Abstract
Background:
Over 700 donor livers are discarded annually in the United States due to high risk of poor graft function. The objective of this study was to determine the impact of using normothermic machine perfusion to identify transplantable livers among those currently discarded.
Study Design:
A series of 21 discarded human livers underwent viability assessment during normothermic machine perfusion. Cross-sectional analysis of the Scientific Registry of Transplant Recipients database and cost analysis was performed to extrapolate the case series to national experience.
Results:
21 discarded human livers were included in the perfusion cohort. 11 of 20 (55%) eligible grafts met viability criteria for transplantation. Grafts in the perfusion cohort had a similar donor risk index compared to discarded grafts (n=1402) outside of New England in 2017 and 2018 (median [IQR]: 2.0 [1.5, 2.4] vs. 2.0 [1.7, 2.3], P=0.40). 705 (IQR 677–741) livers were discarded annually in the United States since 2005, translating to the potential for 398 additional transplants nationally. The median cost to identify a transplantable graft with machine perfusion was $28,099 USD.
Conclusions:
Normothermic machine perfusion of discarded livers could identify a significant number of transplantable grafts, significantly improving access to liver transplantation.
Keywords: machine perfusion, liver, transplant, normothermic, cost analysis, organ shortage
Introduction
Liver transplantation (LT) provides the only effective therapy for end-stage liver disease. Unfortunately, only two-thirds of the approximately 13,000 patients on the waitlist receive a life-saving LT1. Efforts to alleviate the organ shortage include increasing living donation, optimizing allocation models2, and improving utilization of grafts from extended-criteria donors (ECD), such as those from older donors and donation after circulatory death (DCD), or those with macrosteatosis, abnormal liver function tests (LFTs), or significant alcohol or drug use history3. These livers are often declined because poor post-transplant function and graft survival rates as low as 20% pose too great a risk for the recipient4, 5.
Interestingly, a landmark European clinical trial of normothermic machine perfusion (NMP) demonstrated a 50% lower rate of organ discard with NMP than with cold storage (CS), in spite of longer warm ischemic and total preservation times. Availability of functional metrics during NMP likely increased surgeon confidence in these grafts, a judgment subsequently supported by the finding that more aggressive acceptance of organs in the NMP arm was associated with no significant difference in 1-year graft and patient survival. In fact, there was a clinically significant decrease in post-reperfusion syndrome (PRS) and early allograft dysfunction (EAD) among recipients of NMP livers, in spite of lower quality graft metrics prior to NMP6.
Thus, one avenue for increasing utilization of ECD grafts while minimizing recipient risk is systematic NMP of discarded grafts. Several European single center reports of successful transplantation of ECD and discarded grafts with NMP have demonstrated its potential benefit in this cohort7–11. In the United States (US), widespread use of MP has lagged behind Europe by several years due to delays in regulatory approval, though several randomized clinical trials sponsored by device manufacturers are currently ongoing or recently completed. However, these trials are all still limited to relatively standard-criteria livers routinely accepted for transplant and the value of NMP in identifying transplantable organs among those discarded in the US has yet to be evaluated.
Therefore, we assessed graft viability rates during NMP in a cohort of 21 human livers that were rejected for transplantation in the Northeastern United States. In order to extrapolate our experience to a national scale, we compared graft metrics of our cohort to those discarded regionally and nationally. We then estimated the cost per transplantable graft obtained, if NMP were deployed for all discarded livers.
Materials and Methods
Discarded Livers
21 livers declined for transplantation by all centers, with consent for research, between February and October 2019 were included in this study. Livers were not accepted for research if donors were hepatitis C virus (HCV) positive, if there was evidence of cirrhosis, or significant traumatic injury precluding perfusion. All livers were received through two organ procurement organizations: New England Donor Services (NEDS, n=17) and LiveOnNY (n=4). The Massachusetts General Hospital Institutional Review Board, NEDS, and LiveOnNY approved this study (No. 2011P001496). Livers were declined for transplantation due to a combination of factors (Table 1).
Table 1 –
Reasons for initial discard in cohort of grafts undergoing normothermic machine perfusion
| Liver # | Region of Origin | Reason for discard |
|---|---|---|
| 1 | 1 | Significant steatosis on biopsy |
| 2 | 1 | DCD with prolonged WIT, suspected significant steatosis on visual |
| 3 | 1 | Suspected significant steatosis on visual |
| 4 | 1 | DCD, older age |
| 5 | 1 | DCD with prolonged WIT |
| 6 | 1 | DCD with prolonged WIT, suspected steatosis on visual |
| 7 | 1 | Significant steatosis on biopsy |
| 8 | 1 | Significant steatosis on biopsy |
| 9 | 1 | DCD, older age |
| 10 | 1 | Significant steatosis on biopsy |
| 11 | 1 | DCD, older age |
| 12 | 1 | Elevated LFTs |
| 13 | 1 | Malignancy in liver after procurement |
| 14 | 9 | DCD, older age |
| 15 | 9 | Significant alcohol abuse |
| 16 | 1 | DCD, older age |
| 17 | 9 | DCD, older age |
| 18 | 9 | DCD, suspected significant steatosis on visual |
| 19 | 1 | DCD with prolonged WIT, older age |
| 20 | 1 | DCD, steatosis on visual, significant alcohol abuse |
| 21 | 1 | Significant steatosis on biopsy |
DCD, donation after circulatory death; LFTs, liver function tests; WIT, warm ischemic time; Region 1, New England; Region 9, New York. “Significant steatosis on biopsy” indicates that the local pathologist at the donor hospital suspected % macrosteatosis as being 30% or greater on frozen section histology.
