NOTE

Effect of dietary Lorenzo’s oil and docosahexaenoic acid treatment forZellweger syndrome

Yasuhiro Arai1,2, Yohei Kitamura3, Masaharu Hayashi4, Kyoichi Oshida5, Toshiaki Shimizu1, and Yuichiro Yamashiro1

1Department of Pediatrics, Juntendo University, School of Medicine, 2Department of Child Neurology, Tokyo Metropolitan Tobu MedicalCenter for Persons with Developmental/Multiple Disabilities, 4Department of Clinical Neuropathology, Tokyo Metropolitan Institute forNeuroscience, 5Department of Nutrition, Kemin Japan K.K., Tokyo, and 3Nutritional Science Laboratory, Morinaga Milk Industry,Kanagawa, Japan

ABSTRACT We investigated the possible therapeuticeffect of decreasing plasma levels of very-long-chain fatty acids(C26:0) with a synthetic oil containing trioleate and trielucate(Lorenzo’s oil) as well as increasing docosahexaenoic acid(DHA) in red blood cells (RBC) with DHA ethyl ester in fourpatients with Zellweger syndrome. We investigated serialchanges of plasma C26:0 levels and DHA levels in RBC mem-branes by gas-liquid chromatography/mass spectrometry (GC/MS). After death, the fatty acid composition of each patient’scerebrum and liver was studied. Dietary administration ofLorenzo’s oil diminished plasma C26:0 levels. Earlier adminis-tration of Lorenzo’s oil was more effective and the response didnot depend on the duration of administration. DHA was incor-porated into RBC membrane lipids when administrated orally,and its level increased for several months. The final DHA levelwas correlated with the duration of administration and was notrelated to the timing of initiation of treatment. DHA levels in thebrains and livers of treated patients were higher than inuntreated patients. Early initiation of Lorenzo’s oil and thelong-term administration of DHA may be useful for patientswith Zellweger syndrome.

Key Words: docosahexaenoic acid, Lorenzo’s oil, very-long-chainfatty acids, Zellweger syndrome

INTRODUCTION

Zellweger syndrome (ZS) is the prototype of the generalized per-oxisomal disorders involving the central nervous system (CNS),and its features include severe psychomotor retardation, retinopa-thy, liver disease, and early death (Martinez 1992). Impaired differ-entiation and migration of neuronal cells are the major pathologicalfindings in the CNS of ZS patients (Martinez 1992). In ZS and othersimilar disorders, peroxisomes are not formed normally and defi-ciency of peroxisomal enzymes occurs.

In ZS, the peroxisomal enzymes involved in b-oxidation of very-long-chain fatty acids (VLCFA) 26:0 are defective, and these fattyacids accumulate in the plasma (Martinez & Vazquez 1998).Because plasmalogen synthesis is also catalyzed by peroxisomalenzymes, plasmalogen levels are low in ZS (Martinez 1989). More-over, these patients also show impaired synthesis of bile acids andcholesterol (Martinez et al. 1993). Patients with ZS have very low

levels of docosahexaenoic acid (DHA) in the brain, retina, and otherorgans (Martinez et al. 1993). Such marked DHA deficiency couldbe related to defective b-oxidation (Martinez & Vazquez 1998).In this study, we investigated the possible therapeutic effect ofdecreasing plasma C26:0 levels by administering a synthetic oilcontaining trioleate and trielucate (Lorenzo’s oil) and the effect ofincreasing DHA levels in red blood cell (RBC) membranes bytreatment with DHA ethyl ester in four patients with ZS.

We investigated serial changes of plasma C26:0 levels and DHAlevels in RBC membrane by using gas-liquid chromatography/massspectrometry (GC/MS).

CASE REPORTS

Patient 1 was a female infant, born in gestational week (GW) 39with a weight of 2634 g. The baby was the second child of unrelatedhealthy parents. She had congenital aortic arch interruption anddied of disseminated intravascular coagulation on day 43.

Patient 2 was a female infant, born in GW 38 with a body weightof 1668 g.

