Author + information
- Received April 6, 2016
- Revision received July 26, 2016
- Accepted July 28, 2016
- Published online December 1, 2016.
- Vincent Algalarrondo, MD, PhDa,b,∗ (, )
- Teresa Antonini, MDb,c,
- Marie Théaudin, MD, PhDb,d,
- Denis Chemla, MD, PhDe,
- Anouar Benmalek, PhDf,
- Catherine Lacroix, MDd,
- Denis Castaing, MDb,c,
- Cécile Cauquil, MDb,d,
- Sylvie Dinanian, MDa,b,
- Ludivine Eliahou, MDa,b,
- Didier Samuel, MD, PhDb,c,
- David Adams, MD, PhDb,d,
- Dominique Le Guludec, MD, PhDg,
- Michel S. Slama, MDa,b and
- François Rouzet, MD, PhDg
- aCardiology Department, Antoine Béclère Hospital, Assistance Publique Hôpitaux de Paris (AP-HP), UMR-S 1180, University of Paris-Sud, Clamart, France
- bFrench Referral Center for Familial Amyloidotic Polyneuropathy and Other Rare Peripheral Neuropathies, Bicêtre, France
- cHepato-Biliary Center, Paul Brousse Hospital, AP-HP, UMR-S 785, University of Paris-Sud, Villejuif, France
- dFilière Neuromusculaire, Neurology Department, Kremlin Bicêtre Hospital, AP-HP, Bicêtre, France
- ePhysiology Department, EA4533, University of Paris-Sud, Le Kremlin Bicêtre, France
- fSchool of Pharmacy, University of Paris-Sud, Chatenay Malabry, France
- gNuclear Medicine Department and Département Hospitalo Universitaire Fibrose Inflammation et Remodelage en pathologies cardiovasculaires, Bichat Claude Bernard Hospital, AP-HP, University of Paris VII, UMR-S 1148, Paris, France
- ↵∗Reprint requests and correspondence:
Dr. Vincent Algalarrondo, Cardiology Department, Antoine Béclère Hospital, 157 Avenue de la Porte de Trivaux, Clamart, France 92140.
Objectives This study sought to compare techniques evaluating cardiac dysautonomia and predicting the risk of death of patients with hereditary transthyretin amyloidosis (mATTR) after liver transplantation (LT).
Background mATTR is a multisystemic disease involving mainly the heart and the peripheral nervous system. LT is the reference treatment, and pre-operative detection of high-risk patients is critical. Cardiovascular dysautonomia is commonly encountered in ATTR and may affect patient outcome, although it is not known yet which technique should be used in the field to evaluate it.
Methods In a series of 215 consecutive mATTR patients who underwent LT, cardiac dysautonomia was assessed by a dedicated clinical score, time-domain heart rate variability, 123-meta-iodobenzylguanidine heart/mediastinum (123-MIBG H/M) ratio on scintigraphy, and heart rate response to atropine (HRRA).
Results Patient median age was 43 years, 62% were male and 69% carried the Val30Met mutation. Cardiac dysautonomia was documented by at least 1 technique for all patients but 6 (97%). In univariate analysis, clinical score, 123-MIBG H/M ratio and HRRA were associated with mortality but not heart rate variability. The 123-MIBG H/M ratio and HRRA had greater area under the curve (AUC) of receiver-operating characteristic curves than clinical score and heart rate variability (AUC: 0.787, 0.748, 0.656, and 0.523, respectively). Multivariate score models were then built using the following variables: New York Heart Association functional class, interventricular septum thickness, and either 123-MIBG H/M ratio (SMIBG) or HRRA (Satropine). AUC of SMIBG and Satropine were greater than AUC of univariate models, although nonsignificantly (AUC: 0.798 and 0.799, respectively). Predictive powers of SMIBG, Satropine, and a reference clinical model (AUC: 0.785) were similar.
Conclusions Evaluation of cardiac dysautonomia is a valuable addition for predicting survival of mATTR patients following LT. Among the different techniques that evaluate cardiac dysautonomia, 123-MIBG scintigraphy and heart rate response to atropine had better prognostic accuracy. Multivariate models did not improve significantly prediction of outcome.
