COMMENTARY

Dapagliflozin and Glucosuria: Weight Loss and Risks

Lynda A. Szczech, MD

Disclosures

April 02, 2012

The US Food and Drug Administration (FDA) held an advisory committee meeting in July 2011 to evaluate the new drug application for dapagliflozin, a novel diabetes drug, and voted against recommending approval.[1] In February 2012, the FDA notified the drug manufacturers, AstraZeneca and Bristol-Meyers Squibb, that it would not approve the drug at that time.[2] These 2 companies intend to work with the FDA to determine the next steps to work toward approval. Thus, it is relevant and essential that potential prescribers know the mechanism of action of dapagliflozin, the clinical trial results, and the major issues that have been raised.

Mechanism of Action

The kidneys filter and resorb approximately 180 g of glucose per day, using the 2 sodium-dependent glucose transporters (SGLT1 and SGLT2) in the proximal tubule. Of these, SGLT2 is responsible for the transport of glucose from the proximal convoluted tubular lumen into the blood; it plays a greater role in resorption than SGLT1 and is more specific to the kidneys. Dapagliflozin is among the first in the new class of medications known as SGLT2 inhibitors.

Glucose is freely filtered in both diabetic and nondiabetic persons. In someone with normal blood glucose levels, the proximal tubule resorbs all the glucose that is filtered, resulting in a negative urine dipstick test. However, when the amount of glucose filtered increases because of elevated blood glucose levels, the delivery of glucose overwhelms the ability of SGLT2 to resorb the greater load. Dapagliflozin blocks the SGLT2 receptor and prevents resorption of glucose, even at normal serum glucose levels and a normal filtered load.

Dapagliflozin produces a dose-dependent and sustained reduction in both fasting and postprandial serum glucose levels through increased urinary glucose excretion, and subsequently it promotes lipolysis. Longer-term use results in reduced hepatic glucose production and improved insulin sensitivity with dose-dependent reductions in body weight, which is thought to be related to calorie loss in the urine and a mild osmotic diuresis. In studies of dapagliflozin among patients with type 2 diabetes, patients lost approximately 3 kg of body weight during therapy.

Efficacy in Clinical Trials

In clinical studies,[2] dapagliflozin reduced A1c values by 0.54% to 0.89%. Thus, the efficacy of dapagliflozin is similar to that of older agents, but it is less potent than newer agents. The advantage lies in the effect of the glucosuria on metabolic syndrome. Dapagliflozin removes glucose from the system similar to the way in which a diuretic removes sodium. It neither increases insulin levels nor decreases insulin sensitivity.

Safety in Clinical Trials

Because of its mechanism of action, particular attention was focused on renal safety in the clinical development program. In the phase 3 trials, glomerular filtration rate (GFR) decreased immediately by 2.6 to 4.2 mL/min, depending on the dapagliflozin dose among patients receiving study drug and compared with those receiving placebo. The decline may have represented osmotic diuresis, and GFR was similar to that in the placebo arm by about 24 weeks. However, in the combined trials comprising approximately 1200 patients receiving dapagliflozin and 1100 patients receiving placebo, the number of patients experiencing a renal event (defined as a 50% increase in creatinine or an adverse event related to renal impairment) was low and similar between treatment arms, at 1.2% for dapagliflozin and 1.4% for placebo.

In the comprehensive evaluation of safety,[2] a higher risk for certain infections was seen among persons receiving dapagliflozin. Although the incidence of urinary tract infections was similar between the active group (4.34%) and the control group (3.54%), the occurrence of vulvovaginitis, balanitis, and other genital infections differed significantly. These infections occurred in 4.78% of the active treatment group but only 0.80% of control participants (difference, 4.23%; 95% CI, 2.70%-5.82%). This risk for infection represents a signal with respect to a medically minor complication that is easily diagnosed and treated; however, as another safety signal, it also supports the biological plausibility of an increased risk for bladder cancer.

Incidence rates for the diagnosis of new cancer in the dapagliflozin vs control arms were similar for all cases and types of cancer taken together, at 1290 and 1280 cases per 100,000 patient-years, respectively. However, small imbalances were observed when individual cancers were examined separately in patients who developed breast or bladder cancer.

Nine women in the dapagliflozin group but only 1 woman in the control group were diagnosed with breast cancer during the course of the studies. Although these numbers are small, the safety signals that have resulted in other medications being pulled from the market have often been related to events so infrequent that not even a single case was seen in the phase 3 trials. Fulminant hepatic failure is a classic example for which the FDA has now developed guidance as to which worrisome changes in liver function tests should prompt attention to this possibility.[3] Thus, consideration must be given to even the smallest of differences, regardless of their likelihood of being related to treatment or to chance (a type I error). The cases of breast cancer among women receiving dapagliflozin were a mixture of stage I and stage II, and all were diagnosed within 1 year of treatment initiation, with 2 diagnosed in the first 2 months. The scientists who evaluated these data suggest that given the time span during which breast cancer develops, this supports the presence of the lesions in a clinically undetectable state at the time of enrollment.

With respect to bladder cancer, 9 men were diagnosed in the dapagliflozin arms and 1 man in the control arms. Of these 9 men, 5 had hematuria (ranging from trace to 3+) at baseline, supporting the possibility that disease was present before enrollment in at least some of the cases. Only 1 patient received this diagnosis within 2 months of participation; 3 cases were diagnosed within the first year, and 5 were discovered more than 1 year after enrollment (range, 393-727 days).

When considering the plausibility of safety signals such as these, the preclinical (animal) data can supplement the evaluation of a potential signal. With respect to dapagliflozin, neither breast nor bladder cancer, nor preneoplastic or neoplastic changes, were detected in 2-year carcinogenicity studies in mice and rats. However, in light of the higher risk for genitourinary infections and the potential that the change in urine composition (higher concentration of urinary glucose) plausibly could affect the function of transitional cells of the bladder, such signals cannot be discounted.

Conclusions

The history of the clinical development of dapagliflozin is interesting for many reasons. First and foremost, as a nephrologist, I find the mechanism of action of dapagliflozin fascinating, and I wonder whether this is what it was like when the concept of the sodium diuretic was first proposed. In my own experience, after decades of using glucose in the urine to reflect proximal tubular dysfunction, particularly among patients with HIV who develop Fanconi syndrome while taking certain antiretrovirals, the therapeutic value of induced glucosuria will take some getting used to.

However, even more important is the safety evaluation process that must be rigorously attended to in the development of new therapeutic targets. Small signals in settings of very limited power must be considered carefully. Investigators must consider details supporting their individual likelihood of being related to therapy in conjunction with preclinical data and the potential plausibility of the relationship. It is a slow process but one that appropriately reflects the rather obvious statement that even one avoidable cancer death should definitely be avoided.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
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.
Post as:

processing....