Mumbling orphans—a bit more

Mark McQuain has raised the persistent, vexing issue of the pricing of drugs for rare diseases—in the case at hand, Sarepta’s eteplirsen (Exondys 51) for Duchenne Muscular Dystrophy, the disease over which the late comedian Jerry Lewis lost sleep every Labor Day weekend for years.

Mark provided an excellent summary (he calls it “crude,” but it’s anything but that).  In this case, the concern is not just price for a truly rare disease, but whether the drug showed sufficient evidence that it worked for FDA to approve it.  In the absence of alternative treatments, that was the truly tempestuous issue for Sarepta.  (Recall that under the 1962 Kefauver-Harris amendments to the Federal Food, Drug, and Cosmetic act, drug manufacturers in the U.S. may not sell a drug unless the FDA finds it not only safe, but effective—a standard that generally applies worldwide.)  It’s one thing for a drug to have a high price, but rather another if it doesn’t work, or doesn’t work very well.  (I decline to comment publicly about the Sarepta data; outside my expertise.  Those seeking a case in point may wish to consider Avastin for breast cancer.)

And to be sure the high price concern dogs other treatments that appear to work quite well—such as high-profile ones for cystic fibrosis or for cancer.  A case can be made that such drugs are worth the price, that too much government heavy-handedness risks stifling innovation, and that a search for the “just price” is misguided, but also, for sure, that society should share the costs of some of these drugs, that measures should be taken to limit out-of-pocket costs to disease sufferers, and that reimbursement approaches are ripe for overhaul.  In that last bucket: if drugs work only some of the time, only pay for the cases in which they do work; foster true competition (rather than having the costs of all drugs in a class go up when a new one is introduced, as if drugs were houses); eliminate the middle man (i.e., pharmacy-benefit managers that take a cut—that appears on the horizon); and the “biggie,” having government payers push back harder on prices.  At least some of these measures seem likely, and at least some seem warranted.

But overall, high costs for truly innovative treatments are justifiable, where no alternatives existed previously and especially when other, more expensive and quite possibly less effective medical treatments may be obviated (see: drug treatment for hepatitis C vs liver transplantation).  This is not to endorse price gouging for existent, cheap drugs that fall into an incidental monopoly (in which case, BTW, elimination of said monopoly, through regulatory facilitation of alternative sources, is warranted).

Public input into gene-editing decisions

Lyme disease is caused by a type of bacteria that lives in mice, which are considered a “reservoir” for the disease-causing agent.  Ticks bite the mice, pick up the bacteria, and then infect people when they bite them.  (Ticks are called the “vector” for the disease.)

If mice were immune to the bacteria, their immune systems would destroy them, and there’s be no reservoir, and no Lyme disease.  If scientists genetically engineered mice to make them immune—for example, by editing their genes—they could make progress toward that goal.  But to work, the mouse population would have to be predominantly made of bacteria-immune mice.  That could be accomplished by using “gene drive,” an approach that would make the altered gene spread preferentially and rapidly in the population.  However, doing that could alter the environment in unpredictable ways.

Because of the risks, scientists on the “Mice Against Ticks” project are determined that even if they succeed in genetically altering mice as suggested, they will not release those mice into the wild without full public awareness and approval.  They are holding public meetings—specifically, in Martha’s Vineyard and Nantucket—well in advance of the project coming to full fruition.  And they are trying to figure out, with the public, what level of communication and acceptance constitutes public approval.

Similarly, scientists in New Zealand would like to use a form of gene drive to greatly reduce the population of rats, possums, and other destructive predators that are decimating the environment.  And their public deliberations include seeking advice and, before taking action, buy-in from a network of Maori leaders.  Those conversations are so sensitive that the Maori objected when the scientists published a “what-if” type of article discussing the issues raised by the technology.  Among the concerns: some readers got the impression that gene editing of the animals was imminent, not hypothetical, as it still is.  Some of the news coverage of the Nuffield Council’s recent deliberations about the potential acceptability of heritable human gene editing seemed, likewise, to create the impression that the birth of the first gene-edited human is upon us—which it is not, not quite yet.

The public discussions above are two commendable moves toward true public involvement in decision-making about gene editing.  They were described in a recent Wall Street Journal article.  If you have subscription access, by all means read it.

