You can access the Consensus publication here
International consensus recommendations position CGM Time in Range (TIR) metrics as next-generation endpoints in diabetes clinical trials
Time In Range (TIR) is becoming established as an empowering continuous glucose monitoring (CGM) metric that can inform clinical decisions and support self-management of diabetes.1,2 It is also emerging as a valuable new clinical trial endpoint that can add value to research findings by providing deeper and nuanced insights into diabetes treatment efficacy.3
An international consensus of recommendations for the standard use of CGM time in range data in clinical trials was published in January 2023 in Lancet Diabetes & Endocrinology following expert panel discussions.3 The diaTribe foundation, a non-governmental organisation in the US, invited panel members including expert healthcare professionals experienced with using CGM in clinical trials, academic institutions, the US Food and Drug Administration (FDA), the National Institutes of Health, the American Diabetes Association (ADA), and various other associations.3 Traditionally, prospective studies of novel diabetes treatments have reported HbA1c as the measure of efficacy. These landmark recommendations recognise the overwhelmingly positive perception of TIR expressed by HCPs during independent global quantitative research and indicate the measure’s potential to help shape the future of diabetes treatment.2,3
The potential benefit of meticulously monitoring glucose during diabetes clinical trials is seemingly obvious; it yields an abundant supply of data offering detailed insight into treatment efficacy, collected in a relatively effortless manner.4 CGM could also reduce the duration of studies and number of participants required, but it’s important that the glucose monitor is used correctly by participants and that appropriate endpoint metrics are selected.4 The consensus recommendations offer clear guidance for both. In particular, they advise how to use core CGM metrics in prospective clinical studies — either as study endpoints or to complement HbA1c — to support decision makers authorising new treatments.3
The recommendations have been endorsed by prominent global stakeholders including the ADA, the Association of Clinical Endocrinologists, the Association of Diabetes Care and Education Specialists and the European Association for the Study of Diabetes.3 For the ADA, the publication represents a big step forward in the field and could further expand the use of CGM and Time in Range.5
Why continuous glucose monitor data add value to diabetes clinical trials
HbA1c has been fundamental to the approval of diabetes treatments but its limitations are increasingly acknowledged, particularly its lack of insight into glycaemic variability, asymptomatic hypoglycaemia and detailed hyperglycaemia.3,6,7 At the same time, the benefits of CGM are becoming widely known and CGM is now recommended by the ADA for all patients with type 1 diabetes and those with type 2 diabetes on basal insulin treatment.6 By generating 24-hour interstitial glucose readings, CGM can provide comprehensive and detailed insight into glycaemic variability and hypo- and hyperglycaemic episodes of all magnitudes.1,6
Adoption of both real-time CGM (rtCGM) and intermittently scanned CGM (isCGM), has grown rapidly over the past few years.1 Consensus recommendations now identify ten core CGM metrics, including TIR, as most helpful both for managing diabetes and as validated clinical endpoints.1,3 This means TIR could influence the success of next-generation diabetes treatments, and we at Novo Nordisk are firmly committed to including the metric as a key measure of efficacy in prospective clinical trials. With increasing use of TIR as an endpoint, we expect regulators, payors, and clinical guidelines will increasingly look to TIR efficacy parameters when making decisions and recommendations about treatments including insulin, GLP-1, and oral antidiabetic drugs (OADs).
Continuous glucose monitor selection and use
CGM devices and their data can have two useful roles in clinical trials; comparative monitoring of the clinical impact of the intervention treatment and recording of endpoint metrics.3 Below are some of the most important recommendations for use:3
- The CGM system should have features appropriate to the aims of the study, the protocol and the diabetes management needs of participants.
- CGM can be configured in unblinded (wearer can see their real-time glucose levels and take corrective action if necessary) or blinded mode (CGM data are transmitted for unbiased review).
- Blinding reduces any confounding effects and is more appropriate when participants are new to continuous glucose monitoring, but unblinded CGM should be used where clinically necessary.
- Data should ideally be collected throughout the entire study period, but may be collected at baseline, prespecified timepoints during the intervention or at study end.
