https://immattersacp.org/archives/2011/06/diabetes.htm

Experts offer advice on oral drugs, insulin for diabetes

Any type 2 diabetic is eventually going to need oral medications, and some on oral medications may eventually progress to needing insulin. The regimens are complex and vary from patient to patient. Learn how to individually tailor such decisions.


Progression of type 2 diabetes is almost inevitable, and with it usually comes a need for different medications. To help internists sort out which therapy to use and when, two speakers at Internal Medicine 2011 offered algorithms for oral medications, advice on using insulin, and pearls for managing patients when guidelines aren't always clear-cut.

Animal insulin has been available since the 1920s, and sulfonylurea agents since the 1940s. When Anne Peters, FACP, CDE, did her endocrinology fellowship in the 1980s, she worked on clinical trials that brought metformin into the U.S. Since then, a dozen new drugs have been introduced, “and trust me, that's going to double in the next 10 years,” she said.

Anne Peters FACP Photo by Kevin Berne
Anne Peters, FACP. Photo by Kevin Berne

Dr. Peters, director of the clinical diabetes programs at the University of Southern California, pointed out that almost all diabetic patients end up on multidrug therapy if they've had diabetes for 10 years or more. “And these [drugs] are no longer based on algorithms that are simple. It's always hard to figure out new guidelines.”

Her best advice is to treat diabetes early and aggressively. “The earlier we start treating our patients with type 2 diabetes, the better we normalize their blood sugars, the less likely it is to progress,” Dr. Peters said.

Metformin is the preferred first-line treatment for type 2 diabetes. It's well known, inexpensive, and has the best profile whether the goal is lowering A1c (by 1% to 2%) or avoiding weight gain or hypoglycemia, Dr. Peters said. Its maximal clinical effect is achieved at dosages of 1,500 to 2,000 mg/d.

Metformin does have gastroenterological side effects in about 10% of patients but is not advised in cases of renal insufficiency, she noted.

The second line of defense includes the addition of sulfonylureas, thiazolidinediones, DPP-4 inhibitors and/or GLP-1 agonists, depending on the patient.

Sulfonylureas are the most common second-line agent. They're inexpensive, but they have limited durability because they can affect beta-cell function. While this drug class creates a 1% to 2% A1c drop in the first year, A1c levels then begin to steadily increase in years two through 10.

“They're good drugs potentially initially if you just want to get somebody down, but they're not my first choice anymore because we have better drugs,” Dr. Peters said.

She more often uses thiazolidinediones in patients who show markers of insulin resistance such as dyslipidemia, central obesity or established cardiovascular disease. There's some cardiovascular benefit to these drugs in patients with established cardiovascular disease, she said.

She'll even add them as a first-line adjunct to metformin. The drugs have a durable span of effect; Dr. Peters has patients who've used them for 15 years. Risks include significant weight gain, edema and heart failure, long-bone fracture and a possible increase in myocardial infarction for rosiglitazone, “although the data for that last outcome are complex.”

She prefers pioglitazone to rosiglitazone and will sometimes prescribe Monday-Wednesday-Friday dosing in the belief that it diminishes side effects and lowers the overall drug costs while still lowering A1c levels. Dr. Peters prescribes 15 or 30 mg, not the more typical 45 mg, to reduce the risk of edema. She has not seen congestive heart failure associated with thiazolidinediones in her practice recently.

As a third drug class (or a second step, in obese patients), Dr. Peters will consider GLP-1 agonists. The drugs help patients avoid excess food intake by promoting satiety and reducing appetite. They enhance glucose-dependent insulin secretion and beta-cell functional mass, regulate gastric emptying and lower glucagon secretion after eating.

Dr. Peters will add GLP-1 agonists to the previous two classes as her “favorite all-time combination,” because of their effect on A1c (about 1% lowering), weight loss (in 80% of patients) and beta-cell function preservation. The drugs target post-meal blood glucose levels. Her patients lose 10 to 15 pounds on average over time, she said.

In considering exenatide versus liraglutide, Dr. Peters said FDA trials have shown more A1c control and weight loss with the latter but the trials differed from common practice in that they followed up on compliance with phone calls and pill counts. In real life, Dr. Peters works with the patient to determine which drug to use, especially considering that liraglutide is taken once a day and exenatide twice. “I see this as a choice for my patients,” she said.

DPP-4 inhibitors (sitagliptin and saxagliptin) increase endogenous levels of GLP-1, she said. They can achieve an additional A1c reduction of 0.5% to 0.8%, with a limited side effect profile, no weight gain or loss, and no hypoglycemia. Dr. Peters especially prefers using these drugs in elderly patients who are more fragile or in those who are somewhat closer to target. Some patients respond well, others less so which makes it important to retest an A1c after 3 months to assess whether or not the patient is responding to therapy.

