Porphyrias

Frequency.

ADP is an autosomal recessive disorder resulting from loss of function mutations of ALA-D. The disorder is exceedingly rare, and fewer than 10 cases have been reported, two in one family. Most of the cases have represented compound heterozygotes. Eleven disease causing ALA-D mutations have been described, 10 of which are point mutations leading to amino acid substitutions. The sole exception is the deletion of two sequential bases resulting in a frame shift yielding a premature stop codon. Estimates of the frequency of heterozygosity for loss-of-function alleles of ALA-D, based on half-normal or less enzyme activity in erythrocytes of asymptomatic northern Europeans range from 1% to 2%. The extreme rarity of homozygosity or compound heterozygosity for ALA-D mutations suggests a high incidence of embryonic lethality when two mutant alleles of ALA-D are present.

Clinical Manifestations.

The clinical manifestations of ADP resemble those of the neurovisceral porphyrias but vary greatly in severity. The first two cases reported by Doss and colleagues in Germany were unrelated 15-year-old boys. Both suffered recurrent attacks of abdominal pain, nausea, vomiting, constipation, tachycardia, and a motor neuropathy causing paralysis. A third German case reported by Doss and colleagues also developed symptoms at age 15, including abdominal pain and a severe polyneuropathy. All three German patients were compound heterozygotes for ALA-D mutations.

The youngest reported case of ADP was a Swedish boy in whom muscle hypotonia was recognized at birth. At the age of 3, he developed recurrent attacks of abdominal pain, vomiting, and a motor polyneuropathy. This child also proved to be a compound heterozygote for ALA-D mutations. A liver transplant was performed at the age of 6 with the assumption that the metabolic defect was most profound in the liver. No change in the excretion of ALA was observed after the transplant, strongly suggesting that the red cells were the major source of this porphyrin precursor. The child died at the age of 9 of pneumonia and respiratory deficiency.

The oldest reported case of ADP was a Belgian man who developed a motor neuropathy and markedly increased urinary ALA excretion at the age of 63. ALA-D activity in erythrocytes was dramatically reduced to 1% of normal controls. The syndrome developed together with a clonal myeloproliferative disorder, namely polycythemia vera. Molecular studies revealed heterozygosity for a loss-of-function mutation (G133R) present in the polycythemia clone, which became clinically manifest with clonal expansion. This represents the only case in which a fully penetrant phenotype occurred in a heterozygote and supports the suggestion that the source of excess ALA in ALA-D porphyria is the erythron.

In two other cases, partial biochemical evidence of ALA-D deficiency were detected in heterozygotes but by a completely different mechanism. The first was a healthy Swedish newborn undergoing neonatal screening for hereditary tyrosinemia. Erythrocyte ALA-D activity was 12% of normal, and there were slight elevations of urinary ALA and coproporphyrin, as well as an increase in erythrocyte protoporphyrin. The child proved to be heterozygous for an F12L mutation, a catalytically inactive mutant that also affects secondary structure of the protein that favors formation of (with wild type protein) the low activity hexameric form of ALA-D. In spite of the ALA-D defect, the child remained healthy and developed normally.

The second case of partial biochemical expression in a heterozygote ALA-D mutant occurred in a 23-year-old American man and was also due to the F12L mutation. This patient presented with an acute porphyric attack clinically and biochemically compatible with hereditary coproporphyria (HCP). Indeed, a mutation of coproporphyrinogen oxidase (CPO) was identified. Because of a greater than usual increase in urinary ALA excretion and an increase in erythrocyte zinc protoporphyrin, an additional defect in ALA-D was suspected. Subnormal erythrocyte ALA-D activity was detected, and molecular studies revealed heterozygosity for the F12L mutation.

Pathogenesis.

