Article
Predictors and Outcomes of Non-Small Cell Lung Carcinoma
Patients Following Severe Immune Checkpoint Inhibitor
Toxicity: A Real-World UK Multi-Centre Study
Umair Mahmood 1,2,3, *, Eleni Josephides 2,4 , Meenali Chitnis 1 , Michael Skwarski 2 , Spyridon Gennatas 2 ,
Sharmistha Ghosh 2 , James Spicer 2,4 , Eleni Karapanagiotou 2,4 , Tanya Ahmad 3 , Martin Forster 3 ,
Mariam Jamal-Hanjani 3,5,6 , Sarah Benafif 3 , Charles Swanton 3,6,7 , Siow-Ming Lee 3 , Dionysis Papadatos-Pastos 3 ,
Alexandros Georgiou 2,4 and Nicholas Coupe 1, *
1 Oxford University Hospitals NHS Foundation Trust, Oxford OX3 7LJ, UK
2 Guy’s and St. Thomas’ NHS Foundation Trust, London SE1 3SS, UK
3 Department of Medical Oncology, University College London Hospitals NHS Foundation Trust,
London NW1 2PG, UK
4 Guy’s Hospital, King’s College London, London SE1 1UL, UK
5 Cancer Metastasis Laboratory, University College London Cancer Institute, London WC1E 6DD, UK
6 Cancer Research UK Lung Cancer Centre of Excellence, University College London Cancer Institute,
London WC1E 6DD, UK
7 Cancer Evolution and Genome Instability Laboratory, The Francis Crick Institute, London NW1 1AT, UK
* Correspondence: (U.M.); (N.C.)
Simple Summary
Treatment of non-small cell lung carcinoma (NSCLC) with immune checkpoint inhibitors
(ICIs) has improved survival rates. However, ICIs can cause severe side effects leading to
hospitalization and are typically managed with steroids. We reviewed how NSCLC patients
with severe ICI toxicity are managed in real-world settings across UK hospitals and assessed
Academic Editor: Constantin the benefits they received from ICIs. Current treatment for ICI toxicity involves high-dose
N. Baxevanis
steroids, which can cause health complications. We aimed to refine this management
Received: 16 July 2025 approach by identifying specific side effects or patient groups that might benefit from lower
Revised: 26 August 2025
steroid doses or early addition of other steroid-sparing agents. We also evaluated how
Accepted: 26 August 2025
stopping or continuing immunotherapy after recovering from side effects impacts lung
Published: 28 August 2025
cancer prognosis and the risk of future ICI toxicity. This is informative for clinicians as the
Citation: Mahmood, U.; Josephides,
decision to continue or stop ICI remains a very difficult clinical decision and an area that
E.; Chitnis, M.; Skwarski, M.;
Gennatas, S.; Ghosh, S.; Spicer, J.;
lacks clinical trial evidence.
Karapanagiotou, E.; Ahmad, T.;
Forster, M.; et al. Predictors and Abstract
Outcomes of Non-Small Cell Lung
Purpose: Evaluation of predictors and outcomes in NSCLC patients treated with an im-
Carcinoma Patients Following Severe
Immune Checkpoint Inhibitor Toxicity:
mune checkpoint inhibitor (ICI) following a severe immune-related adverse event (irAE).
A Real-World UK Multi-Centre Study. Methods: We included all NSCLC patients receiving ≥1 ICI cycle and corticosteroids for
Cancers 2025, 17, 2819. CTCAE Grade ≥3 irAEs between 2017 and 2023 from three UK NHS teaching hospitals.
https://doi.org/10.3390/ Progression-free survival (PFS) and overall survival (OS) after the 1st irAE, best overall
cancers17172819
response (BOR) to ICI, and predictors of clinical benefit were evaluated. Kaplan–Meier,
Copyright: © 2025 by the authors. Cox and logistic regression models, and Wilcoxon tests were used. Results: We screened
Licensee MDPI, Basel, Switzerland. 1658 NSCLC patients and identified 80 eligible subjects. The majority of patients had
This article is an open access article
metastatic (n = 50, 63%) vs. localized (n = 30, 37%) NSCLC. Most patients developed a
distributed under the terms and
single ≥Grade 3 irAE on 1st line ICI (n = 71, 89%). Overall, 14 (18%) patients developed
conditions of the Creative Commons
Attribution (CC BY) license
2nd irAEs, 7 after rechallenge with ICIs. In the complete cohort, median OS after 1st irAE
(https://creativecommons.org/ was 15.84 months (95% CI, 12.45–26.91). Lower neutrophil-to-lymphocyte ratio (NLR),
licenses/by/4.0/). patients receiving >4 cycles of ICI or median duration of ICI of >2.76 months before 1st
Cancers 2025, 17, 2819 https://doi.org/10.3390/cancers17172819
,Cancers 2025, 17, 2819 2 of 16
irAE were associated with improved OS (p < 0.05), the latter two with PFS (p < 0.05).
