Clinical analysis
After excluding neonates with the conditions described above, a total of 69 full-term ABO HDNs infants were admitted to the hospital on the 3rd day after birth (range, 1–10 days). The average peak serum total serum bilirubin level (TBIL) was 335.1 µmol/L (123.3-652.1 µmol/L). Among 69 ABO HDNs, 31 cases had peak TSB ≥ 342 µmol/L, in which 15 neonates had TSB ≥ 427 µmol/L. All the neonates received phototherapy. Most of the neonates were discharged without complication, except for two infants with bilirubin encephalopathy symptoms.
There were significant differences in the bilirubin levels, but no differences in average gestational age, birth weight, gender and feeding pattern between the two groups of neonates divided according to the c.211 genotypes of UGT1A1– the most common UGT1A1 variant in Asian population (Table 1).
Table 1
Demographic and clinical features among the neonates with ABO hemolytic disease (ABO HDN) in UGT1A1 c. 211 G > A mutation group VS c.211 normal group (N = 69).
| c.211 mutation/ABO HDN | c.211 normal /ABO HDN | P |
Sex | | | NS |
Male | 6(30.0) | 25(51.0) | |
female | 14(70.0) | 24(49.0) | |
Gestational week | 39.42 ± 0.23 | 39.24 ± 0.13 | NS |
Birth weight(kg) | 3.30 ± 0.076 | 3.19 ± 0.056 | NS |
Peak serum bilirubin levels (umol/L) | | | |
TBIL | 404.15(219.02-556.07) | 315.00(145.8-585.2) | 0.028 |
IBIL | 380.70(201.04-527.73) | 303.10(128.85-552.05) | 0.037 |
DBIL | 14.30(4.88–69.07) | 12.08(3.30–35.20) | NS |
Feeding | | | NS |
Breast fed | 3(15.0) | 11(22.4) | |
Breast and formula | 2(10.0) | 9(18.4) | |
Formula | 14(70.0) | 20(40.8) | |
unknown | 1(5.0) | 9(18.4) | |
Delivery method | | | NS |
Vaginal | 7(35.0) | 27(55.1) | |
Cesarean | 13(65.0) | 22(44.9) | |
Data are presented as n (%), mean standard deviation, or median (95% Confidence Interval). |
NS: no significance. |
Ugt1a1 Variant Results
In addition to the two common variants of UGT1A1 gene, TA7 polymorphism (UGT1A1*28, rs8175347) in the promoter and c.211 G > A mutation (UGT1A1*6, p.Arg71Gly, rs4148323) in exon 1, another coding variant c.1091C > T (UGT1A1*73, p.Pro364Leu, rs34946978) was observed in the neonates. Specifically, heterozygote of TA7 promoter polymorphism (TA6/TA7) was detected in 9 neonates, with no homozygote for TA7 polymorphism (TA7/TA7) observed. The frequency of heterozygous (G/A) and homozygous (A/A) genotypes of c.211 G > A mutation were 0.275 (19/69) and 0.014 (1/69), respectively. Three cases were observed with heterozygous c.1091C > T mutation (Table 2).
Table 2
Minor allelic, genotypic, and haplotype distributions of UGT1A1 polymorphism in studied patients (N = 69).
