J. Inher. Metab. Dis. 14 (1991) 691-697 © SSIEM and Kluwer Academic Publishers. Printed in the Netherlands

Possible Deleterious Effect of L-Carnitine Supplementation in a Patient with Mild Multiple Acyl-CoA Dehydrogenation Deficiency (Ethylmalonic-adipic Aciduria) A. GREEN 1, M. A. PREECE 1, C. DE SOUSA 2 and R. J. POLLITT 3 iDepartment of Clinical Chemistry, Children's Hospital, Ladywood Middleway, Birmingham B16 8ET, UK; 2Department of Child Health, St. George's Medical School, Tooting, London SW17 ORE, UK; 3Neonatal Screening Laboratory and University Department of Paediatrics, Children's Hospital, Sheffield SIO 2TH, UK

Summary: A patient with riboflavin-responsive mild multiple acyl-CoA dehydrogenation deficiency of the ethylmalonic-adipic aciduria type experi- enced a recurrence of spontaneous hypoglycaemic episodes whilst being given supplementary L-cartinine. This phenomenon is explicable in terms of the known biochemical features of this condition and suggests caution in the carnitine supplementation of patients with defective oxidation of medium- or short-chain fatty acyl-CoA esters, This patient excreted excessive phenylpropion- ylglycine after an oral phenylpropionic acid load, Thus the phenylpropionic acid loading test is not completely specific for primary medium-chain acyl-CoA dehydrogenase deficiency as has been supposed.

INTRODUCTION

Secondary carnitine deficiency is a feature of many disorders of organic acid metabolism and L-carnitine is becoming widely used as an adjunct to the treatment of acute episodes or as part of long-term maintenance therapy in these conditions (Editorial, 1990). Experience of carnitine supplementation in the fl-oxidation defects is limited. Supplementation is suggested as a prophylactic measure in medium-chain acyl-CoA dehydrogenation deficiency (McKusick 22274) (Roe and Coates, 1989) but there have been conflicting reports on its value. The patient reported by Glasgow et al. (1980), now known to have medium-chain acyl-CoA dehydrogenase deficiency, continued to experience muscle weakness and hypoglycaemia on fasting despite the administration of D,L-carnitine. Wabner et al. (1984) reported that L-carnitine supplementation was effective in preventing metabolic decompensation during controlled fasting in a single case. A more recent study, on a patient who was less severely carnitine depleted, showed that supplementation did not change the

MS received 24,7.90 Accepted 31.1.91

691

692 Green et al.

biochemical or clinical response to fasting (Treem et al., 1989). In a patient with multiple acyl-CoA dehydrogenation deficiency (due to deficiency of electron-transfer- flavoprotein dehydrogenase activity) reported by Mandel et aI. (t988), L-carnitine administration led to the resolution of liver dysfunction, hypotonia and ataxia that had persisted after a severe acute attack. However, clinical exacerbations still occur during infection in this patient despite the normalization of free carnitine concentrations in plasma.

We report here a possible adverse effect of L-carnitine administration in a patient with a mild multiple acyl-CoA dehydrogenation deficiency.

CASE HISTORY

The patient, D.H., now aged 8½ years, has been described previously (Green et at., 1985). He has a relatively mild form of multiple acyl-CoA dehydrogenation deficiency characterized by an extremely variable urinary organic acid profile of the ethyl- malonic-adipic aciduria type. His underlying defect has not been defined, though similar patients have shown partial deficiences of either electron-transfer flavoprotein or electron-transfer-flavoprotein dehydrogenase (electron-transfer-flavoprotein-- ubiquinone oxidoreductase) (Rhead et at., 1987; Loehr et al., 1990). His clinical manifestations were less severe than in the patient described by Mandel et al. (1988), with later onset and less marked growth retardation. He presented at the age of 9 months with the first of a series of hypoglycaemic attacks that stopped only after he was given riboflavin and since the age of 2 years he has been maintained on a normal diet with riboflavin supplements, initially 100 mg twice daily, increasing to 200 mg twice daily and then to 200mg three times daily (Figure 1). His weight has closely followed the 10th centile, with height between the 10th and 25th centiles. His father and mother are close to the 10th and 25th height centiles, respectively.

