Dr. S.M.Sadikot.
Hon. Endocrinologist,
Jaslok Hospital and Research Centre,
Mumbai 400026

Life was so much simpler 25 years back. That is all the more true when one considers oral agents used in treating diabetes. I remember when I passed my M.B.,B.S., that if one knew a little bit about tolbutamide and chlorpropamide, that was more than enough to satisfy the examiners. Things really did not change even during the M.D. examinations, with the exception that glibenclamide was then the "hot" oral agent!

That was then.

Today, five classes of oral agents are available for treating Type 2 diabetes and at last count more than 40 formulations were in the market! And they keep increasing day by day.

Is any one oral agent better than the others?

More importantly, when treating a person with diabetes, how does one choose the type of drug to be used!

In order to make some sense of the plethora of oral agents available, it becomes essential to basically understand a few important points.

The first, and probably, the most important point is that none of the oral agents is insulin, no matter what name may be given to the drug by a company.

One should also understand that Type 2 diabetes is characterized by three basic abnormalities that contribute to the development of hyperglycemia:

• Impaired insulin secretion by the pancreas
• Peripheral insulin resistance mainly in the skeletal muscle
• Excessive glucose production by the liver

Type 2 patients that one sees in practice would have a combination of these three mechanisms which cause the high blood glucose levels. The problem is that the extent and severity of each of these mechanisms varies in different individuals, and the oral agent which would be most optimal for any patient would depend on which of these three mechanism plays a major role in their hyperglycemia.

Although there is no hard and fast rule for this, it is widely accepted that in lean type 2 patients, impaired insulin secretion is the predominant defect, while insulin resistance tends to be less severe than in the obese variety. Insulin resistance and hyperinsulinemia are the classic abnormalities of obese individuals with type 2 diabetes.

All the oral agents available do not have the same mechanism of action. Thus, one must know how a class of oral agent acts in order to choose the appropriate drug.

Before, we discuss the modalities of how to manage a patient using an oral agent, it would be worthwhile to understand the types of oral agents available for use.

Sulfonylureas

The sulfonylureas have been available since the 1950s and have historically been the first line of pharmacologic therapy for diabetes.

During World War II, French scientists working on the antibiotic potential of modified sulfonamides found that the mice that they were experimenting upon died unexpectedly. Such unexplained deaths also occurred amongst a few of the soldiers who had been administered these modified sulfonamides as an antibiotic. On closer examination, the cause of the deaths was found to be due to hypoglycemia. Unfortunately, in the milieu of the war scene, this effect of the sulfonamides was not paid much attention to and it was only as late as 1955 that the first oral sulfonylurea, carbutamide, was first introduced in the market as an oral hypoglycemic agent. The following year, tolbutamide and chlorpropamide were also introduced. In later years, the so-called "second generation" sulfonylureas, glibenclamide and glipizide amongst them, were brought into the market. Relative late comers on the sulfonyurea scene are gliclazide and glimepiride.

The sulfonylureas are often classified as belonging to the first or second generation. The first generation sulfonylureas are rarely used now.

Sulfonylureas work primarily by stimulating pancreatic insulin secretion, which in turn reduces hepatic glucose output and increases peripheral glucose disposal. There is some evidence that the newer sulfonylureas such as gliclazide and glimepiride may also have some action in reducing insulin resistance.

The precursors of the biguanides have been around for much longer. Way back in 1918, guanidine derivatives were used as oral hypoglycemic agents. Synthalin A, a homologue of guanidine, continued to be used for many years and would have been still more frequently used if insulin had not become available for routine use. Synthalin A is the precursor of the modern day biguanides, two of which are commonly used here being phenformin and metformin. Although some doctors continue to use phenformin, it's use has declined to such an extent that for most purposes, when one takes of biguanides, one is usually referring to metformin.

Metformin works primarily by inhibiting hepatic glucose production and improving insulin sensitivity. It does not stimulate insulin secretion from the pancreas and, when used alone, does not cause hypoglycemia or hyperinsulinemia. Metformin may increase glucose utilization in peripheral tissues to a lesser degree, by decreasing insulin resistance in muscle cells. Metformin may also improve glucose levels by reducing intestinal glucose absorption. Metformin should be started with a single daily dose of 500 mg, taken with a meal. The patient's response to therapy can be checked at one- to two-week intervals and will guide how the dosage should be increased. Metformin dosage can be increased by one tablet (500 mg) per day at one- to two-week intervals until glycemic control is obtained or a total daily dose of 2,000 mg is reached. The maximal daily dosage should not exceed 2,550 mg; most patients get little additional benefit from dosages of more than 2,000 mg per day.

