Since 1865, when Trousseau first described hemochromatosis and demonstrated its genetic nature, many studies have shed light on its putative pathophysiological mechanism. The most important breakthrough was the discovery of the gene that encoded human hemochromatosis protein (HFE). This made it possible to diagnose the most common form of hereditary hemochromatosis (HH), the p.Cys282Tyr mutation (C282Y) in the HFE gene. In addition, these discoveries enabled discrimination of HFE-related HH from various secondary causes of iron overload and other rare genetic entities. The underlying mechanism of HH was initially thought to be related to the role of HFE in regulating iron absorption at the intestinal crypt, which was supported by the observation that excessive iron absorption occurred in HH. However, this hypothesis has not been confirmed thereafter.
Another major breakthrough in understanding iron metabolism in HH was the discovery of hepcidin and its role in iron metabolism.[6, 7] Mainly secreted by the liver, this small peptide was shown to interact with ferroportin, the only known cellular iron exporter, and this interaction induced ferroportin internalization and degradation. Through its regulation of ferroportin, hepcidin can reduce iron export from macrophages and enterocytes into the bloodstream. Consequently, the liver represents the major regulator of iron metabolism. In patients with HH, serum hepcidin levels were shown to be inappropriately low.
The molecular impact of the HFE C282Y mutation has not been fully elucidated. It is thought to inhibit the ability of HFE to interact with β2 microglobulin and transferrin receptor 1 (TFR1). The interactions between HFE and TFR1 and TFR2[11, 12] play important roles in iron-sensing at the hepatocyte membrane, which regulates hepcidin expression. Mice with Hfe gene knockouts display a hemochromatosis phenotype with inappropriately low hepcidin expression.[13, 14] Mice with a liver-specific Hfe knockout presented the same phenotype. Conversely, hepatic-specific expression of wild-type HFE in HFE knockout mice increased hepcidin expression and lowered liver iron levels. This finding suggests that the liver is the central regulatory site for iron metabolism regulation. In a recent study, Garuti et al. reported that, when a liver with wild-type HFE was transplanted into an Hfe knockout mouse, the hemochromatosis iron loading phenotype was reversed and hepcidin expression was normalized at the transcriptional level. From a clinical viewpoint, human case studies have not provided clear results. In some studies, a normal liver was transplanted into a patient with HH, and in other studies an HH liver was transplanted into normal patients.[18-22] The results did not show the expected evolution of iron burden due to these LTs. Moreover, very few cases of HH have been reported, and among those assessments of the iron burden were often incomplete, and the evolution of hepcidin levels was not documented.
In the setting of LT, poor survival rates have been generally associated with iron overload, particularly in patients with HH, compared to patients with liver diseases of alcoholic or viral origin. It was suggested that hemochromatosis increased the propensity for infections and associated cardiovascular diseases.[23-27] However, those studies were limited by various biases; for example, earlier studies were performed before the availability of HFE[23, 24] testing, and thus, associations with hemochromatosis were likely to reflect various conditions that caused iron overload. More recent studies have identified patients with confirmed HFE mutations, with a family history of HH, or with an HLA A3B7 haplotype; in those studies, either the diagnosis was retrospective and included historical cases with a highly severe phenotype,[26, 27] or the control populations were not representative of other liver diseases. Consequently, it remains a matter of debate whether survival after LT is affected by HFE C282Y homozygosity and iron overload.
Therefore, the present study aimed to (1) evaluate the role of the liver in regulating iron metabolism, based on the long-term evolution of iron metabolism after LT in patients homozygous for HFE C282Y; and (2) analyze survival after transplantation in patients homozygous for HFE C282Y compared to patients who received transplants for other liver diseases, but underwent similar surgical procedures and medical management over the same time period.
This study is the first to report long-term follow-up data on serum hepcidin and iron metabolism profiles after LT in a well-defined group of patients homozygous for HFE C282Y. In addition, it is the first to definitely demonstrate the paramount role of liver in human HH. Our results showed, as expected, that patients with HH had low serum hepcidin levels, and that serum hepcidin levels were normalized after LT. Moreover, we found that, after a median follow-up of 5 years without iron depletion treatment, in patients without confounding factors, normal serum hepcidin levels were associated with normal biological iron metabolism and normal liver iron content, based on MRI. The patient and graft survival rates after LT were not significantly different between patients with HH and the overall population of patients that received LTs.
