Research in context
Evidence before this study
We searched the scientific literature for long-term survival outcomes with approved PD-1 and PD-L1 inhibitors in the treatment of previously treated advanced NSCLC. We used the search terms “atezolizumab” AND “nivolumab” AND “pembrolizumab” AND “non-small cell lung cancer” AND “overall survival” AND “previously treated” to search for publications in PubMed from Jan 24, 2014 to Jan 22, 2019, and for congress abstracts presented in 2018 at annual congresses of the American Association of Cancer Research, the American Society of Clinical Oncology, the European Society for Medical Oncology, and the World Conference on Lung Cancer. Our search results showed that monotherapy with PD-1 or PD-L1 inhibitor antibodies in randomised controlled trials improves survival compared with docetaxel in patients with advanced NSCLC previously treated with chemotherapy. Key benefits of these immunotherapies are the durability of responses and the improved long-term survival compared with docetaxel. In addition, greater maximum reduction in tumour target lesions (depth of response) has been associated with longer progression-free survival and overall survival. However, the effect of response status on long-term survival has not been evaluated in detail or in a large patient population to provide more robust analyses.
Added value of this study
Our analyses of 4-year survival outcomes in pooled populations from four studies, including two randomised studies comparing nivolumab and docetaxel, represent, to our knowledge, the longest follow-up to date for immunotherapy in a large population of patients with previously treated advanced NSCLC. The results show a continued long-term survival advantage with nivolumab versus docetaxel up to and beyond 4 years. Six-month landmark analyses of overall survival by response category show that patients with a response (at 6 months) had the greatest post-landmark survival benefit. Additionally, stable disease at the landmark provided a survival benefit versus progressive disease. A survival benefit of response was maintained even after loss of response. In nivolumab-treated patients with disease progression, those who had a previous complete or partial response had the longest overall survival post-progression. A smaller post-progression survival benefit was noted in patients with disease progression after initial achievement of stable disease.
Implications of all the available evidence
Although it has been known for some time that nivolumab is associated with much longer duration of response than docetaxel in patients with previously treated advanced NSCLC, to our knowledge, the current analyses show for the first time that an objective response to nivolumab is associated with an exceptionally durable survival benefit, with the median not reached at 4 years from the time of response. Part of this sustained benefit is attributable to extended survival even after disease progression. Corresponding survival benefits with docetaxel were substantially smaller. Overall, the results show that responses to nivolumab compared with docetaxel are more durable and have a more sustained effect on survival, even after loss of response.