Age, gender, stage, KRAS mutation, PD-L1 and ICI type were not associated with PFS
or OS. Pneumonitis as 1st irAE had the worst PFS and OS (p < 0.05). Median starting
corticosteroid dose of ≤60 mg for 1st irAE had an improved OS (p = 0.04). Post 1st irAE
response associated with better PFS and OS (p < 0.05). Number and duration of irAEs and
additional immunosuppressive agents (14% of patients) were not associated with PFS or
OS. Conclusions: In ≥Grade 3 irAEs patients managed with corticosteroids, lower baseline
NLR, longer ICI use, response to ICI after 1st irAE, and a ≤60 mg corticosteroid dose had
promising outcomes.
Keywords: non-small cell lung carcinoma; immunotherapy; toxicity; progression-
free survival; overall survival
1. Introduction
Immune checkpoint inhibitors (ICIs) have served as standard of care treatment for
patients with recurrent or metastatic disease across multiple solid and hematologic ma-
lignancies. They involve interaction with immune checkpoint pathways such as the
PD-1/PD-L1 axis, which plays a key role in tumor immune evasion. T cell activation
is initiated when the T cell receptor (TCR) recognizes tumor-derived peptides presented by
major histocompatibility complex (pMHC) molecules on the tumor surface [1]. However,
PD-L1, expressed on tumor cells, binds to the PD-1 receptor on activated T cells, delivering
an inhibitory signal that suppresses T cell activity and limits anti-tumor immunity [1,2].
Therapeutic blockade of this pathway using anti-PD-1 or anti-PD-L1 antibodies can con-
tribute to restoration of T cell function and enhance immune-mediated tumor clearance [1].
In a distinct immunoregulatory mechanism, the CTLA-4 pathway regulates early T cell
activation in lymphoid tissues. Effective T cell priming requires not only TCR engagement
with pMHC, but also co-stimulatory signaling through CD28 binding to B7-1/2 ligands
on antigen-presenting cells (APCs) [1]. CTLA-4, which is upregulated on activated T cells,
competes with CD28 for B7-1/2 binding, thereby inhibiting T cell activation [1,3]. The use
of anti-CTLA-4 antibodies blocks this inhibitory interaction, preventing CTLA-4 from out-
competing CD28, and thus facilitates T cell activation through positive co-stimulation [1].
ICIs such as anti-PD-L1 and anti-CTLA4 antibodies currently represent a widely
adopted immunotherapy approach for patients with lung cancer [4]. The incorporation
of these agents—either as monotherapy or in combination with chemotherapy—has led
to improvements in most clinical efficacy endpoints compared to chemotherapy alone
in patients with locally advanced or metastatic NSCLC lacking oncogenic driver alter-
ations [5]. However, the increasing number of patients with non-small cell lung carcinoma
(NSCLC) receiving ICIs has led to more patients developing immune-related adverse
events (irAEs) in recent years. A meta-analysis of data from 38 trials of advanced-stage
lung cancer patients reported incidences of Grade 3–5 irAEs for ICI monotherapy (6.6%),
an ICI plus chemotherapy (11.4%), dual ICIs (13.8%), and dual ICIs plus chemotherapy
(13.5%) [6]. Current ESMO Clinical Practice Guidelines suggest hospitalization for most
patients experiencing Grade 3 or higher irAEs with corticosteroid treatment delivered as
1–2 mg/kg/day [7]. However, the impact of irAEs among NSCLC patients on treatment
response to ICI therapy is currently limited. Additionally, there is a paucity of data regard-
ing survival outcomes in the setting of immunotherapy following irAEs. These patients
represent an important clinical population where real-world evidence regarding outcomes
following the development of irAEs can help guide clinicians as to the optimal patient
management strategy.
