X-Linked Adrenoleukodystrophy: TheAustralasian Experience

Edwin P.E. Kirk, Janice M. Fletcher,* Peter Sharp, Bill Carey, and Alfred PoulosDepartment of Chemical Pathology, Women’s and Children’s Hospital, Adelaide, Australia

Our objective was to review the Australa-sian experience of X-linked adrenoleuko-dystrophy (ALD), to compare the spectrumof disease seen in Australasia with previ-ously published data from elsewhere, and toassess the reliability of carrier testing.Study design was a retrospective review ofrecords collected over a 15-year period, thesetting was an international referral labora-tory for the study of metabolic disease, andthe subjects were all known cases of ALDdiagnosed in Australia and New Zealand be-tween 1981 and 1996 and their families. Weestimate that the combined incidence ofALD and its variants in Australasia is atleast 1.6 per 100,000. Of 95 affected males, 51had cerebral adrenoleukodystrophy, 24 hadadrenomyeloneuropathy, 15 had Addison’sdisease only, and 5 remained asymptomaticwhen last examined. However, the distribu-tion of phenotypes among newly diagnosedpatients has changed substantially over thelast 15 years, with cerebral forms of the dis-ease forming a decreasing proportion ofnew diagnoses. The measurement of plasmavery long chain fatty acids (VLCFAs) alonedetects 93% of women who can be proven tobe carriers. The addition of genetic linkagestudies or assay of VLCFAs in cultured fi-broblasts improved this detection rate tothe point that there were no obligate carri-ers who could not be detected using a com-bination of two or more techniques. Am. J.Med. Genet. 76:420–423, 1998.© 1998 Wiley-Liss, Inc.

KEY WORDS: X-linked adrenoleukodystro-phy; incidence; phenotype;carrier testing

INTRODUCTION

X-linked adrenoleukodystrophy (ALD) is a variableentity. Phenotypes that have been identified includechildhood cerebral (CCALD), adolescent cerebral (Ado-CALD) and adult cerebral disease (ACALD), adreno-myeloneuropathy (AMN) and Addison’s disease only,without neurological changes [Moser et al., 1995]. Thecerebral forms of the disease are characterized by re-lentless white matter degeneration with dementia andultimately fatal neurological decline. Adrenomyelo-neuropathy is also a disease of white matter but affectsprimarily the spinal cord with progressive long tractsigns. Adrenal insufficiency is a variable component ofALD in all its forms and may antedate or follow theonset of neurological disease, or may be the only mani-festation of the condition in an individual.

Female carriers of ALD may also manifest symptomsof the disease, almost always of the adrenomyeloneu-ropathy type, and without adrenal insufficiency [Moseret al., 1995]. In addition, some males who express thebiochemical defect have no clinical abnormalities, al-though it seems likely that with time they will developmanifestations of the condition.

The gene, which is mutated in patients with ALD,was identified in 1993 and is located at Xq28 [Mosser etal., 1993; Sarde et al., 1994]. The protein for which itcodes, ALDP, is a peroxisomal ATP binding cassettetransporter. A number of mutations have been identi-fied, most of which are unique to the family in whichthey were identified [Feigenbaum et al., 1996; Braun etal., 1995; Ligtenberg et al., 1995; Kraseman et al.,1996.]. No genotype/phenotype correlation has beenpossible. This is not surprising, as it has been recog-nized for some time that affected individuals within afamily may have completely different phenotypes[Moser et al., 1984]. The reasons for this phenotypicvariability are not yet understood.

For the purposes of this paper, ‘‘ALD’’ will be used torefer to all of the clinical variants of the disease, con-sidered as a group. This does not include symptomaticfemale carriers unless explicitly stated.

Since 1981, the Peroxisomal Diseases Research Unit

Presented in part at the Australasian Society for Inborn Errorsof Metabolism Annual Scientific Meeting, Adelaide, Australia,September 1996, and at the 7th International Congress of InbornErrors of Metabolism, Vienna, May 1997.

Contract grant sponsor: Australian Leukodystrophy SupportGroup.

