As of today, no drug, supplement or lifestyle measure has been shown to delay onset of the genetic prion diseases – familial Creutzfeldt-Jakob Disease (fCJD), Fatal Familial Insomnia (FFI) and Gerstmann Straussler Scheinker syndrome (GSS). Those of us carrying these genetic prion diseases in our families would like very much to have some way of delaying onset. Even treatments that couldn’t delay onset forever would buy us more time to find a cure. This post will briefly review past efforts, current thinking and future research directions. I am not a physician and this post does not include medical advice.
In this recent post I reviewed the list of drugs that have been attempted as therapeutics in symptomatic prion disease patients. The short answer is that none of them worked. Most of these drugs are covered in a published review [Stewart 2008], who summarizes by saying that the only positive result reported for any drug was that flupirtine appeared to help preserve cognitive function in sporadic CJD patients, but did not extend survival at all [Otto 2004]. The results of some more formal clinical trials for quinacrine [Collinge 2009] and pentosan polysulfate [Bone 2008, Tsuboi 2009] have come out since Stewart’s review was published, so here are quick summaries of those drugs.
Quinacrine reduces prion formation in cell culture models [Doh-Ura 2000, Korth 2001], but no study has ever shown quinacrine to reduce prion formation, extend survival or delay onset of prion disease in any animal model [Collins 2002, Barret 2003, Doh-Ura 2004]. All reports on human use declared no therapeutic effect [Furukawa 2002, Benito-Leon 2004, Haik 2004, Collinge 2009]. It was later shown that even in cell culture, quinacrine’s effects are only transient, apparently because it merely selects for certain strains of prions [Ghaemmagami & Ahn 2009].
Pentosan polysulfate is highly effective in mice, at least against some prion strains and if administered early in the disease course [Doh-Ura 2004]. It cannot cross the blood-brain barrier, and so in both mouse and human trials it was infused directly into the brain with an implanted device, a risky procedure which resulted in complications in several patients. It did not demonstrate any therapeutic effect in clinical trials in sporadic and variant CJD patients in the U.K. and Japan [Bone 2008, Tsuboi 2009].
The only clinical trial that is ongoing as of May 2013 is for doxycycline. Doxycycline and the related compound tetracycline delayed disease in certain animal models [Forloni 2002, Luigi 2008] and doxycycline was used in symptomatic sCJD patients, where it showed a small but possibly statistically significant effect. The results of those human trials have never been formally published as a study, but some results have been presented at conferences or mentioned in other articles [Luigi 2008, Stewart 2008, Zerr 2009 (ft), see also tetracyclines post]. Based on the animal results and the marginal results from symptomatic patients, doxycycline moved on to a clinical trial in 11 asymptomatic FFI carriers. This clinical trial is not expected to report results until 2022.
That date – 2022 – reflects the extreme difficulty of running clinical trials for genetic prion diseases. The onset of these diseases is late and highly variable, and the number of patients is small. Prion diseases are rare even as a whole – ~250 deaths/year in the U.S. for instance [Holman 2010] – and then just ~15% of prion disease patients have genetic mutations [Appleby & Lyketsos 2011]. And then they’re not all the same mutations: you have E200K and D178N and P102L and so on – so any study will need to either control for this difference or only enlist people with a particular mutation (D178N in the doxycycline study). Finally, because patients are perfectly healthy up until onset and will be taking the drug for years or decades, it’s only realistic to consider approved drugs with excellent safety profiles. For drugs with nastier side effects (rapamycin) or unapproved experimental compounds (anle138b), the risks would probably outweigh any benefits in genetic carriers.
The gold standard of biomedical research is the double-blind, placebo-controlled clinical study in a large number of people. That’s probably never going to happen in genetic prion diseases. There are too few patients, and too few would agree to be randomized and maybe get a drug and maybe get a placebo. And even for the tiny (11 people), non-double-blind, non-placebo-controlled clinical trial that is underway for doxycycline, it’s a tall order to ask anyone to wait until 2022 to get answers.
This leads naturally to the possibility of off-label use - of doctors prescribing to patients a drug which is already approved for other uses, even though clinical evidence is not yet in place to demonstrate the drug’s efficacy in prion diseases.
Informally, the National Prion Disease Pathology Surveillance Center in Cleveland tells patients to take doxycycline just in case it turns out to work. When my wife Sonia and I received her genetic test results from the Surveillance Center in December 2011, indicating she carries the D178N mutation, the manager of the Center at that time (who was not an MD) recommended she take doxycycline.
I have asked around a bit, and though no one agreed to be quoted for this blog post, physicians appear to be divided on the issue of whether to prescribe doxycycline to genetic prion disease carriers. On one hand, doxycycline has a rather low rate of adverse events [Smith & Leyden 2005] and an overall strong safety record, even in long-term use for rosacea [Valentin 2009]. Other physicians, however, feel that drugs shouldn’t be prescribed for experimental uses outside of a clinical trial setting.
