Anne L. Peters, MD


October 28, 2015

This feature requires the newest version of Flash. You can download it here.

Hi. I'm Dr Anne Peters, and today I'm going to discuss using sodium/glucose cotransporter 2 (SGLT2) inhibitors in people with type 1 diabetes.

First off, let me say that managing type 1 diabetes is difficult for both the provider and for the patient because it is so variable. People with type 1 diabetes have blood sugars that go up and down. Often, patients themselves can't really predict what is going to happen. They may feel they do the same thing and eat the same thing, yet their blood sugars are like a sine wave, up and down all the time.

The SGLT2 inhibitor class is a class of drugs designed for treating people with type 2 diabetes, but they work in people with type 1 diabetes because they work in a beta-cell–independent way. They basically make individuals urinate out more glucose, which made experts believe these drugs would also be effective in people with type 1 diabetes.

From the beginning, I've used SGLT2 inhibitors—off-label—in my patients with type 1 diabetes, and I really saw a benefit. Where my patients' blood sugar used to be up and down and up and down, it became much smoother. The variation became less, and patients were able to reduce their insulin. They were able to lose some weight, and they had fewer episodes of hypoglycemia.

Then, we found—and this was published not that long ago—that people on SGLT2 inhibitors with type 1 diabetes primarily, and in some cases with type 2, developed episodes of diabetic ketoacidosis (DKA).[1] In some cases, these patients were euglycemic, meaning that their blood glucose levels were 250 mg/dL or less, and they did not recognize that they were becoming ketotic. People with type 1 diabetes are used to seeing their blood sugars going high and then worrying about DKA, but in the case of patients on these agents, they often did not see that increase.

All About the Dose?

A phase 2 trial looking at the SGLT2 inhibitor canagliflozin in the treatment of type 1 diabetes was just completed.[2] I want to point out something very important when it comes to looking at data in patients with type 1 diabetes: If you reduce the glucose variability, you're going to reduce both highs and lows, so the medication's impact on A1c may be somewhat smaller than you would expect because you're taking away some of the lows. The result is better for the patient, but overall, the A1c may not change that much.

In this trial, they used placebo and compared it with 100 mg of canagliflozin and 300 mg of canagliflozin. They found there was a significant reduction in the endpoint that encompassed A1c reduction, a reduction in rates of hypoglycemia, and weight loss.

The overall A1c reduction was, on the average, about 0.4%, but that happened in conjunction with the other two—weight loss and a reduction in hypoglycemia—which to patients and to me is actually quite significant, because it makes type 1 diabetes easier to manage. Patients like it.

On the other hand, there is that concern about diabetic ketoacidosis. In our study, we showed that yes, rates of diabetic ketoacidosis are increased in patients with type 1 diabetes when taking an SGLT2 inhibitor. The rate was close to 10% on the 300-mg dose of canagliflozin, but closer to 5% (about one half as much) in the patients taking the 100-mg dose. Rates of diabetic ketoacidosis were zero on the placebo-treated dose.

What we believe is that the initial dose of SGLT2 inhibitors—that dose meant for patients with type 2 diabetes—is too high a dose for patients with type 1 diabetes. We need to use lower doses, and in fact, lower doses are going to be studied to see if we can reduce the risk for DKA.

Managing the Risks in the Meantime

If you ask me whether I still use SGLT2 inhibitors off-label in my patients with type 1 diabetes, the answer is "yes." It's "yes" because my patients want to use it. Even those who went into DKA still want to stay on the drug because they love that reduction in variability.

I have developed a way to use these agents that I feel is safe, and in fact, since I have been doing it, I haven't had any patients with type 1 diabetes go into DKA.

First of all, I start patients out with ketone monitoring. Less than 20% of adult patients with type 1 diabetes do any ketone monitoring, and yet we teach that often for children. It is not that hard, and in fact, you can get blood ketone monitoring test strips.

You can give patients a meter, so that they check their serum ketone levels first thing in the morning when they check their glucose. Anything that is less than 0.6 mmol/L, I consider acceptable; above 0.6, I start worrying that someone is going to become ketotic.

I have patients test their fingerstick ketone level each morning for a week so that I can get a baseline, and some people actually run a little more ketotic than others, but I want to see whether it is 0.1 or 0.2 or 0 when they test. Then, I start with not only the lowest dose of the SGLT2 inhibitor, but I have patients break the tablet, and I will have them use one half of a dose or even one third of a dose—so a tiny dose, meaning 25-50 mg of canagliflozin.

I reduce the basal insulin by 50%, and then I have the patient keep testing for ketones as long as I'm dose-adjusting. As long as their insulin dose is going down, I have them check their morning ketones.

Once the patient is stable on a new lower dose of insulin, I don't need them to continue to test their serum ketones unless something changes—for example, if they become ill, get viral gastroenteritis, become dehydrated, start exercising more, or suddenly go on a diet and cut way back on their insulin dose. Anything that I think could put them at risk for developing ketoacidosis is a time when I am going to have them test themselves. If their ketones in the morning are positive, I have them not take the medication that day, and then I believe they are safe.

Should a patient develop strongly positive ketones on an SGLT2 inhibitor, treat the patient by having them consume simple carbohydrates and take insulin, because a lot of the time, the glucose levels are not going to be elevated.

If a patient has a glucose of 125 mg/dL and you ask them to take insulin, they are going to fall too low. I make sure that they have some sort of sugar-containing sports drink or soda pop on hand, so that they can increase their carbohydrate intake and raise their glucose level and they can take some more insulin to overcome the risk for ketosis.

I do it this way, having the ketone monitor as part of what I do, and making sure patients understand that if they develop strongly positive ketones, they need both carbohydrate-containing fluids and insulin, and then follow up with their ketone testing.

I really think that SGLT2 inhibitors are a very useful drug for patients with type 1 diabetes as well as, obviously, patients with type 2 diabetes. It is important to be aware of the risk. There may be some crossover patients—patients who seem to have type 2 diabetes, but actually are more of a LADA [latent autoimmune diabetes in adults] patient shifting over to more of a type 1 patient, who could develop ketosis. We need to keep in mind that not all type 2s are type 2s.

That being said, as long as both we and our patients are aware of this risk, I think that these drugs are going to become increasingly used in both type 1 as well as type 2 diabetes patients.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.