Journal Information
Share
Share
Download PDF
More article options
Visits
...
Original article
Open Access
Available online 10 April 2022
Predictors of immune-related adverse events and outcomes in patients with NSCLC treated with immune-checkpoint inhibitors
Visits
...
M. Serinoa,
Corresponding author
mariana.serino@gmail.com

Corresponding author at: Centro Hospitalar Universitário de São João: Centro Hospitalar Universitario de Sao Joao, Portugal.
, C. Freitasc, M. Martinsa, P. Ferreiraa, C. Cardosoa, F. Veigab, V. Santosa, D. Araújoa, H. Novais-Bastosc, A. Magalhãesa, H. Queirogac, G. Fernandesc, V. Hespanholc
a Pulmonology Department, Centro Hospitalar Universitário São João, Porto, Portugal
b Faculty of Medicine, University of Porto, Porto, Portugal
c Pulmonology Department, Centro Hospitalar Universitário São João; Faculty of Medicine, University of Porto, Porto, Portugal
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Figures (1)
Tables (4)
Table 1. General characteristics of patients.
Table 2. Immune-related adverse events associated by type of ICI.
Table 3. Univariate and multivariate analyses to determine risk factors for IRAEs.
Table 4. Univariate and multivariate analyses of PFS and OS.
Show moreShow less
Additional material (1)
Abstract
Objective

To identify predictors of immune-related adverse events (IRAEs) in patients with non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs). Assess associations between outcomes and the development of IRAEs.

Methods

Retrospective analysis of patients with NSCLC treated with ICIs between 2016 and 2020 in the Pulmonology Department of our hospital. Patients with and without IRAEs were compared. A logistic regression analysis was performed to determine predictors of IRAEs. Progression-free survival (PFS) and overall survival (OS) curves were calculated using the Kaplan-Meier method, and the long-rank test was used to assess survival differences between groups. Univariate and multivariate Cox proportional-hazards regression models were used to identify factors associated with PFS and OS. The value considered statistically significant was p≤0.05.

Results

A total of 184 patients (77.7% men, mean age 66.9±9.5 years) treated with ICIs were analyzed. During follow-up, 49 (26.6%) patients developed IRAEs and 149 (81.0%) died. According to the multivariate logistic regression analysis, treatment with statins (OR:3.15; p = 0.007), previous systemic corticosteroid therapy (OR:3.99; p = 0.001), disease controlled as response to ICI (OR:5.93; p < 0.001) and higher hemoglobin values (OR:1.28; p = 0.040) were independent predictors for the development of IRAEs. Patients who developed IRAEs had significantly longer medians of PFS (41.0 vs 9.0 weeks, p < 0.001) and OS (89.0 vs 28.0 weeks; p < 0.001).

Conclusions

Patients treated with statins, pre-ICI systemic corticosteroids, higher baseline hemoglobin value and controlled disease as initial response to ICI had a higher risk of developing IRAEs. The development of IRAEs was associated with better outcomes.

KEYWORDS:
NSCLC
immune checkpoint Inhibitors
Immune-related adverse events
outcomes
Full Text
Introduction

Non-small cell lung cancer (NSCLC) accounts for the majority of all lung cancers,1,2 which in most cases present with metastatic disease at diagnosis1. Lung cancer is the leading cause of cancer deaths worldwide.1,2 Advances in understanding the biology of cancer and the mechanisms of oncogenesis have showed that NSCLC is a heterogeneous disease and multiple molecular and immunohistochemical targets have been recognized.3

Immune checkpoint inhibitors (ICIs) targeting either programmed cell death protein 1 (PD-1) or programmed cell death ligand 1 (PD-L1) have become part of the clinical approach for management of NSCLC.2,4,5 ICIs are indicated as first-line treatment in patients with advanced NSCLC harboring high PD-L1 expression (PD-L1≥50%) in absence of targetable drive oncogene molecular variant, or in combination with platinum-based chemotherapy in patients with advanced NSCLC and PD-L1<50%.4,5

ICIs are associated with a specific spectrum of immune-related adverse events (IRAEs),6-9 by increasing the activity of the immune system.7,8 IRAEs can involve any organ or system, most commonly skin, colon, lungs, endocrine glands and liver.6-9 Most IRAEs are self-limiting or resolve with systemic corticosteroids, but in some cases, these can be severe and life-threatening and require permanent discontinuation of treatment.6,7

Currently, the pathophysiological mechanisms of IRAEs are poorly understood6,7 and biomarkers and/or patient risk factors that predict toxicity are not recognized. However, with the increasing use of immunotherapy, it is important to identify patients at increased risk for IRAEs in order to better predict and manage them.

Therefore, in this retrospective cohort study, we aimed to identify predictors of IRAEs in patients with NSCLC treated with ICIs. In addition, we assessed associations between outcomes, including progression-free survival (PFS) and overall survival (OS), and the development of IRAEs.

Material and methodsStudy population

A retrospective analysis was performed of patients with NSCLC treated with antagonist monoclonal antibodies against programmed death-1 (PD-1) (Nivolumab and Pembrolizumab) and the programmed death ligand-1 (PD-L1) (Atezolizumab), between 2016 and 2020, in the Department of Pulmonology at the Centro Hospitalar Universitário de São João, Portugal. Patient data were entirely anonymized and authorized by the body Responsible for Access to Information (RAI) of Centro Hospitalar Universitário de São João. The registration protocol complies with the ethical guidelines of the Declaration of Helsinki and it was approved by the Ethics and Health Committee of Centro Hospitalar Universitário de São João on July 7, 2021.

