Cytokine release syndrome (CRS) is a well-characterised complication of cellular therapies, particularly chimeric antigen receptor T-cell (CAR-T) therapy, in haematological malignancies [6, 10]. It results from uncontrolled immune activation with excessive cytokine release, producing systemic inflammatory manifestations ranging from fever and malaise to shock and multi-organ dysfunction. Although once considered largely restricted to cellular therapies, growing evidence suggests that immune checkpoint inhibitors (ICIs) can also precipitate CRS in both haematological and solid tumours [7, 11]. Agents implicated include anti-programmed cell death protein 1 (PD-1) and anti-programmed death-ligand 1 (PD-L1) antibodies (e.g., nivolumab, pembrolizumab) as well as anti-cytotoxic T-lymphocyte antigen-4 (CTLA-4) antibodies (e.g., ipilimumab). However, the majority of cases arose from non-gastric cancers and detailed clinical characterisation was lacking. This report uniquely adds to the limited body of evidence describing recurrent CRS in metastatic gastric adenocarcinoma, a setting where the phenomenon remains under-recognised.

Compared with haematological malignancies, the incidence and severity of CRS in solid tumours appear lower. For example, an early trial of CAR-T in colorectal cancer and glioblastoma reported CRS in 34.6% of patients, though the majority were Grade 1–2 with only isolated severe cases [11]. Similarly, real-world pharmacovigilance data covering 2,672 patients with solid tumours, including gastric cancer, identified a 1.1% incidence of CRS, again mostly mild, but with occasional Grade ≥ 3 or fatal cases [12]. In contrast, CRS has been reported in 77–93% of patients with acute lymphoblastic leukaemia or diffuse large B-cell lymphoma receiving CAR-T, with up to 13% developing severe or life-threatening events [6, 13]. The gastric tumour microenvironment (TME) may partly explain the lower overall incidence: it is characterised by chronic inflammation, immunosuppressive stromal and immune cell populations, and antigen heterogeneity, all of which may dampen the systemic cytokine surge seen in haematological cancers where malignant cells are more directly accessible to T-cell–mediated killing [14,15,16]. These differences underscore the importance of documenting cases in gastric cancer to clarify patterns of presentation and risk.

A handful of case reports have linked ICIs to CRS in gastro-oesophageal malignancies, including nivolumab-associated events [17,18,19]. Other PD-1 inhibitors, such as tislelizumab, have also been implicated [20], and similar cases have been described in lung, cervical, and head and neck cancers [21,22,23,24]. This report adds important nuance by demonstrating recurrent CRS in metastatic gastric adenocarcinoma temporally associated with nivolumab–FOLFOX therapy. Unlike most published cases that describe a single CRS episode, our patient developed reproducible, treatment-linked episodes across consecutive cycles, providing rare support for causality. The reproducibility of symptoms in the absence of infection strongly supports an immune-driven mechanism and highlights the potential for cumulative immune sensitisation with continued ICI exposure.

Our case raises the possibility that chemotherapy may potentiate CRS risk by enhancing tumour antigen release, priming immune responses, and facilitating checkpoint inhibitor–mediated hyperinflammation. In this context, the combination of FOLFOX chemotherapy with nivolumab may have synergistically triggered the recurrent inflammatory cascades. Whereas CAR-T–associated CRS is typically driven by rapid T-cell proliferation and high levels of interleukin-6 and interferon-gamma [8, 25], chemotherapy-induced antigen release may act as an amplifier, precipitating delayed CRS. The timing here (after the third cycle) aligns with other reports of ICI-associated CRS, which often manifests later in treatment rather than at initiation [19, 27]. This pathophysiological model warrants further exploration, as ICIs are increasingly combined with cytotoxic backbones in gastrointestinal oncology.

Diagnosing CRS in solid tumour patients is particularly challenging because its manifestations overlap with neutropenic sepsis, tumour-related fever, and chemotherapy toxicity. In this case, the absence of an infectious source on microbiology and imaging, combined with rapid recovery after supportive care, supported an immune-mediated aetiology. The reproducibility of symptoms across successive treatment cycles further strengthens diagnostic certainty and suggests a priming effect, where repeated exposure heightened the inflammatory response. Development of novel diagnostic strategies, such as rapid cytokine profiling or cancer-specific clinical scoring systems, may allow earlier and more confident differentiation of CRS from sepsis in this population [26].

Whilst cytokine panels measuring IL-6, IFN-γ, and soluble IL-2 receptor levels were not used during this case, they may provide valuable diagnostic clues in similar cases, as elevations in these cytokines have been consistently observed in ICI- and CAR-T-associated CRS and correlate with severity [27,28,29]. IL-6 and ferritin have been reported to rise acutely in CRS compared with sepsis, aiding differentiation [30]. Conversely, infection biomarkers such as procalcitonin and C-reactive protein can help distinguish bacterial infection from immune-mediated inflammation: elevated procalcitonin supports the presence of bacterial infection rather than CRS, whereas CRP may rise in both contexts [31, 32]. Integrating these biomarkers into early assessment algorithms could improve diagnostic accuracy and guide timely intervention.

Clinicians should maintain a high index of suspicion for CRS in patients receiving ICIs, particularly when systemic symptoms such as fever, hypotension, or multi-organ dysfunction appear without an identifiable infectious source. Early steps include:

Baseline and early cytokine monitoring (IL-6, IFN-γ, sIL-2R, ferritin) in high-risk patients or those presenting with atypical systemic inflammation [27,28,29,30].

Concurrent infection workup, including procalcitonin, cultures, and imaging, to rule out bacterial causes [31, 32].

Rapid multidisciplinary involvement, including oncology, microbiology, immunology, and intensive care, if systemic inflammatory features worsen [33].

Supportive care, including fluid resuscitation, temporary treatment hold, and close monitoring of organ function [6, 10, 26].

Corticosteroid therapy for recurrent or severe CRS, weighing risks of immunosuppression against benefits [6, 10, 34].

Documentation and pharmacovigilance reporting to refine clinical understanding and guide future management strategies [35].

The potential contributory role of granulocyte colony-stimulating factor (G-CSF) should be acknowledged. In this case, CRS recurred despite the absence of G-CSF, suggesting a secondary rather than primary effect. Nevertheless, G-CSF has been associated with immune-related adverse events, including cytokine-driven phenomena, in other solid tumour contexts [36,37,38]. Its role should therefore not be dismissed and warrants further systematic investigation.

This case highlights the need for structured monitoring and management protocols for CRS in gastric cancer chemoimmunotherapy. Protocolised approaches incorporating dose modifications, corticosteroid prophylaxis in selected patients, or treatment discontinuation after recurrent CRS should be explored in prospective studies [6, 10, 34]. Expanding pharmacovigilance and registry data will be critical to refining evidence-based guidelines and improving patient safety in this emerging treatment landscape [35].