Saturday , August 18 2018
Home / Conditions / Type 2 Diabetes / Joslin’s Diabetes Deskbook, Updated 2nd Ed., Excerpt #44: Pharmacotherapy of Type 2 Diabetes, Part 3

Joslin’s Diabetes Deskbook, Updated 2nd Ed., Excerpt #44: Pharmacotherapy of Type 2 Diabetes, Part 3

Apr 28, 2014

Richard S. Beaser, MD


This week’s excerpt covers the following topics:

  • Which medications reduce insulin resistance
  • The action of biguanides
  • How thiazolidinediones work
  • The risks of thiazolidinediones
  • Liver toxicity with medications

Medications that Improve Insulin Action

One could make the argument that all patients with type 2 diabetes could benefit from using a medication that reduces insulin resistance. Insulin resistance is felt to be the hallmark of type 2 diabetes and should be present in everyone who is properly diagnosed with this type of diabetes. In fact, it is present in people with type 2 diabetes before diabetes becomes clinically apparent.

Medications in this group of classes share the following characteristics:

  • They require the presence of endogenous or exogenous insulin to have their effect.
  • The patient must have insulin resistance for these medications to be effective.
  • As these medications reduce insulin resistance rather than increase insulin quantity, they tend to be more effective at higher glucose levels. They are also not as likely to cause significant hypoglycemia as treatments that increase insulin levels.
  • General comments: The one biguanide currently available in the United States is metformin, although in the past phenformin was available and others have been tried in various parts of the world. The primary mode of action of medications in this class is to decrease hepatic production of glucose, therefore, the most predominant clinical manifestation is lowering of the fasting glucose level, which is the parameter most reflective of hepatic glucose production. To a lesser extent, these medications may also increase peripheral glucose utilization.
  • Anticipated efficacy: Studies have suggested that treatment with metformin can result in a reduction of the A1C level by about 1.5% to 1.8%. In addition, data suggest that metformin may have some small beneficial effect on lipid profiles and can promote weight loss. The most common adverse events are gastrointestinal, especially diarrhea, affecting 30% of people starting metformin. However, it is usually mild, lasting about 2–3 weeks after initiation of therapy, and results in only a 4% to 5% drug discontinuation. The incidence of diarrhea may be reduced by taking the medication during or after meals and titrating the medication dose more slowly. This problem may also be reduced by use of the extended-release formulation. Subclinical reductions in vitamin B12 levels requiring no clinical intervention can also occur, affecting about 7% of people using metformin.
  • Clinical dosage: Metformin is usually started as 500 mg orally twice daily, although 500 mg once daily can be used for a few weeks in patients in whom gastrointestinal disturbances occur. The use of the extended-release formulation would allow once-daily dosing and often cause less gastrointestinal side effects. Also, formulations with glyburide or glipizide provide for easier combination treatment. Dose response increases as the total daily dose is increased to 2000mg daily. However, if the desired efficacy is not achieved at this dose level, adding a second agent to the therapeutic regimen should be considered. Doses higher than 2000 mg daily do not generally increase efficacy.
  • Adverse effects: Lactic acidosis was a concern with a previous biguanide, phenformin, which is no longer on the market. Compared with phenformin, lactic acidosis associated with metformin therapy is extremely rare due to differences in metabolic pathways. The reported incidence of lactic acidosis in metformin-using patients is 0.03 cases/1000 patient-years. While the mortality rate for those people who do develop lactic acidosis can be quite high, the actual numbers of people dying from metformin-induced lactic acidosis is comparable to the mortality from sulfonylurea-induced hypoglycemic coma. Metformin-associated lactic acidosis occurs primarily in patients with specific contraindications to drug use or with conditions known to increase the risk of developing lactic acidosis. These include:
    • renal disease (serum creatinine: >1.5 males, >1.4 females, or abnormal creatinine clearance
    • liver dysfunction
    • acute congestive heart failure (CHF) or chronic CHF under drug treatment
    • history of alcohol abuse/binge drinking
    • acute or chronic metabolic acidosis

Metformin should be withheld in conditions predisposing to renal insufficiency and/or hypoxia, including:

  • cardiovascular collapse
  • acute myocardial infarction
  • acute CHF
  • severe infections.

Metformin should be temporarily discontinued at the time of or prior to the use of iodinated contrast media or major surgical procedures and withheld for 48 hours subsequent to the procedure. It should be reinstituted only after renal function has been re-evaluated and found to be normal. An earlier requirement to stop metformin 48 hours prior to a contract study has been eliminated.