Eligibility criteria were age ≥18 years, histological or cytological confirmation of NSCLC and treatment with ICI according to clinical practice. A minimum follow-up time of 4 weeks after starting the ICI was required for inclusion in the study. All patients with metastatic NSCLC and a targetable driver oncogene molecular variant (EGFR, ALK, or BRAF) had previously received targeted therapy for this oncogene prior to treatment with ICI.

Pembrolizumab and Atezolizumab were administered intravenously (iv) every 3 weeks at doses of 200 mg and 1200 mg, respectively. Nivolumab was administered iv at a dose of 240 mg every 2 weeks.

Data collection

Demographic and clinical characteristics of patients before starting ICIs were reviewed, including smoking history, comorbidities, systemic corticosteroid therapy in the 3 months prior to ICIs, Eastern Cooperative Oncology Group (ECOG) performance status (PS), tumor histology, molecular results and PD-L1 expression level (when available), disease stage, previous treatments, and baseline blood laboratory results. Baseline blood laboratory results were defined as the most recent (within 2 weeks) before the start of ICI, and were used to calculate NLR (absolute neutrophil count/absolute lymphocyte count) and PLR (platelet count/absolute lymphocyte count). The association between NLR≥5 and PLR≥180 values and outcomes were tested. These cutoff points were chosen according to literature references.10,11

Study assessments

The primary aim was to determine predictors of development of IRAEs. IRAEs were considered as adverse events related to an immune dysregulation that required specific monitoring or treatment. The Common Terminology Criteria for Adverse Events (CTACAE) of the National Institute (version 4.03) was used to assess patients’ adverse events. Based on IRAEs, patients were divided into two groups (IRAEs group and non-IRAEs group). Demographic and clinical characteristics and blood laboratory results were compared between groups.

ICIs were maintained until evidence of disease progression, death, severe or life-threatening IRAEs. Disease control rate (DCR) and pattern of disease progression during treatment with ICI were assessed. The Response Evaluation Criteria in Solid Tumors (RECIST) was applied for the evaluation of disease response. RECIST may underestimate the therapeutic benefit of ICIs, due to low accuracy in detecting pseudoprogression.4 However, in all cases, considering clinical and radiological findings, progression and DCR were admitted after consensus in a multidisciplinary lung cancer team meeting. DCR was defined as the proportion of participants with complete and partial response or disease stability for at least 6-8 weeks from onset of ICI.

The secondary objective was to assess the impact of developing IRAEs on PFS and OS. Progression-free survival (PFS) was calculated as the time from the start of ICI treatment to the date of radiographic or clinical progression or patient death. Overall survival (OS) was calculated from the time of initiation of the ICI until death from any cause or last follow-up.

Statistical analysis

Categorical variables are presented as frequencies and percentages, and continuous variables as means and standard deviations (SD), or medians and interquartile ranges (IQR) for variables with skewed distribution. Normal distribution was tested using skewness and kurtosis. We used the chi-square test to compare categorical variables. Independent-samples t-test was used to evaluate differences in continuous variables with normal distribution and Mann-Whitney U tests were used to evaluate differences in continuous variables with skewed distribution. A univariate and multivariate binary logistic regression analysis were performed to determine predictors of development of IRAEs. PFS and OS curves were calculated using the Kaplan-Meier method, and the long-rank test was used to assess survival differences between groups. Univariate and multivariate Cox proportional hazards regression models were used to identify factors associated with PFS and OS. Factors which were statistically significant in the univariable/unadjusted analysis were incorporated into the multivariable/adjusted analysis. The p-value considered for statistical significance was 0.05. Data was stored and all statistical analyses were performed using the Statistical Package for Social Sciences (SPSS, IBM Corp, Chicago, IL, USA) software, version 26.0.

Results

A total of 184 patients with NSCLC treated with ICIs were included. Patients had a mean age of 66.9±9.5 years and the majority were male (77.7%), current or former smokers (83.5%) and had an ECOG-PS of 0-1 (82.1%), non-squamous histology (67.4%), stage IV disease (98.4%) and PD-L1 expression level ≥1 (Table 1). Most patients were treated with ICIs in first- or second-line, 70 (38.0%) patients were treated with Pembrolizumab, 77 (41.8%) with Nivolumab and 37 (20.1%) with Atezolizumab (Table 1).

Table 1.

General characteristics of patients.

