• The Phase II POPLAR trial of atezolizumab (MPDL3280A) versus docetaxel as second-line treatment for locally advanced or metastatic NSCLC demonstrated a pattern of improved OS (11.4 versus 9.5​
 months) with atezolizumab correlating with PD-L1 expression.
• In a Phase I/II trial of MEDI4736 in a variety of advanced solid tumors, antitumor activity in the NSCLC cohort was encouraging and durable (ORR = 16%) and the toxicity profile was manageable.
• Phase III adjuvant studies of both of these agents are ongoing.

At ASCO we saw more data on anti-PD-L1 antibodies in NSCLC, and although these agents are earlier in development than nivolumab, the findings to date seem similar. Unlike anti-PD-1 antibodies, which bind the PD-1 receptor, these agents bind the PD-L1 ligand and thus do not inhibit PD-L2, but thus far this mechanistic distinction has not translated into a different safety or efficacy profile. Of interest, PD-L2 on normal lung tissue is thought to be important in the rare event of pneumonitis, and it has been hypothesized that the risk of this complication may be less with anti-PD-L1 antibodies.

Dr Spigel commented on the time course of response with checkpoint inhibitors and noted that although some patients will demonstrate objective regressions in the first few weeks of treatment, other responses may be delayed and in some cases are preceded by what appears to be tumor progression. In many cases it may be difficult to determine whether this is true progression or some type of inflammatory infiltrate, but the usual clinical approach is to continue treatment if the patient is clinically stable and no new lesions have developed. A related issue is the optimal duration of therapy, and Dr Spigel notes that some of the checkpoint inhibitor trials call for treatment cessation after a predefined period, particularly if the patient is in complete clinical response. Notably, some patients who have discontinued these agents because of toxicity have on occasion remained stable off treatment.