  • General comments: The thiazolidinediones are a relatively new class of diabetes medications; the first member of this class, troglitazone, was introduced in the United States in 1997, but withdrawn in 2000 due to liver toxicity. The two remaining medications in this class, pioglitazone and rosiglitazone, do not share this significant adverse characteristic. Thiazolidinediones are thought to work primarily by decreasing peripheral insulin resistance, affecting tissues such as muscle and adipose. Their effect is to increase basal and insulin-stimulated glucose transport.

Many feel, however, that all of the efficacy of these medications cannot be explained solely on the basis of the peripheral effects, and there is evidence that they may also have some effect in reducing hepatic glucose output as well. Thiazolidinediones have no glucose lowering effect in the absence of insulin, do not produce hypoglycemia when used alone, and have been shown to have some lipid lowering and antihypertensive effects as well.

All compounds contain the thiazolidinedione structure, with additional moieties added to affect bioavailability and potency. The apparent mechanism of action involves binding to nuclear receptors that regulate gene expression at the transcriptional level. Thiazolidinedione receptors include the peroxisome proliferator-activated receptors (PPAR family).

There has been some suggestion that this class of medications may have some effects beyond direct glucose control. Early studies suggest that use of these medications may lead to prolongation of the survival time of β-cells and could reduce the risk of endothelial damage. However, more data are needed before we can conclude that these beneficial effects can be ascribed to this medication class, or quantitate the impact. Other, more recent studies suggest some benefits by medications in this class in slowing the progression of the development of diabetes or blunting macrovascular events.

Medications of a different class that bind to multiple PPAR receptors ("Dual PPAR’s") and may have significant effects on other related metabolic functions such as lipid metabolism were under development in the past, but development was stopped due to concern about adverse events. At this writing, there is some ongoing interest in finding a member of this class that might pass the safety screening, as they could be useful pharmacologic tools.

  • Anticipated efficacy: At this writing, there are few conclusive head to-head studies. As studies of these medications individually vary according to inclusion/exclusion criteria, starting level of glucose control, and study design, comparisons are difficult. However, variations may be determined in the future if more head-to-head studies are performed.
  • Clinical dosage: There are differences in clinical doses that result from variations in binding affinity among the medications in this class. These doses result in similar clinical effect:

_ Pioglitazone doses: 15 to 45 mg/day _ Rosiglitazone doses: 4 to 8 mg/day

  • Adverse experiences:
    • The most significant side effect, which is seen with both medications, is a tendency for edema, which can result in a dilutional anemia or exacerbate congestive heart failure (CHF). These medications are absolutely contraindicated in patients with NYHA class

111 or 1V CHF and should be used with caution in people who may be at risk for the development of CHF.

Patients using these medications also tend to gain weight, often greater amounts than might be expected from the reduction in glycosuria and the fluid retention. However, that weight gain tends to be more subcutaneous fat rather than the potentially metabolically active visceral (intra-abdominal) fat.

A meta-analysis of studies of rosiglitazone suggests a greater risk of coronary artery events. However, a subsequent double-blind study did not find any increase risk with the use of rosiglitazone, throwing into question the findings of the meta-analysis. Yet, to further compound the issue, subsequent studies question whether rosiglitazone does, in fact, have the same level of safety as pioglitazone, and the controversy continues at this writing.

While most people tend to find comfort in a randomized double blind study over a meta-analysis, the ongoing debate should be monitored for further findings on these issues, and potential impact on usage of this medication. This concern does not seem to be shared by piaglitazone. ACCORD and the VA Diabetes study showed NO increased risk for MI’s in rosiglitazone-treated people.

A concern about increased fracture rates and osteoporosis in post-menopausal women has been noted, and use of these drugs in such patients should be accompanied with caution and appropriate monitoring.

  • Drug interaction profiles: Differences in drug interactions among these medications result from differences in cytochrome pathways for metabolism. (See Medication Summaries at the end of this chapter.)
  • Liver toxicity: There has been considerable concern about the relationship between use of medications in this class, particularly troglitazone, and idiosyncratic hepatic injury. Small numbers of people treated with troglitazone were reported as having mild elevations of liver transaminase enzymes, but there were a few with more significant liver damage, and some deaths. With the availability of other thiazolidinediones, pioglitazone and rosiglitazone, which seemed to have much better safety profiles, troglitazone was voluntarily removed from the market on March 21, 2000.

