Does dapagliflozin regress left ventricular hypertrophy in patients with type 2 diabetes? A prospective, double-blind, randomised, placebo-controlled study

Background Patients with diabetes have a two to fourfold increased risk for development of and death from cardiovascular disease [CVD]. The current oral hypoglycaemic agents result in limited reduction in this cardiovascular risk. Sodium glucose linked co-transporter type 2 [SGLT2] inhibitors are a relatively new class of antidiabetic agent that have been shown to have potential cardiovascular benefits. In support of this, the EMPA-REG trial showed a striking 38% and 35% reduction in cardiovascular mortality and heart failure [HF] hospitalisation respectively. The exact mechanism (s) responsible for these effects remain (s) unclear. One potential mechanism is regression of Left ventricular hypertrophy (LVH). Methods The DAPA-LVH trial is a prospective, double-blind, randomised, placebo-controlled ‘proof of concept’ single-centre study that has been ongoing since January 2017. It is designed specifically to assess whether the SGLT2 inhibitor dapagliflozin regresses left ventricular [LV] mass in patients with diabetes and left ventricular hypertrophy [LVH]. We are utilising cardiac and abdominal magnetic resonance imaging [MRI] and ambulatory blood pressure monitoring to quantify the cardiovascular and systemic effects of dapagliflozin 10 mg once daily against standard care over a 1 year observation period. The primary endpoint is to detect the changes in LV mass. The secondary outcomes are to assess the changes in, LV volumes, blood pressure, weight, visceral and subcutaneous fat. Discussion This trial will be able to determine if SGLT2 inhibitor therapy reduces LV mass in patient with diabetes and LVH thereby strengthening their position as oral hypoglycaemic agents with cardioprotective benefits. Trial registration Clinical Trials.gov: NCT02956811. Registered November 2016.


Background
Patients with type 2 diabetes mellitus [T2DM] have double the risk of cardiovascular death [CVD] compared with patients without T2DM [1,2]. Hyperglycaemia itself contributes to both the pathogenesis of atherosclerosis and heart failure [3]. While intensive glucose control reduces the risk of microvascular complications it appears to be insufficient to reduce cardiovascular [CV] events [4]. Three large randomized controlled trials [RCTs] ADVANCE, ACCORD and VADT failed to demonstrate any significant effect on macrovascular events of more intensive glycaemic control in patients with longstanding T2DM when compared with standard medical care [5][6][7].
The EMPA-REG OUTCOME trial was a landmark trial as it demonstrated for the first time that a glucose lowering agent could reduce CV events [8]. It was a multicentre, randomised, double blind, placebo-controlled trial performed in 7020 patients with T2DM at high cardiovascular risk comparing the SGLT2 inhibitor empagliflozin to placebo. In the empagliflozin group there were significantly lower rates of death from cardiovascular causes and heart failure hospitalisations, by 38 and 35% respectively. The exact mechanisms responsible for these effects remains unclear.

Left ventricular hypertrophy
One potential mechanism is regression of left ventricular hypertrophy [LVH]. Left ventricular hypertrophy is thought to be present in up to 70% of patients with T2DM [9] It is a strong independent predictor of cardiovascular deaths and events and is even worse than triple vessel coronary disease [10,11]. The reason why LVH is so adverse is because it predates so many different cardiovascular events i.e. LVH is intrinsically arrhythmogenic and causes sudden death, it impedes left ventricular [LV] filling and leads to diastolic heart failure, it reduces coronary perfusion reserve and causes ischaemia and it causes left atrial enlargement leading to atrial fibrillation [AF], and cardio-embolic strokes [12].
How does one cause regression of LVH? Controlling blood pressure [BP] and using a drug that blocks the renin-angiotensin system [RAS] are the standard approaches to the management of LVH but this approach is only partially effective since 44% of all patients with T2DM are normotensive patients with LVH [9]. Thus normotensive LVH is very common [9,13]. Indeed, BP only contributes 25% to the variability in LV mass seen in a population [14]. Despite a "normal" BP, normotensive LVH is just as risky as is hypertensive LVH [15]. Nevertheless, we do know that regressing LVH irrespective of BP changes is an effective way to reduce the incidence of all major cardiovascular [CV] events including specifically sudden deaths, heart failure hospitalisations, new onset AF and strokes [16][17][18][19][20][21][22][23]. The LIFE trial provides conclusive proof that in diabetes, LVH regression per se reduces future cardiovascular events [by 24%], reduces CV deaths [by 37%] and reduces total deaths [by 41%] irrespective of BP [24].
Since controlling BP and using an angiotensin enzyme inhibitor or angiotensin receptor blocker is only partially effective at regressing LVH, we now need additional ways of regressing LVH. Insulin resistance is another mediator of LVH. The literature is awash with observational studies linking insulin resistance to LVH. The large studies are mostly positive which includes the Framingham Study, the Whitehall trial, the Strong Heart trial and the Women's Health Initiative trial while Hyper-GEN is the one large negative trial [25][26][27][28][29]. Therefore, it is likely that glycaemia contributes to LVH. However, there is little evidence to date that glycaemic control alone affects the risk of CV events and thus key ancillary properties of each anti-glycaemic drug will be necessary to deliver the CV benefits we so badly need in diabetes [4,30,31]. A separate albeit related factor associated with LVH is also obesity [25,32,33].

