Nivolumab Can Trigger Deadly Cytokine Storm, Largest Case Review Finds

Immunotherapy has transformed the treatment of many advanced cancers, but the same immune activation that destroys tumors can, in rare cases, turn against the patient. A new retrospective analysis published in Supportive Care in Cancer examines one of the least recognized complications of the widely used checkpoint inhibitor nivolumab: cytokine release syndrome, a systemic inflammatory storm best known as the signature toxicity of CAR T-cell therapy. By pooling published clinical reports, researchers led by Shaoli Zhao, Wei Sun, Liping Peng and Chunjiang Wang of the Third Xiangya Hospital of Central South University and Changsha Medical University have assembled the clearest picture yet of how this syndrome presents, how it is treated, and how often it proves fatal when the trigger is a PD-1 inhibitor rather than an engineered immune cell product.

The study collected clinical reports of nivolumab-induced cytokine release syndrome published before March 31, 2026, extracting clinical data for retrospective analysis. Thirty-five patients were included, spanning an extraordinarily wide age range from 3 to 86 years, with a median age of 55. That breadth alone carries a message for oncologists: the complication is not confined to any single tumor type, age group or treatment setting. It can emerge in children with refractory Hodgkin lymphoma and in elderly patients with solid tumors alike, wherever nivolumab is deployed to release the brakes on T-cell immunity.

Perhaps the most clinically consequential finding concerns timing. The median time from the start of nivolumab to the appearance of cytokine release syndrome was 54.5 days, with cases reported as early as 0.2 days and as late as 391 days after initiation. The median number of infusion cycles before onset was three, but the range stretched from one to seventeen. This late and highly variable onset distinguishes checkpoint inhibitor-associated cytokine release syndrome from the fulminant, early-onset syndrome seen after CAR T-cell infusion, where fever and hypotension typically erupt within days of cell transfer. A patient who develops fever and falling blood pressure months into nivolumab therapy may not immediately suggest a cytokine storm to the treating team, and the authors argue that the possibility of CRS should be explicitly considered throughout the entire course of administration, not just in the first weeks.

The clinical signature the researchers documented is dominated by fever, which affected 85.7 percent of patients. Hypotension followed in 57.1 percent, tachycardia in 42.9 percent, and fatigue, rash and impaired consciousness each appeared in 25.7 percent of cases. This constellation mirrors the physiology of uncontrolled immune activation: activated T cells and macrophages flood the circulation with inflammatory mediators, producing capillary leak, vasodilation and hemodynamic instability. Neurological involvement in the form of impaired consciousness underscores that the brain is not spared when cytokine levels climb, a feature also recognized in severe CAR T-cell-associated toxicity. In practical terms, a patient on nivolumab who presents with unexplained fever plus hypotension or altered mental status should raise immediate suspicion of CRS, even if the last infusion was months earlier.

Laboratory findings reinforce the inflammatory diagnosis. The reviewed cases commonly revealed liver injury, renal injury, elevated C-reactive protein, elevated ferritin and elevated interleukin-6. Each of these markers has a mechanistic rationale. Interleukin-6 is a central driver of the syndrome, amplifying the inflammatory cascade and contributing to vascular permeability. Ferritin, an acute-phase reactant produced abundantly by activated macrophages, has been proposed in prior literature as a diagnostic and prognostic marker for immune-related adverse events. C-reactive protein rises rapidly under interleukin-6 stimulation and serves as a readily available readout of systemic inflammation. Hepatic and renal abnormalities reflect both direct inflammatory organ injury and the hemodynamic consequences of vasodilatory shock, in which reduced perfusion compounds cytokine-mediated tissue damage.

Treatment in the collected cases followed two main pillars: discontinuation of nivolumab and immunosuppressive or immunomodulatory therapy directed at the cytokine cascade itself. Glucocorticoids and tocilizumab, a monoclonal antibody that blocks the interleukin-6 receptor, were widely utilized as the primary therapeutic approaches. After these interventions, 71.4 percent of patients showed improvement in symptoms and clinical markers. The remaining 28.6 percent died. That mortality figure, drawn from published case reports that may overrepresent severe presentations, nevertheless signals that nivolumab-induced cytokine release syndrome can lead to multi-organ functional impairment and can be fatal. The authors emphasize that the syndrome is not a benign infusion reaction but a potentially life-threatening emergency requiring prompt recognition.