All donor livers were procured based on the standard technique of in situ cold flush using University of Wisconsin (UW) preservation solution for both donation after brain death (DBD) and circulatory death (DCD). Livers were subsequently transported to the laboratory under cold storage. On arrival, back table preparation of the graft was performed as previously described12. Cold ischemic time (CIT) was defined as the period from in situ cold flush to start of NMP; total warm ischemic time (tWIT) was defined from extubation to cold flush; functional warm ischemic time (fWIT) was defined from circulatory arrest to cold flush.
Perfusion Solutions
The circulating perfusate consisted of O+ packed red blood cells, human albumin, lactated Ringer’s (LR) solution, and heparin. The perfusate was titrated with sodium bicarbonate and calcium gluconate during priming to obtain a pH > 7.2 and ionized calcium > 1.05. Bile salts (taurocholic acid) and parenteral nutrition (amino acids and electrolytes supplemented with insulin and glucose) were continuously infused into the perfusate during perfusion.
Immediately prior to perfusion, grafts were flushed with 3 liters of cold LR (2L via portal vein, 1L via hepatic artery), supplemented with 80mcg/L of methylprednisolone and 1mcg/L of epoprostenol. Detailed perfusate component information is provided in the S1 Methods.
Machine Perfusion Settings
The Liver Assist device (Organ Assist, Groningen, Netherlands) was used for each perfusion. A mixture of 95% oxygen and 5% carbon dioxide gas was used, with gas flow rate titrated to physiologic arterial carbon dioxide concentrations. After priming and titration of the perfusate, perfusion was initiated at 22°C and temperature was gradually increased to 37°C during the first 30 minutes. Target portal vein (PV) and hepatic artery (HA) pressures were 6–10 mmHg and 65–95 mmHg, respectively. PV and HA pressure were titrated to obtain an approximate 2:1 PV:HA flow ratio. Epoprostenol was infused via the arterial inflow to obtain a HA flow >300 mL/min when possible and weaned off when target flows were self-sustained. Grafts were perfused up to 12 hours. Blood, bile, and tissue biopsies were collected and analyzed according to the protocol in S2 Methods. Histology was assessed by a blinded pathologist (SGS).
Viability Assessment
Viability assessment was based on the criteria described by Laing et al.13. Grafts were deemed viable if they demonstrated all of the following criteria during the course of perfusion: trough lactate ≤3.0 mmol/L, stable perfusate pH > 7.3, bile production, sustained HA flow > 300 mL/min, and PV flow > 700 mL/min.
National Graft Discard Demographics
The Scientific Registry of Transplant Recipients (SRTR) database was queried for data regarding donor liver discards by region in 2017 and 2018. Discarded livers were defined as grafts recovered but not transplanted. In an effort to evaluate the generalizability of our pre-clinical results, discarded liver demographics were compared between the NMP cohort, Region 1 (New England), and the rest of the nation (Regions 2–11). Donor risk index (DRI)14 for discarded livers in the database was estimated by assuming 10 hours of CIT for all grafts based on the average CIT in the NMP cohort (9.3 hours).
Economic Impact of NMP
The cost of deploying NMP for a discarded liver was calculated using the sum of direct and indirect costs. To determine the range of perfusion costs per graft, the minimum, median, and maximum values per item were calculated in addition to incorporating perfusion durations of 4, 8, and 12 hours, respectively.
Direct costs included the cost of the perfusion device disposable, perfusate components (including infusions), and point-of-care equipment. The device costs were based on the three commercial liver perfusion devices currently available for clinical use in Europe or North America (Liver Assist, Groningen, Netherlands; OrganOx Metra, Oxford, United Kingdom; TransMedics, Andover, USA) and the hourly operational cost was calculated using the reported life expectancy of each device (each usage period defined as 12 hours). Indirect costs included personnel and facility fees, depreciation of the perfusion device, and an additional 5% for unexpected costs. Operating room costs for the time needed to prepare the liver for perfusion was included in the facility fees15. For personnel costs, the salary of a transplant surgeon and nurse for the time needed to prepare the liver for perfusion were included, whereas the perfusionist’s salary was calculated for the entire duration of perfusion16. Benefits, employer taxes, and training costs were also included17.
The cost to yield one discarded liver viable for transplantation was modeled as a function of the NMP viability rate and cost ranges per graft. The NMP viability rate was defined by the percentage of livers that met viability criteria during perfusion. To account for additional procurement procedures for potential grafts that were eligible for retrieval but not recovered, we further included a procedural cost range18.
A detailed summary of the cost calculation model, including all input parameters and assigned values in the minimum, median, and maximum cost scenarios is provided in Table S1 & S2.
Statistical Analysis
Categorical variables are provided as percentages and continuous variables as median with interquartile ranges (IQR). Cost data are provided as median with the minimum-maximum (min-max) range. Comparisons between groups were performed using Pearson’s Chi-squared test for categorical variables and Wilcoxon rank-sum test for continuous variables. A two-tailed P value < 0.05 was considered statistically significant. Statistical analysis was performed in StataSE 15.1 (StataCorp, College Station, USA).
