Unintended consequences: When medicine hurts more than helps

Severe liver damage caused by medications is rare, but when it does occur, it can be catastrophic, leading to liver failure and the need for a rapid transplant, or death.

Why this phenomenon happens -- why some people are injured by a drug that most can take without a problem -- is a medical mystery that several investigators are now trying to solve.

The liver, of course, is the body's guardian, providing protection from potentially harmful substances even as it stores vitamins and other nutrients from food and removes waste. But it is also vulnerable. For example, it is well known that the liver can be harmed by viruses (particularly hepatitis), excessive alcohol use and the abuse of certain drugs, especially acetaminophen.

It is also known, however, that prescription drugs and herbal products used as directed can cause severe liver injury to some people. The reasons behind these rare, idiosyncratic injuries are now being pursued by investigators. Is it genetic, environmental or a combination of both?

The answers to such questions could have wide-ranging implications.

Drugs were recently named by the Food and Drug Administration as the most frequent cause of acute liver failure in the United States, exceeding all other causes combined. And liver injuries are the major reason for the withdrawal of a drug from the market, for a new medication's failure to gain FDA approval, for restrictions on a drug's use and for warnings to physicians.

Considering this information in tandem with the important role drugs play as treatments and drivers of health costs underscores the need for answers.

"Drug-induced liver injuries are a challenging problem to solve, and research is key," said Mark Avigan, MD, director of the FDA's division of drug risk evaluation.

"Among the questions to be answered are which drugs are potentially toxic, what are the causes of rare idiosyncratic liver injury and what is the phenomenon of adaptation where the liver compensates after a mild drug-induced injury. This protective mechanism appears to be deficient in patients who develop severe, life-threatening liver damage," Dr. Avigan said.

In an attempt to answer some of these questions, the National Institutes of Health recently established a network of five clinical centers plus a coordinating center to conduct studies of patients who have experienced liver injury linked to prescription or over-the-counter drugs, nutritional supplements, alternative medicines or herbals.

One objective of the study group, known as the Drug-Induced Liver Injury Network, is to develop standardized definitions and instruments to identify and characterize cases of liver injuries.

Study objectives

Such standardization could aid communication among researchers.

"Drug-induced injuries can mimic any liver disease from acute hepatitis to liver cancer," said Leonard Seeff, MD, senior scientist for hepatitis research at the National Institute for Diabetes and Digestive and Kidney Diseases. "So there is an attempt to bring some science to this."

There are three important reasons for mounting a large study, said Paul Watkins, MD, DILIN steering committee chair and a professor of medicine and pharmacotherapy at the University of North Carolina, Chapel Hill.

First is the catastrophic nature of the problem. It can be fatal without a liver transplant.

It also presents a scientifically fascinating challenge, he said. "You'll have a drug that 10,000 or 20,000 people can take with no problems, get all the benefits and remain totally healthy, but then the next person who takes it can turn yellow and die."

Plus, the liver is a good model to study because individuals are getting exactly the same amount of exactly the same, pure substance -- the medications, he noted.

Injuries also boost the cost of drugs. "Companies can design drugs that will very effectively do what they want them to do," Dr. Watkins said. "However, after testing the drugs on thousands of people at a cost of hundreds of millions of dollars, the drug can go out into the real world and these very rare problems can surface."

"Until we understand what is going on in these rare people, why they are different and why this particular drug or natural product is having this potentially catastrophic effect, it's still going to cost hundreds of millions of dollars to develop new drugs, and drugs will remain very expensive," he added.

Within DILIN there are two separate studies. One is a retrospective study to establish a registry of patients who have since 1994 taken one of four specific drugs and developed a liver injury. The second is a prospective study that will enroll patients who recently had an adverse liver reaction after taking any drug or herbal medicine.