, Cancers 2025, 17, 2819 3 of 16
We conducted a multi-institutional review of NSCLC patients developing ICI-related
toxicity and managed with corticosteroid agents. Our team aimed to evaluate treatment
response, progression-free survival, and overall survival after onset of irAEs. We also
assessed real-world use of corticosteroids and other novel immunosuppressive agents for
the treatment of Grade 3 or higher ICI toxicities, where published data is currently lacking.
The current management approach involves most patients being subjected to high-dose
corticosteroids to treat irAEs, which itself is associated with significant side effects. We
aimed to find groups of patients experiencing improved survival outcomes with specific
doses of corticosteroid therapy to optimize therapeutic benefit. irAEs frequently necessitate
cessation of ICIs, thereby restricting subsequent treatment options for NSCLC patients [8].
Therefore, we also hoped to clarify the clinical benefit of terminating vs. continuing ICI
therapy following irAEs. Identification of such patients is important given the potential for
long-term benefit and reduced side effects of ICIs compared to other systemic therapies.
This would allow patients to continue ICIs to ensure greater duration of treatment benefit.
We also aimed to identify patients who will not benefit from ICI continuation following
irAEs, thereby aiding physicians to recommend other treatment options or terminating ICI
therapy while transitioning to surveillance in cases of adequate disease control to avoid
subsequent ICI-related toxicities.
2. Materials and Methods
2.1. Patient Selection
We screened adult NSCLC patients to identify eligible patients treated with high-dose
corticosteroids for immunotherapy toxicity as described in Figure 1. We abstracted data
from patient medical records, including clinicians’ notes, radiology, radiation oncology,
operative, and pathology reports.
Figure 1. Flow diagram depicting the selection process of eligible patients for subsequent data
analysis. Abbreviations: CTCAE, Common Terminology Criteria for Adverse Events; ICI, Immune
Checkpoint Inhibitor; irAE, immune-related Adverse Event; NSCLC, Non-Small Cell Lung Carci-
noma; UK, United Kingdom.
Predictors and Outcomes of Non-Small Cell Lung Carcinoma
Patients Following Severe Immune Checkpoint Inhibitor
Toxicity: A Real-World UK Multi-Centre Study
Umair Mahmood 1,2,3, *, Eleni Josephides 2,4 , Meenali Chitnis 1 , Michael Skwarski 2 , Spyridon Gennatas 2 ,
Sharmistha Ghosh 2 , James Spicer 2,4 , Eleni Karapanagiotou 2,4 , Tanya Ahmad 3 , Martin Forster 3 ,
Mariam Jamal-Hanjani 3,5,6 , Sarah Benafif 3 , Charles Swanton 3,6,7 , Siow-Ming Lee 3 , Dionysis Papadatos-Pastos 3 ,
Alexandros Georgiou 2,4 and Nicholas Coupe 1, *
1 Oxford University Hospitals NHS Foundation Trust, Oxford OX3 7LJ, UK
2 Guy’s and St. Thomas’ NHS Foundation Trust, London SE1 3SS, UK
3 Department of Medical Oncology, University College London Hospitals NHS Foundation Trust,
London NW1 2PG, UK
4 Guy’s Hospital, King’s College London, London SE1 1UL, UK
5 Cancer Metastasis Laboratory, University College London Cancer Institute, London WC1E 6DD, UK
6 Cancer Research UK Lung Cancer Centre of Excellence, University College London Cancer Institute,
London WC1E 6DD, UK
7 Cancer Evolution and Genome Instability Laboratory, The Francis Crick Institute, London NW1 1AT, UK
* Correspondence: (U.M.); (N.C.)