*Correspondence to: Dr. Janice Fletcher, Department ofChemical Pathology, Women’s and Children’s Hospital, 72 KingWilliam Rd., North Adelaide SA 5006, Australia. E-mail:jmfletch@medicine.adelaide.edu.au

Received 8 July 1997; Accepted 3 September 1997

American Journal of Medical Genetics 76:420–423 (1998)

© 1998 Wiley-Liss, Inc.

at the Women’s and Children’s Hospital, Adelaide, hasacted as the national referral laboratory for investiga-tion of individuals suspected of having ALD, and hasalso served New Zealand in the same way. Until early1996, ours was the only laboratory in Australasia thatprovided diagnostic testing for ALD, based on measure-ments of saturated very long chain fatty acids (VL-CFAs) in plasma and cultured fibroblasts and, latterly,on genetic linkage studies. Therefore, our experiencerepresents close to all of the known cases in Australa-sia in the last 15 years. We were particularly interestedin the relative frequency of the different phenotypes ofALD. With regard to female carriers of ALD, we wereinterested in the frequency of neurological disease andin the reliability of plasma VLCFAs as a method foridentifying carriers.

MATERIALS AND METHODS

The results of VLCFA measurement are usually ex-pressed as the ratio of C26:0/C22:0 and C24:0/C22:0,rather than as absolute values. VLCFAs were deter-mined as previously described [Berkovic et al., 1983].In 1993 the method was modified with the introductionof direct transesterification, eliminating the need forplasma lipid extraction [Poulos et al., 1994]. The diag-nostic indicators of C26:0/C22:0 and C24:0/C22:0 ratioswere not significantly altered by this modification.

Linkage studies were done using the DXS52 variablenumber of tandem repeat locus probed with St14 [Ri-chards et al., 1991] and with the restriction fragmentlength polymorphisms associated with the factor VIIIgene [Chan et al., 1988], as previously described.

We reviewed the clinical and laboratory findings inall Australasian families known to be affected by ALD.The source of the clinical information was the referralletters and subsequent correspondence from the clini-cians involved in the care of the families. In some in-stances, only limited clinical information was avail-able. Patients were classified, where possible, into oneof the phenotypic groups referred to in the Introductionusing clinical criteria described by Moser et al [1995].In some instances, it was not possible to clearly deter-mine the phenotype of patients who were shown to be

affected biochemically. Attempts were made to obtainfurther information from the clinicians involved butthere is still a small group of patients in whom a phe-notypic classification cannot be made with confidence.

RESULTS

Data were analyzed from 61 pedigrees from Austra-lia and New Zealand, each of which includes one ormore members affected by ALD. Information was avail-able on 815 relatives. Plasma VLCFAs were analyzedin 328 individuals from these families, and skin fibro-blast VLCFAs in 137 individuals. Linkage studies wereperformed in 123 individuals from 22 families.The dataincluded 104 affected males and 118 carriers, of whom8 were clinically affected to varying degrees.

Of the 104 affected males, 76 had their diagnosismade during the 15 years covered by this study, theremainder having been diagnosed prior to 1981 on thebasis of clinical and/or autopsy findings. Given a com-bined annual birthrate for Australia and New Zealandduring that period of approximately 310,000 (Austra-lian and New Zealand Bureaus of Statistics), we esti-mate a minimum annual incidence of ALD of 1.6 per100,000 live births.

The clinical spectrum of ALD seen in this study,taken as a whole, is similar to that seen in previouslypublished data from France [Aubourg and Chaussain,1991] and the United States [Moser et al., 1995] (TableI). However, when the pattern of referrals over the pe-riod of the study is examined, the distribution of phe-notypes appears to be changing (Table II).

Concordance of phenotype within families was vari-able. There were 29 families in which there were two ormore affected males whose phenotype could be deter-mined. In 13 of these families, all affected individualshad the same clinical presentation; 10 families had twoor more boys with childhood cerebral ALD, two had twoor more men with Addison disease only, and one hadtwo men with AMN. Six of the remaining 16 familieshad one or more members with a clear clinical diagno-sis and one child who was, as yet, too young to classify.There were 10 families in which a discordance of phe-notype was seen.

