The Difficulty with Rapid Communication

This editorial is accompanied by an article. Please see: Thomson M et al. Generalised mitochondrial cytopathy is an absolute contraindication to orthotopic liver transplant in childhood. J Pediatr Gastroenterol Nutr 1998;26:478-481. Fulminant hepatic failure (FHF) during infancy is fortunately a rare event. Less than 2000 cases per year of FHF occur in all age groups in the United States (1). In adults, FHF is defined as the development of encephalopathy within 8 weeks after the onset of symptoms of liver dysfunction (2). FHF during infancy has no such definition (3). Many infants identified with FHF have evidence of preexisting liver disease, or intrauterine insults, and physicians have great difficulty in detecting the onset of hepatic encephalopathy in infants. Although congenital or perinatally acquired viral infections such as herpes simplex or adenovirus, and “metabolic” disorders such as hereditary fructose intolerance, tyrosinemia, or galactosemia, may account for many cases, there remains a significant number of cases in which the cause of liver disease is unknown. Thompson and colleagues report in a Rapid Communication in this issue on a newly identified class of diseases, the mitochondrial cytopathies (MC), in which FHF in infancy may be one of the presentations (4). It is symbolic that the authors should have submitted this manuscript for consideration as a Rapid Communication, which is generally reserved for papers which have the potential for immediate impact on a given scientific field. Rapid communication is precisely what is required clinically in the setting of FHF of infancy: rapid recognition of the condition, rapid initiation of supportive therapy, rapid transfer to an intensive care unit (preferably at a liver transplantation center), rapid and continuous assessment of neurologic status, and in those with unremitting progressive liver failure, rapid availability of organ transplantation. But it is in this last point that we have the greatest difficulty, given the shortage of available donors for this age group. Alternatives such as living-related liver transplantation, reduced-size liver transplantation, and split-liver transplantation have in part been developed to meet this need. However, it is also true that in the setting of FHF in all age groups, the long-term success of transplantation is limited compared to more “elective” circumstances(3). Because of the modest success of transplantation for FHF of infancy in the setting of donor organ shortage, it becomes crucial that absolute contraindications to liver transplantation be reliably and rapidly identified. In this way, the use of organs for presently unsalvageable infants would not limit the opportunity for meaningful life in other equally ill infants, children, adults, or even potential living donors. This is a task that is fraught with difficulty and doubt. It is to this controversial discussion, that Thompson and colleagues voice their opinion that patients with generalized MC should not be transplanted. The details of the mitochondrial genome and structure of the mitochondria are beyond the scope of this Editorial and can be reviewed elsewhere(5). However, it is important to note that mitochondria maintain a separate genome. Not all of the components of the metabolically crucial molecular complexes of the mitochondria are encoded by mitochondrial DNA, but rather some are derived from nuclear DNA. Thus, mutations of either mitochondrial or nuclear DNA can result in abnormalities of oxidative phosphorylation or other mitochondrial function. Because there are multiple mitochondrial DNA copies in each mitochondria, and there are thousands of mitochondria in each cell, a state of heteroplasmy, or coexistence of normal and mutant mitochondrial DNA, can occur in a single cell. The phenotypic expression of that cell is determined by the relative proportions of normal or mutant genomes (6). Although the majority of mitochondrial diseases have been neuromuscular in nature, recognition of liver involvement has increasingly been reported(7-10). Most patients who have MC with liver involvement have had concurrent neuromuscular disease. Typically, affected infants have a marked lactic acidosis, an elevated lactate-to-pyruvate ratio, only moderately elevated transaminases, and significant coagulopathy without significant cholestasis(6). The obvious neurologic involvement preceding FHF in these rare infants has usually precluded transplantation. A single report of liver transplantation for isolated hepatic MC has been reported with good initial results (8). The condition of heteroplasmy, however, may account for the latest variant in mitochondrial cytopathies described by Thompson and colleagues: those with the sudden development of FHF in the absence of any other neuromuscular or developmental abnormality. This latest report is worrisome because the typical neurologic and biochemical abnormalities found in these disorders were not detectable untilafter liver transplantation, yielding tragic and devastating posttransplantation consequences. On the basis of their experience with two such cases of, if the reader will permit, “occult” generalized MC, the authors propose investigations to exclude generalized MC in all infants presenting with FHF. They suggest that clinicians obtain muscle or liver biopsy to determine histology, mitochondrial ultrastructure by EM, and respiratory chain enzyme activity, along