PAS and “plain old” suicide

Last week, the folks from the Manhattan Declaration (whose key concern is freedom of religion) sent an email with a series of links—perhaps expanding their remit a smidge—one of which dealt with doctor-assisted suicide.  “Doctor-assisted suicide is contagious,” it said, along with this: “Doctor-assisted suicide increases overall suicide rates among the non-terminally ill everywhere it’s made legal.” And it also linked to a 2015 article on the subject from the Southern Medical Journal.

The SMJ authors were looking at whether overall and non-assisted suicide rates were different in states with legalized assisted death than in other U.S. states.  By their own admission, they did not have as much information as they would have liked.  They reviewed available information between 1990 and 2013. Two of the states were Vermont, which legalized assisted suicide in 2013 and had no cases that year, and Montana, which was not keeping count of assisted suicides.  That left them with assisted suicide data from Oregon (since 1998) and Washington state (since 2008).  (Since then, assisted suicide has also been made legal in Colorado, D.C., Hawaii, and California.)

In those two states, the total number of suicides and the number of non-assisted suicides had increased after the respective dates of legalization of assisted suicide.  However, the number of non-assisted suicides had also increased in the other states combined.  When I looked at their graph of the data, going back to 1990, there had been a decline in non-assisted suicide in Oregon, Washington, and other states between 1990 and 2000, with increases after that.  The number of these deaths per 100,000 population were higher in Oregon and Washington (and, indeed, much higher in Montana) than in the other states, but the slopes of the curves—the rates of increase—in Oregon and Washington looked similar to the other states.  Montana’s increase appeared sharply higher, but it’s hard to conclude anything about assisted suicide in that case because there were no data (nor for Vermont).

The authors commented that there had been estimates of how many surreptitious assisted suicides may have occurred outside of legalization, but admitted that firm conclusions were not possible.  They attempted to establish a statistical association between assisted suicide and nonassisted suicide, but the argument seemed inconclusive.  Follow the above link, read the article, and judge for yourself.

The core conclusion was that assisted suicide in its early years had not decreased overall suicide rates, as some of its advocates had argued it would.  Any relationship may become clearer as more information is available, assuming that numbers of assisted and non-assisted suicides are consistently and completely counted and recorded.  But that assumption is questionable at best.

In the meantime, it is important not to overinterpret limited data sets.  It’s also important to remember there are at least 5 reasons why assisted suicide is a bad idea:  it fundamentally alters the nature of medicine as a healing, life-preserving art and profession; its application cannot be reliably limited to those who freely request death or who are terminally ill (the slippery slope); it risks diverting energy and priority from true palliative care; if understood as a “right to die,” or a “right to be made dead,” it creates a duty for someone else to kill; and the notion of a “right to die” is self-contradictory if rights rest on preservation of life and well-being, as is the classical (i.e., pre-20th century) understanding.

In 2012, in his essay “Four Myths about Doctor-Assisted Suicide,” Ezekiel Emanuel wrote: “Patients themselves say that the primary motive [for assisted suicide] is not to escape physical pain but psychological distress; the main drivers are depression, hopelessness and fear of loss of autonomy and control. Dutch researchers, for a report published in 2005, followed 138 terminally ill cancer patients and found that depressed patients were four times more likely to request euthanasia or physician-assisted suicide. Nearly half of those who requested euthanasia were depressed.

“In this light, physician-assisted suicide looks less like a good death in the face of unremitting pain and more like plain old suicide. Typically, our response to suicidal feelings associated with depression and hopelessness is not to give people the means to end their lives but to offer them counseling and caring.”

And “plain old suicide” is, as we know, a problem that is getting worse all over the country.

Britain’s experts on gene-edited babies

by Jon Holmlund

Some of the cable news shows ran segments on the report released this week by Britain’s Nuffield Council on Bioethics, “Genome editing and human reproduction: social and ethical issues.”  Full disclosure: I have not yet read the full report, only the short summaries (all of which are available for free download at the link here).

