WASHINGTON – Short-acting insulin analogues appear to be safe in insulin-requiring pregnancies and have clinical advantages, including increased freedom of meal timing, better matching of insulin dose with meal content, and improved glycemic control with a reduction in the frequency of hypoglyemic events, said Dr. Marshall W. Carpenter.
The analogues lispro (Humalog) and aspart (Novolog) may be advantageous, for instance, for the pregnant woman “with a toddler around who may not know when she is going to sit down to eat,” Dr. Carpenter said at the meeting.
“The benefit is reflected in the [higher, faster] peak insulin values seen with both lispro and aspart compared with human insulin,” said Dr. Carpenter of the department of obstetrics and gynecology at Tufts University, Boston.
Dr. Virginia R. Lupo, who chairs the department of obstetrics and gynecology at the Hennepin County Medical Center in Minneapolis, offered a different take on the utility of short-acting insulin analogues.
Diabetes disproportionately affects women who have annual household incomes below $25,000 and who are more likely to be black, Hispanic, American Indian, or Asian Pacific Islanders than white, she said. For many of these women, a regimen including short-acting insulin analogues is too complex for their lifestyle, eating habits, functional health literacy, and other life circumstances.
“A lot of my patients eat by grazing – there are no distinct meal times,” she said. And because of evening-long food availability and ingestion, these patients “require elevated basal insulin through the evenings.”
“I'm not convinced that insulin analogues are the right way to go,” she said. “I like the idea of NPH twice a day, before breakfast and before supper. It's better to take [insulin] twice a day than not take it five times a day.”
The utility of long-acting insulin analogues in pregnancy, Dr. Carpenter said, is still uncertain considering the lack of substantive real-life clinical experience with these analogues and the safety implications raised in the literature thus far. “And really,” he added in an interview after the meeting, “there's no evidence that glargine (Lantus) offers any benefit over NPH insulin – and NPH insulin is cheaper.”
Questions about the safety of analogues overall stem from the molecular modifications involved in their creation, and specifically the implications of modifying the C-terminal end of the insulin beta chain. Such modifications appear to increase affinity for the insulinlike growth factor–1 (IGF-1) receptor – a receptor that has “a broad array of effects,” from induction of mitogenesis and inhibition of apoptosis, to stimulation of angiogenesis, he said.
Analogues' increased “stimulation of [this receptor] has thus appropriately raised concerns about safety,” he said.
While the current safety profile of the short-acting analogues “suggests no independent effect on retinopathic change or carcinogenesis,” there is reason to be cautious about long-acting analogues until more data are available, he said.
A study published in 2000 comparing the toxicopharmacologic properties of insulin analogues showed that glargine had a six- to eightfold increase in IGF-1 receptor affinity and associated mitogenic potency compared with human insulin, he noted. The two rapidly acting insulin analogues resembled human insulin on all parameters, except for a slightly elevated IGF-1 receptor affinity for lispro (Diabetes 2000;49:999–1005).
A possible association of lispro with proliferative retinopathy was “put on the map” more than a decade ago when Dr. John L. Kitzmiller and his colleagues reported that 3 out of 10 lispro-treated patients with no detected background retinopathy progressed to proliferative retinopathy during pregnancy (Diabetes Care 1999;22:874–5), Dr. Carpenter said.
Studies and commentary since then have shown no adverse impact of insulin analogues on the progression of diabetic retinopathy in pregnant patients, he said. A Finnish study of 69 pregnant women treated with either lispro or conventional human insulin, for instance, showed no significant differences in retinopathy progression (Diabetes Care 2003;26:1193–8).
Experts have also noted that the hemaglobin A1c levels in women in Dr. Kitzmiller's series were initially high, indicative of poor prepregnancy metabolic control, which raises the question of whether the rapid change to euglycemic control may have been the primary contributor to the advancing retinopathy among these patients rather than a specific lispro effect.
Regardless of insulin choice, rapid tightening of glycemic control is among the predictors of proliferative diabetic retinopathy during pregnancy, along with the duration of diabetes, HbA1c or plasma glucose at the onset of care, and other factors, he said. “We really ought to have informed consent for the rapid achievement of normal blood sugars from a nonpregnant state to a pregnant state … for patients who are in denial before becoming pregnant, with very poor metabolic control, and who are then enlisted in very careful management to dramatically improve their glycemic control,” Dr. Carpenter said in an interview. “These are the patients we know are at risk of worsening retinopathy.”