Main

Propionic acidemia is a genetic disorder of amino acid metabolism caused by a deficiency of propionyl-CoA carboxylase(1–3). The first clinical presentation may be manifest at any age from the neonatal period on. The spectrum of possible clinical symptoms is wide. Many patients exhibit poor feeding, vomiting, and lethargy leading to coma if onset of treatment is delayed. Metabolic derangement may be precipitated by infection or excessive protein intake(4). Ketoacidosis and hyperammonemia are prominent laboratory findings, in particular during episodes of metabolic decompensation. The reported neurologic symptomatology is diverse. Frequently observed neurologic signs are axial hypotonia, extrapyramidal manifestations, and seizures (see Table 1). Surtees et al.(5) recently reported the neurologic outcome of 20 patients with propionic acidemia. Extrapyramidal movements were a prominent finding in both the early-onset group and the late-onset group (Table 1). In the late-onset group, bilateral basal ganglia hypodensities were observed on cranial computer tomography when performed shortly after the onset of the movement abnormalities. These lucencies resolved completely in most patients over a 1-3-mo period after the onset of dietary treatment(5). The pathophysiology of these basal ganglia abnormalities is unknown.

Table 1 Clinical symptoms and CT/MRI findings in propionic acidemia patients
Full size table

We report proton (1H) MRS studies of the basal ganglia in three patients with propionic acidemia (one with an early onset, two with the late-onset form). The aim of the study was to investigate whether 1H MRS may contribute to the insight of the underlying mechanisms accounting for the basal ganglia damage in propionic acidemia and whether 1H MRS may contribute in the monitoring of treatment.

METHODS

Three children with propionic acidemia were studied. The diagnosis was made on the basis of clinical presentation, urinary metabolite excretion pattern, and in vitro enzyme assays. Propionyl-CoA carboxylase activity in cultured skin fibroblasts ranged from 0.2 to 1 nmol/h/mg protein (normal 15-75). The clinical symptoms are summarized in Table 1. Patient 1 presented at the age of 6 mo with hypotonia and choreoathetoid movements and was diagnosed at 9 mo of age. Patient 2 came to attention at the age of 1.5 mo when she became comatose after nasogastric tube feeding because of failure to thrive. Patient 3 was diagnosed at the age of 5 d when she presented with feeding difficulties, hypotonia, and respiratory insufficiency. Patients 2 and 3 required hemodialysis or peritoneal dialysis. Once the diagnosis was made, all of the patients were treated with a low protein diet and were given a trial of biotin, which was discontinued when they were shown to be unresponsive. Patient 1 was followed up to his present age of 36 mo and is still severely retarded and choreoathetotic. Patient 2 presented a normal development at the age of 3 mo. Patient 3 had initially repeated episodes of metabolic decompensation, but has remained stable now for 1 y. At 32 mo of age she was still retarded, but she had improved considerably during the last year. At no time choreoathetosis was noted. MRI and 1H MRS of the brain were performed in all three patients during stable clinical and biochemical conditions.

Organic acids in plasma, urine, and CSF were quantitated by gas chromatography(6). A standard amino acid analyzer was used to determine the amino acids in plasma, urine, and CSF. CSF concentrations of free and total GABA were determined by using a stable isotope dilution gas chromatography/mass spectrometry assay(7).

MRI and MRS were performed on a 1.5 Tesla MR instrument. Sagittal T1-weighted images and transverse T1- and T2-weighted images were made. The standard imaging head-coil was used for 1H MRS. A 2 × 2 × 2-cm voxel was chosen in the area of the caudate nucleus, putamen, globus pallidus, and thalamus on the left side, containing a mixture of white and, in particular, central gray matter. The STEAM sequence was used with a repetition time of 2500 ms, an echo time of 20 ms, and 128 acquisitions[see Frahm et al.(8) for all technical details]. The averaged measurements were zero-filled to 2048 data points, exponentially filtered to give 1 Hz line-broadening before fast-Fourier transformation. The spectra were quantified by peak area measurements using system software (Siemens SP63). The following peaks were quantified: NAA (2.02 ppm), Glx (2.1-2.5 ppm), “creatine” representing creatine and phosphocreatine (Cr, 3.02 ppm), “choline” representing choline-containing compounds (Cho, 3.20 ppm). and mI (3.56 ppm). To determine a baseline, four fixed areas were used: left foot of the NAA peak (2.02 ppm), area between the Cr peak (3.02 ppm) and the NAA peak at 2.70 ppm, left foot of the mI peak (3.56 ppm), and the area at the right of the Cr peak at 3.93 ppm. The peaks of spectra were normalized to the Cr peak (3.02 ppm). Creatine age- and area-dependent values were used as known from the literature(9, 10). An exception was made for the region between 2.1 and 2.5 ppm, containing multiple overlapping peaks representing glutamine, glutamate, and some GABA (together Glx). The ratio of the integral of this area and the creatine peak was used as a rough indication of the concentration of the contributing metabolites. The control group for the MRS investigations consisted of eight healthy children varying in age from 3 to 48 mo.