Acarbose is an alpha-glucosidase inhibitor that slows down the breakdown of disaccharides and polysaccharides and other complex carbohydrates into monosaccharides. The enzymatic generation and subsequent absorption of glucose is delayed and the postprandial blood glucose values, which are characteristically high in patients with type II diabetes, are reduced with acarbose. AGIs do not prevent the absorption of carbohydrates and complex sugars, but they do delay their absorption.

Delaying the absorption of carbohydrates is a unique mechanism among oral diabetic medications for lowering HgbA1c levels. The effectiveness of this mechanism is one of the physiologic characteristics of type 2 diabetes. Patients with type 2 diabetes demonstrate a delayed or sluggish insulin response from the pancreas to a glucose (a meal) load. By delaying the absorption of glucose, the insulin response is more matched to the serum glucose, resulting in less postprandial hyperglycemia and a lowering of the HbA1c. The AGIs also demonstrate a lowering of total insulin output of the pancreas, increased insulin sensitivity, a variable but mild decrease in triglycerides, with no effect on patient weight.

One disadvantage with the use of acarbose is that it is to be taken along with the first bite of a meal. Moreover, it has to be taken three times daily with meals. These factors often lead to non compliance and a decrease in the efficacy of the drug. Acarbose may be started at 25 mg three times daily, taken at the beginning of each main meal. The dosage can be adjusted at four- to eight-week intervals. The maximal recommended dosage is 100 mg three times daily (50 mg three times daily if the patient's body weight is 132 lb [60 kg] or less).

Repaglinide from the meglitinide drug class, acts like an extremely short-acting SU (an insulin secretagogue) and is potentially useful as a SU replacement. The effect of repaglinide on the pancreas is very similar to that of the SUs. Repaglinide, like the SUs, blocks the potassium channels on the pancreatic islet beta-cells, which causes an influx of calcium into the cell and increasing the secretion of insulin. There appears to be 2 similar potassium channels on islet beta-cells, one of which is predominantly affected by repaglinide and the other is predominantly affected by SUs.

The repaglinide-affected potassium channel appears to be glucose dependent, which may partially explain why repaglinide is associated with a much lower incidence of hypoglycemia.

What makes repaglinide clinically different from the SUs is its ultra-short half-life (1 hour). Repaglinide is taken just before or with meals, and the stimulation of the pancreas is limited only to a brief time around meals. Because of the short duration, the patient does not have continuous high levels of insulin and the resulting adverse effects.

Its biggest advantage over the other oral hypoglycemic medications is that it allows for flexible timing and missed meals.

Repaglinide stimulates the release of insulin from the pancreatic beta cells by closing ATP-sensitive potassium channels. It is rapidly absorbed, allowing for a rapid onset of action. It was found to be three to five times more potent in stimulating insulin release than the sulfonylurea glibenclamide. However, it does not stimulate the release of insulin in the absence of glucose. Repaglinide augments the early insulin response and decreases prandial and postprandial glucose excursions.

Repaglinide's rapid onset and short duration of action make multiple daily doses necessary. The physician can often turn this characteristic to good use by instructing patients to take repaglinide immediately before each meal. This approach is especially useful for patients who eat at irregular hours or sometimes skip meals altogether. For patients who are just beginning treatment with oral medications or whose HbA1c is less than 8 percent, the recommended starting dose is 0.5 mg taken 15 minutes before each meal. For patients who have previously been receiving oral medication for diabetes and who have an HbA1c of 8 percent or more, a starting dose of 1 or 2 mg taken 15 minutes before each meal is recommended. The maximal recommended dosage is 4 mg per meal, for a maximum of four meals (16 mg) a day. If a patient skips a meal, he or she should not take the tablet. If a patient adds a meal, a tablet should be added for that meal.

The thiazolidinedione (TZD) class of oral hypoglycemics ( popularly known as glitazones) was developed in 1997 and offers a new mechanism for treatment of type 2 diabetes. The first, troglitazone was taken off the market in 1999 because of its association with hepatic toxicity. Rosiglitazone and pioglitazone have been available since 1999.

The primary effect of TZDs is peripheral, with increasing insulin sensitivity and increased glucose uptake. The TZDs have some effect on hepatic glucose uptake and sensitivity to a lesser degree. They do not stimulate the pancreas to produce more insulin.

TZDs are hepatically metabolized and thus can be used safely in patients with renal dysfunction. They can be dosed once daily, although rosiglitazone works better with twice-daily dosing. Reports have suggested that rosiglitazone works better in women, but the reason for this is not known.

Both pioglitazone and rosiglitazone are approved for monotherapy and in combination with metformin, SUs, and insulin.

Besides their effect in lowering the blood glucose levels, both drugs also have notable effects on lipids. The current data show that pioglitazone has a minimal effect on low-density lipoprotein (LDL) cholesterol levels and a favorable effect on high-density lipoprotein (HDL) cholesterol and triglyceride levels. Rosiglitazone is thought to be basically lipid neutral.

