Are you at risk for an adverse medication reaction?

Many patients express concern about being on long term medications. In my view, their concerns are well-founded.  At times the treatment can be worse than the disease.  According to CDC statistics 82 percent of adults are on one or more medications, and 29 percent are on five or more.  Polypharmacy (the use of many drugs together, or excessive medications) is a significant problem of the elderly, and of those with chronic illness.  These populations are at increased risk for drug-related adverse reactions.   How can a patient assure the safety of his or her medication? Many turn to alternative medicine, under the false impression that these substances are somehow safer than those that are brought to market by the pharmaceutical industry.  Others rely on medical professionals—doctors, nurses, and pharmacists-- to warn them about the possibility of drug interactions and toxicity. Electronic prescribing has improved drug safety by automating cross-checking and alerting prescribers when two drugs interact.  However, in my experience electronic systems can establish such low filters for reporting drug interactions that virtually every drug prescribed may cause an alert to pop up. Which of these interactions is clinically relevant? Often clinicians must use their best guess as to whether two or more drugs in combination will be safe for a particular patient.

The cytochrome P450 enzymatic system is involved in the metabolism of many drugs. Although there are more than 50 of these enzymes, only six of them are responsible for the metabolism of 90 percent of drugs. Many significant adverse drug events result from issues that involve this pathway. Ingested substances, whether it’s grapefruit juice, a cup of coffee, an herbal product, or a prescribed medication, can alter metabolism by inducing or inhibiting the activity of the P450 enzymes. Moreover, research has revealed that there is significant genetic variation in their activity from one individual to the next.  I’ve had many a patient tell me of his or her unique sensitivity to drugs.  These circumstantial observations may well be founded in science, and pharmacogenetics is an emerging field that describes the genetic variation in responses to medication while one patient may have particularly efficacious P450 enzymes, another may have P450s that are slower to operate. Two or more drugs that are metabolized by the same P450 may compete and one drug may reduce the metabolism of another, causing high levels of the “substrate” drug to accumulate, and potentially cause toxicity.  Other drugs may up-regulate the digestive enzyme and cause a drug to be cleared more rapidly than normal, reducing its concentration and therapeutic efficacy.

Take the popular cholesterol lowering medication simvastatin (Zocor).  Simvastatin is used to lower cholesterol and has been associated with important clinical outcomes in patients who are treated with it-- including a reduction in cardiovascular death. However, its use has also been linked with an adverse drug reaction--myopathy, or muscle damage.  At its extreme myopathy is known as “rhabdomyolysis,” a process that can lead to kidney failure and even death.  Rhabdomyolysis occurs at a rate of 4.4 cases per 100,000 patients exposed to a “statin”-type medication (also including atorvastatin, rosuvastatin, pravastatin). The risk of myopathy is dose related and recently the FDA has warned against using the 80 mg dose of simvastatin for treatment of elevated cholesterol. 

Simavastatin’s metabolism occurs in the liver with the P450 enzyme CYP3A4.  Numerous other medications affect the activity of this enzyme.  The calcium channel blocker amlodipine (Norvasc) is processed by the same enzyme. Patients who take amlodipine and simvastatin simultaneously may have reduced clearance of simvastatin, and may be more prone to muscle damage from the drug.  Consequently the FDA advises limiting simvastatin dosing in this population to the 20 mg dose. However, enzymatic activity of CYP3A4 is genetically determined.  Within the population certain individuals may be rapid or poor metabolizers of the drug, impacting the generalizability of the FDA recommendations from one person to the next.

Genetic testing for cytochrome P450 enzyme polymorphisms is not yet recommended. Yet, we are moving in that direction, and no doubt the genetic polymorphisms may prove to provide valuable insight into why particular patients may not respond to standard treatments.  For example the drug Plavix (clopidogrel) is an important blood thinner that effects platelet activity and is indicated in patients who have had a stroke, or who have had stents placed for coronary artery disease.  Clopidogrel is a pro-drug-- it must be converted in the liver to its active form and CYP2C19 is the predominate enzyme responsible for this conversion. Patients who are poor metabolizers of Plavix do not effectively convert the drug to its active form. In these patients, the drug is less effective at preventing heart attacks, strokes, and cardiovascular death.  It is estimated that 2 to 14% of the population are poor metabolizers of Plavix; the rate varies based on racial background.  With this finding, some have advocated genetic testing of all patients who need Plavix for its important indication.

Another P450 issue has emerged with Clopidogrel.  The popular, and now over the counter, proton pump inhibitor (PPI) omeprazole is metabolized by the same hepatic enzyme and is an inhibitor of the enzyme, blocking the conversion of clopidogrel to its active form.  However, not all PPIs have the same degree of inhibitory effect on the enzyme (CYP 2C19). The drug pantoprozole (Protonix) may be a less strong inhibitors, and therefore safer for concomitant use with Plavix.

