Layout

LEARNING FROM PRACTICE Dapagliflozin: Clinical practice compared
with pre-registration trial data
ANDREW P MCGOVERN1-3, NINA DUTTA1, NEIL MUNRO1-4, KENNETH WATTERS1,2,4, MICHAEL FEHER1,2,4 Abbreviations and acronyms
Background: Dapagliflozin is the first sodium-glucose co-
transporter 2 (SGLT2) inhibitor to be approved in Europe
glucagon-like peptide-1 receptor agonist and represents a new class of agents developed as oral
general practitioner diabetes medications. Improved glycaemic control and
glycated haemoglobin weight loss have been demonstrated in clinical trials but
oral antidiabetic drug effectiveness outside of the trial environment has not yet
standard error of the mean sodium glucose co-transporter 2 urinary tract infection Method: A systematic clinical case note audit of type 2
diabetes patients initiated on dapagliflozin in a diabetes
specialist outpatient centre of a London teaching hospital.
duced into UK clinical practice in December 2012.
Results: Of the 96 people included, 42% had a reduction
The pre-registration clinical trials of dapagliflozin have in glycated haemoglobin (HbA
demonstrated improvements in glycaemic control when used as 1c) of >1%, 29% had no
reduction; 15% had weight loss >5kg, 3% had weight loss
monotherapy1 and with metformin,2-4 sulfonylureas,5,6 or >10kg and 24% had no weight reduction. Improvements
insulin.7,8 There have also been clinically meaningful reductions in weight and blood pressure. However, these clinical trials were 1c, weight, and blood pressure were consistent with
those reported in clinical trials. The rate of discontinuation
performed on selected patient groups and therefore the trial of dapagliflozin due to side effects (22%) was higher than
results may not be matched in "real-world" clinical practice. The reported in trials (3-4%), but 52% of people tolerating
Association of British Clinical Diabetologists nationwide exe- dapagliflozin were able to stop or reduce one or more
natide audit demonstrated a differing real-world efficacy profile other diabetes medications.
from that reported in registration clinical trials.9 Conclusions: Dapagliflozin is effective at improving gly-
We hypothesised that routine use of dapagliflozin in clinical caemic control. It also reduces blood pressure, results in
practice may vary in efficacy and side effect rates from those weight loss, and reduces the need for concomitant
reported in clinical trials. We present the results of a phase 4 diabetes medications. However, it is not as well tolerated
clinical study in the form of treatment observation of type 2 in real-world patients as in participants of clinical trials.
diabetes patients treated with dapagliflozin.
Br J Diabetes Vasc Dis 2014;14:138-143
Key words: dapagliflozin, real-world, SGLT2, type 2 diabetes
The design was a real-world, observational, non-randomisedsingle centre study. The study comprised of a systematic clinical audit of notes and electronic records from people with type 2 Dapagliflozin is the first SGLT2 inhibitor drug to gain approval in diabetes with dapagliflozin initiated in the diabetes specialist Europe for the management of type 2 diabetes and was intro- outpatient centre of a London teaching hospital.
The records of all people prescribed dapagliflozin before 14th May 2014 were analysed (n=122). People with no follow-up 1 Beta Cell Diabetes Centre, Chelsea and Westminster Hospital, London, data were excluded (n=26).
Anonymised data were collected on patient demographics, 2 Diabetes Therapies Evaluation Network, Chelsea, London, SW3 5HW, UK3 disease profile, concurrent medications, and outcomes. The Department of Health Care Management and Policy, University of SurreyGuildford, GU2 7PX, UK demographic and disease information collected comprised 4 Warwick Medical School, Warwick University, Coventry, CV4 7AL, UK patient age, gender, diabetes type and duration, renal function,and baseline measurement of HbA Address for correspondence: Dr Michael Feher
1c, weight, BMI, and blood Beta Cell Diabetes Centre, Chelsea and Westminster Hospital, pressure. All recorded clinical measurements were taken from 369 Fulham Road, London, SW10 9NH, UK routine data. Where no baseline measurement was available for Tel: +44 (0)20 3315 8000 x8821 HbA1c, weight, or blood pressure, the most recent available measurement within the preceding three months was used.
