Dual-Edged Therapy: Corticosteroids in the Era of Immunotherapy

Abstract

Corticosteroids are widely prescribed in oncology for symptom palliation; however, their use in the era of effective immunotherapy has been a matter of debate for the past decade. Polyakov et al., in Cancer Research Communications, investigated the impact of baseline corticosteroid use on immune checkpoint inhibitor (ICI) efficacy and plasticity of circulating immune cells. This translational study involved 277 patients with stage II-IV non-small cell lung cancer (NSCLC) treated with anti-PD-1 therapy at two academic centers. Corticosteroid use at treatment initiation was associated with significantly reduced progression-free and overall survival and emerged as the most influential prognostic factor on multivariate analysis, surpassing histology, stage, and brain metastases. Higher baseline doses were particularly detrimental to survival, whereas patients who discontinued corticosteroids before ICI initiation tended to have better outcomes. Corticosteroids also abrogated the predictive utility of circulating early on-treatment biomarkers such as the neutrophil-to-lymphocyte ratio and the CX3CR1 score. In mouse models, this study revealed that corticosteroids attenuated the increase of circulating CX3CR1+CD8+ T cells, associated with impaired ICI efficacy. These findings highlight the complex role of corticosteroids in oncology: corticosteroids can interfere with ICI treatment response and impact survival yet can be indispensable for symptom management.

Immune checkpoint inhibitors (ICIs) are a cornerstone in the treatment of non-small cell lung cancer (NSCLC), with their primary mechanism arising from their ability to enhance immune function by relieving the inhibitory effect of tumors on T lymphocytes [1–3]. At the same time, corticosteroids are among the most frequently prescribed drugs in oncology. Polyakov et al. in Cancer Research Communications found that baseline use of corticosteroids, known to cause neutrophilia and lymphopenia, dampen the effects of ICIs [4]. Corticosteroids are used to palliate cancer-related symptoms such as cachexia, fatigue, pain, and cerebral edema from brain metastases [5,6]. They are also a mainstay for managing chemotherapy-induced nausea, cancer-related pain, and appetite loss [7]. This wide use has created a tension between two important tools in cancer care: ICIs, which rely on robust immune activation, and corticosteroids, which are immunosuppressive. Polyakov et al. investigated this clinical dilemma, and since publication, the study has generated discussion and controversy, as oncologists are faced with the competing goals of controlling symptoms and optimizing response to immunotherapy.

This translational study retrospectively evaluated outcomes of 277 patients with stage II-IV NSCLC treated with ICI therapy, specifically anti-PD-1 antibody monotherapy or combination regimens, at two academic institutions: Roswell Park Comprehensive Cancer Center (RPCCC) and the University of Southern California (USC). In this cohort, 8% of the study population were taking corticosteroids at baseline. The results revealed that patients on baseline corticosteroids were found to have significantly reduced progression-free survival (PFS) and overall survival (OS) (for RPCCC: PFS HR = 2.090; 95% CI, 0.96–4.05; P = 0.0423 and OS HR = 2.224; 95% CI, 0.96–4.51; P = 0.0397; for USC: PFS HR = 3.300; 95% CI, 1.61–6.03; P= 0.0003 and OS HR = 2.484; 95% CI, 1.05–4.96; P = 0.020). On multivariate analysis, use of corticosteroids at baseline was the most influential prognostic factor, proving to be stronger than histology, staging, and even the presence of brain metastases. Furthermore, high dosage of corticosteroid administration at initiation of ICI therapy were also factors in ICI efficacy and patient outcomes. Higher doses of baseline corticosteroids, defined as >=30 mg prednisone equivalent, were associated with significantly decreased response rates and worse OS and PFS (for RPCCC: PFS HR = 3.222; 95% CI, 1.11–7.41; P = 0.014 and OS HR = 3.693; 95% CI, 1.26–8.63; P = 0.007; for USC: PFS HR = 5.450; 95% CI, 2.26–11.16; P < 0.0001 and OS HR = 10.48; 95% CI, 3.96–23.11; P < 0.0001). Discontinuing corticosteroids before the induction of ICI therapy tended to have a better treatment response compared to patients who received corticosteroids on the day of ICI therapy initiation (RPCCC: P = 0.0013; USC: P = 0.0329).

Not only do corticosteroids affect survival and treatment response, but they can also attenuate the prognostic performance of biomarkers. Neutrophil-to-lymphocyte ratio (NLR) is a readily accessible biomarker that can capture evidence of systemic inflammation and/or lymphopenia. When elevated, it is associated with worse survival and response to therapy in patients with NSCLC not taking corticosteroids [8,9]. In patients taking corticosteroids, however, Polyakov et al. found that the prognostic performance of NLR was abrogated due to corticosteroids altering peripheral leukocyte profiles.