These two examples demonstrate the complex determinants of drug metabolism—genetic and environmental—and highlight the importance of using individualized treatment plans in order to optimize therapy and reduce the risk of medication related toxicity. 

*The FDA website offers a drug Index of Postmarket Safety Information for Patients and Providers. I found the website useful for specific drug information.

Should Doctors Stop Using PSA to Screen for Prostate Cancer?

The United States Preventive Services Task Force (USPSTF) says yes. 

Recently the influential Task Force gave the PSA (prostate specific antigen) screening test a “D” level rating, meaning that doctors should recommend against using the test for the purpose of early detection.  Prostate cancer is the second leading cause of cancer death in men after lung cancer.  Early detection of prostate cancer through screening with PSA  was introduced into clinical practice in 1986.  Over the past several decades prostate cancer mortality has fallen by approximately 30%. Statistical modeling suggests that this might correlate with the advent of screening with PSA.  In addition, over the past several decades there has been an increase in prostate cancer incidence and a shift toward the detection of earlier stage cancers—likely also a result of PSA screening.

The purpose of health screening is to detect disease in an earlier stage so that preventive measures can be initiated to improve health outcomes related to that condition.  However, in order to justify population level screening, the benefit associated with early detection must outweigh the harm that it potentially causes.  Healthy Americans and their physicians have embraced the practice of early detection, particularly with respect to cancer, which along with Alzheimer’s disease, is the most feared illness in our population.  The USPSTF analyzed available data on prostate cancer screening and treatment, including the results of the two largest clinical trials—one done in the United States and the other in Europe.  The American study (PLCO)  found no mortality benefit with PSA screening after a 7 to 10 year period of follow up.  Its results have been criticized with the possibility that findings were “contaminated” with too many men in the control group who received screening before and during the study, reducing the apparent efficacy of PSA screening.   In contrast, the European ERSPC study showed a statistically significant, but small reduction in prostate cancer mortality (20%), in men who were screened every 4 years over 9 years. Researchers found that 1,410 men would have to be screened and 48 additional cancers would have to be detected to prevent one death from prostate cancer.  This efficacy is similar in magnitude (slightly greater) to the effectiveness of using mammography to screen women in their forties for breast cancer.  A potential shortcoming of these studies is that the study time could have been too short to detect benefit given that prostate cancer may be very indolent. 

Needless to say urologists as a professional group object to the USPSTF’s recommendation, stating on the American Urologic Association website: 

The Task Force is doing men a great disservice by disparaging what is now the only widely available test for prostate cancer, a potentially devastating disease. We hold true to our current position as supported by the AUA's Prostate-Specific Antigen Best Practice Statement that, when interpreted appropriately, the PSA test provides important information in the diagnosis, pre-treatment staging or risk assessment and monitoring of prostate cancer patients. But not all prostate cancers are life-threatening. The decision to proceed to active treatment or use surveillance for a patient's prostate cancer is one that men should discuss in detail with their urologists.

The argument against using PSA to screen, itself an inexpensive blood test, is that it leads to over-diagnosis of clinically insignificant prostate cancer.  Postmortem data indicate that approximately one third of men between the ages of 50 and 65 have microscopic evidence of prostate cancer. This percentage increases to over 50 percent for men in their 70s and 80s. Analysis of data suggests that early detection leads to over-zealous treatment—including radiation and radical prostatectomy.  Over-treatment puts men in harm’s way with side effects related to treatment --loss of continence (ability to control urination) and erectile dysfunction.  I recently read a nice discussion the magnitude of risk associated with screening by Dr. Albert Fuchs. Both physicians and patients are participants in opting for aggressive interventions for early stage cancer.

Active surveillance or “watchful waiting” is an acceptable strategy for managing elevated PSA or even biopsy proven prostate cancer.  In practical terms this means that once cancer is detected it may be reasonable to do nothing aside from monitoring for evidence of prostate cancer progression. In some, this might entail monitoring the PSA velocity (or rate of increase) to help determine an appropriate time to treat.  Or, in others, it might include waiting to treat until the cancer becomes symptomatic with urinary symptoms or symptoms related to prostate cancer that has spread to the bone.  However it’s very difficult for patients and their doctors to sit tight with the knowledge that they might be living with cancer.  I’ve discussed the difficulty with watchful waiting with respect to breast cancer in the blog: “Is Watchful Waiting too Difficult?” 

As for me, I tend to find the USPSTF’s PSA screening recommendation a little too reactionary.  As stated in a recent New England Journal of Medicine editorial a “C” level rating for this test might have been more appropriate.   I am not completely ready to jump onto the anti-PSA bandwagon that many of my fellow internists seem to be espousing and go back to the 1970s.  However, I do see that a more cautionary approach to prostate cancer detection and treatment is warranted.  This will require discussion, courage, and culture change on the part of primary care physicians, urologists, and their patients alike as we try to effectively convey and accept the message that early detection of cancer is good, but that not all early cancer, or “pre-cancer,” needs early and aggressive intervention.

