Expression of aromatase P450(AROM) in the human fetal and early postnatal cerebral cortex.

Aromatase (P450(AROM)), the enzyme responsible for the conversion of testosterone (T) into 17-ß estradiol (E(2)), plays a crucial role in the sexual differentiation of specific hypothalamic nuclei. Moreover, recent findings indicate that local E(2) synthesis has an impact on other brain areas including hippocampus, temporal cortex and cerebellum, and may thus influence also cognitive functions. Numerous studies have described the expression and the distribution of P450(AROM) throughout ontogenesis and postnatal development of the central nervous system in several mammals, but data referring to humans are scarce. In the adult human brain, P450(AROM) has been detected in the hypothalamus, limbic areas, and in the basal forebrain, and described in glial cells of the cerebral cortex and hippocampus. In this study we report the expression, distribution and cellular localization of P450(AROM) in the human fetal and early postnatal cerebral cortex. In our series of fetal brains of the second trimester, P450(AROM) expression appeared at gestational week (GW) 17 and resulted limited to groups of cells localized close to the growing neuroepithelium in the ventricular and subventricular zones. At GWs 20-24, scattered P450(AROM) immunoreactive (-ir) neural cells were identified in the intermediate plate and subplate, and in the parietal cortical plate. In perinatal and early postnatal individuals the quantity of P450(AROM)-ir elements increased, and revealed the morphology typical of glial cells. Double labeling immunostaining with anti-GFAP and anti-P450(AROM) antisera, and subsequent confocal analysis, confirmed this observation. Our data show that the expression of P450(AROM) in the fetal cortex starts approx at the end of the fourth gestational month, but increases steadily only in the last trimester or in the early postnatal period. This temporal trend may suggest that P450(AROM) could act as a differentiation-promoting factor, based on timing of the steroid actions.

Investigation of the transport of trace elements across barriers in humans: studies of placental and mammary transfer.

UNLABELLED: The barrier function of the human mammary gland for selected trace elements was evaluated by analysing 27 maternal sera and corresponding colostrum samples for 14 trace elements. To investigate the impact of the human placenta at the end of gestation on the transfer of 17 trace elements from the mother to the baby, 29 maternal and corresponding umbilical cord sera (UCS) were studied. The uptake of trace elements from the UCS by the fetus was investigated in nine pairs of arterial and venous UCS. In colostrum, the concentration of Cu was 19%, of Se 47%, of Co 80%, of Mg 146%, of Ca 222%, of Sn 228%, of Mn 275%, of Mo 814%, and of Zn 1470% of that of the maternal sera. For Cd and Pb the corresponding values were 200% and 325%, respectively. These data show that the mammary gland can exert an activating as well as an inhibiting effect on the trace element transfer. A concentration gradient mode of action for the transfer of Li, Mo and Sr could be found. In UCS the concentration of Cu was 20%, of Se 55%, of Co 60%, of Sn 85%, of Mo 100%, of Mg 105%, of Ca 120%, of Zn 148%, and of Mn 150% of that of the maternal sera. The corresponding values for Cd were 66% and for Pb 50%. CONCLUSIONS: These findings indicate that the placenta can exhibit an activation or inhibition on transfer as well as a gradient mode of action as for Mo, Cs, Li and Sr. The uptake of essential trace elements from venous UCS by the unborn ranged from 2.5% for Ca to 16.7% for Mo. Both the placenta and the mammary gland can exert an activating, inhibiting or gradient mode of action for selected trace elements, the biological impact of which needs to be further elucidated.

Determination of selenium in serum by FI-HG-AAS and calculation of dietary intake.

A method was developed for the determination of selenium concentration in serum by flow injection-hydride generation-atomic absorption spectrometry (FI-HG-AAS) following microwave digestion of serum samples and reduction of selenate to selenite. The detection limit of the method was 0.3 microg Se/L and the characteristic concentration, corresponding to the 0.0044 absorbance signal, was 0.12 microg Se/L. The results from the analysis of two Seronorm standard reference materials showed good agreement with the certified values. The method was then used to analyze selenium in sera of Austrian and Slovenian people for the calculation of dietary intakes. The selenium concentrations in sera of mothers at delivery, their neonates, and the male and female adults were 71+/-14, 42+/-6, 75+/-21, and 65+/-16 microg/L for the Austrians and 62+/-15, 34+/-7, 70+/-12, and 66+/-15 microg/L for the Slovenians. The dietary intakes of selenium of the mothers and the male and the female adults were calculated as 52, 37, and 46 microg/d for the Austrians and 45, 38, and 32 microg/d for the Slovenians.

