Drugged and Docile?

    This weekend, the New York Times uncovered new, seemingly damaging information about nursing homes, this time unrelated to their mishandling of the Covid-19 epidemic. “Phony Diagnoses Hide High Rates of Drugging at Nursing Homes,” is a detailed investigative essay by three reporters that reveals that nursing  homes regularly under-report the frequency with which they prescribe sedating, antipsychotic medications for residents with dementia. Such medications, while useful for controlling  paranoia and delusions (which may afflict people with dementia), have not been shown to be more generally helpful in controlling the behavioral symptoms of dementia. They are, however, associated with a two-fold increase in mortality and other adverse effects. As a result, nursing homes have been under pressure for years to limit their use of “chemical restraints,” medications that suppress agitation--including general sedatives and antipsychotics. 

            The campaign against the use of anti-psychotics in dementia began with the Nursing Home Reform Act of 1987 (OBRA-8)7, legislation asserting that residents have the right to be free from physical and chemical restraints that are “not required to treat specific medical symptoms.” Then, nearly a decade ago, CMS announced a new approach, the National Partnership to Improve Dementia Care and Reduce Unnecessary Antipsychotic Use in Nursing Homes. This strategy provided for training modules for nursing home staff on how to handle residents with dementia; the strategy also involved including as a “quality indicator” the proportion of long-stay nursing home residents receiving an antipsychotic medication. In 2015, this quality indicator was added to the list of measures that comprise the overall rating of nursing homes that CMS publishes on its website, Nursing Home Compare.

            The last 34 years have seen a marked decline in the use of antipsychotic medications in nursing home residents. The Nursing Home Reform Act of 1987 led to a 27 percent reduction in antipsychotic use. The introduction of psychoactive drug use as a quality indicator led to a further ten percent fall in prescribing. Today, CMS statistics assert that 15 percent of long-stay nursing home residents regularly receive antipsychotics for some problem other than Tourette’s syndrome, schizophrenia, or Huntington’s Chorea, for which antipsychotics are approved. The new report by the NY Times suggests that the correct figure is more like 21 percent, with the excess accounted for by bogus diagnoses of schizophrenia: the implication is that doctors want to control their demented patients by sedating them, but are discouraged from doing so by CMS regulations, so they get around the rules by falsely labeling their patients as schizophrenic.  

            The Times argues that “caring for dementia patients is time- and labor-intensive. Workers need to be trained to handle challenging behaviors like wandering and aggression. But many nursing homes are chronically understaffed and do not pay enough to retain employees, especially the nursing assistants who provide the bulk of residents’ daily care.” All true. The Times goes on to argue that nursing homes with poor staffing ratios (facilities that get a 1 or 2 star rating for the adequacy of staff: patient ratios) dole out more antipsychotic medications than those with better staffing ratios (facilities with 4 or 5 star ratings for staffing). 

While this graph strongly suggests an inverse relationship between staffing and antipsychotic use, what is equally striking is that all the facilities, regardless of staffing, administer antipsychotics to upwards of 15 percent of residents, without appropriate justification. What this suggests, contrary to the NYT implication, is that nursing homes find that no amount of training and no  number of staff members can consistently and reliably treat all the behavioral symptoms of dementia.

            The unfortunate reality is that we know very little about how to care for people with moderate to advanced dementia who exhibit problematic behaviors. The behaviors I am referring to are not merely inconvenient to staff, such as “wandering” off the nursing unit unattended. They include aggression, paranoia, and delusions. That means hitting or spitting; it means slugging or biting the well-intentioned caregiver who sought to bathe or feed a resident. These behaviors are disturbing to persons with dementia, their caregivers and, in a nursing home setting, other residents. 

            The Alzheimer’s Association, the American Psychiatric Association, and others have come up with guidelines for managing these symptoms. Their recommendations advise beginning with psychosocial interventions: first tryi to determine the cause of the behavior (perhaps the person slugged the aide who tried to give him a shower because the water was too cold) and then either address the root of the problem (for example, adjusting the temperature of the shower water) or engage in distraction (taking the individual to a “quiet room” with a box of trinkets and other treasures to examine). But then comes the caveat: “Unfortunately, large population-based trials rigorously supporting the evidence of benefit for non—pharmacological therapies are presently lacking.” The evidence for these approaches is largely anecdotal and commonsensical--as is true for anti-psychotic medication.

            What is the solution? The Times seems to suggest we should either penalize physician prescribers, fine nursing homes that over-prescribe, add further regulations, or all three. Another alternative is to substitute “Green Houses” for conventional nursing homes. The Green House is a model of nursing home care that seems to be successful in many domains where traditional nursing homes have failed abysmally, so there is some reason to believe it may be a good way to deal with dementia. The Green House project restricts homes to no more than 12 residents, employs a home-like focus, and cross-trains staff members to meet any and all resident needs rather than using a rigidly siloed model. Green House nursing homes may be the answer: they are associated with high rates of family satisfaction as well as a superior track record in controlling the Covid epidemic. I hope Green Houses have the solution, but I can find no published data on their ability to manage severe dementia. In fact, it’s far from clear Green Houses, which currently provide care to a total of 3000 people (out of a total nursing home population of 1.3 million), in 300 facilities (out of a total of 15,600 nursing homes nation-wide), actually care for people who have dementia and troublesome behavior. Their limited numbers, small size, and high cost (45 percent of residents pay out of pocket compared to 22 percent of residents in usual nursing homes) suggest they may be able to cherrypick their residents.

            I am not advocating blanket use of psychotropic medications in nursing home residents. Nor am I endorsing mislabeling patients with psychiatric diagnoses they do not have. Antipsychotics should only be used in people with dementia in well-defined and relatively rare circumstances. Perhaps informed consent by the patient's health care proxy should be a prerequisite. But blaming nursing homes or pointing the finger at nursing home doctors or devising new regulations are not the best ways to serve people afflicted with dementia. What we need is to find better ways to alleviate the suffering of people with dementia. We should recognize that the doctors who prescribe antipsychotics are not necessarily lazy or devious, though any who are should be disciplined; many of them are simply desperate, desperate to provide relief to their patients. Medicare is poised to spend billions of dollars on aducanumab, a drug recently approved by the FDA for treatment of Alzheimer’s disease despite the paucity of evidence that it works and the abundance of evidence of its association with severe side effects. Perhaps we should instead devote resources to what is likely to be the more tractable problem, the symptomatic relief of Alzheimer’s disease and other dementias.

 

 

 

 

 

  

All Eyes are on Medicare

  

            When the FDA approved Biogen’s new Alzheimer’s drug, aducanumab (brand name Aduhelm) on June 7, the reaction was surprise, dismay and, in some quarters, enthusiasm. But everyone was shocked by the drug company’s audacity in setting the price for the medication at $56,000 per year. As one STAT article put it, the only question about the consequences for Medicare, the insurer for close to 97 percent of Alzheimer’s sufferers, was whether the impact would be big, huge, or catastrophic. 