Nevertheless, persistent concerns over hepatotoxicity lead to recommendations for hepatic monitoring with the use of rosiglitazone and pioglitazone. Such recommendations were included in their initial package inserts upon release in 1999. As one might predict from statistical probabilities of occurrences, rare patients using these other medications have and will suffer some hepatic dysfunction and hepatic enzyme elevations (> 3 times the upper limit of normal). However, subsequently, and as of this writing, there seems to be enough experience to suggest that there is not likely to be a significant causal relationship between medication use and serious hepatic failure. Therefore, the liver function monitoring guidelines were reduced, and the current recommendation is to check the liver enzymes (ALT and AST) prior to the initiation of therapy, and then periodically thereafter, per the clinical judgment of the healthcare professional. However, as these types of recommendations are subject to further change, practitioners using these medications should watch for updated reports on hepatic function or precautions, should they be released, and for any changes in screening recommendations for all members of this medication class.

These are important medications in the control of diabetes. Their use, either alone or in combination with other treatments, has allowed considerable numbers of patients to achieve more effective glucose control than they ever had previously. Ironically, many of the elevations in liver enzymes seen are the result of fatty liver, rather than drug effects, and treatment to improve glucose control often leads to improvements in liver enzyme elevations, rather than worsening. 

Initial doses and titration instructions are outlined in the Medication Summaries. As a general rule, it takes a minimum of two months to get a full sense of the efficacy of these medications in a given patient, and for some patients an even longer period of time may be needed. Thus, the titration is relatively slow, usually performed at intervals dictated by the timing of liver function monitoring.

If patients are using other diabetes medications and a thiazolidinedione is added, the dose of the other medication(s) should be continued at the same dose initially. However, as the dose of the thiazolidinedione is titrated, the dose of the other medication(s) may need to be reduced if glucose levels begin to fall, particularly below 120 mg/dl. Patients utilizing thiazolidinediones often differ in their response to treatment. Some will show a substantial response to therapy, while others have a minimal response or no response at all. For those who respond, the efficacy can be impressive. Much of the data reported in the literature about thiazolidinedione efficacy is based on "intent-to-treat" studies in which patients are included in the drug treatment group based on random assignments. Thus, in an intent-to-treat analysis, responders and nonresponders are combined in the final efficacy analysis. While this is appropriate for clinical research and required for FDA submissions, in a clinical practice, patients who do not respond to a medication are usually not treated with that medication for a prolonged interval. Thus, over time, those patients who continue to use thiazolidinedione therapy are usually those with a clinically useful, and often quite impressive, response.

Weight gain has been seen as a side effect of thiazolidinedione therapy. This has been ascribed to both reduced glycosuria from the improvement in glucose control and the edema that can occur with treatment with these medications. However, many people think that the weight gain may be a result of other drug effects, such as increasing pre-adiposite differentiation. Most of the weight gain is felt to be subcutaneous fat rather than the more atherogenic and metabolically active intra-abdominal fat.

Studies of lipid effects have varied according to initial lipid and glucose levels of the study populations and therefore are difficult to compare. Most medications in this class lower free fatty acids and triglycerides, which may be part of the mechanism resulting in the drug’s glucose lowering effect; these lipid changes lead to increased glucose uptake in muscle and decreased hepatic glucose production.


The medications that improve insulin action are effective as treatment for anyone who has diabetes and insulin resistance. As the medications reduce insulin resistance — allowing the insulin that is present to work more effectively — they also require the presence of insulin in order to be effective. However, the source of that insulin may be either endogenous or exogenous. Patients who have become insulinopenic and who are receiving insulin injection therapy, but who have significant insulin resistance that impacts their diabetes control, might benefit from treatment with these medications.

Next Issue: Medications that Slow Glucose Absorption

Copyright © 2010 by Joslin Diabetes Center. All rights reserved. Reprinted with permission. Neither this book nor any part thereof may be reproduced or distributed in any form or by any means without permission in writing from Joslin. Note: Joslin does not endorse products or services.

You can purchase the updated 2nd Edition of JOSLIN’S DIABETES DESKBOOK at:

Please Note: Reasonable measures have been taken to ensure the accuracy of the information presented herein. However, drug information may change at any time and without notice and all readers are cautioned to consult the manufacturer’s packaging inserts before prescribing medication. Joslin Diabetes Center cannot ensure the safety or efficacy of any product described in this book.

Professionals must use their own professional medical judgment, training and experience and should not rely solely on the information provided in this book to make recommendations to patients with regard to nutrition or exercise or to prescribe medications.

This book is not intended to encourage the treatment of illness, disease or any other medical problem by the layperson. Any application of the recommendations set forth in the following pages is at the reader’s discretion and sole risk. Laypersons are strongly advised to consult a physician or other healthcare professional before altering or undertaking any exercise or nutritional program or before taking any medication referred to in this book.