4 JMed Genet 1996;33:444-449

Recurrence of neonatal haemochromatosis inhalf sibs born of unaffected mothers

Alain Verloes, I Karen Temple, Anne-Frederique Hubert, Peter Hope, Stephen Gould,Christian Debauche, Gaston Verellen, Jean-Luc Deville, Lucien Koulischer,Etienne M Sokal

Centre for HumanGenetics,CHU Sart Tilman,(B) 4000 - Liege,BelgiumA VerloesL Koulischer

Wessex ClinicalGenetic Service,Princess AnneHospital,Southampton, UKI K Temple

Departments ofPaediatrics andHistopathology, JohnRadcliffe Hospital,Oxford, UKP HopeS Gould

Department ofPaediatricGastroenterology andNICU, CatholicUniversity of Louvain,St Luc Hospital,Brussels, BelgiumA F HubertC DebaucheG VerellenE M Sokal

Department ofObstetrics andGynaecology, Pelzer-Tourelle RegionalHospital, Verviers,BelgiumJ L Deville

Correspondence to:Dr Verloes.

Received 7 August 1995Revised versionaccepted for publication12 February 1996

AbstractWe report two families in which neonatalhaemochromatosis was observed in halfsibs. In the first family, two successivegirls were born of different fathers. In thesecond family, an affected brother andsister were followed by an affected halfbrother born after donor insemination.These observations, as well as a previousabstract describing two affected half sis-ters, revive the debate over the inheritanceof neonatal haemochromatosis. Incom-plete penetrance or gonadal mosaicism fora dominant disorder, a maternal "en-vironmental factor", or mitochondrial de-fect may be more suitable explanationsthan autosomal recessive inheritance inthis condition. Alternative modes of fer-tilisation, such as donor insemination or

in vitro fertilisation with donor eggs,

should be considered with caution.

(JtMed Genet 1996;33:444-449)

Key words: neonatal haemochromatosis; half sibs; ma-ternal inheritance.

Neonatal haemochromatosis (NH) is a poly-visceral iron storage disorder of prenatal onset.It is characterised by a rapidly progressive hep-atic insufficiency with prenatal or perinatalonset and a specific distribution of iron over-

load similar to that seen in adult chromosome6 linked haemochromatosis, in the absence ofany known cause of prenatal liver disease.`'5Serum iron, ferritin, and iron binding capacitysaturation are usually moderately raised,whereas total transferrin is low. Liver histologydiscloses striking haemosiderin deposition inhepatocytes and significantly less in Kupffercells and biliary epithelium. Liver damage maybe widespread with diffuse fibrosis, oftenamounting to frank cirrhosis, together withregenerative changes, ductular proliferation,and cholestasis. Multinucleated hepatocytesare commonly observed, although inflam-matory infiltrate is not striking. Other commonsites of iron storage include adrenal cortex,endocrine and exocrine pancreas (together withLangerhans islet hyperplasia), and the epithelialcells ofrenal tubules, thyroid follicles, and mostexocrine glands, but not the reticuloendothelialsystem (Kupffer cells, spleen, lymph nodes,bone marrow). Total hepatic iron is not alwayssignificantly greater than in control groups6although most cases do show an absolute over-

load4 that can now be quantified in vivo byMRI or CT scanning.7

NH is thus established as a specific patho-logical diagnosis (that is, a phenotype) but itsgenetic or environmental (viral or toxic) basesare still unknown, no specific metabolic dis-order is recognised, and aetiological hetero-geneity is suspected. NH is usually consideredas an autosomal recessive disorder (MIM231 100). We report here on two sibships wheretypical neonatal haemochromatosis was ob-served in half sibs, indicating that at least insome instances NH should not be consideredas a recessive disorder.

Case reportsFAMILY 1Patient 1This female child was born to non-con-sanguineous, healthy, Belgian parents. An oldersister was healthy. The pregnancy was un-eventful and birth weight was 3500 g. From 4hours of life, the child presented with diffusemucosal bleeding and progressive jaundice. Sherapidly deteriorated, with signs of liver in-sufficiency. On admission to a paediatric neo-natal intensive care unit (NICU) on day 12,the child was deeply jaundiced, with ascitesand a swollen abdomen. The liver and spleencould not be palpated. Bleeding diathesis waspatent. Laboratory investigations showed: totalbilirubin 243 iimol/l (normal < 17), direct bili-rubin 92 iimol/l (normal <17), aspartate trans-aminase (AST) 52 IU/l (normal <25), alanineaminotransferase (ALT) 30 IU/l (normal <32),blood ammonium 95 gmol/l (normal: 18 to59), prothrombin time 10% (normal >65%),blood glucose 0-9 mmol/l (normal >3 3). Des-pite intensive management, the child died onday 13 from diffuse uncontrolled cutaneousand mucous bleeding. Liver histology at nec-ropsy showed micronodular cirrhosis with pos-itive iron staining (Perl's coloration). Ironoverload was mainly observed in hepatocytesand giant hepatocytes.