Concentrations of selected trace elements in human milk and in infant formulas determined by magnetic sector field inductively coupled plasma-mass spectrometry.

Magnetic sector field inductively coupled plasma-mass spectrometry (ICP-MS) was applied to the reliable determination of the 8 essential trace elements cobalt (Co), chromium (Cr), copper (Cu), iron (Fe), manganese (Mn), nickel (Ni), selenium (Se), and vanadium (V) as well as the 7 nonessential and toxic elements silver (Ag), aluminum (Al), arsenic (As), gold (Au), platinum (Pt), scandium (Sc), and titanum (Ti) in 27 transitory and mature human milk samples and in 4 selected infant formulas. This advanced instrumentation can separate spectral overlaps from the analyte signal hampering significantly the determination of many trace elements by conventional ICP-MS. Moreover, superior detection limits in the picogram per liter range can be obtained with such magnetic sector field instruments. Therefore, this is the first study to report, the concentrations of the elements Ag, Au, Pt, Sc, Ti, and V in human milk and in infant formulas. Concentrations of Ag (median: 0.41 microg/L; range: < 0.13-42 microg/L) and Au (median: 0.29 microg/L; range 0.10-2.06 microg/L) showed large variations in human milk that might be associated with dental fillings and jewelry. Pt concentrations were very low with most of the samples below the method detection limit of 0.01 microg/L. Human milk concentrations of Co (median: 0.19 microg/L), Fe (380 microg/L), Mn (6.3 microg/L), Ni (0.79 microg/L), and Se (17 microg/L) were at the low end of the corresponding reference ranges. Concentrations of Cr (24.3 microg/L) in human milk were five times higher than the high end of the reference range. For Al (67 microg/L), As (6.7 microg/L), and V (0.18 microg/L), most of the samples had concentrations well within the reference ranges. All elemental concentrations in infant formulas (except for Cr) were approximately one order of magnitude higher than in human milk.

Concentrations of trace elements in extensively hydrolysed infant formulae and their estimated daily intakes.

The 18 trace elements Ba, Be, Bi, Cd, Co, Cs, Cu, La, Li, Mn, Mo, Pb, Rb, Sb, Sn, Sr, Tl, and Zn were determined in three extensively hydrolysed formulae by inductively coupled plasma mass spectrometry. Two formulae were whey hydrolysates, whereas one was based on soy-bovine collagen hydrolysate. Two skim milk powder reference materials, analyzed to ensure the analytical precision and accuracy of the applied procedure, showed good agreement with the certified values. Most of the elemental concentrations in this three formulae were comparable to each other. However, concentrations of Cs in one formula were approximately 20 times higher than in the other two formulae. Another formulae had much higher concentrations of La and Mn compared to the other two formulae. As regards Rb, all three formulae had distinctly lower concentrations than the four pre- and nine follow-up formulae investigated in a previous study. For the essential trace element Sn, much higher concentrations were found in the three extensively hydrolysed formulae ( approximately 8 microg/kg) when compared to pre- and follow-up formulae (most of them <0.44 microg/kg). The concentrations of the toxic trace elements Cd, Pb, Sb, and Tl in extensively hydrolysed formulae did not exceed the concentrations in cow-milk-based formulae. Daily intakes provided by the investigated formulae differ by a factor of 1.6 for Co, of 3.2 for Cu, of 8.7 for Mn, of 4.8 for Mo, and of 1.5 for Zn. Adequate daily intakes for Cu, Mo, and Zn are guaranteed by the use of most formulae, whereas only one hydrolysed formula fulfilled these requirements for manganese. Two hydrolysed formulae provide only approximately 50% of the recommended intakes for Mo.

Maternal selenium status in Slovenia and its impact on the selenium concentration of umbilical cord serum and colostrum.