Pharmaceutical manufacturers figured out some time ago that they could bring out “specialty drugs,” typically targeted against a single relatively rare disease, if they charged ten or even a hundred times more than for an average medication. The list price of crizotinib (brand name Xalkori), for example, used against a relatively uncommon type of lung cancer found in non-smokers, is just under $20,000 for a one-month supply—and patients usually take the drug until they die or develop resistance to it. But aducanumab is intended for all people with Alzheimer’s disease, and according to recent Alzheimer Association estimates, that means 6.2 million people over age 65.

            The high price is especially disturbing because it’s not even clear that the drug works. A number of studies have been carried out with similar drugs, other monoclonal antibodies that, like aducanumab, were designed to mop up abnormal brain amyloid deposits, which are the hallmark of Alzheimer’s disease—but none of those drugs proved helpful in practice. The trials of aducanumab were likewise discontinued because interim analysis showed the drug was ineffective. Then, in a surprise move, the manufacturer nonetheless applied to the FDA for approval after a reanalysis of the data showed some evidence of benefit when the drug is given in high doses. The independent scientific review panel convened by the FDA to evaluate the data was not convinced, however, with 10 out of 11 members rejecting approval and one abstaining—but the FDA nonetheless approved the drug.

            Even if aducanumab does work, “work” means slowing the rate of decline slightly, not stopping or reversing the disease process. And the potential side effects of the drug are considerable: 40 percent of patients experienced brain swelling, in some cases of sufficient magnitude to cause nausea, vomiting, confusion, or visual changes. 

            Patients and their families, who are desperate for an effective drug against this progressive, ultimately fatal disease, are eager to try something with promise, anything. But they are worried about the side effects of aducanamab, about the need for regular MRI scans to monitor for those effects, and about its high cost.

            Most critiques of the new drug—and there are many, the NY Times alone has published eight articles on the subject between June 7, when the FDA announced approval of the drug, and July 9 , and STAT has published at least 16—assume that since it has been approved by the FDA it will necessarily be paid for by health insurers. In the case of aducanumab, that will principally be Medicare. In fact, CMS is not obligated to provide coverage for the drug just because the FDA approved it.

            Medicare, like all other health insurance companies, can decide what tests, procedures, and treatments it will cover. The relevant part of Medicare that will be responsible for paying for aducanamab, if Medicare covers the drug, will be Part B: oral medications fall within the jurisdiction of Medicare Part D plans (prescription drug plans), but medications administered intravenously in a physician’s office, such as aducanumab, fall under Medicare Part B. Most determinations of whether to provide coverage for this kind of treatment are made locally, by the private carriers that process Medicare claims. But the decision about coverage can be made nationally if requested by CMS, by the manufacturer, or by members of the medical profession, in which case the decision becomes binding on all the private carriers. Such “National Coverage Decisions” are reviewed by an internal arm of CMS, the Special Coverage and Analysis Group. For particularly controversial decisions, especially if they have social or ethical implications, CMS may request the guidance of the quasi-independent committee, MEDCAC, the Medical Evidence Development and Coverage Advisory Committee. This is a group of 100 experts including economists, ethicists, physicians, scientists and others, from whom a subgroup of 15 is selected to provide in-depth analysis on the particular test or treatment under consideration. 

            Since its inception, MEDCAC (or its predecessor, the Medicare Coverage Advisory Committee), has issued 348 National Coverage Decisions. These comprehensive analyses have addressed topics as diverse as cardiac pacemakers, Pap smears, and lipid testing. On rare occasions, they have dealt with drugs, for example, an intravenous medicine used in the treatment of heart failure, Nesitiride. When MEDCAC deliberates Medicare coverage for a particular intervention, it can recommend covering the intervention, not covering it, or restricting its use in specific ways. For instance, it advocated coverage of the Left Ventricular Assist Device, an invasive treatment that is almost but not quite an artificial heart, but it required a detailed informed consent process that included a social worker and palliative care expert along with the patient, family, and cardiac surgeon. Ultimately, Medicare approved coverage for the device but set reimbursement at $70,000 (the manufacturer’s price was closer to $200,000) and limited insertion of the device to a handful of medical centers across the country. 

            Medicare is required by federal law to provide coverage for anything that is “reasonable and necessary” for “the diagnosis or treatment of an illness or injury.” Despite years of often contentious debate, there is no precise definition of what this means. The FDA, by contrast, approves drugs and devices if they are “safe and effective.” In the case of aducanumab, it is arguable whether the drug is truly safe and effective, but surely it would be reasonable and necessary for Medicare to restrict the use of aducanumab to early disease (the only group in whom it was tested) and to require an elaborate informed consent process. While Medicare, by established custom, does not reject coverage based on cost-benefit analysis, it could set the price at a level comparable to those of the existing, only marginally beneficial drug treatments for Alzheimer’s disease, drugs such as rivastigmine (brand name Exelon, which has a yearly retail cost, when given as the brand name drug, of $823) and donepezil (brand name Aricept, which has a yearly retail cost, when given as the brand name drug, of $5380).

            Given the controversy swelling around Biogen’s new Alzheimer’s drug, the case for Medicare initiating the National Coverage Decision process is strong. The only reason for failing to do so is external pressure, whether by the manufacturer, by members of Congress under the influence of the pharmaceutical industry, or by the public. If CMS opts against this path or convenes MEDCAC only to reject its advice,* as the FDA did with its advisory committee, that would be a compelling reason to make CMS an independent agency, along the lines of the National Science Foundation, that is under control of a bipartisan board and whose director is independent of the President.

 

            *Between when this essay was drafted on July 9 and edited for publication today, CMS has in fact decided to proceed with a National Coverage Decision.

Have We Been Barking Up the Wrong Tree?

More of my blog posts deal with dementia than with any other subject and the news about Alzheimer’s disease over the years has been largely dispiriting, so who would have thought that I would leap at the opportunity to write about a new diagnostic test. But with so much of the medical literature relentlessly focused on COVID-19, it’s reassuring to realize that research on other subjects is continuing. The new study does not report a treatment, let alone a cure for Alzheimer’s disease. Furthermore, the prospect of screening healthy individuals to determine their future risk of developing progressive cognitive impairment is ethically fraught. Nonetheless, in the current climate, this report is good news.

It’s good news, and not just because it indicates that not all medical scientists have retooled as corona virus researchers, though it does that. It’s good news, and not just because it means it will be possible to target intervention studies to high risk individuals will permit studies to be carried out on smaller numbers of people and over a shorter period of time, though it means that. It’s good news because it shines a bright light on a long-neglected character in the Alzheimer’s story, the tau protein. 

Back in 1906, when Alois Alzheimer peered into his microscope at tissue from the brain of a patient who had died of the disorder of cognition that would one day bear his name, he identified two unusual substances that he described as plaques and tangles. The plaques, which were located between neurons, would ultimately be found to be composed of a protein known as amyloid. The neurofibrillary tangles, which were located inside the nerve cell bodies, would eventually be identified as a protein called tau. These two substances have been recognized as the hallmarks of Alzheimer’s disease for over a century.

