Construction of a checkpoint inhibitor-related pneumonia diagnostic model based on exhaled nitric oxide: a prospective observational study
Original Article

Construction of a checkpoint inhibitor-related pneumonia diagnostic model based on exhaled nitric oxide: a prospective observational study

Yimei Gao1#, Tingyue Luo1#, Danhui Huang1#, Zeyu Fu1, Shudong Ma2, Li Lin2, Haohua Huang1, Tiantian Liu1, Jinming Zhang1, Xiaoxiao Jiang1, Yanmei Ye1, Junwei Chen1, Junjie Xi1, Jinzhong Zhuo1, Kaijun Chen1, Jingqi Ai1, Laiyu Liu1, Shaoxi Cai1, Hangming Dong1

1Chronic Airways Diseases Laboratory, Department of Respiratory and Critical Care Medicine, Nanfang Hospital, Southern Medical University, Guangzhou, China; 2Department of Oncology, Nanfang Hospital, Southern Medical University, Guangzhou, China

Contributions: (I) Conception and design: Y Gao, T Luo, H Dong; (II) Administrative support: S Ma, L Lin, Y Ye, L Liu, S Cai, H Dong; (III) Provision of study materials or patients: T Luo, D Huang, Y Ye, S Ma, L Lin, L Liu, S Cai, H Dong; (IV) Collection and assembly of data: Y Gao, H Huang, T Liu, J Chen, J Xi, J Zhuo, K Chen, J Ai; (V) Data analysis and interpretation: Y Gao, T Luo, D Huang, H Huang, T Liu, J Zhang, H Dong; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.

#These authors contributed equally to this work.

Correspondence to: Hangming Dong, PhD; Shaoxi Cai, PhD; Laiyu Liu, PhD. Chronic Airways Diseases Laboratory, Department of Respiratory and Critical Care Medicine, Nanfang Hospital, Southern Medical University, 1023-1063 Shatai South Road, Baiyun District, Guangzhou 510515, China. Email: dhm@smu.edu.cn; hxkc@smu.edu.cn; liulaiyu@sina.com.

Background: Checkpoint inhibitor-related pneumonia (CIP) is a complication of immune checkpoint inhibitors (ICIs) with high mortality. There is still a lack of effective biomarkers to identify CIP. Exhaled nitric oxide (eNO), an airway inflammatory marker, can be obtained by non-invasive methods, but its value in CIP is unknown. The purpose of this study was to investigate the value of eNO in CIP.

Methods: Lung cancer patients who received ICIs were included at Nanfang Hospital, Southern Medical University. Fractional eNO at expiratory flow rates of 50 and 200 mL/s (FeNO50 and FeNO200) were measured. The alveolar concentration of nitric oxide (CaNO) was calculated based on the two-compartment model of airway and alveoli. The optimal CaNO cut-off value was determined by the receiver operating characteristic (ROC) curve. eNO, clinical characteristics, and laboratory tests were analyzed to find out the risk factors for CIP by logistic regression analysis. A multi-indicator model based on best risk factors for CIP was developed and internally validated.

Results: CaNO was significantly elevated in the CIP group [8.1±5.0 vs. 4.9±3.1 parts per billion (ppb), P<0.001]. The area under the curve (AUC) of CaNO to differentiate CIP was 0.728 [95% confidence interval (CI): 0.670–0.786; P=0.001]. The best cut-off value of CaNO was 6.350 ppb. Increased CaNO [odds ratio (OR), 1.30; 95% CI: 1.19–1.43; P<0.001], emphysema reported on chest computed tomography (CT) (OR, 2.54; 95% CI: 1.41–4.60), a small amount of pleural effusion reported on chest CT (OR, 2.48; 95% CI: 1.37–4.50), pre-existing radiotherapy (OR, 3.89; 95% CI: 1.96–7.73) and the lower counts of lymphocyte cell in peripheral blood (OR, 0.69; 95% CI: 0.44–1.10) were independently associated with CIP. The five factors were incorporated into a multi-indicator model with a good predictive accuracy of 0.821.

Conclusions: CaNO may be a new marker for identifying CIP. Increased CaNO, pre-existing radiotherapy and emphysema reported on chest CT, a small amount of pleural effusion reported on chest CT, and lower count of lymphocyte cell in peripheral blood are independently associated with CIP.

Keywords: Lung cancer; checkpoint inhibitor-related pneumonia (CIP); exhaled nitric oxide (eNO); diagnostic model

Submitted Nov 14, 2024. Accepted for publication Mar 05, 2025. Published online May 20, 2025.

doi: 10.21037/tlcr-2024-1085

Highlight box

Key findings

• The study is the first to reveal the phenomenon of elevated alveolar nitric oxide concentration (CaNO) in checkpoint inhibitor-related pneumonia (CIP) and CaNO may be a new biomarker to distinguish CIP.

What is known and what is new?

• Exhaled nitric oxide (eNO) has been widely used in inflammatory diseases such as asthma and interstitial lung disease, obtained by a noninvasive method. In particular, CaNO can show the level of inflammation in the alveolar area; however, the value of eNO in CIP is unclear.

• Different parameters of eNO were analyzed in cancer patients using immune checkpoint inhibitors (ICIs). It was found that the levels of CaNO in patients with CIP were higher than those in patients without CIP, regardless of the presence or absence of complications such as pulmonary infection or tumor progression. A multi-indicator diagnostic model based on CaNO was constructed. Clinicians can use this tool to screen patients with CIP.

What is the implication, and what should change now?

• eNO detection technology was applied to the identification of CIP for the first time, supplementing the means available in the field of immunotherapy toxicity monitoring. In the future, CaNO may be used to dynamically monitor the pulmonary toxicity during immunotherapy, and a non-invasive and real-time monitoring system may be established.

Introduction

Immune checkpoint inhibitors (ICIs) have brought revolutionary breakthroughs in the treatment of many advanced malignant tumors (1-4). However, during the process of ICIs regulating the immune response to kill tumor cells, excessive immune activation may lead to immune-related adverse events (irAEs) that result in the involvement of various organs (5). Among the different target organs, the lung is the most frequently associated with adverse events (6,7). The incidence of checkpoint inhibitor-related pneumonia (CIP) has been reported to be less than 5% in clinical trials (6,8,9), whereas the data in the real world are much higher than those in clinical trials, as high as 9.5–19% (10-12). Wang et al. (13) found that deaths caused by CIP accounted for about 35% of programmed death receptor 1/programmed cell death ligand 1 (PD-1/PD-L1) inhibitor-related deaths. However, the diagnosis of CIP remains a challenge for clinicians due to it lacking typical clinical symptoms and imaging features.

In the prediction, diagnosis, and differentiation of CIP, a large number of biomarkers have sprung up (12,14,15). At present, there are also clinical diagnostic models (16-18), but most of the biomarkers involved in these models come from blood and bronchoalveolar lavage fluid which are invasive. Cytokines derived from blood samples are often affected by the patient’s own metabolism and complications; elderly and frail patients are often unable to withstand the impact of invasive procedures such as bronchoscopy.

