April 2009
Feature Story

Karen Lusky

People who develop chronic physical pain often require a combination of the art and science of medicine to reclaim or maintain a normal life. And the clinical laboratory is increasingly weighing in on the scientific side of the pain management equation with testing that can improve the safety and efficacy of opioid and other pain medications.

The lab’s role in pain management includes verifying that patients are taking their prescribed medications as directed and providing therapeutic drug monitoring (drug concentrations) to help determine if the medication is doing what it’s intended to do—“give people relief from pain,” says Paul Jannetto, PhD, assistant professor of pathology and director of clinical chemistry toxicology at the Medical College of Wisconsin in Milwaukee.

The No. 1 reason urine opiate drug screening is done by pain management physicians is to monitor patients for compliance, said Dr. Jannetto in a presentation at the 2008 American Association for Clinical Chemistry meeting. But some clinicians are also using therapeutic drug monitoring and pharmacogenomic testing as part of pain management, especially in difficult cases.

While statistics show that pain testing is underused, says Robert Middleberg, PhD, laboratory director at NMS Laboratories in Willow Grove, Pa., and a consultant to the CAP Toxicology Resource Committee, more and more physicians are interested in it. That’s evident, he adds, by the growing number of pain management labs “sprouting up.” In addition, “insurers are covering the testing at profitable rates for laboratories,” he says, which tells you they see an actuarial benefit to anteing up.

Yet pain management testing is no slam-dunk in terms of always providing easy yes or no answers, even in the realm of urine drug screening, which many physicians require patients who take opioids to undergo continually, Dr. Jannetto said.

On the upside, urine drug screening, which provides a two- to three-day detection window for most opiates, is “fast, automated, inexpensive, and noninvasive” for monitoring patients, he noted. But “screening” is the operative word. The screening tests are typically competitive immunoassays based on competition between drug in the patient’s sample and labeled drug for the antibodies’ binding sites. The antibodies are directed against drug groups or classes—opiates, for example. Thus, they can’t be used to identify or quantitate the concentration of a specific opiate such as morphine, codeine, or heroin. A more specific alternative chemical method (GC/MS) must be used to obtain a confirmed analytical result.

A patient taking prescribed hydrocodone who tests positive on the urine opiate screening assay could have sold that medication and used heroin, morphine, codeine, or even oxycodone, all of which can turn the screening test positive, Dr. Jannetto cautions.

He advises physicians to order confirmatory testing to pinpoint what the patient is really taking in cases, for example, where the person displays erratic behavior in the office or is suspected of diverting the medication.

Confirmatory testing uses mass spectrometry, which, as NMS’ Dr. Middleberg explains, provides “a molecular fingerprint of a compound based on structural characteristics of the substance.” The testing identifies not only the parent drug but also its metabolites, which can be tricky for physicians to interpret without guidance from the laboratory.

As an example, Dr. Jannetto recounted a case involving a 45-year-old truck driver for whom morphine was prescribed for worsening back pain. The physician ordered a drug screen on the man to test for opiates and illicit drugs. The opiate screen came back positive, as expected, but the physician wanted to make sure the man was actually taking morphine and not heroin or another opiate. So he ordered confirmatory testing (GC/MS), which came back positive for morphine and hydromorphone.

The question then became whether the patient may have actually been taking hydromorphone obtained from another source since it is a separate pain medication (trade names: Palladone, Dilaudid) and a minor metabolite of morphine. To determine if that might be the case, the laboratory and clinician had to look at the relative concentrations of morphine and hydromorphone.

If the hydromorphone level is less than the morphine level, then it could just be a metabolite of morphine, Dr. Jannetto said. If it’s much higher, it’s more likely that the patient has been taking hydromorphone. The patient’s concentrations, however, showed that the hydromorphone level was minor compared with morphine and, thus, consistent with the patient just taking his morphine.

What if the patient had been taking heroin, which also metabolizes into morphine? To detect that, Dr. Middleberg says, the lab does confirmatory testing that looks for both morphine and 6-monoacetylmorphine (6-MAM), an intermediate metabolite that’s formed as heroin is converted into morphine. “The 6-MAM is more stable in the blood than heroin and thus can show up in the urine,” Dr. Middleberg explains. And “once in the urine, it remains relatively stable as long as bacteria and other potential changes,” such as pH, do not occur.