The pathogenesis of the symptom complex associated with ADP is poorly understood. It has been suggested that ALA may act as a neurotoxin based on structural similarities between ALA and glutamate and ALA and γ-aminobutyric acid (GABA). High concentrations of ALA may interact with GABA receptors or inhibit glutamate uptake. ALA has been demonstrated to bind to GABA receptors in vitro, and GABA receptors are highly expressed by oligodendrocytes. In addition, oligodendrocytes are highly sensitive to glutamate-dependent metabolic pathways, which could be impaired by the effect of ALA on glutamate uptake. A possible role for ALA as a neurotoxin is supported in patients with hereditary tyrosinemia, who develop neurovisceral symptoms associated with marked inhibition of ALA-D by succinylacetone.

Laboratory Diagnosis.

Although ADP is rare, it should be suspected when an acute attack of abdominal pain suggests the possibility of a porphyric attack, especially when a prominent distal motor neuropathy is present. The dramatic reduction of ALA-D activity (to levels of 5% or less) results in the urinary excretion of large amounts of ALA and normal or minimally increased levels of PBG, a finding which distinguishes ADP from other neurovisceral porphyrias. Urinary excretion of coproporphyrin is markedly increased. The mechanism responsible for the increase in urinary coproporphyrin excretion is not known, but similar findings occur when ALA-D is inhibited in lead poisoning and when ALA is administered orally to normal individuals. Zinc-protoporphyrin accumulates in erythrocytes in patients with ADP, again by a poorly understood mechanism.

Heterozygous parents and siblings of probands with ADP have approximately half-normal ALA-D activity as measured in red cell lysates. Erythrocyte ALA-D assays are available in reference laboratories in the United States and Europe. Gene sequencing is available for ALA-D mutation analysis through the Department of Human Genetics at the Mount Sinai School of Medicine, New York, N.Y.

It is important to emphasize that ALA-D activity can be subnormal in disorders other than ADP. Patients with lead poisoning often develop clinical signs and symptoms similar to those found in patients with ADP and other neurovisceral porphyrias. Lead in plasma is taken up by red cells and inhibits ALA-D activity by displacing zinc atoms from the enzyme. The result is excess ALA in blood and urine, excess urinary excretion of coproporphyrin, and accumulation of zinc protoporphyrin in erythrocytes, findings also associated with ADP. The abnormal accumulation and excretion of porphyrins may be due, in part, to inhibitory effects of lead on CPO and FECH but is more likely due to disposition of the excess ALA, as occurs in ADP. Lead poisoning can be distinguished from ADP by measuring the blood lead level.

Patients with hepatorenal tyrosinemia type I, an autosomal recessive disorder due to deficiency of fumarylacetoacetate hydrolase (FAH), manifest acute episodes of porphyric-like neurologic crises and secrete large amounts of ALA in the urine. The FAH defect results in the generation of succinylacetone, a structural analog of ALA and a potent inhibitor of ALA-D.

A single patient with an acquired reversible, unidentified inhibitor of ALA-D, high urinary ALA levels, and a porphyria-like acute attack has been reported. In addition, both styrene and bromobenzene have been shown to inhibit ALA-D in rats. Humans exposed to these compounds have moderately decreased erythrocyte ALA-D activity but do not suffer clinical consequences.

ADP is a rare disorder and must be differentiated from other causes of ALA-D deficiency. Biochemical and molecular evidence is essential to establish the diagnosis of ADP.

Treatment.

Clear recommendations for the treatment of ADP are difficult because the number of reported cases is small, and the results of therapy have been variable. Suppression of ALA-S1 by the infusion of heme (heme arginate) proved effective in three German cases (all adolescent males), and infusions of glucose, narcotic analgesics, and hemin were effective in a 14-year-old American boy. Hemin infusion proved ineffective in the Swedish infant previously described. There have been no reports of marrow transplantation in a patient with ADP. Avoidance of drugs known to induce acute attacks in other neurovisceral porphyrias, generally by depleting hepatic heme pools and inducing ALA-S1, is suggested but has not been proven to be useful.

Frequency.