Results
Discarded Liver Characteristics and Viability Assessment
Table 1 lists the reasons for each liver being turned down for transplantation. The most common reasons were steatosis (suspected or biopsy proven), DCD recovery in an older donor, and prolonged WIT. One graft (liver #13) was initially allocated but subsequently discarded after biopsy of a suspicious nodule identified a carcinoid tumor. Donor liver characteristics of viable and nonviable livers are shown in Table 2. There were no significant differences in warm ischemic times, donor age, or BMI between viable and nonviable livers. Interestingly, viable livers tended to have longer cold ischemic times compared to nonviable livers (671 [489–752] vs. 367 [358–557], P=0.060), while more nonviable livers were from diabetic donors (33 vs. 0%, P=0.031). Of note, nonviable livers had both significantly larger mass compared to viable livers (2.5 [2.3–2.9] vs. 1.9 [1.7–2.3] kg, P=0.023) and were more likely to have ≥30% macrosteatosis (66.7% vs. 8.3%, P=0.005). Detailed donor demographic data are provided in Table S3. Excluding the graft with malignancy, 11/20 (55%) grafts met viability criteria for transplantation (Table 3).
Table 2 –
Discarded liver characteristics undergoing normothermic machine perfusion
| Viable (n=12) |
Nonviable (n=9) |
P value | |
|---|---|---|---|
| DCD recovery | 7 (58.3%) | 7 (77.8%) | 0.35 |
| Total WIT (min) | 28 (24–33) | 26 (23–34) | 0.65 |
| Functional WIT | 9 (8–12) | 9 (9–11) | 0.75 |
| CIT (min) | 671 (489–752) | 367 (358–557) | 0.06 |
| Age (years) | 44 (39–59) | 55 (38–59) | 0.91 |
| Gender (female) | 3 (25%) | 5 (55.6%) | 0.15 |
| BMI (kg/m2) | 27.7 (24.3–31.6) | 29.9 (27.4–34.6) | 0.36 |
| Liver weight (kg) | 1.9 (1.7–2.3) | 2.5 (2.3–2.9) | 0.023 |
| Steatosis (≥30% large droplet macrosteatosis) | 1 (8.3%) | 6 (66.7%) | 0.005 |
| Smoker (active) | 5 (41.7%) | 5 (55.6%) | 0.53 |
| Heavy alcohol use | 2 (16.7%) | 2 (22.2%) | 0.75 |
| Drug use (active or former) | 5 (41.7%) | 2 (22.2%) | 0.35 |
| Donor risk index | 1.9 (1.6–2.15) | 2.1 (1.7–2.4) | 0.34 |
| Diabetes | 0 (0%) | 3 (33.3%) | 0.031 |
| Hypertension | 2 (16.7%) | 3 (33.3%) | 0.37 |
Total warm ischemic time (WIT) defined as extubation to flush, functional WIT defined as circulatory arrest to flush. DCD, donation after circulatory death; CIT, cold ischemic time; BMI, body mass index. Heavy alcohol use defined as more than 10 drinks per day. Pearson’s Chi-squared test used for categorical variables; Wilcoxon rank-sum test used for continuous variables. N(%) for categorial variables, median (IQR) for continuous variables.
Table 3 –
Viability assessment in discarded grafts undergoing normothermic machine perfusion
| Liver # | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Trough lactate | 0.6 | 1.9 | 10.9 | 6.6 | 19.8 | 2.2 | 14.7 | 0.6 | 6.6 | >20 | 1.0 | 4.7 | 3.0 | 0.8 | 1.6 | 0.7 | 0.8 | 0.6 | 7.6 | 13.9 | 2.6 |
| Stable pH>7.3 | Y | Y | N | Y | N | Y | N | Y | N | N | Y | Y | Y | Y | Y | Y | Y | Y | N | N | Y |
| Bile production | Y | Y | Y | Y | Y | Y | Y | Y | Y | N | Y | Y | Y | Y | Y | N | -- | Y | Y | N | Y |
| HAF>300 | Y | Y | Y | Y | N | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | N | Y |
| PVF>700 | Y | Y | Y | Y | N | Y | Y | Y | Y | N | Y | Y | Y | Y | Y | Y | Y | Y | Y | N | Y |
| Viable | Y | Y | N | N | N | Y | N | Y | N | N | Y | N | Ya | Y | Y | Y | Y | Y | N | N | Y |
viable but excluded due to malignancy, -- indicates unavailable data. HAF, hepatic artery flow (mL/min); PVF, portal vein flow (mL/min); Y, Yes/met criteria; N, No/did not meet criteria. Lactate values provided as mmol/L.
National Graft Utilization
The number of discarded livers in the United States has been relatively constant since 2005, with a median 705 grafts discarded annually (IQR 677–741). Based on the 55% viability rate in the perfusion cohort, if all discarded grafts underwent NMP an estimated 398 additional liver could meet viability criteria for potential transplant. We next compared graft characteristics of the perfusion cohort to all discards in Region 1 and Regions 2–11 for 2017–2018 (Table 4) to determine whether the perfusion cohort was a representative sample. Grafts discarded in Region 1 had a small but statistically significant lower DRI compared to the rest of the nation (1.8 [1.5–2.2] vs. 2.0 [1.7–2.3], P=0.012), with anoxia being the most common cause of death (COD). However, grafts in the perfusion cohort had a similar DRI to discarded grafts in Regions 2–11 (2.0 [1.5–2.4] vs. 2.0 [1.7–2.3], P=0.40) and were more often DCDs (66.7 vs. 30.1%, P=0.003).