The TV teasers—”U.K. bioethics council says that gene-editing children may be morally acceptable” were accurate.  The key conclusion is that “the use of heritable genome editing interventions to influence the characteristics of future generations could be ethically acceptable in some circumstances” (emphasis theirs).  But the news folks made it sound like an attempt to birth an edited baby is around the corner, or at least fully green-lighted by Nuffield.

The summary of the report reads more modestly, acknowledging that such attempts are currently banned by law most places, and that making them legal could require “a long and complex legislative pathway.”  But the Council does take the view that at least some attempts, such as those to try to repair a lethal disease gene such as the dominant gene for Huntington’s disease, might be justifiable.  This blog has considered such an argument in the case of sickle cell anemia—single gene defect, well understood, circumscribed attempt to repair only that gene.  An argument can be made.

The Nuffield Council’s summary really is a list of general statements that, taken individually, are hard to take issue with, and are in some cases almost platitudinous.  The overall impression is, “yes, heritable human gene editing could be ethical, and probably should be considered, but only after a long public deliberative process, appropriate regulation, etc., etc.”  Nuffield offers two stipulations for ethically acceptable heritable human gene editing:

  • “Intended to secure, and is consistent with, the welfare of a person who may be born as a consequence” of the effort, and
  • Social justice and solidarity are upheld; that is, discrimination or social division should not be a consequence.

These statements are both too broad to be helpful.  In the first case, the Council acknowledges that some efforts could be attempts to enhance a person’s natural characteristics, not just treat a recognized disease, and that, except for the most genetically straightforward cases, the scientific and technical challenges are substantial.  In the second case, it would seem that pressures for discrimination based on social attitudes or economics (ability to pay for the procedure, medical insurance reimbursement issues) will be unavoidable.

Scientifically and socially, there will be unintended—or at least undesirable—consequences.  These may be known but considered acceptable.  For example, how many human embryos will need to be created and destroyed to perfect the procedure?  How many generations will need to be followed to rule out some late complication?  Can we really guarantee that “having babies the old-fashioned way” won’t become a thing of the past?  And, in spite of the laudable desire to bring healthy children into the world, wouldn’t this be a wholesale acceptance of the basic assumption that only the people we want to be born, should be born?

For these reasons and others previously articulated on this blog, heritable human gene editing falls into a small but critical group of biomedical undertakings that should not be pursued.

And, BTW, the remaining bugs in the system include, as reported this week, that gene-editing techniques appear to introduce errors more frequently than previously appreciated.  Given that heritable human editing involves more than just a few cells in a dish, a “presumption to forebear” should apply.

The TV news gave this about 5 minutes this week.  That’s the breadth and depth of our “public deliberation” beyond a few experts.  At the end of one segment, the host looked into the camera and said, “next up: are liberals or conservatives happier?”

As Neil Postman said:  “now this…”

Raiding the CRISPR

BY JON HOLMLUND

A couple of gene-editing news items from this week’s science literature:

First, Nature reports that a group in my “back yard,” at the University of California San Diego, has tested gene editing using the CRISPR approach in mice.  Recall that CRISPR is an acronym for a particular molecular mechanism, first discovered in bacteria, that is particularly efficient—though not perfectly so!—at editing genes.  The idea is to find a “bad” gene that you’d like to replace, for example to prevent or treat a disease, and edit it to be the normal version of that gene.

The kicker in this particular case in mice is that it tested something called “gene drive.”  In classical genetics, humans (and other higher organisms) have two copies of each gene.  In sexual reproduction each parent passes one copy of the gene to offspring, so the chance of a particular gene being handed down is 50%.

“Gene drive” is a technique designed to change those odds, and make a particular gene “selfish,” and much more likely to be passed on.  In fact, the idea is that transmission would be 100%, or nearly so.  If that worked, then a new gene would soon take over a population of organisms, and every member would, in a few generations, have that gene.

Why might that be a good thing?  Suppose you are interested in pest control, and you could use the technique to make, say, mosquitoes infertile.  Then they would soon all die off.  Or if you had some other “desirable” characteristic, you could make it so all members of a species (rodents?  Cattle?  People?) have that characteristic.  Assuming it’s determined by one gene, that is.

And assuming that the technique works.  In the mouse experiment, efficiency was only 73%.