OBJECTIVE: The aim of this study was to assess serum selenium (Se) level of Slovenian mothers at birth and to elucidate its impact on the Se content of umbilical cord serum of their newborns and of that of colostrum. SUBJECTS AND METHODS: In sera of 20 Slovenian mothers at delivery and in the corresponding umbilical cord sera of their newborns Se concentrations were determined by hydride generation atomic absorption spectrometry (HG-AAS). In addition Se levels in 10 healthy female blood donors were determined. Colostrum samples of 18 mothers collected on the second and third day post partum were also investigated for their Se content. RESULTS: Serum Se concentrations of mothers showed to be 62+/-15 microg/l. Umbilical cord sera had 34+/-7 microg/l, which amounts to 55% of the maternal content. Concentrations of Se in colostrum ranged from 17 to 48 microg/l with a mean of 29+/-10 microg/l. In the sera of female blood donors the mean was 66+/-15 microg/l. A significant correlation (P<0.002) between the Se content of maternal and umbilical sera could be established. No significant correlation was found between maternal serum Se concentration and that of colostrum. CONCLUSIONS: Our data show that dietary Se intake for pregnant women in Slovenia is borderline.

[Pericardial effusion in celiac disease--an incidental finding?]. / Perikarderguss bei Zöliakie--Ein Zufallsbefund?

OBJECTIVE: Ultrasound revealed evidence of pericardial effusion in 13 out of 26 children with coeliac disease. In a prospective study, we tried to analyse the causes underlying this high incidence of pericardial effusion. PATIENTS AND METHODS: Twenty-six patients were evaluated. Coeliac disease was diagnosed by intestinal biopsy. All children underwent sonography and a laboratory work-up including endomysial antibodies and serum selenium and iron concentrations. RESULTS: Patients with pericardial fluid showed no difference compared to those without effusion in regard to ECG, chest x-ray, red and white blood cell count, serum enzymes, serum protein as well as iron levels. The mean value of serum selenium was lower and endomysial antibody titre was higher in patients with pericardial effusion. However, due to the wide range, a clear distinction between the two groups was impossible. In all other investigated parameters there was no difference between patients with and without pericardial effusion. Patients with effusion had a higher frequency of viral infection. CONCLUSION: The high incidence of pericardial effusion in patients with coeliac disease appears to be governed by a multifactorial mechanism. A high endomysial antibody titre as well as selenium deficiency may play a role as a predisposing factor. Viral infection due to reduced immunological competence in conjunction with a hampered ability to eliminate toxic free radicals might cause blood vessel dysfunction, resulting in (asymptomatic) pericardial effusion. The fact that most of these patients were diagnosed during the cold season, with anamnestic evidence of viral infection shortly before the diagnosis, and the fact that adult patients with dilative cardiomypathy show a greater prevalence of coeliac disease, supports the view that coeliac disease is systemic in nature.

Determination of selenium in human milk by hydride cold-trapping atomic absorption spectrometry and calculation of daily selenium intake.

A technique of hydride cold-trapping atomic absorption spectrometry following microwave digestion was developed and optimized for the determination of selenium in human milk. The method was validated by the analysis of two standard reference materials (CRM milk powder). The detection limit was 0.5 ng mL(-)(1). The method was then used to analyze 78 milk samples from 38 Austrian mothers throughout their first 10 months of lactation. The mean concentration of selenium in the mother's milk decreased with the days postpartum from 23.9 +/- 12.0 microg L(-)(1) in colostrum to a plateau of 11.4 +/- 3.0 microg L(-)(1) in mature milk. On the basis of the milk selenium concentrations, the selenium intakes of the fully breast-fed infants and the lactating mothers were calculated. The selenium intake of the infants during their first 3 months of life was >8.2 microg day(-)(1). The selenium intake of the lactating mothers was 48 microg day(-)(1). Compared to the recommended dietary allowance, the fully breast-fed infants received sufficient selenium but the lactating mothers obtained less than the recommended.

Trace element transfer from the mother to the newborn--investigations on triplets of colostrum, maternal and umbilical cord sera.