Nitric oxide (NO) is an important cellular signaling molecule that can be induced by airway epithelial cells, vascular endothelial cells, resident macrophages, and inflammatory cells through inducible nitric oxide synthetase (iNOS) (19). At the same time, inflammatory factors can also promote the production of NO including interleukin and endothelin (20). Multi-flow NO detection can relatively distinguish between bronchial and alveolar-derived NO. Fractional exhaled nitric oxide at expiratory flow rates of 50 mL/s (FeNO50) represents bronchial main airway inflammation, whereas fractional exhaled nitric oxide at expiratory flow rates of 200 mL/s (FeNO200) and CaNO, calculated by the “two-compartment model” (21), represent distal airway and alveolar inflammation. Exhaled nitric oxide (eNO) can be obtained by non-invasive detection methods, and one of the indicators, alveolar nitric oxide concentration (CaNO), has been considered a marker of distal airway inflammation. The CaNO has been widely studied in chronic airway diseases such as asthma, cough, chronic obstructive pulmonary disease, and interstitial lung disease (ILD) (22-24). However, the value in CIP is unknown. The purpose of this study was to investigate the value of CaNO in CIP. In addition, we analyzed the influence of CaNO and clinical characteristics in CIP, and constructed a multi-indicator model using these best risk factors to identify CIP with internal validation. We present this article in accordance with the TRIPOD reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2024-1085/rc).

Methods

Patients

This prospective study was conducted to determine the role of eNO in CIP at the Nanfang Hospital of Southern Medical University between May 2021 and October 2024. Patients who were diagnosed with lung cancer by cytology or histology and treated with ICIs during any treatment line were included; at the same time, eNO testing was performed. The highlighted sentences revised as “The time of eNO testing for the CIP group patients was when pneumonia was developing or when the patient was receiving hormone therapy but was not cured. For the non-CIP group, the time was while the patient was receiving immunosuppressive therapy. Comorbidities include infection or progression. Patients were excluded if they could not complete the eNO test, had allergic rhinitis or acute asthma attacks, or were complicated with tuberculosis, ILDs, rheumatoid arthritis, or sleep apnea syndrome. Comorbidities included infection or progression. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Ethics Committee of Nanfang Hospital (No. NFEC-2021-397 & No. NFEC2023-544). All study participants provided written informed consent.

Data collection

Patients treated with ICIs were regularly hospitalized for anti-tumor treatment according to the original treatment plans. During hospitalization, the eNO measurements were taken for each patient and the clinical information was collected including patient demographic (gender, age, smoking habits), complication of lung diseases [obstructive pneumonia, emphysema, atelectasis, pleural effusion, pulmonary embolism, small airway dysfunction (SAD)] in chest computed tomography (CT) reported during the three months before and after the examination of eNO, pre-existing lung diseases [chronic obstructive pulmonary disease (COPD), ILD, asthma], usage of bronchodilators, tumor types, tumor-node-metastasis (TNM) stage, histologic subtype of oncology, pre-existing radiotherapy, therapeutic regimens, blood routine, the results of spirometry, and the measurements of eNO. The results of laboratory testing were only collected within 24 hours before and after the measurements of eNO. The results of complications were recorded, including pulmonary infection or tumor progression on chest CT during hospitalization. For the study, we defined previous smoking as ≥6 months of smoking cessation, current smoking as current smoking or <6 months of smoking cessation, and heavy smoking as a smoking index score ≥400 pack-years. Smoking index = number of cigarettes smoked per day (pack) × duration of smoking (year). According to the results of spirometry, SAD was defined as the presence of two or more items <65% of the predicted value: maximal expiratory flow at 75% of forced vital capacity (MEF75), maximal expiratory flow at 50% of forced vital capacity (MEF50), and maximum mid-expiratory flow (MMEF) (25). COPD was defined as a ratio of forced expiratory volume in 1 second (FEV1) to forced vital capacity (FVC) of <0.70 after bronchodilator use, as measured by spirometry (26). Each patient completed at least one eNO test. All patients were followed up until 30 October 2024.

Measurements of eNO

The Nano Coulomb Breath Analyzer (Sunvou-CA2122, Sunvou, Wuxi, China) was used to measure NO levels in real time under the guidance of professional personnel according to the 2005 American Thoracic Society/European Respiratory Society (ATS/ERS) technical standards (27) and the 2017 ERS expiratory marker measurement technical standards (28). No strenuous activity, food intake, smoking, or spirometry were allowed within 1 hour before eNO examination. In particular, broccoli, kale, lettuce, celery, water radish, and smoked or pickled foods are prohibited. The principle of the device was to mark eNO at exhaled flow rates of 50 and 200 mL/s, which were recorded as FeNO50 and FeNO200, respectively.

According to the two-compartment model proposed by Tsoukias et al. (21), the device calculates CaNO online using the formula: FeNO = CaNO + apparent bronchial nitric oxide production (J'awNO)/expiratory flow rate (V'E), making use of J'awNO and the NO parameters independent of gas flow NO parameters. A linear equation consisting of two flow rates (50 and 200 mL/s) was calculated by the Nano Coulomb Breath Analyzer. Within error, the slope of the linear equation is approximated by CaNO and the intercept by J'awNO. Due to the limited computational power of the equipment, precise values of CaNOs less than 1 parts per billion (ppb) could not be obtained, and values of CaNO <1 ppb were recorded as 1 ppb in this study. According to the 2017 ERS expiratory marker measurement technical standards (29) and the Nano Coulomb NO Breath Analyzer reference values, the cut-off values of normal FeNO50 <25 ppb, FeNO200 <10 ppb, and CaNO <5 ppb were applied (30,31).

Diagnosis of CIP

The clinical manifestations of CIP are not specific and there is no specific serological marker at present. The diagnosis is often based on comprehensive analysis of clinical history, symptoms, laboratory tests, imaging, and other relevant data, and the therapeutic diagnosis is usually adopted. That is, the clinical symptoms were neither relieved nor aggravated after antibiotic treatment, or the lung shadow still progressed. After steroid treatment, the symptoms were relieved and the lung shadow was absorbed, and the diagnosis of CIP was retrospectively considered.

Statistical analysis

Continuous variables were presented as medians (interquartile range) and categorical variables as numbers (%). For comparison between groups, the Mann-Whitney test or independent sample t-test was used to analyze continuous variables, and either the Pearson χ2 test or Fisher’s exact test was used to analyze categorical variables. Missing data were incomplete with the use of means. In this study, FeNO results were collected only once from patients, and no follow-up was required. There was no loss to follow-up. The receiver operating characteristic (ROC) curve was used to analyze the ability of CaNO to identify CIP in patients with ICIs, and the best CaNO cut-off value was determined. Univariate logistic regression was used to analyze the role of each single factor in identifying CIP in the population at first, and then those significant variables with a P<0.05 of the univariate analysis were analyzed in multivariate logistic analysis to determine independent factors of CIP and building a multi-indicator model based on CaNO. The odds ratio (OR) and 95% confidence interval (CI) of each factor were calculated. The single-indicator model and the multi-indicator model were built by the “rms” package and its discriminative ability was demonstrated by the ROC curve plotted using the pROC package in R (R Foundation for Statistical Computing, Vienna, Austria). The area under the curve (AUC) of the ROC curve, also called the concordance statistic (C-statistic), was applied for appraising validity of the two models respectively. In addition, the bootstrap method (B=1,000) was used to select random samples for internal validation to assess the stability of the multi-index model. Calibration curves were drawn to show the prediction results between the multi-indicator model prediction and the actual observed probability. All P values were based on two-sided tests, and P values of less than 0.05 were considered statistically significant. Statistical analyses were performed using R software version 4.3.3 and GraphPad Prism 10.0 (GraphPad Software, La Jolla, CA, USA).

