Effects of Early Short-Course Corticosteroids on Immune-Related Adverse Events in Non-Small Cell Lung Cancer Patients Receiving Immune Checkpoint Inhibitors Open Access

Immune checkpoint inhibitors (ICIs) alone or in combination with chemotherapy or targeted therapy bring revolutionary improvement to survival and life quality across various types of malignancies at different disease settings [1]. However, ICIs are associated with unique toxicities due to unintended activation of the immune system, namely immune-related adverse events (irAEs) [2‒4]. The occurrence of irAEs is unpredictable, and severe irAEs could lead to treatment discontinuation, prolonged exposure to corticosteroid, increased risks for opportunistic infections and even death [4]. There are no preventive measures for irAEs to date.

Corticosteroids are the recommended initial treatment for moderate to severe irAEs [5‒8]. Some clinical evidence showed that corticosteroids may suppress irAEs without compromising effectiveness of ICIs [9, 10], indicating that autoimmunity and antitumor immunity may respond to corticosteroids differently [11]. This difference creates an opportunity to examine whether early use of corticosteroids, which are administered before or within a few days after each ICI injection, could prevent irAEs.

Short-course corticosteroids are a common prescription in oncology to prevent or alleviate the side effects of chemotherapy agents, especially platinum, pemetrexed, and taxanes [12‒14]. Recently, ICIs in combination with platinum-based chemotherapy have been approved globally as the first-line treatment for patients with advanced non-small cell lung cancer (NSCLC) without actionable driver mutations [15‒17]. In both pivotal phase 3 trials of immunochemotherapy for advanced NSCLC, KEYNOTE-189, and IMpower150, short-course corticosteroid administered soon after ICI injection and before chemotherapy was recommended to all patients [18, 19]. Therefore, patients receiving immunochemotherapy are often exposed to short-course corticosteroids as premedication for chemotherapy. In sharp contrast, patients receiving ICI monotherapy are seldom exposed to corticosteroids.

The differential exposure to corticosteroids in patients with advanced NSCLC receiving ICIs-based treatment provides an ideal model to evaluate whether early short-course corticosteroids following each ICI injection would reduce the occurrence of clinically significant irAEs, especially those that require systemic corticosteroid use or lead to treatment discontinuation. In this retrospective study, we aim to compare frequencies of clinically significant irAEs between NSCLC patients receiving ICIs with and without exposure to early short-course corticosteroids.

This is a retrospective cohort study. Patients with advanced NSCLC who received at least one cycle of anti-programmed cell death-1/programmed cell death-ligand 1 (PD-1/PD-L1) ICI with or without chemotherapy between January 1st, 2015, and December 31st, 2020 at National Taiwan University Hospital (NTUH) and National Taiwan University Cancer Center, Taiwan, were identified through Integrated Medical Database and Cancer Registry of NTUH [20]. Patients were excluded if there was any presence of corticosteroid use within 4 weeks before ICI initiation, concurrent use of other ICIs, double primary cancer, uncertain histology, prior organ transplantation, or known underlying autoimmune diseases. Patients with tumors harboring epidermal growth factor receptor (EGFR) mutations or anaplastic lymphoma kinase (ALK) fusions must have failed to respond to available corresponding targeted agents. Early short-course corticosteroids were defined as corticosteroids administered following ICI injection and before chemotherapy on the same day and no longer than 3 days afterward, with a maximal daily dose of less than 12 mg dexamethasone, aiming to reduce adverse effects of chemotherapy. Eligible patients who received immunochemotherapy with early short-course corticosteroids were categorized as “corticosteroid group,” whereas patients who received immunochemotherapy or ICI monotherapy without corticosteroids as “non-corticosteroid group.” The demographics, clinicopathological characteristics, adverse events, medication, and overall survival (OS) were gathered through electronic medical records. PD-L1 expression was assessed by PD-L1 immunohistochemistry assays (Dako 22C3, Dako 28-8, Ventana SP263, or Ventana SP142). PD-L1-high was defined as PD-L1 expression in equal or more than 50% of tumor cells, while PD-L1-low was defined as PD-L1 expression in less than 50% of tumor cells. The study protocol has been approved by the Research Ethics Committee of NTUH (202110031 RINB). The study was conducted in compliance with the Declaration of Helsinki and other applicable ethical guidelines.

