The utility of living with a transplanted kidney varied according to age and was 0.83 for people aged 18 to 34 years, decreasing to 0.66 for people aged 65 years and older. The reduction in utility of living with dialysis was 0.12. Therefore, a person aged 18 to 34 years on dialysis had a utility of 0.71 and a person aged 65 years and older had a utility of 0.54. Renal registry data were used to model patient survival on dialysis and with a transplant; this rate was also varied according to age.

Data from the PPART study were used to model the cost effectiveness of LifePort compared with Belzer UW storage solution for the preservation of kidneys from non-heart-beating donors. The results of the deterministic analyses suggested that the LifePort kidney transporter was associated with greater cost than Belzer UW storage solution (£141,319 versus £139,205) and fewer QALYs (9.13 versus 9.19). Probabilistic sensitivity analyses predicted that over a range of willingness-to-pay levels (£0 to £100,000) the probability of LifePort being cost effective was about 40%.

Data from the Machine Preservation Trial study were used to model the cost effectiveness of LifePort compared with Belzer UW storage solution for the preservation of kidneys mainly from deceased heart-beating donors. The results of the deterministic analyses suggested that Belzer UW storage solution was associated with greater cost than the LifePort kidney transporter (£142,805 versus £139,100) and fewer QALYs (9.58 versus 9.79). Probabilistic sensitivity analyses predicted that over a range of willingness-to-pay levels (£0 to £100,000) the probability of LifePort being cost effective was 80%.

Data from the cohort study (described in section 4.1.8) were used to model the cost effectiveness of LifePort compared with Marshall's hypertonic citrate for the preservation of kidneys from controlled non-heart-beating donors. The results of the deterministic analyses suggested that Marshall's hypertonic citrate solution was associated with greater cost than the LifePort kidney transporter (£144,332 versus £132,953) and fewer QALYs (8.55 versus 9.54). Probabilistic sensitivity analyses predicted that over a range of willingness-to-pay levels (£0 to £100,000) the probability of LifePort being cost effective was 95%.

Data from the retrospective record review (described in section 4.1.10) were used to model the cost effectiveness of Marshall's hypertonic citrate compared with Belzer UW storage solution for the preservation of kidneys from deceased donors. This study analysed kidneys by duration of cold ischaemia. The cost-effectiveness analyses were based on kidneys that had 19 to 24 hours of cold ischaemic time. For these kidneys graft survival at 3 years was reported as 79.5% and 77.7% in the Belzer UW storage solution and Marshall's hypertonic citrate solution groups, respectively. The results of the deterministic analyses suggested that Marshall's hypertonic citrate solution was associated with greater cost than Belzer UW storage solution (£151,826 versus £151,001) and fewer QALYs (8.57 versus 8.62). Probabilistic sensitivity analyses predicted that over a range of willingness-to-pay levels (£0 to £100,000) the probability of Marshall's storage solution being cost effective was 40%.

The appraisal committee reviewed the data available on the clinical and cost effectiveness of machine perfusion systems and cold static storage of donated kidneys, having considered evidence on the nature of the condition and the value placed on the benefits of improvements in access to viable kidneys for transplantation by people with established renal failure, those who represent them, and clinical specialists. It was also mindful of the need to take account of the effective use of NHS resources.

The committee considered the process of retrieving donated organs and the methods for their storage. The committee was aware that it was important to minimise the length of ischaemic time regardless of the storage method used, in order to reduce the detrimental impact of ischaemia on the donated kidney. It also recognised that kidneys could be obtained from different types of donors and that type of donor is associated with differences in rates of delayed graft function and overall graft survival. The committee understood that minimising primary non-function and early and late graft loss was important. Unsuccessful transplantation has a significant physical and psychological effect on the recipient of the kidney and could reduce the chance of successful reimplant because of exposure to antigens. The committee understood that kidneys from brainstem-dead (that is, deceased heart-beating) donors are allocated nationally and that kidneys from non-heart-beating donors are allocated locally. However, it was aware that transplant services may be reorganised as a result of recent recommendations to the government from the Organ Donation Taskforce. The committee concluded that kidneys from different types of donor needed to be considered separately and that the mechanism for allocating kidneys could influence the choice of storage method.