For years, the roles of amyloid and tau were hotly debated. Some researchers felt that amyloid was the result of Alzheimer’s; others were confident it was the cause. Some scientists were more interested in studying amyloid; others directed their efforts towards tau. But over the course of the last 25 years, amyloid has gained the upper hand. Study after study has sought to improve cognition in Alzheimer’s disease by ridding the brain of amyloid-laden plaques—and each time, the approach failed. 

A great deal of excitement was engendered by immunotherapy back in 2001: the idea was to stimulate the body to create antibodies against amyloid with what was essentially a vaccine—but the study had to be stopped because a subset of patients developed meningitis. Then there was enthusiasm about the use of monoclonal antibodies. Several such antibodies have made it to phase 3 trials in which their efficacy was compared to placebo. In 2014, two studies of Bapineuzumab showed no benefit. In 2018, Solznezumab was tried for individuals with mild Alzheimer’s and it was unsuccessful. In the same year, additional negative results were reported for Verubecestat in people with mild to moderate Alzheimer’s. 

All these negative studies don’t exonerate amyloid. Maybe the trials are initiated too late in the course of these disease’s development. Maybe the dose is too low. But with anti-amyloid strategies repeatedly striking out, I can’t help but wonder, as have others who know much more about the science than I do, that we’re looking at the wrong target.

Which is why the new study that focuses on tau is exciting. The authors found that their tau antibody test was able to diagnose Alzheimer’s disease as well as or better than more invasive existing tests—when they used the test in patients all of whom had some kind of neurodegenerative disease. That is, the test did well in answering the question: is this person being tested more likely to have Alzheimer’s or, say, Parkinson’s? That’s a very different question from: is this person normal or does he have Alzheimer’s? Not only was the population in which the test was studied composed exclusively of patients with some neurologic condition, not only did the population include a much larger proportion of people with Alzheimer’s than would be found in the general population, but the subjects were far from ethnically or racially diverse. So, it’s a long way from the article in JAMA to a widely useful diagnostic test.

Despite the test’s preliminary nature, it is a compelling piece of evidence that tau should get more attention. Two weeks before the on-line publication of the JAMA study, a small Swiss pharmaceutical company, AC Immune, announced that together with Johnson & Johnson, it was launching a trial of a vaccine designed to stimulate the body to produce antibodies against tau—leading its stock price to soar by 18.9 percent. Just a few weeks earlier, the giant Swiss pharmaceutical company, Roche, announced it, too, was investing in the development of an anti-tau vaccine. 

It’s too early to say whether the attack on tau will fizzle, much like the previous attacks on amyloid. But maybe, just maybe, it will be a rousing success.

Rescinding the DNR order--for anti-amyloid drugs

The pharmaceutical firm Biogen issued a stunning press release this week: it would seek FDA approval for its anti-Alzheimer’s drug, Aducanumab—the very same drug the company had pronounced a failure just last March, after preliminary analyses indicated the drug was very unlikely to achieve its objectives. Dennis Selkoe, a prominent Alzheimer’s researcher at the Brigham and Women’s Hospital in Boston, wrote in STAT (a health-oriented news website owned by the Boston Globe) that he believed that “aducanumab is the breakthrough we have been waiting for.” But is it?

Aducanumab is by no means the first drug studied that targets beta amyloid, a protein widely believed to play an important role in the development of Alzheimer’s disease. An article in Nature Reviews (Drug Discovery), published shortly after the first analysis of the Aducanumab data, put the drug in perspective: a monoclonal antibody that binds to beta amyloid, it is the fourth such drug to make it to a Stage III clinical trial. And all three of its predecessors, as well as another five anti-amyloid drugs that have a different mechanism of action, were abysmal failures leading Nature Reviews to proclaim "Anti-Amyloid Failures Stack Up as Alzheimer's Antibody Flops." Maybe what this new analysis of the Aducanumab data has done is to reverse the Do-Not-Resuscitate label previously attached to monoclonal antibodies and instead put them on life support. So, what, exactly, does the new data analysis show? 

The details will be released to the FDA when Biogen files with the FDA to approve the drug. Thus far what we know is what Biogen has included in its press release. Biogen conducted two randomized, double blind, controlled studies, giving the monoclonal antibody (presumably by injection, like other monoclonal antibodies) or placebo to a total of 3285 patients and following them, some for 78 weeks. One study, sporting the acronym ENGAGE, produced negative results even upon re-analysis, although a subset of ENGAGE patients who received a high dose of the drug appeared to respond. The other study, named EMERGE, showed a reduction in the rate of clinical decline in treated patients compared to those receiving placebo. In other words, the study patients did not improve nor did they plateau; rather, they got worse more slowly. Unfortunately, the press release doesn’t report the absolute magnitude of the change, opting instead to say that the scores on the Mini-Mental State Examination (MMSE), a commonly used measure of cognition, fell 15 percent less with treatment than with placebo. Similarly, the rating on the ADAS-Cog13, another measure of cognitive function, declined 27 percent less in those on the drug. Finally, and perhaps most encouragingly, the decline in the ability to perform basic daily tasks (as measured by the Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory) was 40 percent lower in those receiving Aducanumab than in their counterparts who received placebo. 

These are not spectacularly impressive results—the patients continued to worsen and the difference between treated individuals and untreated individuals, at least on measures such as the MMSE, was very small. But what we don’t know is whether the slowed rate of decline will persist over time or whether, as with drugs that are currently on the market such as donepezil (Aricept), there is a one-time effect. We don’t know whether the effect would be greater if the medication were used earlier in the progression of Alzheimer’s disease, perhaps in people with mild cognitive impairment. We don’t know whether the drug will prove to be more effective in those with “pure” Alzheimer’s disease, that is, without concurrent vascular changes. 

What we can predict is that if the drug is approved by the FDA, Biogen, whose stock price soared after the press release, will aggressively market the drug. An estimated 5.8 million Americans have Alzheimer’s disease, according to Alzheimer’s Association statistics released in 2019: since the time from diagnosis to death is 5-7 years, roughly one-fifth of the affected individuals or over a million people likely have early Alzheimer’s and are potential Aducanumab consumers. This could be a dream come true for the pharmaceutical company--a blockbuster specialty drug. (Blockbuster drugs are prescribed to at least one million people a year and specialty drugs are very expensive biologicals such as monoclonal antibodies). And if the sales price for the drug is anything like that of other newly marketed monoclonal antibodies, it may be upwards of $1000 per month (Dupilumab, brand name Dupixent, a new monoclonal antibody used in the treatment of asthma and/or nasal polyps, has a list price of $37,000 per year). 

Biogen stockholders will benefit enormously; how much people with Alzheimer’s disease will benefit is less clear. But on a more positive note, as Dennis Selkoe commented in STAT, for the first time we have evidence that it is possible both to design molecules that target amyloid and to administer them to real people with minimal toxicity and some evidence of benefit. That truly is good news.

