Objective
Women with maturity-onset diabetes of the young (MODY) are often first identified and diagnosed with diabetes during pregnancy. Genetics and hyperglycemia play an important role in determining fetal size in MODY pregnancies. The principal objective of the current study is to determine the outcomes and clinical management of hyperglycemia in pregnancies complicated by glucokinase gene (GCK) and hepatocyte nuclear factor (HNF)-1α MODY mutations.
Study Design
A retrospective chart review of 37 women with a GCK/HNF-1α mutation was conducted. Data on variables such as birthweight, mode of delivery, and the treatment of hyperglycemia were available on 89 pregnancies.
Results
The birthweight in unaffected GCK offspring was significantly higher than in the affected GCK offspring (4.8 [4.1-5.2] kg vs 3.2 [3.1-3.7] kg; P = .01). Seven – point home blood glucose monitoring over a 7-day period in each trimester demonstrated higher fasting and postprandial glycemic excursions in the first trimester of GCK pregnancies when compared to HNF-1α pregnancies (fasting 104 [90-115] mg/dL vs 84 [77-88] mg/dL; P = .01 and postprandial 154 [135-196] mg/dL vs 111 [100-131] mg/dL; P = .04) despite insulin treatment. There was a higher percentage of miscarriages in the GCK group when compared to the HNF-1α MODY group (33.3% vs 14%; P = .07), which was similar to the background population. Insulin initiated at an early gestation appeared to lower the incidence of macrosomia in GCK unaffected offspring.
Conclusion
Hyperglycemia in HNF-1α pregnancies is easily managed with current insulin protocols; in contrast, glycemic excursions are difficult to manage in GCK pregnancies. There was an increased percentage of miscarriages in GCK pregnancies highlighting the importance of a diagnosis of GCK-MODY in women prior to conception and the necessity for preconception care.
Maturity-onset diabetes of the young (MODY) is a monogenic autosomal dominant form of diabetes with the most common causative mutations being those in hepatocyte nuclear factor (HNF)-1α and a mutation in the glucokinase gene (GCK). In contrast to type 1 diabetes mellitus, MODY mutation carriers usually are autoantibody negative (islet cell antibody/antiglutamic acid decarboxylase), are nonketotic, and have detectable C-peptide suggesting preserved beta-cell function. Clinically, they differ from those with type 2 diabetes mellitus and gestational diabetes mellitus (GDM) by being lean and lacking features of the metabolic syndrome. GCK-MODY is specifically associated with persistent fasting hyperglycemia. GCK is proposed to be a stable state of hyperglycemia outside of pregnancy, while HNF-1α MODY is reported to have a similar microvascular and macrovascular complication rate to type 1 diabetes mellitus. The reported prevalence of MODY among women with GDM varies between 0.1-6%. Due to its relative rarity it has not been possible to perform a prospective study on the clinical management of MODY pregnancies. The literature available is therefore based primarily on case reports or retrospective analysis of diabetes cohorts in pregnancy.
Clinical experience of treating MODY in pregnancy is also limited as currently the majority of women with GCK and HNF-1α remain undiagnosed during pregnancy. In GCK it has been reported that fetal genetics are at least as important as maternal glucose control in determining fetal size and hence complications. If both the mother and infant have a glucokinase deficiency then the infants are on average 540 g lighter at birth. If the infant does not inherit the mutation and is exposed to untreated hyperglycemia in utero then birthweight can be increased by up to 600 g. The question arises therefore if it is more harmful to treat all women with GCK with insulin or more prudent to withhold insulin treatment. The initiation of insulin treatment in early gestation in pregnancies complicated by GCK has not been studied. To date, there is but 1 case report suggesting the harmful consequence of insulin in GCK pregnancies. In this report, the author suggested that insulin resulted in a small for gestational age (SGA) offspring.
The principal utility of insulin during pregnancy is to achieve euglycemia and prevent excessive fetal growth with its inherent complications. In women with GCK, fasting hyperglycemia predates the pregnancy and should theoretically increase the risk of congenital anomalies and rates of miscarriage. Hyperglycemia in the first trimester is also a risk factor for macrosomia. These women are exposed to hyperglycemia both preconception and indeed throughout all 3 trimesters.
Treatment of HNF-1α in pregnancy is not fraught with the same dilemma as the management of GCK pregnancies as birthweight is reported not to be affected by mutation status of the offspring.
Studies performed pertaining to pregnancy and MODY to date, have largely focused on birthweight and predicting gene expression in offspring of MODY carriers. The aim of the current study is to determine the outcomes and clinical management of hyperglycemia using home blood glucose monitoring (HBGM) in pregnancies complicated by GCK and HNF-1α mutations.