AIP is transmitted as an autosomal dominant trait, but five symptomatic children homozygous for disease–causing PBG-D mutations have been reported. Symptoms in these children began in infancy and included ataxia, dysarthria, psychomotor delay, and both central and peripheral neurologic manifestations. AIP is the most common of the acute neurovisceral porphyrias, but estimates of incidence vary according to the screening method used (clinical manifestations, erythrocyte PBG-D assays, molecular methods). Estimates based on a clinical diagnosis suggest an incidence of 0.13 per 1,000,000 in Europe. The incidence in Sweden is approximately four times higher, mainly due to a founder effect in northern Sweden. When subnormal erythrocyte PBG-D activity is used for screening, the incidence of AIP in European populations is much higher, but estimates based on erythrocyte PBG-D activity must be interpreted with caution. The normal range is wide, with considerable overlap between low-normal and high activity carrier values. Additionally, improper sample processing, storing, and shipping leads to decreased enzymatic activity. Molecular screening for PBG-D mutations on a large scale in a general population has not been done. It is likely that the incidence of undetected gene carriers in the population is much higher than the incidence of clinically evident AIP because most gene carriers remain symptom free throughout their lives. Recent estimates suggest a penetrance of only 1%. In all series, clinical symptoms occur more commonly in women than men.

Clinical Manifestations.

Clinical manifestations of AIP may occur before puberty in homozygotes or compound heterozygotes for PBG-D mutations, but most cases occur after puberty. Pediatricians are likely to encounter adolescent patients with symptomatic AIP, particularly adolescent girls following menarche. Virtually all of the clinical signs and symptoms that characterize attacks of AIP are due to neurovisceral dysfunction ( Table 12-2 ). Diffuse abdominal pain, often accompanied by nausea, vomiting, abdominal distention, and constipation, is the most common symptom of the acute attack. Abdominal pain is usually unremitting and severe but occasionally may be cramping. Decreased bowel sounds and even signs of ileus are generally present, but some attacks may be associated with enhanced gut motility and diarrhea. Tachycardia is often present. Other symptoms due to sympathetic over-activity or autonomic neuropathy include hypertension, tremors, sweating, and restlessness. Anxiety is a particularly common symptom, both during acute attacks and between attacks. A variety of other neuropsychiatric symptoms associated with AIP have been described, ranging from hypomania to overt psychotic behavior, and psychiatric symptoms may be the only manifestation of biochemically evident AIP.

A peripheral neuropathy may occur with or without an acute attack of abdominal pain. The neuropathy, due to axonal degeneration, is manifest mainly by muscle weakness in the extremities. Nerve conduction abnormalities may be minor, even with symptomatic weakness, but sophisticated studies of slow motor fiber conduction have been abnormal, even in the absence of symptoms. Cranial neuropathies are uncommon, but blindness due to optic nerve involvement or occipital lobe involvement has been reported.

Seizures may occur during an acute attack and are often associated with hyponatremia. When serum sodium levels are normal it is difficult to determine if the seizure is due to porphyria or to coexistent idiopathic epilepsy. In one large study of 268 patients, the lifetime prevalence of seizures was 5.1% in those who had suffered at least one acute attack and 2.2% in known carriers. Seizures occurring during latent periods are likely to be related to AIP. It is essential to establish a firm diagnosis of idiopathic epilepsy in AIP patients who experience seizures because many antiseizure medications may precipitate acute attacks (see later; Table 12-3 ). If seizures are associated with the hyponatremia of an acute attack, prolonged antiseizure therapy is not necessary.

Hyponatremia is a common finding during acute attacks. Hyponatremia maybe due to inappropriate secretion of antidiuretic hormone, but renal salt wasting has been implicated in patients with the lowest serum sodium vales. Serum sodium values below 120 nmol/L are generally found only in severe attacks.