Characteristics  Non-IRAEs group (n = 135)  IRAEs group (n = 49)  All patients (n = 184)  p-value 
Age (years)  66.6±9.6  67.7±9.6  66.9±9.5  0.486 
Sex, n (%)        0.242 
Male  102 (75.6)  41 (83.7)  143 (77.7)   
Female  33 (24.4)  8 (16.3)  41 (22.3)   
ECOG-PS, n (%)        0.437 
0-1  109 (80.7)  42 (85.7)  151 (82.1)   
≥2  26 (19.3)  7 (14.3)  33 (17.9)   
Smoking status, n (%)        0.187 
Current or former  109 (81.3)  43 (89.6)  152 (83.5)   
Never  25 (18.7)  5 (10.4)  30 (16.5)   
Comorbidities, n (%)         
Dyslipidemia/Statin treatment  41 (30.4))  24 (49.0)  65 (35.3)  0.020 
Hypertension  55 (40.7)  24 (49.0)  79 (42.9)  0.318 
Diabetes  22 (16.3)  10 (20.4)  32 (17.4)  0.515 
COPD  22 (16.3)  12 (24.5)  34 (18.5)  0.206 
Previous autoimmune disease  3 (2.3)  2 (4.2)  5 (2.7)  0.478 
Previous systemic corticotherapy*  39 (28.9)  25 (51.0)  64 (34.8)  0.005 
Histology, n (%)        0.728 
Adenocarcinoma  90 (66.7)  34 (69.4)  124 (67.4)   
Squamous cell carcinoma  45 (33.3)  15 (30.6)  60 (32.6)   
PD-L1 expression level, n (%)        0.801 
PD-L1 negative (<1%)  20/93 (21.5)  7/34 (20.6)  27/127 (21.3)   
PD-L1 low (1-49%)  35/93 (37.6)  11/34 (32.4)  46/127 (36.2)   
PD-L1 high (≥50%)  38/93 (40.9)  16/34 (47.1)  54/127 (42.5)   
Oncogenic driver mutations, n (%)  29 (42.0)  8 (32.0)  37 (39.4)  0.379 
Disease stage IV, n (%)  133 (98.5)  48 (98.0)  181 (98.4)  0.791 
Patients’ treatments, n (%)
Pembrolizumab  50 (37.0))  20 (40.8)  70 (38.0)  0.641 
Nivolumab  55 (40.7)  22 (44.9)  77 (41.8)  0.613 
Atezolizumab  30 (22.2)  7 (14.3)  37 (20.1)  0.235 
Chemotherapy prior to ICI  113 (83.7)  41 (83.7)  154 (83.7)  0.996 
Thoracic radiotherapy during ICI  20 (14.8)  6 (12.2)  26 (14.1)  0.658 
Line of ICI, n (%)        0.090 
1-2  90 (66.7)  39 (79.6)  129 (70.1)   
≥ 3  45 (33.3)  10 (20.4)  55 (29.9)   
Disease response to ICI         
Disease controlled, n (%)  52 (46.4)  35 (77.8)  87 (55.4)  <0.001 
Time between last therapeutic line and onset of ICI (weeks)  9 (4.0-26.0)  18.5 (6.8-38.3)  11 (5.0-30.0)  0.020 
Time between diagnosis and onset of ICI (weeks)  55 (26.0-91.0)  66 (26.0-117.0)  57 (26.0-93.5)  0.480 
Hemoglobin (g/dL)  12.2 (10.6-13.3)  12.6 (11.6-13.8)  12.4 (10.7-13.4)  0.012 
Leukocytes (cells/µL)  8.5 (6.2-11.9)  8.5 (6.4-11.3)  8.5 (6.4-11.7)  0.807 
Neutrophils (cells/µL)  5.9 (3.8-9.0)  5.6 (4.2-7.5)  5.8 (3.8-8.5)  0.600 
Lymphocytes (cells/µL)  1.4 (1.0-2.0)  1.5 (1.1-2.1)  1.4 (1.0-2.0)  0.192 
Platelets (cells x103/µL)  264.0 (201.0-351.0)  255.0 (194.0-323.5)  259.5 (199.0-347.5)  0.407 
C-reactive protein (mg/L)  24.8 (9.3-71.1)  9.6 (4.5-27.1)  17.1 (8.1-66.7)  0.013 
LDH (UI/L)  227.0 (190.3-200.5)  187.0 (164.0-220.0)  213.0 (180.5-281.0)  0.012 
NLR ≥ 5, n (%)  57 (42.2)  13 (26.5)  70 (38.0)  0.053 
PLR ≥ 180, n (%)  67 (49.6)  20 (40.8)  87 (47.3)  0.290 
Follow-up (weeks)  28.0 (9.0-62.0)  71.0 (39.0-115.0)  36.5 (13.3-81.3)  <0.001 
Death, n (%)  114 (84.4)  35 (71.4)  149 (81.0)  0.047 

Data are presented as frequencies and percentages for categorical variables, as means and standard deviations (SD) for parametric continuous variables and as medians and interquartile ranges (IQR) for non-parametric continuous variables.

Definition of abbreviations: IRAEs: Immune-related adverse events; ECOG-PS: Eastern Cooperative Oncology Group performance status; COPD: Chronic obstructive pulmonary disease; PD-L1: Programmed death-ligand 1; ICI: Immune checkpoint inhibitors

Previous systemic corticosteroid therapy: defined as systemic corticosteroid therapy within 3 months of onset of ICIs

The median follow-up time of patients after starting the ICI was 36.5 (13.3-81.3) weeks (Table 1). During this period, 49 (26.6%) patients developed IRAEs. The general characteristics of the patients according to the development of IRAEs are shown in table 1. There were no statistically significant differences regarding age, sex, ECOG-PS, PDL-1 expression level, ICI treatment line and burden of comorbidities between groups, except for dyslipidemia and systemic corticosteroid therapy within 3 months of onset of ICIs, which were more prevalent in the group that developed IRAEs. In addition, patients with IRAEs had a significantly higher disease control rate with ICI treatment compared to the group that did not develop IRAEs (p < 0.001).

Regarding baseline blood laboratory results (Table 1), patients with IRAEs had a significantly higher median of hemoglobin (p = 0.012) and eosinophils (p = 0.013) and lower median C-reactive protein (p = 0.013) and LDH (p = 0.012).

