DREAM: Fish oil works
Treatment including fish oil reduces dry eye symptoms.* **
The jury has returned, and the verdict is in: If you treat dry eye disease with a protocol that includes purified fish oil, your patients will feel better. Seriously. It says so right there in the results section of the DREAM study. Patients with severe DED who were treated with all manner of DED protocols and were also taking 3,000 mg of fish oil per day had on average a 31% decrease in their Ocular Surface Disease Index scores; 61% of them felt better. QED.
I know what you are thinking. That is not what you read after the DREAM results were published. For some unfathomable reason, the “writing” authors of DREAM have insisted on telling us that fish oil does not work. They have persisted in a contention that simply cannot be supported by their study. DREAM did not find a statistically significant difference in the abatement of DED symptoms between uncontrolled treated groups in which one took fish oil and the other did not. From this they have extrapolated that fish oil has no effect on DED symptoms.
To say that this is a reach is an understatement. The only proper conclusion from the DREAM data available is that it is impossible to tease out any effect, or lack thereof, from the use of fish oil or artificial tears or Restasis (cyclosporine ophthalmic emulsion 0.05%, Allergan) or other anti-inflammatory medications when any or all of them are assembled into a polymodal intervention for the relief of DED symptoms. How much more helpful would this study have been if the authors had come to the more obvious and optimistic conclusion available to them: If you make a serious effort to treat the symptoms of DED patients, you can make them feel better.
That is pretty much all DREAM says.
Fish oil is big in the news right now. I mean, really big. There is a ton of money in play, and most of it is not in eye care. The real money and the real buzz with fish oil are in “real” medicine, such as cardiology, where patients die of their diseases. Over the years, there have been numerous studies that have found no effect from the use of fish oil on any meaningful outcomes in the arena of heart health. Still, there has been enough theoretical support that additional ideas have been pursued.
Two large, complex trials reported their results in the fall. Both of them are instructive as we in eye care look at fish oil for DED. VITAL and REDUCE-IT both looked at the use of highly purified EPA as an adjunctive treatment for otherwise fully treated patients at risk for cardiac events (including death). Disclosure: I am not a cardiologist, I do not play one on TV, and I have not had a recent stay in a Holiday Inn Express.
VITAL used a dose of 1,000 mg of EPA per day. As I read the reported data, there was no measurable effect, positive or negative, on cardiac outcomes between the treated and untreated groups. REDUCE-IT looked at patients who had elevated triglycerides and an elevated risk for adverse cardiac events despite maximized medical treatment (eg, effective statin dosing). Study patients were given purified EPA 4,000 mg per day (the study was funded by Amarin, which hopes to sell this). The results were stunning. REDUCE-IT reported a 25% decrease in cardiac deaths over the course of 7 years in the study group.
You read that correctly. Taking 4,000 mg of purified EPA resulted in 25% fewer deaths from cardiac events.
Let us apply some of the insights from not only REDUCE-IT and VITAL, but also what we have learned from prior studies of omega-3 fatty acid supplementation. Those who designed DREAM have made an attempt to investigate the effect of fish oil in “real-world” circumstances. I would say that they were successful in doing so. This is similar to what was evaluated in REDUCE-IT in that every patient was maximally treated according to 2007 protocol. It is quite dissimilar, however, in that DED protocols are not in any way, shape or form as tightly conforming to a consensus as those in cardiac/serum lipid care. In the real world of DED, patients and doctors jump all over the place, adding, subtracting and switching all manner of treatment. REDUCE-IT was tightly controlled with only prespecified MD-directed changes allowed. DREAM had no treatment controls.
Two well-designed studies in eye care looked at the effect of omega-3 supplementation in the treatment of DED symptoms. Sheppard and colleagues evaluated HydroEye in a randomized, double-masked, tightly controlled study in which the study group differed from the control only in that it took HydroEye (funded by ScienceBased Health). Neither group used anything other than Refresh Tears (Allergan) for DED. Similar to DREAM, the individuals had severe DED. At the 6-month period, treated subjects had an approximately 50% decrease in OSDI scores, an improvement that was statistically significant in comparison with the control.
More recently, Epitropoulos and colleagues evaluated the short-term effects of a highly purified version of fish oil that is marketed by PRN (which funded the study). Similar to Sheppard, the only other allowable DED treatment was artificial tears. Over the 12 weeks of the study, treated subjects had a 54% decrease in OSDI scores, also statistically significant. In addition, the study found significant improvements in other measures such as tear osmolarity and tear breakup times.
Where DREAM fails is where Epitropoulos and Sheppard shine, in the controls of the studies. In order to show a significant effect from any one factor, it is necessary to do as much as you can to isolate it. There are two ways to do this: take out as many variables as possible or evaluate a very large group over a very large period of time. When you choose the former, you are open to the criticism that your results are too far removed from the reality of everyday care. When you attempt to reproduce (or measure) a “real-world” scenario, it is necessary to look at thousands of subjects. Think the Framingham Heart Study or the Women’s Health Study, both of which followed tens of thousands of individuals.
DREAM does show a separation between study and control subjects. It simply lacks the power to declare that it is significant or not.
Why was there such a gleeful declaration that fish oil does not have an effect on DED symptoms? Heck if I know. All three studies I have mentioned are really too different from one another to be directly compared. Two look at severe DED and one evaluates moderate disease. DREAM, REDUCE-IT, Epitropoulos and Sheppard all used different oils as a placebo. One may have improved symptoms (DREAM: olive oil), and one may have had a negative effect on mortality (REDUCE-IT: mineral oil). Two good studies were published as DREAM was taking shape. Why not use one of the placebos from them in DREAM? Again, heck if I know.
So what next? This part I know. Let one of our self-congratulatory professional fundraising organizations sponsor, design and fund a study that looks at mild, moderate and severe DED patients treated with fish oil or other omega-3 supplementation. Use inert oils (sunflower, linoleic) as placebo. Enroll subjects using simple criteria (OSDI). Measure things that are largely agreed upon in the wider DED-treating community (OSDI, osmolarity, tear breakup time, perhaps MMP-9). Make it ginormous like the Framingham Heart Study if you really want to look at “real-world” results. DREAM big.
For now, the only thing we know is that treating patients with severe DED symptoms with a regimen that includes highly purified fish oil or black currant seed oil reduces their symptoms in a significant and meaningful way. Prescribe one or the other for every DED patient.
*Unlike pretty much everywhere else in the DED world, I do not have any conflicts of interest here.
**There is too much detail in DREAM and the other studies for a granular review. Look for upcoming blog posts for deeper insights.
- Epitropoulos AT, et al. Cornea. 2016;doi:10.1097/ICO.0000000000000940.
- Sheppard JD Jr, et al. Cornea. 2013;doi:10.1097/ICO.0b013e318299549c.
- For more information:
- Darrell E. White, MD, can be reached at SkyVision Centers, 2237 Crocker Road, Suite 100, Westlake, OH 44145; email: firstname.lastname@example.org.
Disclosure: White reports he is a consultant to Allergan, Shire, Sun, Kala, Ocular Science, Rendia, TearLab, Eyevance and Omeros; is a speaker for Shire, Allergan, Omeros and Sun; and has an ownership interest in Ocular Science and Eyevance.