Impact of Glucocorticoids on Immune Checkpoint Inhibitor Efficacy and Circulating Biomarkers in Non–Small Cell Lung Cancer Patients

Lauren Polyakov, Angelina Lim, Alexandra Meyer, Aubree Mades, Joshua Ni, Ryan Cooper, Shirley Ye, Ryutaro Kajihara, Takaaki Oba, Leslie Contreras, Eihab Abdelfatah, Joy Sarkar, Junko Matsuzaki, Ming Li, Rajeev Sharma, Brahm H. Segal, Robert Hsu, Hongbin Chen, Jorge J. Nieva, Fumito ItoCancer 2025 · DOI

TL;DR

This study reveals that the common practice of prescribing corticosteroids to non-small cell lung cancer (NSCLC) patients undergoing immune checkpoint inhibitor (ICI) therapy significantly hinders treatment efficacy and survival. Furthermore, baseline steroid use appears to blunt the predictive power of key blood-based biomarkers, suggesting a need for caution when interpreting these markers in steroid-treated patients.

What Problem Does This Paper Address?

Immune checkpoint inhibitors (ICIs) have revolutionized NSCLC treatment, but not all patients benefit. Identifying factors that predict response and resistance is crucial. Corticosteroids (like prednisone) are frequently given to NSCLC patients for symptom management (e.g., brain metastases, pain, breathing difficulties). While their use to manage ICI-induced side effects (immune-related adverse events) doesn't seem to harm outcomes, their use before or during ICI treatment has been anecdotally linked to worse results. However, the extent of this negative impact and the underlying mechanisms remained unclear. This paper aims to systematically investigate how baseline corticosteroid use affects ICI efficacy and survival in NSCLC patients and to explore its influence on promising blood-based biomarkers.

Methods

The researchers conducted a retrospective analysis of 277 NSCLC patients treated with ICIs at two academic institutions (Roswell Park Comprehensive Cancer Center and University of Southern California). They compared clinical outcomes (overall response rate, progression-free survival, overall survival) between patients who were on baseline corticosteroids and those who were not.

To understand the biological mechanisms, they employed a "bedside-to-bench" approach. This involved:

  • Clinical Data Analysis: Analyzing patient demographics, treatment regimens, and outcomes.
  • Blood Sample Analysis: Prospectively collecting blood samples from patients to assess the frequency of specific immune cells, particularly CD8+ T cells expressing the CX3CR1 receptor, which is associated with T-cell effector function.
  • Preclinical Model: Using a mouse model (MC38 colon adenocarcinoma) to experimentally test the impact of dexamethasone (a corticosteroid) on anti-PD-L1 therapy efficacy and T-cell differentiation.

Datasets

The study utilized clinical data from two independent patient cohorts:

  • Roswell Park Comprehensive Cancer Center (RPCCC): 88 NSCLC patients.
  • University of Southern California (USC): 189 NSCLC patients.

Blood samples were collected from a subset of these patients for biomarker analysis. A preclinical mouse model was also used, but no specific dataset name or size was provided for this aspect.

Key Insights and Ideas

  • Steroids are a significant roadblock: Baseline corticosteroid use is a powerful, independent negative predictor of outcomes in NSCLC patients receiving ICIs, even more so than factors like smoking status or performance status.
  • Dose matters: Higher doses of corticosteroids appear to have a more detrimental effect on ICI efficacy than lower doses.
  • Timing is critical: While continuous steroid use is detrimental, discontinuing steroids before ICI initiation might mitigate some of the negative impact, though this needs further investigation.
  • Biomarker interference: Steroids seem to suppress the frequency of circulating CX3CR1+ CD8+ T cells, a promising biomarker for ICI response. This suppression can diminish the predictive accuracy of these markers, leading to potential misinterpretations.
  • Immune suppression mechanism: The study suggests that steroids directly impair T-cell differentiation and effector function, which is a key mechanism behind their negative impact on ICI therapy.

Key Results

  • Reduced Response Rates: Patients on baseline steroids had significantly lower overall response rates (ORR) compared to those not on steroids (e.g., 38% vs. 45% at RPCCC, 33% vs. 48% at USC).
  • Shorter Survival: Baseline steroid use was associated with markedly shorter progression-free survival (PFS) and overall survival (OS) in both cohorts. For example, median PFS was 3.2 months for steroid users vs. 10.7 months for non-users at RPCCC.
  • Independent Risk Factor: In multivariate analyses, steroid use was the only significant independent risk factor for disease progression and mortality in both cohorts.
  • NLR loses predictive power: While a low neutrophil-to-lymphocyte ratio (NLR < 5) is generally a good prognostic indicator for ICI therapy, this predictive value was absent in patients receiving steroids.
  • CX3CR1+ CD8+ T cells suppressed: Patients on steroids had a substantially lower baseline frequency of circulating CX3CR1+ CD8+ T cells. In the preclinical model, dexamethasone treatment attenuated the increase of these cells during anti-PD-L1 therapy, correlating with reduced tumor control.

Limitations and Future Work

  • Steroid Frequency: The study noted a lower frequency of patients on steroids (8%) compared to some previous reports, which might be due to increased awareness and efforts to taper steroids before ICI initiation.
  • Small Subgroups: The number of patients on medium-dose steroids or those who discontinued steroids before ICI was small, requiring cautious interpretation of these specific findings.
  • Preclinical Model Specificity: The preclinical study used only one mouse model, and the duration of steroid treatment might differ between mice and humans.
  • Biomarker Data Availability: Not all patients had complete data for all circulating biomarker analyses.
  • Diverse Patient Population: The USC cohort was more diverse, with a higher frequency of Hispanic and Asian patients, and further research is needed to explore potential ethnic differences in ICI response.

The authors suggest that future research should focus on optimizing steroid tapering strategies before ICI initiation and further investigate the precise mechanisms by which steroids interfere with T-cell function and biomarker utility. They emphasize the need for caution when interpreting blood-based biomarkers in patients on corticosteroids.

Summary updated

Abstract

The impact of corticosteroids, widely prescribed for palliation of cancer-related symptoms, on ICI therapy remains unclear. This study shows that baseline steroid use is a negative independent prognostic factor in patients with NSCLC undergoing ICI therapy and provides insights into the decreased T-cell effector differentiation and utility of predictive blood-based markers by steroids.
BibTeX
@inproceedings{polyakov2025impact,
  author = {Lauren Polyakov and Angelina Lim and Alexandra Meyer and Aubree Mades and Joshua Ni and Ryan Cooper and Shirley Ye and Ryutaro Kajihara and Takaaki Oba and Leslie Contreras and Eihab Abdelfatah and Joy Sarkar and Junko Matsuzaki and Ming Li and Rajeev Sharma and Brahm H. Segal and Robert Hsu and Hongbin Chen and Jorge J. Nieva and Fumito Ito},
  title = {{Impact of Glucocorticoids on Immune Checkpoint Inhibitor Efficacy and Circulating Biomarkers in Non–Small Cell Lung Cancer Patients}},
  booktitle = {Cancer},
  year = {2025},
  doi = {10.1158/2767-9764.crc-25-0051},
}
Figure