Introduction A new UK Lung Allocation Scheme (UKLAS) was introduced in 2017, replacing the previous geographic allocation system. Patients are prioritised according to predefined clinical criteria into a three-tier system: the super-urgent lung allocation scheme (SULAS), the urgent lung allocation scheme (ULAS) and the non-urgent lung allocation scheme (NULAS). This study assessed the early impact of this scheme on waiting-list and post-transplant outcomes.
Methods A cohort study of adult lung transplant registrations between March 2015 and November 2016 (era-1) and between May 2017 and January 2019 (era-2). Outcomes from registration were compared between eras and stratified by urgency tier and diagnostic group.
Results During era-1, 461 patients were registered. In era-2, 471 patients were registered (19 (4.0%) SULAS, 82 (17.4%) ULAS and 370 (78.6%) NULAS). SULAS patients were younger (median age 35 vs 50 and 55 for urgent and non-urgent, respectively, p=0.0015) and predominantly suffered from cystic fibrosis (53%) or pulmonary fibrosis (37%). Between eras 1 and 2, the odds of transplantation within 6 months of registration were increased (OR=1.41, 95% CI 1.07 to 1.85, p=0.0142) despite only a 5% increase in transplant activity. Median time-to-transplantation during era-1 was 427 days compared with waiting times in era-2 of 8 days for SULAS, 15 days for ULAS and 585 days for NULAS patients. Waiting-list mortality (15% era-1 vs 13% era-2; p=0.5441) and post-transplant survival at 1 year (81.3% era-1 vs 83.3% era-2; p=0.6065) were similar between eras.
Conclusion The UKLAS scheme prioritises the critically ill and improves transplantation odds. The true impact on waiting-list mortality and post-transplant survival requires further follow-up.
- Lung Transplantation
- Thoracic Surgery
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
Donor lung allocation on the basis of recipient geographic location has been associated with a high waiting-list mortality of 25% within 2 years of listing for lung transplantation.
WHAT THIS STUDY ADDS
Shifting from a geographic to a clinical urgency based lung allocation scheme, prioritising the sickest patients, improved odds of transplantation with comparable waiting-list mortality and early post-transplant survival rates.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Lung allocation schemes based on clinical urgency can replace geographic allocation schemes.
In the UK, there are five adult lung transplant centres. Until 2017, for the purpose of donor lung allocation, the UK was divided into five geographic zones around each of these centres. Donor lungs were first allocated to the zonal lung transplant centre and then nationally to the other centres. There was no national ranking by clinical status. This scheme (figure 1), along with a shortage of suitable donor lungs, was associated with a high national waiting-list mortality of 25% within 2 years of being listed for lung transplantation and a survival rate of less than 50% at 5 years from listing with significant variation between centres.1
The National Health Service Blood and Transplant (NHSBT) Cardiothoracic Advisory Group proposed a revised UK Lung Allocation Scheme (UKLAS), where listed candidates in whom survival without lung transplantation is estimated to be less than 90 days are given priority access to the national donor lung pool, irrespective of geographic zones (figure 1). Two criteria-based urgent tiers were adopted; the super-urgent lung allocation scheme (SULAS) tier for patients requiring venovenous extracorporeal membrane oxygenation (VV-ECMO) and the urgent lung allocation scheme (ULAS) tier for patients with deterioration in disease-specific physiological parameters (online supplemental appendix 1). Mechanical ventilation was not used as a criterion for urgent or super-urgent listing because of its association with poor post-transplant outcomes particularly when compared with the alternative ‘awake’ ECMO strategy.2 3 For patients not meeting agreed criteria for urgent status, an appeal to the UKLAS adjudication panel could be made. The revised UKLAS was introduced on 18 May 2017.4 Lungs from consented donors-after-brain-death aged <70 years or potential donors-after-circulatory-death aged <65 years, without absolute contraindications,5 were offered via the sequence shown in figure 1. A separate sequence existed for donors aged <16 years.6
This primary objective of this study was to assess the impact of the change in national lung allocation policy from zonal allocation to the new UKLAS, on waiting-list outcomes and post-transplant outcomes.
Data were extracted from the UK Transplant Registry (UKTR) held by NHSBT on 19 September 2019 regarding all adult patients registered for lung transplantation in the following two time periods pre-UKLAS and post-UKLAS: 18 March 2015 to 17 November 2016 (era-1) and 18 May 2017 to 17 January 2019 (era-2). A 6-month (‘washout’) gap was left between both eras to minimise any crossover effect. Later periods were not chosen to avoid the confounding impact of the SARS-CoV-2 pandemic, which resulted in a 48% decrease in the number of donors and a significant reduction in lung utilisation, 24% to 10% during the early pandemic period in the UK.7 Patients registered for combined heart–lung transplantation were excluded.