Table 4 –
Regional comparison of donor demographics of discarded and non-recovered grafts in United States, 2017 and 2018
| Discarded | Eligible but Not Recovered | |||||||
|---|---|---|---|---|---|---|---|---|
| NMP Cohort (n=21) |
Region 1 (n=47) |
Regions 2–11 (n=1402) |
P valuea | P valueb | Region 1 (n=139) |
Regions 2–11 (n=3074) |
P valueb | |
| Age (years) | 47 (38, 59) | 41 (27, 52) | 48 (34, 57) | 0.66 | 0.02 | 46 (36, 55) | 46 (32, 55) | 0.26 |
| Gender (female) | 8 (38.1%) | 20 (42.6%) | 569 (40.6%) | 0.82 | 0.79 | 49 (35.3%) | 1221 (39.7%) | 0.29 |
| Race | ||||||||
| White | 20 (95.2%) | 32 (68.1%) | 928 (66.2%) | 0.09 | 0.20 | 109 (78.4%) | 2219 (72.2%) | 0.23 |
| Black | 1 (4.8%) | 3 (6.4%) | 173 (12.3%) | 8 (5.8%) | 317 (10.3%) | |||
| Hispanic | 0 | 12 (25.5%) | 230 (16.4%) | 19 (13.7%) | 424 (13.8%) | |||
| Asian | 0 | 0 | 44 (3.1%) | 3 (2.2%) | 67 (2.2%) | |||
| Other | 0 | 0 | 27 (1.9%) | 0 | 47 (1.5%) | |||
| BMI (kg/m2) | 28 (25, 33) | 29 (25, 36) | 29 (25, 34) | 0.97 | 0.36 | 29 (24, 34) | 28 (24, 34) | 0.73 |
| Cause of death | ||||||||
| Anoxia | 6 (28.6%) | 31 (66.0%) | 593 (42.3%) | <0.001 | <0.001 | 87 (62.6%) | 1389 (45.2%) | <0.001 |
| CVA | 6 (28.6%) | 4 (8.5%) | 483 (34.5%) | 27 (19.4%) | 757 (24.6%) | |||
| Trauma | 7 (33.3%) | 9 (19.1%) | 284 (20.3%) | 22 (15.8%) | 795 (25.9%) | |||
| Other | 2 (9.5%) | 3 (6.4%) | 42 (3.0%) | 3 (2.2%) | 133 (4.3%) | |||
| Diabetes | 3 (14.3%) | 7 (14.9%) | 264 (18.8%) | 0.60 | 0.50 | 11 (7.9%) | 421 (13.7%) | 0.051 |
| Hypertension | 5 (23.8%) | 20 (43.5%) | 645 (46.4%) | 0.043 | 0.70 | 45 (33.1%) | 1197 (39.5%) | 0.14 |
| CDC Increased Risk | 5 (23.8%) | 11 (23.4%) | 327 (23.3%) | 0.96 | 0.99 | 58 (41.7%) | 749 (24.4%) | <0.001 |
| Hepatitis C virus positive | 0 (0%) | 3 (6.4%) | 103 (7.4%) | 0.20 | 0.80 | 31 (22.3%) | 267 (8.7%) | <0.001 |
| DCD recovery | 14 (66.7%) | 14 (29.8%) | 422 (30.1%) | 0.003 | 0.96 | 94 (67.6%) | 1696 (55.2%) | 0.004 |
| Donor risk index | 2.0 (1.5, 2.4) | 1.8 (1.5, 2.2) | 2.0 (1.7, 2.3) | 0.40 | 0.012 | 2.3 (1.9, 2.7) | 2.2 (1.8, 2.7) | 0.19 |
BMI, body mass index; CVA, cerebrovascular accident; CDC, Center for Disease Control; DCD, donation after circulatory death. Pearson’s Chi-squared test used for categorical variables; Wilcoxon rank-sum test used for continuous variables. Categorical variables presented as N(%), continuous variables presented as median (interquartile range).
indicates comparison between NMP cohort and Regions 2–11;
indicates comparison between Region 1 and Regions 2–11.
In addition to livers recovered but not transplanted (discarded), a larger pool of grafts was eligible for recovery but not procured (Table 4). As the reasons for failure to procure are not adequately specified in the database, we considered failure to recover kidneys as a surrogate for factors that would preclude donation other than organ-specific quality (such as malignancy or intra-abdominal sepsis). In 2018, 1747 donors were eligible for liver retrieval but not procured, of which 3.5% also did not have kidneys recovered, leaving 1685 additional grafts potentially eligible for procurement and viability assessment with NMP. There are no data on viability rates of these organs since none were offered for research, but we note that the DRI in unrecovered organs trends higher than in discarded organs in both Region 1 and Regions 2–11, although this difference is not statistically significant (Table 4). Although unrecovered organs were more likely to be DCD than discarded organs, the percentage DCD in our NMP cohort was similar to that of unrecovered organs in both Region 1 and Regions 2–11.
Economic Impact of NMP
The estimated median NMP cost per graft in a back-to-base setting was $15,454 USD (min-max range: $7,725-$49,855) of which 89.8% (min-max: 89.7–94.8%) were fixed and 10.2% (min-max: 5.2–10.3%) variable with duration of perfusion. The distribution between the different cost categories is shown in Figure 1a and Table S4. The cost of the device and its disposable unit had a substantial impact on the total NMP costs per graft and disproportionately varied between the minimum, median, and maximum cost scenarios compared to all other costs.
Figure 1 – Itemized cost breakdown and cost effectiveness of NMP of donor livers.
a) The minimum, median, and maximum cost scenarios to perform NMP on one liver graft is shown in US dollars. The itemized cost ranges demonstrate relatively minor changes between cost scenarios except for the device materials that substantially change the total cost in each scenario. b) The minimum, median, and maximum costs to identify one viable graft from a pool of discarded livers is provided as a function of the NMP salvage rate. Costs to identify one viable graft with NMP decrease substantially as the viability rate approaches 100%. Solid colored lines represent cost scenarios were the procurement costs are excluded as an estimate of the pool of grafts already recovered but discarded; colored dashed lines represent cost scenarios that include procurement costs as an estimate of the additional pool of grafts eligible for retrieval. The dotted black line represents the estimated costs at the 55% viability rate demonstrated in this study. Min, minimum; Max, maximum; NMP, normothermic machine perfusion; USD, US dollars.
