by Jennifer Phillips, Ph.D.
In the summer of 2010, a small observational study of the effects of Valproic Acid was published online in the British Journal of Ophthalmology1. This study reported on seven patients with autosomal dominant retinitis pigmentosa (ADRP) who were given daily doses of a drug called (in its generic form) Valproic Acid, or VPA. According to the authors, these patients showed some improvements to their visual field and to their visual acuity after a few months on the drug. The authors used this result, coupled with other, in vitro studies by the same research group, to justify approval for a larger clinical trial to come that will test the effects of Valproic Acid on visual function in ADRP patients. There has been a great deal of interest, and more than a smidge of controversy, about this story in the vision research community, so I’ve put together a ‘FAQ’ about the work that I hope will be of interest to our readers.
1. What is Valproic Acid?
It is a fatty acid, chemically similar—but not identical—to the active ingredient found in the herb valerian. It has been FDA approved since 1976 for the treatment of epilepsy, and is also approved for use in treating schizophrenia, bipolar disorder and migraine. More information can be found here.
2. Why did anyone think such a drug would be useful for treating Retinitis Pigmentosa?
Valproic acid is an effective treatment for the above neurological conditions because of its known effects on neurobiology. It is known to impact brain function by inhibiting certain neurotransmitters and also has a role in modifying gene expression by inhibiting an enzyme that regulates the conformation of DNA molecules. More recent work with this compound has shown that it and other similar molecules also have an effect on reducing the inflammation that results when nerve cells are damaged or ailing. Finally, the authors of this study report that they conducted experiments in cell culture to test the ability of VPA to associate with rhodopsin, a light sensitive pigment found in the outer segment of rod photoreceptors. Some forms of RP are caused by mutations in opsin genes, resulting in poorly formed or shaped proteins that can’t get to the right places in the cell because they are unable to bind to the chaperone proteins responsible for delivering the opsins to their proper position within the photoreceptor. These researchers previously published a report testing whether various small molecules could act as substitute chaperones for these poorly folded proteins2. In the current clinical report, the authors refer to their unpublished result that VPA was similarly tested and found to be an effective chaperone.
3. How was this study conducted?
The authors report that seven patients at the University of Florida Ophthalmology department clinic were treated “off-label” (meaning that the condition—RP—for which they were receiving the drug is not an FDA approved use for this drug) with VPA and subjected to ‘before and after’ visual function tests during a brief period of between 2 and 6 months. All patients in the study were being seen for an unspecified autosomal dominant form of retinitis pigmentosa (ADRP), so there were no Usher patients included. Mutations in any one of 17 known genes can lead to ADRP, and it wasn’t clear how many different types of ADRP were represented in this study. The authors define this as a ‘retrospective study’.
4. What is a retrospective study?
This type of study involves reviewing previously recorded patient histories and looking for correlations between patient disease profiles and some other factor—some difference in the diet, or behavior, or genetic background that may change the outcome or progression of a particular disease. This is in contrast to a ‘prospective study’, where a research question is posed and patients with a particular condition are given treatment, monitored for changes, and, usually, compared to another group of patients with the same condition who did not receive that treatment.
In general, prospective studies are considered to be more rigorous, because they can be carefully thought out and designed to include all the appropriate controls and statistical bells and whistles that make the findings more believable. However, prospective studies aren’t always possible, and retrospective studies can often yield useful health data. At a minimum, a retrospective study can give one a jumping off point for new investigations. If, for instance, a correlation is noted between consuming Factor X and having a milder manifestation of disease Q, more formal studies of the effects of X on Q could be justified.
That said, I question whether this was truly a retrospective study in the usual sense of the term. If a chart review of ADRP patients had shown that those who had been prescribed VPA for some other reason (unrelated to the RP) had a slower decline in vision, than ADRP patients of a similar age and disease profile who were not taking VPA, that would be a retrospective study. In this case, however, someone decided, ahead of time, to give these particular ADRP patients a dose of VPA for something other than the conditions it is normally used to treat. Indeed, the “off-label” dosage prescribed to all seven patients is reduced by 25-50% from the dose usually prescribed to treat the neurological conditions for which VPA is approved. The authors spend some time justifying how they chose the starting dose, and explain that in two of the seven cases they decided to increase the dose by 25% at some point during the testing period. Baseline visual measurements were taken, and additional measurements were collected from each patient at a follow-up visit. Then someone perused the patient charts, extracted the relevant data, and wrote up the paper. So, either there is some dramatically different definition of ‘retrospective’ in play here, or there is some information missing.
Another odd thing about the origins of these data is revealed when the authors state: “The data reviewed from these patient records were not collected originally with the intent to study.” Why were ADRP patients given off-label doses of VPA, if not to study the effects? The authors never explain this, and although in and of itself it doesn’t affect the outcome or impact of the results to any great extent, I find it unusual, to say the least.
5. What were the results of the study?
Briefly, data were collected from 13 eyes from 7 individuals with ADRP. The authors explain that data from the 14th eye were not included in the analysis due to ‘poor quality’. When the results were averaged together, there was a small improvement in visual acuity, comparable to going from about 20/50 vision to about 20/30 vision. There was also an increase of about 10% reported in the average visual field areas of these patients. Considering the 13 eyes of these 7 patients individually, 9 eyes were judged to have increased visual field areas, 2 had no change in visual field, and 2 got worse, with diminished visual field areas between the initial exam and follow-up. The results per individual eye on the ‘eyechart’ test that measures visual acuity were a mixture of improved (7 eyes) and no change (6 eyes). These changes were judged to be statistically significant, but because there was no control group, the ‘controls’ used to compare the treatment group to were calculated as ‘no change’ in visual field area or visual acuity.