There was a similar rate of IRAEs in patients treated with Nivolumab (28.6%) and Pembrolizumab (28.6%), but a lower rate in patients treated with Atezolizumab (18.9%). The distribution of IRAEs by type of ICI is described in table 2.

Table 2.

Immune-related adverse events associated by type of ICI.

Adverse-effect  Nivolumab (n = 22)  Pembrolizumab (n = 20)  Atezolizumab (n = 7)  All (n = 49)  p-value 
Hepatic, n (%)  3 (13.6)  1 (5.0)  0 (0.0)  4 (8.2)  0.413 
Neurologic, n (%)  0 (0.0)  2 (10.0)  0 (0.0)  2 (4.1)  0.221 
Skin-related events, n (%)  2 (9.1)  6 (30.0)  1 (14.3)  9 (18.4)  0.207 
Endocrine, n (%)  8 (36.4)  5 (25.0)  4 (57.1)  17 (34.7)  0.299 
Pneumonitis, n (%)  3 (13.6)  2 (10.0)  2 (28.6)  7 (14.3)  0.478 
Colitis, n (%)  6 (27.3)  4 (20.0)  0 (0.0)  10 (20.4)  0.296 
Arthralgies/myalgies, n (%)  1 (4.5)  4 (20.0)  0 (0.0)  5 (10.2)  0.169 
Severity of the adverse event (AE)
Grade 1, n (%)  3 (13.6)  8 (40.0)  2 (28.6)  13 (26.5)  0.153 
Grade 2, n (%)  13 (59.1)  8 (40.0)  3 (42.9)  24 (49.0)  0.438 
Grade 3, n (%)  4 (18.2)  2 (10.0)  2 (28.6)  8 (16.3)  0.494 
Grade 4, n (%)  2 (9.1)  1 (5.0)  0 (0.0)  3 (6.1)  0.658 
Grade 5, n (%)  0 (0.0)  1 (5.0)  0 (0.0)  1 (2.0)  0.477 
Systemic corticosteroid*, n (%)  15 (68.1)  10 (50.0)  5 (71.4)  30 (61.2)  0.403 
Definitive discontinuation of ICI, n (%)  13 (59.0)  8 (40.0)  2 (28.6)  23 (46.9)  0.267 
Time between start of ICI and IRAEs (weeks)  23.0 (11.3-38.0)  17.5 (5.3-47.8)  4.0 (2.0-25.0)  21.0 (5.5-36.0)  0.060 

Data are presented as frequencies and percentages for categorical variables and as medians and interquartile ranges (IQR) for non-parametric continuous variables.

Definition of abbreviations: ICI: Immune checkpoint inhibitors; IRAEs: Immune-related adverse events.

Systemic corticosteroid required by IRAEs.

The majority (75.5%) of IRAEs were mild or moderate, but there were life-threatening IRAEs in 2 patients treated with Nivolumab and in 1 patient treated with Pembrolizumab and one fatal IRAE in 1 patient treated with Pembrolizumab. Thirty patients required systemic corticosteroid therapy and 23 definitive discontinuations of ICI due to IRAEs. Patients treated with Atezolizumab tended to have earlier IRAEs (p = 0.060), mostly endocrine-IRAEs.

According to the univariate analysis (Table 3), patients with dyslipidemia, systemic corticoid therapy prior to ICI, controlled disease and higher hemoglobin values had a significantly higher risk of developing IRAEs. These associations were confirmed by multivariate logistic regression analysis, showing that dyslipidemia (OR: 3.15; 95% CI: 1.38-7.22, p = 0.007), previous systemic corticosteroid therapy (OR: 3.99; 95% CI: 1.73-9.23, p = 0.001), disease controlled as response to ICI (OR: 5.93; 95% CI: 2.39-14.69, p <0.001) and higher hemoglobin values (OR: 1.28; 95% CI: 1.01-1.63, p = 0.040) were independent predictors for the development of IRAEs (Table 3). Age, sex, ECOG-PS, histology, PD-L1 expression level, oncogenic driver mutations (including KRAS mutation), line of ICI and other analytical parameters were not associated with increased risk of IRAEs. The severity of IRAEs between dyslipidemic and non-dyslipidemic patients was evaluated, a statistically significant association was not found (p=0.456) (Fig. 2 in supplement).

Table 3.

Univariate and multivariate analyses to determine risk factors for IRAEs.

VariableCategoryUnivariateMultivariate
OR  95% CI  p-value  OR  95% CI  p-value 
Sex  Male  1.66  0.71-3.89  0.245       
Age  ≥ 70 years  1.20  0.62-2.32  0.595       
ECOG-PS score  0-1  1.43  0.59-3.55  0.439       
Dyslipidemia/Statin treatment  yes  2.20  1.13-4.30  0.021  3.15  1.38-7.22  0.007 
Previous systemic corticotherapy*  yes  2.56  1.31-5.02  0.006  3.99  1.73-9.23  0.001 
Histology  Adenocarcinoma  1.13  0.56-2.29  0.728       
PD-L1 expression levelNegative (<1%)  0.95  0.36-2.49  0.911       
High (≥50%)  1.29  0.58-2.84  0.532       
Oncogenic driver mutations  yes  0.65  0.25-1.71  0.381       
KRAS mutation  yes  1.35  0.51-3.60  0.548       
Line of ICI  First-line  0.91  0.36-2.28  0.83       
Disease control rate  Controlled disease  4.04  1.82-8.94  0.001  5.93  2.39-14.69  <0.001 
Time between last line and start of ICI  weeks  1.01  1.00-1.02  0.116       
Hemoglobin  g/dl  1.28  1.06-1.55  0.009  1.28  1.01-1.63  0.040 
Eosinophils  cells/µL  1.79  0.66-4.86  0.252       
NLR baseline  ≥5  0.49  0.24-1.02  0.055       
PLR baseline  ≥180  0.70  0.36-1.36  0.291       

Definition of abbreviations: IRAEs: Immune-related adverse events; ECOG-PS: Eastern Cooperative Oncology Group performance status; PD-L1: Programmed death-ligand 1; ICI: Immune checkpoint inhibitor; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; OR: odds ratio; CI: confidence interval.