Simple Summary
Treatment of non-small cell lung carcinoma (NSCLC) with immune checkpoint inhibitors
(ICIs) has improved survival rates. However, ICIs can cause severe side effects leading to
hospitalization and are typically managed with steroids. We reviewed how NSCLC patients
with severe ICI toxicity are managed in real-world settings across UK hospitals and assessed
Academic Editor: Constantin the benefits they received from ICIs. Current treatment for ICI toxicity involves high-dose
N. Baxevanis
steroids, which can cause health complications. We aimed to refine this management
Received: 16 July 2025 approach by identifying specific side effects or patient groups that might benefit from lower
Revised: 26 August 2025
steroid doses or early addition of other steroid-sparing agents. We also evaluated how
Accepted: 26 August 2025
stopping or continuing immunotherapy after recovering from side effects impacts lung
Published: 28 August 2025
cancer prognosis and the risk of future ICI toxicity. This is informative for clinicians as the
Citation: Mahmood, U.; Josephides,
decision to continue or stop ICI remains a very difficult clinical decision and an area that
E.; Chitnis, M.; Skwarski, M.;
Gennatas, S.; Ghosh, S.; Spicer, J.;
lacks clinical trial evidence.
Karapanagiotou, E.; Ahmad, T.;
Forster, M.; et al. Predictors and Abstract
Outcomes of Non-Small Cell Lung
Purpose: Evaluation of predictors and outcomes in NSCLC patients treated with an im-
Carcinoma Patients Following Severe
Immune Checkpoint Inhibitor Toxicity:
mune checkpoint inhibitor (ICI) following a severe immune-related adverse event (irAE).
A Real-World UK Multi-Centre Study. Methods: We included all NSCLC patients receiving ≥1 ICI cycle and corticosteroids for
Cancers 2025, 17, 2819. CTCAE Grade ≥3 irAEs between 2017 and 2023 from three UK NHS teaching hospitals.
https://doi.org/10.3390/ Progression-free survival (PFS) and overall survival (OS) after the 1st irAE, best overall
cancers17172819
response (BOR) to ICI, and predictors of clinical benefit were evaluated. Kaplan–Meier,
Copyright: © 2025 by the authors. Cox and logistic regression models, and Wilcoxon tests were used. Results: We screened
Licensee MDPI, Basel, Switzerland. 1658 NSCLC patients and identified 80 eligible subjects. The majority of patients had
This article is an open access article
metastatic (n = 50, 63%) vs. localized (n = 30, 37%) NSCLC. Most patients developed a
distributed under the terms and
single ≥Grade 3 irAE on 1st line ICI (n = 71, 89%). Overall, 14 (18%) patients developed
conditions of the Creative Commons
Attribution (CC BY) license
2nd irAEs, 7 after rechallenge with ICIs. In the complete cohort, median OS after 1st irAE
(https://creativecommons.org/ was 15.84 months (95% CI, 12.45–26.91). Lower neutrophil-to-lymphocyte ratio (NLR),
licenses/by/4.0/). patients receiving >4 cycles of ICI or median duration of ICI of >2.76 months before 1st
Cancers 2025, 17, 2819 https://doi.org/10.3390/cancers17172819
,Cancers 2025, 17, 2819 2 of 16
irAE were associated with improved OS (p < 0.05), the latter two with PFS (p < 0.05).
Age, gender, stage, KRAS mutation, PD-L1 and ICI type were not associated with PFS
or OS. Pneumonitis as 1st irAE had the worst PFS and OS (p < 0.05). Median starting
corticosteroid dose of ≤60 mg for 1st irAE had an improved OS (p = 0.04). Post 1st irAE
response associated with better PFS and OS (p < 0.05). Number and duration of irAEs and
additional immunosuppressive agents (14% of patients) were not associated with PFS or
OS. Conclusions: In ≥Grade 3 irAEs patients managed with corticosteroids, lower baseline
NLR, longer ICI use, response to ICI after 1st irAE, and a ≤60 mg corticosteroid dose had
promising outcomes.
Keywords: non-small cell lung carcinoma; immunotherapy; toxicity; progression-
free survival; overall survival
1. Introduction
Immune checkpoint inhibitors (ICIs) have served as standard of care treatment for
patients with recurrent or metastatic disease across multiple solid and hematologic ma-
lignancies. They involve interaction with immune checkpoint pathways such as the
PD-1/PD-L1 axis, which plays a key role in tumor immune evasion. T cell activation
is initiated when the T cell receptor (TCR) recognizes tumor-derived peptides presented by
major histocompatibility complex (pMHC) molecules on the tumor surface [1]. However,
PD-L1, expressed on tumor cells, binds to the PD-1 receptor on activated T cells, delivering
an inhibitory signal that suppresses T cell activity and limits anti-tumor immunity [1,2].