Use of CGM time in range metrics in clinical trials
TIR indicates the proportion of each day that a person with diabetes spends with their glucose levels in target range (TIR; 70–180 mg/dL [3.9–10.0 mmol/L]). TIR is accompanied by time above range (TAR) and time below range (TBR), which are defined by International Consensus Time in Range guidelines.1
The recommendations prioritise incorporating TIR, TAR and TBR within endpoints for all prospective clinical studies using CGM, and highlight the particular importance of TBR, which indicates hypoglycaemia.3 The metrics can be categorised into temporal subgroups (e.g. within a 24 hr period, diurnal or nocturnal), and classified as core, secondary or composite endpoints.3 The various TIR endpoints recommended for use are shown below:3
Recommended TIR endpoints for clinical trial
| Core endpoints | Units and quantity |
|---|---|
| Time in range 70-180 mg/dL (3.9-10.0 mmol/L) | Percentage of time below range; amount of time (hours and minutes) |
| Time below range <70 mg/dL (<3.9 mmol/L), including readings of <54 mg/dL (<3.9 mmol/L) | Percentage of time below range; amount of time (hours and minutes) |
| Time below range <54 m/L<3.0 mmol/L) | Percentage of time below range; amount of time (hours and minutes) |
| Time above range >180 mg/dL (>10.0 mmol/L), including readings of >250 mg/dL (>13.9 mmol/L) | Percentage of time above range; amount of time (hours and minutes) |
| Time above range >250 mg/dL (>13.9 mmol/L) | Percentage of time above range; amount of time (hours and minutes) |
| Coefficient of variation | Percentage coefficient of variation intraday (ie, within 24 h) and interday (ie, over multiple days) |
| SD of mean glucose | SD |
| Mean sensor glucose | mg/dl (mmol/L) |
| Secondary endpoints (continuous outcomes) | |
| Time in tight range 70-140 mg/dL (3.9-7.8 mmol/L) | Percentage of time in tight range; amount of time (hours and minutes) |
| Change in Glucose Management Indicator | Absolute mean change in mmol/mol or percentage |
| Extended hypoglycaemic event rate <70 mg/dL (<3.9 mmol/L) | Number of events with sensor glucose <70 mg/dL (<3.9 mmol/L) lasting at least 120 min; event ends when glucose returns to ≥70 mg/dL (≥3.9 mmol/L) for ≥15 min |
| Extended hyperglycaemic event rate >250 mg/dL (>13.9 mmol/L) | Number of events with sensor glucose >250 mg/dL (>13.9 mmol/L) lasting at least 120 min; event ends when glucose returns to ≤180 mg/dL (≤10.0 mmol/L) for ≥15 min |
| Secondary endpoints (binary outcomes) | |
| Proportion of participants with time in range 70-180 mg/dL (3.9-10.0 mmol/L) for >70% of each day | Percentage of participants |
| Proportion of participants with time in range 70-180 mg/dL (3.9-10.0 mmol/L) with ≥5% points improvement from baseline | Percentage of participants |
| Proportion of participants with time in range 70-180 mg/dL (3.9-10.0 mmol/L) with ≥10% points improvement from baseline | Percentage of participants |
| Proportion of participants with time below range <70 mg/dL (<3.9 mmol/L) for <4% of each day | Percentage of participants |
| Proportion of participants with time below range <54 mg/dL (<3.0 mmol/L) for <1% of each day | Percentage of participants |
| Proportion of participants with time above range >180 mg/dL (>10.0 mmol/L) for <25% of each day | Percentage of participants |
| Proportion of participants with time above range <250 mg/dL (>13.9 mmol/L) for <5% of each day | Percentage of participants |
| Composite endpoints | |
| Proportion with improvement in HbA1c >0.5% points without an in increase in TBR <54 mg/dL (<3.0 mmol/L) oF >0.5% | Percentage of participants |
| Proportion of participants with >10% points improvement in percentage of time in range 70-180 mg/dL (3.0-10.0 mmol/L) without an increase in time below range <54 mg/dL (<3.0 mmol/L) of >0.5% | Percentage of participants |
| Proportion of participants with mean glucose <154 mg/dL (<8.6 mmol/L) and <1% time below range <54 mg/dL (<3.0 mmol/L) | Percentage of participants |
| Proportion of participants with >70% time in range 70-180 mg/dL (3.0-10.0 mmol/L) and <4% time below range <70 mg/dL (<3.9 mmol/) | Percentage of participants |
| Proportion of participants with >70% time in range 70-180 mg/dL (3.0-10.0 mmol/L) and <1% time below range <54 mg/dL (<3.0 mmol/) | Percentage of participants |
Table 3: Recommended CGM-derived endpoints for clinical trials
Adapted from Battelino et al, 20233
TIR is immediately responsive to changes in medication, diet, and lifestyle and so is a useful measure of glucose control that can be visualised in a clinical trial setting, with TBR and TAR as accompanying endpoints.3
Assessing treatment efficacy using Time in Range
The efficacy of a particular treatment can be evaluated by examining the differences in percentage TIR between treatment and control groups.3 A difference of ≥5% (absolute percentage points) is defined as clinically meaningful for a study participant, whereas a difference of 3% between a treatment and control group shows a clinically meaningful effect on glycaemia.3 Studies can be powered to detect this magnitude of change.3
The landmark InRange trial was the first to use Time in Range as a primary endpoint
The InRange trial was the first to use TIR as the primary endpoint, and results demonstrate the depth of insight that can be gained by using CGM.8,9 InRange was a comparison study of second-generation insulin analogues glargine 300 U/mL (IGlar-300 ) and degludec 100 U/mL (IDeg-100) for type 1 diabetes management. CGM data were collected over 20 consecutive days and the primary endpoint was percentage TIR at 12 weeks.8,9
Both treatments led to similar outcomes (52.74 and 55.09% TIR, respectively) and it has been reported that the rate of hypoglycaemia (TBR) detected by continuous glucose monitoring was 3–6 times higher than detected with a self-measured blood glucose monitor in the same data-collection period.3,9 When presenting these results at ATTD 2022 one of the investigators, Dr Richard Bergenstal, commented that only 11% of clinical trials were currently employing CGM and strongly advocated for increased use.10 The use of TIR endpoints can identify selective treatment targets related to hypoglycaemia, and the InRange findings not only serve as a benchmark for their use but also illustrate this potential.3,8,9
HQ23NNGG0004 July 2023
References