Physicians sometimes also receive unexpected negative results from urine drug screens for patients who are supposed to be taking an opioid or other medication. To prevent false-negatives, Dr. Jannetto said, clinicians and labs have to understand what the assays they are using will detect. Interpretation of results must also take into account that urine concentrations can vary extensively with fluid intake and other biological variables. In addition, they should realize that cross-reactivity varies among test manufacturers’ kits.

As a cautionary tale, Dr. Jannetto recounted how an elderly patient who was taking methadone for chronic pain at one of the pain management clinics repeatedly tested negative on the urine opiate screen. The physician accused her of diverting or not taking her medication and dismissed her from the pain management program. But the woman returned to the clinic with legal counsel and claimed that she had indeed been taking her medication and wanted to continue treatment at the clinic.

At that point, the physician decided to consult with the laboratory for clarification, Dr. Jannetto reported. “I told him the urine opiate assay doesn’t cross-react with methadone at all,” which is why it’s testing negative. “And we actually have a methadone screening and a confirmatory assay. And sure enough when the original [urine] sample was retested using the methadone screen­ing and confirmatory assay, it was positive for methadone.” The doctor thus had to apologize to the patient and reinstate her in the pain management program.

Lesson learned, Dr. Jannetto said: Physicians should never act solely based on a drug screening result. Screening tests provide only a preliminary analytical test result.

Oxycodone urine screening assays also cause confusion for the physicians, Dr. Jannetto said, because the lab’s urine opiate screening assay may not cross-react with oxycodone at all or very well. “So doctors will say, ‘My patient is on oxycodone and sometimes tests positive and sometimes not on your assay—what’s going on?’ The answer is that whether the person tests positive may have to do with his or her hydration state and the level of oxyco­done in the urine. The lab actually has a separate oxycodone screen with a lower cutoff that specifically detects oxycodone and its metabolite [oxymorphone],” he said.

Of course, says Gwen McMillin, PhD, the most common reason a drug screen comes back negative is that the patient is not taking the drug or taking it less frequently than prescribed. But “perhaps the patient didn’t realize he had to produce a urine sample until he got to the office, and then drank a lot of fluids to produce the sample,” says Dr. McMillin, assistant professor in the University of Utah Department of Pathology and medical director of toxicology at ARUP Laboratories. “Or the patient could have accelerated metabolism or drug elim­ination, or have an enzyme inducer that’s ramped up his metabolism. In rare cases, a person with Celiac disease or Crohn’s disease might not absorb a medication.”

Some pain patients adulterate their urine samples so they’ll test negative for illicit drugs. Dr. Jannetto sees such samples about five percent of the time in his laboratory. The patient can drink a lot of water to dilute the urine, which drops the drug below the cutoff on a screening assay. The alternative is to add products to the urine that contain acid, nitrites, and various other compounds that interfere with the immunoassays, he said. Third, patients can buy drug-free urine on the Internet.

Pain management physicians typically do not do witnessed urine collections, Dr. Jannetto noted. But simple countermeasures can help flag potentially adulterated samples. You can test the urine-specific gravity to identify very dilute urine, and look at the pH to make sure the patient hasn’t added acid to the sample. “If the pH is less than 3, that’s suspicious.”

Dr. Jannetto routinely advises physicians doing pain management testing to run creatinine on urine samples, which is very cost-effective.

“If the creatinine is less than 20 mg/dL, that person is well hydrated,” which the physician might consider when interpreting negative results, he advised. Some samples his labor­atory receives “have no creatinine [<5 mg/dL], urea, no nothing ...” And that, of course, isn’t human urine.

Urine screening is most vulnerable to sample tampering that affects results. But some forms of adulteration, Dr. Middleberg says, can also affect mass spectrometry results, even when the lab is using state-of-the-art technology. That’s one reason labs do adulteration testing in the form of pH, nitrites, oxidants, and creatinine, he says. Nitrites, which are in products used to adulterate urine testing, “tend to have a particular effect on the ability to measure can­nabinoids in urine by mass spectrometry,” he says, “especially depending on the pH of the urine.”