Sodium glucose linked co-transporter2 [SGLT2] inhibitors and their potential to regress LVH
In this study, we hypothesize that SGLT 2 inhibitors may be able to lead to LVH regression. Firstly, SGLT2 inhibitors employ a novel mechanism to lower blood glucose by enhancing urinary glucose excretion by competitively blocking the sodium glucose linked co-transporters in the proximal renal tubules, thus preventing the reabsorption of filtered glucose and sodium, resulting in glycosuria and natriuresis [34,35]. This is in contrast with other antidiabetic medications which focus on restoring β-cell activity, insulin sensitivity and tissue glucose uptake to reduce plasma glucose levels. Accordingly, SGLT2 inhibitors are expected to maintain their potency as beta cell function declines with disease progression. Secondly, the glycosuric effects of SGLT2 inhibitors result in around 240-400 kcal/ day loss through the urinary tract [36]. This calorific loss results in an average weight loss of around 2-3 kg that could help lead to LVH regression [37]. Finally, the natriuretic effect and subsequent osmotic diuresis could also be significant in patients with cardiovascular disease. This diuretic effect should reduce preload on the heart. The SGLT2 inhibitors lower blood pressure by 7-10 mmHg, reduce arterial stiffness and afterload [38][39][40].
In summary, SGLT2 inhibitors may improve cardiac structure because they appear to reduce the four main causes of LVH: glycaemia/insulin resistance, weight, preload and afterload [blood pressure] [41]. Our hypothesis is that dapagliflozin will regress LVH in normotensive patients with T2DM. If so, this could be a large part of explaining why such drugs reduce CV events in the EMPA-REG OUTCOME trial.
The issue of how SGLT2 Inhibitors reduces CV events in diabetes is a hot topic following EMPAREG OUT-COME. A large ongoing trial [DECLARE -TIMI 58] is also in progress assessing the effect of Dapagliflozin on CV events. If DECLARE-TIMI 58 shows clearly that dapagliflozin reduces CV events, then our trial if positive will have revealed a possible contributing mechanism to the reduced CV events i.e. LVH regression.