TABLE I. Phenotypes of Affected Males: Comparison of French, American, Dutch, andAustralasian Data

Phenotype

Aubourg andChaussin

[1991](n 4 185) (%)

Moser et al.[1995]

(n 4 1,475) (%)

van Geel et al.[1994]

(n 4 77) (%)This study

(n 4 96) (%)a

CCALD/AdoCALD 57 53 31 52ACALD 3 3 1 2AMN 28 25 46 25Addison disease only 8 10 14 16Asymptomatic 4 8 8 5

aNote that there were 104 affected patients identified in this study, but for 9 of them (who were testedand found to be affected on biochemical grounds) there was insufficient clinical information available toclassify them with confidence. Of these, 6 probably had childhood cerebral ALD, 2 probably had adultcerebral ALD or AMN with marked cerebral involvement, and 1 probably had Addison disease only. Ofthe patients with childhood cerebral or adolescent cerebral disease, only one had adolescent cerebraldisease. Of the ‘‘asymptomatic’’ patients, only one is an adult, a man aged 39 years. The other 4 patientsare children aged 12 years (a boy who is on ‘‘Lorenzo’s Oil’’ therapy), 6, 3, and 3 years. It is possible thatany of these could yet change to another classification, particularly the three younger boys.

ALD: The Australasian Experience 421

One hundred and eighteen female carriers of ALDwere identified, either by testing or because they wereobligate carriers. Of these, eight had neurologicalsymptoms which could be attributed to ALD. The re-ported manifestations ranged from mild difficultieswith walking to severe, disabling disease similar toAMN. None of the women who were clinically affectedhad adrenal insufficiency. There was no signficant dif-ference between the plasma C26:0/C22:0 and C24:0/C22:0 ratios in affected and unaffected carriers (two-tailed Student’s t-test, P > 0.5).

Of the 98 carriers who had plasma VLCFAs mea-sured, 93% had plasma VLCFA ratios higher than theupper limit of the reference range. The 7% of carrierswho had normal plasma VLCFAs were identified ascarriers on the basis of measurement of VLCFAs incultured skin fibroblasts, by linkage or because theywere obligate carriers.

Of the 118 carriers, 30 were identified on pedigreeanalysis as being obligate carriers. Of these, 20 hadplasma VLCFAs measured. Five obligate carriers hadnormal plasma VLCFAs. Of these five, three had notbeen tested further and two had had both linkage stud-ies and VLCFA measurement in cultured skin fibro-blasts done; all of these further studies confirmed car-rier status in these two women.

There were also two obligate carriers who had abnor-mal VLCFAs in plasma but normal VLCFAs in cul-tured skin fibroblasts. This raises the possibility thatthere may be some carriers who have normal VLCFAsin plasma and cultured fibroblasts.

Forty-five women who were tested and found to havenormal plasma VLCFAs have had no further investi-gations. Applying the figure of 93% carrier detectionusing plasma VLCFAs alone, it can be calculated thatapproximately three of these women are carriers buterroneously believe that they are not.

DISCUSSION

X-linked adrenoleukodystrophy is the commonest ofthe peroxisomal disorders; however, there are rela-tively limited data on its incidence. Moser et al. [1995]estimated an incidence of 1.1 per 100,000 using a simi-lar approach to that described here. Van Geel et al.[1994] estimated a minimum incidence of 1 per 100,000males, i.e., 0.5 per 100,000 births. The figure of 1 per20,000 has been quoted elsewhere [Mosser et al., 1993;Sarde et al., 1994; Jorge et al., 1994; Moser, 1993], butit is not clear how this figure was derived. The figure of1.6 per 100,000, which we derived, is similar to that ofMoser et al. [1995]; this is a minimum incidence, both

because ascertainment of cases is unlikely to be 100%and because the number of new cases diagnosed annu-ally appears to have increased over the last 15 years(Table II). If the 36 patients diagnosed in the last 6years represent close to 100% ascertainment, the trueincidence is probably 2 per 100,000, or 4 per 100,000males.

When the figures for the number of cases in eachdiagnostic group are considered together, the distribu-tion of phenotypes in Australasia appears strikinglysimilar to that in the United States [Moser et al., 1995]and France [Aubourg and Chaussain, 1991] (Table I).However, over time, the pattern of phenotypes amongnewly diagnosed patients has changed substantially,primarily due to an increase in diagnosis of the non-cerebral forms of the disease. It may be that the truepattern is more like that seen in The Netherlands. VanGeel et al. [1994] suggested that the difference betweentheir figures and the French and American data couldbe related to ease of access to medical services in TheNetherlands. While it is true that in Australia and NewZealand there is easy access for most of the populationto a high standard of health care, an alternative expla-nation may be that there is an increasing degree ofawareness among Australasian physicians of the non-cerebral forms of ALD. There may be real differencesbetween the phenotypes that occur in these four coun-tries. It is probable that some of the patients who arecurrently listed as ‘‘Addison disease only’’ will developneurological disease in the future; it is likely to takeseveral decades with close to full ascertainment beforea completely accurate picture of the distribution of phe-notypes can be obtained.