A small number of people had dapagliflozin stopped by their THE BRITISH JOURNAL OF DIABETES & VASCULAR DISEASE LEARNING FROM PRACTICE Table 1 Changes in outcome measures at final follow-up visit in patients with type 2 diabetes treated with dapagliflozin
Outcome measure (units)
Mean baseline
Mean change from
95% confidence
value (SEM)
-0.84 (0.23)
-9.2 (2.5)
-2.2 (0.5)
Systolic BP (mmHg) -3.9 (2.3)
Diastolic BP (mmHg) -3.9 (1.3)
GP and there was no documented reason for this discontinua- 71%). Those with incomplete data for an outcome measure tion in the secondary care notes. In these cases the GP was were excluded from the analysis of that outcome measure.
contacted for additional information and asked about any drug The included cohort (n=96) attended one or more follow- related adverse effects. Where people stopped taking up appointments and comprised 43% women, mean age 58.9 dapagliflozin, data from subsequent visits were excluded from (range 31-85) years, with a mean duration of type 2 diabetes of the analysis here.
15.1 (range 1-37) years. The mean weight of the cohort was All collected data were sense checked and apparently anom- 94.7 (range 49-154) kg and BMI 33.7 (range 22-52) kg/m2. alous values were rechecked in the clinical records. The mean duration of follow-up was 152 (range 7-431; stan- dard deviation 115) days. The shortest duration of follow-up was Statistical analyses an urgent appointment requested because the patient had To evaluate change in the outcome parameters of HbA1c, developed a widespread rash. Dapagliflozin was stopped at this weight, and blood pressure over time, patient follow-up data visit. The mean number of follow-up appointments included was were grouped by three month periods; 0-3, 4-6, 7-9, and 10-12 1.57 (range 1-4).
months. The median change in each outcome parameter wascalculated for each three month group. Where people attended more than one follow-up appointment in a single three month Of the included people, 72 (75%) had follow-up within the first period, the last appointment data was used. three months of starting dapagliflozin, 50 (52%) at 4-6 months, We performed a linear regression analysis to identify predic- 22 (23%) at 7-9 months, and 12 (13%) at 10-12 months. Aver- tors of reduction in HbA1c, weight, and blood pressure in people age changes at the follow-up visit are shown in Table 1. Seven taking dapagliflozin. People who failed to start dapagliflozin or people had no response to dapagliflozin with no improvement who stopped taking it during the follow-up period were in HbA1c and no weight reduction (Table 2). The HbA1c response excluded. Change from baseline for the outcome measures of was sustained (Figure 1). Weight and blood pressure continued HbA1c, weight, and blood pressure were calculated. We assume to improve during the year of follow-up (Figure 2). a significance level of p <0.05 and report model performance Of those people who tolerated dapagliflozin through the follow using R-square and adjusted R-square values. The analysis was -up period, 36 (52%) were able to stop or reduce one or more other undertaken using SPSS version 20.0.