The committee considered the different machine perfusion systems. It was aware that the RM3 is not portable and therefore does not replace cold static storage if transportation is needed. The committee noted that although clinical effectiveness evidence comparing the RM3 with the LifePort kidney transporter was available, it had methodological limitations because it was based on retrospective record reviews rather than prospective studies. The committee recognised that the assessment group had been unable to complete any cost-effectiveness analyses that included the RM3 because no cost data were made available by the manufacturer. The committee concluded that it could only issue recommendations for storage methods whose cost is known.

The committee specifically considered the use of machine perfusion to assess the viability of kidneys before implant. The committee heard from clinical specialists that there was no clear experimental evidence to support testing the viability of the kidney using the machine, but that they considered viability testing to be potentially important. The committee heard that clinical experience of machine perfusion systems, and knowledge of kidney function after transplantation, may enable clinicians to identify factors associated with poor viability. The committee was aware that a retrospective analysis had been proposed as part of the Machine Preservation Trial study but had not yet been completed. The committee concluded that although viability testing is potentially important, currently there is insufficient evidence to make this a deciding factor in choice of storage methods.

The committee considered the differences in clinical effectiveness between the 2 cold static storage solutions. The committee noted that the analysis by the assessment group suggested no statistically significant differences between the 2 solutions. The committee noted that Marshall's hypertonic citrate solution is used to store kidneys in a large proportion of centres, although choice of solution may depend on several factors. The committee heard from clinical specialists that there are differences in viscosity between Marshall's hypertonic citrate solution and Belzer UW storage solution, which affected the choice of solution in some cases. The committee additionally heard that for multiorgan donation that included the pancreas, Marshall's hypertonic citrate solution was not considered suitable for cooling of the organs in situ. The committee also heard that where a longer cold ischaemic time is anticipated, clinicians consider Belzer UW storage solution to be more suitable than Marshall's hypertonic citrate solution. However, cold ischaemic times greater than 24 hours are generally avoided in the UK, unless an organ is reallocated, or national allocation means that the kidney has a long transport time. These factors are, however, difficult to predict beforehand. The committee concluded that the clinical effectiveness evidence did not support a general preference for one storage solution over another but that both these clinical and logistical considerations need to be taken into account when choosing between storage solutions.

The committee considered the clinical effectiveness evidence for the use of the LifePort kidney transporter for the storage of kidneys from deceased heart-beating donors. In considering the clinical effectiveness evidence, the committee was mindful of comments from a consultee about the additional time required to attach the donated kidney to the LifePort kidney transporter, and the impact that this may have on the retrieval process for the kidneys and other organs. The committee noted the results of the Machine Preservation Trial study. The committee was aware that this study included mainly kidneys from deceased heart-beating donors, which does not reflect the type of kidneys for which the machines are usually used in the NHS. The committee considered that this study suggested a small statistically significant benefit in terms of graft survival favouring the use of machine perfusion. The committee heard from clinical specialists that these small benefits in rates of graft survival are important, but that factors such as discard rates before implant are also important. The committee heard clinical specialists express concern about the exclusion of a large number of kidneys from the statistical analysis in the Machine Preservation Trial study, and the effect that these exclusions may have had on results. The committee also heard from clinical specialists that the potential advantages of machine perfusion are not necessarily considered greater for the storage of kidneys from deceased heart-beating donors because success rates with cold storage solutions are already high. The committee concluded that machine perfusion may be marginally more clinically effective than Belzer UW storage solution for the storage of kidneys from deceased heart-beating donors. However, it was mindful of possible clinical considerations for choosing between machine perfusion and cold static storage and also the clinical specialists' comments that further evidence was required before the benefits of the LifePort kidney transporter over cold static storage can be fully demonstrated.

