Efficacy of the treatment with dapagliflozin and metformin compared to metformin monotherapy for weight loss in patients with class III obesity: a randomized controlled trial

Dapagliflozin is a selective sodium-glucose cotransporter type 2 inhibitor (iSGLT2) that blocks glucose resorption in the proximal tubule of the kidney, thereby increasing urinary glucose excretion and reducing blood glucose levels. It is currently indicated in the management of patients with type 2 diabetes (T2D) because of its sustained effect on the reduction of glycemia and glycated hemoglobin (HbA1c) [1]. As described by Baker et al. in their systematic review and meta-analysis, dapagliflozin also decreases systolic blood pressure by 4–5 mmHg and results in a 1.7 kg weight loss (95% CI 1.33 to 2.11) [2]. Furthermore, Jayawardene et al. [3] reported that dapagliflozin increases high-density lipoprotein cholesterol (HDL-c) concentrations by 1.8% to 4.4% and decreases triglyceride concentrations by 2.4% to 6.2%. Dapagliflozin effects in patients with prediabetes and other types of diabetes are currently the subject of intense research [4]. Current meta-analyses report that its adverse effects include an increase in the risk of genital tract infections (OR 3.48, 95% CI 2.33 to 5.20) [5] and a low risk of euglycemic ketoacidosis [6], among others. Most of these adverse effects are mild and do not require drug withdrawal.

Metformin is a biguanide that acts via both adenosine monophosphate-activated protein kinase (AMPK)-dependent and AMPK-independent mechanisms, inhibiting enzymes related to gluconeogenesis and lipogenesis, while also having antineoplastic effects, delaying the aging process, and regulating gut microbiota [7]. Besides being the most common drug for T2D treatment, it has been reported by some studies as having effects on weight loss. Moguel et al. prescribed metformin for a year at doses of 1500 mg/day to nondiabetic women with a body mass index (BMI) of 25 to 32.9 kg/m² (group 1) and doses of 2000 mg/day to women with a BMI 33 to 41.7 kg/m² (group 2) in addition to a hypocaloric, carbohydrate-modified diet. They observed that the first group decreased 8.06 ± 0.96 kg (p < 0.001), and the second group decreased 15.1 ± 3.3 kg (p = 0.011) [8].

In murine models, the glycosuria effect of dapagliflozin has been associated with a caloric deficit between 80 to 340 kcal/day [9]. As monotherapy in patients with T2D, weight loss ranges from 2 to 3 kg; meanwhile, in a 104-week, randomized, double-blinded study of dapagliflozin versus glipizide as add-on therapies to metformin, weight reductions of up to 5.1 kg (95% CI -5.7 to -4.4 kg) without regain (at least for 2 years) have been reported [10]. Other studies have shown that dapagliflozin is as effective as metformin for weight loss. In a 12-week, randomized, parallel-group, double-blind, placebo-controlled study, patients assigned to dapagliflozin 10 mg daily presented a 2.7% reduction in body weight (95%IC -3.5 to -1.8%) as compared to a weight loss of 1.7% in those assigned to the metformin group (95%CI -2.4 to -0.9%) and 1.2% in patients assigned to placebo (-2.0 to -0.4%) [11]. In a 48-week, randomized, double-blinded, placebo-controlled, parallel-group study, when used in combination with pioglitazone, dapagliflozin resulted in a 2.3 kg weight loss (95% CI -3.37 to -1.23 kg) [12].

In a systematic review by Orme et al., in patients previously treated with sulfonylurea, dapagliflozin induced a 1.54 kg weight loss (95%CI -2.16 to -0.92 kg), compared to patients treated with GLP-1 (loss: -0.65 kg, 95% CI -1.37 to 0.07 kg) and a DPP4 inhibitor (gain of 0.57 kg, 95% CI 0.09 to 1.06 kg) [13]. Furthermore, the addition of dapagliflozin in patients already treated with metformin and sitagliptin resulted in a 2.1 kg reduction (95% CI -3.2 to -1 kg) compared to placebo [14]. Finally, Bolinder et al. found that the addition of 10 mg dapagliflozin to patients inadequately controlled with metformin resulted in a decrease of 2.08 kg of total weight (-2.8 to -1.31 kg, p < 0.001) and decreased waist circumference by 1.52 cm (2.74 to 0.31 cm, p = 0.014) after 24 weeks of treatment in a randomized, parallel-group, double-blind, placebo-controlled study [15].

