The options for diabetes treatment are tremendous now. Before 1995, we had only one option, sulfonylureas, and everyone was on a sulfonylurea or insulin. You still have sulfonylureas, which stimulate the pancreas, the beta cells, to secrete insulin in both basal and postprandial situations.

You also have other medications, including the meglitinides, two of them are being used; repaglinide and nateglinide. These oral medications also stimulate the beta cells to secrete insulin, but they work much better on postprandial blood glucose. You also have the alpha-glucosidase inhibitor, acarbose. This medication delays carbohydrate absorption, however it is a weak medication. The biguanides group returned to the US market in late 1994, in the form of metformin. This medication works only on the liver and reduces hepatic insulin resistance and hence hepatic glucose production. You also have two thiazolidinediones (TZDs) on the market, pioglitazone and rosiglitazone. These medications, as well, improve insulin sensitivity.

And now you have the incretins; GLP-1 analog (exenatide) and the dipeptidyl peptidase-4 (DPP-IV) inhibitors (sitagliptin). These medications also lead to improvement in insulin secretion, primarily physiologic insulin secretion, and they also suppress glucagon. Finally, you have insulin, and remember that the best physiologic replacement for beta-cell failure is insulin.

With all these options, let me ask: what is the ideal oral medication? You need medication that is effective. You need medication that will reduce A1C and fasting and postprandial blood glucose. You need medication that is safe; you have heard about the recall of troglitazone, the first TZD on the market, because of hepatic failure and some resultant deaths. You need medication that has no or mild side effects. You need medication that does not cause hypoglycemia.

Also, you want to avoid weight gain. Most of the current antidiabetic medications cause weight gain. In the United Kingdom Prospective Diabetes Study (UKPDS), in over 12 years of follow up of type 2 diabetic patients, all the medications caused a significant amount of weight gain. You also need medications that can be used in combination with other diabetes medications, still be effective, and yet have some synergistic action. You do not need medications that have many contraindications or interactions with other medications, such as statins, antihypertensive medications, or congestive heart failure medications.

You need once-daily medication. You need medication that is safe in elderly people. You need medication that is safe for renal patients. You need medication that delays the progression of the disease. Finally, you need medication that is covered by the patient's insurance.

Do we currently have any medications that fit all these criteria? The bad news is that we do not, but we have to compromise a little bit. The good news is that we are getting closer and closer to developing a very good oral medication that can be effective, can be safe, can have less side effects and less hypoglycemia, does not cause weight gain, and has many other very good benefits.

But if you would like to design treatment for diabetes, the most important thing to know is the physiology of the problem. The physiology starts when people are obese or overweight (80% of diabetics are overweight, obese, or severely obese). Those people, at a certain point, are diagnosed with prediabetes, impaired glucose tolerance, impaired fasting glucose, and later develop diabetes. For approximately 5 to 6 years, you can easily control diabetes with one oral medication or maybe a combination of oral medications. As the disease progresses, it becomes extremely difficult to manage, and you need to add a third medication, insulin, or insulin with oral medications.

So what is going on in the background of that scenario? Once people start to gain weight, they start to be insulin-resistant, and they continue to be insulin-resistant for the rest of their lives. Their beta cells will start to secrete more and more insulin to try to overcome that resistance. But once people are diagnosed with diabetes, the beta cells will continue to fail no matter what medications given to patients.

You have an early scenario during which you still have some insulin secretion; a relative deficiency of insulin. Then the beta cells start to fail and ultimately you have significant failure of the beta cells.

What you need at the beginning of this problem is to improve insulin resistance. We usually start with metformin or one of the TZDs, either pioglitazone or rosiglitazone or a combination of metformin and TZD. As the disease progresses and the beta cells continue to fail, we need to stimulate the beta cells with sulfonylureas or secretagogues. Even with combination oral medications, you cannot stop the deterioration of the beta cells; the beta cells continue to fail. At that time, the patient will need insulin. Initially, you can give long-acting insulin in combination with oral medications. But when that combination is not achieving your target, then you need to switch to split-mix insulin or long- plus short-acting insulins. This scenario is what we had been doing until recently.

But what have we been doing to prevent the beta cells from failing? It is nice to start with metformin and TZDs to improve insulin resistance, but we are currently doing nothing to prevent failure of the beta cells. Now we have medications called DPP-IV inhibitors. These medications restore glucagon-like peptide-1 (GLP-1) to normal levels, stimulate the beta cells to secrete insulin in response to oral feeding, and prevent failure of the beta cells over time. So it makes sense to use these as an earlier medication or in combination with metformin and TZDs.