RESULTS

Results of biochemical studies are summarized in Table 2. In the period of the 1H MRS all patients had normal plasma ammonia values. The urinary 3-OH-propionic acid concentration was elevated in all patients. The free and total GABA in CSF was normal, whereas CSF glycine and alanine concentrations were increased in all patients.

Table 2 Laboratory findings of patients with propionic acidemia under treatment, performed in the same period as1H MRS investigation
Full size table

Initial MRI of all patients showed a mildly enlarged ventricular system and subarachnoid spaces. The MRI of patient 1 at 10 mo of age (Fig. 1) showed an abnormally high signal intensity on T2-weighted images, abnormally low on T1-weighted images in the caput and corpus of the caudate nucleus and in the putamen. Globus pallidus and thalamus were normal. Myelination of the cerebral hemispheres was delayed and irregular with patchy areas of high signal intensity on T2-weighted images and low signal intensity on T1-weighted images. At the age of 24 and 36 mo the repeat MRIs showed progress in myelination and diminishing cerebral atrophy, but the basal ganglia abnormalities remained unchanged.

Figure 1
figure 1

MRI of the brain of patient 1 at the age of 10 mo. The T2-weighted MR images show enlarged subarachnoid spaces and ventricular system, lesions in the putamen and caudate nucleus, and delayed myelination.

Full size image

MRI of patient 2, made at the age of 2 mo, showed, apart from some cerebral atrophy, no other abnormalities. The images showed the beginning of myelination and normal basal ganglia.

At 20 mo of age the MRI of patient 3 showed mildly delayed myelination, which was not yet completed in the temporal lobes and in the directly subcortical areas. Some mild and patchy increases in signal intensity were seen in the frontal and occipital periventricular white matter on T2-weighted images. The basal ganglia were normal. Repeat MRI at the age of 32 mo showed considerable improvement of myelination and diminished cerebral atrophy.

The spectroscopic data from the area of the basal ganglia of the patients were compared with those of the eight healthy controls. The data are summarized in Table 3 (see also Fig. 2). In all patients the level of NAA was decreased. The Cho level in patients 1 and 3 did not differ from the level in controls, but was elevated in patient 2. In patients 1, 2, and to a lesser extent in patient 3, elevated signals were seen in the range of 2.1-2.5 ppm, indicating a rise in Glx. In all patients mI was decreased.

Table 3 1 H MRS findings of basal ganglia in patients with propionic acidemia and normal controls
Full size table
Figure 2
figure 2

Short echo time STEAM spectrum obtained from the area of the basal ganglia on the left side in patient 2 (upper spectrum), compared with a normal spectrum of a control child of the same age(lower spectrum). Note the decrease in NAA (2.02 ppm), increase in Glx (2.1-2.4 ppm), rise in Cho (3.2 ppm), and decrease in mI (3.56 ppm) relative to Cr (3.02 ppm) in patient 2 compared with the control.

Full size image

DISCUSSION

Propionic acidemia is caused by a defect in the carboxylation of propionyl-CoA to methylmalonyl-CoA. Biochemically this disorder is characterized by moderate to severe ketoacidosis, hyperglycinemia, and hyperammonemia. The acidosis is the result of accumulation of various normal and abnormal organic acids, in particular propionate, 3-hydroxybutyrate, methylcitrate, and 3-hydroxypropionate. Propionate is derived from hydrolysis of propionyl-CoA and is detectable in all body fluids. The hyperammonemia during acute episodes is thought to be caused by the inhibitory effects of propionyl-CoA on N-acetylglutamate synthetase, the first enzyme involved in ureagenesis(11).

Although the metabolic routes of the production of abnormal acids in propionic acidemia are known(12) and the effects of these organic acids on different organ systems have been studied, the explanation for the heterogeneous clinical presentation is unknown(13). The differentiation between the early-onset and late-onset group can neither be made on the basis of different genetic mutations(14), nor on the basis of different levels of residual activity of propionyl-CoA carboxylase(13). In general, episodes of clinical deterioration are accompanied by metabolic decompensation, but clinical deterioration and necrosis of the basal ganglia have been described despite documented excellent metabolic control at the time of onset of the cerebral dysfunction(15).

Our three patients had different clinical presentations. Patient 1 presented with choreoathetosis at the age of 6 mo. Despite adequate treatment the choreoathetosis persisted, probably because of irreversible damage of the basal nuclei. The delay in treatment may be important in this respect. MRI of the brain at the age of 10, 24, and 36 mo showed symmetrical lesions in the putamen and caudate nucleus, in addition to a delay in myelination and some cerebral atrophy. Patients 2 and 3 presented clinical symptoms at 1.5 mo and 5 d, respectively. They had no choreoathetosis, nor abnormalities of the basal nuclei.

Bilateral lesions of the basal nuclei have been often noted in propionic acidemia, in particular in the late-onset variant (Table 1). The pathophysiologic mechanisms underlying these lesions, however, are still unknown.

The MRS findings from the basal ganglia in our patients were essentially identical showing decreased NAA, increased Glx, and decreased mI peaks in comparison with the values found in healthy controls. These changes were reproducible in follow-up. The decreased level of NAA is commonly ascribed to neuronal dysfunction, damage, or loss(16). It is possible that the decrease in NAA can be partially ascribed to delayed neuronal maturation.