She had a mild asphyxia, hypotonia, and mild liver dysfunctionwith cholestasis at birth. She died of pneumonia on day 146.

Patient 3 was a female infant, born in GW 40 with a body weightof 2480 g. She had hypotonia and hepatosplenomegaly at birth, anddied of pneumonia on day 210.

Patient 4 was a male infant, born in GW 40 with a body weightof 2516 g. He had a congenital ventricular septum defect, and wasthe younger brother of patient 3. He died of intestinal bleeding onday 63.

Based on the findings, ZS was diagnosed in these patients:typical craniofacial features including an elongated forehead withhypoplastic supraorbital ridges, a large anterior fontanelle withwidely split sutures, large palpebral fissures, and a small nose andears, an increase of plasma VLCFA (C26:0), a low plasma level ofplasmalogen, and lack of peroxisomes in fibroblasts on skin biopsy.Magnetic resonance imaging of the brain showed bilateral germi-nolytic cysts and delayed myelination in all patients. After death,the fatty acid composition of the cerebrum and liver was studied.In all patients, autopsy was carried out within 24 h of death.

TreatmentLorenzo’s oil (glycerol trioleate oil at 2 g/kg per day and glyceroltrierucate oil at 0.5 g/kg per day) (Augbourg et al. 1993) and DHA(22% docosahexaenoic acid) were added to the diet (Martinez et al.1993). In patient 1, middle chain triglyceride (MCT) milk,Lorenzo’s oil, and DHA were administered daily from day 7 for5 weeks (Tables 1,2). In patient 2, MCT milk was started from day

Correspondence: Yasuhiro Arai, MD, PhD, Department of Pediatrics,Juntendo University School of Medicine, 3-1-1 Hongo, Bunkyo-ku, Tokyo113-8421, Japan. Email: aria_trc@mtrc.jp

Received May 7, 2008; revised and accepted August 11, 2008.

180 Congenital Anomalies 2008; 48, 180–182 doi:10.1111/j.1741-4520.2008.00201.x

© 2008 The AuthorsJournal compilation © 2008 Japanese Teratology Society

40, while Lorenzo’s oil and DHA were administered from day 55for 10 weeks. In patient 3, MCT milk was started on day 56,Lorenzo’s oil was administered from day 180 for one week, andDHA was given from day 84 for 16 weeks. In patient 4, Lorenzo’soil was administered from day 20 for three weeks, and DHA wasnot given. Treatments were performed with parental consent. In allpatients, glycyrrhizin and taurine were also administered to treatliver dysfunction and cholestasis. With parental consent, we inves-tigated serial changes of plasma C26:0 levels and DHA levels inRBC membranes by using gas-liquid chromatography/mass spec-trometry (GC/MS). We also measured the DHA levels in severalorgans after death.

RESULTS

Serial changes of plasma C26:0The plasma C26:0 level reached a minimum of 0.91 mg/mL on day42 in patient 1. In patient 2, the C26:0 level decreased transiently to

1.43 mg/mL on day 74, but after that increased gradually in spite oftreatment of Lorenzo’s oil. In patients 3 and 4, C26:0 levelsincreased without relation to Lorenzo’s of oil administration(Table 1).

Serial changes of DHA and arachidonic acid (AA) inRBC membranesIn patient 1, the DHA level in RBC membranes was the highest onday 7 before DHA administration and then decreased gradually inspite of DHA treatment. In patients 2 and 3, we did not measureDHA in RBC membranes before DHA administration. The levelsincreased gradually after the start of treatment for several months.The changes of AA levels were similar to those of DHA (Table 2).

Total fatty acids in the cerebrum and liverDHA levels in the brain and liver of treated ZS patients weremarkedly higher than in untreated ZS patients (Table 3). In the liverof patient 3, the DHA level was even higher than in the controls(Table 3).

Clinical responseIn patient 2, after three weeks of treatment with Lorenzo’s oil andDHA, her muscle tone and strength improved, and she could beweaned from mechanical ventilation. She started to show eye-to-eye contact and horizontal excursion. She also started to inspect hersurroundings and to respond to the sound of a rattle. She no longerneeded nasogastric tube feeding due to improvement of oralsucking.