Amyloidosis due to hereditary transthyretin amyloidosis (mATTR) is a rare life-threatening, autosomal dominant disease caused by mutation of the transthyretin (TTR), a thyroxine transport protein. The mutation is associated with an unstable tetrameric structure of TTR leading to misfolding, promoting the deposition of amyloid fibrils predominantly in the peripheral nerves, heart, eyes, gastrointestinal tract, and kidneys. In endemic regions, Val30Met, Val122Ile, and Thr60Ala are the most prevalent TTR variants, although >100 variants have been reported worldwide. Depending on the mutation, the clinical manifestations of mATTR are as follow: 1) predominantly a polyneuropathy, known as familial amyloid polyneuropathy; 2) predominantly a cardiomyopathy with minimal or no neuropathy, known as familial amyloid cardiomyopathy; or 3) “mixed” forms of the disease, with cardiac and neurological manifestations (1,2). Cardiac involvement includes heart failure due to infiltrative and restrictive cardiomyopathy, cardiac arrhythmias, and cardiac denervation (3). This denervation affects both the sympathetic and the parasympathetic systems. The sympathetic system effect involves a decrease in pre-synaptic catecholamine stores with preserved cardiac β-receptor catecholamine responsiveness, which leads to an early and striking decrease in meta-iodobenzylguanidine (MIBG) uptake. On the other hand, parasympathetic denervation is associated with up-regulation of muscarinic acetylcholine receptors; however, because acetylcholine availability and binding is absent, parasympathetic tone is low and atropine has minimal or no effect on heart rate (4–6). By removing the main source of the circulating mutated TTR, orthotopic liver transplantation (LT) improves the course of the neuropathy and survival (7,8). More recently, pharmaceuticals have been developed as an alternative to LT (9–12). Nevertheless, LT remains the mainstay of familial amyloidotic polyneuropathy therapy and remains the historical reference for the evaluation of new treatments (13,14).
Several factors influence the prognosis after LT (7,15–20), including cardiac dysautonomia, and our group recently proposed a multivariate model to evaluate the risk of death in LT patients (21). This score is based on readily available clinical cardiac and neurological variables that included orthostatic hypotension, a typical symptom of dysautonomia. Numerous techniques exist to evaluate cardiovascular dysautonomia, including 123-MIBG scintigraphy, heart rate variability, and heart rate response to atropine, and previous reports indicate that evaluation of cardiac denervation using these techniques could be of interest to assess prognosis in mATTR patients (18,22–25). To the best of our knowledge, no study thus far has evaluated and compared the prognostic accuracy of these techniques to predict overall mortality in mATTR patients after LT.
In this study, we aimed to evaluate and compare the ability to predict the overall mortality in mATTR patients after LT of 4 different techniques reflecting cardiac dysautonomia: clinical evaluation; time-domain heart rate variability; 123-MIBG cardiac scintigraphy; and heart rate response after atropine injection.
From the database of the French National Reference Center for Familial Amyloidotic Polyneuropathy, we identified 218 consecutive patients who underwent LT for mATTR between January 1, 1993 and January 1, 2011. mATTR was diagnosed by the observation of both amyloid deposits in biopsy specimens and a TTR mutation. After the exclusion of 3 patients who underwent combined heart and liver transplantation, 215 patients entered the analysis, of which 1 was censored at 65 months after LT when undergoing cardiac transplantation. Patient characteristics, clinical evolution, and outcomes, including the causes of death, were analyzed. This study complied with the ethical principles formulated in the Declaration of Helsinki and was approved by the ethics committee of the Hôpital de Bicêtre, Bicêtre, France. All patients who survived to the day of study initiation gave written, informed consent to participate in the registry (Commission Nationale de l'Informatique et des Libertés no. 1470960).
The cardiac evaluation included a physical examination and assessment of the New York Heart Association (NYHA) functional class, a surface electrocardiogram (ECG), and an echocardiogram. Pulmonary artery pressures, capillary wedge pressure, and cardiac index were measured during right heart catheterization in 200 patients. Arrhythmias and intracardiac conduction disorders were evaluated with 12-lead ECG, 24-h Holter ECG recordings, and electrophysiological studies as previously reported (26).
The neurological evaluation included a neuropathy impairment score and the polyneuropathy disability score (PND) (27,28). Nerve conduction studies of the extremities were performed by measuring the sensory action potential amplitude of the ulnar and sural nerves and the compound muscle action potential amplitude of the ulnar and peroneal nerves.
