By Jack D. McCue, MD, Clinical Professor of Medicine, University of California, San Francisco
SAN DIEGO—There were 16 update sessions at the American College of Physicians annual meeting that covered each of the various medicine subspecialties, plus some of the generalist disciplines. These were supplemented by many other presentations, as well as 9 sessions aimed at controversial topics ("multiple small feedings for the mind"). I present an "update of updates" of sorts, resulting from sampling the various presentations of interest to physicians who take care of the increasing number of elderly patients.
This year I concentrated on new opportunities not to treat. Polypharmacy has become an unwelcome fact of life for geriatricians and their elderly patients, whether young-old or old-old. Despite decades of arguing for simplicity in drug regimens, it has become clear to me that, at least for the treatment and prevention of cardiovascular (CV) diseases, lower doses of multiple medications can offer advantages of greater efficacy and reduced adverse drug reactions (ADRs) compared with larger doses of single medications. Still, polypharmacy is more expensive (2 or 3 drugs are usually more expensive than a higher dose of 1 drug), adds new sets of ADRs to the regimen, and increases the risk of drug interactions. Hence, as the number of drugs our elderly patients must take increases, a search for drugs of marginal or no benefit that can be discontinued has become even more critically important.
I present a stereotypic case—one that will be familiar to all primary care physicians. You have assumed care for an 89-year-old patient whose attending physician has decided to retire. She is taking 12 prescription drugs, plus a variety of over-the-counter (OTC) medications and nutritional supplements; the monthly cost is nearly $1000. She has known CV disease, with a history of a coronary artery bypass and atrial fibrillation; cerebrovascular disease, with dizziness; and multifactorial mildly-to-moderately severe dementia, with occasional spells of confusion. She has a thyroid-stimulating hormone (TSH) level of 4.9 mIU/L. You conservatively estimate her life expectancy at 2 to 3 years. Her regimen includes clopidogrel (Plavix), aspirin, alendronate (Fosamax), oxybutynin (Ditropan), and olanzapine (Zyprexa); she also takes OTC dehydroepiandrosterone (DHEA) and glucosamine/chondroitin sulfate. Surely some of her dozen prescription medications can be stopped, you ask; or should you throw caution to the wind and add thyroxine, drug #13, hoping to improve her forgetfulness?
No fewer than 6 of the update/controversies sessions discussed new studies of clopidogrel. All the efficacy studies of clopidogrel have been compromised by the presumption that a placebo-controlled trial would be unethical, true or not. Consequently, the trials have been either comparisons of clopidogrel combined with aspirin versus aspirin, or clopidogrel alone versus aspirin.1
Plavix has been the subject of an unseemly patent battle that involved huge bribes and embarrassing litigation. Let's face it, this is a drug that only a pharmaceutical company or an interventional cardiologist could trust. It is nominally superior to historic, cheap aspirin in preventing ischemic vascular events in high-risk patients—a difference that is of doubtful clinical significance; and the combination of clopidogrel and aspirin is marginally beneficial to aspirin even in acutely ill patients with unstable angina or myocardial infarction. Long-term therapy studies have not been encouraging, moreover. Bhatt et al2 give courage to those who have concluded that clopidogrel in patients with stable CV or neurovascular disease is of little or no benefit; and they concluded that clopidogrel may be harmful in long-term maintenance regimens.
Unless your patient has had an acute CV event in the past year, you are on solid ground in stopping clopidogrel in your elderly patients. Even if you are tempted to continue it despite unconvincing evidence, you might consider the wisdom of Karen C. Johnston, MD, Department of Neurology, University of Virginia Health Sciences Center, Charlottesville, who said that an aspirin a day is much better than an occasional clopidogrel in patients who are reluctant to spend the nearly $4/day for it.
A related study discussed in several sessions at ACP was a comparison of clopidogrel plus aspirin with warfarin prophylaxis for patients with atrial fibrillation.3 The trial had to be stopped early because of relative lack of efficacy and increased risk of bleeding with the combination regimen. My conviction has always been that warfarin is the only drug that really works in atrial fibrillation—but if you can't use warfarin, it is better to use aspirin alone than to increase the cost and risk of therapy by adding clopidogrel.
: I believe that you should have a very good reason for continuing clopidogrel in this patient. Data supporting its use in anything other than acute cardiac syndromes are, at best, weak.
There is no doubt about the benefits of bisphosphonates, most often used as weekly alendronate, in the prevention of vertebral and hip fractures in elderly women with osteoporosis. Furthermore, there is growing doubt about the efficacy of the near-universal use of calcium and vitamin D supplementation. Although probably few geriatricians will be willing to forgo nutritional supplementation until much more convincing data appear, it would be nice to know when it is reasonable to discontinue alendronate—which does cause its share of ADRs and costs about $25 for the single weekly dose.
In one study, 1099 women (average age, 73 years at baseline) were randomized either to take alendronate for 10 years, or to discontinue it after 5 years and take placebo for the remaining 5 years.4 The rate of clinical vertebral fractures did indeed increase (5.3% vs 2.4%) and bone density in hip and spine decreased (3.7% vs 2.4%), respectively, in the placebo-for-5-years group versus the active-therapy-for-10years group. However, hip fractures were identical in the 2 groups.
The good news is that continuing alendronate did work, and the additional good news is that the consequences of discontinuation were relatively minor and clinically insignificant when the greatest cause of mortality and morbidity in women with osteoporosis—hip fractures—was considered. These women were relatively younger than our patient above, but remember that our patient has a life expectancy of <5 years; this means that over a 5-year period, she is unlikely to be harmed by discontinuation of alendronate.