Results

Baseline characteristics of the study population

A total of 414 lung cancer patients were included in this study. There were 307 patients enrolled, including 100 patients (32.6%) with CIP groups (51 CIP patients without comorbidities and 49 CIP patients with comorbidities); and 207 patients with non-CIP (77 non-CIP patients without comorbidities and 130 non-CIP patients with comorbidities) after exclusion of 84 patients with allergic rhinitis, asthma, tuberculosis, sleep apnea syndrome or ILD and 23 patients who could not complete the eNO test (some patients with advanced tumors were weak or patients with severe COPD had poor lung function and lacked sufficient strength to complete the test), as shown in Figure 1. The clinical characteristics of participants are listed in Table 1. The cohort was predominantly male (81.8%) and the median patient age was 62.0 years. There were 230 (74.9%) current or former smokers and 102 patients (33.2%) met the criteria for heavy smoking. At the time of taking the eNO test, chest CT reports of enrolled patients showed obstructive pneumonia in 180 patients (58.6%), emphysema in 143 patients (46.6%), atelectasis in 33 patients (10.7%), a small amount of pleural effusion in 111 patients (36.2%), and pulmonary embolism in 9 patients (2.9%). The cohort had pre-existing pulmonary comorbidities including 79 COPD patients (25.7%), 2 ILD patients (0.7%), and 24 SAD patients (7.8%). A total of 267 (87.0%) patients had never used drug-based bronchodilators. Pre-existing radiotherapy had been administered to 69 (22.5%) patients. The distribution of pre-existing radiotherapy among CIP and non-CIP was significantly different (P<0.001). Patients with adenocarcinoma accounted for about 37.8% of the cohort. ICIs were used in III/IV tumor stage (92.5%). Most participants (86.6%) received first and second lines of therapeutic regimen and in combination therapy (89.3%). In peripheral blood tests results, the cell counts of neutrophils and lymphocyte were comparable between the two groups.

Figure 1 The flow chart showing cohort selection. Comorbidities include infection or progression. CIP, checkpoint inhibitor-associated pneumonitis; ILD, interstitial lung disease.

Table 1

Baseline characteristics of the study population

Variables All patients (n=307) Control group (n=207) CIP group (n=100) P value
Gender 0.07
   Male 251 (81.8) 163 (78.7) 88 (88.0)
   Female 56 (18.2) 44 (21.3) 12 (12.0)
Age (years) 62.0 (55.0, 68.0) 60.0 (54.0, 66.0) 63.0 (56.0, 67.0) 0.09
Smoking habits 0.057
   Current 76 (24.8) 56 (27.1) 20 (20.0)
   Former 154 (50.2) 94 (45.4) 60 (60.0)
   Never 77 (25.1) 57 (27.5) 20 (20.0)
Severe smoking 0.01
   Yes 102 (33.2) 128 (61.8) 77 (77.0)
   No 205 (66.8) 79 (38.2) 23 (23.0)
Complication of lung disease
   Obstructive pneumonia 180 (58.6) 113 (54.6) 67 (67.0) 0.052
   Emphysema 143 (46.6) 83 (40.1) 60 (60.0) 0.002
   Atelectasis 33 (10.7) 19 (9.2) 14 (14.0) 0.27
   Pleural effusion 111 (36.2) 63 (30.4) 48 (48.0) 0.004
   Pulmonary embolism 9 (2.9) 6 (2.9) 3 (3.0) 0.78
   COPD 79 (25.7) 55 (26.6) 24 (24.0) 0.73
   ILD 2 (0.7) 2 (1.0) 0 0.81
   Small airway dysfunction 24 (7.8) 18 (8.7) 6 (6.0) 0.07
Bronchodilators 0.20
   Current/former 40 (13.0) 23 (11.1) 17 (17.0)
   Never 267 (87.0) 184 (88.9) 83 (83.0)
Pre-existing radiotherapy 69 (22.5) 27 (13.0) 42 (42.0) <0.001
Cancer type 0.01
   Adenocarcinoma 116 (37.8) 80 (38.6) 36 (36.0)
   Squamous 84 (27.4) 51 (24.6) 33 (33.0)
   Small cell carcinoma 59 (19.2) 35 (16.9) 24 (24.0)
   Others# 48 (15.6) 41 (19.8) 7 (7.0)
Tumor clinical staging 0.01
   I 7 (2.3) 6 (2.9) 1 (1.0)
   II 16 (5.2) 16 (7.7) 0
   III 75 (24.4) 45 (21.7) 30 (30.0)
   IV 209 (68.1) 140 (67.6) 69 (69.0)
Therapeutic regimen 0.13
   First and second lines 266 (86.6) 184 (88.9) 82 (82.0)
   Third line and above 41 (13.4) 23 (11.1) 18 (18.0)
Treatment with ICI 0.37
   Monotherapy 33 (10.7) 25 (12.1) 8 (8.0)
   Combined therapy 274 (89.3) 182 (87.9) 92 (92.0)
Blood tests
   White blood cell (×109/L) 6.39 (4.86, 7.96) 6.33 (4.75, 7.71) 6.47 (5.16, 8.56) 0.27
   Neutrophils (×109/L) 4.03 (3.10, 4.26) 3.86 (2.88, 5.01) 4.54 (3.33, 6.12) 0.01
   Lymphocyte (×109/L) 1.48 (1.08, 2.01) 1.62 (1.21, 2.06) 1.41 (0.84, 1.80) 0.01
   Eosinophils (×109/L) 0.12 (0.05, 0.22) 0.12 (0.05, 0.22) 0.10 (0.04, 0.21) 0.11
   CRP (mg/L) 7.46 (2.50, 17.64) 6.91 (2.23, 17.64) 8.53 (2.99, 18.22) 0.15
   PCT (ng/mL) 0.16 (0.06, 0.16) 0.16 (0.06, 0.16) 0.16 (0.06, 0.16) 0.98

Data are presented as n (%) or median (range). Severe smoking: defined as smoking index ≥400; Others# included lung lymphocytic epithelioid carcinoma and other pathological types of lung cancers. CIP, checkpoint inhibitor-associated pneumonitis; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; ICI, immune checkpoint inhibitor; ILD, interstitial lung disease; PCT, procalcitonin.

