W ay back in 1995, doctors in the US and patients with diabetes had a new option to treat type 2 diabetes: a drug of a different kind from those already available: up until then, the only available drugs for T2DM in the US were the insulins, and a class of drugs called the sulfonylureas. The new drug was in a class called biguanides, and was called metformin; it was initially available only as the brand-name version Glucophage. It's now available generically, and in varying forms, including a liquid and a long-acting version. Interestingly, there had been an earlier drug in the biguanide class (phenformin) which was withdrawn from sale because of a high risk of a complication called lactic acidosis. As a result, the FDA was very leery of approving metformin, even though it was sold in Europe and many other countries for years before it was approved in the States.
But I digress. Metformin was a different class of agents to treat diabetes than those that were then available, and had an entirely different mechanism of action than the other drugs. As such, patients could be treated with various combinations of insulin, sulfonylureas, and metformin, and were able to obtain better glycemic control than before. One thing that was learned over and over was that better control required using drugs from the 3 different classes -- that is, adding a second sulfonylurea to a patient already on maximal doses of another sulfonylurea was unlikely to have any additional effect to lower blood sugar, as the two sulfonylureas would have the same mechanism of action, and hence, it was a waste of time and money to try adding the second. But adding metformin to a program for a patient on a sulfonylurea (or vice versa) was very likely to have great additional benefit. So treatment programs were devised to add a second drug from another class (note, not switch to the other class) when diabetes needed additional treatment, and then to add a third (insulin was the other class that could be used).
Another class of diabetes drugs became available soon after, in 1996: alpha-glucosidase inhibitors, when the FDA approved a drug called acarbose (brand name Precose in the US, and Glucobay elsewhere). Soon after, another was approved, miglitol (AKA Glyset). These drugs work in the intestine, slowing down the digestion of carbohydrates, and lengthening the time it takes for carbohydrates to convert to glucose, thereby facilitating better blood glucose control, but mainly influence the level of blood sugar after eating. These drugs never became popular, partially because they has minimal glucose-lowering effects, and partially because they need to be taken several times daily.
Then in 1997, another new class of diabetes drugs was approved: the thiazolidinediones, also called TZDs or "glitazones." The first agent in this class, troglitazone (brand name Rezulin), again revised physician thinking about treating diabetes: now, with a new and efficacious oral agent, insulin became a last resort for treating T2DM, as doctors used varying combinations of sulfonylureas, metformin, and troglitazone (and sometimes the alpha-glucosidase inhibitors) with considerable success. Competitors in Rezulin's class soon appeared, rosiglitazone (Avandia) and pioglitazone (Actos), and Rezulin was withdrawn from the market after it became clear that it had a risk of toxic effects on the liver.
About the same time, yet another class of diabetes drugs became available, meglitinides. The first was replaglinide (AKA Prandin); the second was nateglinide (Starlix). These were short-acting drugs that work very similarly to the sulfonylureas, require dosing several times daily, and never became very popular with the medical community.
Then there was a bit of a break in approval of new classes of diabetes drugs, until early 2005, when pramlintide (Symlin), a synthetic analog of the human hormone amylin, became available. Pramlintide was different: it was an injectible medication for people taking insulin, whether they have T1DM or T2DM. Pramlintide was another non-starter: from my point of view, and apparently from that of many others, why should people taking insulin injections also take another injection of pramlintide to supplement their insulin.
But another injectible medication, this one for T2DM patients, has fared somewhat better in the marketplace. The drug, exenatide (brand name Byetta), a drug called an "incretin mimetic", has a side effect that may explain part of its success: weight loss. It also was approved in 2005.
And a year later, the FDA approved another class of diabetes drugs, the "DPP4-inhibitors." The name of the first was sitagliptin, brand name Januvia. Another, vildagliptin (Galvus) is approved in Europe but not in the US.
Then, in 2008, the FDA approved an older drug for a new reason, for the treatment of T2DM. The drug, colesevelam, has been approved since 2000 for treatment of hyperlipidemia, and now is also approved for use in combination with other diabetes drugs. It is in a class of drugs called bile acid sequestrants.
Most recently, 2009, the FDA approved another old drug for the treatment of T2DM. The drug, bromocryptine (brand name Cycloset), is in a class called dopamine receptor agonists, and as of yet hasn't been marketed for the new indication.
So, depending on how you count (I count insulin and the injectibles as separate classes, and I'm only counting drugs approved in the US for treatment of diabetes), there are now eleven classes of diabetes drugs, and several drugs in most of these classes.
Will there be more classes of diabetes drugs? Yes, inevitably there will be. The next may be the SGLT-2 (sodium-glucose transporter-2) inhibitors; and there are others being studied, including a class called glucokinase (or GK) activators.
Much different from 1995