The Dirt on Drugs

Several months ago, in writing about Katy Butler’s generally admirable new book, The Art of Dying Well,I questioned her uncritical acceptance of claims that certain drugs, the anticholinergics, caused dementia. That drugs such as antihistamines and some antidepressants cause delirium, or acute confusion, is well-established. But dementia? A large, well-conceived study just published in JAMA Internal Medicine provides additional evidence that they may well result in dementia. Previous studies were tainted because they tended to be small, observational studies that lumped many different anticholinergic drugs together, some of which are known to be far more potent than others. So how does the new study fare by comparison?

The recent analysis is still an observational study—it is impossible to randomize patients into those who receive anticholinergics and those who do not and then follow them for years—but it is very large, it analyzes different classes of anticholinergics separately, and it focuses on the cumulative anticholinergic exposure over a period of many years. While not perfect, it is probably the best we are going to get and it provides strong, although not definitive, evidence that anticholinergics are a risk factor for dementia.

This “nested case control study” allowed the authors to identify just under 59,000 people over age 55 with dementia from a British data base of 30 million. For each case, they found 5 controls matched along four discrete dimensions. The main variable of interest was the total, cumulative anticholinergic exposure which they calculated according to a well-established protocol that allows standardization across drugs with different dosage regimens. The authors also decided in advance to study the association between the total standardized daily dose (TSDD) for each type of anticholinergic: antihistamines, antidepressants, antipsychotics, antiparkinsonian drugs were considered both separately and collectively. Finally, the study repeated the analysis for patients diagnosed with Alzheimer’s disease, those diagnosed with vascular dementia, and those with some other form of dementia. So what did they find?

The main finding was that the adjusted odds ratio associated with low anticholinergic exposure was 1.06, while the ratio associated with the highest degree of anticholinergic exposure was 1.49, a highly significant difference. Also interesting was the observation that only certain groups of anticholinergics increased the risk of developing dementia. In particular, antidepressants, antipsychotics, antiepileptics, and drugs used to treat Parkinson’s and incontinence were the main offenders (and antihistamines, which geriatric physicians inveigh against as potential causes of delirium, had no effect). Curiously, the effects were more dramatic in patients with vascular dementia than in those with Alzheimer’s disease, a novel finding. Finally, the strongest association was found in people diagnosed before age 80.

As the authors are quick to point out, associations do not demonstrate causality. They can’t. Researchers simply cannot exclude the possibility that some of the drug use in fact reflected early, preclinical effects of as yet undiagnosed dementia. For example, many people with dementia are depressed; it is entirely possible that the depression manifests itself before a formal diagnosis is made, at a time when patients are beginning to detect subtle but disturbing changes in their memory and problem-solving ability. But if anticholinergics are causative, then they may well be responsible for as much as 10 percent of all dementia. The evidence is sufficiently suggestive and the magnitude of the danger sufficiently great that it’s time to be very wary of these drugs, especially in people under age 80. 

What Does Dying Have to Do With It?

What Katy Butler gets spectacularly right in her new book, The Art of Dying Well, is that if we want life's last chapter to be a good one, there’s a great deal more to talk about than death and dying. She understands, which so many writers about aging do not, that maintaining function—the ability to walk, to see, to hear, and a host of other verbs describing the actions that are critical for a fulfilling life—is of paramount importance in this phase of life. She understands that medical tests, procedures, and treatments often do more harm than good and this danger becomes greater as the number of underlying medical problems grows, which happens more and more often with advancing age. So why, then, does she call her book the “art of dying well?”

At first, I speculated that the title had been chosen by the publisher’s marketing division, as often happens, chosen perhaps because books about dying are in vogue, or at least more so than are books about frailty or chronic disease. Then I wondered whether the problem was merely semantic—after all, the formative experience that awakened Butler to the issue of “dying well” was that of her father, which she poignantly describes in her previous book, “Knocking on Heaven’s Door.” Her father had a stroke, only to spend the next seven or so years declining, his life prolonged by medical technology such as a pacemaker. From his daughter’s point of view, that entire period of decline could be viewed as “dying,” even though it was measured in years, not days or months. But Butler says that her goal in her new book is to provide readers with “a step-by-step guide to remaining as healthy and happy as possible, and as medically and unafraid, through the predictable health stages of late life, from vigorous old age to final breath.” Although I would argue with the implication that everyone goes through “predictable health stages”—some people plunge headlong into frailty, for example, whereas others move towards it gradually and others go directly from being robust to dying with virtually no time between the two—she does acknowledge that there’s more to old age than dying. In the very next breath, however, she says that “the goal of each chapter is to help you thrive and keep you on a path to a good end of life.” In other words, a major part of the point is to act today to assure a good death tomorrow. I would emphasize optimizing each day, rather than assuming that the purpose of your behavior today is to prevent a bad death.

The same phenomenon of grasping what old age is all about but not quite getting it is evident in Butler’s misconception about the “goals of care.” In Chapter 4, “Awareness of Mortality,” she asserts that discussing the goals of care is “medical shorthand for exploring what matters most to you [yes!], and how medicine can help you accomplish it [yes!], when time is short and cure is not in the cards [no!]” I think that patients and their physicians need to clarify the goals of care at every stage of life, not just when the end is near. It’s true that most people who are vigorous and are not afflicted with a fatal illness will choose life-prolongation as their main goal. But it is also true that many people who suffer from multiple chronic conditions but who do not have a terminal diagnosis and who can anticipate another ten years of life may choose as their principal goal of care “maximizing function.” Butler is right that for some physicians, discussing the goals of care is a euphemism for moving from treatment that seeks to cure to treatment that seeks to comfort--but goals of care discussions ought to be far more than that.

Then there's Butler's curious discussion of why you should cultivate a network of friendships in old age and find ways to remain engaged with life. Both are decidedly beneficial, as Butler asserts, but not just because they will prove useful “later on.” Relationships and engagement are ways to find meaning in life after the children have grown up and moved away and after retirement. This is yet another instance of the author seeming to understand what’s important as people age but then backsliding into thinking it’s important only as a means to assuring a good death. Befriending your neighbor can be rewarding in and of itself, not just so she will buy groceries for you when you are too ill to do so yourself.

Butler does an admirable job of conveying some of the main insights of geriatrics and palliative care. She understands, for instance, that the hospital is often a perilous environment for an older person, leading to loss of some of the functions most critical to remaining independent. She recognizes that physicians often focus on the benefits of medical technology, whether an implanted cardiac defibrillator (ICD) or an artificial heart valve, and fail to consider their risks. She rightly identifies home care programs, advance care planning, and enrollment in hospice as potentially life-enhancing strategies. But then she makes statements about medicine that are at best misleading and at worst simply wrong. For example, she says that “Benadryl and the sleeping pills are…anticholinergics, an insidious group of commonly prescribed drugs that befuddle thinking and substantially increase the likelihood of developing dementia.” Yes, anticholinergics can result in delirium, a form of acute, reversible confusion. But dementia? 