6. What is the significance of these results?
Taken at face value, the data are interesting but inconclusive. There are a number of factors that limit what can be interpreted from this study. The small size of the study is a big consideration—seven subjects is a miniscule sample size, and given the variation in the genetic basis of their disease, the difference in their ages (between 16 and 56 years), the short duration of the study, and the differences in the degree of RP they were experiencing during the time of the study (some quite a bit more advanced than others), it is challenging to extract a clean signal from all that noise. These differences tend to average out with more individual data points, but in a study of this size, there are ambiguities all over the place. In fairness, the authors do acknowledge the sample size and study duration as limitations in the paper.
7. Will further studies be carried out to confirm these results?
The authors state that further pre-clinical studies have already been carried out, and/or are underway; however, all remain unpublished at this writing. Several studies with rodent models of RP and cell culture were presented at the ARVO meeting last year, but are not yet published in peer-reviewed journals. Approval was granted for a larger Phase II Clinical trial on the basis of these collective results, however. ADRP patients are currently being recruited for this prospective study in which about 100 ADRP patients will receive either VPA or a placebo for one year while their vision is monitored.
8. Can VPA be used to treat RP due to Usher syndrome?
At this point, there is not enough information to justify such treatment. We know very little about how VPA might be working in the ADRP patients, and it will be impossible to extrapolate that potential benefit to other forms of RP until we know more.
9. Should Usher patients start taking VPA, just in case?
Short answer: No. Most importantly, no one should ever take any medication for anything without consulting a qualified physician. No Usher patient should ever take, or be given, a drug or supplement without consulting a doctor who specializes in Usher syndrome.
10. But what’s the harm? VPA is an FDA-approved drug—so isn’t it safe?
One of the strong ‘selling points’ of this particular study was that it explored a new use for an already FDA-vetted pharmaceutical. That saves both money and time that would have been spent on developing a novel drug AND safety testing it in pre-clinical and Phase I clinical trials. We have over 30 years of data on the safety profile of VPA in humans, so any future trials using this drug, including the one now being organized for ADRP patients, can start at Phase II. But does that mean that VPA is ‘safe’? In the sense that it is generally well tolerated, not toxic when taken at prescribed doses, and improves the specific conditions for which it is currently approved, yes. BUT—and this is a really big BUT—safety is relative. VPA has well documented and potentially serious side effects, which must be weighed against any potential benefit. Furthermore, this is a drug with known neurological effects. A choice to take VPA for RP is way, WAY different than just choosing to eat more paella or supplement with DHA.
11. What is the “take-home” message from this study?
That a very small and limited case study has generated some interesting results that bear following up. The authors have been subjected to some criticism by other clinician scientists, who wrote letters to the journal 3,4opposing the strength of the conclusions, based on the available evidence. The authors wrote nearly as much in response to these letters as they did in the original article5,6, and, much like the Vitamin supplementation for RP story, the controversy will likely roil for some time, at least until more robust data are available.
Personally, I am interpreting these results with even more caution than I applied to the supplementation story, given the limitations of the study described above, the lack of peer-reviewed supporting data, and some level of unease about how the study was conducted. I am also, as always, mightily chastened by the findings of John Ioannidis, who, in addition to his work on the impact of translational research on the discoveries of new treatments, has also used impeccable statistical analysis to show that ”for most study designs and settings, it is more likely for a research claim to be false than true”.
It is unquestionably beneficial to research potential new uses for existing drugs. I regard this study and the proposed trial with interest, and I will hope that all due rigor will be applied to the collection and interpretation of the data. It is a reasonable expectation that a bigger study with the proper controls, in addition to some results from animal RP models, can tell us what we need to know about if—and how—VPA works in RP patients, and what kinds of RP it might benefit.
The bottom line is that there is reason to hope—not necessarily for this particular drug to be THE ONE for Usher syndrome, but because lots of similar research is happening every day, and every one of these studies brings us closer to finding a treatment that will work, even if part of the progress results from the process of elimination. Watch this space for further updates.
1Clemson, M. et al., 2011. Therapeutic potential of valproic acid for retinitis pigmentosa. British Journal of Ophthalmology 95 (1): 89-93. (published online July 20, 2010)
2Noorwez, SM et al., 2008. A high-throughput screening method for small molecule pharmacologic chaperones of misfolded rhodopsin. Investigative Ophthalmology and Vision Science 49: 3224-30.
3vanSchooneveld, MJ et al., 2011. The conclusions of Clemson et al. concerning valproic acid are premature. British Journal of Ophthalmology 95 (1): 153-154. (published online October 22, 2010).
4Sandberg, MA et al., 2011. Lack of scientific rationale for use of valproic acid for retinitis pigmentosa. British Journal of Ophthalmology (published online December 3, 2010).
5Clemson, M et al., 2011. Author’s response [to vanSchooneveld, et al.]. British Journal of Opthalmology 95(1): 153-154. (published online October 22, 2010).
6Tzekov et al., 2011. Author’s response [to Sandberg, et al.]. British Journal of Ophthalmology (published online January 9, 2011).