At the lower magnetic field strengths used in clinical studies, the separation of glutamine from glutamate is not possible. The elevation of Glx may represent an increase in glutamine(17–20). Elevated glutamine peaks in MRS of the brain have been described most often in hyperammonemia(17–20). Hyperammonemia leads to enhanced production of glutamine and glutamate from 2-oxoglutarate. Glutamate is a major excitatory neurotransmitter, whereas GABA, formed by decarboxylation from glutamate, is the most important inhibitory neurotransmitter. In the presynaptic neuron, glutamine is converted to glutamate and ammonia by glutaminase. Glutamate can either be converted to GABA by glutamic acid decarboxylase or be released in the synaptic cleft. After release by the presynaptic neuron, glutamate is taken up by the astrocyte where it is converted into glutamine by glutamine synthetase and then transported back to the presynaptic neuron. This cycle operates mainly between the presynaptic neuron, the neuronal cleft, and the astrocyte. Hyperammonemia has a great impact on this cycle by stimulating glutamine synthesis via glutamine synthetase, by possible inhibition of glutaminase, and by inhibition of glutamate re-uptake by the astrocyte. High plasma ammonia levels would lead to an accumulation of glutamine in the brain, mainly in astrocytes. An imbalance between excitatory and inhibitory neurotransmitters may contribute to the cerebral dysfunction. Accumulation of intracellular glutamine is reflected in elevated Glx in 1H MRS(17–20). However, at the time of1 H MRS in our three patients, the plasma ammonia concentration and the plasma and CSF concentrations of glutamine and glutamate were normal or just above normal (see Table 2). This indicates that blood and CSF concentrations may differ from tissue concentrations. Differences between blood, CSF, and brain parenchyma metabolite concentrations have been reported before in metabolic disorders, in particular with respect to lactate in mitochondrial disorders(21, 22). MRS may reveal considerable elevation of lactate in brain tissue, whereas lactate in CSF and blood is normal or only slightly elevated(21). On the other hand, MRS may not show any evidence of elevated lactate, whereas CSF lactate is high(22). Assessment of brain tissue metabolite concentrations by MRS is an additional method for monitoring metabolic stability at tissue level to the biochemical analysis of urine, plasma, and CSF.

Glutamate is a major excitatory neurotransmitter in the CNS and is well known for its excitotoxic effects in many forms of brain damage, particularly in hypoxia-ischemia. Elevated Glx reflecting elevated glutamate has been found in cases of acute cerebral hypoxia-ischemia(23–25). The elevated Glx that we found in our patients could be a sign of active excitotoxic damage of the basal nuclei. Considering the clinical and metabolic stability of the patients at the time of MRS, this hypothesis is less attractive.

Another possible explanation for basal nuclei dysfunction in propionic acidemia could be alterations in mitochondrial energy metabolism causing accumulation of toxic metabolites(15). The presence of elevated cerebral lactate levels is an argument in favor of this hypothesis. However, although we found slightly increased lactate values in CSF, we found no spectroscopic evidence of increased lactate within the brain in our patients.

It has been suggested that one or more of the metabolites accumulating in propionic acidemia is neurotoxic(15). Our data show that 3-OH-propionate is the only metabolite of the propionate derivatives detectable in large quantity in CSF. Unfortunately this metabolite could not be detected by MRS methods.

Reduced cerebral mI as found in our patients has also been observed in patients with hyperammonemia and elevated Glx levels. Glycine co-resonates with mI at 3.56 ppm(26). CSF glycine was elevated in all patients. It cannot be excluded that mI was actually lower than now estimated, because the decrease in peak area was partly compensated by the increase in glycine. However, it is not probable that there was a major contribution of glycine in our patients. In all patients additional spectra were acquired with a longer echo time (135 ms). These did not show the presence of a resonance at 3.56 ppm, which would have been expected in the case of elevation of glycine. The significance of the mI reduction is not understood. It is generally believed that mI plays a role as a precursor of the inositol polyphosphate family of intracellular messengers. It may play an important role in the detoxification reactions of the brain(17, 18), and it has been suggested that low mI may be the result of enhanced use of detoxification processes. Another function of mI may be that of osmoregulator(17, 18). It is possible that the decrease in mI is compensatory for the increase in Glx to maintain osmotic balance in the brain.

The observation of spectroscopic abnormalities in metabolically stable patients is important. Serious, even fatal, neurologic deterioration with basal ganglia necrosis may occur in patients who are in a stable metabolic condition as estimated by conventional biochemical analyses(15). The current study shows that biochemical changes at tissue level as seen by MRI/MRS differ from the results obtained by biochemical analysis of body fluids. MRS of the brain in vivo would provide a valuable, noninvasive tool in monitoring patients with propionic acidemia and other inborn errors of organic acid metabolism. In this respect MRI and MRS would have complementary values, MRI reflecting both temporary and permanent cerebral structural damage, and MRS reflecting actual intracellular biochemical changes at tissue level.