Within the 2017 UKLAS, patients registered in one tier could subsequently be re-registered in another tier if their clinical condition changed. To ensure mutually exclusive cohorts across eras and urgency groups for statistical purposes, patients were categorised by the highest urgency tier they had reached. For instance, if a patient was initially registered in era-1 and then subsequently transitioned to the non-urgent, urgent and then super-urgent tier in era-2, they would only appear in the era-2 cohort as a super-urgent patient. Hence, the same patient would not appear in both eras. Patients were further analysed by their primary lung diagnosis where these were grouped into the following categories: chronic obstructive pulmonary disease (COPD), cystic fibrosis and bronchiectasis (CF/B), pulmonary fibrosis (PF), pulmonary hypertension (PH) and other.
Patient demographics were compared between both eras, using the Fisher’s exact test for categorical variables, the t test for normally distributed continuous variables and the Wilcoxon rank sum otherwise. Additionally, in era-2, demographics were compared across the three urgency tiers using the Fisher’s exact test for categorical variables, analysis of variance (ANOVA) for normally distributed continuous variables and the Kruskal-Wallis test otherwise.
Median waiting time to transplant and unadjusted post-transplant survival were analysed using the Kaplan-Meier method and the log-rank test was used to compare across eras and urgency groups. For this analysis, patients were extracted from the UKTR for the two time periods based on the date of transplantation rather than registration. Transplant and registration cohorts were, therefore, not fully identical (71% of transplant recipients were present in the registration cohort). No direct comparisons were made between the transplant and registration cohorts. The odds of transplant and waiting-list mortality were assessed using logistic regression. Between eras and urgency group comparisons for ischaemic time, intensive and total hospital stay and length of mechanical ventilation were performed using the Kruskal-Wallis test. Transplant type and postoperative primary graft dysfunction (PGD) and mechanical circulatory support (MCS) use were compared using Fisher’s exact test. P<0.05 was considered statistically significant. Loss-to-follow-up and areas of missing data were described and no imputation methods were used when handling such data. All analyses were performed using SAS V.9.4 (SAS Institute, North Carolina).
The final study cohort comprised of 932 patients; 461 (49.5%) registered in era-1 and 471 (50.5%) in era-2. Of those registered in era-2, 370 (79%) were non-urgent, 82 (17%) were urgent and 19 (4%) were super-urgent (figure 2). The majority of patients achieved urgent status by meeting the agreed criteria, while 5% of patients were afforded urgent status after appeal to a national adjudication panel. Thirty-four urgent and eight super-urgent patients were initially registered prior to UKLAS and transitioned to the urgent and super-urgent tiers, respectively, following its introduction (table 1).
Baseline characteristics and distribution of primary diagnoses were similar across both eras (table 2). However, of the 471 patients in era-2, those in the super-urgent tier were noted to be significantly younger than those in the urgent and non-urgent categories (median age 35 years vs 50 and 55 for urgent and non-urgent, respectively, p=0.0015) and like those, in the urgent category, comprised predominantly of patients with diagnoses of CF/B and PF (collectively accounting for 89% and 78% of super-urgent and urgent recipients, respectively, vs 58% of those in the non-urgent tier; p<0.0001; table 1). Contrastingly, COPD accounted for a higher proportion of those in the non-urgent tier (32% vs 2% in the urgent tier and 0% in the super-urgent tier; table 1). There were statistically significant differences between centres in the proportion of registrations in each tier (p<0.0001; table 1).
The most common indication for urgent listing was PF with persisting hypoxia despite continuous oxygen therapy (42%; table 3). Excluding one patient with exceptional circumstances approved by a national adjudication panel, all super-urgent patients were supported on VV-ECMO at time of listing (n=18). In comparison, eight patients were identified as bridged on VV-ECMO in era-1. However, if VV-ECMO was instituted after registration to a patient who did not reach transplant, this was not captured in this era and, hence, may be under-represented.
Transplant and mortality rates
Although follow-up time on the list was longer in era-1 (median 240 days vs 170 days), all patients in era-2 had recorded at least a 6-month waiting-list outcome. During era-1, 8% of patients received a lung transplant within 1 month of registration, 2% died on the waiting list and 90% remained on the waiting list (figure 3). During era-2, 63% of super-urgent and 61% urgent patients received a lung transplant within 1 month of registration.
