Adding to the arsenal – what is the new oral add-on for Type 1 Diabetes?

For the first time, an oral add-on treatment, Dapagliflozin (aka Forxiga from Astra Zeneca) has received regulatory approval for managing type 1 diabetes. While this is a first, what is it, what does it do, and who can get it?

What is it and what does it do?

Dapagliflozin is an oral tablet. It is taken once daily.

It is one of a family of SGLT-2 inhibitors, which means that it inhibits certain transporter proteins that are responsible for about 90% of the glucose re-absorption in the kidneys.

By blocking these transporters, instead of reabsorbing glucose into the blood, it’s simply passed into the urine. There’s a great deal more to this than that simple explanation, but that’s essentially what happens. 

As a person with type 1, you’d be familiar with this mechanism of non-absorption from urine testing. Essentially what Dapagliflozin does is allow you to pee out glucose when your glucose level is at normal levels, not only when it’s particularly high. 

The effects of this, shown in the DEPICT-1 and DEPICT-2 studies are very interesting. (DEPICT stands for “Dapagliflozin Evaluation in Patients With Inadequately Controlled Type 1 Diabetes”, but let’s not go there with the naming shall we…)

The baseline HBa1C data for the group was between 7.7% and 11.0%.

Essentially, they saw:

  • A reduction in Hba1c of around 0.4 percentage points on average
  • A reduction in insulin use by between 7% and 12% (dependent on dose)
  • A reduction in weight of between 3kg and 4kg per person (dependent on dose)
  • A reduction in mean value of 24hr CGM of between 10mg/dl and 12mg/dl (dependent on dose)
  • A reduction in Mean Amplitude of Glucose Excursions (explanation of that here) by between 9mg/dl and 17mg/dl (dependent on dose)
  • An increase in time in range (3.9mmol/l-10mmol/l/70mg/dl-180mg/dl) of between 5% and 9% (dependent on dose)

Given all of these, it looks like a bit of a wonder drug. Well, almost.

There’s also a risk of increased DKA with it (between 2x and 3x greater risk of DKA compared to the control group in the 12 month trial, and “significant” in the 6 month trail, where none of the control group suffered DKA). What’s not clear from the trials data is whether these were euglycaemic, which has been observed in users with type 2.

The question is why? In type 2 patients, this was believed to be due to reductions in insulin dosage that maintained a stable glucose level, but was not enough to counter ketogenesis, although, there’s not a great deal of good data on this. The biggest issue here is that this presented as euglycaemic DKA, instead of hyperglycaemic DKA. 

The process is described in this image:

which is taken from the an ADA paper discussing this issue. The same paper also discusses that euDKA when using SGLT2 inhibitors is still pretty rare, and should be detectable using existing monitoring tools. 

The other thing that is mentioned is that activities that are generally likely to increase free fatty acids may be accentuated by any SGLT2 inhibitor, e.g. a low carb diet or reduction in insulin. So it’s a case of being aware. 

It’s also worth noting the actual incidence of these issues. It’s between 2% and 4% of users, compared with 1% of the control, so  while it’s a 3x risk increase, it’s not an enormous change in absolute terms.

There is also (according to other data from MHRA with people with type 2) a slight increase in risk of Fournier’s Gangrene (which sounds like a very unpleasant experience).

Who can get it?

Now that’s an interesting question. According to the EMA approval, this is the list of things that prescribers need to be aware of in Type 1 patients that indicate it is unsafe to start Dapagliflozin:

  • Patients with low insulin needs.
  • Patient not on optimal insulin dose or who have recent issues with noncompliance or recurrent errors with insulin dosing and who are unlikely to maintain adequate insulin dosing.
  • Patients with increased insulin requirements due to acute medical illness or surgery.
  • Patients who insist on maintaining caloric restriction, carbohydrate restriction or ketogenic diet or who chronically under-dose insulin (e.g. in order to remain in a lipolytic state).
  • Patients with recent or recurrent history of DKA.
  • Patients with elevated ketones levels (BHB reading is greater than 0.6 mmol/L or urine ketones one plus (+)). If ketones are elevated (blood beta-hydroxybutyrate reading 0.6 mmol/L or greater), treatment with dapagliflozin should not be started until the ketone levels are normal
  • Patients unable or unwilling to monitor ketones.
  • Patients with excessive alcohol consumption or who use illicit drugs.