At 3 years he was started on a course of supplementary L-carnitine (25 mg kg- i day-1) lasting 8½ months as described in the Results section. At the age of 8 years 4 months a phenylpropionic acid load test (Rumsby et al., 1986; Seakins and Rumsby, 1988) was performed.

METHODS

Organic acids were extracted from acidified urine using diethyl ether and ethyl acetate and converted to trimethylsilyl derivatives using bis(trimethylsilyl)trifluoroacetamide (containing 1% chlorotrimethylsilane) and pyridine in equal volumes. Extracts were analysed initially by gas chromatography-mass spectrometry.

Quantitation of phenylpropionylglycine was by gas chromatography with flame- ionization detection, using molar response factors based on the methyl/methylene contents, phenylpropionylglycine and the heptadecanoic acid internal standard, and assuming 100% recovery. This will somewhat underestimate the true phenylpropionyl- glycine content of the sample.

Total and free carnitine in plasma was measured by the method of de Sousa et al. (1990).

J. Inher. Metab. Dis, 14 (1991)

Carnitine in Multiple Acyl-CoA Dehydrogenation Deficiency 693

rag/day

Riboflavin ~ 6 0 0 [ / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / ~

~ / / / / / / / / / / / / / / / / / / / A 0

24 - Carnitine 300 rag/day

2O

v t 6

~ 1 2

8

.501h centile

- - ~,. i , . . ~ TM

. - 1 ~ ' - ~ Change of y

I t I , ~ Hypoglycaemic attack

centile

0 ~ [ t l f I 1 l 0 1 2 3 4 5 6 7 8

Age (years)

Figure 1 Weight chart of patient D.H. showing also dietary treatment and hospital admissions with hypoglycaemia

RESULTS

Organic acid excretion The urinary organic acid profiles shown by D.H. were extremely variable (Green et al., 1985) as has been noted by others in mild multiple acyl-CoA dehydrogenation defects. During acute attacks the excretion pattern was similar to that shown by Mandel et al. (1988), with large peaks of glutaric and adipic acids. In partial remission the pattern resembled more that seen in medium-chain acyl-CoA dehydrogenase deficiency, some samples showing recognizable peaks due to hexanoylglycine and suberylglydne. The biochemical response to riboflavin treatment was only partial, with a persistent excretion of ethylmalonic acid in particular (Green et al., 1985), but the variability of urinary organic acid excretion made it a poor indicator of response to treatment. No obvious change in the urinary organic acids occurred on carnitine supplementation, though there has been a tendency to increased excretion of the glycine conjugates of monocarboxylic acids (hexanoic, butyric, isovaleric, and isobutyric) with increasing age.

J. lnher. Metab. Dis. 14 (1991)

694 Green et at.

The phenylpropionic acid loading test resulted in the excretion of enough phenylpropionylglycine, approximately 0.7/~mol/mmol creatinine in the first post- load collection, to give a recognizable peak on gas chromatography. Identity was confirmed by gas chromatography-mass spectrometry. The urine samples collected at this time also contained hexanoylglycine and butyrylglycine (approximately 1 #mol/mmol creatinine) and showed a modest dicarboxylic aciduria, particularly ethylmalonic, adipic and suberic acids.

Clinical response to treatment

The start of riboflavin treatment at the age of 22 months was marked by rapid catch- up growth (Figure 1). There was a further mild hypoglycaemic episode 2 months after starting treatment but recovery was much more rapid than previously (Green et al., 1985).

The concentration of free carnitine in D.H.'s plasma was lower than normal when measured on several occasions (early results not shown). Based on this and relatively poor growth, the decision was taken to supplement L-carnitine intake cautiously, at a quarter of the dosage usually used (Mandel et al., 1988; Treem et al., 1988, 1989). After 4 weeks this had produced an increase in acyl carnitine but no significant change in free carnitine in plasma (Table 1).