That’s probably good news.   This is one of those techniques that could have serious unintended consequences if tried in the field.  Scientists have been warning about that.  It looks like it’s a way off, but something else to fret about.

The second item involves a clinical trial to treat sickle cell anemia.  In this one, blood stem cells from a person with the disease are removed from the bloodstream and gene-edited outside the body to make hemoglobin that is not as damaged as in the disease (SCA is an inherited disease in which the red blood cells have abnormal hemoglobin that doesn’t carry oxygen well).  Then the altered cells become the therapy, and are given back to the patient.

The FDA has put a “clinical hold” on this clinical trial.  Exactly why has not been publicly disclosed (it doesn’t have to be), and it sounds like the trial itself hadn’t started yet, but that the company developing it was getting ready to start.  This is, in my view, an approach to gene editing that does not pose special or particularly worrisome ethical issues, because the genetic changes are done on “adult” stem cells to treat an existing individual with a disease in a way that would not entail transmission of altered genes to future generations.

And, probably, it’s a case of “this too shall pass,” and the FDA’s concerns will be answered and the trial will proceed.

But check out the sidebar reporting this in Nature Biotechnology.  If you follow the link you will probably get a prompt asking for payment but I was able to sneak a free read on my screen.  If you go there, read below the separate quote (itself picked up from The New York Times) from Dr. George Church of Harvard:  “Anyone who does synthetic biology [engineering of biological organisms] should be under surveillance, and anyone who does it without a license should be suspect.”  Apparently he said that in response to “the publication of an experiment recreating a virus that has engendered fears that such information could be used to create a bioweapon. ”

The old “dual use problem,” eh?  We should really fret about that.

Labs are growing human embryos for longer than ever before

BY JON HOLMLUND

That’s only a slight paraphrase of a news feature article this week in Nature.  The clearly-written article is devoid of scientific jargon, with helpful illustrations, open-access online, and readily accessible to the non-specialist.  Check it out.

Key points include:

  • Scientists who do not find it ethically unacceptable to create and destroy human embryos solely for research purposes continue to follow the so-called “14-day rule,” by which such experimentation is limited to the first 14 days after fertilization. At that point, the human nervous system starts to form and the time for twinning is past.
  • The 14-day rule is law in some nations, but until now has not been a practical issue because scientists have been unable to grow human embryos that long in the laboratory.
  • That technical limit has been sufficiently overcome that embryos are now surviving for almost 14 days. Scientists have not directly challenged the 14-day rule yet, but might, and would like to revisit it.
  • Experiments on human embryos in that time have included editing of critical genes to see what happens (sometimes they stop growing), and making hybrids of animal embryos with human cells whose purpose is to “organize” embryonic development rather than remain part of the developing individual.
  • Embryo-like structures, referred to as “embryoids” in the article, and sounding similar to “SHEEFs” (“synthetic human entities with embryo-like features”) are also being created. These entities don’t necessarily develop nervous systems in the same way as a natural embryo, prompting questions of just how much they are like natural embryos, whether the 14-day rule applies, and whether they raise other ethical concerns.

The last paragraph of the article, reproduced here with emphases added, is striking and more than a little ironic in light of arguments that embryos are “just a clump of cells”:

As the results of this research accumulate, the technical advances are inspiring a mixture of fascination and unease among scientists. Both are valuable reactions, says [Josephine] Johnston [bioethicist from the Hastings Center]. “That feeling of wonder and awe reminds us that this is the earliest version of human beings and that’s why so many people have moral misgivings,” she says. “It reminds us that this is not just a couple of cells in a dish.”

A safety concern with gene editing

BY JON HOLMLUND

Hat-tip to Dr. Joe Kelley for bring this to my attention…

As readers of this blog will recall, there is keen interest in exploiting recent discoveries in genetic engineering to “edit” disease-causing gene mutations and develop treatments for various diseases.  Initially, such treatments would likely use a patient’s own cells—removed from the body, edited to change the cells’ genes in a potentially therapeutic way, then return the altered cells to the patient’s bloodstream to find their way to the appropriate place and work to treat the disease.  How that would work could differ—make the cells do something they wouldn’t normally do, or make them do something better than they otherwise do (as in altering immune cells to treat cancer); or maybe make them work normally so that the normal function would replace the patient’s diseased function (as in altering blood cells for people with sickle cell anemia so that the altered cells make normal hemoglobin to replace the person’s diseased hemoglobin).