The autonomic cardiovascular nervous system was evaluated with 4 different techniques. A dedicated and validated clinical score (Compound Autonomic Dysfunction Test) was performed in all patients (29). The score integrated evaluation of postural hypotension, nausea/vomiting, diarrhea/constipation, and sphincter disturbances (Online Table 1). Each component of the score ranged from 0 (severe dysautonomia) to 4 (normal autonomic nervous system), and the Compound Autonomic Dysfunction Test was considered abnormal if the score was below 16. On the basis of 24-h Holter ECG, the mean heart rate and the standard deviation of the normal cycles (SDNN) were measured in 180 patients. The normal value of SDNN in our laboratory is 153 ± 34 ms (4). The sympathetic cardiac innervation was further examined with scintigraphic measurement of the heart-to-mediastinum uptake ratio (H/M ratio) in the anterior view of the chest on a gamma-camera (Infinia, GEMS, Buc, France) equipped with low-energy, high-resolution collimators 4 h after injection of 3 MBq/Kg of 123-MIBG in 133 patients (5,22). In our laboratory, the value of the H/M ratio for healthy subjects is 1.98 ± 0.35 (5). Finally, cardiac parasympathetic tone was assessed in 151 patients according to the modified protocol of Sutton et al. (4,30): an atropine bolus of 0.04 mg/kg intravenously (maximum: 2 mg) was injected after a 15-min rest period and heart rate was continuously monitored for the next 15 min. Heart rate responses to atropine (HRRA) represented the difference between maximal and basal heart rates. In normal subjects, atropine infusion causes an increase in heart rate from 68 ± 10 beats/min to 109 ± 14 beats/min corresponding to a mean increase of 42 ± 12 beats/min or 62 ± 17% (4).
Liver transplantation and follow-up
All patients were ambulatory and free from standard contraindications to LT (31). The liver grafts were harvested from 200 cadavers (93%) and from 15 living, related donors (7%). After surgery, physical examinations were performed daily for 2 weeks after LT, 3× weekly over the following week, then once a week for 1 month, once a month for 3 months, and every 3 months thereafter. The primary study endpoint was all-cause mortality.
Continuous variables are presented as medians (interquartile ranges [IQR]) and categorical variables are presented as counts and percentages. Correlations were calculated using the Spearman correlation test. Cox single variable regression models were used to identify the pre-operative variables associated with the primary endpoint. Receiver-operating characteristic (ROC) curves were constructed using the overall death as the endpoint and the areas under the curve were calculated. ROC curves were compared using contrast tests on ROC and according to DeLong et al. (32). Cutpoints were identified as the values corresponding with the highest average of sensitivity and specificity. Using the 2 best predictor variables in univariate analysis, multiple variable Cox models and risk scores were constructed as previously reported (21). Because 1 aim of the study was to compare the different methods evaluating autonomic innervation, we intentionally excluded the multiple imputation method for compensation of missing data (Online Figure 1). No selection bias occurred because patients with and without missing data had similar clinical profiles (Online Tables 2 to 4). A p value <0.05 was considered statistically significant. MedCalc (MedCalc Software, Ostend, Belgium) and R (R Foundation, Vienna, Austria) were used for the analyses.
A total of 215 consecutive mATTR patients entered the analysis. Median age was 43 years (IQR: 35 to 58 years) and 62% were men. In our cohort, 148 patients (69%) carried the Val30Met mutation, of whom 120 (56%) had the early onset and 28 (13%) the late onset form. We observed 14 different mutations. The median delay between the first reported symptom and LT was 3 years (IQR: 2 to 5 years), ranging between 0 and 15 years. More detailed characteristics are shown in Online Table 5.
The median compound autonomic dysfunction score was 13 (IQR: 11 to 15; range 5 to 16) and was abnormal (score <16) in 82% patients (Figures 1A and 1B). Orthostatic hypotension (score <4) was diagnosed in 40% of patients, nausea (score <4) was documented in 43% of patients, diarrhea (score <4) was documented in 75% of patients, and sphincter disturbances (score <4) were documented in 37% of patients. On 24-h Holter ECG, the median SDNN was 83 ms (IQR: 65 to 106 ms; range 29 to 213 ms) (Figure 1C). The SDNN was below 120 ms in 88% of patients and below 100 ms in 71% of patients. The median HRRA injection was 7 beats/min (IQR: 0 to 17 beats/min; range 0 to 77 beats/min). In 38 patients (25%), the atropine injection was not followed by any significant heart rate response (Figure 1D). On 123-MIBG cardiac scintigraphy, the median H/M ratio at 4 h was 1.49 (IQR: 1.24 to 1.74; range 0.97 to 2.52). The H/M ratio was below 1.6 in 82 patients (62%) (Figure 1E). Of note, cardiac dysautonomia could be observed in patients with no other signs or symptoms of cardiac amyloidosis. Indeed, among patients with 123-MIBG H/M ratio <1.6, 89% were NYHA functional class I, 62% had an interventricular septum thickness ≤12 mm, 92% had left ventricular ejection fraction >55%, and 89% had a normal right heart catheterization. Taken together, all but 6 patients (97%) had at least 1 abnormal test. Among variables that evaluate the autonomic nervous system, significant correlations were observed (Table 1), especially between HRRA and 123I-MIBG H/M ratio (Figure 1F).