With use of glitazones, patience is required. Blood sugar levels may show a significant reduction statistically in as little as two to four weeks, but the maximum effects are not seen until two or three months have passed.

Having discussed some basic aspects of the oral agents available today for use in diabetes, the next step would be the treatment plan itself.

A simple plan is given in the Figure below.

Diet and exercise remain the mainstay of diabetes management and all patients should be advised about the changes in their diet and prescribed an exercise schedule. The next critical step would be to find out those patients in whom it would be mandatory, or better, to initiate further management with insulin. Such patients have been discussed in the section dealing with Insulin Therapy. Having done this, one would then be faced with the vast majority of patients who may need oral hypoglycemic agents.

Theoretically, one should put all such patients through a trial with diet and exercise before starting them on any oral agent. This is precisely what all of us have been taught in medical college. In practice, one needs to have a slightly different approach.

I feel that any patient who comes with an initial fasting blood glucose value which exceeds 250mg% or who is severely symptomatic with polyuria and polydipsia or who is losing weight inspite of an adequate diet, can be started of with an oral agent along with diet and exercise. In fact, some of these patients may do better if insulin is started initially.

If we decide to start an oral agent initially due to these reasons, my oral agent of choice would be a sulfonylurea. Why a sulfonylurea, and not another class of oral agent? Metformin would not be ideal as the patient is already losing a lot of weight and metformin will only increase this weight loss. The glitazones have a significant time lag before showing their full activity whilst the others are too mild for such patients.

For the purposes of our discussion, let us accept that we have decided to initiate treatment with glimepiride, although this should not be taken to mean that glimepiride is the best sulfonylurea available. I usually start with a relatively small dose, say 2mg. per day. The patient is then monitored closely. I usually see them after a week with the fasting and the postmeal blood glucose levels or preferably with results of self monitored blood glucose levels.

In quite a few patients, the blood glucose would have decreased significantly and there is a significant amelioration of the symptoms. If the values have reached acceptable levels, then I usually decrease the dose of the oral agent. In many cases this would mean managing the patient on diet and exercise alone or halving the dose of the sulfonylurea till the next visit after about two to three weeks. Many patients who have been off the oral agent will continue to show acceptable control at this time and they would continue to be managed with diet and exercise alone. Patients who had been given a decreased dose of the oral agent, if they show such an acceptable level of control at the second visit would be taken off this smaller dose to see whether they can subsequently be managed on diet and exercise alone. If at a subsequent visit, the control seems to have worsened in spite of adequate dieting and exercise, then the oral agent should be reintroduced at the small dose at which the patient had been started.

A few of the patients will show a decrease in the blood glucose levels and some amelioration of the symptoms but this may not have reached an acceptable level. Such patients would be continued on the same small dose to see whether this would cause a further lowering of the blood glucose levels. At the next visit, if the control is within acceptable limits, then the patient is managed as discussed in Scenario 1. In case, there is no further improvement in the control, then I would treat the patient as in Scenario 3.

Quite a few patients may show some improvement in the control parameters, but this may not be to any significant extent. Such patients would need an increase in the does of the oral agent. I would increase the dose of the glimepiride to say 4 mg per day and follow them after 3-4 weeks to judge control. At this time, some of these patients may show an adequate control and they would then undergo a decrease in the dosage to see whether they could be managed with the smaller dose. If at the next visit, the control is maintained, then they may be continued on this small dose or one may even try a trial with a still further decrease of the oral agent.

If the control in spite of taking 4mg of glimepiride per day is still not at acceptable levels, it would be possible to further increase the dose to say 6mg of glimepiride per day. In a few patients this increase may lead to an acceptable control, and the patient is then managed with this dose.

It should be realized that only a few patients who did not show an acceptable control at 4mg per day would then go on to show a good control with 6mg. per day. Most of these patients would be better managed as in Scenario 4.

The point that I would like to stress upon is that one should try and achieve adequate control with the SMALLEST possible dose of the oral agent.

A number of patients may not be optimally controlled even when the patient is 4mg of glimepiride. At this stage, it would be more prudent to review the patient clinically rather than make a further blind increase in the dose of the tablets. In my opinion the maximum dosage of the oral agents, given by the drug manufacturer's and also mentioned in the text books is not valid for our patients. These dosages are too high and in Table 1, I have given the maximum doses of the oral tablets as mentioned in the books and literature and the dose at which I feel the patient warrants a close review. For instance, the maximum dose of glibenclamide is reportedly 20mg per day (4 tablets). I feel that this is too high for most our indian patients and that it would be the very rare person who did not show a substantial response to two tablets of glibenclamide (10mg per day) and then would go and show "control" with a higher dose.

