Increased Use of Split-Liver Transplantation

Increased Use of Split-Liver Transplantation Can Relieve Organ Shortage and Bring More Equitable Outcomes for Patients

By Yanhan (Savannah) Tang, Alan Scheller-Wolf, and Sridhar Tayur
Tepper School of Business, Carnegie Mellon University

 

Liver transplantation is the only effective treatment for patients with end-stage liver diseases (ESLD). In the United States, the number of patients awaiting liver transplants far exceeds the number of available donated livers, [1]. Split liver transplantation (SLT)–a procedure that can save two lives using one liver (see Figure 1)—increases the benefit from the limited number of livers available, but it is rarely used in the United States. [5,6]

In our new working paper [8], which won first place in the inaugural INFORMS DEI Student Paper Competition in 2022, we developed a model to help improve liver allocation decisions and evaluate the potential of SLT. We found that increased use of SLT in the United States could yield significant benefits for patient outcomes and equity. The paper is conducted by researchers from Carnegie Mellon University (CMU) and physicians from University of California, San Francisco.

 

Figure 1: Anatomy of a human liver and two methods to split a liver for SLT. The yellow line illustrates an adult-child split where a child receives the smaller left lobe, and an adult receives the extended right lobe; the black line illustrates an adult-adult (or adult-big child) split where an adult receives the right lobe, and an adult (or big child) receives the left lobe. SLT is possible because human liver cells have strong regenerative capacity: When a partial liver is successfully transplanted into a human body, it can grow to its full size and function well with proper medical care [3].

 

Figure 2: The number of liver transplants (blue bars) and eligible patients waitlisted for liver transplantation (orange bars) in the US from 2017 - 2021. Despite the increase in the number of transplants, the organ shortage persists. There has been a steadily-increasing trend in the number of ESLD patients; common causes of ESLD include heavy drinking, viruses, a buildup of fat in the liver, among other reasons. Click here for more information about ESLD.

  

 As liver shortages persist despite countless efforts to bridge the gap between supply and need (see Figure 2), we must take steps to ensure that the livers we have are allocated as efficiently and fairly as possible. This should include expanding the use of SLT: Fewer than 1.5% of livers are split for transplantation in the United States (while more than 10% are medically safe for SLT), whereas the United Kingdom splits all donated livers except in certain circumstances (an “all-split” policy). Barriers to increased use of SLT in the United States include logistical difficulties, lack of surgical expertise, geographical challenges, and complexities associated with donor-recipient matching.

“Increased SLT use benefits physically smaller patients, who are disadvantaged without SLT.”

 

Figure 3: A negative correlation between death proportion on waitlists and patient height. ESLD patients of shorter stature are more likely to die before a liver transplant becomes available [2].

 

Increased use of SLT may improve equity by giving transplant candidates who are physically smaller increased access to liver transplants, as most donated livers are too large to transplant into smaller patients without the use of SLT.  Indeed, previous studies have found that there is a negative correlation between death proportions on waitlists and patient stature (see Figure 3) since transplant candidates of smaller sizes (e.g., children, smaller adults) typically wait longer for a new liver and consequently suffer more waitlist mortality risk.

“The transplant community has an ethical obligation to maximize the outcomes from donated organs, while also promoting equity.” ---OPTN/UNOS Ethics Committee [5]

The Organ Procurement and Transplantation Network (OPTN) oversees all organ procurement and allocation in the US. When a deceased-donor liver becomes available, the organ is sequentially offered to appropriate ESLD patients on the waiting list according to a ranked ordering; patients may accept or reject the organ offer. The US allocation policy prioritizes the sickest patients when other factors, such as sizes, geographical locations, and blood types, are compatible.  By default, livers are allocated as a whole to the sickest; only in exceptional cases where a highly ranked child or small adult requires a partial liver will a liver be split and used for SLTs, if medically safe.

A 2016 OPTN white paper published by OPTN’s ethics committee commented: “The transplant community has an ethical obligation to maximize the outcomes from donated organs, while also promoting equity.” In terms of the choice of SLT and the traditional method, whole liver transplantation (WLT), the white paper said, “The Committee affirms that optimal allocation policies involving whole liver or split liver allografts should reflect a balance between the principles of equity and utility.” [5]

We modeled the deceased-donor liver allocation and dynamic-matching problem using a fluid model, incorporating the use of SLT for the first time. Our model and analysis offer insights into dynamic liver allocation in realistic settings, incorporating fairness concerns. Specifically, we consider a group probability fairness notion, where the probability of getting a transplant for patients in the worst-off size group is lower-bounded by chosen thresholds [4, 8].  We found that increasing the use of SLT could improve utility and equity, confirming statements from the 2016 OPTN white paper [5] that “Split liver transplantation might improve net survival of liver transplant candidates, while also increasing the number of individuals who benefit from transplantation, especially children. Splitting suitable livers for suitable patients is appropriate in centers with adequate experience and outcomes.”