: Our patient, as well as nearly all our nursing home patients (whose average life expectancy is considerably less than 5 years) can safely discontinue their bisphosphonates. Continuation of vitamin D and calcium supplementation is probably still prudent, although data to support efficacy in this population are not very strong.5 In addition, calcium/vitamin D supplementation does carry some risk of ADRs and causes malabsorption (eg, when given with our probably overused fluoroquinolones).
The geriatrics update session by David Reuben, MD, professor of medicine at the University of California, Los Angeles, was filled with provocative, good ideas about drugs to stop. In fact, it seems that quite a few of the medications with which we treat our elderly patients don't really do very much. In fairness to physicians, half of the newly found useless medications he discussed are OTC, and usually our patients don't even tell us when they are taking them.
DHEA and testosterone levels decline with age, and supplements have been encouraged as an antiaging regimen. Replacement doses do indeed normalize DHEA and testosterone hormone levels, but several studies6,7 found that the hoped-for changes in body composition, physical performance, bone density, and glucose intolerance did not occur. Probably the data are not all in, but for the present, unless an older man is clearly testosterone deficient,8 there is little to support the use of testosterone supplementation. No data are available to support use of DHEA or testosterone, either preventively or to treat changes related to normal aging.
Wading deeper into controversy, the latest study of glucosamine and/or chondroitin sulfate or celecoxib (Celebrex) found they were no better than placebo for painful knee osteoarthritis.9 And perhaps yet deeper, saw palmetto (great name!) does not work for symptoms or signs of benign prostatic hypertrophy.10
Atypical antipsychotic drugs, such as olanzapine, risperidone (Risperdal), or quetiapine (Seroquel), which have earned "black box warnings" because of dangerous ADRs and generally have a poor tolerability profile, were found to have no more efficacy than placebo.11 I agree with Dr Reuben, however, that this latter study should not be taken at face value. Patients with dementia plus psychotic symptoms who can tolerate these drugs may benefit from them. I believe that they are potentially valuable drugs whose use requires more "art" than can be captured in a carefully controlled trial.
Finally, the combination of tolterodine (Detrol) (or oxybutynin, probably) and tamsulosin (Flomax), which many of us have used together empirically for our older men with overactive bladder symptoms, does seem to work better (80% improvement) than either drug alone (65% tolterodine; 71% tamsulosin)—although it has to be pointed out that, as in nearly all these urologic drug studies, placebo works quite well too (62% improvement),12 which does not necessarily match up with a paper by the same group in a different journal.13 Personally, I think this is a very suspect literature that has been overly influenced by pharmaceutical firms. This study once again raises the question of whether the anticholinergic drugs are more effective than placebo for overactive bladder when used in tolerable doses. Generally speaking, unless you raise the dose until unpleasant side effects appear, these drugs do not seem to do much.
: Unless a patient convinces me otherwise, I think the "bladder" anticholinergics in our elderly patients should be viewed very skeptically. Just because your elderly patient has been on a drug for years—even for a well-defined clinical indication—do not assume that continuation of that drug is beneficial. Opportunities for fighting polypharmacy abound.
The temptation is irresistible, isn't it? We were taught during our geriatrics training that hypothyroidism is protean in its presentations, and it can cause nearly anything that old people might complain of. Similarly, we have been encouraged to believe that giving small, probably innocuous doses of thyroid hormone to patients with a slightly elevated TSH may help everything from dementia to failure to thrive or pressure ulcers.
There have been no data to support the potential benefit of treating subclinical hypothyroidism, however, which is said to be present in asymptomatic patients with a TSH near the upper range of normal, 5 mIU/L. However, in the real world of clinical medicine, we know that elderly patients are never asymptomatic—there is always some sign or symptom that might be attributable to hypothyroidism, from dry skin to tiredness; so if the TSH obliges by being in the high-normal range, or maybe just a little bit over the upper limit of normal, it is reasonable to think that a therapeutic trial of thyroxine may make a difference. Similarly, we often are tempted to increase the thyroxine dose when routine TSH monitoring shows a drift into the high-normal range or into the low-abnormal 5- to 10-mIU/L range—even though patients are not manifesting symptoms of hypothyroidism (see also, Page 15).
Walsh et al (the folks who also brought us the disappointing news that combined thyroxine/triiodothyronine supplementation did not make our patients feel any better than good old thyroxine14) found that subclinical hypothyroidism and hyperthyroidism in patients taking thyroid replacement are indeed subclinical. Adjusting the dose of thyroid hormone did not help patients with no real symptoms of hypothyroidism or hyperthyroidism feel better, even if their TSH was near the high or low levels for the normal range.15 Their study adds to the controversy about whether the normal range for normal TSH should be left alone at
0.5 to 5.0 mIU/L, or should be narrowed. In conclusion, patients who are clinically stable on a dose of thyroid replacement and who have a TSH that drifts to the limits of, or even a little out of, the normal range do not seem to benefit from fine adjustments in dosing. Of course, this conclusion also raises the secondary question of why we should even test TSH on a periodic basis if patients are asymptomatic.
In a separate, supporting study, Roberts et al16 found that after the confounding effects of comorbid conditions and use of medication were controlled for, subclinical thyroid dysfunction was not associated with depression, anxiety, or cognition.
: Until there are more convincing data, the wisdom of treating more-or-less asymptomatic patients whose TSH is mildly abnormal is questionable. And talk about an immortal drug (drugs that persist at least as long as you do)—when was the last time you had the courage (or time) to stop thyroxine?