OBJECTIVE: To investigate the trace element transfer from the mother to the newborn. DESIGN: The concentrations of the eight essential elements calcium (Ca), cobalt (Co), copper (Cu), magnesium (Mg), manganese (Mn), molybdenum (Mo), tin (Sn), and zinc (Zn), and of the non-essential and toxic elements barium (Ba), beryllium (Be), bismuth (Bi), cadmium (Cd), cesium (Cs), lanthanum (La), lithium (Li), lead (Pb), rubidium (Rb), antimony (Sb), strontium (Sr), and thallium (Tl) were determined in umbilical cord (n = 29) and corresponding maternal sera (n = 29) as well as in colostrum (n = 27). RESULTS: Umbilical cord serum concentrations of Ca, Mn, and Zn were 120%, 150%, and 148% of the maternal value, respectively. Maternal sera had twice the Cu concentrations found in healthy adults and five-times higher Cu than umbilical cord sera. Concentration ratios colostrum/maternal serum and colostrum/umbilical cord serum were approximately one for Co, 1.4 for Mg, two for Ca, Mn, and Sn, five for Cu (maternal serum), eight for Mo, and ten for Zn. Concentrations of the toxic elements Cd and Pb decreased in the order colostrum (Pb 2.6 microg/L; Cd 0.6 microg/L), maternal sera (0.8 microg/L; 0.3 microg/L), umbilical cord sera (0.4 microg/L; 0.2 microg/L). Maternal serum Ba and Rb was 182% and 66% of the umbilical cord value. For Sr and Li, an almost perfect correlation between umbilical cord and maternal sera was found. For Ba, Co, Cu, Mn, Zn none, and for Ca, Cs, Mn, Mo, Rb only weak positive correlations between these two compartments could be established. CONCLUSIONS: The results of this study indicate that an active transport mechanism for the transport of Ca, Mn, Rb, and Zn from the mother to the newborn exists, whereas Cs, Li, and Sr follow concentration gradients. As regards Cu, the placenta showed to have a blocking effect on the transfer from the mother to the baby.

Concentrations of trace elements in sera of newborns, young infants, and adults.

Concentrations of trace elements in newborns, infants, and adults may be significantly different from each other. Serum trace element reference ranges for different age groups are of value for diagnostic purposes. Inductively coupled plasma-mass spectrometry was applied to the determination of the 21 trace elements Ba, Be, Bi, Ca, Cd, Co, Cs, Cu, La, Li, Hg, Mg, Mn, Mo, Pb, Rb, Sb, Sn, Sr, TI, and Zn in a total of 117 sera of individuals representing different age groups. After microwave-assisted acid digestion with high-purity reagents, 20 umbilical cord sera, 5 sera of fully breast-fed infants, 6 sera of formula-fed infants, 66 sera of patients suffering internal diseases, and 20 sera of healthy blood donors were analyzed for trace elements. One serum and two whole-blood reference materials were analyzed for quality control. Experimental concentrations were in good agreement with certified values. Umbilical cord serum concentrations of the essential elements Ca, Co, Cu, and Mg and of the nonessential and toxic elements Ba, Be, Li, Pb, and Sb were elevated compared to the elemental concentrations in the sera of infants and adults. Serum levels of Ba, Ca, Co, Mn, Pb, and Sb of infants were much higher and serum Cu was significantly lower than in adults. Serum Cu increased significantly with age (newborns: 353 microg/L; infants: 755 microg/L; healthy adults: 810 microg/L), whereas for other trace elements no age-dependence could be established.

Changes in the concentrations of trace elements in human milk during lactation.

Inductively coupled plasma mass spectrometry (ICP-MS) was used to determine 18 trace elements (Ba, Be, Bi, Cd, Co, Cs, Cu, La, Li, Mn, Mo, Pb, Rb, Sb, Sn, Sr, Tl, and Zn) in 55 human milk samples from 46 healthy mothers collected during lactation periods extending to 293 days after birth. Se was quantified by hydride generation atomic absorption spectrometry (HG-AAS). To test the accuracy and the precision of the analytical procedure, milk powder reference materials (BCR 063 and BCR 150) were analyzed. The results obtained by ICP-MS and HG-AAS showed good agreement with the certified values. Whenever available, trace element concentrations determined in the human milk samples were compared to reliable literature data. The concentrations of Be (< 0.05 to 0.9 microgram/kg), Bi (< 0.09 to 2.0 micrograms/kg), Cs (1.7 to 7.7 micrograms/kg), La (< 0.05 to 3.7 micrograms/kg), Rb (440 to 1,620 micrograms/kg), and Tl (< 0.08 to 0.5 microgram/kg) are the first to be reported for human milk. The concentrations of the essential trace elements Cu (p < 0.005), Mn (p < 0.05), Mo (p < 0.0005), Se (p < 0.001), and Zn (p < 0.0005) significantly decreased and the concentrations of cobalt significantly increased (p < 0.005) in human milk during the course of lactation. All concentrations for the essential trace element tin in the human milk samples were below the method detection limit of 0.3 microgram/kg. Among the not essential and toxic elements-with the exception of Ba, Pb, and Tl-the trend toward lower concentrations with continuing lactation is much less pronounced than for the essential trace elements. With the exception of Se, the daily intakes of essential trace elements of fully breast-fed infants are considerably lower than dietary recommendations.