Survival: Univariate analysis
Over a median follow-up of 5.9 years (IQR: 2.6 to 9.4 years) after LT, 84 patients died. Among variables assessing the autonomic nervous system, the following were significantly associated with survival: the clinical compound test (especially orthostatic hypotension and urinary incontinence), the HRRA, and the 123I-MIBG H/M ratio (Table 2). Neither SDNN nor mean heart rate were significantly associated with survival. The univariate analyses of the other variables are shown in Online Table 6.
ROC curves and best cutoff values were computed for each of the 4 variables and were compared (Figure 2). For the clinical compound test, the area under the curve (AUC) was 0.656 (95% confidence interval [CI]: 0.589 to 0.720; cutoff value ≤12 points). For SDNN, the AUC was 0.523 (95% CI: 0.448 to 0.598; cutoff value ≤79 ms). For mean heart rate, the AUC was 0.588 (95% CI: 0.514 to 0.659; cutoff value ≤76.3 beats/min). For HRRA, the AUC was 0.748 (95% CI: 0.671 to 0.815; cutoff value ≤11 beats/min). Patients with HRAA ≤11 beats/min had a 5-year survival post-LT of 66% versus 98% for patients with HRAA >11 beats/min (log rank test, p <0.0001) (Figure 3A). For the 123-MIBG H/M ratio, the AUC was 0.787 (95% CI: 0.707 to 0.853; cutoff value ≤1.43). Patients with 123-MIBG H/M ratio ≤1.43 had a 5-year survival post-LT of 64% versus 93% for patients with 123-MIBG H/M ratio >1.43 (log rank test, p < 0.0001) (Figure 3B). Comparisons of the ROC curves showed that HRRA and 123-MIBG H/M ratio both had significantly better predictive values than clinical compound test, mean heart rate, and SDNN (Table 3). Therefore, HRRA and 123-MIBG H/M ratio were integrated into multivariate models and their additional values to predict overall mortality were compared with the reference clinical score, whereas heart rate variability and clinical compound score were not further investigated.
Survival: Multivariate analysis
The reference model (Sclinical) used to predict overall mortality was previously described and includes PND ≥III, orthostatic hypotension, NYHA functional class >I, QRS interval ≥120 ms, and interventricular septum thickness (21). This model had an AUC of the ROC of 0.785 (95% CI: 0.722 to 0.840). Similarly, 2 models were generated using successively HRRA (Satropine) and 123-MIBG (SMIBG) (detailed calculation of the models are provided in the Online Appendix). For these 2 models, the variables PND ≥III and QRS interval ≥120 ms were not significant and thus were removed from the analysis (Table 4). The Satropine model had an AUC of the ROC of 0.799 (95% CI: 0.718 to 0.863) and the SMIBG model had an AUC of the ROC of 0.798 (95% CI: 0.722 to 0.840). These 2 new models were then compared with the reference model (Sclinical) (Figure 2B). Comparison of the 3 models Sclinical, SMIBG, and Satropine did not show significant differences (contrast tests on ROC; p = 0.44). Compared with the univariate models using HRRA and 123-MIBG H/M ratio, the multivariate models Sclinical, SMIBG, and Satropine improved risk prediction, although not significantly (Table 5).
In this large retrospective study, the assessment of the autonomic nervous system was a valuable addition for predicting survival of mATTR patients after liver transplantation. Among the different tools evaluating the autonomic nervous system, 123-MIBG scintigraphy and the HRRA injection provided better prediction of survival than the clinical compound test and the heart rate variability. As compared with univariate models, multivariate models integrating evaluation of the autonomic nervous system did not improve significantly the outcome prediction.