These results shed light on the evolution of iron burden in patients with HH. Although our cohort of patients was relatively small, we used rigorous selection criteria that excluded other causes of iron overload. Moreover, the iron burden assessment was based on biochemical and imaging data that were available over a long follow-up period.
Several case report studies[18-22, 30, 31] have previously attempted to elucidate the role of the liver in HH. However, the results showed discrepancies, and they precluded definite conclusions due to the small numbers of patients, the lack of genetic testing, the short-term follow-ups, and the lack of hepcidin assessments. Kowdley et al. studied LIC after LT of patient with HH. Their results indicated that iron overload did not recur in patients with HH, based on measurements of serum transferrin saturation and serum ferritin levels. However, they lacked LIC assessments for half of the surviving patients. Moreover, that study was performed before the discovery of hepcidin's role in iron metabolism; thus, no data were available on hepcidin secretion.
In the present study, after LT we found no increases in serum transferrin saturation, considered the earliest biochemical event in HH. In addition, we found no hepatic iron overload in the entire study group. Importantly, in our series only one patient received occasional phlebotomy treatment after transplantation, which could have hampered the recurrence of iron overload. We found two patients with modified iron parameters, but they both exhibited comorbidities that influenced iron metabolism. This finding underlined the importance of a careful search for potential confounding cofactors.
One of the major difficulties in describing the outcome of HH after LT has been the lack of definite markers for HH and its variable expression. However, we were able to measure serum hepcidin, which exerts a primary regulatory role in iron metabolism and in hemochromatosis expression.[33, 34] This provided a more precise, relevant evaluation of LT outcome. In mice, Garuti et al. found that transplantation of HFE wild-type livers in HFE knockout mice resulted in normal hepatic hepcidin expression at the messenger RNA (mRNA) level and a trend toward decreased iron burden. Moreover, transplantation of HFE knockout livers in HFE wild-type mice resulted in low hepatic hepcidin expression and the occurrence of iron overload. Here, we used this hepcidin approach for evaluating human LTs. We established for the first time that serum hepcidin levels, which were abnormally low before LT, were fully normalized after LT.
Our finding that serum hepcidin levels were low before LT was consistent with previous studies in patients with HH. Moreover, we confirmed that the hepcidin/ferritin ratio was preferable to the serum hepcidin evaluation. Among five patients (nos. 3, 4, 10, 15 and 18) with normal hepcidin values before LT, three (nos. 3, 4, and 15) had very high ferritin values; thus, the hepcidin/ferritin ratio was low, due to inadequate hepcidin synthesis, and this confirmed the hepcidin-deficient phenotype. Two patients had normal hepcidin levels and hepcidin/ferritin ratios (nos. 10, 18); both were diagnosed with HCC. We speculate that inflammation, related either to the cancer disease or to the effect of cancer treatment applied during the time the patient spent on the waiting list, or abnormal hepcidin secretion by the tumor, contributed to the normal hepcidin levels for these patients. Of note, two patients (nos. 1, 5) had normal ratios, with very low hepcidin and serum ferritin levels. This could indicate that the hepcidin/ferritin ratio may be inadequate in cases with low serum ferritin levels. The same explanation may apply to patients 1 and 18, who, after transplantation, presented with an abnormally low hepcidin/ferritin ratio. Cirrhosis itself has been associated with reduced hepatic expression; however, most of our patients were free of liver failure, therefore it is unlikely to have modified our serum hepcidin value. Since hepcidin levels have been shown to decrease during the induction phase of phlebotomy treatment, we cannot exclude that this factor contributed to the pretransplantation decreased hepcidin values. However, the data obtained with the hepcidin/ferritin ratios indicated that hepcidin remained relatively low when compared to the degree of iron stores, in line with the observations previously reported in patients exhibiting HFE hemochromatosis, and at a similar level to that of the untreated patients.