Trace elements in formulas based on cow and soy milk and in Austrian cow milk determined by inductively coupled plasma mass spectrometry.

With inductively coupled plasma-mass spectrometry (ICP-MS), the 18 trace elements Ba, (Be), (Bi), Cd, Co, Cs, Cu, La, Li, Mn, Mo, Pb, Rb, (Sb), (Sn), Sr, (Tl), and Zn were quantified in the digests of 13 formulas based on cow milk, of two formulas based on soy protein, of two milk powders, from which formulas were prepared, of two samples of Austrian cow milk, and in the water, with which the powders were suspended. Concentrations in parentheses were at or below the method detection limits in the formulas. The accuracy and precision of the analytical procedure tested with milk powder reference materials BCR 063 and BCR 150 were satisfactory. The concentrations of trace elements in the powders vary considerably from batch to batch. The ratios of high to low concentrations ranged from 1.1 to 4.8 and were higher for the essential trace elements Co, Cu, Mn, Mo, Sn, and Zn than for nonessential or toxic elements. The contribution of tap water from the water system of the city of Graz, Austria to the concentrations of trace elements in the formulas ranges from 45% for Pb to 0.2% for Rb and is negligible, for instance, for Cd, Cs, La, Mo, and Sn. Preformulas and follow-up formulas are partly supplemented with the essential trace elements Cu, Mn, and Zn and, therefore, concentrations of these trace elements in the formulas vary considerably. However, supplementation of a formula with a particular element must not necessarily result in higher concentrations compared to non-supplemented formulas. Concentrations of the essential elements were in the following ranges for preformulas, follow-up formulas, soy-based formulas (in microg/kg): Co, 8.3-11.2, 4.5-13, 5.0-5.7; Cu, 330-750, 27-730, 440-530; Mn, 33-580, 40-390, 440-530; Mo, 10-32, 9-39, 44-46; Sn, <0.44-3.8, <0.44-1.0, <0.44-5.8; Zn, 3340-11,380, 4120-7100, 5590-6,840. A preformula supplemented with Mn had a 10 times higher manganese concentration than preformulas without supplementation. Concentrations of all trace elements quantified were lower in cow milk than in formulas and do not meet the dietary requirements of infants.

[Value of ultrasound in diagnosis of celiac disease]. / Bedeutung der Sonographie in der Diagnostik der Zöliakie.

AIM: To evaluate the potential benefits of sonographic assessment in the diagnostic work-up of children suspected of having coeliac disease. METHOD: 39 infants with biopsy-proven coeliac disease were evaluated by sonographic assessment. Ten of them had presented with unusual clinical features such as acute abdomen and underwent sonography as the first diagnostic procedure. RESULTS: Various sonographic anomalies were observed: abdominal fluid in 76%, hyperperistalsis in 82%, pericardial effusion in 47% and unusual appearance of the small bowel wall in 94%. CONCLUSION: Although sonography cannot replace intestinal biopsy, awareness of the sonographic anomalies associated with coeliac disease in children can lead to a quicker diagnosis and prompt introduction of adequate therapy. It should be performed as a part of the diagnostic workup in infants who fail to thrive.

Selenium and glutathione peroxidase levels in healthy infants and children in Austria and the influence of nutrition regimens on these levels.

Serum selenium values were investigated in 56 formula-fed and in 18 wholly breast-fed infants. In 14 of these infants, glutathione peroxidase (GSH-Px) was also investigated. Determination of selenium and GSH-Px was also done in umbilical cord blood of seven healthy newborns. In another 109 infants aged 1-15 yr, serum selenium values were investigated. A continuous fall of serum selenium values was noted in the first 3 mo of life. Low levels continued until the age of 6 mo with a mean of 36% of the umbilical cord vein level. Breast-fed babies of GSH-Px showed a less pronounced fall in selenium and had significantly higher levels of GSH-Px. GSH-Px activity was reduced from age 5 to 8 mo. Feeding of beikost caused a rise in the level of selenium. Children in the age groups 1-15 yr still had a significantly lower serum selenium level than adults.

Endomysium antibodies in coeliac disease: an improved method.

