Once More into the Breach: An Update on Valproic Acid for Retinal Degeneration
May 23, 2013
by Jennifer Phillips, Ph.D.
ARVO 2013 concluded two weeks ago, but there are still a few more noteworthy stories from ARVO to tell, and this is one of them.
As you know, I am not a clinician. I can make pronouncements about potential treatments in lengthy, opinionated blog posts and not have to face the disappointment and despair that clinicians must see, and that their patients—our readers—must experience. But I know science. That’s the whole reason I’m here, after all; to tell you about the science behind what might eventually find its way into your doctor’s office, and to give you my opinion on whether it’s worth getting excited over. Most of what I choose to write about here really IS worth getting excited over, although I’m sure it’s happening agonizingly slowly. The stories I’m about to lay out for you in the next two posts, however, are more akin to cautionary tales.
If you’ve been following this blog for a few years, you might recall my previous post on a short term study where patients with diagnoses of Autosomal Dominant* forms of RP were given “off label” prescriptions of Valproic Acid (VPA), a drug approved by the FDA for use as an anticonvulsant and for treatment of several mood disorders. The results and ensuing interpretations of this small study were quite controversial at the time, with several clinicians writing in to the journal in which the original article appeared to contest the findings. My scientifically informed opinion of this original study was quite skeptical, although I did subsequently report on further VPA testing in a mouse model of severe retinal degeneration at the ARVO meeting two years ago. To date those results have not been published in a peer-reviewed journal. So what has happened to this story in the ensuing years since the Clemson, et al paper appeared in the British Journal of Ophthalmology?
A Phase II Trial is currently being conducted, at multiple institutions, to evaluate the efficacy of VPA in patients with Autosomal Dominant RP. The clinical trials record hasn’t been updated in over a year, at which time they were still recruiting participants for a one-year course of VPA treatment, and nothing is known about the outcome of this study yet.
The small Clemson, et al. study, along with published data from cell culture, (discussed in my coverage of the original Clemson article linked to above) justified this Phase II Trial for ADRP patients. But, the field of Ophthalmology collectively wondered, would VPA work for other types of RP as well?
A follow-up letter to the Clemson, et al. article was published in 2012, providing some additional data to ponder: Motivated by the Clemson, et al. report, Dr. Robert Sisk of the Cincinnati Eye Institute offered VPA to a subset of his RP patients. Three patients in total, after being advised of the risks, elected to take VPA and submit to additional testing of their visual acuity and visual field. Of the three participants, only one had the family history of Autosomal Dominant RP. The other two had more variable family histories that suggested something other than ADRP. Dr. Sisk’s report describes that all three patients experienced “complications or side effects” that warranted stopping the VPA treatment. These adverse effects (extreme light sensitivity and/or rapid, involuntary movement of the pupils, both of which are known side effects of VPA) went away when the drug treatment ceased, but all three patients experienced a decline in visual acuity—a worsening in their vision that was not restored when they stopped taking the drug. Three patients is a very small sample size, even for a clinical study, but the fact that all of them lost vision during the 5 month treatment period is sobering.
Then, just last month, Dr. Sheena Bhalla and other researchers at the University of Florida College of Medicine in Gainsville, the institution where the patients in the original Clemson, et al. study were treated, published the results of a long-term study on the safety and effectiveness of VPA for treatment of RP. This research was presented at the ARVO 2013 meeting two weeks ago, and I’ve since obtained the paper to read over the results in detail.
Some interesting background given in this article included the information that a previous (unnamed) investigator at the University of Florida College of Medicine had prescribed VPA to a wide range of RP patients. Staffing and policy changes at the institution resulted in asking all patients still taking VPA to stop their dosages until more rigorous studies were completed. However, as there now existed a large-ish group of patients who had been taking the drug for an extended period of time, the researchers who authored this report proceeded with a retrospective review of the patients’ medical histories and vision test results to make what they could of the available data. One other interesting tidbit was that, while the researchers were not able to absolutely confirm that the seven patients from the original Clemson, et al study were included in this chart review, they said it was likely that those patients were indeed among the 31 patients represented in this long-term follow-up. Let’s keep that likelihood in mind as we digest the results.
None of these patients had genetic diagnoses of their disease, but the majority (21) were clinically diagnosed with RP, which apparently included autosomal dominant and autosomal recessive forms. Of the remainder, there were a handful of cone-rod or cone degeneration diseases, three patients with Stargardt’s disease and one with Leber’s Congenital Amaurosis.
So, what did Bhalla and colleagues find in this chart review? First, they looked for any visual tests that had been performed before, during and after the VPA treatment started.
Visual Field tests: Out of the 31 total patients, five of them had a pre-treatment visual field test that could serve as a ‘baseline’ for any changes in the visual field that occurred while taking VPA, or after discontinuing it. Of these five, four of them had a decrease in visual field while on the drug. One had a small increase. Two more patients had received VF testing during and after VPA treatment, and while one showed an improvement in both eyes, the other patient got worse.