Urine drug monitoring using mass spectrometry won’t detect patients taking a higher dose of a drug than prescribed, Dr. Middleberg says. The clinician could “attempt to get at that with blood or serum testing, but even there you have to consider [pharmacogenomics] and other pharmacokinetic differences, so it’s never easy.”

It’s also not always easy for phy­sicians to figure out what’s going on with a patient when testing suggests that the person may have been doctor shopping or overtaking medication. In fact, new evidence-based pain management guidelines for use of opioids in chronic non-cancer pain touch on that issue, says Roger Chou, MD, lead author of the guidelines by the American Pain Society and the American Academy of Pain Medicine (Chou R, et al. J Pain. 2009;10[2]:147–159). In such cases, you have to ask, “Is it really abuse and addiction, which has a very specific definition ... or is it due to inadequate pain treatment or depression or other psychological issues that the person is self-treating?”

“There is a lot of nuance there,” Dr. Chou cautions, whereas the previous messages about such issues have been more “black and white” and may have caused some patients to be taken off opioid drugs when “maybe we could have managed it a little bit better.”

While not widely used, therapeutic drug monitoring on blood samples can help physicians manage more complex cases. “Great candidates” for TDM, says Loralie Langman, PhD, associate professor at Mayo College of Medicine and director of the toxicology and drug monitoring laboratory at Mayo Clinic in Rochester, Minn., are patients who are “atypical”—that is, requiring higher or lower doses of a drug or experiencing significant adverse drug reactions.

Physicians who use TDM to troubleshoot, however, will be somewhat in the dark without the patient’s baseline because “what’s toxic in one person may be therapeutic in another,” she notes. Yet that can change for an individual patient as the person develops tolerance to the opioid. “Performing TDM is not a bad idea” when dealing with a drug that doesn’t have a well-established therapeutic range or is known to have a large interpatient variability, especially when compliance is an issue, Dr. Langman says.

“Switching a patient from one opioid to another may also be a good time to do TDM,” Dr. Langman adds, “as the patient’s pain may return and a clinician may want to monitor blood levels to find a therapeutic baseline for that patient or establish compliance.”

TDM can also be done on nonopioid medications used in treating pain. ARUP’s Dr. McMillin thinks it makes sense to use TDM for anti-epileptic drugs and potentially for the anxiolytics, such as the benzodiazepines. For example, the anti-epileptic drug Neurontin (gabapentin) is popular as a pain medication, she notes. “And TDM to identify blood levels may be helpful because the absorption is saturable—that is, you can provide higher dose but the patient won’t absorb any more of it after a certain point.” And “TDM will help you identify the threshold for absorption.”

Nancy Bratanow, MD, clinical director of Midwest Com­pre­hensive Pain Care Center, Milwaukee, uses TDM for patients on Neurontin and other neuromodulators for pain to see if they are absorbing the medication and to identify high levels of the drug in the blood. There are defined levels for these drugs when used to treat seizures, she says, but not for treating pain. “The dose of the neuromodulators required for pain control, for example using Neurontin, can be higher than that required for seizure control.”

As for pharmacogenomic testing, Dr. Jannetto says physicians rarely test for CYP2D6, an enzyme responsible for metabolizing 20 to 25 percent of prescription medication and over-the-counter products. “And in the case of opioids/opiates—oxycodone, hydrocodone, and codeine —it’s an important player.”

A poor metabolizer of CYP2D6, which two to 10 percent of most ethnic groups are, says Dr. Jannetto, will build up higher concentrations of the parent drugs oxycodone or hydro­codone, and may end up with more side effects. “In the case of co­deine, poor metabolizers of CYP2D6 won’t get any pain relief because the analgesic properties of codeine come from its CYP2D6 metabolite morphine.”

As an alternative to doing CYP2D6 genotype testing, the clinician can also identify poor metabolizers by doing TDM. But even if clinicians identify a poor metabolizer, says Dr. Jannetto, there are no formal guidelines for how to adjust the person’s pain medication dosages. Thus, “option B,” he adds, may be to start the person on a low dose of an opioid and titrate it up until the person gets pain relief. Or the clinician could put the person on a medication metabolized by a different enzyme. “Fentanyl is one option, as it’s metabolized by 3A4 and 3A5. Methadone is metabolized by multiple pathways, so that’s another option.”