Previous systemic corticosteroid therapy: defined as systemic corticosteroid therapy within 3 months of onset of ICIs.

During the follow-up period, 149 (81.0%) patients died. Patients with IRAEs had a significantly longer median PFS compared to patients without IRAEs (41.0 weeks vs 9.0 weeks, p < 0.001) (Fig. 1.A). The univariate analysis identified ECOG-PS score of 0-1 (p = 0.002), a higher hemoglobin value (p = 0.001) and the development of IRAEs (p < 0.001) as factors associated with improved PFS; a baseline NLR≥5 (p = 0.030) was associated with worse PFS (Table 4). A higher hemoglobin value (HR: 0.90; 95% CI: 0.82-0.99, p = 0.030) and the development of IRAEs (HR: 0.44; 95% CI: 0.28-0.69, p < 0.001) persisted significantly associated with improved PFS in the multivariate analysis.

Fig. 1.

A Kaplan-Meier curves showing progression-free survival, stratified by groups of patients with and without IRAEs. B Kaplan-Meier curves showing overall survival, stratified by group of patients with IRAEs and group of patients without IRAEs.

(0.24MB).
Table 4.

Univariate and multivariate analyses of PFS and OS.

VariableCategoryPFSOS
UnivariateMultivariateUnivariateMultivariate
HR  95% CI  p-value  HR  95% CI  p-value  HR  95% CI  p-value  HR  95% CI  p-value 
Sex  Male  0.93  0.63-1.38  0.715        0.97  0.67-1.43  0.893       
Age  ≥ 70 years  0.83  0.59-1.17  0.278        0.84  0.60-1.16  0.287       
ECOG-PS score  0-1  0.50  0.32-0.77  0.002  0.63  0.39-1.00  0.052  0.54  0.36-0.82  0.004  1.42  0.68-2.98  0.353 
Dyslipidemia  yes  1.20  0.84-1.70  0.324        0.97  0.69-1.37  0.858       
Previous systemic corticotherapy*  yes  1.05  0.74-1.49  0.777        1.30  0.94-1.81  0.118       
Histology  Adenocarcinoma  1.06  0.74-1.53  0.756        0.80  0.57-1.12  0.200       
PD-L1 expression levelNegative (<1%)  1.50  0.92-2.45  0.102        1.07  0.66-1.74  0.774       
High (≥50%)  0.75  0.49-1.14  0.180        0.77  0.81-1.82  0.201       
Oncogenic driver mutations  yes  1.15  0.70-1.88  0.578        1.22  0.75-1.97  0.425       
Disease stage  IV  1.12  0.28-4.56  0.870        2.16  0.30-15.47  0.128       
Line of ICI  First-line  1.17  0.70-1.95  0.551        0.69  0.42-1.11  0.378       
Time between diagnosis and start of ICI  weeks  1.00  1.00-1.00  0.065        1.00  1.00-1.00  0.847       
Hemoglobin  g/dl  0.86  0.78-0.94  0.001  0.90  0.82-0.99  0.030  0.86  0.79-0.94  <0.001  0.88  0.67-2.12  0.080 
LDH  UI/L  1.00  1.00-1.00  0.057        1.00  1.00-1.00  0.002  1.00  1.00-1.00  0.177 
NLR baseline  ≥5  1.46  1.04-2.07  0.030  1.12  0.76-1.61  0.555  1.78  1.28-2.47  0.001  1.19  0.67-2.12  0.546 
PLR baseline  ≥180  1.08  0.77-1.51  0.668        1.27  0.92-1.75  0.149       
IRAEs  yes  0.39  0.25-0.61  <0.001  0.44  0.28-0.69  <0.001  0.45  0.31-0.66  <0.001  0.41  0.23-0.75  0.003 

Definition of abbreviations: PFS: Progression-free survival; OS: Overall survival; IRAEs: Immune-related adverse events; ECOG-PS: Eastern Cooperative Oncology Group performance status; PD-L1: Programmed death-ligand 1; ICI: Immune checkpoint inhibitors; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; HR: Hazard ratio; CI: confidence interval.

Median OS was also significantly higher in patients who developed IRAEs (89.0 vs 28.0 weeks, p < 0.001) (Fig. 1.B). In the univariate analysis, ECOG-PS score 0–1 (p = 0.004), hemoglobin value (p < 0.001), LDH value (p = 0.002), NLR ≥5 (p = 0.001), and the development of IRAEs (p < 0.001) was significantly associated with overall survival. In the adjusted analysis, only the development of IRAEs was significantly associated with better OS (HR: 0.41; 95% CI: 0.23-0.75, p = 0.003) (Table 4).