diabetes medications, whereas 18 (26%) patients had medication Ethical considerations Table 2 Response to dapagliflozin treatment in patients with
This study was designed as a clinical audit of routine practice for type 2 diabetes (‘real-world data') the purpose of improving patient management and as such didnot require ethics committee review.10 Reduction < 1% All people (n=122) who had been prescribed dapagliflozin were Reduction > 1% initially included for analysis. Three people were excluded as their notes were unavailable. Four people were excluded as they failed Weight loss < 5kg to attend any follow-up appointments. One person was Weight loss > 5kg and <10kg 15Weight loss > 10kg excluded as they had type 1 diabetes initially misdiagnosed astype 2, and dapagliflozin was then stopped. Fourteen people Systolic BP (mmHg) Reduction < 5mmHg were excluded because they were not yet due to attend follow- Reduction > 5mmHg up and four were excluded because they chose not to starttaking dapagliflozin. The majority of participants had complete Diastolic BP (mmHg) Reduction < 5mmHg data (an initial value and follow-up measurements) on HbA1c Reduction > 5mmHg (n=79; 82%), weight (n=88; 92%), and blood pressure (n=68; VOLUME 14 ISSUE 4 l OCTOBER/NOVEMBER/DECEMBER 2014 LEARNING FROM PRACTICE Figure 1. Changes over time in HbA1c in patients with type 2
Figure 2. Changes over time in weight and blood pressure in
diabetes following initiation of dapagliflozin patients with type 2 diabetes following initiation of dapagliflozin n Systolic BP change (mmHg) l Diastolic BP change (mmHg) s Weight change (kg) Time (days)
Time (days)
Table 3 Changes in concurrent diabetes medications during the follow-up period in patients who tolerated dapagliflozin
Concurrent medication type
*Includes patients with changes to >1 medication. Table 4 Clinical predictors of change in HbA1c in patients
Table 5 Clinical predictors of weight change in patients
taking dapagliflozin (n=79) taking dapagliflozin (n=88) (kg) Clinical
Coefficient (95%
Clinical
Coefficient (95%
-0.007 (-0.023 to 0.009) 0.011 (-0.031 to 0.052) -0.437 (-0.747 to -0.126) -0.480 (-1.288 to 0.328) Diabetes duration (years) 0.010 (-0.015 to 0.034) Diabetes duration (years) 0.002 (-0.061 to 0.065) 0.014 (-0.010 to 0.037) -0.144 (-0.208 to -0.081) Baseline HbA1c (%) -0.598 (-0.686 to -0.511) Baseline HbA1c (%) -0.165 (-0.409 to 0.079) Model R-square 0.459, adjusted R-square 0.419 Model R-square 0.072, adjusted R-square 0.006 doses increased or an additional medication added (Table 3).
baseline blood pressure was associated with a greater reduction Higher HbA1c at baseline was associated with a greater re- in both systolic and diastolic blood pressure (Tables 6 and 7).
duction in HbA1c whilst taking dapagliflozin (Table 4). Reduction Older age was weakly associated with a greater reduction in in HbA1c was independent of age, gender, duration of diabetes.
diastolic blood pressure.
Higher BMI at baseline was associated with greater weight loss(Table 5). Baseline weight was not associated with the degree of weight loss (results not shown). Weight loss was independent Adverse effects were recorded for 36 (38%) people. Increased of age, gender, duration of disease, and baseline HbA1c. Higher urine flow (nocturia and polyuria) was the most common adverse THE BRITISH JOURNAL OF DIABETES & VASCULAR DISEASE LEARNING FROM PRACTICE Table 6 Clinical predictors of reduction in systolic blood
Table 7 Clinical predictors of reduction in diastolic blood
pressure (mmHg) in patients taking dapagliflozin (n=68) pressure (mmHg) in patients taking dapagliflozin (n=68) Clinical
Coefficient (95%
Clinical
Coefficient (95%
-0.169 (-0.404 to 0.066) -0.312 (-0.462 to -0.161) 2.186 (-2.029 to 6.401) -0.885 (-3.435 to 1.665) Diabetes duration (years) -0.065 (-0.378 to 0.248) Diabetes duration (years) 0.175 (-0.018 to 0.368) -0.241 (-0.572 to 0.090) -0.209 (-0.410 to -0.008) Baseline HbA1c (%) -0.826 (-2.009 to 0.358) Baseline HbA1c (%) -0.445 (-1.169 to 0.279) Baseline systolic BP (mmHg) -0.528 (-0.664 to -0.392) Baseline diastolic BP (mmHg) -0.274 (-0.406 to -0.143) Model R-square 0.261, adjusted R-square 0.175 Model R-square 0.146, adjusted R-square 0.048 effect (occurring in 17 people) followed by genital candidiasis discontinuation of dapagliflozin was genital candidiasis (4 of 10 (10), postural hypotension (3), UTI (3), dyspepsia (2), thirst (2), people affected). Most people with reported polyuria (6 