From a research standpoint, the evidence for doxycycline’s efficacy is not as strong as the evidence for many other drugs. Doxycycline was able to delay onset of prion disease in infected mice [Luigi 2008], but there are a few caveats to this study. First, the mice were either infected peripherally (“peripherally” meaning “not in the brain”) and treated peripherally, or infected in the brain and treated with direct injections into the brain. So Luigi’s results don’t establish that enough doxycycline crosses the blood-brain barrier to be effective if taken orally for a disease in the brain. Second, Luigi treated the mice starting on the day they were infected. Lots of drugs have shown efficacy if administered on the same day of inoculation, when the prion infection has yet to take hold. It is much harder for drugs to be effective once infection has taken hold. Third, the closely related compound minocycline was shown to have no effect when administered to mice late in prion infection [Riemer 2008 (ft)]. As for the human results from doxycycline, the results have never been published but I have heard it argued that the extension of survival in symptomatic patients treated with doxycycline was small enough that it could potentially be explained by the increased medical attention paid to study participants, or the antibiotic effects of doxycycline helping to prevent chest infections.
But the caveats relating to efficacy (or lack thereof) late in disease may not apply here. Carriers who know their genetic status may actually be able to start taking a drug years or decades before any prion infection takes place, so we cannot rule out the possibility that drugs effective early in disease course in mice (which might include doxycycline) could be helpful.
Still, it is informative to note that several other drugs have been effective in more stringent mouse models – mice infected intracerebrally and treated orally or peripherally late in the disease course. Notably, statins – which also have an excellent long-term safety record [Law & Rudnicka 2006] – have consistently been able to extend survival or delay onset by ~10% in mouse models [Mok 2006, Kempster 2007, Haviv 2008 (ft), Vetrugno 2009 (ft)]. Consistent with that, there is some unpublished evidence that high LDLs (‘bad cholesterol’) are correlated with shorter survival in sporadic CJD.
The fact that the Surveillance Center recommends doxycycline, and not statins, to genetic carriers may owe to a few reasons. First, a mechanism of action has been suggested – tetracyclines such as doxycycline are suspected to bind PrP and inhibit its misfolding and/or aggregation [Tagliavini 2000] – whereas the mechanism of action of statins in prion disease is still mysterious. Second, though results are unpublished and effects are marginal at best, the fact that doxycycline may have extended survival slightly in symptomatic patients provides a layer of human evidence where statins only have mouse evidence. Third, the fact that doxycycline is in a current clinical trial for FFI may make its off-label use seem more justified.
In any event, neither doxycycline nor any other drug has been shown to delay onset of prion disease in patients with genetic mutations, so any off-label use at this point is highly speculative. Indeed, doxycycline (and statins) have never even been tested in genetic mouse models of prion disease. Most drug studies in animals are done in mice infected with prions. But mouse models are now available for several genetic prion disease mutations including D178N [Jackson 2009], E200K [Friedman-Levi 2011] and A117V [Yang 2009].
That’s why one of our foremost research goals at Prion Alliance is to take several of the compounds that have shown some efficacy in prion-infected mice and test them in genetic mouse models of prion disease. We’re currently working on gathering all the data on previous mouse studies in order to objectively pick the best candidates.
In choosing the best candidates for further animal studies, we’ll need to consider not only which compounds are most potentially effective, but also which are most realistic to become treatments for genetic carriers. To date, the only drug that has extended survival in a genetic mouse model of prion disease is rapamycin [Cortes 2012], but rapamycin is an immunosuppressant with very serious complications including a long-term increase in cancer, which makes it a very poor candidate for anyone to take for years or decades in the hopes of delaying prion disease. Experimental new molecules like anle138b [Wagner 2013 (ft)] are not available commercially and their side effects are as yet unknown, which means that their safety and efficacy will need to be shown first in clinical trials in symptomatic patients before any doctor will consider prescribing them to genetic carriers – though it would certainly still be of interest to know whether they are effective in genetic prion diseases. Meanwhile, even delays in onset of a few percent could correspond to meaningful years of a person’s life, so compounds of small or arguable effect – curcumin [Riemer 2008 (ft)] or astemizole [Karapetyan & Sferrazza 2013] for example – shouldn’t be ruled out automatically, but rather evaluated on the basis of efficacy as well as safety and availability.
Sometime in the next few months, expect a blog post reviewing the different studies to date and evaluating which compounds might be most worthy of further study. For now, unfortunately, little effort has gone into studying potential treatments in genetic mouse models, so we don’t have much animal data to go on. And no treatment has been shown to delay the onset of genetic prion disease in humans.
As mentioned above, I’m not a physician and nothing in this post is medical advice. As a researcher and as someone affected personally by genetic prion disease, I wrote this post to gather the available information into one quick reference. If readers would like to discuss their questions with a physician, Dr. Brian Appleby of the Cleveland Clinic has agreed to be contacted by any genetic prion disease carriers seeking information on this subject (). Still, note that only your own physician will be able to provide you with medical advice.