Discussion

ICIs are currently one of the mainstays of NSCLC treatment. However, not infrequently, patients treated with ICIs develop IRAEs.6-9

In this study, our purpose was to identify predictors of IRAEs in patients with NSCLC treated with ICIs. Similar to other studies,13,14 in our analysis more than a quarter of patients developed IRAEs (grade≥2 in 73.5% of cases). In most cases with IRAEs, systemic corticosteroid therapy was required and almost half of them had to permanently discontinue ICI. Lower rates of definitive discontinuation of ICIs due to IRAEs have been reported,6,15,16 but the majority of studies assessing IRAEs had a shorter follow-up period than ours.

In addition, we showed that there was a similar rate of IRAEs in patients treated with Nivolumab and Pembrolizumab and a lower rate in patients treated with Atezolizumab. However, a smaller number of patients treated with Atezolizumab were analyzed, so the association between the type of ICI and increased risk of IRAEs cannot be established based on our results.

Taking into account our purpose, we found that dyslipidemia, pre-ICI systemic corticosteroids, disease controlled as response to ICI and a higher baseline hemoglobin value were independent predictors of the development of IRAEs. Regarding these results, we emphasize that all dyslipidemic patients were treated with statins. Statins have been associated with an immunomodulatory effect by preventing protein prenylation, increasing antigen presentation, T cell activation and cytolytic response.17 Some studies have even suggested that statins may act synergistically with ICIs,18 potentiate their effects18 and be associated with better outcomes in lung cancer patients.19,20 So, the increased risk of IRAEs in patients with dyslipidemia that we found may be mainly related to treatment with statins. Despite this, we did not find a statistically significant association between dyslipidemia/statin use and the severity of IRAEs, and dyslipidemia/statin use did not have a relevant impact on PFS or OS. Thus, the association between the use of statins and the risk of IRAEs, as well as the impact on outcomes in patients treated with ICIs, should be further explored.

Concerning corticosteroid therapy, Arbor et al.21 previously reported that the use of ≥10 mg prednisone equivalent at the time of initiation of ICI is associated with a worse outcome in patients with NSCLC. Here, about a third of patients were treated with systemic corticosteroids prior to initiation of ICI, in most cases to control symptoms directly or indirectly related to the disease. Despite this, at the start of ICI, all patients had stopped or received an equivalent dose of prednisone <10 mg daily. So, we assessed the impact of systemic corticosteroid treatment before starting ICI on the development of IRAEs and outcomes, and we found that pre-ICI corticosteroid treatment was associated with a higher risk of IRAEs, but not with worse PFS or OS. In our opinion, this association may identify a subgroup of patients with T-cells that are more responsive in proliferation and differentiation to the modulating effect of ICIs and corticosteroids. Although corticosteroid treatment at the start of ICI may be associated with a negative impact on clinical outcomes, based on our results, there is no association if corticosteroid use is prior to ICI. In fact, studies that showed an association between the use of corticosteroids (at the beginning of ICI) and worse outcomes do not exclude the possibility that these patients had more comorbidities and a poorer baseline prognosis.21,22

Furthermore, in our analysis, a higher baseline hemoglobin value was associated with the development of IRAEs and better outcomes. Other studies had reported the hemoglobin value as a strong predictor of response (independent of performance status) to treatment with ICIs in NSCLC patients.23 Anemia is a common abnormality in patients diagnosed with advanced cancer,24 and there is evidence that anemia and hypoxemia may be implicated in tumor growth, anti-apoptosis and angiogenesis mechanisms, and reduced effects of therapies.25 Thus, it is plausible that higher hemoglobin values may be associated with a greater effect of ICI's, better outcomes, but also with more IRAEs. Therefore, correction of anemia before starting ICIs may have clinical benefits.

In NSCLC, inflammatory biomarkers have shown to correlate with a poor prognosis and a low therapeutic response to conventional treatment.10,12 Some studies have even shown an association between baseline NLR and PLR and the risk of IRAEs and prognostic outcomes.10,12 Using published cutoff points of NLR11 and PLR,12 we found an association between baseline NLR≥5 and worse PFS and OS in univariate analysis, but this association was not confirmed in multivariate analysis. In addition, the association between baseline NLR or PLR and the risk of IRAEs was not found. Nevertheless, these results may be due to the considerable number of patients treated with systemic corticosteroids pre-ICI, by modulating peripheral blood immune cells profile.

Moreover, we showed that patients with IRAEs had a significantly higher disease control rate in response to ICI and significantly longer median PFS and OS compared to patients without IRAEs. The association between the development of IRAEs and better PFS and OS was supported by univariate and multivariate analyses. Several studies have already shown an association between IRAEs and better outcomes in patients with metastatic melanoma26-28 and NSCLC.14,29-35 Regarding this, Biagio Ricciuti et al.35 showed that the development of IRAEs was even a strong predictor of PFS and OS outcomes in NSCLC patients treated with Nivolumab.

As mentioned above, although the mechanisms of IRAEs are not fully understood, this is certainly related to the hyperstimulation of the immune system by ICIs. Therefore, patients with a greater immune system response to ICI will have better antitumor responsiveness and a higher risk of developing adverse effects.

There are some limitations to our study. This is a unicentric retrospective study so information bias cannot be excluded. IRAEs were considered when recorded by patients' physicians and mild IRAEs may not have been valued and not recorded. The small number of cases in our study may have affected some results and the interpretation of predictors and outcomes. The individual impact of each therapeutic strategy on outcomes has not been assessed.

