Gilbert's syndrome and
jaundice in glucose-6-phosphate dehydrogenase deficient
neonates
Michael Kaplan, Cathy Hammerman
Department of Neonatology, Shaare Zedek Medical Center; Faculty
of Medicine of the Hebrew University, Jerusalem, Israel.
Correspondence: Michael Kaplan, M.D., Department of Neonatology,
Shaare Zedek Medical Center, PO Box 3235, Jerusalem 91031, Israel;
Phone: international +972-2-6666154, 655-5551 - Fax: international
+972-2-6520689. E Mail: kaplan@cc.huji.ac.il
We read the article by Iolascon et al.,1 studying
the effect of Gilbert's Syndrome on the risk of developing
hyperbilirubinemia in Sardinian neonates, with great interest. The
authors concluded that homozygosity for the variant gene promoter
for the bilirubin conjugating enzyme, UDP glucuronosyltransferase
(UGT), seen in Gilbert's Syndrome, does not play a significant
role in moderating the incidence of hyperbilirubinemia in
glucose-6-phosphate dehydrogenase (G-6-PD) deficient newborns in
their population. They contrasted their findings to a recent
report of ours, in which we did find, in Sephardic Jews in Israel,
that the presence of UGT gene promoter polymorphism with seven TA
repeats, rather than the usual six, in combination with G-6-PD
deficiency, resulted in a significantly higher incidence of
neonatal hyperbilirubinemia than newborns with the identical UGT
genotype, but without G-6-PD deficiency.2 An effect was
observed not only in those homozygous for the promoter
polymorphism, but in the heterozygotes as well. However, there are
a number of differences in methodology in the Iolascon et al study
and ours which might have accounted for some of these
discrepancies, while dissimilar statistical analysis may have led
the authors to conclude that there was no effect of the promoter
polymorphism on the incidence of hyperbilirubinemia. Comparison
between our two studies is clouded by the fact that our data was
collected as a cohort of consecutively born babies, while the 45%
incidence of hyperbilirubinemia, instead of an incidence of 25% to
30% usually reported in their population, implies that Iolascon et
al's data was not reflective of a cohort. Iolascon et al. assumed
that, if promoter polymorphism for the UGT gene did cause an
increase in the incidence of hyperbilirubinemia in the entire
population, then the gene frequency for the seven repeat promoter
would also be higher in the population subset of G-6-PD deficient
neonates who became hyper-bilirubinemic. In their study, Iolascon
et al analyzed the frequency of homozygosity for the variant UGT
gene promoter in two highly selected subgroups of their infant
population: those who had developed hyperbilirubinemia, and those
who did not develop significant jaundice. Homozygosity was found
in 11% of 56 hyperbilirubinemic, and in 17% of 46
non-hyperbilirubinemic G-6-PD deficient infants, suggesting the
antithesis of what we had previously found. Our study, however,
contrasts with the present one, in that we prospectively studied
an entire population cohort and determined the incidence of
hyperbilirubinemia for each of the three UGT promoter genotypes in
the G-6-PD deficient and G-6-PD normal populations, respectively.
Using this method, and focussing, for the purpose of this
discussion, as did Iolascon et al, on those who were homozygous
for the variant promoter, we found that 6/12 (50%) in the G-6-PD
deficient group, compared with 5/34 (14%) of those with the
equivalent UGT genotype in the control group, developed
hyperbilirubinemia (p=0.03). Having ascertained these findings in
the population based study, we too expected to find a
preponderance of homozygosity for the variant UGT gene promoter
among the subset of G-6-PD deficient infants who had developed
hyperbilirubinemia. However, reanalysis of our data shows that,
had we studied only those who became hyperbilirubinemic, we would
have arrived at a very different conclusion: of the 30 G-6-PD
deficient neonates in our study who developed hyperbilirubinemia,
6 (20%) were homozygous for the variant UGT gene promoter, while 5
(22%) of the 22 hyperbilirubinemic control neonates had this
genotype. Thus, while in the same population cohort, a
prospective, population based study did provide evidence of
interaction between Gilbert's syndrome and G-6-PD deficiency in
exacerbating neonatal hyperbilirubinemia, analysis of the highly
selected hyperbilirubinemic subset of the same population group
did not show this effect.
The apparent lack of effect when studying the hyperbilirubinemic
subgroups may have been exacerbated by the relatively small
numbers of patients, especially when the statistical analysis
focussed on the smallest of the UGT genotypes. It is, of course,
possible that genetic differences in the populations studied, as
suggested by Iolascon et al., do exist, or that environmental
influences may indeed have resulted in the discrepancies noted. We
hope that this discussion will stimulate a prospective, population
based study in order to determine definitively, whether the gene
interaction, crucial to the development of hyperbilirubinemia in
Sephardic Jewish G-6-PD deficient neonates, also occurs in
Sardinian newborns.
