Journal Logo

In View: Special Article

Big Data in Transplantation Practice—the Devil Is in the Detail—Fontan-associated Liver Disease

Kim, Michelle H. MD1; Nguyen, Ailene BS1; Lo, Mary MS2; Kumar, Subramanyan Ram MD, PhD3,4; Bucuvalas, John MD5; Glynn, Earl F. MS6; Hoffman, Mark A. PhD6; Fischer, Ryan MD7; Emamaullee, Juliet MD, PhD, FRCSC1,8

Author Information
doi: 10.1097/TP.0000000000003308

Abstract

INTRODUCTION

Surgical palliation of congenital univentricular heart disease is achieved via a Fontan circuit, characterized by passive venous return to the pulmonary circulation. Advances in surgical and medical management have allowed a significant proportion of patients with single right or left ventricle physiology to survive into adulthood. Despite these improvements, long-term follow-up has revealed serious noncardiac sequelae including progressive hepatic fibrosis and cirrhosis, termed “Fontan-Associated Liver Disease” (FALD).1

Surgical management of single ventricle physiology includes a staged approach over the first 3–5 y of life, whereby the functional single ventricle is recruited to support pulsatile systemic circulation. Venous return from the inferior and superior vena cavae is diverted to the pulmonary arteries bypassing the heart, thereby constructing a total cavopulmonary connection.1 The combination of chronic nonpulsatile hepatic outflow and elevated central venous pressures leads to venous congestion and impaired hepatic blood flow.1 FALD is, hence, thought to be a form of congestive hepatopathy. Recent single-center biopsy series revealed that virtually 100% of patients post-Fontan have some degree of clinically silent hepatic fibrosis, with bridging fibrosis reported in 38% by adolescence.2-4

Liver tests, imaging, and biopsy fail to identify patients who will develop clinically significant chronic liver disease or hepatocellular carcinoma (HCC). Also unclear is when to consider combined heart-liver transplant (CHLT) versus heart transplant alone.1,5-8 Single-center series of FALD and CHLT have provided limited, but important, data showing good outcomes thus far.2,4,9,10 Prospective, multicenter studies evaluating noncardiac complications in patients post-Fontan remain necessary.11,12

Recognizing these difficulties, a multidisciplinary group of American Society of Transplantation Liver and Intestine Community of Practice members analyzed several administrative datasets to better understand the prevalence and outcome of FALD in the United States. Herein, we review the challenges encountered. Additionally, we describe strategies to facilitate future studies in patients post-Fontan. The projected mean age of Fontan patients will be 23 y by 2025, with an estimated global population of 70 000+ Fontan patients.11,13 Thus, it is imperative that we implement changes for better post-Fontan care.

MATERIALS AND METHODS

Please refer to the Supplementary Methods (SDC, http://links.lww.com/TP/B944).

RESULTS

Using Databases to Identify FALD

The ideal dataset to identify patients post-Fontan with FALD would include (1) years of longitudinal data, (2) children and adults, and (3) inpatient and outpatient encounters. We considered several potential databases, outlined in Table S1 and the Supplementary Methods, SDC, http://links.lww.com/TP/B944. Our pilot efforts targeted the Vizient database, which contains longitudinal data, has pediatric and adult patients, and captures 107 academic medical centers. Several patients with known FALD at one center (J.E.) were identified, but none of these patients were captured using ICD codes in our single institution query.

A different and promising dataset that transcends pediatric ages and adulthood is the Cerner Health Facts database. The Health Facts dataset contains deidentified patient data from nearly 70 million patients with >500 million patient encounters since 2000. Data including laboratory results, medications, and diagnostic and procedural coding are integrated in the Cerner system. We identified patients with an included ICD-9, ICD-10, or CPT-4 code for single ventricle heart disease or an associated procedure as detailed in Supplementary Methods (SDC, http://links.lww.com/TP/B944). As shown in Figure 1, 5402 unique patients were identified with these diagnoses (from 2009 to 2017) and/or procedures (from 2013 to 2017) (panel A), with overlap of codes identifying 1937 of these patients. Most patients with compatible codes were <10 y old (panel B). In parallel, we identified patients with ICD-9, ICD-10, or CPT-4 codes for liver disease (Table S2, SDC, http://links.lww.com/TP/B944). Here, the majority with diagnostic or procedure codes for liver disease were >40 y old (panel C). When the cohorts were combined, 163 patients were identified with single ventricle and liver disease diagnosis code or procedure overlap as shown in an “upset graph” (panel D).14 This suggests a FALD incidence of 3.0%, severely underestimating the burden in the single ventricle population.1,3,11 The challenges can be illustrated by using Children’s Hospital-Los Angeles, a Cerner institution, as a reference. There are >500 known Fontan patients in that system, of which >100 are followed for FALD.4