What are the glycemic goals? The American Diabetes Association goals for fasting plasma glucose are between 90 and 130 mg/dL. For 2-hour postprandial plasma glucose, which is very important, the goal is less than 180 mg/dL, with A1C less than 7%. The American College of Endocrinology (ACE) goals, which I believe are much better because they encourage you to strive for more aggressive control, are a fasting plasma glucose of less than 110 mg/dL, a postprandial plasma glucose of less than 140 mg/dL, and A1C less than 6.5%. But, what we usually tell our patients verbally is that we need their fasting glucose to be around 100 mg/dL and their 2-hour postprandial glucose to be less than 200 mg/dL.

We need the A1C as low as possible without hypoglycemia and without side effects. If we lower A1C as much as we safely can, I think we are moving in the right direction. And I want to stress that initially you need to work on the fasting glucose. As A1C starts to go down, you also need to work on the postprandial glucose, since around 70% of the contribution to an A1C of less than 7.3% is from postprandial glucose. So if a patient has an A1C of 7.5% or 8.0%, it will be impossible to lower that A1C without starting to work on the postprandial blood glucose level. This may explain why many of our patients are still not at the goal.

Now, we will review the available oral compounds, how we can use them, and the benefits and the drawbacks of each one of them.

The sulfonylureas are one of the older compounds used for treatment of diabetes. You have first-generation compounds; but I believe no one is still using them. The newer generation includes glimepiride, glyburide, and glipizide in the US, and in Europe and elsewhere outside the US, they have gliclazide and glibenclamide. These medications are all generic now.

You also have the benzamido compounds, which include repaglinide and nateglinide, which are not generic yet.

These medications try to replace the natural stimulus for beta cells to secrete insulin.

The problem is that these medications stimulate the pancreas to secrete basal and postprandial insulin, which means they may cause severe and recurrent hypoglycemia if used at high doses. Unlike the hypoglycemia that can occur with insulin therapy, which is easily corrected, the hypoglycemia caused by sulfonylureas and glinides may last for a longer time and may be recurrent even after correction. Of course, for these medications to work, they require some functioning beta cells, because they stimulate beta cells.

Sulfonylureas and glinides are very potent medications. They reduce A1C by approximately 1% to 2%; but this depends on how high the A1C is when you start. The immediate problems with sulfonylureas are weight gain and allergy, because they are mostly sulfa. And as I mentioned, their main risk is hypoglycemia.

When we dose these medications, we usually start with one eighth to one fourth of the maximum dose, which you can increase as you go. Keep in mind that the maximum effective dose is not usually the maximum dose. For some sulfonylureas and glinides, the maximum effective dose is half of the maximum dose. We titrate these medications on a weekly basis, or sometimes on a monthly basis.

If you give, for example, glimepiride 1 mg, you see a drop in blood glucose levels. If you give 4 mg, you get an additional decrease; but at 8 mg, you do not get much of a decrease in blood glucose levels. So increasing dose does not mean that you will always increase the effect.

The duration of action of glinides and sulfonylureas is a little bit different. For example, nateglinide increased the insulin secretion after a meal; breakfast, lunch, and supper. Long-acting glyburide stimulated insulin secretion not only during the meal, but for maybe 2 or 3 hours after the meal as well, which can lead to hypoglycemia before the next meal. The problem is that patients during this period may eat some additional snacks to avoid, or in response to, hypoglycemia, which cause an additional weight gain.

Other compounds include the alpha-glucosidase inhibitors. They are commonly used outside the United States, but unfortunately not used as commonly here. I believe one of the reasons is that they only decrease A1C by 0.5% to 1%. These medications also cause a lot of adverse events like gas bloating and diarrhea, and have not been very successful in the United States.

Metformin is one of the most commonly used medications in the US and also outside the US. Just to give you some background, this medication was in the US market long ago. It came with a biguanide called phenformin. During that period, phenformin caused serious side effects and unfortunately, some fatalities from lactic acidosis. So the US Food and Drug Administration (FDA) recalled phenformin, but also decided to recall metformin at the same time, even though it was not associated with as many cases of lactic acidosis as phenformin. Then they realized this was a mistake; everywhere else, metformin was still used. In December 1994, they put metformin back on the market. Since then metformin has become the number one medication prescribed for type 2 diabetes.

Metformin is a medication working only on the liver. It prevents the liver from pushing too much glucose into the circulatory system, especially during the fasting situation. It requires insulin, because it improves the sensitivity of the liver to insulin, which then suppresses hepatic glucose production.