The cost to identify a viable discarded liver for transplantation is dependent on the NMP cost per graft and the viability rate (Fig. 1b). Using this study’s viability rate of 55%, the cost to identify one viable discarded graft for transplantation is $28,099 USD (min-max: $14,045-$90,646). For livers that otherwise would not have been recovered, incorporating the additional procurement costs raises the total cost to identify a viable graft to $44,957 (min-max: $29,649-$108,759).
Discussion
Functional assessment of livers declined for transplantation with NMP from the New England and New York regions revealed that 55% of grafts met viability criteria being used in European centers for subsequent transplantation with excellent outcomes. Extrapolation to similar discarded livers nationwide indicates the potential for an additional 400 livers to be transplanted each year in the United States if NMP were routinely used to assess declined livers. This could eliminate up to one-third of annual waitlist deaths, making a significantly larger impact than liver redistribution proposals recently put into effect with much controversy19. There are likely additional livers that would be found to be transplantable among the 1700 donor livers never recovered. In the absence of NMP experience with those livers, it is hard to predict the viability rate among those, though it is likely less than 55% given the slightly higher DRI of that cohort. This cohort may reflect severely injured grafts that would not respond to conventional short-term NMP but may benefit from longer-term perfusion strategies20, adjunct therapeutics aimed at graft rehabilitation21–23, or techniques that avoid additional ischemic injury such as ischemia-free LT24.
Although a portion of the discarded livers in the perfused cohort may have been considered transplantable without NMP at some transplant centers, these grafts were nonetheless declined by multiple centers in regions with long waitlists and high MELD patients, where these factors are associated with more aggressive donor organ use25. Whether they were declined for reasons not apparent from SRTR data or for logistic reasons (inability to minimize CIT or locate an appropriate recipient), NMP could have increased the likelihood of these grafts being transplanted in a suitable recipient, providing both functional assessment and mitigating any deleterious impact of prolonged preservation time26. Traditional cut-offs for graft acceptance, such as donor age over 50 or total WIT over 30 minutes, are likely to be de-emphasized with improvements in objective viability assessment27 and as experience with transplantation of NMP grafts increases7, 9, 11, 28. Moreover, NMP would deescalate the time-sensitive race to match, locate, and prepare a recipient for the operating room, especially in situations of rapid DCD procurement or last-minute organ offers.
The cost of NMP is often cited as a barrier to its widespread adoption, especially for standard criteria livers where NMP shows no significant clinical benefit over standard cold storage6. In this study, we consider the substantial clinical benefit of increasing the supply of donor organs and access to liver transplantation, and additionally report the first cost analysis of back-to-base NMP. The median cost to perform NMP on an already discarded graft was $15,454 USD (min-max: $7,725-$49,855), while the cost to identify one viable discarded graft with NMP was $28,099 USD (min-max: $14,045-$90,646), based on a viability rate of 55%. In addition to preventing a wait list death, this is only slightly more than the estimated monthly Medicare expenditure of $22,68529 for the care of a patient with MELD 30. As the viability rate with NMP increases, the cost to identify one viable graft will decrease, indicating that efforts to improve the success rate are needed. For example, steatotic grafts may benefit from functional rehabilitation with adjunct therapies during NMP30, 31. The expected benefits of NMP will be further evident in regions with longer waitlist times and disproportionately higher MELD scores.
One notable finding in our analysis was the significant variance in perfusion device costs; whether this will continue to represent a barrier to adoption following regulatory approval in the US remains to be seen. As NMP and its potential adjunct therapies continue to expand32, the added cost of incorporating perfusion into the total cost of a live-saving LT is a relatively small increase compared to the potential decreases in waitlist mortality33. Given the high startup cost needed to create a perfusion program, it may not be economically feasible to expect every transplant center to adopt this technology. Development of regionalized perfusion programs34 will likely be a more efficient use of resources, though the logistics of this effort will need to be established. Routine access to perfusion would also significantly decrease cold ischemic times, which were prolonged in this study due to its research nature.
Finally, there were several limitations to our study. First, the DRI of discarded grafts in Region 1 was noted to be lower than the rest of the nation. Anoxic cause of death related to the prevalence of opiate use disorder and HCV are higher in Region 1 than the rest of the country, and these variables are not fully captured in the DRI formula. More importantly, graft characteristics of livers in the NMP cohort closely mirror those of discarded grafts nationwide, and it is the transplantability rate among our perfused cohort that was extrapolated to a parallel national cohort. It is also worth noting that each discarded liver was turned down by at least six liver transplant centers before designation for research, indicating that discard patterns were not unique to a single center. Second, as there is currently no FDA approved perfusion device or clinical trial evaluating NMP for discarded livers, we were unable to transplant the grafts that met viability criteria to obtain post-transplant data. However, reports of clinical outcomes in other countries with similar criteria are largely positive6, 7, 35. It should also be noted that assessment of biliary tree health was limited to presence of bile production, though there is increasing evidence that bile production alone is not indicative of biliary health36, 37. The viability criteria used in this study were based on those used by Mergental et al. in a recently published clinical trial of transplantation of discarded livers after viability assessment with NMP38. This substantial study demonstrated 100% 90-day survival in 22 recipients of discarded livers that met viability criteria during NMP, though 4 (18%) recipients developed biliary strictures that required re-transplantation, highlighting the importance of biliary assessment in DCD livers. Therefore, conclusions about the long-term outcomes of grafts that met viability criteria in this study cohort should be interpreted cautiously, though short-term survival is expected to be excellent. As experience with various modalities of perfusion increases globally and in the United States, different perfusion techniques may be utilized to decrease the risk of ischemic cholangiopathy to the benefit of the recipient39, 40. Additional perfusion modalities or their combinations will ultimately serve to further decrease discards and improve organ utilization while maintaining the safety of the recipient.