with CSF lactate levels while using “vigorous correction of coagulopathy.” On the basis of these results, they would exclude from transplantation the identified patients with “occult” generalized MC. In support of this protocol, they offer the results of its implementation in their series of four presumably consecutive infants. Of these infants with FHF, three were successfully transplanted, and one was excluded from transplantation on the basis of having identified generalized MC. There are multiple goals to be achieved quickly by diagnostic laboratory testing in FHF. First, is the practical aim of testing of vital functions and their monitoring. Second, the identification of potentially treatable and reversible conditions (drug toxicity, or galactosemia as examples). Third, the identification of diseases that may be individually variably treatable, but are important for prognostic purposes or for future genetic counseling (i.e., tyrosinemia or neonatal hemochromatosis). It is this last group of tests to which the proposed protocol belongs. However, to be clinically useful, such testing must be very sensitive and specific, especially if one is toexclude potential interventions on the basis of the results. Furthermore, the results need to be readily available in a relatively short period of time to make an impact on clinical decision-making. The protocol as proposed illustrates some of the difficulties of rapid communication in the setting of infantile FHF. Importantly, the incidence of infants with generalized MC without generalized symptoms or signs is unknown, and the Rapid Communication by Thomson et al. is not designed to answer this difficult question. The incidence of a disease significantly affects both the sensitivity and specificity of the diagnostic tests designed to detect it. The lower the incidence of a disease, the more likely one is to have false positive and false negative results, a less than desirable possibility in a patient with FHF awaiting liver transplantation. Another concern is that the performance of invasive procedures in the setting of FHF is not without risk. The question must be asked-Is that risk worthwhile when the incidence of the disease being sought is unknown? Thompson and colleagues suggest that patients with “occult” generalized MC presenting with FHF will have abnormal tissue biopsies and an abnormal CSF lactate at the time of presentation. This may or may not be the case. The abnormalities in both tissue and CSF in their first two patients were not determined prior to transplant. As such, one cannot assume that these abnormalities would be equally discriminating, or even present, in patients with “occult” generalized MC on day one of FHF. It is impossible to exclude the effect of some confounding toxicity from FHF therapy, immunosupressive drugs, or sepsis post-transplant on the tissue samples or CSF lactate levels in these first two patients. It may be that the generalized nature of the MC was expressed phenotypically only subsequent to some factor in the transplantation process which altered the heteroplasmic balance in the sampled tissues. Other disorders have also been shown to produce mitochondrial injury in the liver (6). This problem can be only resolved by prospective assessment. Since the histories and biochemical data on the subsequent four cases presented by the authors is limited, it is not clear whether other clinical or laboratory information might have led to the suspicion of MC in the single confirmed case. Additionally, the most definitive tests for MC, tissue examination for histology, electron microscopy, and respiratory chain enzyme activity determination, may be difficult to apply in a timely fashion in the setting of FHF. The latter is reliably performed only in a handful of centers. Thus, the structure of the report makes it difficult to conclude, with the absolute certainty needed, that Thomson and colleagues' proposed protocol is sufficiently sensitive and specific to exclude from transplantation only those patients with“occult” generalized MC. The article by Thompson and colleagues raises an important question. How do clinicians identify unsalvageable patients in a high-risk setting for liver transplantation and thus ensure the best outcomes both in the use of scarce resources, and patient survival? The authors are surely correct to point out that certain infants with FHF may hold tragic surprises for transplant centers. In this respect, it is appropriate that they “rapidly communicate” with their colleagues regarding their experience. However, the challenge of “rapidly communicating” the cause of FHF in infancy still remains problematic, and exclusion of transplantation for some infants on the basis of the previously mentioned test results remains of unproved reliability. In such a scenario, it is likely that mistakes will continue to occur from time to time, and occasional infants with“occult” generalized MC will continue to be transplanted because of diagnostic difficulties. As a result of the discovery of MC and the potentially dismal results of transplantation for these disorders, the pressure on pediatric gastroenterologists to be right about the cause of FHF in infancy has increased. To this end, the development of reliable, quick, and accessible genetic testing for MC will be a great clinical tool. Given the impact that this development will have on our field, this information should definitely be “rapidly communicated” to us all.