Metformin decreases A1C as well by 1% to 2%. We are not sure for how long this effect lasts; it appears that it lasts only a few months. Most adverse events are related to the gastrointestinal tract: diarrhea, nausea or GI upset. The most serious adverse event with metformin is still the lactic acidosis, but its frequency is very, very low.

We also have some contraindications for this medication, including patients with a renal problem, hepatic problem, or cardiac problem. If someone is going to have radiology work with intravenous contrast, we usually stop metformin for 48 hours before that procedure.

We usually start metformin at 500 mg/day, once- or twice-daily dosing to minimize side effects, and then we start to titrate the dose up. The effective dose of metformin is 2000 mg/day, although the maximum dose is 2550 mg/day. We have not seen any big difference between 2000 mg/day and the maximal dose. In fact, the larger dose, sometimes, is not as effective as the smaller dose, and you will get more side effects with the larger dose. In order to prolong the action of metformin and to minimize the side effects, the manufacturer has a slow-release formulation as well.

The UKPDS study, which showed that metformin failed over time, also has some very interesting information. Metformin reduced cardiovascular events and cardiovascular-related mortality in treated patients. People may think metformin is just a diabetes medication; in reality, metformin has some additional benefits. Recently, there are data that show improvement in endothelial function, improvement in fibrinolytic activity, and some other additional benefits as well.

Metformin still has a problem with safety. In terms of patients with renal impairment, you cannot prescribe metformin to any man with serum creatinine above 1.5 mg/dL, or any woman with serum creatinine above 1.4 mg/dL. The problem is in patients who are 80 years of age and older; there is a very high risk that metformin may accumulate in their circulatory system, and this can lead to life-threatening lactic acidosis.

Patients with congestive heart failure or with recent myocardial infarction have to be very cautious using metformin. Patients with liver problems, patients who are alcohol abusers, and patients with metabolic acidosis should not use metformin. So although it is widely prescribed, there are significant safety issues with metformin, especially in the elderly and in patients with chronic conditions. In reality, a lot of our diabetic patients have these problems.

The TZDs are a group of medicines that stimulates peroxisome proliferator-activated receptors (PPARs). There are many of these receptors, including PPAR-alpha, PPAR-gamma, and PPAR-delta. Fibrates, medications that improve triglycerides, target PPAR-alpha receptors. Stimulating PPAR-gamma receptors yields very interesting results as well. Once you stimulate the PPAR-gamma receptors, you improve insulin sensitivity not just in the liver, but everywhere in the body, especially in the adipose tissue and the skeletal muscles. Also, PPAR-gamma differentiates newer fat cells. So patients taking medications targeting PPAR-gamma receptors usually gain some weight, an effect that was seen with pioglitazone and rosiglitazone. PPAR-beta and -delta raise high-density lipoprotein (HDL) cholesterol, but unfortunately we do not have any compounds that have that effect on PPAR-delta.

What are the problems with and what is the benefit of TZDs? They are very effective medications that also decrease A1C by 1% to 2%. So it appears that whether you use sulfonylureas, metformin, or TZDs, you get a very nice decrease in A1C. Again, the percentage of reduction depends on how high the starting A1C was. If A1C is significantly high, you get a 2% reduction or even more. But if A1C is close to 7%, you might not get even a 1% reduction. The major problems with TZDs are edema and weight gain. Patients gain approximately 6 to 7 pounds on average over a 3-year period.

Troglitazone has caused a lot of liver problems. Because of this, it has a black-box warning that you need to measure and follow liver enzymes in those patients on TZDs. It is not a good idea to start someone who has very high liver enzymes and active liver disease on a TZD. However, new data show that pioglitazone reduces intrahepatic fat.

The initial dose is 15 mg/day for pioglitazone or 2-4 mg/day of rosiglitazone; the maximum doses are 45 and 8 mg/day, respectively. Most patients usually end up taking 30 mg of pioglitazone or 8 mg of rosiglitazone.

These are slower medications; you do not see an immediate effect. It may take a couple of weeks before you see a drop in blood glucose. You need to keep an eye on the blood glucose and based on that, titrate the medication accordingly.

In our practice, we usually tell our patients that there is a possibility that their liver enzymes will go up, and we are very cautious about this. At baseline we usually measure the liver enzymes. We also tell patients to report any weight gain or any swelling in their lower limbs. We usually start with a small dose and ask patients to come back in a short period of time to see what is going on.

