Introduction

Lung cancer is the second most commonly diagnosed cancer and the leading cause of cancer death worldwide, of which approximately 85% are non-small cell lung cancer (NSCLC).1,2 The overall survival (OS) of patients with advanced NSCLC was significantly prolonged with immune checkpoint inhibitors (ICIs) targeting the programmed cell death-1 (PD-1) and programmed death-ligand 1 (PD-L1) axis.3,4,5 For early-stage lung cancer, the 5-year survival rate for patients ranges from 80% in stage IA to 41% in stage IIIA, and many cases relapse after surgical resection.6 Currently, multiple clinical trials have manifested the encouraging efficacy of neoadjuvant immunotherapy in stage I-IIIA resectable NSCLC.7,8,9 However, the effect of immunotherapy in ultra early-stage NSCLC patients with micro-invasive or even pre-invasive lesions remains unclear.

With the implementation of computed tomography (CT)-guided lung cancer screening, there has been a gradual increase in the detection of pulmonary nodules.10 They are classified as solid or sub-solid, with the latter further divided into pure ground-glass opacity (GGO) and part-solid, based on CT appearance.11 There is remarkable difference in biological behavior between lung cancers manifesting as different radiological types. Compared with lung cancers presenting with solid nodules, GGO-associated lung cancers have an indolent clinical course, and are characterized by a less active metabolism and a less active immune microenvironment.12,13

In recent years, an increasing number of multifocal lung cancers have been diagnosed.14,15 Multiple primary lung cancer (MPLC) often presents as multiple GGOs in CT, most of which are minimally invasive or pre-invasive lesions.16,17,18,19 One routine option is to resect the major lesion(s), followed by close surveillance of the remaining lesions.20,21 Generally, it was extremely difficult to remove all lesions for MPLC patients, considering their pulmonary function, comorbidities, and multiple lesions in different lobes, etc. As a clinical dilemma, there is no consensus on the management of unresected lesions with a high risk of progression for MPLC patients after primary surgery.17,6

MPLC diagnosis

The occurrence of two or more primary lung cancers in the same individual is known as MPLC.75,76 For patients with more than one site of lung cancer, distinguishing between MPLC and intrapulmonary metastasis (IPM) is crucial in clinical practice.77 The judgment of MPLC is commonly based on a multidisciplinary team (MDT), taking into account clinical, radiologic, and (if available) tumor cytologic/histologic/genetic features. In this study, the included MPLC patients were diagnosed by the MDT, according to the diagnostic criteria from ACCP guidelines.78

T/B/NK-cell subpopulations

The T-cell, B-cell, and NK-cell subpopulations in the peripheral blood were detected in enrolled patients using flow cytometry: 1) Take the antibody reagents from 20uL CD3/CD8/CD45/CD4 Assay Kit (Agilent) or CD3/CD16 + CD56/CD45/CD19 Assay Kit (Agilent) at room temperature, and add them to the FACS tube; 2) add 50uL of fully mixed anticoagulated peripheral whole blood to the tube; 3) gently shake the tube for 5 s using a vortex mixer and incubate for 15 min (18–25 °C); 4) add 450 uL hemolysin; 5) gently shake it for 5 s using a vortex mixer and incubate for 15 min (18–25 °C). Then the samples were analyzed using the flow cytometer (Agilent NovoCyte) and software (NovoExpress).

In this examination, CD8+ T cells showed CD45+CD3+CD8+; CD4+ T cells showed CD45+CD3+CD4+; B cells showed CD45+CD3CD19+; and NK cells showed CD45+CD3CD56+.

TCR-seq, cytokines, exosomal RNA and mIHC

We performed the examination of TCR-seq, cytokines, and exosomal RNA by using blood samples of responders and non-responders. The mIHC analysis was performed on their resected tumors.

TCR-seq

PBMCs from blood samples of patients were isolated by Ficoll density gradient centrifugation and extracted total RNA using the RNeasy Plus Mini kit (Qiagen). The total RNA (500 ng) of each sample was amplified through multiplex PCR (mPCR) using the Oncomine™ TCR Beta‑SR Assay Kit (Thermo) according to the manufacturer’s instructions. Further, TCR libraries were quantified using the Ion Library TaqMan® Quantitation Kit (Thermo) and sequenced by the Ion GeneStudio™ S5 System (Thermo). The data analysis was performed using R v.4.1.3.