Comparison of eNOs between CIP and non-CIP and in different subgroups

Compared to the non-CIP group, the CIP group exhibited markedly higher CaNO (8.1±5.0 vs. 4.9±3.1 ppb, P<0.001), but there was no significant difference in FeNO50 (24.0±30.6 vs. 23.7±26.7 ppb, P=0.32) and FeNO200 (12.0±8.8 vs. 10.8±8.8 ppb, P=0.06) between the CIP group and non-CIP group (Figure 2). The expression is not correct, modified as “in subgroup analyses”. For the values of CaNO, the CIP group was higher than the non-CIP with comorbidities group (8.1±5.0 vs. 5.3±3.2 ppb) (P≤0.05). The CIP group was higher than the non-CIP without comorbidities group (8.1±5.0 vs. 4.8±2.8 ppb) (P≤0.001). The CIP without comorbidities group was higher than the non-CIP with comorbidities group (7.4±5.1 vs. 5.3±3.2 ppb) (P≤0.05). The CIP without comorbidities group was higher than the non-CIP without comorbidities group (7.4±5.1 vs. 4.8±2.8 ppb) (P≤0.01). The CIP with comorbidities group was higher than the non-CIP with comorbidities group (8.3±5.4 vs. 5.3±3.2 ppb) (P≤0.05). The CIP with comorbidities group was higher than the non-CIP without comorbidities group (8.3±5.4 vs. 4.8±2.8 ppb) (P≤0.001). For the values of FeNO50, there was no significant difference between the CIP group and the non-CIP with comorbidities group (24.0±30.6 vs. 20.7±17.5 ppb) (P>0.05). The CIP group was lower than the non-CIP without comorbidities group (24.0±30.6 vs. 28.9±36.9 ppb) (P≤0.05). There was no significant difference between the CIP without comorbidities group and the non-CIP with comorbidities group (27.6±36.9 vs. 20.7±17.5 ppb) (P>0.05). The CIP without comorbidities group was lower than the non-CIP without comorbidities group (20.5±17.0 vs. 28.9±36.9 ppb) (P≤0.05). There was no significant difference between the CIP with comorbidities group and the non-CIP with comorbidities group (20.5±17.0 vs. 20.7±17.5 ppb) (P>0.05). There was no significant difference between the CIP with comorbidities group and the non-CIP without comorbidities group (20.5±17.0 vs. 28.9±36.9 ppb) (P>0.05). For the values of FeNO200, there was no significant difference between the CIP group and the non-CIP with comorbidities group (12.0±8.8 vs. 10.6±8.1 ppb) (P>0.05). The CIP group was lower than the non-CIP without comorbidities group (12.0±8.8 vs. 11.0±9.8 ppb) (P≤0.05). There was no significant difference between the CIP without comorbidities group and the non-CIP with comorbidities group (11.2±7.4 vs. 10.6±8.1 ppb) (P>0.05). The CIP without comorbidities group was lower than the non-CIP without comorbidities group (11.2±7.4 vs. 11.0±9.8 ppb) (P≤0.05). There was no significant difference between the CIP with comorbidities group and the non-CIP with comorbidities group (12.9±10.0 vs. 10.6±8.1 ppb) (P>0.05). There was no significant difference between the CIP with comorbidities group and the non-CIP without comorbidities group (12.9±10.0 vs. 11.0±9.8 ppb) (P>0.05), as shown in Figure 3.

Figure 2 eNO concentration in the CIP group and non-CIP group. (A) FeNO50; (B) FeNO200; (C) CaNO. ns, P>0.05; ***, P<0.001 in comparison with two groups. CaNO, alveolar nitric oxide concentration; CIP, checkpoint inhibitor-associated pneumonitis; eNO, exhaled nitric oxide; FeNO, fractional eNO; FeNO50, FeNO at expiratory flow rates of 50 mL/s; FeNO200, FeNO at expiratory flow rates of 200 mL/s; ppb, parts per billion.
Figure 3 eNO concentration in subgroups. (A,D,G) FeNO50; (B,E,H) FeNO200; (C,F,I) CaNO. ns, P>0.05; *, P≤0.05; **, P<0.01; ***, P<0.001 in comparison between two groups. CaNO, alveolar nitric oxide concentration; CIP, checkpoint inhibitor-associated pneumonitis; eNO, exhaled nitric oxide; FeNO, fractional eNO; FeNO50, FeNO at expiratory flow rates of 50 mL/s; FeNO200, FeNO at expiratory flow rates of 200 mL/s; ppb, parts per billion.

Risk factors for the occurrence of CIP

The results of univariate regression analyses showed that emphysema reported on chest CT (OR, 2.24; 95% CI: 1.38–3.65), a small amount of pleural effusion reported on chest CT (OR, 2.11; 95% CI: 1.29–3.45), pre-existing radiotherapy (OR, 4.83; 95% CI: 2.74–8.51), CaNO (OR, 1.27; 95% CI: 1.17–1.38), and the count of lymphocyte cell in peripheral blood (OR, 0.59; 95% CI: 0.40–0.87) were associated with an increased risk of CIP. However, gender, age, severe smoking, obstructive pneumonia, atelectasis, pulmonary embolism, COPD, SAD, usage of bronchodilators, types of ICIs, FeNO50, FeNO200, the count of blood cell (white cells, neutrophils, and eosinophils), C-reactive protein (CRP), and procalcitonin (PCT) were not found to be significant as predictors of ICI-related pneumonitis. In multivariate analysis, CaNO (OR, 1.30; 95% CI: 1.19–1.43, P<0.001), pre-existing radiotherapy (OR, 3.89; 95% CI: 1.96–7.73, P<0.001), a small amount of pleural effusion reported on chest CT (OR, 2.48; 95% CI: 1.37–4.50, P=0.003), and emphysema reported on chest CT (OR, 2.54; 95% CI: 1.41–4.60, P=0.002) were independent predictors associated with CIP in all patient groups. The count of lymphocyte cell in peripheral blood was not statistically different (OR, 0.69; 95% CI: 0.44–1.10, P=0.12) (Table 2). In subgroup analyses, CaNO was identified as an independent predictor associated with CIP in the CIP without comorbidities group vs. non-CIP group (OR, 1.27; 95% CI: 1.15–1.40, P<0.001), in the CIP without comorbidities group vs. the non-CIP with comorbidities group (OR, 1.23; 95% CI: 1.11–1.36, P<0.001), in the CIP without comorbidities group vs. the non-CIP without comorbidities group (OR, 1.39; 95% CI: 1.21–1.60, P<0.001), in the CIP with comorbidities group vs. the non-CIP group (OR, 1.26; 95% CI: 1.14–1.39, P<0.001), in the CIP with comorbidities group vs. the non-CIP with comorbidities group (OR, 0.86; 95% CI: 0.79–0.92, P<0.001), in the CIP with comorbidities group vs. the non-CIP without comorbidities group (OR, 1.41; 95% CI: 1.21–1.64, P<0.001), and in the non-CIP without comorbidities group vs. the non-CIP with comorbidities group (OR, 0.89; 95% CI: 0.81–0.98, P=0.02) (Table 3).