In addition, they don’t recommend that it is prescribed to those with a BMI below 27 kg/m2

So there are a few that shouldn’t be prescribed it. There’s going to be a clear need for assessment before issuance and not everyone will get it.

Does that sound like something else that was made available in Type 1 care recently, such as the Libre?

How easy will it be to get this prescribed?

Now that’s a good question. I’ll start with the cynical point of view. Easier than Libre…

Why do I say that? Because it’s a drug, and many clinicians find prescribing drugs easier than technology, even basing it on the limitations in the last section. 

NICE has produced the following guidance for prescribing, which is a fairly in depth document covering what is described in the previous section.

And to top that off, Dapagliflozin is £37 per month on the UK tariff compared to £70 per month for Libre, with roughly half the reduction in Hba1C. Where does this look like it might go?

Or in other words, I’ve no idea how easily it will be prescribed. It’s now indicated for Type 1, it’s fairly easy to prescribe (although NICE does recommend that users know how to recognise DKA and have had structured education) and there are clearly some concerns over safety. I think it’s a case of wait and see.


It’s good to see a new tool in the type 1 treatment armoury, and a relatively inexpensive one at that. It remains to be seen whether it’s effective and whether NICE’s concerns about the likelihood of the UK type 1 population being more likely to suffer DKA as a result of using it. 

I’m also intrigued as to what the impact on the kidneys of processing all that glucose in the long term will be (there’s no data around that as of yet, but there are renal contraindications in the use recommendations). 

So while I cautiously welcome an additional “technology” to the game, I’m not 100% convinced that it’s something that I’d want to take.

As a footnote, AstraZeneca announced that the FDA refused approval for the same drug on 15th July. Details of why are not in the public domain. 


  1. Thanks for sharing about this. From the sounds of it, this drug may be useful for me on a medium term basis. I’m going to ask about it next time I’m at Guys.

  2. Thank you for your interesting and thoughtful presentation.
    I note that you mention the risk of euglycaemic DKA increases under certain conditions, including “..those who underdose on insulin ( e.g. to maintain a lipolytic state)” – from reading, experience but not medical opinion or advice: to me it seems less likely that people underdose on insulin for this reason than the fact that the underdosing is due to conditions like diabulimia. And diabulimia would appear to be a very important contraindication to the drug in question.
    The wording you choose above also could be interpreted as a criticism of those TIDs that choose to undertake a ketogenic diet. If they are demonstrating a HBA!C of say, >4.8% and less than 5.8%, keep well hydrated and have minimal BSL variation that would suggest their insulin usage is satisfactory.

    • Thanks Tony. If you read the article again, you’ll see that the points you are picking up on are part of the EMA guidance related to prescribing and is a direct quote, rather than wording that I’ve chosen.

      The mechanism by which SGLT-2 works increases the risk of DKA. That’s clear from the trials. Combining it with a state of ketosis could push a person following a ketogenic diet into euglycaemic DKA which is why both the EMA and NICE have concerns and have included that in their guidance.

      If someone has an Hba1C between 4.8% and 5.8%, I’m not sure they are indicated for the use of this add-on drug, so I’d argue that it’s a moot point.

      • Thank you. I would argue that the EMA guidelines need better wording and some include a term like diabulimia or similar as they would appear to be at risk of the usual DKA.

        • I’m afraid I have to disagree. Someone can be underinsulinised without having an eating disorder, and the key point in that is that it’s insulin stopping DKA, so anyone who can be considered to not be taking enough insulin should not be given a drug that increases the risk of DKA.

          I also wouldn’t correlate someone following a ketogenic diet with someone who is underinsulinised. The lack of glucose variability and Hba1C outcomes would suggest completely the opposite, that they are in fact, perfectly insulinised.

Leave a Reply

Your email address will not be published.