At 118 days after commencing carnitine supplementation, D.H. presented to the casualty department with a history of vomiting during the previous 3 days. On examination he was pale, drowsy and had evidence of a throat infection. Blood glucose was 2.2 mmol/L. He responded well to the intravenous glucose and made a rapid and full recovery. Eighty-eight days later he had a similar episode (blood glucose 2.0 mmol/L) also associated with an upper respiratory tract infection.

Two subsequent hypoglycaemic episodes occurred at intervals of 26 and 2t days (plasma glucose 0.Smmol/L and 1.4mmol/L respectively). On both of these latter

Table 1 Trial of L-carnitine therapy. The carnitine and glucose results are from plasma

Age Status Glucose Carnitine (pmol/L) (years: days) (mmol/L) Total Free

2:193 Well - 23.9 11.6 3:3 Well. Immediately before - 38.6 20.6

start of carnitine 3:31 Well - 54.0 17.8 3:121 Acute admission 2.2 - - 3:209 Acute admission 2.0 - - 3:235 Acute admission 0.8 50.9 13.9 3:240 Well. Riboflavin increased - 56.3 37.4

to 200 mg twice daily 3:256 Acute admission 1.4 - - 3:262 Carnitine discontinued - - -

Normal for children 2-5 years (mean, range): total carnitine 34.2, 28.9-41.4/~mol/L; free carnitine 29.5, 23.0-39.0/~mol/L

J. Inher. Metab. Dis. 14 (1991)

Carnitine in Multiple A c yl-Co A Dehydrogenation Deficiency 695

two occasions he presented with vomiting, food refusal and drowsiness but there was no evidence of an infection. Recovery was rapid and uneventful following treatment with intravenous glucose.

His dose of riboflavin was doubled 16 days before the last attack and a blood sample taken at this time showed the concentration of free carnitine to be at the upper end of the normal range. This is probably more representative than the low result obtained during crisis 5 days previously since the concentration of free carnitine in plasma falls during prolonged fasting (see also Mandel et al., 1988). Carnitine was withdrawn 6 days after the last episode and D.H. then remained symptom free for 4 years. He had a further crisis at the age of 7½ years when his blood glucose fell to 2.4 mmol/L after a 12-hour attack of recurrent vomiting. Following treatment of this episode, the riboflavin dosage was increased to 600mg/day to take account of increased body weight.

DISCUSSION

Despite the marked clinical improvement following riboflavin supplementation, the urinary organic acid pattern in D.H. continued to show abnormalities, agreeing with the experiences of Gregersen et al., (1986) in a similar patient. Consistent with this is a continuing abnormality in medium-chain acyl-CoA dehydrogenation as revealed by the phenylpropioinic acid loading test. This test is essentially qualitative, the excretion of sufficient phenylpropionytglycine to be detected by HPLC (Seakins and Rumsby, 1988) or by gas chromatography (flame-ionization detection) indicating defective activity of medium-chain acyl-CoA dehydrogenase. In D.H. the loading test was positive by this criterion, though the amount of phenylpropionylglycine produced was less than is usually seen in medium-chain acyl-CoA dehydrogenase deficiency.

It has been suggested that the phenylpropionic acid load test is specific for primary medium-chain acyl-CoA dehydrogenase deficiency (Seakins and Rumsby, 1988). However, the positive result from D.H. is not too surprising in view of the similarity of some of the urinary organic acid profiles in this patient to those in medium-chain acyl-CoA dehydrogenase deficiency (Green et al., 1985). In vitro, in intact cultured fibroblasts, the fl-oxidation defect in D.H. is most severe at the C8 chain-length, as in medium-chain acyl-CoA dehydrogenase deficiency (Manning et al., 1990). The discrepancy between our result and the finding of Rinaldo et al. (1988) that the phenylpropionylglycine content of random urine in multiple acyl-CoA dehydrogen- ation deficiency was usually in the normal range may perhaps be explained by the variable, and usually much lower, amount of phenylpropionic acid contributed by gut flora compared to that given in the formal loading test.