The likelihood of full recovery after an acute attack has increased dramatically over the last 25 years. This is undoubtedly due to the introduction of effective therapy with hemin or heme arginate (see “Treatment” section). The risk of death, however, has not been eliminated. A review of 136 patients hospitalized with acute attacks at the University of Minnesota between 1940 and 1998 (i.e., both before and after the introduction of hemin therapy) found that 29 (21%) had died during an acute attack. In a study from Finland and Russia, 7 of 55 patients (13%) died as the result of an acute attack between 1966 and 2002. (Hematin therapy was first employed in 1971, but not widely used until 1978. ) In another study of 47 Finnish patients ascertained between 1967 and 1989, 6 deaths (12.7%) occurred as the result of an acute attack.

An unexplained finding in patients with AIP, variegate porphyria (VP), and HCP is a 35- to 70-fold increased risk of developing hepatocellular carcinoma. Most of these tumors occurred in individuals with a history of acute attacks. It has been suggested that DNA damage induced by ALA may be responsible for these observations. An unusual feature of the hepatocellular carcinomas arising in porphyric patients is the absence of associated liver cirrhosis. Chronic sustained hypertension is also prevalent in AIP and may be associated with renal impairment.

Pathogenesis.

Heterozygotes for PBG-D mutations have approximately half-normal enzyme activity in all tissues, which is usually sufficient for heme homeostasis. The enzymatic defect becomes manifest when demand for hepatic heme is increased, and ALA-S1 is induced. The result is increased production of ALA and PBG, creating a situation where the half-normal activity of PBG-D becomes rate limiting, and PBG (and ALA) accumulate in hepatocytes, circulate in plasma, and are excreted in the urine. Drugs are the most commonly recognized factors that increase the demand for hepatic heme, especially those metabolized by hemoprotein enzymes of the cytochrome P450 family. Barbiturates, anticonvulsants, and steroid hormones are well recognized precipitating factors in patients with AIP, but other factors not directly connected with P450 metabolism have also been incriminated, including fasting, dieting, and the stress of an acute illness. Most drugs taken by patients with AIP and other neurovisceral porphyrias have been classified as unsafe, potentially unsafe, probably safe, or safe. Isolated case reports have resulted in many drugs being classified as unsafe or potentially unsafe, resulting in a need for caution in drawing conclusions. Some drugs, however, have repeatedly been associated with acute porphyric attacks and must be avoided. A selection of these is shown in Table 12-3 .

The current prevailing view is that symptoms of the porphyric attack result from the toxic effects of ALA and/or PBG rather than from a deficiency of heme in neurons. The most striking evidence in support of this is the beneficial effect of allogeneic liver transplant in patients with repetitive and disabling acute attacks refractory to more conservative treatments. An allogeneic liver transplant rapidly normalizes urinary excretion of ALA and PBG and eliminates recurrent neurovisceral attacks. Furthermore, when explanted livers from AIP patients were transplanted into nonporphyric patients, the recipients promptly developed biochemical and clinical evidence of AIP. These impressive reports support the causative role of hepatic overproduction of ALA and PBG in acute neurovisceral attacks, but it is inappropriate to consider liver transplantation as a broadly applied treatment, especially with the efficiency of hemin or heme arginate infusions (see later).

Laboratory Diagnosis.

The signs and symptoms of an acute porphyric attack are quite nonspecific, and the correct diagnosis requires both consideration of a neurovisceral porphyria and a highly specific laboratory test. An acute porphyria is often not considered until fruitless, expensive imaging studies (and often unnecessary surgical exploration) are complete.

Increased urinary excretion of PBG is the key diagnostic finding in patients suffering an acute attack of AIP, HCP, or VP. Most methods used to quantify urinary PBG excretion rely on formation of a violet pigment when the sample is reacted with p-dimethyl aminobenzaldehyde (Ehrlich's aldehyde). Typical reference ranges for PBG excretion are 0-18 µmol/day (0 to 4 mg/day). Values for ALA range from 0 to 60 µmol/day (0 to 8 mg/day). In contrast, urinary PBG excretion during an acute attack generally range from 220 to 880 µmol/day, and ALA excretion ranges from 100 to 450 µmol/day. Recently liquid chromatography-tandem mass spectroscopy methods have been developed that are rapid, specific, highly sensitive, and simultaneously measure ALA and PBG in urine and plasma.