UGT1A1 polymorphism | Location | N | Frequency | PH−W* |
TATA box(rs8175347) | promoter | | | 1 |
TA7 | | 9 | 0.07 | |
TA6 | | 129 | 0.93 | |
TA7/TA7 | | 0 | 0 | |
TA6/TA7 | | 9 | 0.13 | |
TA6/TA6 | | 60 | 0.87 | |
C.211 G > A(rs4148323) | Exon 1 | | | 1 |
A | | 21 | 0.15 | |
G | | 117 | 0.85 | |
A/A | | 1 | 0.01 | |
G/A | | 19 | 0.28 | |
G/G | | 49 | 0.71 | |
c.1091 C > T(rs34946978) | Exon 4 | | | 1 |
T | | 3 | 0.02 | |
C | | 135 | 0.98 | |
T/T | | 0 | 0 | |
C/T | | 3 | 0.04 | |
C/C | | 66 | 0.96 | |
Haplotype(rs8175347-rs4148323-rs34946978) | | | | |
TA6GC | | | 0.76 | |
TA6AC | | | 0.15 | |
TA7GC | | | 0.065 | |
TA6GT | | | 0.018 | |
TA6AT | | | 0.0037 | |
* Hardy-Weinberg Equilibrium test p value. |
There was a strong pairwise LD between the UGT1A1 promoter polymorphism and exon mutation (|D|’>0.8), but none of the polymorphisms in our study had a statistically significant deviation in the HWE test. Haplotype analysis showed that the TA6GC (rs8175347-rs4148323-rs34946978) was the predominant haplotype among the study subjects (75.85%), (Table 2).
Co-inherited Ugt1a1 Variant On Bilirubin Levels In Abo Hdns
When analyzing the peak bilirubin levels according to UGT1A1 genotypes, the average peak bilirubin levels (TBIL and indirect serum bilirubin level (IBIL)) of ABO (+) neonates with both heterozygous and homozygous c.211 G > A coding mutation were higher as compared to those with normal UGT1A1 genotype (P = 0.028 for TBIL, P = 0.036 for IBIL), whereas direct serum bilirubin level (DBIL) showed no statistical difference among the three genotypes (Table 1, Fig. 1). No significant difference in bilirubin levels was observed in the presence of either heterozygous of TA7 promoter polymorphism or heterozygous of c.1091C > T mutation in the neonates.
After adjusting the potential covariance (age, gender, and feeding method), c.211 G > A mutation was still associated with the increased bilirubin levels (ORadj=78.24, 95%CI 14.68–141.80, P = 0.019 for TBIL, ORadj=73.26, 95%CI 12.84-133.69, P = 0.021 for IBIL) (Table 3). Moreover, haplotype association analysis showed that the TA6AC (rs887829-rs4148323- rs34946978) was significantly associated with increased bilirubin levels (ORadj=84.01, 95%CI 23.2-144.82, P = 0.009 for TBIL, ORadj=79.02, 95%CI 21.40-136.64, P = 0.009 for IBIL). Haplotype TA6GT also showed significant association with increased bilirubin levels (ORadj=107.89, 95%CI 93.65-122.14, P < 0.001 for TBIL, ORadj=107.01, 95%CI 93.58-120.44, P < 0.001 for IBIL) (Table 3).
Table 3
The associations between serum bilirubin level and different types of UGT1A1 mutation and genotypes adjusted by age, gender, and feeding practice: Line regression analysis (N = 59*).