Butler goes on, a few pages later, to report on a “landmark study” that found that people who used anticholinergics heavily were 50 percent more likely than those who took few to develop dementia. What she doesn’t say is that it’s very misleading to cite relative risk rather than absolute risk: going from a risk of 1 in a 100 to a risk of 1.5 in a 100 constitutes a 50 percent increase in risk, but the outcome in question remains very rare. She doesn’t say that this study lumped many different medications with anticholinergic activity together, including a variety of drugs that are no longer in widespread use, such as the tricyclic antidepressants. She also does not mention that drugs that block acid production (the proton pump inhibitors such as Prilosec and Prevacid) have also been associated, statistically, with developing dementia, and so have anti-anxiety agents. Is it really the case that all these drugs “cause” dementia? Or might it be that people who take certain kinds of drugs—perhaps because they are already exhibiting the earliest signs of dementia—are more likely to go on and develop the full-blown disease? Before jumping to conclusions, observational studies of the kind Butler cites (as opposed to a randomized controlled trial) need to be replicated or, ideally, followed up with a study in which some people are given anticholinergics for a given condition and others, chosen at random, are given something else. 

To be fair, the author of the “landmark study” has gone on to carry out many other observational studies. Her most recent report on this subject appeared in the British Medical Journal last year partially confirmed her earlier findings. But expert analysis of this paper is mixed. At best, it is reasonable to conclude that anticholinergic medications might be a risk factor for dementia.

The Art of Dying Well has much to offer. I only wish it had been more scrupulously reviewed by geriatricians before publication.

Beer(s)

I haven’t written a blog post in a while. Not because I’ve been ill or travelling, but simply because I couldn’t find anything I was interested in writing about. Now, at last, I came upon some appropriate material. Next week I hope to blog about Katy Butler’s new book, The Art of Dying Well. This week I want to say a bit about the “2019 American Geriatrics Society Beers Criteria for Potentially Inappropriate Medication Use in Older Adults.” 

The Beers criteria were first promulgated in 1991 by Mark Beers and published in the Archives of Internal Medicine. The idea of reviewing the literature on adverse drug reactions in the elderly and publicizing a list of the worst offenders was widely applauded—but the methodology used in the original list was severely criticized. The American Geriatrics Society subsequently adopted the project and has been responsible for recent revisions. The newest list was just issued.

As with earlier AGS revisions, the expert panel charged with evaluating medications used evidence-based criteria for its judgments. For every medication about which it makes a recommendation, it indicates the quality of the evidence and the strength of the recommendation. Moreover, the panel distinguishes among three groups of potentially inappropriate medications: 1) those that should be avoided by older people in general, either because of a high risk of adverse effects, limited effectiveness, or the availability of better alternatives; 2) those that should be avoided by people with certain medical conditions (for example, renal failure); and 3) those in which the risks generally outweigh the benefits but which may be useful for particular individuals.

Physicians should keep all the tables on hand as a reference; because the medications are sorted by therapeutic category, i.e. “anti-infective” or “cardiovascular” and only generic names are used, the list is less useful for patients and families. I am going to summarize some of the main points for the general reader.

One group of medications that the AGS strongly recommends avoiding (although interestingly, the quality of the evidence is rated as “moderate”) are the first- generation antihistamines, drugs such as diphenhydramine (benadryl) and hydroxyzine (atarax) that are used against allergies and itching. They cause dry mouth, constipation, and confusion. To a large extent they have been replaced by the second-generation antihistamines (fexofenadine or Allegra, loratadine or Claritin, and cetirizine or Zyrtec) and their isomers, sometimes called third- generation antihistamines (levocetirizine or Xyzal). 

A number of first-generation antidepressants are similarly to be avoided because they, like the antihistamines, are anticholinergic (amitriptyline or Elavil, imipramine or Tofranil, and desipramine or Norpramin)—though in this case the quality of the evidence is high. These drugs have largely been superseded by newer antidepressants such as the selective serotonin reuptake inhibitors (SSRIs including fluoxetine or Prozac, citalopram or Celexa, and sertraline or Zoloft) and serotonin and norepinephrine reuptake inhibitors (SNRIs such as duloxetine or Cymbalta and venlaxafine or Effexor). It’s important to note that all these drugs can cause confusion; a few highly publicized studies notwithstanding, the AGS does not claim that any of these medications actually cause dementia. It’s also worth commenting that while all the antihistamines are available over the counter, so older patients might mistakenly choose one of the first-generation drugs over less toxic agents, the antidepressants are all prescription drugs and the first-generation agents are very seldom prescribed by physicians.

A second group of medications that AGS singles out are the antipsychotics, both the first-generation variety (drugs such as haloperidol or Haldol and fluphenazine or Prolixin) and second-generation agents (olanzapine or Zyprexa, quetiapine or Seroquil, and risperidone or Risperdal). All these drugs are deemed risky and are to be avoided except in people with schizophrenia or in people with dementia who are exhibiting dangerous behavior that has failed to respond to other alternatives. This is important because physicians continue to prescribe antipsychotics for the behavioral manifestations of dementia, despite compelling evidence that they are ineffective and/or risky.

Finally, both the benzodiazepines (used to treat anxiety or sleep disorders) and non-steroidal anti-inflammatory pain medications (drugs such as ibuprofen or Motrin and naproxen or Naprosyn but not celecoxib or Celebrex) receive an “avoid” recommendation, though in both cases the quality of the evidence is moderate but the recommendation is deemed strong. This is noteworthy because both groups of medicines continue to be prescribed by physicians and many of the second group are available over-the-counter.

The newest iteration of the Beers list is not perfect. But at the very least, there should be an awfully good reason for an older person to take any of the drugs the AGS says to avoid.

Treatment for Alzheimer's or False Alarm?

The most promising treatment of Alzheimer’s disease to date uses monoclonal antibodies to rid the brain of amyloid plaques. These plaques are widely thought to be crucial in the development of Alzheimer’s disease, the form of dementia that afflicts over 5 million Americans—and, unless we find a way to cure or prevent it, will afflict millions more in the coming years. Researchers hope that these drugs will prove to be “disease-modifying,” that they will disrupt the cascade of events that produces the symptoms of the disease. 

So far, the studies of monoclonal antibodies in this setting have been very disappointing: Bapinezumab: 3 failed trials. Solanezumab: 3 failed trials. Crenezumab: 1 failed trial. In the past year, three major pharmaceutical companies, Eli Lilly, AstraZeneca, and Merck, all stopped development of the monoclonal antibodies they were testing against Alzheimer’s disease. Another company, Pfizer, closed its neurodegenerative disease research section entirely. Now we learn that a Phase II trial of the as yet unnamed agent “BAN2401” actually “looks promising.” How promising?

Like its cousins, BAN2401 is a “humanized monoclonal antibody.” That means an antibody that was developed in mice and modified so it wouldn’t be rejected as alien by the human immune system. Now mice don’t normally get Alzheimer’s disease but certain strains have been genetically engineered so that they do. Using the genetically engineered mouse model, a monoclonal antibody was created that binds to amyloid. Today, this chemical is called BAN2401; when it enters the next stage of assessment, it will presumably acquire a name that ends in “zumab” to indicate it is just such a humanized monoclonal antibody. So, what do we know about this nameless entity?