Visual Acuity tests: 21 of the patients had received VA (“eye chart”) tests before and during, or after, VPA treatment. In total, the researchers were able to study VA data from 41 eyes (this number should be 42, but one patient had such poor vision in one eye that no test data could be obtained). Overall, 21 eyes had no change in visual acuity, 16 eyes got worse during or after the treatment, and 4 improved.
In addition to looking at vision testing results, they also examined records of the patients’ general heath while taking VPA.
Side Effects: 12 of the 31 patients (nearly 40%) experienced a variety of side effects while on the medication. Of these 12, 10 patients stopped taking the drug altogether because the side effects were so severe, unpleasant or detrimental to their general health.
Conclusions of Bhalla, et al.: Overall, VPA treatment in these 31 patients with various forms of retinal degenerative diseases showed no benefit. In fact, the majority of participants experienced a decline in their vision over the course of their treatments (the average treatment time was about 9 months). While it is impossible to say for sure, based on this study, that the VPA caused vision to get worse in these patients, the role of VPA in retaining or improving vision in a generalized population of RP patients is absolutely not supported by these findings. The authors concluded their discussion with the recommendation that VPA not be prescribed to patients with RP or other types of retinal degeneration until better studies have been conducted to show which types of RP—clinically or genetically—might benefit from VPA.
Follow up discussion: As your resident science reporter, I would have gleaned all of the above from the recently published paper. However, being in attendance for the oral presentation of this study carried the added benefit of listening to the questions and discussions that followed the talk. The first attendant up to the microphone, a clinician I won’t identify, called for “a more flexible approach” to prescribing VPA to RP and RD patients, citing positive patient anecdotes and the lack of any other therapies to offer these patients. I’ll comment on this a bit later, but note here that the presenting author, in response, stood by her data and her conclusion that VPA should not be widely prescribed at this time.
Interestingly, another presenter at the conference (researcher at Oakland University in Michigan) then stood up and made reference to a poster presentation at this meeting that presented data on VPA treatment studies in mouse models of Retinal degeneration. This study hasn’t been published in a peer-reviewed journal yet, but the work presented at the meeting (I went and looked at it later) reported on two mouse strains with mutations in the Pde6b gene. Pde6b plays an important role in the ability of photoreceptor cells to respond to light. Defects in PDE6B in humans (and in mice) lead to photoreceptor degeneration. The two mouse strains, known as RD1 and RD10, respectively, have mutations at different locations within the Pde6b gene. RD1 and RD10 mice were given VPA and their rates of photoreceptor degeneration were assessed after a period of time. RD1 mice on VPA did show an improvement—i.e. measurable decrease in the rate of photoreceptor loss—compared to untreated RD1 mice, but photoreceptors in RD10 mice taking VPA degenerated more quickly than in untreated RD10 mice.
Two different mutations affecting the same gene. Two different responses to VPA. This is pretty astonishing, and, in combination with the new findings of Bhalla and colleagues, and from Sisk, seems to me to underscore the caution issued to clinicians dealing with RP and RD patients by Bhalla and colleagues. There are many genetic causes of retinal degeneration. The complexity is immense, such that, frankly, it would be surprising indeed if one drug could effectively treat so many different problems. The lure of a broad-spectrum cure is strong, but the potential risks of overinterpreting—and overapplying—preliminary results that seem promising warrants equal consideration. Lack of a better treatment option is not a valid reason for offering patients an unproven drug—especially one with the serious side effects of VPA. Worse, while it’s not ironclad evidence, these reports present a clear possibility that VPA can actually worsen vision in some RP/RD patients.
The fact that there are clinicians out there who appear to be ignoring caution and embracing the unlikely advent of a one-size-fits-all treatment for diseases as multifaceted as retinal degenerations is, frankly, disturbing. The number of practitioners this describes might be small—I certainly hope it is—but it’s a number greater than zero, and that gives me pause. I can only imagine how frustrating it must feel to have patients lose their vision while under your care and not be able to offer a single thing to slow down the progression of the disease. It must be a truly horrible and humbling experience that is anathema to a career dedicated to providing health and healing. But remember: First, do no harm. Offering hope, in and of itself, is not harmful, but when that hope is attached to prescribing a drug with known, strong physiological effects and unproven potential as a treatment for this particular disease, it isnot harm-free. Prescribing a drug that might actually worsen the condition you’re trying to treat certainly isn’t either. Please proceed with care, and thanks for listening.
* The descriptor “Autosomal Dominant” refers to the type of chromosome the defective gene is found on and how it is transmitted. Most of our chromosomes are autosomal, the exception being our sex chromosomes—the X’s and Y’s. There are a number of X-linked diseases that manifest mainly in males, so if a disease isn’t autosomal, it’s usually called “X-linked”. The words “dominant” or “recessive” that follow “autosomal” refer to the way in which symptoms of genetic diseases show themselves. In recessive forms of disease, like Usher syndrome, patients need to inherit two defective copies—one from each parent—of a given gene order to show disease symptoms. The parents of these patients are carriers of the disease genes, but don’t show symptoms themselves. In dominant forms of disease, only one bad copy of a gene is required in order to generate disease symptoms, so often in these cases there is a strong family history of the disease as the faulty gene is passed from generation to generation.