The close temporal association between carnitine supplementation and the recur- rence of hypoglycaemic attacks (Figure 1) suggests that the two might be causally related. The initial symptom-free period after starting supplementation could reflect the time taken to fully replete the tissue stores, and thereafter there were increasingly frequent attacks until carnitine was discontinued. An alternative explanation for the observed course of events is that D.H. was 'growing out' of his riboflavin supplement,

J. Inher. Metab. Dis. 14 (1991)

696 Green et al.

as his weight had increased by 40% since the start of treatment. Doubling the riboflavin dosage did not prevent a further attack occuring 16 days later, and this seems less likely as an explanation for the resumption of hypoglycaemic attacks than a direct effect of carnitine supplementation. In any case it is clear that supplementary carnitine had no significant protective effect.

Primary carnitine deficiency due to defective carnitine transport (Treem et al., 1988; Stanley et al., 1990) results in impaired fatty acid oxidation in vivo and in vitro. Presumably secondary carnitine deficiency has similar effects and at first sight it seems paradoxical to suggest that carnitine supplementation might have been responsible for the resumption of hypogtycaemic attacks in D.H. However, in defects of the main fi-oxidation spiral, of which medium-chain acyl-CoA dehydrogenase deficiency is the classic example, the deficits in direct energy production and ketone- body generation are compounded by the accumulation of acyl-CoA esters. This, and the consequent sequestration of free coenzyme A are thought to play an important additional role in the general collapse of mitochondrial metabolism seen during the Reye-like attacks characteristic of many of these disorders. In patients with mild multiple acyl-CoA dehydrogenation deficiency of the ethylmalonic-adipic aciduria type, the fl-oxidation defect is most severe at the medium-chain level: in fibroblasts C8-C12 monocarboxylic acid derivatives accumulate when palmitic acid is a respiratory substrate (Manning et al., 1990). Thus a mild degree of carnitine deficiency, leading to a reduction in the maximal rate of supply of long-chain fatty acyl-CoA esters to the mitochondrial matrix, might be beneficial in patients with mild multiple acyl-CoA dehydrogenation deficiency as there will be less tendency for toxic accumulation of medium-chain intermediates during catabolic states.

CONCLUSIONS

The evidence for a protective effect of L-carnitine in fl-oxidation defects in general is conflicting and it is possible that in our patient carnitine supplements were harmful. Thus, the decision whether to give carnitine in disorders of straight-chain fatty acid oxidation is not straightforward. Primary carnitine deficiency leads to cardiomyopathy and generalized muscle weakness, often with demonstrable lipid myopathy, in a substantial proportion of cases (Eriksson et at., 1988; Stanley et aL, 1990). Thus the case for such supplementation is strongest in patients with chronic problems of this sort. However, a special degree of caution is warranted in conditions where there is impaired dehydrogenation of short- or medium-chain acyl-CoA esters, or where other intermediates such as 3-hydroxyacyl-CoA esters might be expected to accumulate. Perhaps only severe carnitine depletion should be treated in such cases and the dosage adjusted to maintain a slightly subnormal concentration of free carnitine in plasma.

These considerations will not apply to diseases such as propionic acidaemia, isovaleric acidaemia, or glutaric aciduria type I where carnitine is not involved in the rate-limiting step of the affected pathway.

J. Inher. Metab. Dis. 14 (1991)

Carnitine in Multiple Acy l -CoA Dehydrogenation Deficiency 697

ACKNOWLEDGEMENTS

We thank N. J. Manning for GC-MS analyses and Dr J. Insley for allowing studies of his patient. These studies were partly supported by the Medical Research Council.