FIGURE 1.
FIGURE 1.:
Analysis of Cerner Health Facts Database to identify patients with FALD. The Cerner Health Facts Database includes over 500 million patient encounters from both pediatric and adult hospitals across the U.S. (A) illustrates the total number of patients identified with single ventricle ICD-9 or -10 codes or CPT-4 codes (note: 70 encounters with CPT-4 33615 were noted, but all were found at frequencies less than required to show in the top 15 combinations). In (B), the distribution of patient age with single ventricle ICD-9, -10, or CPT-4 codes demonstrates that the majority of patients with these diagnostic codes are under age 10. The distribution of patients with liver disease ICD-9, -10, or CPT-4 codes by patient age in (C) illustrates that the majority of patients are older adults. D, In the Cerner Health Facts Database, there are only 163 patients with both a liver disease and a single ventricle diagnostic or procedure code. Please see Methods for specific ICD codes used in search strategy. FALD, Fontan-associated liver disease.

Following this, we interrogated the Scientific Registry of Transplant Recipients (SRTR), which includes all patients listed for or have received solid organ transplants in the United States since 1987. There are no SRTR diagnostic codes specific to Fontan or FALD. A recent publication utilized the cardiac diagnosis of “congenital heart disease with surgery” to identify 27 combined-heart liver transplant recipients, which presumably included FALD patients.15 Based on these efforts, we selected patients in the heart transplant recipient database with the same code. Their unique patient identifier was crosslinked to the liver transplant waitlist and recipient database. As shown in Table 1, 10 patients who underwent heart transplant for “congenital heart disease with surgery” were identified on the liver waiting list. There were 6 adults, 1 teenager, and 3 infants <1 y old. It is unlikely that any of the infant patients were post-Fontan since this procedure is rarely performed during infancy. One of the adults received a combined heart-lung transplant procedure before listing for liver transplant. Four adults and 1 adolescent were listed for liver transplant within months of heart transplant. The young adult patient was later relisted for a diagnosis of HCC. Among the 6 adult candidates, 4 died on the waiting list. Only 1 pediatric patient ultimately received a liver transplant.

TABLE 1. - Summary of heart transplant recipients with diagnosis “Congenital Heart Disease: with surgery” who were subsequently placed on the liver transplant waiting list in the SRTR
Patient #1 #2 #3 #4 #5 #6 #7 #8 #9 #10
Age at heart transplant (y) 35a 23 49 30 7 mo 6 mo 17 54 9 mo 48
Heart transplant (y) 2007 2006 2006 2006 2009 2011 2014 2013 2017 2018
Heart transplant waiting time (d) 203 235 22 81 89 155 447 30 13 123
Liver transplant waitlist addition (y) 2008 2006 (2 mo postheart), 2009 2006 (9 mo postheart) 2008 2014 2015 2015 2013 (2 mo postheart) 2018 (2 mo postheart) 2018 (1 mo postheart)
Liver diagnosis Cryptogenic cirrhosis Cardiac liver cirrhosis/ HCC (2009) Hepatic congestion Cardiac cirrhosis Cardiac cirrhosis Budd-Chiari Cardiac cirrhosis NASH Biliary atresia Liver failure
MELD/PELD at listing 16 30, 22 21 10 PELD 1 PELD 0 6 40 PELD 5 36
Current status Death 8 mo after liver listing Death 40 mo after liver listing Death 23 d after liver listing, MELD 29 Alive 135 mo after liver listing Liver Tx 4 mo after listing Inactive 4 mo after liver listing, PELD 2 Alive 46 mo after liver listing Death 13 d after listing, MELD 40 Alive 9 mo after listing Alive 10 mo after listing
aHeart-lung transplant recipient.
HCC, hepatocellular carcinoma; MELD, model of end-stage liver disease; NASH, nonalcoholic steatohepatitis; PELD, pediatric end-stage liver disease; SRTR, Scientific Registry of Transplant Recipients.