of 8) dry mouth (2), rash (2), erectile dysfunction (1), fatigue (1), back elected to continue taking dapagliflozin. Similarly, most people pain (1), palpitations (1), and urinary incontinence (1). The with nocturia (8 of 9) elected to continue. person reporting urinary incontinence had a previous history ofincontinence which returned whilst taking dapagliflozin. The two people reporting rash had developed a widespread erythema- This study of real-world data demonstrates that dapagliflozin is tous rash with associated pruritus within the first three days of effective at reducing HbA1c with 42% of the people who tolerated taking dapagliflozin necessitating discontinuation. One person dapagliflozin achieving a reduction in HbA1c >1%. Significant was admitted to hospital with a confirmed E. coli UTI and reductions in weight and blood pressure were also confirmed in our bacteraemia whilst taking dapagliflozin. This person had a study. The clinical response to dapagliflozin was independent of history of multiple UTIs although none had previously required age and duration of diabetes and the greatest improvement in hospital admission. This suggests a history of recurrent UTIs may HbA1c was seen in those with the poorest control at baseline.
limit the use of this drug.
Similarly the greatest reduction in weight and blood pressure wasseen in those with the highest BMI and blood pressure at baseline.
BMI and not baseline weight was associated with weight loss, A total of 27 (28%) people stopped taking dapagliflozin during which suggests that the degree of obesity is the predictor of weight the follow-up period. Four people were advised to stop because loss. As considerable numbers could either reduce or stop other of deterioration in their renal function (three of these had an oral therapies or insulin, there may be additional cost benefits from improvement in glucose control prior to stopping), two because a pharmaco-economic perspective. they felt it added to an already large pill burden (which may be Dapagliflozin improves glycaemic control by preventing improved in future by fixed dose combinations) and 21 because glucose reuptake by SGLT2 in the proximal tubule of the of adverse effects. The most common adverse effect leading to kidney.11,12 Inhibition of SGLT2 prevents renal reabsorption of Table 8 Outcomes in clinical trials compared with real-world data in patients with type 2 diabetes treated with dapagliflozin
Duration of follow-up (weeks)
Participant demographics
Total participants (n)
Mean duration of diabetes (years) Outcomes (means)
HbA1c change (%)
Weight change (kg) Systolic BP change (mmHg) Diastolic BP change (mmHg) VOLUME 14 ISSUE 4 l OCTOBER/NOVEMBER/DECEMBER 2014 LEARNING FROM PRACTICE glucose. The resulting glucosuria also promotes weight loss.4 Theaction is independent of insulin secretion and insulin action anddoes not predispose to hypoglycaemia.3 Inhibition of SGLT2 can Key messages
promote urinary sodium loss13,14 which, along with weight reduc-tion and an osmotic diuresis, may be responsible in part for theblood pressure lowering effects observed.15 People who experienced polyuria or nocturia were likely to In a ‘real world' setting: continue to take the agent, although several adverse effects were Dapagliflozin responders had a marked reduction in observed leading to 22% of the cohort discontinuing glucose, weight and BP dapagliflozin therapy. The glycosuria induced by dapagliflozin can The greatest benefit occurred in those with the potentially lead to UTIs and genital infections, predominantly can- poorest glucose control, highest BMI and BP didiasis12,16 - the adverse effect most likely to lead to discontinu- Increased urine flow (polyuria and nocturia) was the most common side effect although genital candidiasiswas most likely to lead to discontinuation Comparison with pre-registration trial resultsOur patient cohort was a similar in age to people included in weight, and blood pressure changes is limited by small sample clinical trials, although the mean duration of diabetes was longer size and heteroskedacity in both outcome measures. The higher than that of trial participants (Table 8). The effectiveness of variability in HbA1c, weight, and blood pressure change with dapagliflozin, despite the longer duration of diabetes, in our higher initial values is likely to be an intrinsic property of the data.
study was similar to that reported in clinical trials of similarduration (Table 8).