The ultra structural binding sites of endomysium antibodies have been studied on human umbilical cord tissue. The sensitivity and specificity of IgA endomysium antibodies were compared with recently described methods using basement membrane of smooth muscle of monkey oesophagus. Thirty adults affected by coeliac disease (10 in remission) and 75 healthy adult controls with normal intestinal mucosa (35 false antigliadin positive) were investigated. Sensitivity and correlation of endomysium antibodies with total villous atrophy in untreated coeliac disease patients were 100% on the human umbilical cord smooth muscles, and only 90% on the muscular layer of primate oesophagus. Indirect immunofluorescence was superior to peroxidase staining in detecting these IgA antibodies. The easy availability and enhanced testing sensitivity of the umbilical cord is an advance towards a better diagnostic tool for coeliac disease.

Simple decomposition procedure for determination of selenium in whole blood, serum and urine by hydride generation atomic absorption spectroscopy.

A digestion procedure for selenium determination by hydride generation atomic absorption spectroscopy (AAS) in whole blood, serum and urine is described, it employs sulfuric acid, hydrogen peroxide and vanadium (V) sulfuric acid reagent solution. The method is rapid, uses no explosive reagents and can be performed at a constant temperature of 100 degrees C. Therefore, it is easily applicable in a routine clinical laboratory for a large amount of samples. The coefficient of intra-assay variation was 4.3-5.6%, the coefficient for inter-assay variation was 5-5.9% in the medium and high concentration range, and 5.8-8.6% in the low range. In analyzing several commercial reference materials our results showed good agreement with the target values. Analytical recovery by addition of sodium selenite and seleno-DL-methionine to samples ranged between 97 and 104%. The correlation between the described digestion procedure and the nitric, sulfuric and perchloric acid digestion procedure recommended by the International Union of Pure and Applied Chemistry showed good agreement for whole blood, serum and for urine. We determined selenium in serum (n = 58) and whole blood (n = 50) in a collective of healthy children from 1 to 5 years living in Styria, Austria. The low values in serum (35 +/- 11 micrograms/L) and whole blood (42 +/- 6 micrograms/L) at one year of life increased significantly to 48 +/- 13 mu/L (p = 0.033) and 55 +/- 6 micrograms/L (p = 0.004) at three years of life in serum and whole blood, respectively. The selenium concentration showed no further increase up to five years of age.(ABSTRACT TRUNCATED AT 250 WORDS)

Sonographic findings in celiac disease.

In 43 children who had been admitted to our department due to symptoms of malabsorption, sonography of the small bowel was performed before biopsy. Seventeen of the patients were shown to have celiac disease. Twenty-six patients had gastroenteritis, postenteritis syndrome, and enteropathic cow's milk allergy, and one patient had sarcoidosis. At the time of diagnosis, 16 children with celiac disease showed a sonographically abnormal appearance of the small-bowel wall structure. In addition to hyperperistalsis, slight ascites, pericardial fluid, or liver tissue texture changes could be found in some of the patients. Relating the findings of abnormal small-bowel wall structures to the data of the small-bowel biopsy, we found a sensitivity of 94% and a specificity of 88% for sonographic detection of changes related to celiac disease. Our findings indicate that sonographic data can be of great help in exploring children with signs of malabsorption. These data provide a further argument for the decision to perform a biopsy of the small bowel.

Human leucocyte class I and II antigens in coeliac disease: a study in an Austrian paediatric population.

Regional variations in the human leucocyte antigen (HLA) distribution patterns of coeliac disease (CD) have been reported. This study focuses on phenotype frequencies of a cohort of Austrian paediatric CD patients in comparison with those recorded in the literature. HLA class I and II typing was performed in 136 CD patients and 667 healthy controls from the general population of the same geographical area. The HLA phenotypes of our controls agreed with those published for Caucasians. In our patients the relative risks (RR) were 6.43 for DR3 and 2.52 for DR7, the aetiologic fractions being 0.58 respectively 0.24. The highest RR (7.78) was found for DR3/DR7 heterozygotes. The RR for DR5 was increased in heterozygosities, either with DR3 (3.34) or DR7 (5.53), yet not for DR5 alone. Of our patients, 10% were lacking both DR3 and DR7 as well as B8, 82% of them were positive for DQw3. In these DR3 and DR7 negative patients, DR4 and DR5 were significantly more frequent than in the others. DR5 was also significantly more common in these patients compared to controls lacking the same antigens, whereas this did not hold true for DR4. Prospective studies are required to determine any link between these HLA heterogeneities and long-term progression of the disease.