The commonly used oral hypoglycemic agents, the dosage at which it would be prudent to "review" the patient, as well as the maximum advisable dose which may be used are:

In the detailed review that I carry out in such circumstances, the first point is to see whether the patient is complying with his diet and exercise. This is obviously not easy to judge as most of the patients would tend to swear that they are following their prescribed diet. In the section dealing with diet, I have discussed ways through which it would be possible to judge the patients dietary compliance. If one finds that the patient is not keeping to his diet, then the focus of management would be to increase the dietary compliance rather than increase the dose of the oral agent.

When faced with a patient who is complying with the diet and exercise prescription and who does not show the neccessary response to say, 4mg. per day of glimepiride, it would be worthwhile to rule out the presence of associated conditions which may interfere with the action of the oral agents and not allow for good control. The commoner of these conditions are detailed in Table 2 and have been discussed in detail in the section on "Insulin therapy".

After having ruled out the presence of these conditions, one would be faced with one of the following three scenarios in the vast majority of such cases.

The patient may be an insulin-requiring diabetic. In this condition, although the patient does not go into ketosis if not given insulin, he cannot be optimally controlled without the addition of some exogenous insulin. These patients, and there are a number of such patients in our country, are usually normal or even slightly underweight. The therapeutic strategy in such circumstances would be to give the patient a small dose of exogenous insulin in addition to a small dose of the oral agent. In practice, I usually start with a small mixture of a short acting and an intermediate acting insulin (say 8 units of each) given before breakfast along with an oral agent. Often, with slight modification of the doses, one finds that the patient can not only be optimally controlled, but the patient himself says that he feels much better!

If the patient is overweight, then I consider using a small dose of metformin or a glitazone along with the sulfonylurea. In practice, I give the patient 250mg of metformin two to three tijes a day with the meals in addition to the sulfonylurea. If a glitazone is used, this could be 15mg. of pioglitazone given once daily or 2mg. of rosiglitazone given twice a day. The combination often shows good results and the doses are then carefully adjusted so that the patient maintains the control with the smallest possible doses of the tablets.

If the patient shows a relatively high postprandial levels, then addition of an alpha-glucosidase inhibitor, such as acarbose may help Managing the Overweight (Obese) Patient

A stringent regimen of diet and exercise should be the main stay therapy of all obese patients.

If the patient does not show the desired results,

Metformin is the drug of choice as the initial oral agent in patients who are overweight, especially those with a raised waist to hip ratio, if diet and exercise fail to achieve the aims of control. Phenformin is available in India but its use is on the decline and metformin should be preferred to phenformin when a biguanide is decided upon.

In order to minimize the side effects such as nausea and other abdominal problems, I usually initiate therapy with small doses of metformin, 250mg once or twice daily, taken with meals. If desired result is obtained, decrease the dose of metformin by half tablet (250mg) decrements, to maintain control at smallest possible dose of the OHA.

If the target control is not observed, the importance of diet and exercise is re-enforced. Increase metformin in small doses (250mg) at a time and monitor control. Further increments are usually made every 2-4 weeks till the dose of metformin (500 mg twice a day) is reached.

At this time, check for diet and exercise compliance. Change in weight is an excellent parameter to judge diet and exercise compliance.

If patient is following his diet and exercise schedule:

Check for the presence of any condition which may interfere with the action of metformin. These conditions are given above. If any of these conditions are present, they should be treated. After treatment, patient may show optimal control. In this case, the procedure outlined in Scenario A is followed.

In the absence of any condition interfering with the action of metformin, re-enforce diet and exercise regimens, the dose of metformin can be gradually increased to the maximum. But this rarely leads to better control in most patients.

Usually when the dose of metformin has reached around one gram per day in divided doses and the desired control is not seen, it would be better to combine this regimen with other drugs. In an obese person, a glitazone would be ideal but care is to be taken about the weight gain which is associated with the use of glitazones. 15mg of pioglitazone daily, or 2mg. of rosiglitazone taken twice a day, often leads to a significant lowering of the blood glucose levels. Combination therapy with a sulfonylurea can be started. In this case, a small dose of metformin (250mg two to three times a day) is combined with, say, glimepiride ( 2mg daily). Such a combination therapy with slight adjustment in doses also often leads to target levels.

If the patient shows very high postprandial levels with fairly acceptable fasting and preprandial levels, one could think of the additional use of an alpha-glucosidase inhibitor such as acarbose. If the desired levels are achieved, doses of the oral agents should be slightly decreased gradually to observe if control can be maintained at the smaller doses. The aim, again, is to manage the patient with the smallest doses of the drugs.

Some patients do better with a combination of metformin with insulin.

I would now like to discuss some of the more important side effects of the oral agents.