As we struggled to identify patients post-Fontan, we considered performing a data linkage between a dataset that captures the Fontan diagnosis and 1 of the above-mentioned datasets. The Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD) was specifically explored. As of the last retrieval, the STS-CHSD has data on >500 000 cumulative congenital cardiac operations representing 127 congenital heart surgery hospitals (Table S1, SDC, http://links.lww.com/TP/B944).16 Although the STS-CHSD would be an excellent source to establish the denominator of Fontan patients, it was designed to capture discrete surgical events and does not reliably track individual patients. Beyond that, there is no precedent for linking the STS adult or STS-CHS databases with the SRTR data, although linkages to other databases such as the Pediatric Health Information Systems and Centers for Medicaid Services have successfully been performed with indirect identifiers.17-19

DISCUSSION

Recognition of Challenges and Future Directions

Our AST Working Group has outlined challenges to understanding the long-term epidemiology of FALD. We explored datasets including Vizient, Cerner Health Facts, the STS Congenital Heart Surgery Database, and the SRTR. A lack of diagnostic codes for Fontan and FALD made it difficult to identify these patients. As shown in Figure 1, examination of the most comprehensive dataset available, Cerner Health Facts, did not identify a significant FALD population. This may result not only from inadequate coding options, but also from incomplete “patient trajectories” that are broken when patients visit different health systems. Without changes in diagnostic coding and data capture, examining patients post-Fontan in existing datasets will be extremely limited.

That said, several collaborations and registries are following these patients prospectively (summarized in Table S3, SDC, http://links.lww.com/TP/B944). These include registries such as the Pediatric Heart Network Fontan Cross-Sectional Study, the Alliance for Adult Research in Congenital Cardiology Fontan Liver Health Study, the American College of Cardiology Impact Registry, and Cardiac Networks United. Outside of North America, the Australian and New Zealand Fontan Registry represents the largest and most comprehensive Fontan cohort published to date, with approximately 1500 patients enrolled as of July 2018.20 The collaborative efforts have produced data important to understanding FALD. This includes a reported rate of cirrhosis of 8% in a Pediatric Heart Network subcohort of 373 patients ~18 y post-Fontan surgery, and a high rate of hepatic fibrosis (97%) in a recent publication from Alliance for Adult Research in Congenital Cardiology and a subanalysis of the Australian registry.21,22 Unfortunately, these registries were not designed to follow liver-related complications including fibrosis, cirrhosis, HCC, and liver transplant in FALD.

The transplant community will continue to face complexities of providing evidence-based care to patients with FALD. Several interventions will be necessary to help. First, in the next iteration of the International Classification of Diseases by the World Health Organization (ICD-11), scheduled for 2022, it is imperative that codes specific to “history of Fontan” and FALD be developed.23 Next, transplant registries should create a heart-specific diagnostic code for “history of Fontan” and a liver-specific code for FALD. Programs are increasingly faced with consideration for heart transplant alone versus CHLT and will need to track these diagnoses to understand posttransplant outcomes in these patients. In our analysis of the SRTR, adult patients listed for liver transplant after undergoing heart transplant had a high waitlist mortality.

Lastly, a prospective, collaborative, national, or international level registry designed to follow secondary morbidities of Fontan, including FALD, plastic bronchitis, and protein-losing enteropathy should be established to generate meaningful answers to the questions posed in this article. A data registry used for this purpose would require the following qualities: (1) diagnostic coding to identify patients post-Fontan, (2) data from childhood into adulthood, (3) a multicenter study population, (4) granular and clinically relevant data, and (5) the ability to link to a transplant data source with indirect identifiers. This type of endeavor will require collaboration among societies including the American Society of Transplantation, American College of Cardiology, and Society of Thoracic Surgeons and have the greatest chance for success with support from the National Institutes of Health.

ACKNOWLEDGMENTS

This manuscript is a work product completed by the FALD Research Group that formed out of the American Society of Transplantation Liver and Intestinal Transplant Community of Practice.

