Double-edged sword.

Top Line: There are at least a couple of reasons immune checkpoint inhibitors (ICI) aren’t going to eradicate cancer any time soon, namely that they don’t work in some cases and they work too much in others.
The Study: This ingenious review of Japanese patients receiving PD-1 axis blockade with nivolumab (n=99) or pembrolizumab (n=38) for non-small cell lung cancer (NSCLC) aimed to capitalize on more than medical records. They analyzed pretreatment blood samples to query the ability of simple serum markers to predict immunotherapy response. Specifically, post-hoc tests were performed on samples from patients without a known autoimmune disorder to detect pre-existing rheumatoid factor, antinuclear antibody, antithyroglobulin, or antithyroid peroxidase. Interestingly, over half (58%) of patients had at least one autoimmune maker. And these people had half (0.53) the hazard of progression or death on ICI. But they also had over 1.5x the risk of adverse reaction (45 → 73%), most commonly dermatologic. What’s more, specific autoimmune markers were highly predictive of specific adverse events. For example, harboring rheumatoid factor doubled the risk of developing skin toxicity (24 → 47%) while thyroid trouble occurred virtually only among those with antithyroid markers (20%) versus without (1%). Notably, there was a pretty even spread ranging from no to strong tumor expression of PD-L1 with no apparent association with pre-existing autoimmune markers.
Bottom Line: Autoimmune markers may provide an easy new insight into predicting ICI response and toxicity. | Toi, JAMA Oncol 2018


Popular Posts