Despite this, we present some relevant results, highlighting the value of hemoglobin and the use of statins rather than inflammatory biomarkers, which have been more traditionally reported as predictors of IRAEs and outcomes in patients treated with ICIs.

Conclusions

In conclusion, the most important results of our study were that patients treated with statins, pre-ICI systemic corticosteroids, higher baseline hemoglobin values, and controlled disease as an initial response to ICI had an increased risk of developing IRAEs. Furthermore, patients who developed IRAEs had better outcomes compared to patients who did not. Further studies with a larger number of patients are needed to validate and complement these results, as well as to better clarify the mechanisms of how ICIs act, in order to achieve the best efficacy and minimal toxicity with this therapy.

CRediT authorship contribution statement

M. Serino: Conceptualization, Methodology, Investigation, Data curation, Resources, Software, Formal analysis, Writing – original draft, Visualization, Writing – review & editing. C. Freitas: Conceptualization, Methodology, Investigation, Data curation, Resources, Formal analysis, Writing – original draft, Visualization, Writing – review & editing. M. Martins: Conceptualization, Investigation, Resources, Visualization. P. Ferreira: Conceptualization, Investigation, Resources, Visualization. C. Cardoso: Conceptualization, Investigation, Resources, Writing – original draft, Visualization. F. Veiga: Investigation, Resources. V. Santos: Conceptualization, Methodology, Writing – original draft, Visualization. D. Araújo: Conceptualization, Methodology, Writing – original draft, Visualization. H. Novais-Bastos: Conceptualization, Methodology, Investigation, Data curation, Resources, Formal analysis, Writing – original draft, Visualization, Writing – review & editing. A. Magalhães: Conceptualization, Methodology, Writing – original draft, Visualization. H. Queiroga: Conceptualization, Methodology, Writing – original draft, Visualization. G. Fernandes: Conceptualization, Methodology, Investigation, Data curation, Resources, Formal analysis, Writing – original draft, Visualization, Writing – review & editing. V. Hespanhol: Conceptualization, Methodology, Investigation, Data curation, Resources, Formal analysis, Writing – original draft, Visualization, Writing – review & editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References
[1]
RL Siegel, KD Miller, HE Fuchs, A Jemal.
Cancer statistics, 2021.
CA Cancer J Clin, 71 (2021), pp. 7-33
[2]
C Zappa, SA Mousa.
Non-small cell lung cancer: current treatment and future advances.
Transl Lung Cancer Res, 5 (2016), pp. 288-300
[3]
MA Bareschino, C Schettino, A Rossi, et al.
Treatment of advanced non-small cell lung cancer.
[4]
ESMO: Metastatic non-small cell lung cancer Clinical practice guidelines for diagnosis, treatment, and follow-up (2018, updated 2020).
[5]
National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology. Non-small cell lung cancer. https://www.nccn.org/professionals/physician_gls/pdf/nscl.pdf (Accessed on September 12, 2021).
[6]
LDD Khoja, T Wei-Wu Chen, LL Siu, AR Hansen.
Tumour- and class-specific patterns of immune-related adverse events of immune checkpoint inhibitors: a systematic review.
Ann Oncol, 28 (2017), pp. 2377-2385
[7]
MA Postow, R Sidlow, MD Hellmann.
Immune-related adverse events associated with immune checkpoint blockade.
NEJM, 378 (2018), pp. 158-168
[8]
J Remon, L Mezquita, J Corral, N Vilariño, N Reguart.
Immune-related adverse events with immune checkpoint inhibitors in thoracic malignancies: focusing on non-small cell lung cancer patients.
J Thorac Dis, 10 (2018), pp. S1516-S1533
[9]
A Cortellini, R Chiari, B Ricciuti, et al.
Correlations Between the Immune-related Adverse Events Spectrum and Efficacy of Anti-PD1 Immunotherapy in NSCLC Patients.
Clin Lung Cancer, 20 (2019), pp. 237-247
[10]
S Diem, S Schmid, M Krapf, et al.
Neutrophil-to-Lymphocyte ratio (NLR) and Platelet-to-Lymphocyte ratio (PLR) as prognostic markers in patients with non-small cell lung cancer (NSCLC) treated with nivolumab.
Lung Cancer, 111 (2017), pp. 176-181
[11]
SJ Bagley, S Kothari, C Aggarwal, et al.
Pretreatment neutrophil-to-lymphocyte ratio as a marker of outcomes in nivolumab-treated patients with advanced non-small-cell lung cancer.
Lung Cancer, 106 (2017), pp. 1-7
[12]
X Gu, S Sun, XS Gao, et al.
Prognostic value of platelet to lymphocyte ratio in non-small cell lung cancer: evidence from 3,430 patients.
Sci Rep, 6 (2016), pp. 23893
[13]
GC Leonardi, JF Gainor, M Altan, et al.
Safety of programmed death-1 pathway inhibitors among patients with non-small-cell lung cancer and preexisting autoimmune disorders.
J Clin Oncol, 1 (2018), pp. 1905-1912
[14]
A Pavan, L Calvetti, A Dal Maso, et al.