The four drugs to be included in the retrospective study are isoniazid, which is prescribed for tuberculosis; phenytoin and valproic acid, drugs used for seizures; and clavulanic acid/amoxicillin, for infections. These were chosen because they are widely prescribed and tend to have a characteristic presentation of liver injury that makes it relatively easy to diagnose. In addition, they have been well-studied and exhibit a variety of mechanisms that cause injury, Dr. Watkins said.

Acetaminophen is not included in DILIN because it is not a classic idiosyncratic toxin, he explained. "The way to avoid having a liver problem with [acetaminophen] is to take it as directed. These other drugs when taken as directed can cause a problem -- though only rarely."

DILIN researchers are seeking to enroll 50 to 100 people who were injured by each of the four drugs and the same number of controls, or people who took the drugs for the same length of time and were not injured. By late July, 70 people had enrolled overall.

Participants will be asked to provide blood and urine samples that will be stored in a data base along with such information as a patient's age, ethnicity and underlying health conditions.

Crystal ball not needed

"We hope with this kind of study we will begin to develop ways of predicting who is at risk for developing injury from a specific drug," said Herbert Bonkovsky, MD, professor of medicine at the University of Connecticut Medical School in Hartford, a DILIN site.

The registry will allow patients to be tracked for several years so if new hypotheses are advanced, the material will be available to test them, Dr. Bonkovsky noted.

The prospective study will enroll all people who have had a severe liver injury within the previous six months that is believed to be due to a drug or an herbal product other than acetaminophen. These patients will be followed over time to find out what happens to them as a result of their injury.

When at least five people with injuries from the same drug or herb have been enrolled, the plan is to recruit controls who took the drug or herb without injury, Dr. Watkins said.

The physicians running the centers are looking to their colleagues in primary care to refer patients they believe have had liver injury from a drug, either one of the four being scrutinized most closely, or another drug or herbal product.

Subjects are not easy to find because the condition is rare and the diagnosis so tricky. But physicians could raise their index of suspicion that a drug could have caused a liver injury, Dr. Bonkovsky said.

"Anytime you have someone with abnormal liver tests or with symptoms suggesting a liver problem, such as nausea, loss of appetite, pain in the upper abdomen, fatigue or excessive sleepiness, the physician should think, 'Does my patient have a liver problem and could this be due to a drug or other chemical?' "

A careful medical history also is important, Dr. Bonkovsky said. Physicians should ask what drugs and herbal products a patient is taking. Sometimes herbal products are overlooked.

Dr. Avigan also urges physicians to report such injuries to the FDA's online safety monitoring system, MedWatch. "But the report can't be perfunctory. It has to have good clinical narration," he advised.

"We want to know if the patient has been worked up and whether a problem with a drug has been indicated or did they have some other disease or condition. It's important that we get good reporting, because it helps us to get to the problems quickly," Dr. Avigan said. "And this is where good citizenship by physicians is important."

Pharmacogenetics gets a boost

Solving the contentious problems of drug-induced liver injury also holds the potential for gains in individualized medicine, or pharmacogenetics. "The whole long-term dream of this general field is to individualize the selection of drugs for individual patients," Dr. Bonkovsky said.

There is a great deal of enthusiasm for individualized medicine across the professional infrastructure, said Mark Linder, PhD, vice president of operations at Pharmacogenetics Diagnostic Laboratory at the University of Louisville in Kentucky.

Pharmaceutical companies now embrace the concept; the FDA is developing guidance policies and documents; and congressional staffers "know what you're talking about when you say pharmacogenetics," Dr. Linder said. "These are important indicators that the storm is coming and everything is getting in place."

Although the map of the human genome has been useful in this field, pre-genomic-era findings also have led to many discoveries, including which enzymes are required to metabolize which medications, Dr. Linder said.

"But with sequencing of the human genome we can much more quickly identify where there are differences between individuals."

Dr. Linder's lab is working toward identifying mechanisms that aid the liver in developing tolerance for a drug. Such mechanisms play a crucial role in protecting the organ.

"This is important, because we've identified an individual marker that disrupts this natural process. So that marker could be identified as a[n individual's] risk factor for liver injury," he said.