The major outcomes of interest were irAEs, which were determined and graded per the Common Terminology Criteria for Adverse Events (CTCAEs) version 5.0 by individual treating physicians after excluding other possible etiologies. Frequency of irAEs requiring systemic (oral or intravenous) corticosteroid use regardless of grade and irAEs leading to permanent ICI discontinuation were compared between the two groups. Time to systemic corticosteroid use (defined as the interval between initiation of ICI and initiation of systemic corticosteroid for irAEs), duration of systemic corticosteroid treatment (defined as the interval between initiation and the last dose of systemic corticosteroid), and OS (defined as the interval between initiation of ICI and death or the last follow-up) were also analyzed.

Categorical variables between exposure groups were compared using the χ², or Fisher’s exact test, and the results were reported as frequencies. Continuous variables were compared using Mann-Whitney U test. Logistic regression analysis was performed to evaluate the effects of other potential factors on the development of irAEs of interest. OS was calculated using Kaplan-Meier analysis and compared between subgroups using the log-rank test. Cox regression models stratified to the first-line treatment and the use of concurrent chemotherapy were performed on selected factors (exposure to corticosteroid, smoking, gender, age, first-line treatment, concurrent chemotherapy, and squamous cell histology) to assess their effects on survival. All factors were regarded as binary variables. R (version 3.3.3 for Mac OS; R foundation, Vienna, Austria) was used to conduct all statistical analyses. Tests were performed at a significance level of α = 0.05, and values were considered statistically significant if p value <0.05. The data cutoff for this analysis was March 15, 2022.

A total of 252 eligible patients were identified, 137 in the corticosteroid group and 115 in the non-corticosteroid group. The patient characteristics are summarized in Table 1, with a median age of 61.3 (range 31.0–86.5) years and male predominance (n = 130, 55.2%). The majority of patients had adenocarcinoma (n = 184, 73%), followed by squamous cell carcinoma (n = 33, 13.1%), and other histological types including pleomorphic and sarcomatoid carcinoma (n = 33, 13.1%). Among 228 patients who had tumor samples tested for genetic alterations, 82 patients (36%) had druggable genetic alterations. Of 197 patients who had tumors assessed for PD-L1 expression, 81 patients (41.1%) had PD-L1-high tumors. All variables were balanced between the two groups except for treatment lines and the use of concurrent chemotherapy. More patients received ICIs as the first-line treatment ICIs in the corticosteroid group (n = 75; 54.7%), as compared to the non-corticosteroid group (n = 28; 24.3%) (p < 0.001). Only 19 patients (16.5%) in the non-corticosteroid group received concurrent chemotherapy. The details of immunotherapy and chemotherapy partners are summarized in Supplementary Tables 1 and 2 (for all online suppl. material, see https://doi.org/10.1159/000534350). The most commonly prescribed regimen of early short-course corticosteroids was intravenous betamethasone 4 mg before chemotherapy followed by oral dexamethasone 8 mg daily for 3 days (n = 98). The other regimens included intravenous betamethasone 4 mg before chemotherapy (n = 19) and intravenous betamethasone 4 mg before chemotherapy followed by oral dexamethasone 8 mg daily for less than 3 days (n = 21).

Thirty patients (21.9%) in the corticosteroid group and 35 patients (30.4%) in the non-corticosteroid group developed irAEs requiring systemic corticosteroid use (odds ratio [OR], 0.64; 95% confidence interval [CI], 0.35–1.18; p = 0.15) (Table 2). The most common irAE requiring systemic corticosteroid use was pneumonitis (n = 37), followed by skin rash (n = 13), adrenal insufficiency (n = 8), and hepatitis (n = 7) in all patients. The frequency of immune-related pneumonitis requiring systemic corticosteroid use was numerically lower in the corticosteroid group although statistical significance was not met (corticosteroid group vs. non-corticosteroid group, 10.9 vs. 19.1%; OR, 0.52; 95% CI, 0.24–1.12; p = 0.076). No other clinical factors were significantly associated with the development of irAEs requiring systemic corticosteroid in logistic regression analysis (online suppl. Table 3).

The median time to irAEs requiring systemic corticosteroids was 18.9 weeks (interquartile range [IQR]: 7.6–31.9) in the corticosteroid group versus 12 weeks (IQR: 2.2–34.3) in the non-corticosteroid group (p = 0.13). The median duration of steroid treatment was 6.8 weeks (IQR: 3.4–13.8) in the corticosteroid group versus 5.1 weeks (IQR: 2–11.9) in the non-corticosteroid group (p = 0.25). Among 197 patients with available PD-L1 expression data, 20 patients (24.4%) with PD-L1-high tumors and 30 patients (26%) with PD-L1-low tumors developed irAEs requiring systemic corticosteroids (p = 0.87).