DiscussionAn increase of C26:0 plays an important role in the pathogenesis ofneurological deficits in ZS (Evrard et al. 1978). ZS patients have asevere and characteristic defect of neuronal migration, which is themajor cause of their profound mental retardation and intractableseizures (Evrard et al. 1978). C26:0 is not only derived from thediet, but also from endogenous synthesis by a microsomal systemthat elongates long-chain fatty acids (Bourre et al. 1976). Simplerestriction of dietary VLCFA does not lead to biologic or clinicalimprovement (Brown et al. 1982). Excessive levels of C26:0 play arole in impairment of neuronal migration or promotion of neuronalapoptosis (Evrard et al. 1978). Unsaturated fatty acids (Lorenzo’soil) have been used to decrease the levels of saturated VLCFA by

Table 1 Plasma C26:0 concentrations (mg/mL) in the four patients

Patient 1 Patient 2 Patient 3 Patient 4

(before Lorenzo’s oil treatment*)

Day 7 3.75* Day 35 2.43* Day 90 2.7* Day 1 1.7*

120 2.0* 8 2.2*

14 1.62 74 1.43 187 3.9 23 3.0

21 2.23 89 2.74 – – – –

28 2.44 100 2.33 – – – –

35 1.94 115 2.01 – – – –

42 0.91 – – – – – –

Lorenzo’s oil administration

From Day 7 Day 55 Day 180 Day 20

Duration 5 weeks 10 weeks 1 week 3 weeks

Day of death Day 43 Day 146 Day 210 Day 63

Table 2 Ratio of DHA(22:6n-3) (%)/AA(20:4n-6) (%) concentra-tion in red blood cell membranes

Patient 1 Patient 2 Patient 3

Day DHA/AA Day DHA/AA Day DHA/AA

(before DHA treatment*)

7 6.6/12.3* Not performed

before treatment

Not performed

before treatment

14 3.2/8.2 74 1.3/0.7 90 1.45/8.37

21 4.9/10 89 1.04/4.18 120 5.43/10.3

28 3.7/10.2 100 1.20/4.56 187 8.51/9.65

35 1.9/6.2 114 1.57/5.32

42 1.3/3.0 120 2.49/6.87

131 3.34/6.55

134 5.34/8.68

146 4.38/9.30

DHA administration

From Day 7 Day 55 Day 84

Duration 5 weeks 10 weeks 16 weeks

AA, arachidonic acid; DHA, docosahexaenoic acid.

Lorenzo’s oil and DHA for Zellweger syndrome 181

© 2008 The AuthorsJournal compilation © 2008 Japanese Teratology Society

competition for the fatty acid chain elongation enzyme (Moser &Borel 1995). In a previous study, the mean plasma C26:0 levels ofcontrols and ZS patients were 3.30 mg/mL and 0.29 mg/mL, respec-tively (Moser & Moser 1996). In this study, Lorenzo’s oil only hadan effect in patient 1. In the other patients, this oil did not reduce theplasma C26:0 levels. Assuming that the long-chain fatty acids aresynthesized by the same enzyme, elongenase, earlier administrationof Lorenzo’s oil would more effectively compete for binding withthe enzyme. In the present study, the earlier administration ofLorenzo’s oil seemed to be more effective for decreasing plasmaC26:0 levels. It seems that the duration of treatment is not relevantto the effect on C26:0.