Most of the side effects associated with sulfonylurea therapy are mild, infrequent, and occur less often with the second-generation agents; they include:

• Mild gastrointestinal upset
• Skin reactions
• Rashes
• Purpura
• Pruritus
• Weight gain (this effect can be minimized or prevented by increasing emphasis on dietary habits)

Hyponatremia, fluid retention, and the Antabuse reaction have also been reported with the use of chlorpropamide. The major complication of sulfonylurea therapy is severe hypoglycemia, which has been more of a problem with chlorpropamide than with any other agent because of its long half-life and duration of action. Hypoglycemia is also more common in individuals who consume large amounts of alcohol and skip meals. Glipizide is reported to have a lower risk of hypoglycemia because it is metabolized to inactive by-products that do not have hypoglycemic activity. Other reactions are rare and include:

• Hematologic reactions
• Leukopenia
• Thrombocytopenia
• Hemolytic anemia
• Cholestasis (with and without jaundice)

I would like to discuss sulfonylurea associated hypoglycemia in a little more detail. Whilst the manifestations and the management of hypoglycemia have been discussed in a seperate section, there is one point that I would like to stress on. This is the hypoglycemia that occurs with the use of chlorpropamide. This drug continues to be relatively widely used here, as many doctors feel that they are well versed in using this particular sulfonylurea.

Although the text books mention that the effectiveness of chlorpropamide lasts for around 36 hours, from a clinical viewpoint, and especially in the elderly and those with renal problems, the effect may be prolonged up to 4-5 days. Therefore, any patient who has a hypoglycemic reaction of any severity should be closely monitored for at least 4-5 days, as they are prone to go into recurring episodes of hypoglycemia. This is the reason why I usually admit all patients who have had a severe reaction of hypoglycemia with this drug. They often have to be kept on glucose drips for 72-96 hours so that all traces of chlorpropamide are washed out from the body. There can be nothing more embarrassing than to be called in to treat a hypoglycemia, give intravenous glucose, see the patient recover and go away only to be called back after a few hours for the same problem! If one were to do the same thing again, it would be distinctly possible that another physician would be called in if the hypoglycemia occurred for the third time!

A disulfiram (Antabuse) like effect can occur with the use of sulfonylureas, especially chlorpropamide. Antabuse is a drug used in the treatment of alcoholism. When a patient is given this drug and then he takes alcohol some or all of the following reactions can occur: a-feeling of warmth or extreme heat and redness, flushing, nausea, vomiting, sweating and thirst. In severe cases, there may be difficulty in breathing, chest pain which may mimic a myocardial infarction, hypotension and syncope. Some patients who are on sulfonylureas and partake of alcohol may also manifest similar symptoms. This is especially true of chlorpropamide and the reaction is popularly known as the "Chlorpropamide- Alcohol Flushing Syndrome (CPAF)"

The following case history of one of my patients illustrates the syndrome adequately.

"Mr. K.D., male aged 32 years was admitted to the hospital having being brought by some friends with the history that the patient had complained of feeling unwell, very hot in the face and throbbing headache. He had complained of feeling breathless and of quite severe chest pain. He had vomited twice prior to the admission. No other details were forthcoming. The C.M.O. made a diagnosis of an acute coronary episode and the patient was admitted into the I.C.C.U. When he was examined by the Registrar, it was noticed that the face was flushed, the pulse was 110/ minute, feeble, the BP was 90/60 but the physical examination was otherwise negative. His ECG was within normal limits as was his Chest X-ray. The random blood glucose was 102mg% and there was no sugar or acetone in the urine. His blood was collected for cardiac enzymes and along with the other relevant treatment, a glucose saline drip was administered. Within a couple of hours, the patient was much better and seemed to recover completely. At this time, a further history was elicited from the patient. He was a newly diagnosed diabetic and had been put on Tab. Chlorpropamide (250mg) twice daily about 3-4 weeks prior to the admission and he had been taking the tablets regularly. The patient volunteered that a similar, but milder episode had occurred a week back when he had been to another party and had taken alcohol. That episode which had been milder and dominated by nausea and vomiting had been attributed to "gastritis".

In view of this statement, the diagnosis of CPAF syndrome was made. The blood reports were all normal and there was no further change in the ECG. The patient was taken off the oral agent and controlled with insulin which in my opinion should have been used in the first place in view of the weight and age of the patient. The patient has now been followed up for two years and remains well without any problems when he does take an occassional drink."

It is felt that this syndrome is seldom seen in our country. I feel that often the diagnosis is missed, especially in the mild cases. Often when the patient complains of headaches and flushing, it is attributed to the drinking itself or if he manifests gastrointestinal symptoms, as in this case, the diagnosis is something like "gastritis"! The point to be remembered is that such a reaction can occur and one should be aware that we may be facing this problem when diagnosing a patient with a relevant history.