Patient 2Two years after the death of patient 1, themother sought genetic advice for a new preg-nancy. She was divorced from the father ofpatient 1. It was clearly stated by the motherthat her new partner was not the father ofpatient 1 and that she had no more contact withher former husband. In view of this situation, avery low recurrence risk was given, assumingthat the disease of patient 1 was probably auto-somal recessive. Patient 2, a girl, was born at39 weeks of gestation with a birth weight of

444

on April 26, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.33.6.444 on 1 June 1996. Dow

nloaded from

Recurrence of neonatal haemochromatosis in half sibs born of unaffected mothers

3070 g. From the sixth hour of life, she de-veloped multiple petechiae and mucosal bleed-ing. It soon appeared that she also had severeliver failure. On admission on day 1, she wasjaundiced, with a distended abdomen, a hardenlarged liver 1-5cm below the right costalmargin, no splenomegaly, and multiple pe-techiae.

Biological investigations on day 1 showed:total bilirubin 102 [tmol/l, direct bilirubin7,mol/l, AST 84 IU/l, ALT 16 IU/l, pro-thrombin time 15%, blood glucose 2 mmol/l,ammonium 41,mol/l, lactic acid 4-4 mmol/l.Total bilirubin increased to 530,mol/l, directbilirubin to 148 gmol/l, AST to 237 IU/l, ALT69 IU/l, ammonium 171 jmol/l, and pro-thrombin time decreased to <10% during thefollowing days. As for patient 1, screening forneonatal cholestasis and early liver failure wasperformed using appropriate investigations andwas not contributory. Tyrosinosis was ex-cluded. Serum iron concentration on day 2was 62 jmol/l (normal 14 to 30), serum ironbinding capacity 28 gmol/l (normal 47 to 68-5),serum ferritin 425 gg/l (normal 10 to 300),transferrin 4 g/l (normal 2 2 to 4), and ot feto-protein 2 jg/l (normal 0 01 to 0-13). Liverultrasound showed increased echogenicity andnodular transformation. Liver biopsy was notperformed because of the clotting defect.A diagnosis of NH was considered and the

child was treated with desferrioxamine (Des-ferralO) by a continuous subcutaneous infusionof30 mg/day. Large amounts ofiron were foundin the urine: from 8 gmol/l before treatment to326,mol/l under treatment. However, serumiron increased to 228 itmol/l and liver in-sufficiency persisted. She had disseminatedintravascular coagulation with fibrin de-gradation products at 160 gg/ml (normal <20).In addition to standard treatment for liver fail-ure, she was given continuous infusion factorsII, V, VII, IX, and fibrinogen at a dosageof 50 IU/kg/day without improvement. Deathresulted from massive intracranial haemorrhageand multivisceral failure on day 5.Postmortem liver biopsy showed micro-

nodular cirrhosis, mild ductular proliferationand portal inflammation, extramedullaryhaematopoiesis, and multinucleated hepa-tocytes. Iron overload was obvious and wasconcentrated in hepatocytes, without iron ac-cumulation in Kupffer cells. Iron overload wasfound at necropsy in the pancreas, stomach,adrenal glands, kidneys, thyroid, heart, andsalivary glands.

AETIOLOGICAL INVESTIGATIONSAs recommended elsewhere,8 appropriateblood and urine tests were performed beforedeath in both children. Combined with post-mortem pathological examination, tyro-sinaemia, galactosaemia, hereditary fructoseintolerance, organic acidaemias, Niemann-Picktype C, glycogenosis I, III, and IV, ocl anti-trypsin deficiency, peroxisomal disorders, prim-ary defects of bile acid synthesis, cystic fibrosis,Byler disease, Wolman disease, and neonatalinfections were excluded.