Study design
The DAPA-LVH trial is a prospective, double-blind, randomised, placebo-controlled 'proof of concept' singlecentre study conducted in NHS Tayside, Scotland designed to evaluate the efficacy of 12 months of the SGLT2 inhibitor dapagliflozin compared to placebo on left ventricular hypertrophy [LVH] in 64 normotensive participants with diabetes identified to have LVH. A recruitment window of 1.5 years from December 2016 has been set. Participants will be enrolled in this trial for a period of 12-13 months, [ Fig. 1]. Therefore, the overall trial end date will be May 2019.
At the screening visit an initial medical history and clinical examination will be performed following informed consent. Participants will have an electrocardiogram performed and bloods taken for safety analysis. Vital signs including blood pressure will be recorded to confirm eligibility prior to enrolment. Blood pressure will be taken using an Omron M10-IT blood pressure monitor and eligible patients will have an office blood pressure of 145/90 mmHg averaged over three readings. Patients who require optimisation of their blood pressure will do so but will have to be stable on their current antihypertensive medications for 3 months prior to enrolment. Patients with borderline office blood pressure will undergo ambulatory blood pressure measurement (AMBP) This will be performed using a Spacelab 90,217 ambulatory blood pressure monitor. Inclusion will be possible with a 24 h mean blood pressure < 140/85 mmHg. Participants will also be screened for echocardiographic evidence of left ventricular hypertrophy [LVH] by the standard American Society of Echocardiology [ASE] criteria. This will be performed using a Philips Epiq 7 machine by a fully trained operator. Eligible participants identified to have LVH on echocardiography will be recruited. The full inclusion criteria are as listed below.
Recruited patients will return for a cardiac magnetic resonance imaging [CMRI] at the Clinical Research Centre, Ninewells Hospital, Dundee, within 3 weeks of the planned baseline [randomisation] visit. At the randomisation visit participants will have vital signs, body mass index, waist circumference and waist to hip ratio recorded. Participants will also be asked to undergo 24 h ambulatory BP monitoring using a Spacelab 90,217 ambulatory monitor. Examinations with greater than 50% successful readings will be deemed an acceptable exam. Bloods for safety analysis and research purposes [BNP, FIRI and Uric acid] will also be taken. During the visit, participants will also be randomly be assigned to either dapagliflozin 10 mg or matching placebo. The first dose will be administered during this visit and participants will be educated on the symptoms of both hypoglycaemia and diabetic ketoacidosis and given written instructions of how to manage it if either event occurs.
To reduce the likelihood of hypoglycaemia in participants taking insulin, participants who are already on insulin at time of recruitment will have their total daily dose of insulin reduced by 10% on the day they are randomised. Further dose titration will be done by the study team or GP based on the participant's symptoms, home and laboratory-based blood sugar levels. Down-titration of therapy will be done in a stepwise manner starting with insulin. Other anti-diabetic agents will only be down-titrated once insulin has been discontinued.
In order to make the two groups comparable, a target HbA1c of ≤ 53 mmol/mol will be set for all participants. New onset diabetic patients will not be included in this study as SGLT2 inhibitors are currently only licensed as second line therapy. We will therefore be comparing a dapagliflozin [mostly as a second drug after metformin] based group against a conventionally treated group but without a SLGT2 inhibitor.
This will ensure that any difference in LV mass between groups is because dapagliflozin and all its ancillary cardiac properties and not because the two groups differed in glycaemic control.
With regards to BP, the main criteria will be that the baseline office BP is < 145/90 mmHg. However, the investigator will have clinical discretion to change antihypertensive drugs during the trial for safety reasons, under two circumstances. Firstly, if the systolic BP rises to above 140 mmHg on 2 consecutive visits during the trial then the participant can be started on extra antihypertensive drugs to re-achieve a systolic BP of < 145 mmHg. Secondly, if the participant suffers from dizziness and/or their systolic BP has fallen either by ≥ 25 mmHg or to an absolute level of ≤ 110 mmHg, then the attending physician can reduce or stop one of their antihypertensive drugs. These criteria serve two functions: firstly, to copy normal clinical practice and secondly to maintain participant safety.
Participants will return for three visits throughout the year to have safety and research bloods taken and to have vital signs, BMI, waist to hip ratio and waist circumference recorded. They will also be assessed for adverse events and to alter diabetic/antihypertensive therapy [if applicable].
The biomarker samples will be centrifuged and decanted into an aliquot which will be stored at -80°C. Uric acid will be analysed with an elisa method using SIGMA-ALDRICH assay, UK. BNP will be measured by a MULTI-ARRAY system. The kit is from MESO SCALE DISCOVERY, USA. FIRI will be analysed with an elisa method using an ALPCO assay UK.
At the end of the 1 year study period, participants will return for repeated assessment of vital signs, BMI, waist circumference and waist to hip ratio, ambulatory blood pressure, echocardiography and CMRI. These values will be compared with their baseline tests to determine if any significant change has occurred with each of the two arms of the study populations. [See Table 1 for an overview of all visits scheduled for the trial].

Study population
Diabetic patients with LVH will be identified for this study and all potential participants who meet the following criteria will be eligible for the trial: Inclusion criteria: If a participant has had a clinical echo done within the previous 6 months the results from this will be used for assessing eligibility, otherwise an echo will be performed at the screening visit d U&E, FBC, LFT, cholesterol, HDL-cholesterol e See section 7.2 f HbA1c, FIRI, BNP, glucose, uric acid. A sample to be held for future research will be taken at the last visit g The MRI scan may be performed ± 3 weeks from the baseline visit and may therefore require a separate visit h If the participant as not yet had their MRI scan they will be asked not to start their study medication until they have had their scan. Their Visit 3 will be delayed until the patient has had at least 3 weeks of study medication and packaged by AstraZeneca and labelled by our onsite clinical trials pharmaceutical pharmacy. Randomisation will be carried out via our Tayside Randomisation System [TRuST], a Good Clinical Practice [GCP] compliant web-based system run by the Tayside Clinical Trials Unit [TCTU], to preserve allocation concealment. This will securely backup both the randomisation seed and the randomisation allocation and have it available in the onsite 24 h emergency unblinding facility.
Once randomised, the participant will continue to take the trial medication once daily for 1 year, if tolerated. Compliance will be checked and documented, by the dispensing pharmacy, using tablet counts at each visit. If non-compliant, they will be encouraged to become compliant. If study drug needs to be stopped due to intolerance or adverse events, they will remain in the study in order to do an "intention to treat" analysis. To assess the tolerability of dapagliflozin in this patient group