Sometimes patients accumulate a lot of fluid. Some are at very high risk for congestive heart failure or have undiagnosed congestive heart failure. If someone develops edema, we usually give diuretics, perhaps furosemide or hydrochlorothiazide, and we keep them under observation because it is important that they not develop congestive heart failure.

Interest in this medication has started to rise because, in reality, it has many additional benefits other than being just a diabetes medication. There are data that show it reduces endothelial dysfunction and carotid intimal-medial thickness (IMT). There are data showing that these medications improve inflammation and increase adiponectin. Glitazone has many additional benefits. Most weight gain in patients using glitazone is related to accumulation of peripheral fat and fluid retention.

A newcomer to this field is the DPP-IV inhibitor; sitagliptin, the first medication in this category. It is an oral medication approved by the FDA in October 2006 to be used as monotherapy or in combination with metformin or TZDs. Because this medication has a very good effect on the early stage of disease, preventing deterioration of beta cells, the indication of using it as monotherapy was granted.

This medication requires a reserve of functioning beta cells to respond to it. It may not be as effective late in the course of the disease when functioning beta-cells are mostly gone. This medication should be given as early as possible in the course of the disease. It cannot be combined with sulfonylureas because both medications stimulate the beta cells. Of course, it cannot be used in type 1 diabetic patients or in patients diagnosed with diabetic ketoacidosis.

It comes in a very convenient single dose, 100 mg, taken once every day. Patients with moderate renal impairment should reduce the dose to 50 mg/day and patients with end-stage renal disease should take only 25 mg/day. It is not contraindicated in renal impairment, but it should be used in much smaller doses.

What data on sitagliptin were presented to the FDA? Two studies were presented, one 18 weeks and one 24 weeks. In the 18-week study, investigators found that this medication reduced A1C by 0.6%, and in the 24-week study, by 0.8% when given to drug-naïve patients. Monotherapy with sitagliptin may be enough in newly diagnosed, mild diabetes.

Again, the reduction of A1C percentage depends on how high the baseline A1C is. If the A1C is very high, above 9%, sitagliptin will decrease A1C by almost 1.4%. If the A1C is lower, you will get much less effect with sitagliptin; 0.7% or 0.6% if baseline A1C is less than 8%.

If you combine sitagliptin with metformin, you get an additional 0.7% reduction in A1C. If you combine it with a TZD (e.g., pioglitazone) you will also get an additional 0.7% reduction.

This slide illustrates the most important point. With metformin alone, you can see that blood glucose spikes after breakfast, after lunch, and after supper. When you add sitagliptin to metformin, you see that the peak in postprandial glucose is much less. One additional benefit that we can get with combining metformin with sitagliptin is much better postprandial physiologic blood glucose levels.

One of the problems that we have seen with oral medication is weight gain. For example, after 12 weeks, glipizide causes approximately 1 kg of weight gain (2.2 pounds). But with sitagliptin and placebo, you do not see similar weight gain, but you also do not see weight loss either. Sitagliptin is weight-neutral; it does not cause weight gain or weight loss.

What is very interesting is the possibility that sitagliptin may preserve beta cells, and these are data in experimental animals. Shown are the islets in nondiabetic mice and diabetic mice. In diabetic mice, the alpha cells, which secrete glucagon, are increased, and the volume of the islets as a whole is much less. When you put diabetic mice on sitagliptin, the islets are close to normal. So it appears this medication, in experimental animals at least, can stop the apoptosis of beta cells, and may also stimulate proliferation of new beta cells.

The inescapable problem is that, whatever medication you give to your type 2 diabetic patient, over time you will need to add additional oral medications. Whatever medication you use, you get a bit of improvement at the beginning and it persists for some time, but eventually these medications fail over time. This slide shows the results of 10 years of follow-up, and you can see the failure of these medications. That is why a lot of companies are coming out with combinations of medications.

Combinations include glyburide with metformin, glipizide with metformin, rosiglitazone with metformin, rosiglitazone with glimepiride, and pioglitazone with metformin. You also have pioglitazone with glimepiride, and sitagliptin in combination with metformin.

Why start our patient on combination therapy? With 2000 mg of metformin, A1C is significantly reduced. With 2500 mg of metformin, you do not see as much effect as with 2000 mg. What you will see is an increase in side effects. So you do not necessarily always get much additional benefit by increasing the dose.

This is a very nice example. A group of patients on glyburide are switched to metformin, and there is no improvement in fasting plasma glucose. But if you add metformin to the glyburide, you see a very nice decrease in the fasting plasma glucose level. So it appears that giving a combination is much better than switching or increasing the medication.