Materials and Methods
Subjects
A total of 37 (n = 12, GCK; n = 25, HNF-1α) women participated in the study; these women were identified as part of the MODY cohort screening study in the Mater Hospital. The GCK mutations were p.D311fs, A378V, p.I293R, L146fs, D160N, and p.Y61X (6 pedigrees). The HNF-1α mutations were P379T, p.L502fs, V351fsdelG, c.1502_6 G>A, p.G292fs, S335X, Ser352fs, p.R159Q, p.R200Q, and P.E230fs (11 pedigrees).
We retrospectively studied the pregnancy outcome in these 37 women. The mutation status was known prepregnancy in 18.9% of the women studied. Therefore, the majority of the women were diagnosed with GDM and treated as such. This analysis involved both review of maternal charts and maternal recall. There were 132 pregnancies in these women. Data available in 89 cases included birthweight, gestation at delivery, and mode of delivery. Complications that included congenital anomaly, birth injury, admission to neonatal intensive care unit, and prolonged neonatal hypoglycemia were recorded. Genotyping of 36 offspring from these 37 women was known and a presumptive diagnosis of a GCK mutation was made in an additional 5 of the offspring of GCK mothers based on a fasting blood glucose of >100 mg/dL on 2 separate occasions. The presumptive diagnosis was made in offspring aged ≥10 years.
Diabetes care in Dublin is delivered exclusively by a single team of endocrinologists, diabetes nurse specialists, and specific obstetricians at each clinical site. All women who are diagnosed with pre-GDM or GDM not responding to dietary modification are treated with insulin. Women are treated with short-acting insulin analogues premeals and neutral protamine hagedorn insulin up to 4 times daily (premeal and bedtime). All women perform HBGM using glucometers 7 times daily (before and 1 hour postmeal and at bedtime). Glucose targets are fasting premeal and at bedtime of <90 mg/dL and 1-hour postmeal readings of <126 mg/dL. The same principals were applied to the clinical management of women diagnosed with GCK or HNF-1α MODY.
The HBGM recordings of 15 women, who were all managed according to our current guidelines for diabetes in pregnancy, with a GCK or HNF-1α mutation referred to 1 of the 3 tertiary centers, were analyzed. The median of the highest and lowest blood glucose value for each day, at each time point, was averaged over a 7-day period. The relative use of each insulin was determined. In addition, data on severe hypoglycemic episodes were recorded. A severe hypoglycemic episode was defined as one necessitating third-party assistance. Serum fructosamine corrected for albumin was checked at clinic visits. Fructosamine is a glycated serum protein complex that reflects the average blood glucose concentration over the preceding 1-3 weeks. Fructosamine is therefore particularly useful given the rapid fluctuations in insulin dosage required during pregnancy. The treatment target for fructosamine is <250 μmol/L. Data on birthweight in the general GDM cohort were available through the annual report for 2013 (rotunda annual report). Outcome variables recorded were rate of first-trimester spontaneous abortion (miscarriage rates) following booking to clinic, major congenital anomalies rate classified as per European Registry of Congenital Anomalies guide, preterm delivery (<37 weeks), and infant birthweight. Centile growth was estimated using standard centile growth charts to eliminate the confounding factor of fetal sex and gestational age at delivery. Macrosomia was defined as being a birthweight >95th centile for that sex and gestation. SGA was defined as a birthweight <10th centile.
Assays
All laboratory analyses were performed with commercially available standardized methods. The HBGM was performed on Ascensia Contour (Bayer, Leverkusen, Germany) glucometers. Fructosamine was determined using a colorimetric assay (Roche Diagnostics, Basel, Belgium).
Genetic analysis
Analysis of the GCK and the HNF-1α gene was performed by polymerase chain reaction amplification of highly purified genomic DNA, followed by semiautomated unidirectional DNA sequencing of all exons including the highly conserved flanking intronic sequences of the exon-intron splice junctions. Genetic analysis was performed by IntegraGen GmbH (Bonn, Germany) in 2000 through 2007 and the Molecular Genetic Laboratory (Exeter, United Kingdom) from 2008 through 2013.
Statistical analysis
Quantitative variables were expressed as median and interquartile range; qualitative variables were expressed as absolute number or percentages. Due to the study population number normal distribution could not be determined and nonparametric tests were applied. Mann-Whitney U test was used to compare quantitative variables, the χ 2 test was used for frequency comparisons. Only 2-tailed P values < .05 were considered significant.
