Miriam E Tucker

September 12, 2017

LISBON, PORTUGAL — Use of real-time continuous glucose monitoring (CGM) produces significant long-term improvement in HbA1c among adult patients with type 1 diabetes, regardless of whether they use insulin pumps or multiple daily insulin injections (MDI), new research shows.

Two-year data from the Comparison of Different Treatment Modalities for Type 1 Diabetes Including Sensor-Augmented Insulin Regimens (COMISAIR) study were presented September 11, 2017 here at the European Association for the Study of Diabetes (EASD) 2017 Annual Meeting by Jan Šoupal, of the Charles University, Prague, Czech Republic.

The real-world, nonrandomized study compared changes in HbA1c among 65 patients using one of four treatment regimens: insulin pumps with or without real-time CGM (n = 15 and 20, respectively) and MDI with or without real-time CGM (12 and 18, respectively), with all patients using self blood glucose (finger-stick) monitoring. 

At 2 years, use of CGM, with or without the pump and with MDI provided "significant, comparable, and stable improvement of HbA1c."

In fact, the combination of CGM and MDI was actually more effective than insulin-pump therapy alone in reducing HbA1c, Dr Šoupal said.

"In this trial, we're showing it's not so important how insulin is injected. What's more important is how patients monitor their glucose," he told Medscape Medical News, adding that the 2-year data are the longest-ever for a trial of real-time CGM.

[Real-time CGMs are the systems that patients use, allowing them to view their data continuously, as opposed to the blinded CGM systems that physicians download data from in their office, thereby reviewing the data in retrospect].

Although the findings were presented in a Dexcom-sponsored satellite symposium at the EASD meeting, the company didn't fund the trial and not all patients in the trial used Dexcom's CGM. Rather, COMISAIR was sponsored by the Agency for Healthcare Research of the Czech Republic, with no industry support, Dr Šoupal stressed.

He also explained that the study was not randomized to minimize dropouts, since patients tend to leave randomized trials when the treatment doesn't meet their needs, and COMSAIR was planned as a long-term trial lasting up to 5 years.

Also, he believes there was an ethical consideration, since several professional organizations now advise that real time-CGM be at least considered in all patients with type 1 diabetes, so it can be viewed as standard of care.

"Is CGM in type 1 diabetes like metformin in type 2?" he queried.

CGM Not Quite Ready for Standard of Care

Asked to comment, Dr Pilar Beato-Vibora, a diabetes consultant at Badajos University Hospital, Spain, said that the study's findings align with her experience. "I think these are really provocative data regarding the benefits of CGM."

Dr Beato-Vibora has several patients who use real-time CGM along with MDI rather than pumps for a variety of reasons. Some patients view the pump as more complicated than the CGM, or they have hypoglycemic unawareness and primarily seek out the CGM for the low-blood-glucose alarms, she noted.

However, she doesn't entirely agree with the strong endorsement of CGM as standard of care for all type 1 diabetes patients at this point.

"We will see — maybe in the future CGM will be the standard of care for type 1 diabetes, but at the moment I don't think all type 1 diabetes patients need CGM or are willing to use it. But I think that as clinicians our responsibility is to offer each patient what they need, want, or could benefit from."

She also noted that, despite the rationale for the lack of randomization, it is a study limitation. "The patients got the treatment they wanted. A real-life study is good, but it's not comparable to a randomized trial from a statistics point of view and so it's not applicable to all patients."

Second Year of Data Further Supports CGM

The 1-year COMISAIR data were published in September 2016 (Diabetes Technol Ther. 2016;18:532-538).

At the study start, all patients received a 4-day training program similar to the UK's Dose Adjustment for Normal Eating (DAFNE), in which they received instructions in timing and dosing of preprandial insulin, prevention of hypoglycemia, carbohydrate counting, and bolus calculations. Patients could then choose the treatment group they preferred.

Of the 65 enrolled, 60 completed 2 years of the study, with all of the patients using CGM using it more than 70% of the time, Dr Šoupal reported.

At 2 years, HbA1c levels in the no-CGM groups had not changed significantly, from 8.4% to 8.0% for the pump group (= .07) and 8.3% to 8.1% in the injection group (NS).

However, both real-time CGM groups experienced significant HbA1c reductions, from 8.2% to 7.1% in the pump/CGM group and 8.5% to 7.2% in the MDI/CGM group (both < .001).  

More than half (54%) of the two CGM groups achieved an HbA1c below 7% at 2 years, compared with 13% of the no-CGM groups. At the same time, the proportion of time spent in hypoglycemia was reduced from 8% to 6% (< .01) in the two CGM groups, with no significant difference between them.

There were no episodes of severe hypoglycemia with real-time CGM, vs two in each of the no-CGM groups. However, there was one episode of ketoacidosis in the pump/CGM group.

Glycemic variability was also significantly reduced in the two CGM groups (< .05), with no difference between them.

Tips for Using CGM, Including Avoiding "Alarm Burnout"

Dr Šoupal offered some practical suggestions for success in using real-time CGM with MDI: Patients should be willing to take more bolus doses to correct high blood glucose levels, use short-acting insulin for boluses — including pens that can deliver half units for lean patients — and consider different types of basal insulins from those the patient may already be using.

In addition, for those with pumps or MDI, he advised that careful attention be given to the appropriate CGM settings for high- and low-glucose alarms and target ranges in order to avoid "alarm burnout." This can be tricky, he noted, since alarm settings that are "too benevolent" can impair glycemic control, but so can excessive alarms.

"We should not forget there is an important psychological aspect.…In clinical practice, if the patient has not improved HbA1c in 3 or 6 months from real-time CGM initiation, this is one of the first things I check," he said.

Overall, he said, the choice of treatment — pump vs MDI, real-time CGM or just finger-stick monitoring — must be based on individual characteristics, including physical-activity level, hypoglycemia awareness, and personal preference about wearing devices.

"There is a huge variety of patients with type 1 diabetes.…Individualization is crucial."

Dr Šoupal is an advisor, consultant, and/or speaker for Novo Nordisk, Eli Lilly, Boehringer-Ingelheim, AstraZeneca, Medtronic, Roche, and Dexcom. Dr Beato-Vibora has no relevant financial relationships.  

Šoupal J. CGM combined with either MDI or CSII is superior to standalone MDI or CSII in T1D: 2 years of follow-up in the COMISAIR study [Dexcom-sponsored industry symposium]. European Association for the Study of Diabetes 2017 Annual Meeting. September 11, 2017. Lisbon, Portugal.

For more diabetes and endocrinology news, follow us on Twitter and on Facebook.


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.