We know that BAN2401 made it through a Phase I trial that tested its safety, tolerability, and pharmacokinetics in 80 people with mild to moderate dementia. We know it then entered Phase II of testing in people aged 50-90 with evidence of amyloid on their PET scans and either mild Alzheimer’s disease or mild cognitive impairment (a condition that is not dementia but often evolves into dementia). Five different dose regimens were administered intravenously and compared to placebo in terms of their effects on cognition and on brain amyloid deposits on PET scan. We know that when the preliminary results were announced last December, no cognitive benefit was reported. But now, after 18 months of study, the final results are in. They were presented at the Alzheimer’s Association International Conference in Chicago this past week.

What we don’t know is how this study will fare when subjected to peer review. We don’t know what the final paper will look like. But here’s what I can glean from the press release by Eisai, the company that collaborated with Biogen to develop and test the drug, and the reports by the news media:

--The amount of amyloid in the brain (as measured by PET scan) decreased with all the doses tested. The more drug, the greater the change. The findings were statistically significant. 

--There was no discernible effect on cognitive function except in the 161 people who were treated with the highest dose of BAN2401. In those 161 people, cognitive function deteriorated over time, just as it did in all the other groups, but it deteriorated less.

--Cognitive function was measured in a way that differs from the way it has been measured in other studies of treatments for Alzheimer’s disease. The researchers used a composite measure made up of several scales, each of which has been individually validated as a way to assess mental status, but only one of which is regularly used to report the outcome of clinical trials of Alzheimer’s drugs. 

--The patients tested included a mix of people with mild dementia and people with mild cognitive impairment (who don’t actually have dementia but aren’t entirely normal either) and the study had no way to determine whether the effectiveness of BAN2401 was any different in the two subgroups.

What does all this mean? The PET scan changes indicate that the monoclonal antibodies were successful in destroying amyloid deposits in the brain. That doesn’t prove that whatever effect the drug had was due to its amyloid-busting, but it’s suggestive. It also indicates that to the extent that the drug was a failure, it wasn’t a failure because it didn’t destroy amyloid. 

How should we interpret the cognitive changes? It’s hard to know, given that the researchers used a non-standard means of measuring clinical decline (a test they called ADCOMS). We cannot, for example, say that BAN2401 is better than one of the other monoclonal antibodies that was deemed a failure since nobody looked at the effect of the other drug on the “ADCOMS.” What we can say is that a 30 percent change in the rate of decline may be statistically significant but clinically, it’s a lot less impressive. Everyone  deteriorated, including those given the highest dose regimen; they just deteriorated a little less. How much of a difference actually occurred depends on the absolute decline: suppose the score went from 100 at baseline to 80 at the end of the study in the controls (I don’t know what the actual numbers are, so this is hypothetical). A 30 percent difference in the rate of decline means that the score in the treated group dropped from 100 to 86.  Is that clinically meaningful? Probably not.

So, no, BAN2401 is not likely to likely to dramatically change the course of Alzheimer’s disease. I hope the FDA does not allow the drug’s manufacturers catapult it into a clinical trial prematurely, before it goes through Phase 3 testing. But it has taught us a few valuable lessons. 

We now know that monoclonal antibodies can be designed that destroy amyloid plaques in the brain. We know that only the highest tolerated dose of the drug has any chance of being clinically useful. We should also realize the importance of studying patients with early dementia and those with MCI separately—it’s possible that using monoclonal antibodies once dementia has set in is too late, and I suppose that it’s also possible that using them before dementia has developed is too early. Finally, the study reminds us of the importance of a uniform methodology in conducting this kind of work. If the test that was previously regarded as the gold standard, the ADAS-Cog, is not the right test to use, then researchers need to agree on that and decide collectively what test to use instead. 

ABC News reported the drug as “giving patients hope” and showing “big promise.” Fortune magazine asserted that “world leaders want to end Alzheimer’s by 2025; a new drug breakthrough means we just might.” Investors are closer to the mark: Eisai’s stock price fell 10 percent after the data were reported.

False Hope

This past week, the House passed its “Right to Try” bill, which would give dying patients easier access to experimental drugs. The Senate has already passed a version of this ill-conceived legislation and Trump is an enthusiastic supporter, viewing it as a means of defanging the FDA and of demonstrating his great compassion for the American people. The bill itself, if it is signed into law, will not affect very many people and most of them are not likely to be elderly. But it is the first step in a campaign to destroy drug regulation. That is an issue for all of us, and for older people, as the greatest per capita consumers of prescription medications, in particular.

A STAT article on the subject quotes Andrew McFadyen, executive director of a non-profit patient advocacy group: “I think this is the first step, for sure. Tear down as many regulations as possible, take away all oversight, and let it be the Wild West of medicine.” He continues that its Republican proponents “opened the door to Koch brothers and Goldwater to rip apart the FDA, and then other government bodies after that.”

It’s important to recognize three things: 1) the FDA plays an enormous role in patient safety. Tearing up the regulatory framework, as Trump and his ilk favor, would have devastating consequences for health and well-being; 2) the history of granting desperate patients access to experimental drugs is one of offering false hope and enormous suffering; 3) the track record of the experimental drugs that the current legislation would support is mediocre.

The FDA approval process keeps us safe. When the FDA was first created in 1906, all it required was that drug labels accurately report the contents of medication. No longer was it permissible to sell “tonics” such as “Pinkham’s Vegetable Compound” or “Mrs. Winslow’s Soothing Syrup” without divulging on the label that the former contained alcohol and the latter morphine. 

The idea that FDA approval required that a drug be demonstrably safe and effective came much later. This amendments to the law were finally passed in 1938--after 5 years of debate--after the sulfa disaster of 1937: a liquid form of the new antibiotic, sulfa, had become widely available and was marketed for use in children. Because there were no safety tests performed, nobody seemed to have noticed that it was made with ethylene glycol (the active ingredient of antifreeze), a toxic chemical that causes liver failure and death—and that killed over 100 children. 

The requirement that drug companies actually perform randomized controlled trials to determine the safety and efficacy of proposed new drugs had to await the thalidomide disaster of the early 1960. Only after pregnant women treated with sleeping pills gave birth to babies with severe congenital malformations was were the Kefauver-Harris amendments to the FDA passed in 1962.

The track record of experimental drugs is poor. While the proponents of the new legislation, conservative groups such as the Goldwater Institute (named after Barry Goldwater) focus on “free choice” as the rationale for making untested substances available to the public, the widespread assumption is that substances entering a “phase III” clinical trial are all but proven drugs—just a few more bureaucratic hurdles, and they will be FDA approved. In fact, a phase III trial is designed to determine whether a drug that has shown some evidence of efficacy in a smaller, earlier trial (one that was focused primarily on looking for side effects) is actually effective. Many drugs that reach the Phase III stage do not move on in the approval process because they are not found to work. For cancer drugs, only 40% move ahead.