REFERENCES

de Sousa, C. English, N. R., Stacey, T. E. and Chalmers, R. A. Measurement of L-carnitine and acylcarnitines in body fluids and tissues in children and adults. Clin. Chim. Aeta 187 (1990) 317-328

Editorial. Carnitine deficiency. Lancet 1 (1990) 631-633 Eriksson, B. O., Lindstedt, S. and Nordin, I. Hereditary defect in carnitine membrane transport

is expressed in skin fibroblasts. Eur. J. Pediatr. 147 (1988) 662-668 Glasgow, A. M., Eng, G. and Engel, A. G. Systemic carnitine deficiency simulating recurrent

Reye syndrome. J. Pediatr. 96 (1980) 889-891 Green, A., Marshall, T. G., Bennet, M. J., Gray, R. G. F. and Pollitt, R. J. Riboflavin-responsive

ethylmalonic-adipic aciduria. J. Inher. Metab. Dis. 8 (1985) 67--70 Gregersen, N., Christensen, M. F., Christensen, E. and Kolvraa, F. Riboflavin-responsive

multiple acyl-CoA dehydrogenase deficiency. Assessement of three years of riboflavin treatment. Acta Paediatr. Scand. 77 (1986) 676-681

Loehr, J. P., Goodman, S. I. and Frerman, F. E. Glutaric acidemia type II: Heterogeneity of clinical and biochemical phenotypes. Pediatr. Res. 27 (1990) 311-315

Mandel, H., Africk, D., Blitzer, M. and Shapira, E. The importance of recognising secondary carnitine deficiency in organic acidaemias: Case report in gtutaric acidaemia type II. J. Inher. Metab. Dis. 11 (1988) 397-402

Manning, N. J., Olpin, S. E., Pollitt, R. J. and Webley, J. A comparison of [9,10-3H]palmitic 3 and Eg,10- H]myristic acids for the detection of defects of fatty acid oxidation in intact

cultured fibroblasts. J. Inher. Metab. Dis. 13 (1990) 58~68 Rhead, W. J., Wolff, J. A., Lipsom, M., Falace, P., Desai, N., Fritchman, K., Moon, A. and

Sweetman, L. Clinical and biochemical variation and family studies in the multiple acyt- CoA dehydrogenation disorders. Pediatr. Res. 21 (1987) 371-376

Rinaldo, P., O'Shea, J. J., Coates, P. M., Hale, D. E., Stanley, C. A. and Tanaka, K. Medium chain acyt-CoA dehyrogenase deficiency: Diagnosis by stable isotope dilution measurement of urinary n-hexanoylglycine and 3-phenylpropionylglycine. N. Engl. J. Med. 319 (1988) 1308-1313

Roe, C. R. and Coates, P. M. Acyl-CoA dehydrogenase deficiencies. In Scriver, C. R., Beaudet, A. L., Sly, W. S. and Valle, D. (eds.), The Metabolic Basis of Inherited Disease, 6th edn, McGraw-Hill, New York, 1989, pp. 889-914

Rumsby, G., Seakins, J. W. T. and Leonard, J. V. Simple screening test for medium-chain acyl-CoA dehydrogenase deficiency. Lancet 2 (1986) 467

Seakins, J. W. T. and Rumsby, G. The use of phenylpropionic acid as a loading test for medium- chain acyl-CoA dehydrogenase deficiency. J. lnher. Metab. Dis. 11 (Suppl. 2) (1988) 221-224

Stanley, C. A., Treem, W. R., Hale, D. E. and Coates, P. M. A genetic defect in carnitine transport causing primary carnitine deficiency. In Tanaka, K. and Coates, P. M. (eds.) Fatty Acid Oxidation: Clinical, Biochemical and Molecular Aspects. A. R. Liss, New York, 1990, pp. 457.-464

Treem, W. R., Stanley, C. A., Finegold, D. N., Hale, D. E. and Coates, P. M. Primary carnitine deficiency due to a failure of carnitine transport in kidney, muscle, and fibroblasts. N. Engt. J. Med. 319 (1988) 1331-1335

Treem, W. R., Stanley, C. A. and Goodman, S. I. Medium-chain acyl-CoA dehydrogenase deficiency: Metabolic efficacy of long-term L-carnitine supplementaion. J. Inher. Metab. Dis. 12 (1989) 112-119

Wabner, L., Francomano, C., Brusilow, S., Valle, D., Frerman, F. and Goodman, S. Medium- chain acyl-CoA dehydrogenase (MCD) deficiency. Pediatr. Res. 18 (t984) 302A

J. Inher. Metab. Dis. 14 (1991)