Table 2

Logistic regression analysis of potential for CIP

Variables Univariate analysis Multivariate analysis
OR (95% CI) P value OR (95% CI) P value
Gender (female vs. male) 1.98 (0.99–3.94) 0.052 – –
Age (years) 2.21 (1.36–3.61) 0.55 – –
Severe smoking (no vs. yes) 1.02 (0.49–2.09) 0.96 – –
Complication of lung disease
   Obstructive pneumonia 1.69 (1.03–2.78) 0.059 – –
   Emphysema 2.24 (1.38–3.65) 0.001 2.54 (1.41–4.60) 0.002
   Atelectasis 1.61 (0.77–3.36) 0.20 – –
   Pleural effusion 2.11 (1.29–3.45) 0.003 2.48 (1.37–4.50) 0.003
   Pulmonary embolism 1.04 (0.25–4.23) 0.96 – –
   COPD 0.87 (0.50–1.52) 0.62 – –
   Small airway dysfunction 0.67 (0.26–1.74) 0.41 – –
Bronchodilators 3.24 (0.89–11.75) 0.07 – –
History of radiotherapy 4.83 (2.74–8.51) <0.001 3.89 (1.96–7.73) <0.001
Lung cancer type (squamous vs. non-squamous) 1.51 (0.89–2.54) 0.12 – –
Therapeutic regimen (1st–2nd lines vs. 3rd line & above) 1.76 (0.90–3.43) 0.09 – –
Treatment with ICI (combination vs. monotherapy) 1.58 (0.69–3.64) 0.28 – –
FeNO50 1.00 (0.99–1.01) 0.93 – –
FeNO200 1.02 (0.99–1.04) 0.25 – –
CaNO (ppb) 1.27 (1.17–1.38) <0.001 1.30 (1.19–1.43) <0.001
Blood tests, median (range)
   White blood cell (×109/L) 1.00 (0.94–1.07) 0.99 – –
   Neutrophils (×109/L) 1.03 (0.96–1.11) 0.36 – –
   Lymphocyte (×109/L) 0.59 (0.40–0.87) 0.007 0.69 (0.44–1.10) 0.12
   Eosinophils (×109/L) 0.31 (0.08–1.20) 0.09 – –
   CRP (mg/L) 1.00 (1.00–1.01) 0.28 – –
   PCT (ng/mL) 2.91 (0.91–9.27) 0.07 – –

CaNO, alveolar concentration of nitric oxide; CI, confidence interval; CIP, checkpoint inhibitor-associated pneumonitis; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; eNO, exhaled nitric oxide; FeNO, fractional eNO; FeNO50, FeNO at expiratory flow rates of 50 mL/s; FeNO200, FeNO at expiratory flow rates of 200 mL/s; ICI, immune checkpoint inhibitor; OR, odds ratio; PCT, procalcitonin; ppb, parts per billion.

Table 3

Univariate analysis of CaNO for CIP in different groups

Groups Beta SE Z OR (95% CI) P value
CIP vs. non-CIP 0.24 0.04 5.82 1.27 (1.17–1.38) <0.001
CIP without comorbidities vs. non-CIP 0.24 0.05 4.91 1.27 (1.15–1.40) <0.001
CIP without comorbidities vs. non-CIP with comorbidities 0.2 0.05 4.01 1.23 (1.11–1.36) <0.001
CIP without comorbidities vs. non-CIP without comorbidities 0.33 0.07 4.59 1.39 (1.21–1.60) <0.001
CIP with comorbidities vs. non-CIP 0.23 0.05 4.53 1.26 (1.14–1.39) <0.001
CIP with comorbidities vs. non-CIP with comorbidities −0.16 0.04 −3.99 0.86 (0.79–0.92) <0.001
CIP with comorbidities vs. non-CIP without comorbidities 0.34 0.08 4.46 1.41 (1.21–1.64) <0.001
Non-CIP without comorbidities vs. non-CIP with comorbidities −0.12 0.05 −2.33 0.89 (0.81–0.98) 0.02

CaNO, alveolar concentration of nitric oxide; CI, confidence interval; CIP, checkpoint inhibitor-associated pneumonitis; OR, odds ratio; SE, standard error.

Prediction power and threshold of CaNO to identify CIP

From the results of multivariate regression analysis, we knew that CaNO was an independent risk factor for CIP, but whether it was able to distinguish CIP patients was not clear. We further adopted the ROC curve to analyze the ability of CaNO to distinguish CIP and to determine the optimal cut-off value of CaNO. The results showed that CaNO possesses the capacity to recognize CIP with an AUC of 0.728 (95% CI: 0.670–0.786; P=0.001). The optimal threshold of CaNO for predicting the occurrence of CIP by Youden index was 6.350 ppb. In subgroup analyses, the capacity of CaNO to recognize CIP also had a good performance (Figure 4). The AUC in the CIP without comorbidities group vs. non-CIP group was 0.725 (95% CI: 0.648–0.802; P<0.001), in CIP without comorbidities group vs. the non-CIP with comorbidities group was 0.691 (95% CI: 0.606–0.776; P<0.001), in the CIP without comorbidities group vs. the non-CIP without comorbidities group was 0.714 (95% CI: 0.590–0.830; P<0.001), in the CIP with comorbidities group vs. the non-CIP group was 0.723 (95% CI: 0.648–0.798; P<0.001), in the CIP with comorbidities group vs. the non-CIP with comorbidities group was 0.685 (95% CI: 0.601–0.769; P<0.001), and in the CIP with comorbidities group vs. the non-CIP without comorbidities group was 0.787 (95% CI: 0.707–0.868; P<0.001), as shown in Table 4.

Figure 4 Prediction power of CaNO for occurrence of CIP in different subgroups. The optimal cutoff value of CaNO by ROC curve analysis with AUC =0.728 in the CIP group and the non-CIP group. (A) ROC curve of CaNO in the CIP without comorbidities group and different non-CIP subgroups; (B) ROC curve of CaNO in the CIP with comorbidities group and different non-CIP subgroups. AUC, area under the curve; CaNO, alveolar nitric oxide concentration; CI, conffdence interval; CIP, checkpoint inhibitor-associated pneumonitis; ROC, receiver operating characteristic.

Table 4

The AUC and the optimal threshold of CaNO recognizing CIP in different groups

Groups AUC 95% CI P value Optimal threshold (ppb) Specificity Sensitivity
CIP vs. non-CIP 0.724 0.665–0.783 <0.001 6.75 0.768 0.590
CIP without comorbidities vs. non-CIP 0.725 0.648–0.802 <0.001 6.35 0.725 0.627
CIP without comorbidities vs. non-CIP with comorbidities 0.691 0.606–0.776 <0.001 6.55 0.708 0.608
CIP without comorbidities vs. non-CIP without comorbidities 0.714 0.590–0.830 <0.001 6.40 0.714 0.745
CIP with comorbidities vs. non-CIP 0.723 0.648–0.798 <0.001 7.05 0.802 0.592
CIP with comorbidities vs. non-CIP with comorbidities 0.685 0.601–0.769 <0.001 7.05 0.762 0.592
CIP with comorbidities vs. non-CIP without comorbidities 0.787 0.707–0.868 <0.001 5.40 0.714 0.776

AUC, area under the curve; CaNO, alveolar concentration of nitric oxide; CI, confidence interval; CIP, checkpoint inhibitor-associated pneumonitis; ppb, parts per billion.