In a meta-analysis, Zhang et al. found that the use of dapagliflozin resulted in 20% less weight reduction than expected (2.5 kg instead of 3.05 kg) [16]. This could be related to the increase in appetite observed in DIO (diet-induced obesity) rats by Devenny et al. [7]. The effect of dapagliflozin in appetite has not been evaluated in humans.

Notably, all studies have been conducted in patients with T2D and a BMI under 40 kg/m², so the effects dapagliflozin on weight and other metabolic factors have not been studied in populations with extreme obesity. Furthermore, the effects of dapagliflozin on the cytokine profile and on appetite markers like ghrelin have not been studied in this population.

Patients under evaluation for bariatric surgery have been suggested to decrease 10% of their excess weight to improve the control of some of the comorbidities that are associated with a higher surgical risk such as diabetes, hypertension, or sleep apnea [17]. This also allows evaluating the patients’ adherence to medical therapy and their capability to follow a diet. International guidelines suggest the use of drugs for weight loss when BMI is higher than 30 kg/m², but only few have been approved due to their adverse effects, including increased blood pressure, anxiety, tachycardia and psychiatric effects [18]. Patients with severe obesity use a daily average of 4.4 ± 4.1 medications to control comorbidities such as diabetes, including psychoanaleptics and agents acting on the renin-angiotensin system (12%, 11.3%, and 8.2% of all medications, respectively) [19]. Therefore, most patients may not be candidates for some of these weight loss drugs, which will limit their chances to undergo surgery. SGLT-2 inhibitors have gained a privileged position in the management of T2D in the last few years due to their ability to control glucose as well as other cardiovascular risk factors including blood pressure, weight, and the lipid profile. Currently, diabetes guidelines place these inhibitors as the first choice for patients with diabetes and high cardiovascular risk [20].

The aim of the study is to determine if combined treatment with dapagliflozin and metformin (D/M) is more effective than monotherapy with metformin (M) for weight loss in patients with class III obesity and prediabetes or diabetes who are waiting for bariatric surgery.

Mexico has one of the highest prevalence rates of overweight and obesity worldwide. According to the National Health and Nutrition Survey, Mid Step (ENSANUT MC: Encuesta Nacional de Salud y Nutrición de Medio Camino 2016), 33.3% of Mexicans older than 20 years are obese, with a prevalence of 27.7% in males and 38.6% in females. This survey also showed that the prevalence of class III obesity in males was 1.7% and 4.1% in females (2.4-fold higher in the latter) [28].

In 2008, a multidisciplinary clinic for the treatment of severe obesity and its related comorbidities was inaugurated at Hospital de Especialidades of Centro Médico Nacional Siglo XXI. One of the purposes of the Clinic is to offer bariatric surgery to qualifying patients, who are defined as those who achieve an appropriate weight loss (EBWL of 10%). Several studies have proven that weight reduction prior to surgery decreases surgery-related complications and decreases hospitalization time [17]. Unfortunately, as previously observed in a quasi-experimental study performed at our Obesity Clinic, 46% of our patients do not achieve the required weight loss after a year of dietary treatment. Furthermore, 46% of our patients have prediabetes or T2D, 66% are hypertensive, and 33% have dyslipidemia [29]. T2D treatments available in our clinic include metformin, pioglitazone, acarbose, and insulin (Neutral Protamine Hagedorn insulin, glargine and ultra-rapid insulin analogs). The patients with diabetes and obesity could benefit from the use of weight loss pharmacological treatment as recommended in 2015 by the Endocrine and Obesity Society [18]. However, Mexico’s Federal Commission for the Protection against Sanitary Risks (COFEPRIS: Comisión Federal para la Protección contra Riesgos Sanitarios) has approved only two treatments: orlistat and liraglutide [30]. The former has a low adherence rate associated to its secondary effects, and the latter is not covered by social security due to its high cost. Therefore, we propose the use of a low-cost and secure treatment such as dapagliflozin for weight reduction in patients with grade III obesity and T2D or prediabetes. In our institution, the most common dose of metformin prescribed for glucose control is 850 mg bid [31]. To assess if dapagliflozin has a higher efficacy, we decided to compare its addition to the usual treatment instead of using combination therapies such as Xigduo, which contains higher amounts of metformin (dapagliflozin 5 mg with metformin 1000 mg). Nowadays, both treatments are only indicated for glucose lowering in patients with diabetes or prediabetes. However, we aim to probe that beyond this effect, antihyperglycemic drugs such as metformin and dapagliflozin could also decrease weight in patients with higher grades of obesity.