In pre-registration trial patients taking dapagliflozin 10mg The real-world data presented here may have greater generalisabil- single agent therapy, genital infections were reported in 9.7%, ity to clinical practice, than data from clinical trials as patients with UTIs in 8.1%, nocturia in 1.6%, hypotension in 1.1%, and any multiple co-morbidities or on multiple oral antidiabetic agents were adverse event leading to discontinuation in 4.3%.1 In patients included in the population analysed and are usually excluded from taking concurrent metformin and dapagliflozin (10mg), genital clinical trials. This study looked at a population of patients with infections were reported in 8.8%, UTIs in 8.1%, and any adverse diabetes referred for specialist management in secondary care and event leading to discontinuation in 3.0%.2 In patients taking con- therefore may not be applicable to all patients in the primary care current insulin and dapagliflozin (10mg), polyuria was reported in 8.3%, and any adverse effect leading to discontinuation in 4.2%.
These data confirm that dapagliflozin is effective for use in By contrast, the current study had slightly higher rates of gen- clinical practice but clinicians should account for a higher level of ital infections (10.4%) and postural hypotension (3.1%), and intolerance to the side effects of dapagliflozin in clinical practice higher rates of increased urine flow (polyuria/nocturia) (17.7%) than that which is reported in clinical trials. In those who tolerated than reported in these trials. The most substantial discrepancy was dapagliflozin, the commonly clustered metabolic risk factors of the rate of discontinuation due to side effects which we found to poor glycaemic control, obesity, and hypertension were all signif- icantly improved. There was also a reduction in the use of otheroral antidiabetic agents and insulin in this population. Limitations of this real-world studyUsing routinely collected data to assess the prevalence of side- Conflict of interest AMcG and ND were funded by the Diabetes Ther-
effects is likely to lead to some underestimation of their apies Evaluation Network. NM, KW, and MF receive financial support for frequency, as minor side effects may not be reported in clinics.
research, speaker meetings, and consultancy from MSD, Merck, BMS, However major adverse events or effects that lead to patients AstraZeneca, Pfizer, Novo, Eli-Lilly, and Sanofi-Aventis. AMcG, ND, NM, KW, wanting to discontinue dapagliflozin are almost certainly docu- and MF wrote the manuscript. AMcG and ND collected and analysed the mented. The routinely collected data also had some missing values for the outcome measures of HbA1c, weight, and bloodpressure. We cannot determine retrospectively if there was a bias References
1. Ferrannini E, Ramos SJ, Salsali A, et al. Dapagliflozin monotherapy in
towards missing data in a particular subpopulation although we type 2 diabetic patients with inadequate glycemic control by diet and suspect that missing data are mostly random.
exercise: a randomized, double-blind, placebo-controlled, phase 3 trial.
Our real-world data are confounded by changes to other med- ications and there is no control group for comparison. However 2. Bailey CJ, Gross JL, Pieters A, et al. Effect of dapagliflozin in patients with type 2 diabetes who have inadequate glycaemic control with metformin: the general trend observed during the follow-up period was a randomised, double-blind, placebo-controlled trial. Lancet 2010; towards less concomitant medication, suggesting that the beneficial effects seen in this group of patients are mostly attrib- 3. Bailey CJ, Gross JL, Hennicken D, et al. Dapagliflozin add-on to met- utable to dapagliflozin.
formin in type 2 diabetes inadequately controlled with metformin: a ran-domized, double-blind, placebo-controlled 102-week trial. BMC The linear regression analysis of factors associated with HbA1c, THE BRITISH JOURNAL OF DIABETES & VASCULAR DISEASE LEARNING FROM PRACTICE 4. Strojek K, Yoon KH, Hruba V, et al. Dapagliflozin added to glimepiride Diabetologists nationwide exenatide audit. Diabetes Obes Metab 2011; in patients with type 2 diabetes mellitus sustains glycemic control and weight loss over 48 weeks: a randomized, double-blind, parallel-group, 10. National Research Ethics Service. Defining Research 2009. Available placebo-controlled trial. Diabetes Ther 2014;5:267-83.