Establishment and internal validation of a multi-indicator model for CIP

The independent risk factors with statistical significance (P<0.05) were selected from the results of multivariate analysis to establish a multivariate model for identifying CIP. Independent risk factors, the value of CaNO, pre-existing radiotherapy, emphysema reported on chest CT, a small amount of pleural effusion reported on chest CT, and the count of lymphocyte cell in peripheral blood were finally included in the new model. The multivariate model is presented in the form of alignment diagram in Figure 5. The Bootstrap resampling method was used for internal validation of the multi-indicator model, and the calibration curve was used to evaluate the calibration of the model. The multi-indicator model demonstrated great predictive accuracy that the AUC was 0.821 (95% CI: 0. 0.772–0.869) (Figure 6A). The sensitivity and specificity were 0.770 (95% CI: 0.688–0.852) and 0.758 (95% CI: 0.700–0.817), respectively. The positive and negative predictive values (PPV and NPV) were 0.606 (95% CI: 0.521–0.691) and 0.872 (0.95% CI: 0.823–0.921), respectively. Good agreements could be found between observed and predicted probabilities in the calibration plot for identifying CIP (Figure 6B).

Figure 5 Presentation of the multivariate model with an alignment diagram for identifying CIP. The first line is a reference line for reading scoring points for each prediction parameter. The cumulative results of the five scores correspond to the total score line. The score on the total score line represents the likelihood that the patient has CIP. CaNO, alveolar nitric oxide concentration; CIP, checkpoint inhibitor-associated pneumonitis; ppb, parts per billion.
Figure 6 Establishment and internal validation of the multivariate risk model for CIP. (A) ROC curve analyzes the multi-indicator model for identifying CIP. The AUC is 0.821; (B) calibration curves for identifying CIP. The diagonal line is the reference line, indicating the probability of an ideal nomogram. AUC, area under the curve; CI, confidence interval; CIP, checkpoint inhibitor-associated pneumonitis; ROC, receiver operating characteristic.

Discussion

In this study, we found that CaNO, a marker of alveolar inflammation, was different between those who developed CIP and those who did not develop CIP among lung cancer patients who received the ICIs, with a statistically significant difference (P<0.001). As far as we know, this is the first study to evaluate the potential value of eNO in CIP. When CIP occurs, activated immune cells to attack normal lung tissues, leading to a series of physiological processes including inflammatory cell infiltration, cascade inflammatory response, and alveolar structural damage (32). In the process of epithelial or endothelial injury healing and repair, some cells (fibroblasts, epithelial cells, macrophages and other inflammatory cells) can produce NO through different pathways, so NO is regarded as a key mediator in the process of injury repair (33,34). Larsen et al. (35) systematically detailed the histopathologic features of 9 patients with CIP and they found foamy macrophages accumulation was one of the essential characteristics. The accumulation of macrophages in the lung tissue of CIP patients may produce large amounts of NO, which may lead to increased alveolar NO.

CIP is a feared complication of ICIs therapy. When in the acute phase, patients develop acute lung injury as a status dominated by inflammation and exudation, and if it is not well controlled, CIP may enter a chronic phase as fibrous tissue hyperplasia or even transform into interstitial pneumonia (36). In this prospective study which explored eNO parameters at different flow rates, we found that CaNO, a marker of inflammation in the alveolar area, was able to discriminate between CIP and non-CIP according to the CaNO cut-off of 6.350 ppb. Many clinical studies have investigated the potential value of eNO as a parameter in the differential diagnosis of ILD, and showed a good differential effect (37-39). The mean of CaNO in CIP patients was 8.1±5.0 ppb in this study. This result was consistent with that of CaNO in other ILDs. Tiev et al. reported that CaNO was 6.96±5.5 ppb in systemic sclerosis (SSc) patients and found that CaNO can predict the event of future lung function deterioration or death with a CaNO cut-off at 5.3 ppb (22). Oishi et al. displayed that the values of CaNO in acute cryptogenic organizing pneumonia (COP) patients was 6.2 (3.9–8.6) ppb and acute hypersensitivity pneumonitis was 5.8 (4.5–6.2) ppb (39). Another study showed a higher CaNO levels in patients with ILD: connective tissue disease-related ILD (CTD-ILD) was 15.6±7.2 ppb, idiopathic pulmonary fibrosis was 10.9±5.4 ppb and nonspecific interstitial pneumonia was 10.1±5.1 ppb (20). Clinicians should be alert to the occurrence of CIP when encountering patients who are using ICIs and have high CaNO.

In addition, this study showed that pre-existing radiotherapy was an independent predictor of CIP. This result is consistent with previous studies (40-44). In the phase III clinical PACIFIC trial (gov number, NCT02125461), patients with lung cancer who received durvalumab after chemoradiotherapy had a higher rate of CIP than those who received placebo (33.9% vs. 24.8%) (45). Oxidative damage of DNA and proteins happened when lung tissue was irradiated with radiation, which induced the release of tumor antigens and inflammatory factors. The process induced more cytokines and inflammatory cells to accumulate in alveolar space and produce inflammatory response (46). ICIs may also enhance the response of anti-tumor immunity by inducing lymphocyte differentiation and up-regulating the levels of cytokines and autoantibodies, resulting in higher levels of cytokines and more immune cells entering the irradiated lung tissue (47). In the aspect of clinical characteristics, we found that emphysema and a small amount of pleural effusion reported and on chest CT was the risk factor of CIP, which was consistent with previous studies (32,48,49). However, many CIP patients rarely experience dyspnea and other symptoms due to pleural effusion. The correlation between pleural effusion reported by chest CT and CIP still needs more evidence to confirm. Finally, CaNO combined with the other clinical features were included to construct a multi-indicator predictive model. Outstanding calibration and discrimination were shown with internal validation in the model’s predictive performance, with the C-statistic of 0.808. This model may be a simple and effective tool to identify CIP and assist clinicians in monitoring the adverse reactions of ICIs.

In conclusion, CaNO has great potential value of application in identifying CIP. Patients with CaNO elevation, previous history of lung radiotherapy, emphysema reported on chest CT, a small amount of pleural effusion reported on chest CT, and low lymphocyte cell count in peripheral blood should be alert to the occurrence of CIP during immunotherapy. For clinicians, monitoring distal alveolar inflammation during the use of ICIs may enable early to detecting CIP and giving prompt treatment. Previous studies have constructed several CIP-related models based on clinical features and cytokines (16-18), but cytokines are not universally detected at present and obtained from blood samples by non-invasive means. For example, Chao et al. established a nomogram model to predict CIP involved in baseline plasma interleukin-8 (IL-8) levels non-routinely detected in clinical diagnosis (17). This study explored the potential diagnostic role of eNO testing for CIP, but there are still some limitations. We only collected the data of eNO at the onset of CIP, and did not collect the values of eNO before and after of CIP. As a result, we were not able to compare the differences in CIP across periods longitudinally, and more prospective data are needed for verification. The universality of the CaNO threshold of this study is still in doubt. The best CaNO cut-off value in this study is based on the data of a single center, and the differences such as region and race have not been considered. The direct application of different populations may exist risk, and the best CaNO cut-off value needs to be further determined in a larger population. More prospective research data are needed to explore in the future. At the same time, the model we constructed has not been verified in an external population, and a larger population is still needed to verify the performance of the model in the future. In addition, the population we included may not be representative of all patients using ICIs in a single-center, and more multi-center studies are warranted in the future. Another limitation is that multiparameter eNO technologies are not widespread. Comprehensive trace monitoring of patients on immunosuppressive agents using CaNO may currently be difficult to achieve. In the future, the development and generalization of this technology are still needed to obtain more data to analyze CaNO for painless identification of CIP patients.