Eight patients (5.8%) in the corticosteroid group, as compared to 18 patients (15.7%) in the non-corticosteroid group, permanently discontinued ICIs due to irAEs (OR, 0.34; 95% CI, 0.12–0.85; p = 0.013) (Table 2). Pneumonitis was the most common reason for ICI discontinuation in each group (5.8% in the corticosteroid group; 10.4% in the non-corticosteroid group). Among patients with available PD-L1 expression data, 11 patients (13.4%) with PD-L1-high tumors and 10 patients (8.7%) with PD-L1-low tumors permanently discontinued ICIs due to irAEs (p = 0.35). Logistic regression analysis was performed to further evaluate the influence of various clinical factors on irAEs leading to treatment discontinuation. In both univariate and exploratory multivariate analyses (including factors significant at p < 0.15 in univariate analysis), corticosteroid exposure emerged as the sole significant factor associated with the development of irAEs leading to treatment discontinuation (OR, 0.34; 95% CI, 0.14–0.82; p = 0.017) (Table 3).

The median OS was 28 months and 18.8 months in the corticosteroid group and non-corticosteroid group respectively (online suppl. Fig. 1. A). However, a direct comparison of survival between the two groups was not performed due to inherent differences in patient characteristics. Univariate Cox regression analysis revealed that treatment line, exposure to early short-course corticosteroids, and concurrent chemotherapy were potential confounding factors affecting survival. Multivariate analysis demonstrated that only treatment line was independently correlated with OS (Table 4).

In the subgroup analysis of 103 patients who received ICIs in the first-line setting (online suppl. Table 4), the median OS was 28.7 months and 32 months in the corticosteroid group and non-corticosteroid group, respectively. Exposure to early short-course corticosteroids was not associated with poorer survival (hazard ratio [HR], 1.06; 95% CI, 0.56–2; p = 0.9) (online suppl. Fig. 1. B). We also conducted an exploratory subgroup analysis in patients who received concurrent immunochemotherapy. Among the 156 patients who received immunochemotherapy regardless of treatment lines (online suppl. Table 5), the median OS was 25.9 months in patients who received early short-course corticosteroids (n = 137) and 12.5 months in patients who did not (n = 19). The exposure to corticosteroids was not associated with an increased risk of death (HR, 0.72; 95% CI, 0.37–1.45; p = 0.4) (online suppl. Fig. 1. C); however, the result should be interpreted with caution due to disproportionate sample size.

Despite the growing use of ICIs in modern oncology practice, effective preventive strategies for irAEs remain limited. The current study showed that in NSCLC patients receiving ICIs with or without chemotherapy, early use of short-course corticosteroid as premedication for chemotherapy is significantly associated with reduction in irAEs leading to ICI discontinuation and marginally associated with reduction in irAEs requiring systemic corticosteroid use, particularly in cases of pneumonitis. Besides, there was no obvious signal to indicate that the early short-course corticosteroid compromises treatment efficacy. Our results suggest that the preventive effects of prophylactic corticosteroids on irAEs in cancer patients receiving ICIs are worthy of further prospective investigation.

To the best of our knowledge, the current study is the first one to examine the protective effects of early short-course corticosteroids following ICI injection on irAEs using real-world data of NSCLC patients receiving ICIs with or without chemotherapy. Our study provides clinical evidence that early short-course corticosteroids administered within 3 days following ICI injection have the potential to reduce clinically significant irAEs and prevent permanent discontinuation of effective treatment. In line with our findings, a retrospective study showed that baseline corticosteroid use before initiation of immunotherapy was significantly associated with reduction in ICI discontinuation due to irAEs in patients with various types of cancers receiving ICIs [21]. Taken together, prophylactic corticosteroids administered before and within a few days after each ICI injection have great potential to prevent or limit devastating effects of irAEs.