5. Nauck MA, Del Prato S, Meier JJ, et al. Dapagliflozin versus glipizide as 11. Fujita Y, Inagaki N. Renal sodium glucose cotransporter 2 inhibitors as a add-on therapy in patients with type 2 diabetes who have inadequate novel therapeutic approach to treatment of type 2 diabetes: Clinical data glycemic control with metformin: a randomized, 52-week, double-blind, and mechanism of action. J Diabetes Investig 2014;5:265-75.
active-controlled noninferiority trial. Diabetes Care 2011;34:2015-22.
12. Rahmoune H, Thompson PW, Ward JM, et al. Glucose transporters in 6. Strojek K, Yoon KH, Hruba V, et al. Effect of dapagliflozin in patients human renal proximal tubular cells isolated from the urine of patients with type 2 diabetes who have inadequate glycaemic control with with non-insulin-dependent diabetes. Diabetes 2005;54:3427-34.
glimepiride: a randomized, 24-week, double-blind, placebo-controlled trial. Diabetes Obes Metab 2011;13:928-38.
13. Komoroski B, Vachharajani N, Feng Y, et al. Dapagliflozin, a novel, se- lective SGLT2 inhibitor, improved glycemic control over 2 weeks in pa- 7. Wilding JP, Norwood P, T'Joen C, et al. A study of dapagliflozin in pa- tients with type 2 diabetes mellitus. Clin Pharmacol Ther 2009;85:513-9.
tients with type 2 diabetes receiving high doses of insulin plus insulin sensitizers: applicability of a novel insulin-independent treatment. Dia- 14. Tahrani AA, Barnett AH. Dapagliflozin: a sodium glucose cotransporter 2 inhibitor in development for type 2 diabetes. Diabetes Therapy 8. Wilding JP, Woo V, Soler NG, et al. Long-term efficacy of dapagliflozin in patients with type 2 diabetes mellitus receiving high doses of insulin: 15. Ptaszynska A, Hardy E, Johnsson E, et al. Effects of dapagliflozin on car- a randomized trial. Ann Intern Med 2012;156:405-15.
diovascular risk factors. Postgrad Med 2013;125:181-9.
9. Thong KY, Jose B, Sukumar N, et al. Safety, efficacy and tolerability of 16. Plosker GL. Dapagliflozin: a review of its use in type 2 diabetes mellitus.
exenatide in combination with insulin in the Association of British Clinical Nationwide Audit Now Launched!
ABCD launched a nationwide audit of dapagliflozin in the UK.
This audit is particularly important with dapaglifozin being the first of a new class of drugs for diabetes, the SGLT2 inhibitors. We have a chance to assess real clinical efficacy and safety of this new type of treatment by pooling our experience nationwide Does your centre use dapagliflozin (Forxiga)?
If yes, REGISTER YOUR CENTRE! by contacting [email protected]
l you are able to analyse your local data easily
l you are invited to enter your patients' data into the online tool on N3 (the NHS secure network)
l the data will be automatically added to the national data in anonymised form
l we can provide easy-to-complete paper proformas for use in clinic if preferred Please remember: - the more data, the more complete our understanding of this new treatment will be
- all contributors will be listed in publications arising from data submission
VOLUME 14 ISSUE 4 l OCTOBER/NOVEMBER/DECEMBER 2014

Source: http://www.mcgov.co.uk/publicwork/%5BArticle%5D%20-%20Dapagliflozin%20in%20clinical%20practice.pdf