Furthermore, although we did not enroll HCV positive livers due to laboratory biohazard restrictions, these grafts are being used clinically with excellent results41, indicating that serology alone likely does not play a large role in functional quality. Lastly, costs were obtained from a range of sources in the United States, making generalizability of these data limited to a US cohort. Given significant cost variance in other parts of the world, the potential benefits of NMP examined here may not apply abroad and international health systems should determine the benefits individually.
In conclusion, NMP in a US cohort of extended-criteria and discarded livers is likely to yield a significant number of grafts available for transplantation, thereby substantially increasing access to LT and potentially decreasing waitlist mortality. The projected cost of NMP to identify a viable graft is comparable to the monthly cost of care for waitlist patients with high MELD scores and should be expected to decrease as NMP viability rates improve with experience and adjunct therapies.
Supplementary Material
Acknowledgements:
We would like to gratefully acknowledge the patients, families, and staff of New England Donor Services (NEDS) and LiveOnNY for making this research possible.
Funding: Funding from the US National Institutes of Health (grants R01DK096075, R01DK107875, R21EB020819, R01DK084053 and R01DK114506) and support from the Shriners Hospitals for Children is gratefully acknowledged.
Footnotes
Conflicts of Interest: Dr. Uygun is inventor on pending patents relevant to this study and has a provisional patent application relevant to this study. Uygun has a financial interest in Organ Solutions, a company focused on developing organ preservation technology. Dr. Uygun’s interests are managed by the MGH and Partners HealthCare in accordance with their conflict of interest policies.
Supplementary Materials
Additional Supporting Information may be found online in the supporting information tab for this article.
S1 Methods – Detailed perfusate component information, concentrations, and source
S2 Methods – Perfusion protocol with data collection and sampling scheme
Table S1 – Itemized costs to perform normothermic machine perfusion per graft
Table S2 – Detailed personnel cost calculations
Table S3 – Detailed characteristics of discarded grafts undergoing normothermic machine perfusion
Table S4 – Cost scenarios to perform normothermic machine perfusion per graft
References
- 1.Organ Procurement and Transplantation Network (OPTN). https://optn.transplant.hrsa.gov/data/view-data-reports/. Accessed December 7, 2018.
- 2.Gentry SE, Massie AB, Cheek SW, Lentine KL, Chow EH, Wickliffe CE, Dzebashvili N, Salvalaggio PR, Schnitzler MA, Axelrod DA, Segev DL. Addressing geographic disparities in liver transplantation through redistricting. Am J Transplant. 2013;13(8):2052–2058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Vodkin I, Kuo A. Extended Criteria Donors in Liver Transplantation. Clin Liver Dis. 2017;21(2):289–301. [DOI] [PubMed] [Google Scholar]
- 4.Schlegel A, Kalisvaart M, Scalera I, Laing RW, Mergental H, Mirza DF, Perera T, Isaac J, Dutkowski P, Muiesan P. The UK DCD Risk Score: A new proposal to define futility in donation-after-circulatory-death liver transplantation. J Hepatol. 2018;68(3):456–464. [DOI] [PubMed] [Google Scholar]
- 5.Linares I, Hamar M, Selzner N, Selzner M. Steatosis in Liver Transplantation: Current Limitations and Future Strategies. Transplantation. 2019;103(1):78–90. [DOI] [PubMed] [Google Scholar]
- 6.Nasralla D, Coussios CC, Mergental H, Akhtar MZ, Butler AJ, Ceresa CDL, Chiocchia V, Dutton SJ, Garcia-Valdecasas JC, Heaton N, Imber C, Jassem W, Jochmans I, Karani J, Knight SR, Kocabayoglu P, Malago M, Mirza D, Morris PJ, Pallan A, Paul A, Pavel M, Perera M, Pirenne J, Ravikumar R, Russell L, Upponi S, Watson CJE, Weissenbacher A, Ploeg RJ, Friend PJ, Consortium for Organ Preservation in E. A randomized trial of normothermic preservation in liver transplantation. Nature. 2018;557(7703):50–56. [DOI] [PubMed] [Google Scholar]
- 7.Mergental H, Perera MT, Laing RW, Muiesan P, Isaac JR, Smith A, Stephenson BT, Cilliers H, Neil DA, Hubscher SG, Afford SC, Mirza DF. Transplantation of Declined Liver Allografts Following Normothermic Ex-Situ Evaluation. Am J Transplant. 2016;16(11):3235–3245. [DOI] [PubMed] [Google Scholar]
- 8.Watson CJE, Kosmoliaptsis V, Pley C, Randle L, Fear C, Crick K, Gimson AE, Allison M, Upponi S, Brais R, Jochmans I, Butler AJ. Observations on the ex situ perfusion of livers for transplantation. Am J Transplant. 2018;18(8):2005–2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Watson CJ, Kosmoliaptsis V, Randle LV, Russell NK, Griffiths WJ, Davies S, Mergental H, Butler AJ. Preimplant Normothermic Liver Perfusion of a Suboptimal Liver Donated After Circulatory Death. Am J Transplant. 2016;16(1):353–357. [DOI] [PubMed] [Google Scholar]