The activities of the five mitochondrial res-piratory chain complexes were assayed forpatient 2. No anomalies were observed either inmuscle or liver samples (Professor A Munnich,Paris). Fumarylacetoacetate lyase activity wasnormal in fibroblasts of patient 2.

FAMILY HISTORYThe mother was aged 22 at the birth of patient1. Her liver function and basal iron metabolism,assessed six months after the birth of patient 2,showed slightly raised iron (2-04 mg/l, normalrange 0 6-1 5) and transferrin (4- 12 g/l, normalrange from 2-4-3- 8 1). On CT scan, attenuationof the liver was normal and the differencefrom splenic attenuation was not significant.Hepatic, renal, splenic, and adrenal iron stor-age, estimated through abdominal MRI scan,by TI, T2, and gradient echo sequences, wasin the normal range. Striated muscle-liver signalratio was higher than 1-6 in T2 and between1-3 and 1-5 in Ti. Iron load was estimatedas <50 tmol/kg. Chronic viral infection andautoimmune disorders were excluded.Because of lack of proper material from

patient 1, paternity testing could not be per-formed by DNA analysis. The parents werevery concerned by the implications of maternalinheritance and accepted all investigations. Wehad the opportunity to discuss genetic aspectsseveral times with the mother, alone or withher husband. The possibility of a single fatherwas categorically excluded and, consideringtheir involvement and distress, the author hasno doubt about that statement.

FAMILY 2Patient 1This boy was born at 39 weeks of gestationafter a normal pregnancy. Birth weight was2877 g, OFC 33-5 cm. He was noted to havespontaneously resolving hypoglycaemia on day1. Jaundice appeared within 24 hours andsevere conjugated hyperbilirubinaemia withsigns of liver dysfunction was obvious by day 7.The liver remained non-palpable. Progressivehepatic failure and convulsive encephalopathyled to death at 5 weeks. Necropsy showedwidespread liver fibrosis and biliary ductularproliferation, the latter containing bile in manyplaces. Large amounts ofstainable iron pigmentwere present predominantly in the hepatocyteswith the macrophages being less affected. Amild chronic lymphomonocytic inflammatoryinfiltrate was seen in the fibrous tissue. Con-siderable iron was also shown in the myo-cardium and a few renal tubular cells. Neitherthyroid, adrenal, or other tissues examinedshowed iron. The pancreas was not availablefor examination. There was no other significantpathology except for severe brain swelling.

Patient 2This girl, the sib of patient 1, was born bynormal delivery with a birth weight of 1984 g.Intrauterine growth retardation had been diag-nosed late in the pregnancy. She died on day

445

on April 26, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.33.6.444 on 1 June 1996. Dow

nloaded from

Verloes, Temple, Hubert, Hope, Gould, Debauche, Verellen, Deville, Koulischer, Sokal

2 of overwhelming liver failure. Blood takenwithin the first day of life showed a serum ironwithin the normal range and a serum ferritinof 1270 mg/l (normal range up to 950 mg/l). Aliver biopsy showed an identical picture to thatofher brother with severe destruction ofhepaticarchitecture and widespread fibrosis and lossof hepatocytes. There was extensive biliarytransformation and hepatocytes containedmuch stored iron. Necropsy was refused.

Patient 3Following genetic counselling, the couple wasgiven a 1 in 4 risk of recurrence based onprobable autosomal recessive inheritance. Theparents decided to opt for artificial in-semination by an anonymous donor (AID). Amale infant with a birth weight of 2536 g wasborn after a normal pregnancy but died offulminating liver disease on day 2. A liver biopsyshowed an identical picture to that of his twohalf sibs. Necropsy was not performed. Pa-ternity by donor was confirmed by a five multi-allelic loci DNA probe analysis.

AETIOLOGICAL INVESTIGATIONSMultiple investigations were undertaken onpatients 1 and 2. Liver function became grosslydisordered in patient 1 over the first sevendays. There was raised aspartate transaminase(654 IU/l on day 1 and 1255 IU/l by day 8).Alpha fetoprotein levels were grossly raised inpatient 1 (>44 800IU/1). There was no evi-dence of haemolytic anaemia and no erythro-phagocytosis on bone marrow biopsy. A sweattest and immunoglobulin estimation werenormal. Metabolic screening excluded tyro-sinaemia, galactosaemia, glutamic aciduria type2, and Niemann-Pick disease type C. No fattyoxidation defects or inborn errors of bile acidswere discovered, excluding most peroxisomaldisorders. There was no evidence of a storagedisorder and bone marrow and spleen speci-mens from the first child were normal. Mito-chondrial DNA investigations showed no majordeletions or any of the common mutations(MELAS, MERRF, and NARP). Viral studiesincluding hepatitis B, CMV, rubella, and Cox-sackie serology were normal. There was noevidence of toxoplasmosis or mycoplasma in-fection.