Recognition of ALD patients who have Addison dis-ease only and those who have AMN is important fortwo reasons. Firstly, ascertainment of families allowstracing of carriers and makes it possible to offer pre-natal diagnosis of affected males. Secondly, affectedmales who have yet to develop neurological diseasemay benefit from the use of glyceroltrioleate/glyceroltrierucate therapy (‘‘Lorenzo’s Oil’’) or frombone marrow transplantation [Korenke et al., 1995;Aubourg et al., 1990]. While ‘‘Lorenzo’s Oil’’ has beenshown not to benefit those who already have neurologi-cal disease [Poulos et al., 1994; Korenke et al., 1995;Aubourg et al., 1993; Rasmussen et al., 1994], it maydelay or prevent progression to cerebral disease inthose who have not yet progressed [Moser, 1993; Ko-renke et al., 1995]. A large international study is cur-rently under way to assess this possibility (Kohler etal., unpublished data).

Carrier detection is an important issue in families

TABLE II. Variation of Phenotypes Among Newly Diagnosed AustralasianPatients Over Time

PhenotypeBefore 1981(n 4 19) (%)

1981–1985(n 4 15) (%)

1986–1990(n 4 25) (%)

1991–1996(n 4 36) (%)

CCALD and AdoALD 100 67 48 22ACALD — — 8 —AMN — 33 24 36Addison disease only — — 12 34Asymptomatic — — 8 8

422 Kirk et al.

with ALD. While our data indicate that a very highdegree of accuracy is obtainable using biochemical test-ing in conjunction with linkage studies, mutationanalysis may supersede this approach. However, thereare no common mutations [Feigenbaum et al., 1996;Braun et al., 1995; Ligtenberg et al., 1995; Krasemanet al., 1996] and most mutations reported to date areunique to the families in which they are found. Whilethe current biochemical approach remains predomi-nant, it is important not to rely on the measurement ofplasma VLCFAs alone, as this detects only 93% of car-riers.

Counselling of women who are carriers of ALD re-mains very difficult. The uncertainty associated withthe variability of phenotype seen in this conditionmeans that the decision to terminate an affected preg-nancy may be more difficult than with other conditionswith more consistent phenotypes. The possibility of go-nadal mosaicism [Graham et al., 1992] is a furthercomplicating factor. The intra-familial variation seenin Australasian families is similar to that seen else-where, and the observation that the childhood cerebralform may be less prevalent than previously observedmay make the decision to terminate an affected fetuseven more difficult. A recent report [Korenke et al.,1996] of affected monozygotic twins who have discor-dant phenotypes suggests that even full understandingof the genetic influences on ALD phenotype may notallow confident prediction of the outcome in a particu-lar case. ALD will continue to be a challenge for thosecounselling affected families for some time to come.

ACKNOWLEDGMENTS

This work was supported in part by funding from theAustralian Leukodystrophy Support Group. We thankthe many clinicians who made the diagnosis of ALD intheir patients and provided the clinical information onwhich this study is based. We also thank Enzo Ranierifor assistance with statistical analysis.

REFERENCESAubourg P, Chaussain J-L (1991): Adrenoleukodystrophy presenting as

Addison’s disease in children and adults. Trends Endocrinol Metab2:49–52.

Aubourg P, Blanche S, Jambaque I, Rocchiccioli F, Kalifa G, Naud-Saudreau C (1990): Reversal of early neurologic and neuroradiologicmanifestations of X-linked adrenoleukodystrophy by bone marrowtransplantation. N Engl J Med 322:1860–1866.

Aubourg P, Adamsbaum C, Lavallard-Rousseau M-C, Rocchiccioli F,Cartier N, Jambaque I (1993): A two-year trial of oleic and erucic acids(‘‘Lorenzo’s Oil’’) as treatment for adrenomyeloneuropathy. N Engl JMed 329:745–752.

Berkovic SF, Zajac JD, Warbuton DJ, Merory JR, Fellenberg AJ, Poulos A,

Pollard AC (1983): Adrenomyeloneuropathy: Clinical and biochemicaldiagnosis. Aust NZ J Med 13:594–600.

Braun A, Ambach H, Kammerer S, Rolinski B, Stockler S, Rabl W, GartnerJ, Zierz S, Roscher AA (1995): Mutations in the gene for X-linked ad-renoleukodystrophy in patients with different clinical phenotypes. AmJ Hum Genet 56:854–861.

