DM in Pregnancy and AI - Ben Smarr Ph.D. and Sandy Ramos M.D.

Let me start off by just framing the question, why is insulin management so difficult during pregnancies complicated by diabetes? Well, it's super important to have optimal glycemic control, which is much tighter than outside of pregnancy.

And this is important because we know that hyperglycemia is associated with adverse maternal and fetal complications.

However, it's very difficult to manage during pregnancy because there's a continuum that we see with rapid changes to insulin resistance, which is mediated a lot through placental hormones.

But in the first trimester, for example, there could be a risk of hypoglycemia, which can really adversely affect the mother.

And then in the second and third trimester, rising insulin resistance, which usually peaks in such a way that for type 1 diabetes patients, they can be on 300% higher doses of insulin than pre-pregnancy.

So traditionally, the way we optimize care was through self-monitoring of blood glucose, but this entailed finger sticks up to four to seven times a day, which was painful and burdensome and limited with regards to the data.

So you can see on this figure, we have a patient who's monitoring her glucose, and it looks like the red circles are within the desired range.

However, there's some missed opportunities to check the glucose, as you can see with the open white circles.

However, wouldn't it be wonderful if we were able to see a 24-hour continual set of data where you can find opportunities for treatment with mixed, missed hyperglycemia and also some missed hypoglycemia.

probably linked to over-treatment with insulin.

And so this is where continuous glucose monitors come into play, and this has really changed the way we manage diabetes.

As you know, it measures institutional glucose continuously every one to five minutes, and it's attached to the skin with a sensor, and then we usually use a phone nowadays for receiving that information.

And really this tool became useful in the cohort of patients with type 1 diabetes.

And this was the seminal sort of article about CGM use in pregnancy in those with type 1 diabetes, where they found that CGM alone improved neonatal outcomes with lower rates of large for gestational age infants, hypoglycemia, and NICU stay.

And not many women needed to be wearing CGM in order to see these effects.

And so, the American Diabetes Association has come up with glycemic control targets utilizing CGM.

And you can see here, this is a type 1 diabetes and pregnancy management algorithm.

However, for GDM and for type 2 diabetes, they say it's evidence, it's actually based on expert opinion and not evidence-based.

And honestly, this is where we need more research.

And the American Diabetes Association does not recommend the use of CGM for type 2 and gestational diabetes in pregnancy until we have this data, although people are using it clinically, as you know.

And so we have more than 11 actively recruiting trials currently, and we're lucky enough to be one of those centers that is doing a clinical trial on the use of CGM in type 2 diabetes in pregnancy, and we're currently enrolling.

We also have some trials underway for gestational diabetes here in California, including the CAPL trial.

And some studies are actually using CGM for postpartum screening for type 2 diabetes after gestational diabetes, which is the postpartum trial.

But there's possible uses outside of diabetes care with this technology, which is super exciting.

So investigators have been able to actually define what is normal glycemia in pregnancy.