The choice of tocilizumab as first-line therapy reflects experience imported from the CAR T-cell field, where interleukin-6 receptor blockade has become standard for severe cytokine release syndrome and is embedded in consensus grading systems such as the ASTCT criteria and in clinical practice guidelines from the Society for Immunotherapy of Cancer and the American Society of Clinical Oncology. Glucocorticoids, which broadly suppress T-cell and macrophage activation, are typically layered on when tocilizumab alone is insufficient or when neurological symptoms are present. For refractory cases, the broader literature on immune effector cell toxicities describes additional salvage options, including plasma exchange, but the evidence base for such measures in checkpoint inhibitor-associated CRS remains thin and largely anecdotal.

Herein lies the central limitation the authors acknowledge: the standardized and optimal treatment strategy for nivolumab-induced cytokine release syndrome remains undefined. Because the analysis is retrospective and built from individually published case reports, it is vulnerable to reporting bias, inconsistent grading and inherent clinical confounders. Patients described in case reports are often the sickest, which may inflate the apparent mortality, and concomitant medications, infections, combination immunotherapy regimens and underlying malignancies can all mimic or exacerbate the syndrome. The researchers also note that the available evidence base is limited, meaning that while glucocorticoids and tocilizumab are the de facto standards in clinical practice, no prospective trial has defined the ideal timing, dosing or sequencing of these agents specifically for nivolumab-triggered CRS.

The mechanistic puzzle is also unresolved. Nivolumab blocks programmed death-1, a checkpoint receptor that restrains T-cell activity, and cytokine release syndrome after checkpoint blockade is thought to arise when reinvigorated T cells expand and secrete interferon and other mediators that activate macrophages and endothelial cells. Why only a small fraction of the many thousands of patients treated with nivolumab develop this syndrome is unknown, though prior pharmacovigilance analyses of WHO data and case series of checkpoint inhibitor-associated CRS suggest the event is rare and may be more frequent with combination regimens such as ipilimumab plus nivolumab. Host genetic factors influencing interleukin-6 biology, tumor burden, and individual immune repertoire dynamics are plausible contributors that current data cannot disentangle.

For clinicians, the practical takeaways are concrete. Fever developing during nivolumab therapy, particularly when accompanied by hypotension, tachycardia, rash or confusion, should prompt evaluation for cytokine release syndrome rather than automatic attribution to infection or tumor progression. Measurement of C-reactive protein, ferritin and interleukin-6 can support the diagnosis, and early involvement of intensive care may be warranted given the risk of hemodynamic collapse and multi-organ failure. For researchers, the study defines the agenda: prospective registries, standardized grading using consensus CRS criteria, and controlled evaluation of tocilizumab and corticosteroid protocols are needed to move management from case-report empiricism to evidence-based practice. Until then, vigilance remains the most powerful tool, because the window in which glucocorticoids and interleukin-6 blockade can reverse the inflammatory cascade is likely narrow, and the price of missing it, in roughly one of the patients captured in this analysis, was fatal.

Subject of Research: Cytokine release syndrome induced by the immune checkpoint inhibitor nivolumab

Article Title: Clinical characteristics, treatment and prognosis of nivolumab induced cytokine release syndrome

Article References: Zhao, S., Sun, W., Peng, L., & Wang, C. (2026). Clinical characteristics, treatment and prognosis of nivolumab induced cytokine release syndrome. Supportive Care in Cancer, 34(10), Article 1039. https://doi.org/10.1007/s00520-026-11303-8

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11303-8

Keywords: nivolumab, cytokine release syndrome, immune checkpoint inhibitors, tocilizumab, glucocorticoids, interleukin-6, immune-related adverse events, ferritin, hypotension, immunotherapy toxicity, PD-1 inhibitor, retrospective analysis