FAMILY HISTORYThe parents were healthy, non-consanguin-eous, and British, aged 29 at the birth ofpatient1. Clinical examination was normal. Basal ironmetabolism in the parents (serum iron, ironbinding capacity, and ferritin) was normal otherthan low serum iron levels in the mother(9-5 gmol/l, normal range 14-31 ,umol/l). Liverfunction tests in both parents were normalother than minimally raised bilirubin levelsin the father (25 itmol/l, normal range3-17 ,umol/l). Basic metabolic screening (ureaand creatinine, electrolytes, amino acids,caeruloplasmin) was all normal as were thekaryotypes of both parents. Chronic immune/

infectious disorder was excluded in the mother,who in addition had normal glycosylation oftransferrin.Although its levels were normal, in the

mother (1 2 g/l, normal range 1-2-2-6) andthe father (1 * 3 g/l), a 1 antitrypsin phenotypingindicated that the mother was heterozygous fora new mutation which she had inherited fromher father. The mutant protein was transmittedto patients 2 and 3, but patient 1 had a normalPi M phenotype. The significance of this rareallele is not yet known (D Whitehouse, personalcommunication).

DiscussionA diagnosis of neonatal haemochromatosis wasestablished in the five children reported hereon histopathological and clinical grounds. Atleast 65 patients with NH have been observedand exhaustive reviews are available.459 Themost important features are a high prevalenceof prematurity (median gestational age 36weeks, 55% <37 weeks), IUGR (median birthweight 2300 g, 60% <2500 g), early onset (88%during the first day), and an usually fulminatingcourse of the liver dysfunction (median survivalseven days, 80% dead <40 days using datafrom reference 5). Some typical cases of NHwere diagnosed beyond the perinatal period.5Longer survival was recently reported in earlyonset cases owing to more effective intensivecare or desferrioxamine chelation,'o l1 althoughthe only curative treatment remains orthotopicliver transplantation, successfully performed insome children since 1992.7 12 Some sibs of typ-ical cases have been shown to have a mildercourse and to survive without graft.' 12-14 Thesecases were nevertheless perinatally symptom-atic. 12 13 In a 16 month old surviving girl, controlliver biopsy showed spontaneous disappearanceof iron overload, and hepatic fibrosis.'3Up to now, no consensus has been reached

as to whether the iron storage is the primarydefect or a phenomenon secondary to a prim-itive liver disease, and whether it is causallyheterogeneous. Several physiopathologicalmechanisms have been proposed for NH, in-cluding intrauterine infection with an unknownpathogen, abnormal transplacental iron trans-fer from mother to fetus (possibly related to anabnormal iron metabolism in the mother), oran inherited or acquired disorder affecting thehepatic binding, the intracellular handling, orthe control ofiron distribution.6 Iron does notreaccumulate in transplanted liver,7 suggestingthat the anomaly of iron metabolism is notbecause of a multisystemic defect. The fetalliver plays a crucial role in the pathogenesis,either as the primitive pathological site, or asthe major target of a transplacental transferanomaly. This differs from adult onset, chro-mosome 6 linked haemochromatosis, in whichiron reaccumulates in the allograft.Some cases ofNH appear to differ from the

usual phenotype or show unusual biochemicalfeatures which illustrate the heterogeneity ofNH. For example, in a familial observation,the liver was the only site of iron storage.'5

446

on April 26, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.33.6.444 on 1 June 1996. Dow

nloaded from

Recurrence of neonatal haemochromatosis in half sibs born of unaffected mothers

Reference Sibship Comment

0*0

O*OOOOE----

*.0

OE -

00EE

00 - *"M

0@-

*U-U

Om-@

Om-

*-

00--*O-

Abstract (incomplete pedigrees)

One case surviving at 16 months

3 with histologically proven NH, one with "severe liver disease"(iron not investigated), two babies dead at day 2First affected case died in utero at 34 weeks: no iron depositionin extrahepatic tissue

First child possibly affected (died day 3 with generalisedhaemorrhage)Abstract (incomplete pedigrees)

