Quick Hits

Charting the course of side effects

An algorithm that tracks adverse conditions in patients gives doctors new insights

By Robert Stoneback, Associate Editor

Deducing possible medicinal side effects could become much easier in the future thanks to an algorithm designed to chart the potential causes of such problems.

Nicole Pratt, PhD, and her team at the University of South Australia’s School of Pharmacy and Medical Sciences created the algorithm.

“We look at the link between starting a new medicine and a hospitalization event and determine whether there is an association between those two events,” said Dr. Pratt, in an interview with the news service The Lead South Australia.¹

In an email interview, Dr. Pratt says that there have been numerous examples of medicines released on the market for which safety data are limited. She cited rofecoxib (Vioxx, Merck), used to treat pain and inflammation, as an example. After quickly gaining a global market of more than 200 million people, Merck withdrew after it was found to cause heart attacks. By that time, however, at least 80,000 U.S. patients had experienced myocardial infarctions linked to the medication.

“Rapid methods for adverse drug event detection will help avoid catastrophes such as this and will mean that safety issues are detected early and harms avoided,” says Dr. Pratt.

Finding the link between medicine and illness

In a study published in Ophthalmic Epidemiology, Dr. Pratt and her team used the algorithm to analyze 25,984 veterans of the Australian military given timolol eyedrops for glaucoma treatment. The study investigated if the medicine could aggravate comorbidities. Among the results collected by Dr. Pratt’s team were that 88% of veterans with an airway disease were dispensed timolol eyedrops with the potential to aggravate this condition. Their research also showed an increased likelihood of hospitalization for bradycardia after initiation of timolol treatment.²

These data, created from an algorithm developed by Nicole Pratt, PhD and Libby Roughead, PhD, shows a link between rofecoxib and myocardial infarctions. More people experienced a heart attack after starting the medicine, suggesting an increased risk with the treatment.

Subsequent studies used a variation on the algorithm to examine the risk of stroke associated with the use of ranibizumab (Lucentis, Genetech) to treat macular degeneration and the risk of bradycardia associated with the use of timolol for glaucoma.

Analysis from Dr. Pratt’s team suggested a 90% increased risk of ischemic stroke 31 to 60 days following ranibizumab treatment, based on patient data of about 300,000 eligible study candidates treated from 2007 to 2013.³ Not all of the study candidates were treated with ranibizumab.

The timolol study found the risk of bradycardia was significantly increased 31 to 180 days after timolol initiation. Drawing upon data of Australian veterans, Dr. Pratt’s team found 6,373 hospitalized for bradycardia at some point between 2003 and 2009; of those, 267 had been exposed to timolol. They concluded the medicine poses a potential adverse event for the ailment.⁴

Completing months of work in minutes

The algorithm is free and available upon request, says Dr. Pratt, and can be obtained by emailing her at Nicole.Pratt@unisa.edu.au. It is designed for use in large observational data sets, such as pharmaceutical dispensing datasets.

A 2015 report from the U.S. Department of Veterans Affairs listed 285,000 American veterans as having glaucoma.⁵

Dr. Pratt developed the code with the support of Libby Roughead, PhD, at the Quality Use of Medicines and Pharmacy Research Centre of the University of South Australia. The principles of the algorithm had been described previously, but were not implemented on a large scale until Dr. Pratt and Dr. Roughead developed it in 2008. They originally made it to help improve the efficacy of medicines administered to Australian veterans.

The algorithm can generate results, identifying associations between medicines and adverse events, in minutes, as opposed to the six-month timespan required in a more traditional study to locate such links, says Dr. Pratt.

It has been distributed to multiple countries for use in their health datasets and has been used to cover populations of more than 200 million people.

“We have demonstrated the potential for global surveillance. Our big successes have been the uptake of the tool by researchers and some regulators, including now in Korea, Japan, Hong Kong, Taiwan, China, Sweden and Canada,” Dr. Pratt says. Much of the implementation has been accomplished through collaboration with the Asian Pharmacoepidemiology Network (AsPEN).

The next phase for Dr. Pratt involves improving the algorithm to be a data-mining tool, so that it can rapidly analyze large data sources and detect unknown safety issues.

And, work has begun to use the program to determine the point in time an adverse effect is most likely to occur, such as immediately after medication is administered or after longer-term exposure. This feature of the algorithm has been successfully piloted in Korea, where it was used to examine the risk of cardiovascular events related to attention deficit disorder medication.

