Electronic nose (eNose) technology, which uses several cross-reactive sensors that analyze a complete mixture of volatile organic compounds in exhaled breath, may be useful for identifying patients with advanced-stage non-small cell lung cancer (NSCLC) who are most likely to respond or not respond to immune checkpoint inhibitors, study results in the Annals of Oncology suggest.

Researchers enrolled patients with advanced-stage NSCLC who were eligible for antiprogrammed death ligand 1 (PD-L1) therapy in the observational study (N=143). They included individuals in the training set (n=92) if they had initiated pembrolizumab or nivolumab between March 2016 and April 2017, and assembled a validation cohort of patients who started therapy after April 2017 (n=51).

The investigators collected exhaled breath profiles in duplicate at baseline with a 2-minute interval using the eNose. They conducted the analysis during the same day as the spirometry tests within 2 weeks before the start of anti-PD-1 therapy. Researchers used standard eNose software to process eNose sensor signals, including ambient correction, detrending, filtering, and peak detection. At 3-month follow-up, they assessed efficacy of either pembrolizumab or nivolumab using the Response Evaluation Criteria in Solid Tumors version 1.1.

Continue Reading

Related Articles

In the training set, responders and nonresponders to anti-PD-1 therapy differed at baseline with regard to exhaled breath analysis of sensor 3 (P <.001), sensor 5 (P =.01), sensor 1_BH (P =.04), and sensor 5_BH (P =.001). When the eNose sensors were combined into a single biomarker, the receiver operating characteristic-area under the curve reached 0.89 (95% CI, 0.82-0.96). In the validation set at baseline, PD-L1 expression was different between responders and nonresponders to anti-PD-1 therapy according to the independent t-test analysis (P =.03). Similar to the training set, breath profiles of patients in the validation set differed between responders and nonresponders (receiver operating characteristic-area under the curve 0.85; 95% CI, 0.7-0.96).

Limitations of the study include the lack of a randomized design and its potential for selection bias. 

Findings from this study suggest that the use of the “eNose can potentially avoid application of ineffective treatment” in patients with NSCLC less likely to respond to checkpoint inhibitors while also helping to avoid “unnecessary delays and start treatment with a better alternative.”

Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.


de Vries R, Muller M, van der Noort V, et al. Prediction of response to anti-PD-1 therapy in patients with non-small-cell lung cancer by electronic nose analysis of exhaled breath [published online September 17, 2019]. Ann Oncol. doi:10.1093/annonc/mdz279