Sputum as a new reflective tool of local immune response in lung cancer

Abstract

Introduction

The management of metastatic non-small cell lung cancer (NSCLC) has been revolutionized by immune checkpoint inhibitors (ICI). However, it would be valuable to refine ICI treatment with predictive combined biomarkers models. The tumor microenvironment is profoundly affecting ICI response in NSCLC. Tissue-resident memory T cells (TRM) are a non-circulating lymphocyte population encountered in peripheral tissues with both memory and effector functions. TRM infiltration in lung tumor is correlated with favourable ICI response in NSCLC. But it is not possible to perform iterative lung biopsies to monitor TRM during treatment. The objectives of this study were to evaluate the feasibility of a longitudinal collection of sputum in NSCLC patients and to evaluate TRM cells in comparison with lung biopsies.

Methods

This prospective and exploratory study included 9 patients with a NSCLC treated in first line with pembrolizumab. Before the first ICI infusion and every treatment day until 12-month follow-up, patients received a 15-minute aerosol of isotonic saline solution during which they expectorate. To recover sputum cells, samples were clarified from mucus, stirred, and filtered. TRM cells were quantified by flow cytometry, as CD₄₅ +CD₃ +CD₈ +CD₄₉ +CD₆₉ +CD₁₀₃+ or CD₁₀₃− cells. Spectral multiplex in situ immunofluorescence was used to quantify TRM cells in lung tissue biopsies collected before ICI treatment.

Results

Between May 2021 and September 2022. Nine patients were included: median age was 62 (42–83) years and all patients had smoking history. All patients were able to provide a sputum at each visit and the median follow-up duration was 6.9 months. We were able to identify longitudinally TRM cells in sputum: 80 adequate sputum samples (97.6%) were analyzed by flow cytometry. In order to evaluate the representativeness of TRM cells in sputum, we compared TRM cell proportion determined in lung biopsies by immunofluorescence (method 1) and in the first sputum, by flow cytometry (method 2). The Bland-Altman analysis showed that differences between the 2 methods fitted in the 95% of agreement and bias were close to zero. Thus, there was an agreement between the measurement methods. Moreover, calculation of Pearson coefficient showed a P-value < 0.05 for both CD₁₀₃+ and CD₁₀₃− cells, suggesting that the correlation between the two methods was significant.

Conclusion

Our results indicate that sputum offer a simple and non-invasive indirect estimate of the immune cell infiltration in the pulmonary tissue and allow analysis of TRM cell dynamics during ICI treatment in NSCLC. Sputum collection might help understanding evolution of local immune response during NSCLC and ICI treatment, extending immunological studies to other cell types and function markers.