A simplified method for measuring PBG at the bedside or in the emergency department employs a small anion exchange column packed in a kit that includes all of the required reagents and a color chart. The trace PBG kit (Thermo Trace/DMA, Arlington, TX) can detect PBG at concentrations greater than 6 mg/L, and results correlated well with standard quantitative assays. The kit does not measure ALA.

Urinary porphyrin excretion (mainly uroporphyrin) is increased during acute attacks of AIP, mainly from nonenzymatic polymerization of PBG, but this is a much less specific finding than increased excretion of PBG and ALA. Fecal porphyrin levels are generally normal, but plasma levels may be increased. In contrast, increased urinary porphyrin excretion during an acute attack of HCP and VP consists mainly of coproporphyrin, and stool porphyrin excretion is also increased. Markedly increased stool coproporphyrin and protoporphyrin are found in VP. Fluorescence scanning of diluted plasma at neutral pH readily distinguishes VP from the other neurovisceral porphyrias, even during asymptomatic periods.

A word of caution is required in interpreting urinary porphyrin values. Minimal elevations of urinary coproporphyrin may be found in patients with liver disease, marrow disorders, and a host of other illnesses. An incorrect diagnosis of porphyria is commonly made by over-interpreting minimal elevations of urinary coproporphyrin, leading to a delay in establishing the true cause of abdominal pain and other neurovisceral symptoms.

Assays of PBG-D activity in erythrocyte lysates have been widely employed to confirm the diagnosis of AIP and to detect asymptomatic family members once an index case has been identified. Erythrocyte PBG-D activity is approximately half-normal in most patients with PCT. Normal PBG-D activity, however, does not exclude a diagnosis of AIP because of differential splicing of the PBG-D transcript in the liver and the red cell. Mutations in exon 1 of the PBG-D gene affect activity of the enzyme in the liver, but the erythrocyte enzyme is encoded only by exons 2 through 15. It is important to emphasize that a valid half-normal measurement of erythrocyte PBG-D activity does not distinguish between asymptomatic (latent) and clinically manifest cases of AIP. Only urinary PBG levels can make this distinction.

PBG-D mutation detection by DNA analysis has proven to be more reliable than PBG-D enzyme assays and is now the preferred method for confirming a diagnosis of AIP and for identifying asymptomatic carriers. DNA studies have demonstrated that enzymatic assays may misclassify up to 15% of individuals tested. Mutation detection for AIP patients and their relatives is now readily available through the Department of Human Genetics at the Mount Sinai School of Medicine, New York, N.Y.

Treatment.

Treatment of an acute attack of AIP (and the other neurovisceral porphyrias) is designed to repress the activity of hepatic ALA-S1 and to offer supportive care and relief of symptoms. Intravenous infusions of heme result in a rapid reduction in levels of ALA and PBG in plasma and urine. Heme binds to plasma hemopexin and albumin and is then taken up by hepatocytes, where it expands the heme pool and represses ALA-S1. Many small, uncontrolled trials have demonstrated both biochemical and clinical responses to infusion of heme. In one study involving 22 patients suffering 57 acute attacks, all patients responded favorably, but therapy was most beneficial when administered early in the attack, a finding confirmed by others. A single, placebo-controlled trial found dramatic biochemical benefits of heme infusion, but clinical benefits were more difficult to confirm.