The shift to the new UKLAS scheme improved overall odds of transplantation at 6 months by 41% (OR 1.41, 95% CI 1.07 to 1.85, p=0.0142; table 4). Of 471 patients in era-2, 179 (38%) were transplanted by 6 months in comparison to 140 of 461 (30%) patients in era-1 (figure 3). The overall median waiting time reduced from 427 days (95% CI 312 to 542) in the era-1 to 292 days (95% CI 185 to 399) in era-2. Although deaths on the waiting list at 6 months (8% era-2 vs 10% era-1; OR=0.84, 95% CI 0.54 to 1.31, p=0.4337; table 4) and 1 year (13% era-2 vs 15% era-1; p=0.5441) were proportionally less in era-2, this was not statistically significant.
A total of 278 patients underwent lung transplantation in era-1 and 292 in the era-2, with the latter representing a 5% increase. Although this was coupled with a 15% increase in donor lung offers, donor lung utilisation rates remained low and relatively similar (22% during era-2 vs 24% during era-1). The proportion of single-lung to double-lung transplants did not differ between both cohorts and between urgency tiers (table 5).
Transplant and mortality rate by urgency tier (era-2 only)
When compared with non-urgent recipients, the odds of transplantation for super-urgent (OR: 25.9, 95% CI 9.3 to 72.0, p<0.0001) and urgent recipients (OR: 23.6, 95% CI 12.8 to 43.5, p<0.0001) were significantly higher. The median waiting time to lung transplantation was 8 days (95% CI 1 to 15) in the super-urgent tier, 15 days (95% CI 11 to 19) in the urgent tier and 585 days (95% CI 419 to 751) in the non-urgent tier (figure 4). As one would expect, and despite the shorter waiting time, the odds of death on the list were also significantly higher for super-urgent patients when compared with both urgent (OR: 7.0, 95% CI 1.6 to 29.2, p=0.0079) and non-urgent (OR: 18.5, 95% CI 5.2 to 65.7, p<0.0001) patients with 26% dying on the list within 1 month, reflecting the severity of their condition (figure 3). All super-urgent patients had reached an outcome of transplant, removal from the waiting list or death by 1 month. The median time to death for SULAS registrants was 20 days.
Transplant rate and mortality rate by diagnostic group
The odds of receiving a transplant within 6 months were similar across eras for patients with COPD and PF (p=0.7258 and p=0.5417, respectively; table 4). However, patients with CF/B (OR 1.98, 95% CI 1.18 to 3.33; p=0.0098) and PH/other (OR 2.56, 95% CI 1.10 to 5.95; p=0.0289) were significantly more likely to receive a transplant in era-2. The odds of death on the waiting list within 6 months did not change significantly between eras when analysed by diagnostic group (table 4 and figure 5).
The post-transplant analysis comprised 570 patients across both eras. A small number (n=48; 10%) of those included in the era-1 registration analysis were included in the era-2 transplant outcome analysis due to their transplant occurring during era-2. Median post-transplant follow-up times were 1047 days and 340 days for era-1 and era-2, respectively. There was only one patient lost to follow-up. There was no significant difference in 1-year survival following transplantation between both eras (81.3% era-1 vs 83.3% era-2; p=0.6065; table 5; figure 6) or between tiers in era-2 (p=0.1461). Survival was numerically worse in the urgent tier versus the non-urgent tier, but these differences were not statistically significant (p=0.0604). There were significant differences, between eras, in total ischaemia time, length of hospital stay, length of intensive care stay and length of mechanical ventilation (table 5). Between era-1 and era-2, the overall median total ischaemia time increased from 6.5 hours to 6.9 hours (p=0.0062) with super-urgent recipients generally enduring the longest durations of ischaemia. Moreover, although there was no difference observed between eras in postoperative MCS use, its use in super-urgent patients in era-2 was notably higher than other subgroups (45% vs 11% and 21% non-urgent and urgent recipients, respectively; p=0.0029). The incidence of PGD grade 3 within 72 hours of transplantation reduced significantly in era-2 (25% in era-1 vs 15% in era-2, p=0.0031; table 5). There was a trend, however, towards a higher incidence in the urgent and super-urgent groups compared with the non-urgent group in era-2 although this did not reach statistical significance (p=0.1729). Overall, the length of intensive care stay increased from 5 to 6 days (p=0.0222) between both eras with super-urgent patients requiring longer periods of mechanical ventilation (median 192 hours; p=0.0053). The use of Ex-Vivo Lung Perfusion was low across both periods (10 cases in era-1 and 8 cases in era-2; p=0.5585).