Here is another example of patients initially on 1000 mg of metformin. They increased metformin to 2000 mg in one group and in the other group they added rosiglitazone 8 mg to the 1000 mg of metformin. In the rosiglitazone group, 58% achieve the A1C target vs 48% in the metformin-only group. But most importantly, there were much fewer side effects.

So it is not always a good idea to keep increasing the dose of oral antidiabetic medication. May be by combining medications, you will achieve much better A1C, and also reduce the side effects to a great extent.

This is the same scenario: of patients on metformin only, 18% achieved the A1C target recommended by the American Diabetes Association (ADA). Add sitagliptin to the metformin and 47% achieved the ADA target of <7%. With pioglitazone alone, 23% achieved the ADA target, and if sitagliptin is added to the pioglitazone, 45% achieved that target. So it looks like adding another medication is a good idea.

What is the next step if these oral medications fail? We have injectable exenatide.

Exenatide has been approved by the FDA to be used with metformin, sulfonylureas, or a combination of metformin and sulfonylureas. Recently it was also approved to be used with TZDs. Used in combination, you get an additional 0.6% to 0.8% decrease in A1C.

Exenatide causes weight loss, which effect is greatest in combination with metformin over 30 weeks, and much less in combination with sulfonylureas and in the metformin/sulfonylurea combination itself. Exenatide used in combination with metformin was the most effective for weight loss.

Another option is to add glargine insulin to the oral medication. With glargine insulin or exenatide, you can get a 1% decrease in A1C. The main difference is that with glargine, patients usually gain weight and with exenatide, patients usually lose weight.

Most importantly, with glargine, you still have the pattern of 3 spikes in blood glucose after meals; but with exenatide, you do not see those increases. So it definitely appears that giving those incretins makes a big difference in the postprandial blood glucose level.

The problem with exenatide is that it frequently causes nausea; 40% to 44% of patients get nausea. Most of this nausea is mild or moderate and usually goes away over time. But some patients may get severe nausea or vomiting and cannot tolerate the medication.

The following algorithm may be helpful. When patients are diagnosed with diabetes, we usually start with lifestyle intervention. If they are controlled very well, we see them every 3 to 6 months to measure A1C, and continue as long as they are doing very well.

If they are not doing well, you have 2 scenarios. You have patients with low fasting plasma glucose and high A1C, which means that most of those patients have very high postprandial blood glucose levels, which you have to target. For those patients we have the DPP-IV inhibitors, exenatide, meglitinides, sulfonylureas, and even insulin. So you can target the postprandial blood glucose level if the fasting level is acceptable, but the postprandial level is significantly high.

If the fasting glucose is high, most of those patients have significant insulin deficiency. What do you do? The DPP-IV inhibitors, sulfonylureas, insulin, and exenatide all target that insulin deficiency. And, of course, most of those patients have insulin resistance, for which you have metformin or TZDs.

This is just the beginning of the scenario. But most important, when we design an oral medication strategy, we have to select medication based on our understanding of the pathophysiology of the disease, depending on if the patient is in a stage of insulin resistance, severe beta-cell failure, or early beta-cell failure. We have to target both the beta-cell dysfunction and the insulin resistance.

You have to stage your medication based on the stage of disease. If you know that the patient has significant beta-cell failure, you can start insulin, which is a very good medication that can be used in patients with significant beta-cell failure.

Combination therapy can also be considered earlier. Why not start with a medication for insulin resistance (metformin or TZDs) in combination with medications that preserve beta cells, such as DPP-IV inhibitors? This combination makes sense.

You treat to goal. The goal is not to control fasting and postprandial glucose only; you also need to decrease A1C as much as possible. If you reach 6.5%, you need to go lower; if you reach 6.0%, you can still go lower as long as you do not cause hypoglycemia. You also have to consider the nonglycemic effects of these medications. If the medication has an additional benefit, such as a cardiovascular benefit, it is more likely to be prescribed.

Again, I stress that insulin therapy is a replacement for beta-cell failure. And when insulin is indicated, you must start it.

Finally, there is the scenario of insulin resistance, and beta-cell failure. Early in the course of type 2 diabetes, a combination of TZDs or metformin with a DPP-IV inhibitor can be a very nice tool to preserve the beta cells and improve insulin sensitivity. If the beta cells continue to fail, you still have the sulfonylureas, and finally you have the insulin.

In 1922 the insulin era began. We are part of a new era now—the incretin era. I think incretin can be a new modality that we can use more frequently in our patients, especially at the early stages of the disease progression.