Sample size and power calculations
For the primary outcome of LV mass regression using cardiac MRI, we have powered this study for an absolute change in LV mass based on previous studies that we have conducted in our unit. In our recently published study of LVH regression using allopurinol in participants with ischaemic heart disease [41], we found that allopurinol significantly reduced LV mass by − 5·2 ± 5·8 g compared to placebo − 1·3 ± 4·5 g [p < 0·007]. In per-cent terms, this degree of LVH regression is the same as seen between the two arms of the echo sub-study of the LIFE study where CV events were also different between groups. For an 80% power at a 5% significance level [α = 0·05], to detect a similar change in LV mass, we will require 29 subjects per group. Both our previous studies have shown a 10% dropout rate. Therefore, accounting for this, we will require a total of 64 participants [32 per group]. The 10% dropout rate is standard for such studies and includes those who died and those who withdraw consent.

Cardiac MRI protocol
Baseline and repeat Cardiac magnetic resonance imaging [CMRI] examinations at baseline visit [+/− 3 weeks] and after the final 12 month [+/− 4 weeks] visit will be performed on a 3 T Magnetrom Trio scanner [Siemens, Erlangen, Germany] using body array and spine matrix radiofrequency coils.. Short axis images from the atrioventricular ring to the LV apex will be acquired using a 2D ECG-gated breath hold segmented SSFP cine sequence with retrospective gating. Quantitative measurement of LV mass, ejection fraction (EF), end-diastolic volume (EDV), end-systolic volume (ESV) and stroke volume (SV) will be derived by region of interest contours placed around endocardial and epicardial LV borders at end systole and end diastole. Transmitral flow and the isovolumetric relaxation time will be assessed using through plane phase contrast images with electrocardiographic synchronisation.
At the same time of the CMRI visceral and subcutaneous abdominal fat mass will be assessed. For measurement of subcutaneous adipose tissue (SCAT) and visceral adipose tissue (VAT) two successive axial 3D DIXON volume interpolated breath hold examination sequences will be acquired. These sequences will cover an anatomical area from the diaphragm to the pelvic floor, with a slice thickness of 3 mm and up to 88 slices (dependent on patient size) collected within a single 3D block.
For image analysis the 'fat only' DIXON MR images will be segmented using Analyze (Mayo Clinic) software, and the SCAT and VAT compartments are defined using a signal intensity threshold method with manual correction where required. Epicardial fat structures and fat associated with the vertebrae will both be omitted from the final calculated volumes.
From a single MRI slice at the L2-L3 intervertebral level. This single observer will analyse all the scans. Analysis will be performed offline [Argus Software, Siemens] by a single blinded observer.
The reproducibility of all parameters using MRI will be derived by this observer. A test-retest intra-observer co-efficient of variation of 2.0% is usual in this department's past MRI studies. Should the scanner become unavailable for a prolonged period of time during the study an alternative scanner will be used. MRI methods will be adapted as appropriate to ensure optimal study results can be obtained.