“This information is important for clinicians, patients and parents so that these medicines can be used safely, particularly in children,” says Dr. Pratt. OM

REFERENCES

OIS spotlights ophthalmic therapy innovations

Nearly 1,000 hear about changing eye-care landscape

Burgeoning interest in new ophthalmology treatments — ranging from drugs to devices as well as drug delivery and digital — drew nearly 1,000 attendees to November’s Ophthalmology Innovation Summit in Las Vegas. Attendees of the event watched numerous early-stage companies present their new proposed methods of battling eye disease or restoring vision.

The afternoon discussions examined innovation at large companies such as Alcon and Allergan. Brent Saunders, who will be in second in command when Allergan and Pfizer officially unite this year, delivered a keynote address highlighting the tightening relationship between large ophthalmology companies and start-ups developing new treatments.

“We operate in a vibrant ecosystem,” he told the 800-plus attendees at OIS.

Massive amounts of money are being invested into eye care, he pointed out. What it means is that ideas will reach fruition. Mr. Saunders noted that in 1998, 62% of the ideas that reached fruition presumably came from Big Pharma; in 2013, only 22% came from Big Pharma. Fifty percent came from biotech and start-ups.

“Be agnostic as to where great ideas come from,” he said. “Science is advancing. We are in an industry that used to be about chemistry… but with the human genome” it’s all changed. Now, it’s a matter of, “How do you turn that biology into a great product?”

Along with the daylong presentations and talks, attendees heard about the impressive financial health of the ophthalmology sector, covering the clinical, regulatory and financial fronts. For more information visit OIS.net. OM

Dry eye roundup

A recent discussion in Toronto on effective diagnostic and treatment tools for DED.

By Louise Gagnon

Ophthalmologists should use numerous tools, both objective and subjective, in diagnosing dry eye disease (DED) and evaluating the severity of it, according to Anthony Aldave, MD, professor of Ophthalmology, chief of the Cornea and Uveitis division and director of the Corneal Genetics Laboratory, Jules Stein Eye Institute, Los Angeles, Calif.

Speaking at the annual Walter Wright Symposium in Toronto held in December, Dr. Aldave reviewed current screening tools and treatment options for DED.

He told an audience of fellow ophthalmologists (mostly Canadian) that they can be confident in their use of the Canadian Dry Eye Assessment (CDEA) to evaluate DED. “It’s a good inventory of patient symptoms and how it [DED] impacts daily life,” said Dr. Aldave. This assessment quantifies severity of ocular surface disease, which determines a treatment algorithm.

CDEA is a modified, condensed-version of the Ocular Surface Disease Index (OSDI), another tool that should be integrated in the evaluation of DED because it helps to determine the severity of a patient’s DED, said Dr. Aldave, noting the OSDI score indicates how the condition adversely impacts daily activities.

“The OSDI score is a validated questionnaire that also saves ‘chair time’ because it is administered before you see the patient,” he said. “It may tell us that the patient is, for example, having trouble driving with the air conditioning on [in the car]. It’s much faster than asking patients open-ended questions about their dry eye syndrome.”

In regard to other diagnostic tools, Dr. Aldave stressed that he found no value in using Schirmer’s test to help arrive at a diagnosis of dry eye. “There are so many confounders that can influence the result,” he said. “It is a poor objective measure of dry eye syndrome.”

The treatments of dry eye include autologous serum and cyclosporine, but he pointed to a meta-analysis to support the latter over the former. This included four randomized, controlled trials that compared autologous serum to artificial tears and concluded that the benefit of using autologous serum, designed to be a substitute for lubrication, in treating dry eye is inconsistent.¹

In contrast, a systematic review and meta-analysis looking at the impact of topical cyclosporine A in the treatment of dry eye disease found the topical agent could be effective, particularly in the presence of conjunctival injury.²

Oral omega-3 fatty acids should be part of the armamentarium in treating dry eye, said Dr. Aldave. He pointed to data from a randomized, double-blind clinical trial where patients received placebo or 180 mg of EPA and 120 mg DHA twice daily for one month. The data showed a significant difference in tear break-up time between the actively treated patients and patients who received placebo, as well as greater improvements for the treatment group on secondary outcomes.³

Dr. Aldave noted that conjunctival staining with rose bengal can result in staining of the ocular surface in conditions other than dry eye such as conjunctivocholasis. “It [conjunctivocholasis] can cause symptoms consistent with dry eye syndrome,” he said.

Similarly, in corneal staining with fluorescein, clinicians should ensure the pattern is consistent with dry eye syndrome and not another condition, cautioned Dr. Aldave. “It could be limbal stem cell deficiency.” OM

REFERENCES