Several studies have reported that metformin induces a modest decrease or no change, in weight [32, 33]. In an open randomized cross-over study, Hermann et al. reported a 2.64 kg loss of total weight (95% CI -4.23 to -1.05 kg) after one year of treatment with metformin (patients with an average weight of 76.4 kg) [34]. Similarly, in a quasi-experimental study, Stumvoll et al. observed a total weight loss of 2.7 ± 1.3 kg in a study conducted in 10 T2D patients with an average BMI of 32.1 ± 3.2 kg/m² and treated with metformin at a dose of 2550 mg (850 mg three times a day) [35]. In addition, several studies have demonstrated that dapagliflozin effectively induces weight loss, decreases WC, reduces systolic and diastolic blood pressure as well as TG, and increases HDL-c both as monotherapy and in combination with other antidiabetic medications [10‐15]. No studies have been conducted in patients with a BMI higher than 40 kg/m² to assess the weight loss capacity of the combination of D/M.

In their systematic review and meta-analyses, Zhang et al. calculated that a urinary glucose excretion of 71.2 g/24 h induced by iSGLT2 should decrease 3.05 kg. However, only a 2.5 kg weight reduction was achieved, representing 20% less than expected [16]. This result led to Devenny et al. evaluating possible compensatory mechanisms that attenuate the effect of dapagliflozin. In a group of DIO rats, orally administered dapagliflozin in different doses induced weight and fat mass loss in a dose-dependent way. In this study, the effect on weight and fat mass was greater in those animals with restricted food consumption. Furthermore, in the group of animals fed ad libitum, they observed an increase in appetite as of day 7 and throughout the study. They evaluated appetite as the amount of food consumed each day in grams multiplied by the caloric value of the diet. They concluded that the increased appetite decreases the weight loss capacity induced by dapagliflozin [9]. This conclusion has not been evaluated in humans. In our protocol, we propose to evaluate ghrelin concentration, as a surrogate indicator of appetite [36].

Additionally, few studies have evaluated the effect of dapagliflozin on the concentration of inflammatory cytokines. In a single-arm interventional study, Okamoto et al. observed that dapagliflozin administered for 12 weeks decreases highly sensitive CRP (hs-CRP) and increases adiponectin [37]. Once again, this study included patients with an average BMI of 32.7 ± 6.5 kg/m². According to the effect of decreased weight and WC previously reported in patients with grade I and II obesity treated with dapagliflozin, we expect a decrease in resistin and IL-6 concentrations and an increase in IL-10 and adiponectin concentrations.

Finally, we aim to investigate if dapagliflozin treatment increases insulin and glucagon secretion in patients with obesity. Chronic hyperglycemia leads to a defect in insulin production due to glucotoxicity. The lowering of glucose by dapagliflozin has been suggested to reduce glucotoxicity and improve insulin secretion [38]. Furthermore, the normalization of blood glucose levels as well as the decrease in HbA1c levels and weight reduction induced by dapagliflozin seems to be related with an improvement in insulin sensitivity [38]. Some studies suggest that dapagliflozin could increase endogenous glucose production due to an increase in glucagon secretion by exerting a direct effect on alpha cell function [39]. Those effects remain controversial in people with obesity.

The study is currently recruiting and enrolling participants according to version 2 of the protocol in June 2019. Recruitment began on July 7, 2018, and the approximate date for completion of recruitment will be July 7, 2020.