Specifically, if we can split all splittable livers and then allocate them to the highest-ranking (i.e., sickest) compatible patients (i.e., not changing the current US liver allocation policy), we may be able to significantly reduce waitlist death rates for physically smaller patients with virtually no detrimental effect to the larger sized patient groups (See Figure 4).  Of course, if one can optimize allocation rules as well – which is administratively more difficult – even greater benefits are potentially possible.

 

Figure 4: Under the “sickest-first” organ allocation system (which approximates the current US liver allocation system), splitting all splittable livers and then allocating to the sickest can potentially reduce waitlist death rates for physically smaller patients, especially for children and small adults, with only a tiny increase in the large patients’ waitlist mortality. Only very few livers (fewer than 1.5%) are currently split for SLT (i.e., few split, sickest first). Here S, M, ML, and L stand for Small, Medium, Medium-Large, and Large, respectively.

The larger issue: Who gets livers first?

 The current OPTN organ allocation gives priority to the sickest patients; however, under common transplant objectives (e.g., maximizing overall transplant survival, maximizing quality-adjusted life years, reducing patient deaths, reducing waitlist deaths), this is unlikely to be a very good strategy. Our model provides an approach to optimize the organ allocation and priority rules according to OPTN’s choice of transplant objectives. At a high level, we prioritize the patients that bring the most overall benefits to society (e.g., gain the most quality-adjusted life years out of a transplant while reducing disparities among patient groups). Based on current U.S. data, splitting all splittable livers and then allocating the whole/partial livers according to our model solution would have nearly optimal results in terms of boosting total quality-adjusted years of life while also reducing the total number of patient deaths and increasing equity: A win-win-win situation.

Looking forward

In addition to encouraging the wider use of SLT, we hope our insights inspire more detailed analyses, including cost-benefit analyses, and foster discussions in the transplant community about expanding the use of SLT and incorporating fairness into allocation rules.

Our model also has uses beyond transplantation. It could be used to address issues such as optimal scheduling of proactive service with dynamic patient health conditions, resource sharing among multiple queueing classes with customer abandonment, and hospitals’ management of patient flow, where equity is an important criterion [8].

We are honored to have been selected as the winner of the 2022 INFORMS DEI Student Paper Competition; the competition provides a wonderful opportunity for students to share their DEI-related research, gain exposure, and receive high-quality feedback and potential recognition. We strongly recommend participating in future DEI Student Paper Competitions.  Interested readers are invited to discover more about SLT and organ allocation by reading our paper [8] and related papers [1, 7, 9].

 

References

[1] OPTN/SRTR 2020 Annual Data Report: Liver. Link to report.

  1. Akan, M., Alagoz, O., Ata, B., Erenay, F. S., & Said, A. (2012). A broader view of designing the liver allocation system. Operations research, 60(4), 757-770.
  2. Bernards, S., Lee, E., Leung, N., Akan, M., Gan, K., Zhao, H., Sarkar, M., Tayur, S. & Mehta, N. (2022). Awarding additional MELD points to the shortest waitlist candidates improves sex disparity in access to liver transplant in the United States. American Journal of Transplantation.
  3. Hackl, C., Schmidt, K. M., Süsal, C., Döhler, B., Zidek, M., & Schlitt, H. J. (2018). Split liver transplantation: current developments. World Journal of Gastroenterology, 24(47), 5312.
  4. Kim, T. W., Roberts, J., Strudler, A., & Tayur, S. (2022). Ethics of split liver transplantation: should a large liver always be split if medically safe?. Journal of Medical Ethics, 48(10), 738-741.
  5. OPTN and UNOS. Split Versus Whole Liver Transplantation OPTN/UNOS Ethics Committee. Public Comment Proposal, 2016.
  6. Perito, E. R., Roll, G., Dodge, J. L., Rhee, S., & Roberts, J. P. (2019). Split liver transplantation and pediatric waitlist mortality in the United States: potential for improvement. Transplantation, 103(3), 552.
  7. Tang, Y. S., Li, A.A., Scheller-Wolf, A. A., & Tayur, S. R. (2021). Multi-Armed Bandits with Endogenous Learning and Queueing: An Application to Split Liver Transplantation. Available at SSRN 3855206.
  8. Tang, Y. S., Scheller-Wolf, A. A., Tayur, S. R., Perito, E., & Roberts, J. P. (2021). Split Liver Transplantation: An Analytical Decision Support Model. Available at SSRN 3877523.
  9. Zenios, S. A., Chertow, G. M., & Wein, L. M. (2000). Dynamic allocation of kidneys to candidates on the transplant waiting list. Operations Research, 48(4), 549-569.