REFERENCES

1. Gordon-Walker TT, Bove K, Veldtman G. Fontan-associated liver disease: a review. J Cardiol. 2019;74:223–232doi:10.1016/j.jjcc.2019.02.016
2. Surrey LF, Russo P, Rychik J, et al. Prevalence and characterization of fibrosis in surveillance liver biopsies of patients with Fontan circulation. Hum Pathol. 2016;57:106–115doi:10.1016/j.humpath.2016.07.006
3. Goldberg DJ, Surrey LF, Glatz AC, et al. Hepatic fibrosis is universal following Fontan operation, and severity is associated with time from surgery: a liver biopsy and hemodynamic study. J Am Heart Assoc. 2017;6:e004809doi:10.1161/JAHA.116.004809
4. Emamaullee J, Yanni G, Kohli R, et al. Impact of sex, ethnicity, and body mass index on progression of fibrosis in Fontan-associated liver disease. Hepatology. 2019;70Suppl 11869
5. Munsterman ID, Duijnhouwer AL, Kendall TJ, et al.; Nijmegen Fontan Initiative. The clinical spectrum of Fontan-associated liver disease: results from a prospective multimodality screening cohort. Eur Heart J. 2019;40:1057–1068doi:10.1093/eurheartj/ehy620
6. Rathgeber SL, Harris KC. Fontan-associated liver disease: evidence for early surveillance of liver health in pediatric Fontan patients. Can J Cardiol. 2019;35:217–220doi:10.1016/j.cjca.2018.11.019
7. Reardon LC, DePasquale EC, Tarabay J, et al. Heart and heart-liver transplantation in adults with failing Fontan physiology. Clin Transplant. 2018;32:e13329doi:10.1111/ctr.13329
8. Simpson KE, Esmaeeli A, Khanna G, et al. Liver cirrhosis in Fontan patients does not affect 1-year post-heart transplant mortality or markers of liver function. J Heart Lung Transplant. 2014;33:170–177doi:10.1016/j.healun.2013.10.033
9. Wong TW, Gandhi MJ, Daly RC, et al. Liver allograft provides immunoprotection for the cardiac allograft in combined heart-liver transplantation. Am J Transplant. 2016;16:3522–3531doi:10.1111/ajt.13870
10. Hilscher MB, Johnson JN, Cetta F, et al. Surveillance for liver complications after the Fontan procedure. Congenit Heart Dis. 2017;12:124–132doi:10.1111/chd.12446
11. Schilling C, Dalziel K, Nunn R, et al. The Fontan epidemic: population projections from the Australia and New Zealand Fontan Registry. Int J Cardiol. 2016;219:14–19doi:10.1016/j.ijcard.2016.05.035
12. Rychik J, Atz AM, Celermajer DS, et al. Evaluation and management of the child and adult with Fontan circulation: a scientific statement from the American Heart Association. Circulation. 2019;140:e234–e284doi:10.1161/CIR.0000000000000696
13. Jacobs JP, Mayer JE Jr, Mavroudis C, et al. The Society of Thoracic Surgeons Congenital Heart Surgery Database: 2017 update on outcomes and quality. Ann Thorac Surg. 2017;103:699–709doi:10.1016/j.athoracsur.2017.01.004
14. Lex A, Gehlenborg N, Strobelt H, et al. UpSet: visualization of intersecting sets. IEEE Trans Vis Comput Graph. 2014;20:1983–1992doi:10.1109/TVCG.2014.2346248
15. Bradley EA, Pinyoluksana KO, Moore-Clingenpeel M, et al. Isolated heart transplant and combined heart-liver transplant in adult congenital heart disease patients: insights from the united network of organ sharing. Int J Cardiol. 2017;228:790–795doi:10.1016/j.ijcard.2016.11.121
16. Jacobs JP, Mayer JE Jr, Pasquali SK, et al. The Society of Thoracic Surgeons Congenital Heart Surgery Database: 2019 update on outcomes and quality. Ann Thorac Surg. 2019;107:691–704doi:10.1016/j.athoracsur.2018.12.016
17. Jacobs JP, Edwards FH, Shahian DM, et al. Successful linking of the Society of Thoracic Surgeons adult cardiac surgery database to Centers for Medicare and Medicaid Services Medicare data. Ann Thorac Surg. 2010;90:1150–6discussion 1156. doi:10.1016/j.athoracsur.2010.05.042
18. Pasquali SK, Jacobs JP, Shook GJ, et al. Linking clinical registry data with administrative data using indirect identifiers: implementation and validation in the congenital heart surgery population. Am Heart J. 2010;160:1099–1104doi:10.1016/j.ahj.2010.08.010
19. Prasad A, Helder MR, Brown DA, et al. Understanding differences in administrative and audited patient data in cardiac surgery: comparison of the University HealthSystem Consortium and Society of Thoracic Surgeons Databases. J Am Coll Surg. 2016;223:551–557.e4doi:10.1016/j.jamcollsurg.2016.06.393
20. Iyengar AJ, Winlaw DS, Galati JC, et al. The Australia and New Zealand Fontan Registry: description and initial results from the first population-based Fontan registry. Intern Med J. 2014;44:148–155doi:10.1111/imj.12318
21. Atz AM, Zak V, Mahony L, et al.; Pediatric Heart Network Investigators. Longitudinal outcomes of patients with single ventricle after the Fontan procedure. J Am Coll Cardiol. 2017;69:2735–2744doi:10.1016/j.jacc.2017.03.582
22. Wilson TG, d’Udekem Y, Winlaw DS, et al.; Australian and New Zealand Fontan Registry. Hepatic and renal end-organ damage in the Fontan circulation: a report from the Australian and New Zealand Fontan Registry. Int J Cardiol. 2018;273:100–107doi:10.1016/j.ijcard.2018.07.118
23. World Health Organization Release of ICD-11.

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.