Although, most of the cases have been described with chlorpropamide and therefore the name of CPAF syndrome, this may be seen even with other sulfonylureas.

Whilst chlorpropamide is known to be associated with intrahepatic cholestatic jaundice, this may be seen rarely with other sulfonylureas.

Chlorpropamide and tolbutamide are also associated with the inappropriate secretion of the antidiuretic hormone. This leads to a fall in the serum sodium levels and the patient may present with lethargy and headaches. Often the lethargy is diagnosed as due to the diabetes itself or worse, to the laziness of the patient who, it is thought, is using this as an excuse to eat more or not exercise. In severe cases, the patient may present with stupor, convulsions and coma. This is particularly true of elderly patients, and those that may be taking diuretics for hypertension along with a salt restricted diet. Once again, it is extremely important to recognise the presence of this syndrome so that corrective measures can be taken. This is one diagnosis that is often missed in clinical practice.

This syndrome has not been reported with the second generation sulfonylureas but an antidiuretic hormone action should always be kept in mind as a differential diagnosis when faced a patient who complains of inordinate lethargy, stupor or in whom the serum sodium levels are found to be lower than normal for no apparent cause.

Previously, one of the main concerns with the use of oral sulfonylurea drugs was their association with cardiovascular mortality. This came into focus with the findings of the University Group Diabetes Programme (UDGP) Study. In this, it was reported that an oral sulfonylurea, tolbutamide, lead to a greater incidence of cardiovascular mortality when given to patients as compared to a control group of diabetics given a placebo. This has lead to quite a heated controversy about the role of oral sulfonylureas in the routine management of diabetes. I do not intend to discuss the UGDP study or the controversy in any further detail, except to say that the generally accepted opinion, today, and with which I completely agree, is that oral sulfonylureas when used "appropriately" are safe for use in diabetes. This has also been borne out by many other studies which have NOT shown any increase in the cardiovascular morbidity or mortality when sulfonylureas have been given to diabetic patients, and I continue to routinely use them when required.

Side effects associated with metformin therapy include a metallic taste in the mouth, anorexia, diarrhea, nausea and abdominal discomfort, which may contribute to weight loss. The side effects can be reduced if the drug is taken with meals. Side effects can also be reduced by beginning treatment with a low dosage and increasing the dosage slowly until optimal blood glucose control is achieved. The most serious complication of metformin therapy is lactic acidosis. This problem is rare when the drug is given to appropriately selected patients.

Lactic acidosis is a rare complication of metformin therapy and has a high mortality rate. Most of the cases of metformin-associated lactic acidosis occurred in patients for whom the drug was contraindicated, ie, patients with renal dysfunction. Metformin should not be prescribed if the serum creatinine is greater than 1.5 mg/dL in men or greater than 1.4 mg/dL in women. Metformin is also contraindicated in patients with significant hepatic disease, cardiac insufficiency, alcohol abuse, and any hypoxic condition or history of lactic acidosis. Metformin should be temporarily discontinued 1 to 2 days before any dye studies so that serum metformin levels are low if the patient develops renal failure from the dye.

In any patient who is hospitalized with an acute severe illness, metformin should be temporarily discontinued.

• impaired renal function
• impaired hepatic function
• any serious cardiovascular disease
• patients prone to alcohol abuse
• underweight patients

Side effects of alpha-glucosidase inhibitors include diarrhea, flatulence, abdominal bloating and cramping, which are reversible on discontinuation of the drug.

Gastrointestinal side effects may resolve with continued treatment. A low starting dosage and careful titration can help minimize these side effects.

Acarbose does not cause weight gain or hypoglycemia when used alone. However, symptoms of hypoglycemia may occur when used with a sulfonylurea. Oral glucose should be used to treat hypoglycemia instead of sucrose because acarbose inhibits the breakdown of sucrose into fructose and glucose.

Acarbose, particularly in dosages of 50 mg three times a day, may cause elevations of the serum transaminases ( hepatic enzymes). It is therefore recommended that serum transaminase levels be checked every three months in the first year of therapy and periodically thereafter.

Acarbose is contraindicated in patients with inflammatory bowel disease, colonic ulceration, partial intestinal obstruction or a predisposition to intestinal obstruction, chronic intestinal diseases associated with marked disorders of digestion or absorption, and conditions that may deteriorate as a result of increased gas formation in the intestine.

Acarbose is contraindicated in patients with renal dysfunction because they are excreted primarily by the kidneys. Acarbose is also contraindicated in patients with cirrhosis.

The most common adverse effect of repaglinide is hypoglycemia, although at a lower incidence than that of the SUs. Other adverse effects that occasionally occur include headaches, upper respiratory infection symptoms, arthralgia, and back and chest pain. Other, minor side effects which may be seen include nausea, diarrhea, constipation, vomiting and dyspepsia;

Most side effects reported are mild. Commonly reported ones included upper respiratory tract infections, headaches, sinusitis, muscle pain, tooth disorders, and sore throats.