The committee considered the clinical effectiveness evidence for the use of the LifePort kidney transporter for the storage of kidneys from non-heart-beating donors. The committee noted that the PPART study had shown no statistically significant differences between the LifePort kidney transporter and cold static storage using Belzer UW storage solution. The committee heard from clinical specialists that the results of the PPART study were not consistent with clinical opinion or practice for storing this type of kidney. The committee also noted comments from a consultee about possible limitations in the reproducibility of the results of this study. The committee was mindful of the preliminary nature of the data from the PPART study and considered whether the availability of longer-term data would change the conclusions. The committee heard from clinical specialists that they did not think that the overall conclusion of the PPART study would change as more data become available. The committee noted that the Machine Preservation Trial study reported results from a subgroup of kidneys from non-heart-beating donors in whom death was controlled. The committee recognised that preliminary analyses suggested benefits to delayed graft function, but did not yet suggest differences in primary non-function and graft survival. The committee was also aware that a cohort study had compared the LifePort kidney transporter and Marshall's hypertonic citrate solution for the storage of kidneys from non-heart-beating donors where cardiac death had been expected. The committee noted that this study had shown a statistically significant difference that favoured the LifePort kidney transporter for rate of delayed graft function, but that there were methodological limitations with the design of the study. The committee concluded that the clinical effectiveness evidence did not allow it to distinguish between the LifePort kidney transporter and cold static storage for the storage of kidneys from non-heart-beating donors.

The committee considered the economic modelling carried out by the assessment group and noted the assumptions about the costs of the different storage methods. The committee heard from clinical specialists that clinicians may use different quantities of the storage solution, varying between 2 and 8 litres per kidney. The committee noted that the assessment group had assumed that 2 LifePort kidney transporters were required for each transplant centre, but that in clinical practice more machines may be required to ensure that a machine is readily available. The committee also noted comments from a consultee that in some locations an extra person was employed to supervise the LifePort kidney transporter, and that consumables would be wasted if kidneys were prepared for machine perfusion and then found not to be suitable. However, the committee was persuaded that the upfront costs of storage are much smaller than the costs of dialysis for failed grafts used in the model and that differences in the costs of storage for different methods would have little effect on the results of cost-effectiveness analysis. The committee concluded that while specific methods of storing kidneys may differ between centres, which would affect the cost of the technologies, this would not change the results of the cost-effectiveness analyses.

The committee considered the cost-effectiveness evidence for the use of the different methods of kidney storage. The committee understood that the cost-effectiveness results were driven by differences in the rate of graft survival between storage methods, because better graft survival led to fewer people on dialysis, which reduced costs and improved health-related quality of life.

The committee noted that the results of the cost-effectiveness analyses suggested that Marshall's hypertonic citrate solution is associated with greater cost and fewer QALYs than Belzer UW storage solution. However, the committee noted that the data suggested small differences in clinical effect between the 2 solutions, which led to small differences in both costs and QALYs. The committee concluded that no robust differences in clinical effectiveness had been shown. It recommended that the cheapest solution be used if the solutions are otherwise considered equally suitable, bearing in mind the clinical considerations that might affect the choice (described in 4.3.5).

The committee considered the cost-effectiveness evidence for the LifePort kidney transporter compared with cold static storage solutions. The committee noted that the results of the cost-effectiveness analyses suggested that Belzer UW storage solution was associated with lower costs and more QALYs, when using data from PPART. However, using data from the Machine Preservation Trial study, the LifePort kidney transporter was associated with lower costs and more QALYs. The committee noted these results were also based on small differences in costs and QALYs. Taking into consideration the testimony of clinical specialists and the clinical effectiveness evidence, the committee was not persuaded that the LifePort kidney transporter could be preferentially recommended for the storage of kidneys from deceased donors over other forms of storage. Given that the overall costs and benefits associated with kidney transplantation using either machine perfusion or cold static storage were similar, the committee recommended that the LifePort kidney transporter be considered as an alternative to cold static storage solutions and that the choice of which to use would depend on clinical and logistical factors within both the retrieval team and transplant centres.