Dr. Merrill noted the belimumab trial was very large, upping the statistical power. This helped the researchers get statistically significant findings, even if the treatment response rates were only a little higher than the placebo responses when all lupus patients were considered as a group and not by disease subsets.
Impact of Polypharmacy
Dr. Merrill explained that polypharmacy is another factor that can make it challenging to get positive results in lupus trials. She pointed out that patients come to lupus trials on a variety of background immunosuppressants, but these can impact the response to the studied treatment. For example, whether a patient is on background methotrexate or azathioprine impacts the expression of different lupus genes, altering their immune profiles. This may affect response to a trial agent.
To address the issue of polypharmacy, Dr. Merrill described her biomarkers of lupus disease (BOLD) study, which used an immunosuppressant withdrawal strategy.2 “You take away all background medications, except maybe minimal amounts of steroid or hydroxychloroquine, and you look at time to flare,” she said. When patients flare, they are treated immediately and recorded as non-responders. Executed correctly and by excluding patients with organ-threatening disease, such studies can be done safely, explained Dr. Merrill.
Dr. Merrill also argued that we need to make progress in understanding how older drugs, such as hydroxychloroquine, work at the genomic and molecular levels, to know how they may best be combined with additional drugs in specific patient subsets. “To figure out how drugs work, … we have to separate them out from all the other drugs, study them in a cleaner environment and then figure out which [lupus patient] subset we need to look at, to see if the drug made an impact,” she said. “It’s tough work, but it’s not impossible.”
Immunologic Parameters
Interferon signals have provided an initial approach to addressing immune heterogeneity and finding subsets of patients that may best respond to certain treatments. Dr. Merrill discussed anifrolumab, a monoclonal antibody that inhibits type I interferon signaling. Although patients with increased interferon signals showed a greater difference between treatment and placebo groups, this seemed to be due to a relatively lower placebo response and not an improved response to treatment. Dr. Merrill showed data to suggest that the four-gene signature the anifrolumab development team used to define interferon signals may not be optimal as a diagnostic.3
Dr. Merrill also described analyses from a phase 2 trial of obexelimab, an agonist of the Fc receptor γRIIb, which helps dampen signals going through the B cell receptor. Although the study did not meet its primary end point of percent of patients without flare at day 225, it met a secondary end point of time to flare. Moreover, when Dr. Merrill and colleagues analyzed the results by the same lupus clusters that had previously been described in Oklahoma, they found that drug responsiveness seemed to occur almost exclusively in patients in two of the seven lupus clusters, resulting in nominally statistically significant improvement.4,5
