Date Published: January 25, 2017
Publisher: Public Library of Science
Author(s): Kristopher P. Croome, David D. Lee, Denise Harnois, C. Burcin Taner, Jose Ignacio Herrero.
Several studies have investigated the effects following the implementation of the “Share 35” policy; however none have investigated what effect this policy change has had on waitlist and liver transplantation (LT) outcomes for hepatocellular carcinoma(HCC).
Data were obtained from the UNOS database and a comparison of the 2 years post-Share 35 with data from the 2 years pre-Share 35 was performed.
In the pre-Share35 era, 23% of LT were performed for HCC exceptions compared to 22% of LT in the post-Share35 era (p = 0.21). No difference in wait-time for HCC patients was seen in any of the UNOS regions between the 2 eras. Competing risk analysis demonstrated that HCC candidates in post-Share 35 era were more likely to die or be delisted for “too sick” while waiting (7.2% vs. 5.3%; p = 0.005) within 15 months. A higher proportion of ECD (p<0.001) and DCD (p<0.001) livers were used for patients transplanted for HCC, while lower DRI organs were used for those patients transplanted with a MELD≥35 between the 2 eras (p = 0.007). No significant change to wait-time for patients listed for HCC was seen following implementation of “Share 35”. Transplant program behavior has changed resulting use of higher proportion of ECD and DCD liver grafts for patients with HCC. A higher rate of wait list mortality was observed in patients with HCC in the post-Share 35 era.
The Model for End Stage Liver Disease (MELD) has been used as the method of liver graft allocation since 2002. Liver grafts are distributed geographically based on MELD score in Organ Procurement and Transplant Network (OPTN) sharing areas consisting of local, regional and national tiers. It has been shown that patients with a MELD score ≥ 35 have wait list mortality rates that were similar to Status 1 candidates, who represent a cohort of candidates with acute liver failure most likely to die within 7 days without liver transplant . To facilitate liver graft allocation to those patients with high wait list mortality, regional sharing for patients with a MELD ≥ 35 was implemented in June of 2013 with the goal of increasing life saving liver transplant for the sickest patients and decreasing death on the waiting list (“Share 35”).
After approval from the Mayo Clinic Institutional Review Board, data were obtained and extracted from the United Network of Organ Sharing (UNOS) Standard Analysis and Research file. The study population included all patients on the waitlist for LT in the United States from June 18, 2011 to June 18, 2015. Prior to Share 35 in the United States liver allografts were allocated to patients with the highest MELD score in sequential sharing areas consist of first local, then regional, and finally national tiers. The Share 35 policy was implemented in June 2013 to achieve broader sharing whereby the sickest waitlist candidates (patients with a MELD score ≥ 35) are first prioritized in a tiered manner regionally before any local candidates with MELD scores less than 35 are offered the livers. For the majority of analyses, data were provided for 2 eras; the 2 years pre-implementation of “Share 35” (Era 1) (6/18/2011 to 6/17/2013) and the 2 years post-implementation (Era 2) (6/18/2013 to 6/18/2015). Share 35 was implemented in June 2013 and therefore the dates were chosen so that we had 2 equal time periods with a minimum of 1 year of follow-up.
In Era 1 a total of 12,636 LT were performed of which 2916 (23.0%) were performed for HCC compared to 13533 LT of which 3029 (22.4%) were for HCC in Era 2. There was no difference in the proportion of LTs performed for HCC in the 2 eras (p = 0.18). No difference in the median match MELD score was seen for HCC patients between Era 1 and Era 2 (25 [range 6–40] vs. 25 [range 6–40]; p = 0.12). For HCC patients the median wait-time in era 1 was 185 days compared to 195 days in era 2. No significant difference in wait list time for HCC patients was seen between Era 1 and Era 2 in any of the eleven UNOS regions. Similarly, no difference in wait list time was seen for non-HCC patients in any of the eleven UNOS regions except for Region 1 where wait time increased (206 vs. 277 days; p = 0.02). When a sub-analysis was performed for all patients with a MELD ≥ 35, no difference in wait list time was seen for any of the eleven Regions. There was a slight statistical increase in the proportion of patients newly listed with HCC exception MELD between the 2 eras (14.3% vs.15.0%; p = 0.04); however this statistical difference is of minimal clinical relevance.
Broader regional sharing though the “Share 35” policy change was implemented with the goal of increasing life saving LT for the sickest patients on the wait list. Initial publications examining the effects of “Share 35” have shown several positive results, including reduction in 90-day mortality for patients with MELD scores ≥ 35 [2,12]. While these initial results are encouraging, it is important to fully explore the effects of “Share 35” for all patient groups.