Two maternal half sibs in good healthOlder affected sib: successful liver transplantationYounger affected sib alive and stable aged 6 months

Mother with post-transfusional non-A, non-B chronic hepatitis.Histologically proven giant cell hepatitis without iron overload ina 25 week fetus

Two half sibships. Second boy survived. Abnormal ferritinTwo unaffected sibs died perinatally: one "dysmorphic", the otherwith polycystic renal diseaseAbnormal iron metabolism in parentsOne unaffected anencephalic boyGirl: successful liver transplantation. Complete clearing of ironoverload in exocrine glands 5 months later

Incomplete pedigrees. No family history available

Last affected child in each family is maternal half sib of theprevious case(s)

|OK0 normal, E@* affected, E/®/* unaffected malformed boy/girl/child of unknown sex, . miscarriage.

Familial cases of tpical NH.

Deficiency of A4-3-oxosteroid 5-p-reductasewas postulated in two unrelated patients'6 whopresented with late onset liver insufficiency (1and 6 weeks). Paucity of primary bile acidsand a predominance of some derivatives ofcholenoic acid were noted in the serum.Whether these represent a primary defect, oran anomaly secondary to the liver disease is stillunknown. In the family described byjacknow etal,4 the mother and an affected surviving childwere shown to have low serum transferrin andabsent ferritin, so that a dominant mutation offerritin with abnormal iron affinity and struc-ture facilitating hepatic uptake was suggestedto be causative. A single newborn was reportedwith NH, facial dysmorphism, complex cy-anotic heart disease, syndactyly, and postaxialpolydactyly." We have observed two sibs withdelayed NH, intractable diarrhoea, facial dys-morphism, and hair anomalies.17ANH is usually considered to be a recessive

disorder. Recurrence of NH in sibships hasbeen reported in at least 24 sibships, includingours 12711-14 18-28 However, in familial reports,

parental consanguinity was never mentioned.The total number of affected children is 57. Inaddition, although complete pedigrees wereprobably not always published, the frequencyof fetal losses is striking (figure): among 18sibships where familial data were available,there were 18 probands, 27 affected sibs, 19unaffected sibs, and 10 miscarriages. In anabstract, Jacknow et al'4 was the first to describeNH in half sibs born to the same mother. Agirl died at birth, a male sib, although affectedearly, was alive at the age of 5 years, and amale half sib died on day 4. Another caseoccurring in half sibs is briefly mentioned byKnisely,4 without clinical data. Although noformal paternity testing was performed in fam-ily 1, it was verified in family 2. Consideringthe rarity of NH, recurrence in maternal halfsibs in these two further families requires a lessunlikely explanation than fortuitous recurrentmatings between heterozygotes.The lack of consanguinity in familial reports

of this exceptional disease may also indicatethat autosomal recessive inheritance is not the

18

13

19

20

21

22

23

24

25

12

2

14

11

26

27

7

28

This paper

447

on April 26, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.33.6.444 on 1 June 1996. Dow

nloaded from

Verloes, Temple, Hubert, Hope, Gould, Debauche, Verellen, Deville, Koulischer, Sokal

correct pattern for at least a subgroup of NH.A maternal factor has been suggested by some.In the family of Jacknow et all4 the motherand the surviving child had low serum iron,increased serum transferrin, and undetectableferritin. The authors concluded that a mutantferritin, transmitted as a dominant trait, wasresponsible for the hepatic iron sequestration(although an abnormality of the ferritin meta-bolism rather than a primary defect of ferritinwas also possible). Driscoll et al" hypothesisedthat NH could result from increased trans-placental transport of iron facilitated by ma-ternal heterozygosity and fetal homozygosityfor a recessive gene. Others have suspecteda purely maternal disorder, either genetic oracquired. Rand et al7 suggested the presenceof maternal isoimmune antibodies directedagainst a placental protein regulating the ironflux. A sporadic case was suspected to be relatedto the presence of anti-Ro/SS-A and anti-La/SS-B ribonucleoprotein antibodies in themother who had Sjogren syndrome."9 Becauseof the hepatitic aspect of the liver, an unknownviral infection has been suspected by someauthors. No direct proof has been obtained forthis, although some cases of NH have beenobserved in children whose mother had non-A, non-B post-transfusional hepatitis.'"' Asnormal children were born after the affectedchild in some pedigrees, a persistent immuneor infectious factor is unlikely. Absence of re-currence in the allograft and correction ofserum markers of iron storage disease arguesagainst the viral hypothesis.Mitochondriopathy has recently been ob-