- 10.Ghinolfi D, Rreka E, De Tata V, Franzini M, Pezzati D, Fierabracci V, Masini M, Cacciatoinsilla A, Bindi ML, Marselli L, Mazzotti V, Morganti R, Marchetti P, Biancofiore G, Campani D, Paolicchi A, De Simone P. Pilot, Open, Randomized, Prospective Trial for Normothermic Machine Perfusion Evaluation in Liver Transplantation From Older Donors. Liver Transpl. 2019;25(3):436–449. [DOI] [PubMed] [Google Scholar]
- 11.Pezzati D, Ghinolfi D, Balzano E, De Simone P, Coletti L, Roffi N, Rreka E, Meacci L, Campani D, Mazzoni A, Paolicchi A, Caponi L, Marchetti P, Marselli L, Filipponi F. Salvage of an Octogenarian Liver Graft Using Normothermic Perfusion: A Case Report. Transplant Proc. 2017;49(4):726–728. [DOI] [PubMed] [Google Scholar]
- 12.Karimian N, Matton AP, Westerkamp AC, Burlage LC, Op den Dries S, Leuvenink HG, Lisman T, Uygun K, Markmann JF, Porte RJ. Ex Situ Normothermic Machine Perfusion of Donor Livers. J Vis Exp 2015(99):e52688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Laing RW, Mergental H, Yap C, Kirkham A, Whilku M, Barton D, Curbishley S, Boteon YL, Neil DA, Hubscher SG, Perera M, Muiesan P, Isaac J, Roberts KJ, Cilliers H, Afford SC, Mirza DF. Viability testing and transplantation of marginal livers (VITTAL) using normothermic machine perfusion: study protocol for an open-label, non-randomised, prospective, single-arm trial. BMJ Open. 2017;7(11):e017733. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Feng S, Goodrich NP, Bragg-Gresham JL, Dykstra DM, Punch JD, DebRoy MA, Greenstein SM, Merion RM. Characteristics associated with liver graft failure: the concept of a donor risk index. Am J Transplant. 2006;6(4):783–790. [DOI] [PubMed] [Google Scholar]
- 15.Childers CP, Maggard-Gibbons M. Understanding Costs of Care in the Operating Room. JAMA Surg. 2018;153(4):e176233. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Occupational Employment Statistics. www.bls.gov. Accessed November 5, 2019, 2019.
- 17.Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford: Oxford University Press; 2015. [Google Scholar]
- 18.DuBay DA, Redden DT, Bryant MK, Dorn DP, Fouad MN, Gray SH, White JA, Locke JE, Meeks CB, Taylor GC, Kilgore ML, Eckhoff DE. Resource utilization associated with procurement of transplantable organs from donors that do not meet OPTN eligible death criteria. Transplantation. 2014;97(10):1043–1048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Weaver T, Schladt DP, Pyke J, Thompson B, Herreid A, Lake JR, Kim WR, Snyder JJ, Kasiske BL, Israni AK, Gentry SE. SRTR Liver Simulated Allocation Model (LSAM) Analysis Report: U.S. Department of Health & Human Services; September 24 2018. [Google Scholar]
- 20.Eshmuminov D, Becker D, Bautista Borrego L, Hefti M, Schuler MJ, Hagedorn C, Muller X, Mueller M, Onder C, Graf R, Weber A, Dutkowski P, Rudolf von Rohr P, Clavien PA. An integrated perfusion machine preserves injured human livers for 1 week. Nat Biotechnol. 2020;38(2):189–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhang A, Carroll C, Raigani S, Karimian N, Huang V, Nagpal S, Beijert I, Porte RJ, Yarmush M, Uygun K, Yeh H. Tryptophan Metabolism via the Kynurenine Pathway: Implications for Graft Optimization during Machine Perfusion. J Clin Med. 2020;9(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chin LY, Carroll C, Raigani S, Detelich DM, Tessier SN, Wojtkiewicz GR, Schmidt SP, Weissleder R, Yeh H, Uygun K, Parekkadan B. Ex vivo perfusion-based engraftment of genetically engineered cell sensors into transplantable organs. PLoS One. 2019;14(12):e0225222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Laing RW, Stubblefield S, Wallace L, Roobrouck VD, Bhogal RH, Schlegel A, Boteon YL, Reynolds GM, Ting AE, Mirza DF, Newsome PN, Mergental H, Afford SC. The Delivery of Multipotent Adult Progenitor Cells to Extended Criteria Human Donor Livers Using Normothermic Machine Perfusion. Frontiers in Immunology. 2020;11(1226). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhang Z, Tang Y, Zhao Q, Wang L, Zhu C, Ju W, Wang D, Yang L, Wu L, Chen M, Huang S, Ningxin G, Zhu Z, Zhang Y, Sun C, Xiong W, Shen Y, Ma Y, Hu A, Zhu X, Rong J, Cai C, Guo Z, He X. Association of perfusion characteristics and post-transplant liver function in ischemia-free liver transplantation. Liver Transpl. 2020. [DOI] [PubMed] [Google Scholar]
- 25.Yeh H, Smoot E, Schoenfeld DA, Markmann JF. Geographic inequity in access to livers for transplantation. Transplantation. 2011;91(4):479–486. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jassem W, Xystrakis E, Ghnewa YG, Yuksel M, Pop O, Martinez-Llordella M, Jabri Y, Huang X, Lozano JJ, Quaglia A, Sanchez-Fueyo A, Coussios CC, Rela M, Friend P, Heaton N, Ma Y. Normothermic Machine Perfusion (NMP) Inhibits Proinflammatory Responses in the Liver and Promotes Regeneration. Hepatology. 2019;70(2):682–695. [DOI] [PubMed] [Google Scholar]