Docosahexaenoic acid plays an important role in the structure ofcell membranes, particularly of neuronal tissues and retinal pho-toreceptor cells, which suggests that the DHA deficiency observedin ZS patients could very well be involved in symptoms of thissyndrome (demyelination, psychomotor retardation, and retinopa-thy) (Innis 2003). DHA deficiency is a constant finding in ZS,especially in the brain (Martinez 1992). The beneficial effect ofDHA treatment has been suggested because DHA is a fundamentalconstituent of membrane phospholipids, especially in the brain. Ina previous report, treatment with DHA improved clinical featuresand advanced brain myelination on magnetic resonance imaging(Martinez et al. 1999). In this study, the DHA concentration ofRBC membranes gradually decreased further despite DHA admin-istration for 42 days in patient 1. In contrast, the DHA concentra-tion gradually increased in patients 2 and 3, rising for severalmonths. The present results indicate that DHA can be incorporatedinto membrane lipids when administered orally, and that it takesseveral months for the level to stabilize in erythrocytes. In a pre-vious report, the mean DHA level in the cerebrum and liverof controls was 1693 mg/g and 1092 mg/g, respectively (Martinez1989). In contrast, the mean DHA levels in the cerebrum and liverof untreated ZS patients were much lower (506 mg/g and 63 mg/g,respectively) than in the controls (Martinez 1989). In this study,DHA levels in the brains and livers of treated ZS patients weremuch higher than in untreated ZS patients. These results indicatethat DHA is not only incorporated into cell membranes, but also

into the brain and liver. The DHA level seemed to be correlatedwith the duration of DHA administration and not with the timingof initial administration.

REFERENCESAugbourg P, Adamsbaum C, Lavallard-Rousseau MC et al. (1993) A two-

year trial of oleic and eruic acids (‘Lorenzo’s oil’) as treatment foradrenomyeloneuropathy. N Engl J Med 329: 745–752.

Bourre JM, Daudu O, Baumann N et al. (1976) Nervonic acid biosynthesisby erucyl-CoA elongation in normal and quaking mouse brainmicrosomes: Elongation of other unsaturated fatty acyl-CoAs (mono andpoly-unsaturated). Biochim Biophys Acta 424: 1–7.

Brown FR, Van Duyn MA, Moser AB et al. (1982) Adrenoleukodystrophy:Effects of dietary restriction of very long chain fatty acids and of admin-istration of carnitine and clofibrate on clinical status and plasma fattyacids. Johns Hopkins Med J 151: 164–172.

Evrard P, Caviness VS, Prats-Vinas J et al. (1978) The mechanism of arrestneuronal migration in the Zellweger malformation: An hypothesis basedupon cytoarchitectonic analysis. Acta Neuropathol 41:109–117.

Innis SM (2003) Perinatal biochemical and physiology of long-chain poly-unsaturated fatty acids. J Pediatr 143: S1–S8.

Martinez M (1989) Polyunsaturated fatty acid changes suggesting a newenzymatic defect in Zellweger syndrome. Lipids 24: 261–265.

Martinez M (1992) Abnormal profiles of polyunsaturated fatty acids in thebrain, liver, kidney and retina of patients with peroxisomal disorders.Brain Res 583: 171–182.

Martinez M, Pineda M, Vidal R et al. (1993) Docosahexaenoic acid-A newtherapeutic approach to peroxisomal-disorder patients: Experience withtwo cases. Neurology 43: 1389–1397.

Martinez M, Vazquez E (1998) MRI evidence that docosahexaenoic acidethyl ester improves myelination in generalized peroxisomal disorders.Neurology 51: 26–32.

Martinez M, Vazquez E, Garcia-Silva MT et al. (1999) Treatment of gen-eralized peroxisomal disorders with docosahexaenoic acid ethyl ether.Rev Neurol 28: 59–64.

Moser HW, Borel J (1995) Dietary management of X-linked adrenoleukod-ystrophy. Annu Rev Nutr 15: 379–397.

Moser HW, Moser AB (1996) Very long-chain fatty acids in diagnosis,pathogenesis, and therapy of peroxisomal disorders. Lipids 31: S141–S144.

Table 3 Fatty acid composition in Zellweger (ZS) syndrome compared with controls

Controls (n = 5) ZS (untreated with DHA) Patient 1 Patient 2 Patient 3 Patient 4

Cerebrum

22:6n-3(DHA) 1693 506 693 1462 1041 331 (mg/g)

Liver

22:6n-3(DHA) 1092 63 396 543 1954 65 (mg/g)

The values of controls and untreated ZS are referred from Martinez (1989).DHA, docosahexaenoic acid.

182 Y. Arai et al.

© 2008 The AuthorsJournal compilation © 2008 Japanese Teratology Society