Results
Clinical parameters
The clinical characteristics of the participants are contained in Table 1 .
| Variable | GCK-MODY | HNF-1α MODY | P value |
|---|---|---|---|
| Participants, n | 12 | 25 | n/a |
| Age, y | 22 ± 6 | 24 ± 8 | n/a |
| No. of pregnancies, n | 56 | 76 | n/a |
| No. of miscarriages, % | 33.3 | 14 | .07 |
| No. of live births, n | 41 | 65 | n/a |
| Miscarriage gestation, wk | 7.5 (6.2–8.7) | 12 (10–14) | .003 a |
| First-trimester weight, kg | 76.4 (59.4–81) | 75.3 (62–82) | .9 |
| Third-trimester weight, kg | 94 (76–97) | 85.9 (73–91) | .5 |
| First-trimester fructosamine, μmol/L | 275 (250–306) | 251 (231–283) | .5 |
| Third-trimester fructosamine, μmol/L | 238 (226–252) | 230 (208–279) | .8 |
| No. of pregnancies insulin commenced, % | 26.6 | 26 | ns |
| Gestation insulin commenced, wk | 14 (8–29) | 15 (7–28) | .8 |
| Insulin dose, U/kg | |||
| Trimester 1 | 0.3 (0.1–0.5) | 0.4 (0.2–0.6) | .4 |
| Trimester 2 | 0.4 (0.2–0.7) | 0.6 (0.4–0.7) | .7 |
| Trimester 3 | 0.8 (0.8–1.5) | 0.8 (0.6–0.9) | .7 |
| Delivery gestation offspring, wk | 40 (39–40) | 39 (38–40) | .02 a |
| Birthweight offspring kg/centile | 3.9 (3.2–4.5) 88th (33rd–99th) | 3.6 (3.1–4) 60th (19th–88th) | .4 |
| Macrosomia (>95th centile) | 31.2 | 10 | .01 a |
| Cesarean delivery (insulin treated), % | 57.2 | 37.5 | ns |
| Cesarean delivery (noninsulin treated), % | 42.8 | 14.3 | ns |
There was no significant difference in birthweight in GCK pregnancies when compared to HNF-1α pregnancies (3.9 [3.2-4.5] kg vs 3.6 [3.1-4] kg; P = .4). Insulin doses expressed in U/kg did not significantly differ between the groups throughout the trimesters ( Figure 1 ). In both the GCK and the HNF-1α group 7 injections (3 short-acting insulin analogs and 4 neutral protamine hagedorn injections) were required in the subgroup analyzed. The booking fructosamine levels were higher in the GCK group than in the HNF-1α group (275 [250-306] μmol/L vs 251 [231-283] μmol/L; P = .5). Following insulin treatment, fructosamine levels in the third trimester were similar in both the GCK and the HNF-1α group (238 [226-252] μmol/L vs 230 [208-279] μmol/L; P = .8). The fructosamine at presentation in both the GCK and the HNF-1α groups was higher than that noted in our general GDM cohort (236 ± 15 μmol/L).
There was a higher rate of miscarriage in the GCK group when compared to the HNF-1α group. This rate was higher than that seen in the HNF-1α group (33% vs 14%; P = .07), which is comparable to the background population miscarriage rate of 15% ( www.irishhealth.com ). The miscarriages occurred at an earlier gestation in the GCK group than that noted in the HNF-1α group (7.5 [6.2-8.7] weeks vs 12 [10-14] weeks; P = .003).
Home blood glucose monitoring
In 15 of the pregnancies (n = 5, GCK; n = 10, HNF-1α) HBGM data were studied. Figure 2 , A and B, demonstrate the average glucose recordings for each of these groups of women, respectively, for 1 week during the first and third trimester. Women with GCK have more dramatic glycemic excursions than that noted for HNF-1α pregnancies particularly in the first trimester. These excursions were significantly higher in the morning and early afternoon in the GCK group in the first trimester when compared to the HNF-1α group ( Table 2 ). In the third trimester the prebedtime and fasting HBGM readings were significantly higher in the GCK group when compared to the HNF-1α group ( Table 2 ).
| Variable | GCK-MODY, median (IQR), mg/dL | HNF-1α MODY, median (IQR), mg/dL | P value |
|---|---|---|---|
| First-trimester HBGM | |||
| Prebreakfast | 104 (90–115) | 84 (77–88) | .01 |
| Postbreakfast | 154 (135–196) | 111 (106–131) | .01 |
| Prelunch | 102 (86–122) | 81 (73–90) | .05 |
| Postlunch | 131 (118–160) | 122 (108–138) | ns |
| Predinner | 106 (93–115) | 97 (84–117) | ns |
| Postdinner | 120 (97–156) | 122 (86–156) | ns |
| Prebedtime | 115 (99–126) | 95 (86–115) | ns |
| Third-trimester HBGM | |||
| Prebreakfast | 100 (86–106) | 82 (72–88) | .03 |
| Postbreakfast | 145 (108–153) | 124 (118–153) | ns |
| Prelunch | 86 (70–117) | 86 (84–113) | ns |
| Postlunch | 133 (117–154) | 109 (106–126) | ns |
| Predinner | 93 (86–145) | 90 (86–106) | ns |
| Postdinner | 118 (108–138) | 126 (117–133) | ns |
| Prebedtime | 122 (108–138) | 90 (84–90) | .05 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