Cardiac denervation and prognosis in mATTR
Clinical exam and orthostatic hypotension
The compound autonomic score was specifically built to evaluate dysautonomia in mATTR patients and has been shown to have good inter-rater reliability (29). Here, we demonstrated that its predictive power following LT was average; items that were not significantly associated with survival were included into the score, blunting its overall predictive power. Conversely, the items “orthostatic hypotension” and “urinary incontinence” were significantly associated with mortality, which confirmed previous reports (7,20,21). Moreover, the item “orthostatic hypotension” was included into the previously published reference model (21). In the current study, this reference model had similar predictive power as compared with the MIBG and atropine models. Therefore, the rough evaluation of orthostatic hypotension seemed to be a key element of prognosis in mATTR. Other techniques such as ambulatory blood pressure monitoring (33) or baroreflex sensitivity (34,35) could be useful to further refine this aspect of the patient’s evaluation.
123-MIBG is an analogue of norepinephrine that competes with norepinephrine in the pre-synaptic granules of the sympathetic nerves. Cardiac scintigraphy of radiolabeled 123-MIBG shows a common decreased uptake in mATTR patients, even among those who have no heart failure and normal ejection fraction (5,22). Abnormalities detected by 123-MIBG scintigraphy in mATTR precede the development of clinical manifestations or cardiac thickening on echocardiography (6). Delahaye et al. (4,5,23) reported that mATTR patients with cardiac denervation before LT had worse neurologic evolution following LT. A subsequent study by Takahashi et al. (36) showed that antiamyloid therapy by diflunisal could improve autonomic dysfunction, including the delayed H/M ratio of the 123-MIBG scintigraphy. The prognostic value of 123-MIBG scintigraphy in mATTR patients was studied by Coutinho et al. (18) in a series of 143 patients that included 53 patients with previous LT, although the timing between LT and 123-MIBG scintigraphy was not standardized. In this study, the AUC of the ROC curve was 0.72 and the proposed cutoff value of an H/M ratio <1.60 was based on previous reports that analyzed patients with heart failure. In the present study, the cutoff value of 1.43 for the H/M ratio was tailored for mATTR patients, which may be helpful to identify patients with worse prognosis and to guide therapeutic management. Further studies may refine the quantification of cardiac MIBG uptake by identifying segmental denervation and mismatch between perfusion and innervation in tomographic acquisitions.
Cardiac dysautonomia is only 1 facet of cardiac involvement in mATTR, and thus whether or not MIBG was solely identifying patients with more severe cardiac disease must be discussed. We felt that this hypothesis is unlikely for the following reasons: first, cardiac dysautonomia could be observed in patients with no other signs or symptoms of cardiac amyloidosis. Indeed, among patients with 123-MIBG H/M ratio <1.6, 89% were NYHA functional class I, 62% had an interventricular septum thickness ≤12 mm, 92% had left ventricular ejection fraction >55%, and 89% had a normal right heart catheterization. Furthermore, among patients with 123-MIBG H/M ratio <1.43, 48% met these 4 combined criteria. Second, the multivariate analysis demonstrated that variables evaluating cardiac dysautonomia (123-MIBG H/M ratio, HRRA, orthostatic hypotension) were associated with overall mortality independently from the other covariates that sized the severity of cardiac involvement (NYHA functional class, interventricular septum thickness). Third, none of variables measured during right heart catheterization or echocardiography had stronger predictive performances than 123-MIBG scintigraphy or atropine test (Online Table 7). Finally, pre-operative cardiac screening was performed for all patients before LT and those who had the severe forms of any cardiac diseases (including severe restrictive infiltrative cardiomyopathy) were not transplanted and thus were excluded from the study. Overall, this suggests that cardiac dysautonomia may be an upstream endpoint that could enhance risk evaluation in mATTR, beyond the classical evaluation of the cardiac disease.
Heart rate response to atropine
Cardiac denervation in mATTR affects both sympathetic and parasympathetic pre-synaptic innervation (4). Parasympathetic denervation could be easily detected and assessed by the HRRA injection, which antagonizes in a reversible and competitive manner the post-synaptic muscarinic receptors. In the present study, HRRA was a powerful predictive tool, with results that mirrored those obtained with 123-MIBG scintigraphy. Of note, 25% of patients did not have any increase in heart rate following atropine injection and this emphasizes that this drug is often ineffective in the management of mATTR patients with bradycardia. Even if these results are promising, HRRA has mainly been reported by our group, and further studies will be necessary to externally validate its clinical value.
Heart rate variability
The value of heart rate variability analysis in the management of mATTR patients is still debated. LT seems to have no impact on heart rate variability (23–25) and the prognostic value of heart rate variability after LT has not been reported to the best of our knowledge. In our study, 88% of the patients had SDNN below 120 ms before LT; however, SDNN did not correlate significantly with prognosis.