While other studies have previously associated baseline corticosteroid use with decreased response rates and poor prognosis, Polyakov et al. moved beyond association to explore underlying immunomodulatory mechanisms of corticosteroids in the context of ICIs [10,11]. Having observed lower frequency of circulating terminally differentiated CX3CR1+CD8+ T cells in patients on corticosteroids at baseline, this study evaluated the predictive value of the percent change of the CX3CR1+ subset in CD8+ T cells from baseline (CX3CR1 score), known to correlate with response to ICI therapy [12–15]. Whereas higher CX3CR1 score correlates with response to anti-PD-1 therapy and chemo-immunotherapy (anti–PD-1 therapy: P < 0.0001; chemo-immunotherapy: P < 0.0001), it was not predictive in patients on corticosteroids. Furthermore, employing a bedside-to-bench approach, this study found that corticosteroids attenuated the increase of CX3CR1+CD8+ T cells in mice (P = 0.0245), significantly decreasing the anti-tumor efficacy of ICIs and survival (P < 0.01). These results provide insights into the decreased effector differentiation of CD8+ T cells and utility of predictive blood-based biomarkers by corticosteroids.

Overall, this study indicates that corticosteroids can interfere with ICI treatment response and impact patient survival; however, corticosteroids may be necessary and unavoidable for some patients to manage their cancer-related symptoms, especially patients who encounter symptomatic brain metastasis prior to surgical or radiation therapy. That being said, our study is focused on patients who receive baseline corticosteroids at ICI initiation, not patients treated with corticosteroids as a side effect to ICI therapy. For example, pneumonitis secondary to ICI therapy must be promptly treated with corticosteroids to avoid progression of life-threatening inflammation. Similarly, corticosteroids remain the standard of care for the management of most other immune-related adverse events (irAEs) associated with ICI therapy [16]. Importantly, in this setting, corticosteroids do not appear to compromise outcomes. Indeed, several studies have shown that patients developing irAEs often have favorable responses to immunotherapy, and corticosteroid treatment for irAEs does not abrogate those benefits [17]. This distinction highlights that the timing and indication for corticosteroid use are critical: corticosteroids given at ICI initiation appear detrimental, whereas those used reactively for irAEs are not.

Strengths of this study include its translational approach, which bridges a retrospective clinical analysis with mouse models, allowing the investigators to offer an underlying explanation for the impact of corticosteroids on patient survival. Additionally, clinical results were replicated across separate databases from two large academic cancer centers. The study is limited by its retrospective design and smaller frequency of patients on corticosteroids (8%) compared to previous studies. This limited sample size may reduce statistical power and raise the possibility of unmeasured confounding.

Further research should investigate strategies to reduce or counteract corticosteroid-induced immunosuppression in patients receiving ICIs. Alternative methods for managing cachexia such as GDF-15 inhibitors may be preferred, and the use of opiates or radioligand pharmaceuticals for bone pain may spare patients the need for corticosteroids at diagnosis [18]. For patients who require corticosteroids, further evaluation must be done to determine whether lowering the dose or optimizing tapering schedules can mitigate immunosuppressive effects without compromising symptom control. Beyond NSCLC, the relevance of corticosteroid use extends to other malignancies treated with ICIs, including melanoma, renal cell carcinoma, and head and neck squamous cell carcinoma, in which timing, dose, and indication may similarly influence outcomes. In addition to treatment strategies, further evaluation must be done to evaluate how corticosteroids interact with emerging immune biomarkers. Given that increased CX3CR1 score correlates with efficacy of PD-1/PD-L1 blockade therapy, it may have utility as a biomarker for the development of new targeted agents [13–15]. However, because our study found that corticosteroids diminish the prognostic performance of CX3CR1 and NLR, further studies are needed to examine the effect of corticosteroids on other biomarkers and to identify alternative biomarkers less susceptible to corticosteroid effects.

Overall, baseline corticosteroid use at ICI initiation is associated with impaired biomarker performance and worse survival outcomes. Mouse models further demonstrated that corticosteroids blunt CX3CR1+CD8+ T cell expansion, attenuating the anti-tumor effect of ICIs. Although alternative strategies to mitigate corticosteroid-induced immunosuppression warrant further investigation, the reality remains that many patients require corticosteroids for urgent symptom control. From the patient’s perspective, corticosteroids are often a source of rapid symptom relief and improved quality of life by reducing brain swelling, improving appetite, alleviating fatigue, or treating a possible comorbidity. The tradeoff between immediate symptom control and potential long-term survival is therefore complex. The present study serves to remind clinicians to balance palliation with the promise of durable remission, ensuring that treatment decisions are individualized and aligned with patient goals.