Type 2 diabetes drugs, new and old

This post centers around a recent review of the topic in American Family Physician and contains a sprinkling of some of my own thoughts on this subject. Here are some key points made in the article:

Metformin monotherapy (along with lifestyle modifications) is the first intervention.

The article goes on to say that metformin is the only oral diabetes drug proven to reduce mortality and complications. That is not true in the simple manner in which it was stated and reflects a misinterpretation of the findings of the UK Prospective Diabetes Study which showed a reduction in microvascular complications attributable to treatment with sulfonylureas and insulin. This review is not the first place I've read such a statement. It seems to be a popular myth and I have no idea how it got started unless out of an old controversy regarding possible macrovascular harm attributed to sulfonylureas, as originally reported decades ago in the University Group Diabetes Program study.

There are nuances to this controversy which I have blogged previously in multiple posts. At the risk of oversimplification, evidence seems to show that reduction of blood sugar to a certain target, no matter by what means, reduces microvascular complications but that a number of agents, including insulin, have the potential to cause macrovascular harm. It is beyond the scope of this post to fully explore that controversy but I'll have a bit more to say about it below.

All that being said the evidence points to metformin as having the best efficacy in reducing complications and it is generally accepted as the front line drug.

Initial oral monotherapy can be expected to lower the Hgb A1C by about 1%.

Furthermore each oral drug added on should lower it by an additional 1%. If a patient's initial A1C is very high, e.g. in the double digits, one might consider combination drug therapy or insulin right off the bat. However, a guideline based algorithm shown in the article does not recommend that approach. Instead, metformin monotherapy is started and then drugs are added sequentially one at a time with insulin entering the mix as early as step 2.

Pills are added sequentially in no particular order

---or, as the article puts it, “in a patient centered fashion.” This reflects the lack of high level evidence to guide clinicians in this area. The sequence continues until the A1C goal is reached. Although the algorithm shows basal insulin as an option to add on as early as step 2, complex (multiple injection) regimens are reserved for later steps.

The appropriate A1C goal is controversial.

The UKPDS target was 7. Other studies have pointed to macrovascular harm with more aggressive targets. The review suggests higher targets for patients who are older, have macrovascular disease or multiple risk factors for same.

Newer agents (that is, those other than sulfonylureas and metformin), approved on the basis of safety and their ability to lower blood sugar, have not been shown to improve clinical outcomes.

That doesn't negate the possibility of outcome improvement that may surface years later, as discussed below.

A pearl in hypoglycemia management.

If you have to treat a hypoglycemic episode in a patient whose regimen includes an alpha glucosidase inhibitor (acarbose or miglitol) oral feeding may not work. Oral dextrose or IV therapy will be necessary. Due to the inhibition of alpha glucosidase the patient cannot break sucrose down into monosacharides.

The paradox of macrovascular disease

The concern that diabetes drugs might drive macrovascular disease is controversial. Here's my perspective. The principal macrovascular risk factor associated with DM 2 is the metabolic syndrome. Mere reduction of blood glucose does not alleviate that condition. Macrovascular benefit attributable to diabetes drugs is the exception, confined to metformin and possibly pioglitazone, likely representing pleiotropic effects.

How might glucose lowering drugs cause macrovascular harm? In the exceptional case of rosiglitazone it is probably a pleiotropic effect. But in the case of multiple other drugs, such as insulin and drugs which enhance insulin secretion such as sulfonylureas, it is probably because they promote weight gain and worsen insulin resistance, known drivers of macrovascular disease. More recently it has been found that hypoglycemia, a known consequence of intensive glycemic control, impairs endothelial function and induces hypercoagulability.

Initial concerns about macrovascular harm came from the UGDP study cited above. Serious questions have been raised about the validity of that study. Nevertheless it resulted in a black box warning for sulfonylureas. Adding fuel to the controversy is newer evidence suggesting that if patients whose glucose is lowered intensively with insulin or sulfonylureas are followed long enough macrovascular benefit might eventually be seen. This was illustrated in the 10 year UKPDS follow up. In that study the benefit appeared years after differences in glycemic control had disappeared, suggesting a delayed secondary effect rather than any direct beneficial effect of glucose lowering. The secondary effect (speculation on my part) might stem from prevention of diabetic nephropathy, as CKD is known to be a powerful driver of atherosclerotic complications.