Heme is not soluble at neutral pH, but solubilized hemin (heme-chloride) and heme arginate have been administered to many patients. In the United States a lyophilized preparation of hemin, (Panhematin, Recordati Rare Diseases, Lebanon, NJ) has been approved by the U.S. Food and Drug Administration for the treatment of acute porphyric attacks. Hemin is unstable when reconstituted with sterile water, and degradation may cause an anticoagulant effect and phlebitis at the site of infusion. Phlebitis can be avoided by the use of intravenous catheters that extend to larger veins (e.g., the superior vena cava). Instability and the locally irritating effects of lyophilized hemin can be attenuated by solubilizing hemin in 25% albumin, a method now recommended. The recommended regimen for heme therapy is 3 to 4 mg/kg infused once daily on 4 consecutive days. Clinical improvement generally occurs within 1 or 2 days, but when neuronal damage is severe, complete recovery may take months. The keys to success are to initiate therapy as soon as an attack is recognized and to complete a 4-day course. Rare side effects of therapy include constitutional symptoms such as fever, myalgias, and malaise, but one case of marked hypotension and another of transient renal failure have been reported. The latter case was associated with an excessive dose (1000 mg) of hemin.

Heme arginate (Normosang, Orphan Pharmaceuticals, Paris, France) is more stable in solution than hemin, is equally effective, and is widely available in Europe and South Africa but remains an experimental drug in the United States. Phlebitis is a rare complication with heme arginate, and the shelf life is considerably longer than that of hemin. Accidental overdosing with heme arginate has been associated with acute hepatic failure. Adenoviral-mediated hepatic expression of PBG-D corrects the metabolic defect in mouse models of AIP, but no studies have been done in humans. Intravenous infusions of recombinant human PBG-D have been administered to normal subjects and asymptomatic PBG-D mutation carriers. Transient decreases in plasma concentrations of PBG were noted, and no serious side effects occurred. No trials of recombinant PBG-D for the treatment of symptomatic AIP have been reported.

Intravenous infusions of glucose (300 to 500 g/day, administered as 10% glucose) may be effective when an attack is mild or until hemin is obtained, but glucose is far less effective than hemin. Infusions of glucose in water solutions are contraindicated because they aggravate hyponatremia, which can prove fatal. The mechanism of the glucose effect relates to the previously described finding that transcription of ALA-S1 is upregulated by the peroxisome PGC-1α and transcription of PGC-1α is controlled by glucose availability. Insulin may downregulate PGC-1α, and it has been suggested that the combination of glucose and insulin may be more effective than glucose alone.

Acute attacks require treatment of symptoms in addition to therapy directed at ALA-S1. Pain is often severe, and full doses of morphine or other opioids are both safe and effective. Phenothiazines in modest doses are safe and are useful for the control of nausea and anxiety. Hypertension and tachycardia may respond to β-adrenergic blocking agents, but these agents should be used with caution when patients are hypovolemic because increased catecholamine secretion is often an important compensatory mechanism. Seizures during an acute attack can be a particularly vexing problem because most antiseizure drugs have the potential to worsen the intensity of the attack. Clonazepam and gabapentin may be safe, and benzodiazepines are probably safe. Hyponatremia and hypomagnesemia should be corrected because these abnormalities are common and may be causative for seizures.

The long-term management of patients with acute neurovisceral porphyrias is focused on the prevention of future attacks. Avoidance of unsafe drugs (see Table 12-3 ) is critical as is maintenance of a balanced, calorically adequate diet. In some women acute attacks occur only in the luteal phase of the menstrual cycle. In these cases gonadotropin-releasing hormone (GnRH) analogues can be highly effective. Prolonged use of GnRH analogues in young women may cause irreversible bone loss, but low-dose transdermal estradiol may safely prevent this complication. Both cyclic and noncyclic attacks may be prevented by frequent infusions of hemin, but the expense and the potential for iron overload (8 mg of iron per 100 mg of hemin) makes this an unattractive option.

Levels of progesterone, an inducer of hepatic heme synthesis, are elevated in pregnancy, but pregnancy is well tolerated by most women with acute porphyrias. In a large series reported from Finland of 176 pregnancies in women with AIP or VP, porphyric symptoms occurred in fewer than 10%. Worsening symptoms during pregnancy may be due to harmful drugs, inadequate nutrition, or both.