Each technique measuring the autonomic nervous system has its own intrinsic limitations. Heart rate variability cannot be assessed in patients with pacemakers or if the Holter ECG trace has a poor quality. Atropine injection is contraindicated in patients with glaucoma or prostatic hypertrophy. Some treatments interfere with the uptake of 123-MIBG including antidepressants or painkillers. Here, the study was retrospective, and variables were not fully available for all patients, which may have biased the results; however, patients with and without available variables had similar profiles (Online Appendix, Online Tables 2 to 4). The predictive models have not been validated externally, and thus further prospective studies will be required. Finally, the autonomic nervous system could have been evaluated by additional techniques, such as measuring baroreflex sensitivity, heart rate profile during exercise testing, or blood pressure variability. The application of these techniques in further studies may improve the risk stratification of patients with mATTR.
Cardiovascular dysautonomia is a key predictive element in mATTR following LT. Among the clinical variables reflecting dysautonomia, orthostatic hypotension had the best predictive value. In univariate models, cardiac sympathetic and parasympathetic denervation (respectively measured by 123-MIBG scintigraphy: H/M ratio ≤1.43 and HRRA ≤11 beats/min) had a good predictive power and performed better than models based on heart rate variability or clinical parameters. As compared with univariate models, multivariate models did not improve outcome prediction.
COMPETENCY IN MEDICAL KNOWLEDGE: In mATTR, cardiac dysautonomia is strongly associated with a worse prognosis after LT.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Among techniques that evaluate cardiac dysautonomia, 123I-MIBG scintigraphy and HRRA have the best predictive accuracy to identify patients with higher risk of death after LT.
TRANSLATIONAL OUTLOOK: In transthyretin amyloidosis, few tools have been validated so far to identify high-risk patients under antiamyloid therapy. Current guidelines recommend evaluating cardiac infiltration and identifying patients with clinical heart failure, which could be unsatisfactory in the early stage of the disease. Evaluation of cardiac dysautonomia by 123I-MIBG scintigraphy or atropine test could improve risk stratification in these patients. Whether cardiac dysautonomia could be proposed as an early specific therapeutic endpoint remains to be demonstrated.
The authors gratefully acknowledge Eric Duong for his careful final proofreading.
For supplemental materials, figure, and tables, please see the online version of this article.
Dr. Algalarrondo has received scholarship funding from Medtronic, Biotronik, Boston Scientific, St. Jude Medical, and Sorin. Dr. Théaudin has received a travel grant from Pfizer; and speaking honoraria from Pfizer and LFB Biomedicaments. Dr. Samuel has received consulting fees from Astellas, Bristol-Myers Squibb, Gilead, LFB, Merck Sharp & Dohme, Novartis, Roche, and Biotest. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- area under curve
- confidence interval
- heart-mediastinum ratio
- heart rate responses to atropine
- interquartile range
- liver transplantation
- hereditary transthyretin amyloidosis
- New York Heart Association
- polyneuropathy disability score
- receiver-operating characteristic
- standard deviation of the normal sinus cycles
- Received April 6, 2016.
- Revision received July 26, 2016.
- Accepted July 28, 2016.
- American College of Cardiology Foundation
- Rapezzi C.,
- Arbustini E.,
- Caforio A.L.,
- et al.
- Rapezzi C.,
- Quarta C.C.,
- Obici L.,
- et al.
- Gertz M.A.,
- Benson M.D.,
- Dyck P.J.,
- et al.
- Delahaye N.,
- Le Guludec D.,
- Dinanian S.,
- et al.
- Adams D.,
- Samuel D.,
- Goulon-Goeau C.,
- et al.
- Adams D.,
- Cauquil C.,
- Theaudin M.,
- Rousseau A.,
- Algalarrondo V.,
- Slama M.S.
- Coutinho M.C.,
- Cortez-Dias N.,
- Cantinho G.,
- et al.
- Herlenius G.,
- Wilczek H.E.,
- Larsson M.,
- et al.,
- for the FAPWTR Investigators
- Algalarrondo V.,
- Antonini T.,
- Théaudin M.,
- et al.
- Tanaka M.,
- Hongo M.,
- Kinoshita O.,
- et al.
- Antonini T.M.,
- Lozeron P.,
- Lacroix C.,
- et al.
- ↵(1994) Consensus statement on indications for liver transplantation: Paris, June 22–23, 1993. Hepatology 20:63S–68S.
- Bernardi L.,
- Passino C.,
- Porta C.,
- Anesi E.,
- Palladini G.,
- Merlini G.