Or maybe we could even edit out a gene that causes disease (sickle cell anemia, Huntington’s disease) or increases the risk of disease (e.g., BRCA and cancer) so that future generations wouldn’t inherit it.  Or maybe we could edit genes to enhance certain health-promoting or other desirable qualities.

The recent scientific enthusiasm for gene editing is fueled by the discovery of the relatively slick and easy-to-use (if you’re a scientist, anyway) CRISPR-Cas9 system, which is a sort of immune system for bacteria but can be used to edit/alter genes in a lot of different kinds of cells.

It turns out that cells’ normal system to repair gene damage can and does thwart this, reducing the efficiency of the process.  The key component to this is something called p53, a critical protein that, if abnormal, may not do its repair job so well.  When that happens, the risk of cancer increases, often dramatically.  In cancer research, abnormal p53 is high on the list of culprits to look out for.

Two groups of scientists, one from the drug company Novartis and one from the Karolinska Institute in Sweden, have published on this.  P53’s thwarting of gene editing is particularly active in pluripotent stem cells, that are some, but not the only, candidate cells to be edited to create treatments.  These cells are also constituent cells of human embryos.  If the CRISPR-Cas9 process is used on these cells, p53 usually kills them off—unless it’s lacking or deficient, in which case it doesn’t, but also in which case it means that the altered cells could themselves become cancers, later on.

This is something that has to be monitored carefully in developing cells as medicines, so to speak, with genetic editing.  One does not want the patient to appear to be healed, only to develop a cancer, or a new cancer, later on.  One certainly would want to know the risk of that before editing an embryo—an unborn human, a future baby if placed in the right environment—to create a gene-edited human being.

Yet, as I’ve written here in the past, it appears that experimentation in heritable gene editing is pressing on.  I’ve argued, and continue to argue, that heritable human gene editing is a line that must not be crossed, that would place too much trust in the providence of the scientists/technologists who are the “actors” exerting power over fellow humans who become “subjects” in a deep sense of the term; that the risks to the subjects are undefinable; that it would enable perception of humans as “engineering projects”; that the gift of life would tend to be replaced by seeking to limit birth to “the people we want”; that the people acted upon are unable to provide consent or know what risks have been chosen for them by others, even before birth.  Rather than press ahead, we in the human race should exercise a “presumption to forbear.”

A counter argument is that, in limited cases where the genetic defect is limited and known, the disease is terrible, treatment alternatives are few or none, that the risks are worth it.  The recent papers seem to expose that line as a bit too facile.  How many embryos created (and destroyed) to develop the technique before “taking it live?”  Could we work things out in animals—monkeys, maybe?  How many generations to alter, create, and follow to be sure that a late risk—such as cancer—does not emerge?  Or maybe our animal rights sensibilities stop us from putting monkeys at such risk—maybe mice will do?

The new papers are dense science.  Frankly, I can grasp the topline story but have trouble digesting all the details.  More sophisticated readers will not be so impaired.  The news report, in the English of the general public, can be read here, the Novartis and Karolinska reports read (but not downloaded or printed) here and here, respectively.

More on physician-assisted suicide

Recently, Dr. Arthur Caplan of NYU, on the Medscape service (subscription required), took on the question of whether physician-assisted suicide (PAS) should be allowed for old folks just because they are old, or because they want to die together.  There have been reports of just that.  While he supports PAS for terminal illness but objects that PAS for “suffering” in general is just too fuzzy, and therefore rejects broadening it.  An accompanying poll of doctors reported:  64-36% against PAS for old age, but 69-31% in favor of PAS for terminal illness.  As some advocates of PAS, like the editors of The Economist, have pointed out in the past, however, this distinction is highly difficult to sustain:  if someone is suffering “intolerably,” who are we to overrule that person’s wishes based on a diagnosis of the cause of said suffering?

Better is to recognize, as Neil Skojdlal noted this week, that real palliative care is not PAS, but is the ethical alternative.  And as Mark McQuain noted this week, changing the terminology confuses, rather than clarifies, the issues.  At least Dr. Lo, whose New England Journal of Medicine editorial Mark reviewed, accepted that not all physicians will accept PAS or be willing to offer it or refer for it.  He seemed to make room for that—unlike some advocates.