Genotypes | TBIL | | IBIL | |
| ORadj† (95%CI) | P | ORadj† (95%CI) | P | |
TATA box | | | | | |
TA6/TA6 | 0.00 | 0.57 | | 0.00 | 0.58 |
TA6/TA7 | 30.04(-73.12-133.19) | | 27.55(-70.39-125.49) |
c.211 G > A | | | | | |
G/G | 0.00 | 0.019 | | 0.00 | 0.021 |
G/A-A/A | 78.24(14.68–141.80) | | 73.26(12.84-133.69) |
c.1091 C > T | | | | | |
C/C | 0.00 | 0.16 | | 0.00 | 0.14 |
C/T | 103.06(-39.78-245.90) | | 103.63(-31.65-238.91) |
Haplotype‡ (Frequency) | | | | | | |
TA6GC(75.85%) | 0.00 | | | 0.00 | | |
TA6AC(16.53%) | 84.01(23.2-144.82) | 0.0092 | | 79.02(21.4-136.64) | | 0.0097 |
TA7GC(5.08%) | 70.42(-26.83-167.68) | 0.16 | | 65.65(-26.57-157.87) | | 0.17 |
TA6GT(2.12%) | 107.89(93.65-122.14) | < 0.0001 | | 107.01(93.58-120.44) | | < 0.0001 |
TA6AT(0.42%) | 149.95(147.33-152.58) | < 0.0001 | | 152.42(150.02-154.82) | | < 0.0001 |
* Ten cases with the feeding practice unknown were not taken into the regression analysis. |
†Adjusted for age, gender, and feeding practice. |
‡polymorphisms are in order of: rs8175347-rs4148323-rs3494697 |
Co-inherited UGT1A1 variant on severe hyperbilirubinemia risk in ABO HDNs
The incidence rates of hazardous and severe hyperbilirubinemia in the ABO HDNs were compared in different types of UGT1A1 genotype. Promoter polymorphism and exon mutations were analyzed, separately. Compound heterozygous mutations in the coding sequence were regarded as homozygous mutations. No statistical difference of severe hyperbilirubinemia incidence was found between ABO HDNs with and without the UGT1A1 mutation (P > 0.05). On the contrary, after adjusted by age, gender, and feeding method, ABO HDNs with heterozygous and/or homozygous mutations in the UGT1A1 coding sequence region had a relatively higher risk of developing hazardous hyperbilirubinemia than those with a normal UGT1A1 genotype (ORadj=9.16, 95%CI 1.99–42.08, P = 0.002). Moreover, haplotype association analysis showed that TA6AC was significantly associated with a higher incidence of hazardous hyperbilirubinemia in ABO HDNs (ORadj=9.41, 95%CI 1.80- 49.26, P = 0.011) (Table 4).
Table 4
The associations between risk of severe neonatal hyperbilirubinemia and UGT1A1 coding sequence variants and different type of UGT1A1 haplotype in neonates with ABO hemolysis disease: multivariate logistic regression analysis (N = 59*).
Genotype | | Total | | TBIL > 342 umol/L | | TBIL > 427umol/L |
N(%) | N(%) | ORadj† (95%CI) | P | N(%) | ORadj† (95%CI) | P |
Model 1(n = 59) | | | | | | | | | | |
Wild type | | 38(64.4%) | | 16(57.1%) | 1.00 | | | 4(30.8%) | 1.00 | |
Heterozygous mutation | | 19(32.2%) | | 10(35.7%) | 2.25(0.66–7.61) | 0.076 | | 8(61.5%) | 8.74(1.86–41.09) | 0.008 |
Homozygous mutation | | 2(3.4%) | | 2(7.1%) | NA(0.00-NA) | | | 1(7.7%) | 15.09(0.61-375.49) | |
Model 2(n = 59) | | | | | | | | | | |
Wild type | | 38(64.4%) | | 16(57.1%) | 1.00 | | | 4(30.8%) | 1.00 | |
Heterozygous + Homozygous mutation | | 21(35.6%) | | 12(42.9%) | 2.72(0.82–9.03) | 0.093 | | 9(69.2%) | 9.16(1.99–42.08) | 0.002 |
Halplotype( Frequency) | | | | | | | | | | |
TA6GC(75.89%) | | | | | 1.00 | | | | 1.00 | |
TA6AC(16.58%) | | | | | 3.46(0.97–12.39) | 0.062 | | | 9.41(1.80-49.26) | 0.011 |
TA7GC(5.08%) | | | | | 2.57(0.37–17.89) | 0.35 | | | 8.18(0.80-83.65) | 0.083 |
TA6GT(2.17%) | | | | | 2.38(0.12–48.13) | 0.57 | | | 10.64(0.44-254.56) | 0.15 |
other‡ | | | | | - | - | | | - | - |
* Ten cases with the feeding practice unknown were not taken into the regression analysis |
†Adjusted for age, gender, and feeding practice. |
‡ Other haplotypes had frequencies less than 1%. |