Peripheral blood markers identify risk of immune-related toxicity in advanced non-small cell lung cancer treated with immune-checkpoint inhibitors.
Oncologist, 24 (2019), pp. 1128-1136
[15]
A Rittmeyer, F Barlesi, D Waterkamp, et al.
Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
[16]
L Fehrenbacher, J von Pawel, K Park, et al.
Updated efficacy analysis including secondary population results for OAK: a randomized phase III study of atezolizumab versus docetaxel in patients with previously treated advanced non-small cell lung cancer.
J Thorac Oncol, 13 (2018), pp. 1156-1170
[17]
J Greenwood, L Steinman, SS Zamvil.
Statin therapy and autoimmune disease: from protein prenylation to immunomodulation.
Nat Rev Immunol, 6 (2006), pp. 358-370
[18]
A Rossi, M Filetti, B Taurelli Salimbeni, et al.
Statins and immunotherapy: Togetherness makes strength The potential effect of statins on immunotherapy for NSCLC.
Cancer Rep (Hoboken), 31 (2021), pp. e1368
[19]
M Omori, Y Okuma, T Hakozaki, Y Hosomi.
Statins improve survival in patients previously treated with nivolumab for advanced non-small cell lung cancer: an observational study.
Mol Clin Oncol, 10 (2019), pp. 137-143
[20]
DK Xia, ZG Hu, YF Tian, FJ Zeng.
Statin use and prognosis of lung cancer: a systematic review and meta-analysis of observational studies and randomized controlled trials.
Drug Des Devel Ther, 23 (2019), pp. 405-422
[21]
KC Arbour, L Mezquita, N Long, et al.
Impact of baseline steroids on efficacy of programmed cell death-1 and programmed death-ligand 1 blockade in patients with non-small-cell lung cancer.
J Clin Oncol, 1 (2018), pp. 2872-2878
[22]
M Jove, N Vilariño, E Nadal.
Impact of baseline steroids on efficacy of programmed cell death-1 (PD-1) and programmed death-ligand 1 (PD-L1) blockade in patients with advanced non-small cell lung cancer.
Transl Lung Cancer Res, 8 (2019), pp. S364-S368
[23]
S Benzekry, M Grangeon, M Karlsen, M Alexa, I Bicalho-Frazeto, S Chaleat, et al.
Machine learning for prediction of immunotherapy efficacy in non-small cell lung cancer from simple clinical and biological data.
Cancers, 13 (2021), pp. 6210
[24]
UA Ibrahim, AA Yusuf, SG Ahmed.
The pathophysiologic basis of anaemia in patients with malignant diseases.
Gulf J Oncolog, 1 (2016), pp. 80-89
[25]
SJ Van Belle, V Cocquyt.
Impact of haemoglobin levels on the outcome of cancers treated with chemotherapy.
Crit Rev Oncol Hematol, 47 (2003), pp. 1-11
[26]
C Hua, L Boussemart, C Mateus, et al.
Association of vitiligo with tumor response in patients with metastatic melanoma treated with pembrolizumab.
JAMA Dermatol, 152 (2016), pp. 45-51
[27]
M Freeman-Keller, Y Kim, H Cronin, A Richards, G Gibney, JS Weber.
Nivolumab in resected and unresectable metastatic melanoma: characteristics of immune-related adverse events and association with outcomes.
Clin Cancer Res, 15 (2016), pp. 886-894
[28]
M Sanlorenzo, I Vujic, A Daud, et al.
Pembrolizumab cutaneous adverse events and their association with disease progression.
JAMA Dermatol, 151 (2015), pp. 1206-1212
[29]
J Remon, N Reguart, E Auclin, B Besse.
Immune-related adverse events and outcomes in patients with advanced non-small cell lung cancer: a predictive marker of efficacy?.
J Thorac Oncol, 14 (2019), pp. 963-967
[30]
L Peng, Y Wang, F Liu, et al.
Peripheral blood markers predictive of outcome and immune-related adverse events in advanced non-small cell lung cancer treated with PD-1 inhibitors.
Cancer Immunol Immunother, 69 (2020), pp. 1813-1822
[31]
M Grangeon, P Tomasini, S Chaleat, et al.
Association between immune-related adverse events and efficacy of immune checkpoint inhibitors in non-small-cell lung cancer.
Clin Lung Cancer, 20 (2019), pp. 201-207
[32]
BC Ahn, KH Pyo, CF Xin, et al.
Comprehensive analysis of the characteristics and treatment outcomes of patients with non-small cell lung cancer treated with anti-PD-1 therapy in real-world practice.
J Cancer Res Clin Oncol, 145 (2019), pp. 1613-1623
[33]
K Haratani, H Hayashi, Y Chiba, et al.
Association of Immune-Related Adverse Events With Nivolumab Efficacy in Non-Small-Cell Lung Cancer.
JAMA Oncol, 4 (2018), pp. 374-378
[34]
S Teraoka, D Fujimoto, T Morimoto, et al.
Early immune-related adverse events and association with outcome in advanced non-small cell lung cancer patients treated with nivolumab: a prospective cohort study.
J Thorac Oncol, 12 (2017), pp. 1798-1805
[35]
B Ricciuti, C Genova, A De Giglio, et al.
Impact of immune-related adverse events on survival in patients with advanced non-small cell lung cancer treated with nivolumab: long-term outcomes from a multi-institutional analysis.
J Cancer Res Clin Oncol, 145 (2019), pp. 479-485
Copyright © 2022. Sociedade Portuguesa de Pneumologia
Pulmonology

Subscribe to our newsletter

Article options
Tools
Supplemental materials

Are you a health professional able to prescribe or dispense drugs?