Conclusions

CaNO may become a new potential marker for identifying CIP. Routine monitoring of noninvasive eNOs may help to distinguish CIP. The CaNO-based multi-indicator model may help clinicians to identify CIP patients and give timely treatment to reduce the risk of ICIs-related pulmonary adverse events.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2024-1085/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2024-1085/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2024-1085/prf

Funding: This work was supported by grants from the National Natural Science Foundation of China (Nos. 82470058 and 82270024).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2024-1085/coif). Y.G. reports that the detection instrument used in this study was the Nano Coulomb Breath Analyzer (Sunvou-CA2122, Wuxi, China), which was provided free of charge by Shangwo Medical Electronics Co., Ltd., Wuxi, China. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study complied with the rules of the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Nanfang Hospital (No. NFEC-2021-397 & No. NFEC2023-544). All study participants provided written informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

  1. Ai L, Chen J, Yan H, et al. Research Status and Outlook of PD-1/PD-L1 Inhibitors for Cancer Therapy. Drug Des Devel Ther 2020;14:3625-49. [Crossref] [PubMed]
  2. West H, McCleod M, Hussein M, et al. Atezolizumab in combination with carboplatin plus nab-paclitaxel chemotherapy compared with chemotherapy alone as first-line treatment for metastatic non-squamous non-small-cell lung cancer (IMpower130): a multicentre, randomised, open-label, phase 3 trial. Lancet Oncol 2019;20:924-37. [Crossref] [PubMed]
  3. Kang YK, Chen LT, Ryu MH, et al. Nivolumab plus chemotherapy versus placebo plus chemotherapy in patients with HER2-negative, untreated, unresectable advanced or recurrent gastric or gastro-oesophageal junction cancer (ATTRACTION-4): a randomised, multicentre, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2022;23:234-47.
  4. Ni J, Wang X, Wu L, et al. Sintilimab in combination with stereotactic body radiotherapy and granulocyte-macrophage colony-stimulating factor in metastatic non-small cell lung cancer: The multicenter SWORD phase 2 trial. Nat Commun 2024;15:7242. [Crossref] [PubMed]
  5. Chan KK, Bass AR. Autoimmune complications of immunotherapy: pathophysiology and management. BMJ 2020;369:m736. [Crossref] [PubMed]
  6. Khunger M, Rakshit S, Pasupuleti V, et al. Incidence of Pneumonitis With Use of Programmed Death 1 and Programmed Death-Ligand 1 Inhibitors in Non-Small Cell Lung Cancer: A Systematic Review and Meta-Analysis of Trials. Chest 2017;152:271-81. [Crossref] [PubMed]
  7. Brahmer JR, Abu-Sbeih H, Ascierto PA, et al. Society for Immunotherapy of Cancer (SITC) clinical practice guideline on immune checkpoint inhibitor-related adverse events. J Immunother Cancer 2021;9:e002435. [Crossref] [PubMed]
  8. De Velasco G, Je Y, Bossé D, et al. Comprehensive Meta-analysis of Key Immune-Related Adverse Events from CTLA-4 and PD-1/PD-L1 Inhibitors in Cancer Patients. Cancer Immunol Res 2017;5:312-8. [Crossref] [PubMed]
  9. Naidoo J, Wang X, Woo KM, et al. Pneumonitis in Patients Treated With Anti-Programmed Death-1/Programmed Death Ligand 1 Therapy. J Clin Oncol 2017;35:709-17. [Crossref] [PubMed]
  10. Cho JY, Kim J, Lee JS, et al. Characteristics, incidence, and risk factors of immune checkpoint inhibitor-related pneumonitis in patients with non-small cell lung cancer. Lung Cancer 2018;125:150-6. [Crossref] [PubMed]
  11. Suresh K, Voong KR, Shankar B, et al. Pneumonitis in Non-Small Cell Lung Cancer Patients Receiving Immune Checkpoint Immunotherapy: Incidence and Risk Factors. J Thorac Oncol 2018;13:1930-9. [Crossref] [PubMed]
  12. Chu X, Zhao J, Zhou J, et al. Association of baseline peripheral-blood eosinophil count with immune checkpoint inhibitor-related pneumonitis and clinical outcomes in patients with non-small cell lung cancer receiving immune checkpoint inhibitors. Lung Cancer 2020;150:76-82. [Crossref] [PubMed]
  13. Wang DY, Salem JE, Cohen JV, et al. Fatal Toxic Effects Associated With Immune Checkpoint Inhibitors: A Systematic Review and Meta-analysis. JAMA Oncol 2018;4:1721-8. Erratum in: JAMA Oncol 2018;4:1792. [Crossref] [PubMed]
  14. Lim SY, Lee JH, Gide TN, et al. Circulating Cytokines Predict Immune-Related Toxicity in Melanoma Patients Receiving Anti-PD-1-Based Immunotherapy. Clin Cancer Res 2019;25:1557-63. [Crossref] [PubMed]
  15. Lin X, Deng H, Yang Y, et al. Peripheral Blood Biomarkers for Early Diagnosis, Severity, and Prognosis of Checkpoint Inhibitor-Related Pneumonitis in Patients With Lung Cancer. Front Oncol 2021;11:698832. [Crossref] [PubMed]
  16. Deng H, Deng J, Lin X, et al. A Risk-Scoring Model for Severe Checkpoint Inhibitor-Related Pneumonitis: A Case-Control Study. Clin Drug Investig 2023;43:347-57. [Crossref] [PubMed]
  17. Chao Y, Zhou J, Hsu S, et al. Risk factors for immune checkpoint inhibitor-related pneumonitis in non-small cell lung cancer. Transl Lung Cancer Res 2022;11:295-306. [Crossref] [PubMed]
  18. Li X, Lv F, Wang Y, et al. Establishment and validation of nomogram for predicting immuno checkpoint inhibitor related pneumonia. BMC Pulm Med 2022;22:331. [Crossref] [PubMed]
  19. Kashfi K, Kannikal J, Nath N. Macrophage Reprogramming and Cancer Therapeutics: Role of iNOS-Derived NO. Cells 2021;10:3194. [Crossref] [PubMed]
  20. Cameli P, Bargagli E, Bergantini L, et al. Extended Exhaled Nitric Oxide Analysis in Interstitial Lung Diseases: A Systematic Review. Int J Mol Sci 2020;21:6187. [Crossref] [PubMed]
  21. Tsoukias NM, George SC. A two-compartment model of pulmonary nitric oxide exchange dynamics. J Appl Physiol (1985) 1998;85:653-66. [Crossref] [PubMed]
  22. Tiev KP, Hua-Huy T, Kettaneh A, et al. Alveolar concentration of nitric oxide predicts pulmonary function deterioration in scleroderma. Thorax 2012;67:157-63. [Crossref] [PubMed]