served in sibs with neonatal cholestasis andearly liver insufficiency, without NH.30 As nopaternal half sibs have been observed, a mito-chondriopathy could also explain the unusualpattern of familial occurrence of NH. In-vestigations of the respiratory chain and mito-chondrial DNA screening in our patients failedto confirm this hypothesis. Moreover, there isno report of iron storage in other clinical vari-ants of mitochondriopathies (Kearn-Sayres,Pearson, Wolfram syndromes, etc), and no re-ports mention compatible manifestations inparents.Dominant inheritance with extreme clinical

variation seems the more convincing modelbut no constant biological anomalies of ironmetabolism have been shown in first degreerelatives of NH patients. Abnormal liver func-tion was noted in some parents.9 This has beenconsidered by some as a marker of het-erozygosity," '" but could reflect minimal ex-

pression of a dominant trait as well. In two

instances, liver biopsies were taken from moth-ers. No iron overload was found in thespecimens.""9'4 In most cases, the only ab-normality is a maternal iron depletion (lowserum iron and low saturation) that correctsitself after delivery and is thought to reflect thefetal steal of iron." Gonadal mosaicism, an

imprinted gene with maternal expression, or

cryptic familial translocation are other possiblevariations on the dominant model. Con-firmation of this hypothesis may be difficult to

obtain, as most known surviving cases (either

because of a milder phenotype or after hepaticallograft) are children. Finally, a maternallydependent disorder of fetoplacental iron meta-bolism could be considered.

In summary, a review of familial cases ofNH, and observations of three sets of affectedhalf sibs indicate that at least a subgroup ofNH patients cannot be considered as sufferingfrom a recessive disorder. As causal hetero-geneity is still likely, systematic screening of theparents should be recommended for abnormaliron metabolism and evaluation ofparental ironstorage with MRI. Mitochondrial anomaliesshould still be investigated in all cases. Geneticcounselling of parents and of surviving adultshas to take into account the possibility of ger-minal mosaicism and dominant inheritance,although exact risk figures cannot be given atthis moment. Alternative modes of fertilisation,such as sperm donor or oocyte gift, should beconsidered with caution. Prenatal diagnosis ispossibly feasible through fetal MRI scan ofliver, pancreas, and heart," and eventually per-cutaneous blood cord sampling."

The authors wish to thank Dr S Gosseye for the pathologicalinvestigation of family 1, Dr S Huson for her help with family2, and Dr D Whitehouse for his useful comments on themanuscript.

1 Witzleben CL, Uri A. Perinatal hemochromatosis: entity orend result? Hum Pathol 1989;20:335-40.

2 Hoogstraeten J, de Sa DJ, Knisely AS. Fetal liver diseasemay precede extrahepatic siderosis in neonatal hemo-chromatosis. Gastroenterology 1990;98:1699-701.

3 Silver MM, Cave CT, Kiplani H. Perinatal hemo-chromatosis. Pediatr Pathol 1989;9:203-10.

4 Knisely AS. Neonatal hemochromatosis. Adv Pediatr 1992;39:383-404.

5 Barnard JA III, Manci E. Idiopathic neonatal iron storagedisease. Gastroenterology 1991 ;101: 1420-7.

6 Silver MM, Valberg LS, Cutz E, Lines LD, Phillips MJ.Hepatic morphology and iron quantitation in perinatalhemochromatosis. Am J Pathol 1993;143:1312-25.

7 Rand EB, McClenathan DT, Whitington PF. Neonatalhemochromatosis: report of a successful orthotopic livertransplantation. J Pediatr Gastroenterol Nutr 1992;15:325-9.

8 Burdelski M. Neonatal cholestasis. In: Buts JP, Sokal EM,eds. Management of digestive and liver disorders in infantsand children. Amsterdam: Elsevier, 1994.

9 Knisely AS, Magid MS, Dische MR, Cutz E. Neonatalhemochromatosis. Birth Defects 1987;23(1):75-102.

10 Knisely AS, Grady RW, Kramer EE, Jones RL. Cytoferrin,maternofetal iron transport, and neonatal hemo-chromatosis. Am J Clin Pathol 1989;92:755-9.

11 Jonas MM, Kaweblum YA, Fojaco R. Neonatal hemo-chromatosis: failure of desferoxiamine therapy. J. PediatrGastroenterol Nutr 1987;6:984-8.