Cytokines

Serum cytokines were detected through the Cytokine/Chemokine/Growth Factor 45-Plex Human ProcartaPlex Panel 1 (Thermo) and Immuno-Oncology Checkpoint 14-Plex ProcartaPlex Panel 1 (Thermo) according to manufacturer’s instructions, which included 45 cytokines and 14 immune checkpoints, respectively. The results were measured and analyzed by the Luminex-200 system (Lumiex).

Exosomal RNA

Blood exosomes were isolated by SEC (size exclution chromatography) methods.79 In brief, blood exosomes were eluted and purified using the Exosupur® columns (Echobiotech), then concentrated by 100 kDa molecular weight cut-off Amicon® Ultra spin filters (Merck). The exosomes were verified using nanoparticle tracking analysis (NTA), transmission electron microscopy (TEM) and Western blot analysis. Exosome RNA was extracted and purified with QIAGEN RNeasy Mini Kit (Qiagen). RNA concentration and purity were evaluated using RNA Nano 6000 Assay Kit of the Agilent Bioanalyzer 2100 System (Agilent Technologies). A total amount of 250pg–10ng RNA per sample was used as input material for sequencing libraries using the SMARTer Stranded Total RNA-Seq Kit (Takara Bio) and the index codes were added to attribute sequences for each sample. Library quality was assessed by the Agilent Bioanalyzer 2100 and qPCR. The libraries were then sequenced on an Illumina Hiseq platform, and paired-end reads were generated.

mIHC

There were two panels of 10 biomarkers examined in this study, including panel 1: CD8 (cytotoxic T cells; Clone SP16; ZA0508; Zsbio), CD4 (T helper cells; Clone EP204; ZA0519; Zsbio), PD-1 (programmed cell death-1; Clone UMAB199; ZM0381; Zsbio), PD-L1 (programmed cell death-Ligand 1; Clone SP142; ZA0629; Zsbio), Foxp3 (regulatory T cells; Clone 236A/E7; ab20034; Abcam); and panel 2: CD19 (B cells; Clone UMAB103; ZM0038; Zsbio), CD56 (natural killer cells; Clone UMAB83; ZM0057; Zsbio), CD68 (macrophages; Clone KP1; ZM0060; Zsbio), CD163 (M2 macrophages; Clone 10D6; ZM0428; Zsbio), Cytokeratin (tumor cells; Clone AE1/AE3; ZM0069; Zsbio).

Formalin-fixed and paraffin-embedded (FFPE) samples were cut from surgical specimens, sections of 4 μm thickness. The slides were stained manually according to the instruction using the Opal seven-color IHC Kit (NEL797B001KT; PerkinElmer), including fluorophores 4’,6-diamidino-2-phenylindole (DAPI), Opal 650 (CD8), Opal 570 (CD4), Opal 690 (PD-1), Opal 620 (PD-L1), Opal 520 (Foxp3); Opal 620 (CD19), Opal 520 (CD56), Opal 650 (CD68), Opal 570 (CD163), Opal 690 (Cytokeratin), and TSA Coimarin system (PerkinElmer). Every staining round contained a slide of tonsil as positive control. Stained slides were scanned by the Vectra (Vectra 3.0.5; PerkinElmer). After scanning, a selection of 15 representative images were used to analysis by the inform software (inform 2.3.0; PerkinElmer).

Statistical analysis

The Student’s t test, Wilcoxon’s rank-sum test, and ANOVA were applied to compare continuous variables, including the proportion and absolute counting of circulating immune cells (T/B/NK-cell), the ratio of CD8+/CD4+ T-cell, the TCR clonality and diversity (Shannon-index/evenness/convergence), and the concentration values of various cytokines between responders and non-responders (T1–T4). The ORR and TRAEs were expressed as frequencies and percentages.

To compare the changing trend of proportion of various immune cells over treatment between responded and non-responded patients, the repeated measures analysis of variance was applied, with the change of various indicators as the dependent variable, and response, time, and the response×time interaction as the independent variables.

A two-sided p-value < 0.05 was considered as significant. Clinical analyses were conducted using SPSS Statistics (version 23.0, IBM), R v.4.1.3, Microsoft Excel v.2019 and GraphPad Prism v.8.00. The specific parameters for R v.4.1.3 analyses used in this study were described in the Supplementary materials.