- 27.Linares-Cervantes I, Echeverri J, Cleland S, Kaths JM, Rosales R, Goto T, Kollmann D, Hamar M, Urbanellis P, Mazilescu L, Ganesh S, Adeyi OA, Yip P, Gorynska P, Bojko B, Gorynski K, Grant DR, Selzner N, Wasowicz M, Selzner M. Predictor parameters of liver viability during porcine normothermic ex situ liver perfusion in a model of liver transplantation with marginal grafts. Am J Transplant. 2019;19(11):2991–3005. [DOI] [PubMed] [Google Scholar]
- 28.Manzia TM, Toti L, Quaranta C, Blasi F, Tisone G. Liver transplantation with a normothermic machine preserved fatty nonagenarian liver: A case report. Int J Surg Case Rep. 2019;57:163–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Axelrod DA, Dzebisashvili N, Lentine K, Segev DL, Dickson R, Tuttle-Newhall E, Freeman R, Schnitzler M. Assessing variation in the costs of care among patients awaiting liver transplantation. Am J Transplant. 2014;14(1):70–78. [DOI] [PubMed] [Google Scholar]
- 30.Raigani S, Karimian N, Huang V, Zhang AM, Beijert I, Geerts S, Nagpal S, Hafiz EOA, Fontan FM, Aburawi MM, Mahboub P, Markmann JF, Porte RJ, Uygun K, Yarmush M, Yeh H. Metabolic and lipidomic profiling of steatotic human livers during ex situ normothermic machine perfusion guides resuscitation strategies. PLoS One. 2020;15(1):e0228011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Boteon YL, Attard J, Boteon A, Wallace L, Reynolds G, Hubscher S, Mirza DF, Mergental H, Bhogal RH, Afford SC. Manipulation of Lipid Metabolism During Normothermic Machine Perfusion: Effect of Defatting Therapies on Donor Liver Functional Recovery. Liver Transpl. 2019;25(7):1007–1022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Raigani S, De Vries RJ, Uygun K, Yeh H. Pumping new life into old ideas: Preservation and rehabilitation of the liver using ex situ machine perfusion. Artif Organs. 2020;44(2):123–128. [DOI] [PubMed] [Google Scholar]
- 33.Wai H, Stepanova M, Saab S, Erario M, Srishord M, Younossi ZM. Inpatient economic and mortality assessment for liver transplantation: a nationwide study of the United States data from 2005 to 2009. Transplantation. 2014;97(1):98–103. [DOI] [PubMed] [Google Scholar]
- 34.Yeung JC, Cypel M, Keshavjee S. Ex-vivo lung perfusion: the model for the organ reconditioning hub. Curr Opin Organ Transplant. 2017;22(3):287–289. [DOI] [PubMed] [Google Scholar]
- 35.Ceresa CDL, Nasralla D, Watson CJE, Butler AJ, Coussios CC, Crick K, Hodson L, Imber C, Jassem W, Knight SR, Mergental H, Ploeg RJ, Pollok JM, Quaglia A, Shapiro AMJ, Weissenbacher A, Friend PJ. Transient Cold Storage Prior to Normothermic Liver Perfusion May Facilitate Adoption of a Novel Technology. Liver Transpl. 2019;25(10):1503–1513. [DOI] [PubMed] [Google Scholar]
- 36.Gaurav R, Atulugama N, Swift L, Butler AJ, Upponi S, Brais R, Allison M, Watson CJE. Bile biochemistry following liver reperfusion in the recipient and its association with cholangiopathy. Liver Transpl. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Matton APM, de Vries Y, Burlage LC, van Rijn R, Fujiyoshi M, de Meijer VE, de Boer MT, de Kleine RHJ, Verkade HJ, Gouw ASH, Lisman T, Porte RJ. Biliary Bicarbonate, pH, and Glucose Are Suitable Biomarkers of Biliary Viability During Ex Situ Normothermic Machine Perfusion of Human Donor Livers. Transplantation. 2019;103(7):1405–1413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Mergental H, Laing RW, Kirkham AJ, Perera M, Boteon YL, Attard J, Barton D, Curbishley S, Wilkhu M, Neil DAH, Hübscher SG, Muiesan P, Isaac JR, Roberts KJ, Abradelo M, Schlegel A, Ferguson J, Cilliers H, Bion J, Adams DH, Morris C, Friend PJ, Yap C, Afford SC, Mirza DF. Transplantation of discarded livers following viability testing with normothermic machine perfusion. Nat Commun. 2020;11(1):2939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Schlegel A, Muller X, Kalisvaart M, Muellhaupt B, Perera M, Isaac JR, Clavien PA, Muiesan P, Dutkowski P. Outcomes of DCD liver transplantation using organs treated by hypothermic oxygenated perfusion before implantation. J Hepatol. 2019;70(1):50–57. [DOI] [PubMed] [Google Scholar]
- 40.de Vries Y, Matton APM, Nijsten MWN, Werner MJM, van den Berg AP, de Boer MT, Buis CI, Fujiyoshi M, de Kleine RHJ, van Leeuwen OB, Meyer P, van den Heuvel MC, de Meijer VE, Porte RJ. Pretransplant sequential hypo- and normothermic machine perfusion of suboptimal livers donated after circulatory death using a hemoglobin-based oxygen carrier perfusion solution. Am J Transplant. 2019;19(4):1202–1211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Kapila N, Menon KVN, Al-Khalloufi K, Vanatta JM, Murgas C, Reino D, Ebaid S, Shaw JJ, Agrawal N, Rhazouani S, Navas V, Sheffield C, Rahman AU, Castillo M, Miller C, Quintini C, Zervos XB. HCV NAT positive solid organ allografts transplanted into HCV negative recipients: A real-world experience. Hepatology. 2019. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.