12 Hayes AM, Jaramillo D, Levy HL, Knisely AS. Neonatalhemochromatosis: diagnosis with MT imaging. A3tR 1992;159:623-5.

13 Colletti RB, Clemmons JJW. Familial neonatal hemo-chromatosis with survival. J Pediatr Gastroenterol Nutr1988;7:39-45.

14 Jacknow G, Johnson D, Freese D, Smith C, Burk B. Idio-pathic neonatal iron storage disease. Lab Invest 1983;48:7P.

15 Nezelof C, Watchi JM, Gauter P, Brault A. Cirrhose infantilepigmentaire familiale. Presse Med 1967;75:451-6.

16 Shneider BL, Setchell KD, Whitington PF, Neilson KA,Suchy FJ. Delta 4-3-oxosteroid 5 beta-reductase de-ficiency causing neonatal liver failure and hemo-chromatosis. J Pediatr 1994;124:234-8.

17 Castillo Taucher S, Bentjerodt R, Hbner ME, Nazer J.Multiple malformations in neonatal hemochromatosis.Am3rMed Genet 1994;50:213-14.

17AVerloes A, Lombet J, Lambert Y, et al. Tricho-hepato-enteric syndrome: further delineation of a distinct syn-drome with neonatal hemochromatosis phenotype, in-tractable diarrhea and hair anomalies. Am J Med Genet(in press).

18 Becich MJ, Rothbaum RJ, Keating JP. Non-siderosomeassociated iron in patients with neonatal iron storagedisease: a computer-assisted quantitative morphometricevaluation. Lab Invest 1991;64:107A.

19 Cottier H. Ueber ein der Hamochromatose vergleichbaresKrankheitsbild bei Neugeborenen. Schweiz Med Woch-enschr 1957;37:39-43.

448

on April 26, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.33.6.444 on 1 June 1996. Dow

nloaded from

Recurrence of neonatal haemochromatosis in half sibs born of unaffected mothers

20 Dalhoj J, Kiaer H, Wiggers P, et al. Iron storage disease inparents and sibs ofinfants with neonatal hemochromatosis:30 year follow-up. Am Jf Med Genet 1990;37:342-5.

21 Boissieu D, Checoury A, Barbet P, et al. Hemochromatoseperinatale. Arch Fr Pediatr 1990;47:23-8.

22 Dible JH, Hunt WE, Pugh VW, Steingold L, Wood JHF.Foetal and neonatal hepatitis and its sequelae. PatholBact 1954;67: 195-206.

23 Driscoll SC, Hayes AM, Levy HL. Neonatal hemo-chromatosis: evidence for autosomal recessive trans-mission. Am Hum Genet 1988;43:A232.

24 Ehrlich JC, Ratner IM. Congenital cirrhosis of the liverwith kernicterus. Report of two cases in siblings witha discussion of the so-called neonatal hepatitis to iso-immunization disease. Am Pathol 1955;31:1013-47.

25 Fienberg R. Perinatal idiopathic hemochromatosis: giantcell hepatitis interpreted as an inborn error of metabolism.Clin Pathol 1960;33:480-91.

26 Knisely AS, Harford JB, Klausner RB, et al. Neonatalhemochromatosis: the regulation of transferrin receptorand ferritin synthesis by iron in cultured fibroblastic-linecells. Am Jf Pathol 1989;134:439-45.

27 Laurendeau T, Hill JE, Manning GB. Idiopathic neonatalhemochromatosis in siblings: an inbom error of meta-bolism. Arch Pathol 1961;72:410-23.

28 Silver MM, Beverley DW, Valberg LS, et al. Perinatal hemo-chromatosis: clinical, morphologic, and quantitative ironstudies. Am Jf Pathol 1987;128:538-54.

29 Schoenlebe J, Buyon JP, Zitelli BJ, et al. Neonatal hemo-chromatosis associated with maternal autoantibodiesagainst Ro/SS-A and La/SS-B ribonucleoproteins. Am J7Dis Child 1993;147:1072-5.

30 Goncalves I, Hermans D, Chretien D, et al. Mitochondrialrespiratory chain defect: a new etiology for neonatalcholestasis and early liver insufficiency. Jf Hepatol 1995;23:290-4.

449

on April 26, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.33.6.444 on 1 June 1996. Dow

nloaded from