Therapeutic blockade of this pathway using anti-PD-1 or anti-PD-L1 antibodies can con-
tribute to restoration of T cell function and enhance immune-mediated tumor clearance [1].
In a distinct immunoregulatory mechanism, the CTLA-4 pathway regulates early T cell
activation in lymphoid tissues. Effective T cell priming requires not only TCR engagement
with pMHC, but also co-stimulatory signaling through CD28 binding to B7-1/2 ligands
on antigen-presenting cells (APCs) [1]. CTLA-4, which is upregulated on activated T cells,
competes with CD28 for B7-1/2 binding, thereby inhibiting T cell activation [1,3]. The use
of anti-CTLA-4 antibodies blocks this inhibitory interaction, preventing CTLA-4 from out-
competing CD28, and thus facilitates T cell activation through positive co-stimulation [1].
ICIs such as anti-PD-L1 and anti-CTLA4 antibodies currently represent a widely
adopted immunotherapy approach for patients with lung cancer [4]. The incorporation
of these agents—either as monotherapy or in combination with chemotherapy—has led
to improvements in most clinical efficacy endpoints compared to chemotherapy alone
in patients with locally advanced or metastatic NSCLC lacking oncogenic driver alter-
ations [5]. However, the increasing number of patients with non-small cell lung carcinoma
(NSCLC) receiving ICIs has led to more patients developing immune-related adverse
events (irAEs) in recent years. A meta-analysis of data from 38 trials of advanced-stage
lung cancer patients reported incidences of Grade 3–5 irAEs for ICI monotherapy (6.6%),
an ICI plus chemotherapy (11.4%), dual ICIs (13.8%), and dual ICIs plus chemotherapy
(13.5%) [6]. Current ESMO Clinical Practice Guidelines suggest hospitalization for most
patients experiencing Grade 3 or higher irAEs with corticosteroid treatment delivered as
1–2 mg/kg/day [7]. However, the impact of irAEs among NSCLC patients on treatment
response to ICI therapy is currently limited. Additionally, there is a paucity of data regard-
ing survival outcomes in the setting of immunotherapy following irAEs. These patients
represent an important clinical population where real-world evidence regarding outcomes
following the development of irAEs can help guide clinicians as to the optimal patient
management strategy.
, Cancers 2025, 17, 2819 3 of 16
We conducted a multi-institutional review of NSCLC patients developing ICI-related
toxicity and managed with corticosteroid agents. Our team aimed to evaluate treatment
response, progression-free survival, and overall survival after onset of irAEs. We also
assessed real-world use of corticosteroids and other novel immunosuppressive agents for
the treatment of Grade 3 or higher ICI toxicities, where published data is currently lacking.
The current management approach involves most patients being subjected to high-dose
corticosteroids to treat irAEs, which itself is associated with significant side effects. We
aimed to find groups of patients experiencing improved survival outcomes with specific
doses of corticosteroid therapy to optimize therapeutic benefit. irAEs frequently necessitate
cessation of ICIs, thereby restricting subsequent treatment options for NSCLC patients [8].
Therefore, we also hoped to clarify the clinical benefit of terminating vs. continuing ICI
therapy following irAEs. Identification of such patients is important given the potential for
long-term benefit and reduced side effects of ICIs compared to other systemic therapies.
This would allow patients to continue ICIs to ensure greater duration of treatment benefit.
We also aimed to identify patients who will not benefit from ICI continuation following
irAEs, thereby aiding physicians to recommend other treatment options or terminating ICI
therapy while transitioning to surveillance in cases of adequate disease control to avoid
subsequent ICI-related toxicities.
2. Materials and Methods
2.1. Patient Selection
We screened adult NSCLC patients to identify eligible patients treated with high-dose
corticosteroids for immunotherapy toxicity as described in Figure 1. We abstracted data
from patient medical records, including clinicians’ notes, radiology, radiation oncology,
operative, and pathology reports.
Figure 1. Flow diagram depicting the selection process of eligible patients for subsequent data
analysis. Abbreviations: CTCAE, Common Terminology Criteria for Adverse Events; ICI, Immune
Checkpoint Inhibitor; irAE, immune-related Adverse Event; NSCLC, Non-Small Cell Lung Carci-
noma; UK, United Kingdom.