Donor lung allocation on the basis of urgency scoring has been shown to reduce waiting-list deaths and increase the number of transplants for sicker candidates in the USA.8 The revised UKLAS was developed to ensure timely allocation of donor lungs to those most in need. Geographic boundaries for lung allocation were abandoned in favour of a national waiting list prioritised by clinical urgency into super-urgent, urgent and non-urgent tiers.
Data from the first 20 months following implementation of the new scheme suggest that it has successfully delivered on its remit by reducing overall median waiting times by 135 days and increasing the odds of transplantation at 6 months by 41%. Furthermore, median waiting times to transplant of only 8 and 15 days for SULAS and ULAS registrations, respectively, were reported, which represent a dramatic reduction from the overall median of 427 days in the zonal allocation era. Although not statistically significant, there were proportionally fewer deaths on the waiting list within 6 months of registration (era-2 of 8% vs era-1 of 10%). The lack of clinically significant increases in donor availability and utilisation rates with implementation of the revised UKLAS implies that observed improvements in waiting times and odds of transplantation in era-2 were related to changes in allocation policy rather than significant increases in the donor pool.
Moreover, the 2017 UKLAS did not disadvantage non-urgent registrations with 30% and 27% of listed patients undergoing lung transplantation within 6 months of registration, before and after UKLAS introduction, respectively, with no difference in waiting-list deaths. This may be in-part due to the fact that super-urgent and urgent registrations accounted for a small fraction of the total cohort (4% and 17%, respectively, vs 79% non-urgent). Furthermore, it is conceivable that transplant centres were already prioritising their sickest patients internally during the zonal allocation period pre-UKLAS (era-1), thereby diluting the differences in outcomes between the two different lung allocation schemes.
One potential concern of allocating organs to the sickest patients is the potential for worsened post-transplant outcomes. Reassuringly, however, 90-day and 1-year post-transplant survival were not jeopardised, with no differences in overall survival between both eras. This is consistent with published data from the LAS system in the USA in 20059–13 and its subsequent roll-out in Europe where the negative impact on survival was not experienced.14 However, it is noteworthy that super-urgent and urgent recipients required longer hospital and intensive care stays and endured prolonged durations of mechanical ventilation and increased MCS use. UKLAS is, therefore, likely to result in increased early post-transplant healthcare resource use, similar to other urgency-based allocation systems (US LAS15–19 and the transplant-benefit liver and kidney allocation schemes in the UK.20 21
As donor lungs destined for super-urgent and urgent recipients are distributed nationwide, they tended to be transported further, lengthening donor cold ischaemia times. However, in a country the size of the UK, the observed difference in ischaemia times between the two eras (6.6 vs 6.9 hours) was small and probably clinically insignificant.22 This is in-part supported by the observation of reduced rates of PGD grade 3 in era-2. Shortened waiting-list times coupled with longer ischaemic times, maintained 1-year survival, and unchanged rates of transplant-related complications have similarly been reported in the USA after expanding the primary allocation zone around donor hospitals.23 Therefore, we would argue that the longer intensive care stay and ventilation dependency seen in urgent and super-urgent recipients would more likely reflect the severity of their pre-transplant critical illness and deconditioning.
Another concern was that introduction of new prioritisation mechanisms favouring patients on VV-ECMO may influence clinical practice incentivising an increase in those bridged to transplantation with extracorporeal support as was seen after introduction of the UK urgent heart allocation scheme.24 Reassuringly, however, early experience following the introduction of the 2017 UKLAS has not revealed such a trend with non-urgent transplants still representing the majority.24 Pretransplant VV-ECMO deployment rates did not increase following UKLAS, suggesting that this bridging strategy was not being used to manipulate or ‘game’ the allocation system.25
Inevitably, as sicker patients are being prioritised, UKLAS would likely alter the pattern of lung allocation to patients in certain diagnostic groups. Younger CF patients and older PF patients are known to have a poorer waiting-list prognosis26 and, hence, comprised the majority of super-urgent and urgent registrations. Understandably, CF patients comprised the majority (53%) of those bridged to surgery with VV-ECMO due to their younger age and their overall less deconditioned state making them more suitable candidates for mechanical bridging. Moreover, in the modern era of cystic fibrosis transmembrane conductance regulator modulators, CF patients can be sustained for longer periods, but their condition can decline precipitously, necessitating swift resort to ECMO as a bridge to super-urgent lung allocation.27–29 This raises the possibility that current CF urgency criteria may not be sensitive enough to capture patients who should be prioritised before the point of requiring ECMO rescue.