One of the most dramatic examples of the unfortunate consequences of premature access to an experimental therapy is the case of bone marrow transplant for advanced breast cancer. In the 1990s, more than 30,000 American women received bone marrow transplants at the cost of millions of dollars, extremely unpleasant side effects (including death—treatment-related mortality was reportedly as high as 15%). Insurance companies found themselves pressured to pay for the therapy—until several randomized controlled trials definitively demonstrated that it was no better than conventional chemotherapy.

The sad reality is that dying patients are vulnerable to claims that a cure is lurking around the corner. More often than not, what they are being offered is an opportunity to spend their dwindling resources on a bottle of false hope, often suffering enormously in the process. This is what the “right to try” legislation offers. The House and Senate versions must be reconciled for “right to try” to become law. Tell your senator to halt the assault on the FDA and keep drugs safe and effective.

Down, Down, Down

         The most important recent development affecting older patients, without a doubt, is the issuance of new guidelines for the diagnosis and treatment of high blood pressure. The Wall Street Journal proclaimed “Nearly Half of US Adults Have High Blood Pressure Under New Guidelines,” and venerable health columnist Gina Kolata of the NY Times wrote: “The nation’s leading heart experts on Monday issued new guidelines for high blood pressure that mean tens of millions more Americans will meet the criteria for the condition, and will need to change their lifestyles or take medicines to treat it.” The report, which took me a while to track down and read since most of the references are to summaries of the report or commentaries on the report but not to the actual document, is 175 pages. It was issued or endorsed by eleven organizations, whose initials actually form part of the name of the report: its full name is the “2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPCNMA/PCNA Guidelines for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults.” The subtitle is: “A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.” Now that I’ve had a chance to digest this—what should we make of it?

         I remember the day in 1991 when JAMA published the results of the “SHEP” trial, a study of the treatment of isolated systolic hypertension in the elderly that upended what I’d been taught in medical school. Gone was the belief that high blood pressure was good for older people because they needed greater force to push blood through their stiff arteries; suddenly, there was compelling evidence that high blood pressure was dangerous—it caused strokes and heart attacks. Keeping the systolic blood pressure below 160 reduced the risk of stroke by 36 percent and the risk of a cardiac event by 27 percent. Overnight, the practice of medicine changed.

         To be sure, there were skeptics. Wouldn’t lowering the blood pressure cause older people to become dizzy and to faint? Just how far were we supposed to lower blood pressure anyway? But to me and many others, the choice was clear. Stroke was one of the worst fates that could befall older patients: sometimes it killed them, but more often, it left them impaired, often profoundly, significantly diminishing the quality of life.

         But I also remember a book by physician and historian of science, Jeremy Greene, published in 2008 and called Prescribing by Numbers: Drugs and the Definition of Disease. Greene argued persuasively that the pharmaceutical industry has a vested interest in lowering the cut-off for treatment of a variety of chronic conditions, such as hypertension, high cholesterol, and diabetes. Defining “pre-diabetes” as a real condition warranting treatment or repeatedly dropping the threshold for treating cholesterol with statins had clear implications for the drug companies: they would sell more pills. More volume, more revenue, more profits. So are the new guidelines just another instance of “diagnosis creep,” a way for Pharma to make more money?

         It’s important to realize that what tipped the scales for the guideline writers, what’s different now from the last time they wrote a guideline, three years ago, is the 2015 “SPRINT” study (Systolic Blood Pressure Intervention Trial), a randomized trial of high risk patients over 50 to either a systolic blood pressure target of 140 or 120. The study, which was reported in the New England Journal of Medicine, found a 25 percent reduction in cardiovascular events (such as stroke or heart attacks) when physicians tried to reduce the blood pressure to the lower value.

         As Gilbert Welch of the Dartmouth Institute, co-author of Overdiagnosed: Making People Sick in the Pursuit of Health (2012) and Less Medicine, More Health (2016), commented, that 25 percent sounds pretty good, but it’s the relative risk reduction. In fact, the rate of cardiac events fell from 8 percent in one group to 6 percent in the other—an absolute difference of only 2 percent. Or, looked at differently, 92 percent of people in one group and 94 percent of people in the other group did just fine. Suddenly, the impressive results don’t look quite so impressive. Moreover, all these patients were already at increased risk of cardiovascular disease because of other factors such as cigarette smoking or diabetes. Presumably, had the same study been carried out in people of average or lower risk, the benefits would have been correspondingly smaller.

         So if the new guidelines are based predominantly on the findings from one study, and the one study isn’t quite as compelling as it seemed at first glance, what should we make of the new recommendations? Should we really treat everyone with a systolic blood pressure over 120 who has at least a 10 percent risk of cardiac disease in the next decade? 

         The answer, I think, is yes, but with caveats. First, the guideline writers are at great pains to insist that blood pressure be measured the way it was measured in the SPRINT trial—in a way that it’s almost never measured in the doctor’s office, namely after sitting quietly for five minutes and averaged over three readings. They also advocate use of home monitoring to confirm (or refute) the diagnosis of high blood pressure, as well as to guide medication adjustment once treatment is started. Next, physicians need to be prudent about what medications to use if they’re going to embark on pharmacologic treatment. We have ample numbers of cheap, effective medicines that have been around for years, such as diuretics and beta blockers, and these should be tried before pricier medicines. Finally, both physicians and patients need to be vigilant about medication side effects, which means doctors need to tell their patients what to look for, patients need to report their symptoms, and doctors need to change course if symptoms develop. In addition, non-pharmacologic treatment should be attempted, including exercise, weight loss, and a low salt diet. If we follow all these steps, we can be confident we will be changing how we deal with blood pressure because it’s what’s best for patients, not because it’s what’s best for the pharmaceutical industry.

A Piece of My Mind

Prescription medications cost more in the US than anywhere else in the world and costs have been skyrocketing each year for the past several years. To a growing extent, the burden of the high cost falls directly on consumers, either because their health plan has a tiered system for medications (charging ever larger co-pays for some drugs), because their health plan pays only a percentage of the charge for various drugs (and if the consumer has to pay 20 percent and it’s 20 percent of a very large number, that’s a major outlay), or because they have a high-deductible health plan and the insurer doesn’t pay anything until they have spent $3000—or $5000 or $10,000—on health care. 

By and large, pharmaceutical companies have been blamed for the high cost of medicines, with insurers shouldering some of the blame, thanks to complicated and ungenerous policies. Pharma has tried to justify its sometimes astronomical charges as necessary to support its research efforts, with the most recent industry-endorsed estimate for the cost of developing a new drug and bringing it to market now topping $2.6 billion. Other analyses attack the methodology used in this report to measure costs, arguing that it fails to take into account, for example, that NIH funds much of the research that goes into discovering a new drug, not the pharmaceutical industry. The result is a dramatic over-estimate of the cost borne by industry. Concerns about the role of drug companies and to some extent health insurers are entirely legitimate. But there has been little attention paid to the role of drug stores in contributing to the high cost of medicines.

