SEOUL, SOUTH KOREA – A major challenge for clinical trials in systemic lupus erythematosus (SLE) is how to get the placebo response rate down low enough that the effectiveness of a drug can actually be seen. Better patient selection may be the key.
Speaking at an international congress on SLE, Joan Merrill, MD, professor of medicine at the University of Oklahoma Health Sciences Center, Oklahoma City, presented on how the heterogeneity of lupus is contributing to the ongoing failure of so many potential therapies in clinical trials.
“It’s a miracle that any drug has been successful in clinical trials,” she told the conference, comparing the few drugs approved for the treatment of lupus with the much larger numbers of approved, targeted biologics that are available for rheumatoid arthritis.
Bianca Nogrady/MDedge News
Dr. Joan Merrill
The problem is that placebo response rates in clinical trials for lupus are high – well over 40% – Dr. Merrill said, and trials aren’t showing a big difference in response rates between the treatment and placebo arms. “If the placebo response is 40%, wouldn’t an effective drug help 80%?” she said. “If it also affects only 40%, does that mean it’s a failed drug?”
Dr. Merrill suggested that better patient selection could be key to achieving lower placebo response rates, which could in turn reveal if and in whom a drug might be effective. “If we could get the placebo response rate down, at least we’d be able to see a little bit better whether the drug is effective, even if it only could work in 50% of the patients,” she said.
Data from research done by the Oklahoma Medical Research Foundation suggested that patients with SLE could be loosely categorized into seven different clusters based on patterns of gene expression in areas such as interferon expression and inflammation pathways.
For example, two of those clusters represented patients with high levels of expression for both interferons and inflammation. “Maybe those are the patients who’d want to be put in a trial for interferon inhibition,” Dr. Merrill said.
This was demonstrated in a trial of type 1 interferon inhibitor anifrolumab (Saphnelo), where patients were sorted into groups according to their level of interferon expression – either high or low – based on expression of certain interferon genes. This revealed that patients in the interferon-high group had a much higher treatment effect than patients in the interferon-low group. But the difference lay in the placebo response.
“The efficacy rate was not that different between the interferon-high and the interferon-low patients,” Dr. Merrill said. “The difference was in the placebo response rate – what they had managed to find was a great marker for sicker patients.”
This phenomenon is not limited to interferon-targeted therapies. Dr. Merrill cited another literature review which looked at subset studies within clinical trials that had delivered disappointing results. This showed consistently that patients who were considered more unwell, by virtue of higher SLE Disease Activity Index (SLEDAI) scores, for example, were more likely to show an effect of treatment.
“You begin to see bigger differences between treatment and placebo because the treatment rate might go up, but mostly because the placebo rate goes down,” she said.
Another issue that could be affecting both placebo and treatment response rates is background medication. “Subset analysis of people on less background drugs was showing lower placebo response rates and better differences between treatments and placebo,” Dr. Merrill said. For example, a recent phase 2 study of anifrolumab took the strategy of actively pursuing tapering of glucocorticoids in patients where that could be done safely. That achieved a lowering of the placebo response rate to the point where a greater difference could be seen between the placebo response and the treatment response rates.
The challenge for clinical trials is therefore to identify which patients to include. “If we could figure out which patients would be the most appropriate [to enroll to fit a particular drug’s mechanism of action], then we could really get ahead of the game,” she said.
The unique problem for lupus clinic trials is the heterogeneity of lupus as a disease, Dr. Merrill said in an interview. “We’re going to have to find combinations of treatments that fit right for each patient, and they won’t necessarily be one size fits all,” she said.
Dr. Merrill said that subset analyses at the phase 2 stage could help identify the patients who responded better to the treatment and could therefore be targeted in phase 3 trials. “Once you take that hypothesis, and if you can establish and validate it in phase 3, now you’ve got yourself a biomarker,” she said.
Richard A. Furie, MD, chief of the division of rheumatology at Northwell Health in New York, agreed that the high placebo response rate was a particular nemesis for researchers involved in lupus clinical trials.
Sara Freeman/MDedge News
Dr. Richard A. Furie
Dr. Furie said it could be that selecting sicker patients is a solution to this, as had been suggested in the subset analysis of the anifrolumab studies – which he was involved in – that identified differences in the response rates between interferon-high and interferon-low patients.
But if that was the case, the challenge would be recruiting enough of any particular subset of patients. For example, relatively few patients in the anifrolumab trial were classified as interferon low.
If the interferon expression levels are a marker for patients who are sicker, that could serve as a way to better select patients for clinical trials, he said. But it would also make it harder to achieve recruitment targets.
“I think the major problem in SLE trials is that patients have inflated activity scores, so you can gain SLEDAI scores with a little alopecia and an oral ulcer,” he said. “You can start eliminating those parameters from counting towards entry, but then as soon as you do that, you’re going to have trouble recruiting.”
Dr. Merrill reported consulting for and receiving research support from a range of pharmaceutical companies including Genentech/Roche, GlaxoSmithKline, Pfizer, Janssen, Bristol-Myers Squibb, AbbVie, and anifrolumab manufacturer AstraZeneca. Dr. Furie reported financial relationships with Genentech/Roche, GlaxoSmithKline, Kezar Life Sciences, Kyverna Therapeutics, and Takeda.