Weight gain of about 4 to 7 lbs. occurs with all the glitazones. This is of some concern because weight gain increases insulin resistance.

Pioglitazone and rosiglitazone both appear to cause fluid retention, which can make the effects of congestive heart failure or other fluid overload diseases worse.

Unlike troglitazone, statistically both pioglitazone and rosiglitazone do not have hepatic toxic effects. Only 2 cases of hepatic toxicity have been reported with rosiglitazone use during the past 2 years, and both cases had many other risk factors. To date, pioglitazone has had no reported cases of hepatic toxicity.

Liver functions must be monitored every two months in the first year of therapy and regularly thereafter.

Any person using the glitazones who presents with symptoms such as unexplained nausea, vomiting, abdominal pain, fatigue, anorexia and dark urine or jaundice must have the glitazone stopped immediately and the person investigated for a liver disorder.

This is another aspect that has important clinical implications. Other drugs can affect the hypoglycemic effects of the oral agents, especially sulfonylureas, in two ways. They may either impair glucose tolerance or cause hypoglycemia and thus would be expected to interfere with the action of the oral agents. This interaction may be indirect, since the interfering drug may act by the same mechanisms even in non-diabetics. Diuretics, glucocorticoids, estrogen compounds and diphenylhydantoin compounds impair the hypoglycemic actions of the oral agents. Conversely, drugs such as salicylates, propranalol, oxytetracycline, disopyramide, MAO inhibitors, quinine (and alcohol) can indirectly potentiate the hypoglycemic of effect of the oral agents.

Drugs may also act indirectly, in the sense that they would be known to cause interference only when the patient is taking oral sulfonylureas. Such drugs affect the action of the oral agents by interference with the absorption, distribution, metabolism, or excretion of the oral agent.

Although many drugs may have these actions, the Table below gives a list of some commonly used drugs which have been shown to significantly effect the hypoglycemic potential of oral sulfonylureas. Any patient, who is on an oral agent and if any of these listed drugs are added to the treatment regimen, should be considered to be at added risk for the development of hypoglycemia. I have seen an elderly lady who was tightly managed with chlorpropamide go into hypoglycemia when Septran was added to treat a urinary tract infection! Although such cases may be relatively rare, they do occur and drug interactions should always be kept in mind when managing a diabetic on oral sulfonylureas. Conversely, oral sulfonylureas may also interfere with the action of other drugs. One important drug in this respect are the oral anticoagulants. Sulfonylureas are known to potentiate the action of the oral anticoagulants and the dose of this drug will need to be monitored and adjusted if the patient is put on an oral sulfonylurea agent.

AIthough most of the drugs listed below have been shown to have an interaction with the older sulfonylureas like tolbutamide and chlorpropamide, such interference can also occur with glibenclamide and glipizide although possibly, to a lesser degree.

Commonly used drugs which may interfere with the action of OHAs:

• Glucocorticoids
• Diuretics, especially thiazide group
• Estrogen compounds
• Thyroxine
• Diphenylhydantion sodium

Commonly used drugs which may potentiate the action of OHAs:

• Sulfonamides
• Chloramphenicol
• Coumarin group of drugs
• Phenylbutazone
• Oxyphenbutazone
• Clofibrate
• Salicylates

Rarely, oxytetracycline, propoxyphene, chlorpromazine with orphenadrine.

Some Important Points I have consistently reiterated the point that when starting an oral agent, one should start with small doses, increments should be gradual and that once control is achieved every effort should be made to see of the desired control can be maintained with the smallest possible doses of the oral agent.

I feel that it is important for us to realize that every person with diabetes is different. By this, I mean, that every person will react to the oral agent in a different manner. Let us take three patients who in spite of following their diet and exercise schedule show a fasting blood glucose levels of 200mg%. If we decide to start these three patients on 2mg. of glimepiride each and then judge the effect after a few days, we may find significantly different results. One patient may not show any significant change, whilst the second may show a drop of his fasting values to, say, 150mg%. The third might show a value of 70-80mg%!

Each person responds to the oral agents differently and if we had started all the patients on a higher dose then it is possible that the third person may have gone into a severe, and possibly, life threatening hypoglycemia. How often have many of us seen patients who present with a blood glucose of even 400mg% and then within days of starting a small dose of an oral agent come with levels of 120mg%!

Unfortunately, there is no formula to judge how any one person will react to the oral agent and it would be better to have a "slow and steady" approach to every patient rather than play "Russian Roulette" with your patient's life!

