AMSTERDAM – In patients with autoimmune diseases, cancer treatment with checkpoint inhibitor immunotherapy increases the risk of flares, but these flares are associated with improved cancer outcomes, according to a multicenter, retrospective study presented at the European Congress of Rheumatology.
“Survival was longer in patients who experienced a flare of their preexisting autoimmune disease or any other immune-related adverse event, but this gain was lost if an immunosuppressive therapy was used,” reported Alice Tison, a resident in rheumatology at the Centre Hospitalier Universitaire, Brest, France.
These were some of the mixed messages from this evaluation, which involved 112 patients with preexisting autoimmune disease (PAD) whose data were collected from 11 tertiary care centers in France. Of the cases of PAD represented, the majority involved joint diseases, including psoriatic arthritis (28%), rheumatoid arthritis (18%), and spondyloarthritis (4.5%). However, other types of PAD, including inflammatory bowel disease (13%), were included in the series.
Only 33% of the patients had active disease at the time that checkpoint inhibitor therapy was initiated, and only 21% were taking an immunosuppressive therapy for their disease. Of those on therapy, the majority were taking steroids, but about a third of those on therapy were taking a disease-modifying antirheumatic drug, such as methotrexate.
With the initiation of checkpoint inhibitors, which were offered primarily for the treatment of melanoma (59%) and non–small cell lung cancer (36%), 42% of patients with PAD developed a disease flare. Of these, 30% were considered severe. Other immune-related events not considered related to the underlying disease, such as colitis, were also observed but at rates not clearly different than those observed in patients without PAD.
The activity of checkpoint inhibitors did not appear to be different than that observed in non-PAD patients. For example, the overall response rate was 48% in those with melanoma and 54% in those with non–small cell lung cancer. After a median of 8 months of follow-up, the median progression-free survival was 12.4 months and 9.7 months for the two diseases, respectively. Median overall survival had not been reached in either disease.
However, those with a flare or another immune-related adverse event had significantly better progression-free survival (P = .016) and overall survival (P = .004) when compared with those who did not flare or have an immune-related adverse event. According to Ms. Tison, this has been reported before, but a more surprising finding was that the gain in progression-free survival and overall survival was lost in those treated with an immunosuppressive drug.
Even though non-PAD patients commonly receive steroids for immune-related adverse events such as colitis, the loss of benefit in PAD patients who received immunosuppressive therapies may be caused by, at least in part, cross-reactivity between tumor antigens and autoantigens, Ms. Tison speculated.
Ms. Tison was cautious in drawing conclusions about specific strategies to optimize benefits from checkpoint inhibitors in PAD based on this limited series of patients. However, she did suggest that discontinuation of immunosuppressive therapies prior to initiating checkpoint inhibitors may be prudent in PAD patients, particularly those with inactive disease.
Overall, she emphasized that checkpoint inhibitors “have revolutionized the management of several cancers” and should not be denied to PAD patients who are otherwise appropriate candidates. Although flares are common, more than half of PAD patients in this series did not flare and flares were mild to moderate in most of those who did.
“The response to checkpoint inhibitors in PAD patients is good,” Ms. Tison advised. For those who do flare, “we need prospective studies to understand which strategies provide a good balance of benefit to risk” for cancer immunotherapy and for the options to manage immune-related adverse events.
The study was not industry funded. Ms. Tison reported no potential conflicts of interest.
SOURCE: Tison A et al. . EULAR Congress 2018, Abstract OP0196.