I did a little bit of investigating today. I looked at what two commonly used medications would cost a family like mine who had exceeded the $4000 deductible for their health plan, what they cost today (given that it’s only early March and most people haven’t had the opportunity to spend $4000 on medical care this year), and what they would cost if they were obtained from a Canadian mail-order pharmacy. Here’s what I found for one of the medicines, the widely used nonsteroidal anti-inflammatory drug, Celecoxib.

Celecoxib is used as a treatment for arthritis in people with certain gastrointestinal conditions because it's a little less prone to exacerbate these problems than other anti-inflammatory drugs. It is available generically. The non-profit insurance company Harvard Pilgrim Health Care classifies generic Celecoxib as a “tier 1” drug. That means that the cost of a 3-month mail order supply of the medication (100 mg taken once a day) from Walgreen’s, the pharmacy with which Harvard Pilgrim has a contract, would be $10. But until the deductible is met, Harvard Pilgrim doesn’t pay for medications, so the cost would be a whopping $156.51 for 90 pills (which, incidentally, isn’t even quite a 3-month supply since last I looked, a year has 365 days, not 360 days). So I contacted a pharmacy in Canada, identifying one that is approved by CIPA, the Canadian association of licensed retailed pharmacies.  I found a drug store that will supply 120 pills for $25.99 (plus a small shipping charge). That comes out to $1.74 per pill at Walgreen’s compared to 22¢ at the Canadian competitor. Walgreen’s costs eight times as much as the Canadian pharmacy. And the medication isn’t manufactured in some shady country with questionable oversight. It’s made in the UK.

How can this be? Is Celecoxib a fluke? So I looked at another commonly prescribed medication, this time a drug classified as tier 3. I chose Vagifem, an estrogen suppository, used to treat post-menopausal atrophic vaginitis. The cost of a 3-month supply through the health plan—after using the entire $4000 deductible? $80. The mail order cost from Walgreen’s today, assuming the deductible hasn't been spent? $360. The cost from the Canadian pharmacy? $55. Made in the UK. Walgreen’s is 7 times more expensive.

What’s going on here? I’ll leave that to the policy wonks, but maybe they should look at the behavior of pharmacies as well as drug companies and health insurers. After all, this isn't just a case of generic drugs costing almost as much as their brand name equivalents, which is still another problem for consumers. Meanwhile, importing drugs from Canada is a valuable option. It's not legal to re-import medication for sale or to import restricted drugs such as opioids, but the law on medicines for personal use is a bit fuzzy, or at least its enforcement is. With consumers shouldering an ever increasing proportion of health care costs, and no prospect for relief in sight, there's a strong incentive to look north. 

Three Strikes and You're Out

Three times, Eli Lilly has tried to demonstrate the efficacy of the drug solanezumab in the treatment of Alzheimer’s disease. And three times, the drug failed. What does this mean?

Solanezumab is a monoclonal antibody that selectively attaches to a form of amyloid-beta, widely thought to be the key agent that causes Alzheimer’s disease. The first two studies were phase three, double-blind studies of patients with mild to moderate Alzheimer’s disease. That means earlier studies had found the drug reasonably safe and had identified a dose that had the desired effect on brain amyloid. The results of EXPEDITION1  and EXPEDITION 2 were both published in the New England Journal of Medicine in January, 2014. The authors found that after 18 months of treatment with intravenous solanezumab, there was no improvement in either tests of cognitive function (the Alzheimer’s Disease Assessment Scale) or daily activities (the Alzheimer’s Disease Cooperative Study Activities of Daily Living scale). There was a hint in subgroup analysis that maybe the patients with milder disease might benefit; hence the new study, EXPEDITION3, which evaluated the drug in patients with clinical evidence of mild dementia and abnormalities on PET scan suggestive of amyloid deposition. But alas, the results of this trial are no more encouraging than those of its predecessors.

Solanezumab is not the only anti-amyloid monoclonal antibody that has been tested and found wanting. Two other double blind randomized, placebo-controlled phase 3 trials were conducted with Bapineuzumab in patients with mild to moderate Alzheimer’s, One trial involved carriers of the APOE-4 allele, a gene for a cholesterol carrying protein that has been associated with an increased risk of dementia; the other involved non-carriers. Once again, there was no difference between patients receiving active drug and those getting placebo, whether on tests of cognition or on measures of overall daily functioning.

So now solanezumab has failed three times. Not because it caused terrible side effects, as was the case with earlier attempts to “vaccinate” against Alzheimer’s disease, that were in some instances effective but that caused a debilitating and/or lethal inflammatory response in the central nervous system. Solanezumab failed because it didn’t work. What does this tell us about the use of monoclonal antibodies in Alzheimer’s disease?

An article in Scientific American two years ago on “Why Alzheimer’s Drugs Keep Failing” makes clear that it’s not just a handful of drugs that have fallen short of their promise. According to the article, “dementia has become a graveyard for a large number of promising drugs.” In fact, the failure rate for Alzheimer’s disease “drug candidates” is 99.6 percent (compared to 81 percent for cancer drugs). The problem, in large measure, is that symptoms begin a decade or more after the disease actually starts, by which time intervention may be too late. It’s as though we didn’t treat coronary artery disease until after a patient developed cardiomyopathy, until after so much heart muscle had died that the heart scarcely pumped at all. One strategy for drug development is therefore to identify markers for disease long before there are clinical manifestations, in principle allowing for treatment at a much earlier stage than occurred with any of the solanezumab trials. Another is to target other suspects besides amyloid—abnormal proteins such as tau that also contribute to the development of dementia. Still another strategy is to focus on “disease-modifying treatment,” an approach that slows the progression of the disease rather than one that focuses on cure.

Perhaps another reason that Alzheimer’s disease is such a hard nut to crack is that the brain is an extraordinarily complicated organ, arguably far more complex than the heart. The heart, after all, is basically a pump. An ingenious pump fed by a system of blood vessels and operated by an electrical mechanism, and one that has proved far more complicated than was initially assumed, but nonetheless simple by comparison with the brain. Because the brain is such a remarkable organ and because Alzheimer’s disease strikes its most elusive function, cognition, it is not entirely surprising that treating this disorder has proved so challenging.

The message is not that we should give up. But it is that there’s no cure around the corner. The most effective strategy to date is to decrease the likelihood of developing Alzheimer’s through a combination of smoking cessation, treatment of high blood pressure, and treatment (or prevention) of diabetes—a strategy developed for the prevention of cardiac disease and incidentally found to help stave off dementia. And in the mean time, as scientists continue to struggle to understand the pathogenesis of Alzheimer’s and devise new strategies for intervening in the relentless progression of this dread disease, we must continue to find better ways to support people who are afflicted, allowing them to find as much meaning and satisfaction as possible in their lives.

This is the last post on “Life in the End Zone” in 2016. See you next year!