In a related item, Hastings center president Mildred Solomon “Calls for ‘Moral Leadership’ to Improve End-of-Life Care.”  In essence, she argues that over-emphasis on “autonomy” can be a way for doctors to abdicate their responsibility, and leave patients out to dry without guidance in end of life decision making.  She argues for a more relational approach, rethinking social supports to provide people with broader help in late life.  Makes sense.  She doesn’t address PAS in the brief piece I’m citing here, but I would certainly leave that out of the list of recommendations.

Coming home to roost

Hoo boy.

Scientists who want to study human embryonic development have heretofore been self-limited by a 14-day rule:  embryos can only be experimented on up to 14 days of age, when they start to develop a nervous system.  This is an attempt to avoid censure for unethical experimentation on human subjects, and is seen as something of a concession, since it does not accept that human life begins at conception.

And, inevitably, they seek work-arounds.  One reported this week by Nature is the creation of human chicken hybrid embryos.  Why would someone want to do this?  (Jokes about the San Diego Chicken are NOT called for here.)

Well, apparently 14 days of embryo age is when critical organization takes place, directed by “organizer cells” that don’t appear before then.  So a group of researches did this:  they took embryonic stem cells (which itself might well require creating and destroying an embryo), and made “embryo-like structures” that had cells that either were, or were just like, these “organizer” cells.  (Apparently these structures were not capable of growing into babies, but even if not, ethical issues remain.)  Then they transplanted these cells into chicken embryos, and watched the resulting hybrid grow, and learned something about how human embryos develop.  They figure this is less of an ethical problem than trying to experiment on a fully human embryo older than 14 days, and that hybrids like this might be able to take the place of experimenting on human embryos to answer many of their questions.

Other scientists disagree with this last statement, and still think they must experiment on fully human embryos to get their answers.

Either way, at a minimum it seems that this work will require creating embryos solely for research, and there is in principle no limit on manipulating the human organism in the name of knowledge.  Work is common on some kinds of “hybrid” animals with human cells, such as immune-deficient mice who have human cells transplanted to reconstitute their immune systems.  But that work usually is done with human cells transplanted into fully-formed mice, which appears different from early, hybrid embryos.

The article describing this work says that the hybrid embryos “didn’t live long enough to hatch.”  Wonder what they would have been like if they had.

A pause for doctor-assisted suicide in California

Readers of this blog probably saw this week’s news that a California judge blocked the state’s End of Life Options Act, the one that legalized doctor-assisted suicide in California.  The law passed after apparent failure in a regular California Legislature session, when its sponsors brought it up again in a special session that was supposed to be about Medicaid funding.  The judge said that inserting the assisted suicide law into that session violated the California state constitution.  So doctor-aided suicide is on hold in the state, for the moment.

Sort of a technicality, and celebration of the decision by folks (like me) who are staunch opponents of assisted suicide is likely to be short-lived.  Supporters will certainly challenge the ruling on appeal, perhaps win, perhaps also bring up the law anew in the Legislature, with (re)passage all but certain.

Legal assisted suicide is still bad policy, and assisting another’s suicide is still unethical.  But efforts against it have to address the attitudes and perspectives of our fellow citizens.  Allowing doctors and others to aid suicide poisons the central calling of medicine to protect life and to address human suffering accordingly.  It risks undermining proper palliative care.  It creates a “duty to kill” that someone has to step up to fill—or to be conscripted to fill, against moral objections that will be rejected as “inconsistent with standards of medical practice.”  It cannot logically be limited to the terminally ill (see Steve Phillips’s May 9 post on this blog) and cannot be reliably limited to those who freely and willingly choose death.  And it opens the idea of “rights” to misuse by those who desire death and to misappropriation by those who have reason to think that someone else should desire death.

Opposing assisted suicide is a longer undertaking, more than one vote or even series of votes, more than a court case with appeals.  It requires the cultivation of moral virtue by ourselves and our posterity.  It requires humble compassion subject to valuing the sanctity of human life.  It requires changed hearts.