Insulin analogs in pregnancy

Endocrinol Nutr. 2005;52(9):473-5 473 Several clinical studies1 have shown that control of maternal glycemia before conception reduces the frequency of congenital malformations, which remain the leading cause of mortality and severe morbidity in the infants of diabetic mothers2,3. Good metabolic control throughout pregnancy can also reduce the maternal and other fetal complications typical of pregnancy complicated by diabetes4. The levels of glycemia and HbA1c that should be achieved before conception and maintained during pregnancy, as recommended by the American Diabetes Association5, demand intensive blood glucose self-monitoring and optimized insulin therapy so that these values can be obtained with the lowest possible number of hypoglycemic episodes. Pregnancy is characterized by changes in insulin requirements. To be more precise, the placental passage of glucose and gluconeogenetic substrates causes maternal hypoglycemia in the first trimester (especially at night) and consequently the insulin dosage adopted before conception should be reduced by approximately 10%. During the second and third trimesters, the progressive increase in the production of placental contra-insulin hormones leads to a gradual increase in insulin requirements. Throughout pregnancy, moreover, blood glucose control becomes unstable, with a trend toward low fasting plasma glucose, high postprandial peaks and episodes of nocturnal hypoglycemia. Because of these changes, the dosage of short-acting insulin must be increased to cover the meal and the dosage of intermediate-acting insulin must be finely tuned to guarantee a constant basal rate without hypoglycemic episodes. The rapid-acting insulin analogs, i.e. lispro and aspart, may be useful because they are better able to reduce postprandial hyperglycemia than regular insulin. As for insulin glargine, its potential advantage in the management of pregnancy complicated by type 1 diabetes lies in its ability to reduce nocturnal hypoglycemia, which is particularly common in pregnant women with type 1 diabetes. However, along with the potential benefits, the potential related risks (teratogenicity, embryo toxicity, immunogenicity with transplacental passage, mitogenicity) of using these new insulins in pregnancy must be carefully considered. Concerning teratogenicity, studies in experimental animals have shown no embryotoxic or teratogenic effects of insulin lispro, aspart or glargine. The first report of fetal malformations in diabetic patients treated with insulin lispro involved two type 1 diabetic women with good metabolic control, whose therapy with insulin lispro was started before conception in one patient and in the third week of pregnancy in the other. One pregnancy was aborted at the 20th week of gestation, and the fetus had multiple cardiac malformations; the other ended in term delivery, but the infant died 3 weeks after birth from a congenital diaphragmatic hernia. Since this first report, a series of retrospective clinical studies have been published8. As a whole, 867 pregnant diabetic women treated with insulin lispro (99% with type 1 and 1% with type 2 diabetes) and 221 pregnant diabetic women treated with regular insulin (99.5% with type 1 diabetes) can be compared. Unplanned pregnancies occurred in 41% of the diabetic pregnant women treated with insulin lispro and in 68% of those treated with regular insulin. As for the timing of the insulin therapy, 98.3% of patients were treated with lispro at the time of conception and 1.7% were transferred from regular insulin to insulin lispro between the 7th and 17th weeks of gestation due to poor metabolic control. These studies show that HbA1c levels at conception vary considerably (from 5.2% to 13.1%). Cumulative data analysis shows that the frequency of malformations was 4.8% in patients on insulin lispro and 6.8% in those on regular insulin; this difference was non-significant (p = 0.3). More recently, a retrospective study promoted by Eli Lilly9 evaluated the outcome of pregnancy in 496 patients with pregestational diabetes from 55 centers in different parts of the world. This study examined patients Editorial