Despite being a reasonable candidate for irAEs prevention, corticosteroids prophylaxis has never been prospectively investigated due to a long-standing concern that corticosteroids could attenuate antitumor immunity. To date, several retrospective analyses have been conducted to resolve this concern [21‒28], but all of them evaluated the effects of baseline or concurrent corticosteroids, rather than early short-course corticosteroids, on treatment efficacy. It has been widely accepted that baseline corticosteroids for cancer-related indications, such as brain metastasis or spinal cord compression, are associated with worse survival outcomes, whereas baseline corticosteroids for cancer-unrelated indications are not [25‒27]. On the other hand, the KEYNOTE-407 study, which examined the efficacy of carboplatin plus paclitaxel or nab-paclitaxel with or without pembrolizumab in patients with advanced squamous NSCLC, provides indirect evidence that early short-course corticosteroids as premedication for chemotherapy do not compromise the efficacy of immunochemotherapy [29]. In KEYNOTE-407, early short-course corticosteroids per local guidelines were required for patients who received paclitaxel but not for those who received nab-paclitaxel. Subgroup analyses of KEYNOTE-407 revealed that different taxane partners did not affect the risks of death or progression, although the exact details on corticosteroids administration were not provided. Furthermore, Shen et al. [30] showed that neither low-dose nor high-dose corticosteroid premedication before anti-PD-1 plus anti-cytotoxic T-lymphocyte activation-4 combination attenuated antitumor activities in murine hepatocellular carcinoma models. Therefore, early use of short-course corticosteroids, which has already been a common clinical practice in patients receiving immunochemotherapy, for irAE prophylaxis appears to be a feasible and safe approach that merits further evaluation.

The current study is primarily limited by the retrospective nature of data collection and analyses. We were not able to capture irAEs of any grade; since most of grade 1 or 2 irAEs were poorly defined and sometimes not documented in the medical records. Nevertheless, clinically significant irAEs that required systemic corticosteroid use or led to ICI discontinuation were adequately recorded. Furthermore, our study included a heterogeneous group of patients who received ICIs in different lines or with different chemotherapy partners, making direct survival comparison not feasible. In addition, since most patients in the non-corticosteroid group did not receive chemotherapy, we cannot totally exclude the possibility that the protective effect from corticosteroids observed in our cohort was in fact due to chemotherapy. However, the chemotherapy agents frequently used in our NSCLC cohort, such as platinum, pemetrexed, taxanes, vinorelbine, and gemcitabine, were not considered effective in treating irAEs or autoimmune diseases from previous literature or treatment guidelines. Thus, chemotherapy is unlikely to solely account for the reduction of clinically significant irAEs observed in this study. Finally, the study was conducted in two institutions within a single country and may not be generalizable to other populations or settings.

As immunotherapy combination regimens with higher risks of severe irAEs, such as anti-PD-1/PD-L1 plus anti-cytotoxic T-lymphocyte-associated protein-4 [31‒33] or anti-lymphocyte activation gene-3 (LAG-3) antibodies [34] become standard of care in oncology, the need for preventive strategies for irAEs is increasingly urgent. Our study indicates that early short-course corticosteroids following each ICI injection may reduce the occurrence of clinically significant serious irAEs that lead to ICI discontinuation or require systemic corticosteroid. Further prospective studies are mandated to evaluate the preventive effects of prophylactic short-course corticosteroids on irAEs.

We would like to thank the staff of Department of Medical Research for providing clinical data from National Taiwan University Hospital-integrative Medical Database (NTUH-iMD). We also thank Ms. Hui-Min Lin for her statistical input.

The study protocol has been approved by the Research Ethics Committee of NTUH (202110031 RINB). The study was conducted in compliance with the Declaration of Helsinki and other applicable ethical guidelines. Our study was a retrospective cohort analysis using electronic medical records and did not involve tissue analysis or intervention. The Research Ethics Committee of National Taiwan University Hospital waived the need to obtain informed consent from the study subjects because the study poses no more than minimal risk to subjects and the waiver will not adversely affect the rights and welfare of the subjects.

The authors have no relevant financial or non-financial interests to disclose.

The authors have not received any funding or grants related to this work.

Derek De-Rui Huang and Ying-Chun Shen were responsible for conceptualization and design of the study and were involved in data curation. Derek De-Rui Huang, Bin-Chi Liao, Wei-Hsun Hsu, Ching-Yao Yang, Yen-Ting Lin, Shang-Gin Wu, Tzu-Hsiu Tsai, Kuan-Yu Chen, Chao-Chi Ho, Wei-Yu Liao, Jin-Yuan Shih, Chon-Jen Yu, James Chih-Hsin Yang, Ann-Lii Cheng, and Ying-Chun Shen contributed to patient treatment and clinical data collection. Derek De-Rui Huang performed the formal analysis, visualization, and wrote the original draft. Bin-Chi Liao, Ann-Lii Cheng, and Ying-Chun Shen were involved in review and editing of the manuscript. The work reported in the paper has been performed by the authors, unless clearly specified in the text.

The data that support the findings of this study are available from the corresponding author upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information. Data are not publicly available due to ethical reasons. Further inquiries can be directed to the corresponding author.