Discussion
SGLT2 inhibitors including dapagliflozin improve systemic glucose metabolism, lower blood pressure and lower body weight, thus they ameliorate the metabolic and haemodynamic risk factors heavily implicated in causing LVH. In this study, we propose that the SGLT2 inhibitors may be particularly suitable at regressing LVH. This effect might ultimately explain the reduced CV events seen so far in one large outcome trial with these drugs [Fig. 2].
The primary haemodynamic effect of SGLT2 inhibitors is an osmotic diuresis. Patients treated with dapagliflozin produce approximately 375mls of extra urine per day [36]. Several trials have shown that SGLT2 inhibitors lead to a reduction in systolic BP in a range of 3-5 mmHg and~2-3 mmHg in diastolic BP [37]. This will be further assessed in our trial with ambulatory blood pressure recordings at randomisation and upon completion of the trial. The reason for the observed BP reduction with SGLT2 inhibition is not completely understood but is likely secondary to several different things including the modest diuretic effect, mild natriuresis and weight reduction [39,42]. Data from a mechanistic trial has also demonstrated that empagliflozin reduced arterial stiffness in patients with type 1 diabetes mellitus [43]. These effects on intravascular volume and blood pressure will result in reduced preload and afterload respectively, thereby facilitating a reduction in intra-cardiac pressure and thereby an improvement in cardiac structure [24,44]. Indeed following EMPA-REG Outcome trial there has been a lot of interest in the effects of SGLT2 inhibition on cardiac structure with a number of ongoing trials looking into the effects in patients with both diastolic and systolic heart failure in addition to cardiovascular outcomes [Tables 2, 3, and 4].
Dapagliflozin is known to produce clinically meaningful reductions in HbA1 c. [45] Studies have also shown that treatment with SGLT2 inhibitors improves insulin sensitivity as measured by peripheral glucose uptake [46,47]. One such study showed that insulin mediated tissue glucose disposal increased by around 18% with only 2 weeks of dapagliflozin therapy [47]. Insulin resistance and hyperinsulinaemia have been associated with increased atherosclerosis risk and left ventricular hypertrophy [25][26][27][28][29]48].
Other metabolic effects of the SGLT2 inhibitors include weight loss. With selective SGLT2 inhibition urinary glucose is increased resulting in a negative energy balance and subsequent weight loss [36]. A 24 week study comparing dapagliflozin to placebo showed a 2.5-3.5 kg weight reduction as a result of the calorific loss produced by glycosuria [49]. This is a finding throughout the SGLT2 class [37]. Of potential greater interest is how they change visceral fat mass as this is associated with an increased risk of T2DM and increased risk of CVD and overall mortality [50]. Indeed all the three currently available SGLT2 inhibitors when compared to glimepiride in dedicated body composition studies have shown that the majority of weight loss associated with SGLT2 inhibition was due to a reduction in visceral fat or subcutaneous fat [45,51,52]. Accordingly, we have chosen to also measure visceral and subcutaneous fat mass as a secondary outcome of the DAPA-LVH.
Given these metabolic and haemodynamic effects our hypothesis is that we will see a reduction in left ventricular mass. Indeed, pre-clinical work has shown that SGLT2 inhibitors are capable of reducing LV mass in a rat model with progressive HF [53]. We have therefore selected CMRI measurements of LV mass as our primary outcome measures for the DAPALVH Trial. By ensuring the trial is adequately powered we will determine if treatment with an SGLT2 inhibitor is able to reduce LV mass in diabetic patients with LVH. The EMPA-REG Outcomes trial revealed a reduction in cardiovascular death and HF hospitalisations with the use of empagliflozin in patients with T2DM. However, it is unknown if these effects are seen throughout the SGLT2 inhibitor class. Other cardiovascular outcome trials such as DECLARE-TIMI 58 for dapagliflozin and CANVAS for canagliflozin will reveal whether the cardioprotective effects of SGLT2-inhibitor therapy is seen across the drug class. As described above this study will provide insights into the mechanism of the  positive cardiovascular effects conferred by SGLT2 inhibitor therapy and may also help decide the course of future researchshould LVH be a favoured target?

Limitations
Firstly, this is a relatively small, single centre trial. The use of CMRI though has allowed the power of the trial to be preserved despite the small number of participants. However, given the small numbers some differences observed may still be the result of chance. Secondly, diabetes is a dynamic disease as a patient's glycaemia control may fluctuate and this may necessitate dose adjustments of anti-diabetic medications during the trial which may confound the outcome. However, every measure will be taken to ensure blinding of the investigators is maintained and uniformity in the dose adjustments made.

Conclusion
Historically much attention has focused on the prevention and treatment of the microvascular complications of diabetes. CVD however is still the main co-morbid condition and primary contributor to mortality in patients with diabetes. Besides metformin therapeutic options to optimise glycaemic control which reduce cardiovascular risk are limited. Empagliflozin an SGLT2 inhibitor has been shown to produce significant reductions in cardiovascular mortality and hospitalisation with heart failure [8]. We propose that SGLT2 inhibitors may cause regression of LVH due to their ability to reduce preload/afterload, weight and insulin resistance which may account for their positive cardiovascular effects.
Upcoming major trials will establish if the effect seen with empagliflozin is an SGLT2 class effect. If so, the results of this study if positive will help us understand the mechanisms of the cardioprotective effects of SGLT2 inhibitors and if positive further establish this group of medications as anti-diabetic agents with the added value of protecting the heart.