Chan V, Chan TK, Liu VWS, Wong ACK (1988): Restriction fragmentlength polymorphisms associated with factor VIII:C gene in Chinese.Hum Genet 79:128–131

Feigenbaum V, Lombard-Platet G, Guidoux S, Sarde C-O, Mandel J-L,Auborg P (1996): Mutational and protein analysis of patients and het-erozygous women with X-Linked adrenoleukodystrophy. Am J HumGenet 58:1135–1144.

Graham GE, MacLeod PM, Lillicrap DP, Bridge PJ (1992): Gonadal mo-saicism in a family with adrenoleukodystrophy: Molecular diagnosis ofcarrier status among daughters of a gonadal mosaic when direct detec-tion of the mutation is not possible. J Inherited Metab Dis 15:68–74.

Jorge P, Quelhas D, Oliviera P, Pinto R, Nogueira A (1994): X-linked ad-renoleukodystrophy in patients with idiopathic Addison disease. Eur JPediatr 153:594–597.

Korenke GC, Hunneman DH, Kohler J, Stockler S, Landmark K, HanefeldF (1995): Glyceroltrioleate/glyceroltrierucate therapy in 16 patientswith X-chromosomal adrenoleukodystrophy/adrenomyeloneuropathy:Effect on clinical, biochemical and neurophysiological parameters. EurJ Pediatr 154:64–70.

Korenke GC, Fuchs S, Krasemann E, Doerr HG, Wilichowski E, Hunner-man DH, Hanefeld F (1996): Cerebral adrenoleukodystrophy (ALD) inonly one of monozygotic twins with an identical ALD genotype. AnnNeurol 40:254–257.

Krasemann EW, Meier V, Korenke GC, Hunneman DH, Hanefeld F (1996):Identification of mutations in the ALD-gene of 20 families with adre-noleukodystrophy/adrenomyeloneuropathy. Hum Genet 97:194–197.

Ligtenberg MJL, Kemp S, Sarde C-O, Mandel JL, Chimini G (1995): Spec-trum of mutations in the gene encoding the adrenoleukodystrophy pro-tein. Am J Hum Genet 56:44–50.

Moser HW (1993): (Film review). Lorenzo’s Oil. Lancet 341:544.Moser HW, Moser AE, Singh I, O’Neill BP (1984): Adrenoleukodystrophy.

Survey of 303 cases: Biochemistry, diagnosis, and therapy. Ann Neurol16:628–641.

Moser HW, Smith KD, Moser AB (1995): X-Linked Adrenoleukodystrophy.In Scriver CR, Beaudet AL, Sly WS, Valle D (eds): ‘‘The Metabolic andMolecular Bases of Inherited Disease.’’ New York: McGraw-Hill, pp2325–2349.

Mosser J, Douar A-M, Sarde C-O, Kioschis P, Feil R, Moser HW, PoustkaA-M, Mandel J-L, Aubourg P (1993): Putative X-linked adrenoleuko-dystrophy gene shares unexpected homology with ABC transporters.Nature 361:726–730.

Poulos A, Gibson R, Sharp P, Beckman K, Grattan-Smith P (1994): Verylong chain fats in X-linked adrenoleukodystrophy brain after treatmentwith Lorenzo’s oil. Ann Neurol 36:741–746.

Rasmussen M, Moser AB, Borel J, Khangoora S , Moser HW (1994): Brain,liver and adipose tissue erucic and very long chain fatty acid levels inadrenoelukodystrophy patients treated with glyceryl trierucate and tri-oleate oils (Lorenzo’s Oil). Neuromed Res 19:1073–1082.

Richards B, Heilig R, Oberle I, Storjohann L, Horn GT (1991): Rapid PCRanalysis of the St14 (DXS52) VNTR. Nucl Acids Res 19:1944.

Sarde C-O, Mosser J, Kioschis P, Kretz C, Vicaire S, Auborg P, Poustka A,Mandel J-L (1994): Genomic organization of the adrenoleukodystrophygene. Genomics 22:13–20.

van Geel BM, Assies JA, Weverling GJ, Barth PG (1994): Predominance ofthe adrenomyeloneuropathy phenotype of X-linked adrenoleukodystro-phy in the Netherlands: A survey of 30 kindreds. Neurology 44:2343–2346.

ALD: The Australasian Experience 423