“I like to keep patients in the realm of two drugs, maybe three,” she said. “I don't mind using insulin; I'm an endocrinologist. But for a lot of my patients it's a difficult treatment.”

Indeed, insulin was a confusing enough topic to fill the lecture hall of Irl B. Hirsch, FACP, an endocrinologist and professor of medicine in the division of metabolism, endocrinology and nutrition at the University of Washington School of Medicine in Seattle, who spoke on “Insulin for the Internist.”

Irl B Hirsch FACP Photo by Kevin Berne
Irl B. Hirsch, FACP. Photo by Kevin Berne

He led a crowded-to-capacity room of internists through case examples to reveal his thinking and share some of the clinical pearls he relies on to manage insulin therapy in patients.

But, he added, referral to the endocrinologist just to sort out insulin issues isn't needed.

Dr. Hirsch advocates for early insulin in type 2 diabetes, noting that its use is much higher in Europe (50%) than in the U.S. (21% to 30%, depending upon the survey), where physicians use oral medications more often. As proof, he polled the audience, which in a show of hands said they often won't begin insulin until A1c levels are in the double digits.

Too late, Dr. Hirsch said. Yet, “That's what we see in our clinic: People should have been started on insulin five years earlier, and they have a double-digit A1c.”

It's important to start insulin earlier, in his opinion, because patients with low A1c levels at baseline can ultimately achieve lower levels overall when using basal insulin alone. Three-fourths of patients who begin using insulin with an initial A1c level below 8% can eventually get below 7%, he said, while patients who don't start until their A1c levels are above 9% are unlikely to get to 7% with basal insulin alone.

Dr. Hirsch cited data on 20 patients with type 2 diabetes, a mean A1c of 8.3% and a mean body mass index of 36 kg/m2 who received a single injection of insulin glargine into the abdomen at doses of placebo, 0.5 units, 1 unit, 1.5 units or 2 units per kg of body weight.

The 0.5-unit/kg dose didn't last 24 hours for glycemic control. “You're pretty much out of insulin,” he said. The 1-unit/kg dose showed a larger rise in insulin levels at 7 hours. The doses of 1.5 units/kg and 2 units/kg had much more insulin activity at 24 hours. But, he cautioned, the standard deviations for the 1/kg-, 1.5/kg- and 2-units/kg doses showed that higher doses weren't necessarily more effective. Statistically, there was no additional glucose lowering within 24 hours of the injection for doses of 1 unit/kg and higher.

And when adding in the costs of two units/kg compared to one unit/kg, “According to this study you're wasting a lot of insulin,” he said. “Although it is possible the duration of insulin action is prolonged with increasing doses of glargine, there is no difference in insulin action the 24 hours after injection once the dose is greater than 1 unit per kilo.”

Dr. Hirsch offered another pearl: A higher body mass index does not require a longer needle to deliver insulin. A 4-mm, 32-gauge needle is effective in patients with a body mass index of up to 40 kg/m2, research has shown.

While most internists in the audience said they prescribe an 8-mm, 31-gauge needle, some responded that they use a 12.7-mm, 29-gauge one, which isn't much fun for the patient, Dr. Hirsch said.

Research shows that patients with higher BMIs have a trivial (0.2-mm per 10-kg/m2 increase in BMI) amount of thicker skin and run a higher risk of intramuscular injection when a larger needle is used: 0.5% with a 4-mm needle compared to a 45% risk with a 12.7-mm needle. And the injection site—arm, thigh, abdomen or buttock—doesn't matter for insulin analogues. It does for human insulin, he added.

Dr. Hirsch said he prescribes 4-mm to 5-mm needles for all patients. They appreciate the difference and it may help compliance, he added.

Another area Dr. Hirsch is watching keenly is glucose variability, both because of its possible role in micro- and macrovascular complications and because continuous glucose monitoring has made it easier to measure compared to fingerstick tests.

The standard deviation is what's of interest. A1c is an average measure, “and frankly it's not a very good average,” Dr. Hirsch said.

For an endocrinologist, downloading glucose meters is like taking vital signs. Dr. Hirsch does his while taking vital signs, as a matter of fact. It takes 90 seconds to download data from the meters and print them out, he said.

He's most interested in two formulas. For type 1 diabetes, if the standard deviation × 2 is less than the mean of the blood glucose, the patient is doing a fair job of control. For type 2 diabetes, the standard deviation × 3 should be less than three times the mean. “This may seem very esoteric when first talking about it, but it's not,” he said.

If the standard deviation and multiple are higher than the mean, that patient needs counseling. A patient with type 2 diabetes may be making minimal amounts of insulin, a patient may not be matching his or her food to insulin (requiring a visit from the nutritionist), or a patient may be taking insulin late or not at all, or snacking may be the problem.

“Insulins are far from perfect,” he said, “but if we can be creative our patients can usually do well.”