  23. Cameli P, Bergantini L, Salvini M, et al. Alveolar concentration of nitric oxide as a prognostic biomarker in idiopathic pulmonary fibrosis. Nitric Oxide 2019;89:41-5. [Crossref] [PubMed]
  24. Van Muylem A, Kerckx Y, Michils A. Acinar effect of inhaled steroids evidenced by exhaled nitric oxide. J Allergy Clin Immunol 2010;126:730-735.e2. [Crossref] [PubMed]
  25. Xiao D, Chen Z, Wu S, et al. Prevalence and risk factors of small airway dysfunction, and association with smoking, in China: findings from a national cross-sectional study. Lancet Respir Med 2020;8:1081-93. [Crossref] [PubMed]
  26. Agustí A, Celli BR, Criner GJ, et al. Global Initiative for Chronic Obstructive Lung Disease 2023 Report: GOLD Executive Summary. Eur Respir J 2023;61:2300239. [Crossref] [PubMed]
  27. Kharitonov SA, Yates DH, Barnes PJ. Inhaled glucocorticoids decrease nitric oxide in exhaled air of asthmatic patients. Am J Respir Crit Care Med 1996;153:454-7. [Crossref] [PubMed]
  28. Singh S, Gupta AK. Nitric oxide: role in tumour biology and iNOS/NO-based anticancer therapies. Cancer Chemother Pharmacol 2011;67:1211-24. [Crossref] [PubMed]
  29. Horváth I, Barnes PJ, Loukides S, et al. A European Respiratory Society technical standard: exhaled biomarkers in lung disease. Eur Respir J 2017;49:1600965. [Crossref] [PubMed]
  30. Gelb AF, Taylor CF, Nussbaum E, et al. Alveolar and airway sites of nitric oxide inflammation in treated asthma. Am J Respir Crit Care Med 2004;170:737-41. [Crossref] [PubMed]
  31. Fan X, Zhao N, Yu Z, et al. Clinical Utility of Central and Peripheral Airway Nitric Oxide in Aging Patients with Stable and Acute Exacerbated Chronic Obstructive Pulmonary Disease. Int J Gen Med 2021;14:571-80. [Crossref] [PubMed]
  32. Yin J, Wu Y, Yang X, et al. Checkpoint Inhibitor Pneumonitis Induced by Anti-PD-1/PD-L1 Therapy in Non-Small-Cell Lung Cancer: Occurrence and Mechanism. Front Immunol 2022;13:830631. [Crossref] [PubMed]
  33. Schwentker A, Vodovotz Y, Weller R, et al. Nitric oxide and wound repair: role of cytokines? Nitric Oxide 2002;7:1-10. [Crossref] [PubMed]
  34. Huffman LJ, Prugh DJ, Millecchia L, et al. Nitric oxide production by rat bronchoalveolar macrophages or polymorphonuclear leukocytes following intratracheal instillation of lipopolysaccharide or silica. J Biosci 2003;28:29-37.
  35. Larsen BT, Chae JM, Dixit AS, et al. Clinical and Histopathologic Features of Immune Checkpoint Inhibitor-related Pneumonitis. Am J Surg Pathol 2019;43:1331-40. [Crossref] [PubMed]
  36. Zhou C, Yang Y, Lin X, et al. Proposed clinical phases for the improvement of personalized treatment of checkpoint inhibitor-related pneumonitis. Front Immunol 2022;13:935779. [Crossref] [PubMed]
  37. Cameli P, Bargagli E, Bergantini L, et al. Evaluation of multiple-flows exhaled nitric oxide in idiopathic and non-idiopathic interstitial lung disease. J Breath Res 2019;13:026008. [Crossref] [PubMed]
  38. Krauss E, Froehler M, Degen M, et al. Exhalative Breath Markers Do Not Offer for Diagnosis of Interstitial Lung Diseases: Data from the European IPF Registry (eurIPFreg) and Biobank. J Clin Med 2019;8:643. [Crossref] [PubMed]
  39. Oishi K, Hirano T, Suetake R, et al. Exhaled nitric oxide measurements in patients with acute-onset interstitial lung disease. J Breath Res 2017;11:036001. [Crossref] [PubMed]
  40. Sul J, Blumenthal GM, Jiang X, et al. FDA Approval Summary: Pembrolizumab for the Treatment of Patients With Metastatic Non-Small Cell Lung Cancer Whose Tumors Express Programmed Death-Ligand 1. Oncologist 2016;21:643-50. [Crossref] [PubMed]
  41. Myers CJ, Lu B. Decreased Survival After Combining Thoracic Irradiation and an Anti-PD-1 Antibody Correlated With Increased T-cell Infiltration Into Cardiac and Lung Tissues. Int J Radiat Oncol Biol Phys 2017;99:1129-36. [Crossref] [PubMed]
  42. Hwang WL, Niemierko A, Hwang KL, et al. Clinical Outcomes in Patients With Metastatic Lung Cancer Treated With PD-1/PD-L1 Inhibitors and Thoracic Radiotherapy. JAMA Oncol 2018;4:253-5. [Crossref] [PubMed]
  43. Shibaki R, Murakami S, Matsumoto Y, et al. Association of immune-related pneumonitis with the presence of preexisting interstitial lung disease in patients with non-small lung cancer receiving anti-programmed cell death 1 antibody. Cancer Immunol Immunother 2020;69:15-22. [Crossref] [PubMed]
  44. Kanai O, Kim YH, Demura Y, et al. Efficacy and safety of nivolumab in non-small cell lung cancer with preexisting interstitial lung disease. Thorac Cancer 2018;9:847-55. [Crossref] [PubMed]
  45. Antonia SJ, Villegas A, Daniel D, et al. Overall Survival with Durvalumab after Chemoradiotherapy in Stage III NSCLC. N Engl J Med 2018;379:2342-50. [Crossref] [PubMed]
  46. Hwang WL, Pike LRG, Royce TJ, et al. Safety of combining radiotherapy with immune-checkpoint inhibition. Nat Rev Clin Oncol 2018;15:477-94. [Crossref] [PubMed]
  47. Deng L, Liang H, Burnette B, et al. Irradiation and anti-PD-L1 treatment synergistically promote antitumor immunity in mice. J Clin Invest 2014;124:687-95. [Crossref] [PubMed]
  48. Shibata Y, Murakami S, Kato T. Overview of checkpoint inhibitor pneumonitis: incidence and associated risk factors. Expert Opin Drug Saf 2021;20:537-47. [Crossref] [PubMed]
  49. Cui P, Liu Z, Wang G, et al. Risk factors for pneumonitis in patients treated with anti-programmed death-1 therapy: A case-control study. Cancer Med 2018;7:4115-20. [Crossref] [PubMed]
Cite this article as: Gao Y, Luo T, Huang D, Fu Z, Ma S, Lin L, Huang H, Liu T, Zhang J, Jiang X, Ye Y, Chen J, Xi J, Zhuo J, Chen K, Ai J, Liu L, Cai S, Dong H. Construction of a checkpoint inhibitor-related pneumonia diagnostic model based on exhaled nitric oxide: a prospective observational study. Transl Lung Cancer Res 2025;14(5):1740-1755. doi: 10.21037/tlcr-2024-1085

Article Options

Download Citation

Share