There were also notable improvements in the prospects of PH patients following UKLAS introduction, evidenced by shorter waiting times and a 2.5-fold increase in the odds of transplantation with 60% being registered as urgent or super-urgent. Similar observations were seen in the USA after LAS implementation.30 In contrast, COPD patients represented the minority of urgent registrations and none were mechanically bridged to transplantation. This is consistent with the lack of clear survival gains from lung transplantation for COPD and it is being primarily offered to improve quality-of-life.31 The odds of transplantation or death within 6 months of registration for COPD were unchanged after UKLAS (figure 5) suggesting that COPD patients were not disadvantaged because of the small percentage of donor lungs being redirected to SULAS and ULAS. With time, however, it is anticipated that fewer COPD patients will be listed, a similar trend to that was seen following US LAS implementation.15
Geographical variations in lung transplantation rates and outcomes have been reported in the UK26 32 and elsewhere,33 34 reflecting in-part variations in referral patterns and preferred listing practices. Data presented in this study also revealed significant differences between transplant centres in the proportion of registrations in each UKLAS priority tier.
In contrast to the LAS score-based system, UKLAS is a criteria-based allocation system. Although LAS reduced waiting-list mortality,8 its effect on long-term post-transplant survival has plateaued at survival rates of around 60% at 5 years due to its weighted emphasis on ‘transplant benefit’.11–13 Moreover, LAS remains bound by rigid geographic boundaries leaving certain patients at a disadvantage without medical or logistic reasoning.35 36 In contrast, UKLAS has removed all geographic boundaries for super-urgent and urgent recipients improving access to those most in need irrespective of geographic proximity.
On the other hand, the UKLAS criteria-based listing approach is less flexible than score-based systems. Use of excessively strict criteria for urgent and super-urgent registration may deny rapidly deteriorating patients access to urgent lung allocation and partially explain the lack of a mortality benefit in this early review. Moreover, such stringent criteria may in-part account for the lower postoperative survival reported particularly in the urgent UKLAS group, as eligibility for urgent listing required a more clinically advanced disease process. This practically comprised of an older recipient cohort and a higher proportion of deconditioned PF patients (49% vs 37% in the super-urgent and non-urgent groups, respectively; p<0.0001). UKLAS criteria are currently under review in an iterative process to resolve such issues and achieve the best listing algorithm.
There are limitations to our study. Adult lung transplants were analysed in two time periods, irrespective of when the patient was registered and, therefore, outcomes after registration and transplantation within each era were not from the same exact cohort (71% concordance between registration and transplant cohorts). Our figures may, therefore, not describe a continuous patient journey from listing through surgery and up to a minimum of 6 months postoperatively as initially intended. This is a product of the way UKTR databases are structured. Moreover, given that a proportion of patients included in the UKLAS cohort were initially listed before its introduction and transitioned onto this scheme, the potential impact of lead-time bias on post-UKLAS waiting time cannot be excluded. Furthermore, this study represents an early examination of the UKLAS scheme and its ongoing impact on waiting-list mortality will likely become more apparent with time. The short follow-up duration in this study may have affected the ability to detect a difference between eras if one existed (type-II error). The lack of long-term follow-up data impairs the ability to make definitive conclusions on the impact of the new scheme on post-transplant survival beyond 12 months. The sample size was also not sufficient to detect small changes in mortality or transplant rates across eras, especially when stratified by diagnostic group. Furthermore, as only 11 super-urgent transplants were performed in ECMO-supported patients, a lower survival in this group may only become apparent as more patients accrue. However, several analyses in the USA do not support this speculation,37 38 and data from the UK39 and Germany40 all reported similar early outcomes for those with and without ECMO bridging to lung transplantation.
In conclusion, the new UKLAS scheme has fulfilled its goals of prioritising the most critically ill and improving the odds of transplantation. However, although no clinically relevant improvement was seen in waiting-list deaths and the numbers of lung transplants, the true impact of the new scheme is yet to be seen and will likely continue to evolve as transplant teams adjust their practices to harness its full potential.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Patient consent for publication
Contributors JP and ST were responsible for the conception and design of the study and its critical review. SR contributed to the design, conduct and statistical analysis of the study. RDT, MC, JL, SC and HS contributed in the interpretation of data, and critical review of the study. AA-A and MA-A were responsible for data reporting, drafted the work and revised it to its current format following contribution from all coauthors. JP is the guarantor for this study. All coauthors reviewed and approved the final version of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.