After LT, serum hepcidin was normalized and iron parameters remained normal for 5 years. This confirmed that the liver was the main producer of hepcidin in humans, and that normalization of hepcidin levels in the serum was associated with reversal of abnormal iron metabolism. Two patients with very high ferritin levels before transplant normalized their ferritin values after transplantation. It is possible that taking out most of the iron burden by removing the diseased liver, and restoring normal iron regulation with the new liver, may have led to this evolution. Altogether, these data support the view that hepcidin is a major actor in systemic iron regulation, and agrees with previous data obtained in mice.[17, 34]
It is likely that hepatic transplantation might also cure other genetic forms of hemochromatosis, unrelated to HH; for example, iron overload due to mutations in HJV, HAMP, or TFR2 genes. Indeed, like HFE, both HJV and TFR2 genes are involved in the transduction of signals related to iron sensing. Furthermore, like HFE, these two genes are expressed at the level of the hepatocyte membrane. This observation takes on particular importance when we consider that that these three mutated genes may be responsible for severe forms of juvenile hemochromatosis. Although the HFE gene is expressed in a large number of tissues, our clinical data support the view that HFE expression in nonliver cells is unlikely to play a major role in iron metabolism during genetic hemochromatosis.
This study also addressed the important, debated issue of patient and graft survival after LT in patients with HH. Our results showed that survival rates were not significantly different in patients with HH and the overall population of patients that received liver transplants. However, this result contrasted with previous reports.[23-27] Indeed, LT in patients with HH has been associated with lower survival, due to more frequent infections and cardiovascular diseases. However, some of those studies were conducted before definitive criteria were available for HH diagnosis. For example, Kowdley et al. diagnosed HH in patients prior to LT, but only 43% of patients underwent a genetic test confirmation. It is known that iron overload can occur in various liver diseases, and it can increase with the severity of the disease.[35-37] In a study by Brandhagen et al., only 10% of transplanted patients with iron overload were homozygous for HFE C282Y. Moreover, most previous studies lacked an evaluation of liver disease severity with the Child-Pugh score. Thus, the lack of a definitive diagnosis, combined with the possible selection of severe patients at inclusion, could have induced a selection bias, which influenced the findings that HH was associated with increased complications and mortality.
The present study included patients who were prospectively diagnosed and genetically confirmed for HH. These patients showed survival rates similar to those of the overall population. When we assessed survival according to the underlying liver disease, we did not find differences between patients with HH and those with other liver diseases. Most of our patients were iron-depleted before LT; only six had presented biological signs of iron overload at the time of LT. Thus, we did not have sufficient statistical power to assess survival between iron-depleted and noniron-depleted patients. However, it is likely that the previously described low survival rates of patients with HH were related to iron overload, regardless of the underlying disease, rather than specifically related to the genetic defect. This hypothesis is consistent with the recent study by Yu and Ioannou, which showed that post-LT survival in patients with hemochromatosis was higher in studies conducted more recently compared to survival reported in older studies. However, as acknowledged by the authors, that study lacked definitive genetic diagnoses of patients with HH and did not consider the potential association between HCC and hemochromatosis.
It is noteworthy that, in our population, HCC was the main indication for LT. In addition, only one patient had liver failure associated with excessive alcohol intake. This observation was fully consistent with the natural history of liver cirrhosis in HH, where liver failure is rarely seen in the absence of hepatotoxic cofactors (the case in nearly all our patients). Therefore, this confirmed that the major liver complication in HH was HCC, consistent with our experience and also with a study by Fracanzani et al. A study by Crawford et al. showed that, in patients with genetic hemochromatosis who received liver transplants, mortality was mainly related to HCC recurrence. However, some patients exhibited a tumor that was not within the Milan criteria, because the study was performed before the establishment of HCC guidelines. In contrast, we found that our patients with HH (mainly diagnosed with HCC classified as Child-Pugh A cirrhosis) had survival rates similar to our control population of patients who received liver transplants for treating HCC (also diagnosed with Child-Pugh A cirrhosis). Thus, among patients with and without HH, defined with the same currently accepted criteria, the patients with HH did not exhibit increased risk of HCC recurrence after LT.
In conclusion, our study in humans demonstrated that LT cured the iron overload phenotype and the biological defects in HH by normalizing hepcidin synthesis. These data further support the notion that the liver plays a primary role in iron metabolism. Moreover, our results showed that survival of patients with HH after LT was similar to that of the overall population of patients who received liver transplants.