Other medications a patient is taking can also inhibit CYP2D6, causing the patient to function as a poor metabolizer. For that and other reasons, pain management physicians or those prescribing pain medications should always ask the patient for a complete list of the medications he or she is taking, including over-the-counter medications and herbals, Dr. Jannetto says.

In some cases, therapeutic drug monitoring and pharmacogenomics can provide complementary information to help the physician get to the bottom of what’s going on with a patient. In her AACC presentation, Dr. Bratanow recounted how a patient with chronic pain after multiple spine surgeries required high doses of oxycodone to achieve desired pain relief. TDM showed that the man’s blood levels were in the normal therapeutic range. And genotype testing explained why: He was an ultra-rapid metabolizer of CYP2D6, which metabolizes oxycodone.

TDM and PGx testing combined not only optimize pain management in some cases but also increase safety for patients and for clinicians medico­legally when prescribing opioids.

Dr. Bratanow recalls one case of a patient followed for years at the Medical College and pain center, who was receiving methadone for the treatment of his chronic pain stemming from a gunshot wound that had left him a paraplegic. The man had been tried on many pain medications and finally found pain relief with metha­done, which he had been on at the same dose for five years.

During a summer heat wave, the patient was found dead near his bed. And the medical examiner’s toxicology report found the patient had high levels of methadone, potentially raising questions about the contribution of this drug in the cause of his death.

But the clinic had done its homework in managing the patient’s pain over the years. TDM results had shown that the patient had consistently high levels of methadone in his blood. And genotype testing, which showed he was an impaired meta­bolizer of CYP3A4 and CYP3A5 enzymes, accounted for those chronic high levels and the man’s inability to metabolize methadone well, Dr. Bratanow says. The patient was alert and functional for years with the blood levels shown on TDM, and had no side effects. “And he was comfortable at those levels. If we tried to decrease the dose, his pain would really flare, and he did not want to be on a higher dose than what he was on routinely. He worried about bothering us and asking for more pain medication.”

“In the end his death was thought to be due to a problem with his tricyclic medications, and his inability to regulate his body temperature,” Dr. Bratanow reports. The patient had been taking the tricyclic long-term. “There had been other deaths that the medical examiner was aware of in the city, in the past, related to the heat regulation problem.”

Of course, most pain management patients never end up a case for the medical examiner. Dr. Bratanow points to two studies by the Milwaukee Medical Examiner’s group. One found a higher prevalence of poor and intermediate metabolizers in oxycodone-related deaths.

Another postmortem study tied variations in fentanyl concentrations and the metabolic ratio of fentanyl to norfentanyl to people’s genotype, she says. “There is an ongoing multicenter study looking at this issue with regard to methadone deaths.”

Could pharmacogenomic testing one day help clinicians who treat pain identify patients at risk for drug abuse and addiction? “There may be such a test in the future,” predicts Mayo’s Dr. Langman, who points out that there are genetic polymorphisms associated with increased risk of addiction.

Dr. Bratanow says research suggests people who have a problem with addiction may have altered receptors and drug binding. “And there may be a difference in the way people at risk for addiction metabolize opioids.” For example, she has heard patients with addiction issues say that doses of opioids give them a feeling of bliss that nonaddicted patients don’t report experiencing.

“There is a feeling that approximately 10 percent of the population is at risk for addiction in our society to alcohol or medication,” she says, “and it’d be so helpful to know which 10 percent of that population we are seeing ahead of time, and predict it.” Patients with a higher risk of addiction who develop pain can receive opioid medications for painful conditions such as cancer, but should do so in a controlled setting with careful monitoring, more frequent visits, fewer doses given at a time, followup drug screening, and other precautionary measures, Dr. Bratanow notes.

The bottom line in pain management, she says, is to treat pain safely and adequately so patients no longer have to do what some in her care have described: “expend a huge amount of energy trying to isolate the pain into a corner of their day or life ... like a lion-tamer having to keep the lion at a distance.”