That is also the reason why the dose increments should be done gradually. The time when the maximum effect of the oral agent is seen also differs and abruptly maximizing the dose just because one does not see much of a lowering at the first follow up can be dangerous.

What is the need to review the patient when a certain dose, which is much below the maximum dose prescribed by the manufacturer, is reached without the patient showing the desired response.

I feel that the "review" doses which are given above are adequate for any monotherapy with a single oral agent to show an adequate response. A further blind increase in the dose of the drugs rarely helps in a further lowering of the blood glucose levels but definitely increases the side effects of the oral agent used.

If one does not see the adequate response to the oral agent, it would be more prudent to review the diet and exercise compliance and also look for the presence of any associated condition which may interfere with the desired effectiveness of the oral agent.

There are some other important aspects of treatment which need to be kept in mind.

It should also be realised that a sort of a "Somogyi Effect" can and does occur with the use of oral hypoglycemic agents. In this phenomenon, which is discussed in more detail elsewhere, there occurs a rebound hyperglycemia as a consequence of, possibly, subclinical hypoglycemia. From a clinical viewpoint what occurs is that the patient taking an excessive dose of the oral agent manifests low blood glucose levels but this may not be recognised. As a consequence of these hypoglycemic levels which may even be subclinical or very subtle in its manifestation, the counter-regulatory mechanisms of the body come into play and shoot the blood glucose up to hyperglycemic levels which in certain instances can be quite high! One often comes across patients who have a very high blood glucose level and are taking 4-6 tablets of glibenclamide (or an equivalent high dose of the other oral agents). When the dose of the tablets is reduced to an acceptable level, the blood glucose levels decrease! Therefore, in order to avoid this problem, the dose of the oral agent should be started at the smallest possible value and gradually increased so that excessive medication is avoided.

Another important aspect which is seen fairly commonly in India and in many developing countries is the concept of "insulin requiring" diabetes. In fact, many authorities feel that around 30-40% of our Type 2 patients fall into this category.

What is meant when one is said to have insulin requiring diabetes. These are patients who do have some beta cell function so that they are able to secrete insulin. But they do not have the ability to secrete adequate amounts of insulin. As they can secrete some insulin, they do not go into ketoacidosis routinely in the absence of exogenous insulin therapy. This differentiates them from the typical Type 1 patients.

The vast majority of such patients are underweight or have normal weight, although here again, there is no hard and fast rule.

Let us not forget that none of the oral agents available presently are insulin. Although oral agents which can increase the secretion of insulin from the beta cells do have some effect, this increase in the amount of insulin secreted will never be sufficient to meet the requirements of the patient especially in times of "stress", even if the body is made more sensitive to the insulin secreted through the use of drugs such as glitazones and metformin. Thus, irrespective of the permutations and combinations of oral agents used, even in their maximal dosages, one would not see the desired control.

After all, a lame horse will NEVER win the Derby, no matter how much and how hard you whip it!

The way to manage such patients is to offer a combination of insulin injections along with insulin sensitisers such as glitazones and/or metformin, depending on the individual patient.

The last aspect is that of "OHA Failure".

There are a few patients who have never taken an oral agent, especially the sulfonylureas and meglitinides, and who in spite of being suitable candidates for this therapy, show practically no response to these drugs. Such patients may have what has been termed as showing "primary failure".

Before a patient can be classified into this category, certain other reasons for this failure should be ruled out.

1. Is the patient complying with his diet and exercise regimen?
2. Is there enough insulin secreting capacity for the oral agent to act? Such patients may not neccesarily always show the presence of ketones in the urine.
3. Are there any other problems which prevent the oral agent from manifesting an effective response?
4. In the patient on any medication which is antagonistic to the action of the oral agent?

Once these factors have been ruled out, then there is a distinct possibility that the patient truly shows primary failure. What then is to be done? If the patient is underweight or oven normal weight, then one must control the blood glucose levels with insulin in addition to diet and exercise. If the patient is obese, then I usually put them on a very strict diet and exercise so as to lose weight and possibly add a small dose of metformin or a glitazone provided there are no contraindications for them. Alpha-glucosidase inhibitors such as acarbose may be of help.

More common than those who show a primary failure are those that manifest "secondary failure" These are patients who have shown optimal control with an acceptable dose of an oral agent initially, but who for no obvious reason show a deterioration of the blood glucose control. Once again, before one accepts that the patient is manifesting secondary failure, one should make sure that the patient is complying with his diet and exercise schedule. Non- compliance in this area is the most common cause of the deterioration of the control. Initially, many patients are well motivated and will follow the diet and exercise regimen and thus can be optimally controlled with a small dose of the oral agent. After a while of optimal control, the motivation often decreases and the patient does not keep to his diet and therefore, the blood glucose levels rise.

Some of the causes which need to be ruled out and the corrective actions for these conditions is shown below.