Screening for Inflammation with hsCRP

For several years now I’ve been screening many of my patients for inflammation with their annual physical examination using blood test known as high sensitivity c-reactive protein or hsCRP.  HsCRP is an inflammatory marker that has proven useful as a marker for cardiovascular risk in some individuals.  Inflammation occurs when there is tissue damage.  In general, ongoing inflammation is not good for one’s health.  In the case of arthritis, inflammation affects bones and joints. In the case of infection, inflammation results when the immune system responds to a pathogen.  In the case of cancer, inflammation occurs as cancerous cells invade healthy tissues and cause damage.  More recently, inflammation has been identified as an important factor in atherosclerosis, the process that leads to cholesterol plaque accumulation in blood vessels resulting in heart attack, stroke, and peripheral vascular disease. 

CRP is an acute phase reactant, so its level may go up with infection or trauma.  However, in general, hsCRP levels tend to be relatively stable over time, compared with other inflammatory markers.   Multiple studies have demonstrated the relationship between hsCRP elevation and cardiovascular disease.  In fact, with respect to cardiovascular risk, hsCRP is said to be more predictive of cardiovascular events than LDL cholesterol levels.  Three levels of risk have been identified:

Low risk                                                hsCRP < 1 mg/L

Moderate risk                                     hsCRP 1-3 mg/L

High risk                                               hsCRP  >3 mg/L

In the Jupiter trial healthy men and women with normal LDL cholesterol (<130) but elevated hsCRP (>2 mg/L) were randomized to receive 20 mg of Rosuvastatin or placebo. The trial was halted early when the treatment group was found to have significantly lower risk of cardiovascular events in the 1.9 years that the subjects were studied.  The reduction in risk correlated with a reduction in LDL cholesterol and hsCRP levels.

In my patient population it is my experience that about 25 percent of my patients have hsCRP levels that exceed the 3 mg/L threshold for “high risk.” About 5 to 10% of my screened patients have levels that substantially exceed 3mg/L.  As a generalist, I have been tasked to take action with these particular patients, bringing them back in to the office for a thorough history to exclude occult infection, ordering additional tests to screen for occult rheumatologic disorders, and to make sure cancer prevention guidelines have been followed—and at times doing additional work-ups.

Elevated CRP has also been associated with diabetes and metabolic syndrome.   One patient in her 50s had an hsCRP of 28. This patient also had new onset diabetes, with a hemoglobin A1C of 8.1, LDL cholesterol of 136, and BMI of 42. After losing 70 pounds (over one year) and with resolution of her diabetes my patient’s hsCRP came down to 3.  A statin was started in addition to aspirin therapy.   In this case the crp did not alter my practice, though it did raise my level of concern.

Another healthy patient in her forties had a level of 3.5 mg/L.  The patient, who is vegan, had an LDL of 80, an HDL of 78, a normal glucose, does not smoke, and has a body mass index of 23.  Framingham risk was calculated at less than 1 percent. My patient had astutely read of an association of between elevated CRP and Alzheimer’s Disease risk (which has been described).  Unfortunately, there is no clear and proven intervention to reduce this patient’s potential health risk, which is still likely low. I placed her on aspirin 81 mg daily.

Within the realm of using hsCRP  for the purpose of primary prevention medical knowledge is based primarily on cardiovascular trials and outcomes.  According to one expert author CRP may be at least 50% genetically pre-determined.  In the case of my healthy patient in her 40’s this seems likely. I was reassured to read a summary in Circulation noting that while a relationship between high crp and cardiovascular death has been demonstrated, elevated hsCRP has not been linked to increased mortality from other causes (like cancer). At least this finding will allow me to focus on cardiovascular health when hsCRP is high, as opposed to engaging in a wild goose chase to detect occult illness.

HsCRP measurement is currently recommended only in individuals who are at intermediate risk for cardiovascular disease (defined as 10 to 20 percent chance of having a cardiovascular event in ten years). In this population it can prompt more aggressive management of risk factors, including beginning a statin for only marginally elevated cholesterol.  One review in the Annals of Internal Medicine noted that while high hsCRP levels in women with intermediate or high Framingham risk correlated with worse cardiovascular outcomes, high levels in women deemed to be at low risk by Framingham did not correlate with substantially high vascular risk.  Despite this finding, at times I have still found it helpful to check hsCRP within a low risk population. The test is inexpensive, it can help as a motivator to prompt lifestyle change that could prevent future increased risk, and 20 percent of heart disease may occur in those with no traditional risk factors.  As a novel risk factor hsCRP has become one of many variables known to contribute to cardiovascular health.  However, as an isolated finding it still may have limited utility, raising questions that at this time still have no clear cut answers.