Diagnostic Evaluation and Classification of Cytokine Release Syndrome (CRS)

Diagnostic Evaluation and Classification of Cytokine Release Syndrome (CRS)

Learning Objective

Apply diagnostic and classification criteria to assess a patient with biologic immunotherapy–associated CRS and guide initial management.

1. Clinical Manifestations of CRS

CRS presents acutely after CAR-T or bispecific antibody infusion with fever, hemodynamic instability, hypoxia, and evolving organ dysfunction. Early recognition of these patterns is critical to prompt intervention.

A. Fever Patterns & Constitutional Symptoms

  • Hallmark: Temperature ≥ 38 °C, typically with an onset 1–7 days post-infusion.
  • Patterns: Can be intermittent or sustained with spiking fevers; often precedes the onset of hypotension and hypoxia.
  • Constitutional Symptoms: Rigors, malaise, anorexia, and myalgias are common, mimicking a severe infection, but the tempo is linked to cellular therapy kinetics.
Clinical Pearl

A new fever within 72 hours of CAR-T infusion without a clear alternative infectious focus mandates a formal CRS evaluation, not just empirical antibiotics alone.

B. Hemodynamic Changes: The Hypotension Spectrum

  • Grade 2 CRS: Characterized by fluid-responsive hypotension, where a mean arterial pressure (MAP) ≥ 65 mm Hg can be maintained with ≤ 2 L of crystalloid bolus.
  • Grade 3–4 CRS: Defined by vasoplegia requiring vasopressor support. Norepinephrine is the first-line agent, with vasopressin considered as an add-on for refractory vasodilation.
  • Fluid Strategy: Judicious use of balanced crystalloids is key. The goal is to restore perfusion without exacerbating fluid overload in the context of capillary leak.
Clinical Pitfall

Delaying the initiation of low-dose vasopressors in a patient with progressive hypotension significantly increases the risk of both end-organ hypoperfusion and iatrogenic fluid overload from excessive crystalloid administration.

C. Respiratory Involvement: Hypoxia & ARDS-like Features

  • Grade 2: Requires low-flow oxygen via nasal cannula (NC) at ≤ 6 L/min.
  • Grade 3: Escalation to high-flow nasal cannula (HFNC) or noninvasive ventilation (NIV) is required.
  • Grade 4: Defined by the need for mechanical ventilation. Lung-protective settings (tidal volumes of 4–6 mL/kg ideal body weight and appropriate PEEP titration) are crucial.
Clinical Pearl

Initiating an ICU transfer when a patient’s oxygen requirement exceeds 4 L/min via nasal cannula allows for timely immunomodulatory therapy and escalation of care, preventing crisis-driven transfers later.

D. Multi-Organ Dysfunction

  • Hepatic: Transaminitis (AST/ALT elevations 5–10× the upper limit of normal), hyperbilirubinemia, and coagulopathy are common.
  • Renal: Acute kidney injury (AKI) per KDIGO criteria can develop. Early consideration of continuous renal replacement therapy (CRRT) can aid in volume control and potential cytokine clearance.
  • Neurologic: Mild encephalopathy may occur and can overlap with Immune Effector Cell-Associated Neurotoxicity Syndrome (ICANS). Close monitoring of the ICE score and mental status is essential.

Case Vignette: A 62-year-old patient, post CAR-T infusion, develops an AST of 450 U/L, creatinine of 2.1 mg/dL, and has an episode of transient hypotension responsive to fluids. This clinical picture warrants initiation of the Grade 3 CRS pathway, including consideration of early CRRT for progressive organ injury.

2. Laboratory & Biomarker Assessment

Laboratory assays are crucial for supporting the diagnosis of CRS and stratifying its severity. However, clinicians must be aware of their limitations, including turnaround time, specificity, and availability.

A. Cytokine Assays (IL-6, IL-10, IFN-γ)

  • Interleukin-6 (IL-6): Levels typically peak 2–3 days post-infusion and correlate with CRS severity. However, a 24–48 hour turnaround time limits its utility for real-time decision-making.
  • IL-10 & IFN-γ: These are complementary markers, but they share similar cost and turnaround time limitations as IL-6.
  • Point-of-Care (POC): Emerging POC platforms for cytokine measurement are currently under validation and not yet standard of care.
Clinical Pearl

In a patient with early signs of CRS, a rapid doubling of the IL-6 level over a 6-hour period can predict impending hemodynamic collapse and may serve as a trigger to administer tocilizumab preemptively.

B. Acute Phase Reactants (CRP, Ferritin)

  • C-Reactive Protein (CRP): Levels begin to rise 12–24 hours after CRS onset, closely paralleling the kinetics of IL-6.
  • Ferritin: Can exceed 10,000 ng/mL in severe CRS, but this finding is nonspecific and can be seen in other hyperinflammatory states.
  • Monitoring: The trend of these markers over time is more informative for tracking disease activity and response to therapy than any single absolute value.
Key Point

A greater than 50% drop in the CRP level within 24 hours after administering tocilizumab is a strong predictor of a favorable hemodynamic response and clinical improvement.

C. Organ Function Panels

  • Chemistry: Serial monitoring of AST, ALT, bilirubin, creatinine, and BUN is essential.
  • Coagulation: INR, aPTT, fibrinogen, and D-dimer should be checked to screen for disseminated intravascular coagulation (DIC).
  • Frequency: Daily monitoring in patients with Grade 3–4 CRS is critical to uncover early organ injury and guide supportive care measures.

D. Emerging Biomarkers & POC Testing

Soluble CD25 (sCD25), angiopoietin-2, and microfluidic cytokine assays show promise in research settings but currently lack the multicenter validation needed for routine clinical use.

Editor’s Note

Insufficient source material exists to provide a detailed protocol for point-of-care (POC) assay integration. A complete section would need to cover assay availability, validation status, performance metrics (sensitivity, specificity), and a clear operational workflow for clinical implementation.

3. Classification Systems for CRS

Standardized grading systems are essential for harmonizing clinical trial reporting, facilitating clear communication between providers, and guiding the urgency of management. The American Society for Transplantation and Cellular Therapy (ASTCT) criteria are the current gold standard.

A. ASTCT Consensus Grading (2019)

The single required feature for grading is a fever ≥ 38 °C. Grading is then determined by the severity of hypotension and/or hypoxia.

ASTCT Consensus Grading for Cytokine Release Syndrome
Grade Hypotension Hypoxia Key Intervention
1 None None Supportive care
2 Needs fluids (>2L) or low-dose vasopressor Nasal cannula ≤ 6 L/min Consider tocilizumab
3 Needs vasopressors (any dose) High-flow O₂ or NIV Tocilizumab ± corticosteroids
4 Needs multiple vasopressors; life-threatening Intubation/mechanical ventilation ICU support; high-dose steroids

Note: Pediatric oxygen and fluid thresholds are typically lower. Always check institutional protocols for age-based adjustments.

B. Alternative Grading Scales

While ASTCT is the standard, other systems exist. The Lee et al. (2014) system was an early five-grade scale that incorporated organ dysfunction beyond just hypotension and hypoxia. Frameworks from SITC and in the EU also exist, often adding specific hepatic and renal criteria, but the core differences are minor.

Clinical Pearl

Initiating tocilizumab at the stage of persistent Grade 2 CRS (e.g., requiring low-dose vasopressors or ongoing fluid resuscitation) has been shown to reduce the rate of progression to severe (Grade 3–4) CRS.

C. Impact on Management

  • Grade 1–2: Management is primarily supportive (antipyretics, fluids).
  • Persistent Grade 2 or Grade 3: Warrants early administration of tocilizumab. Corticosteroids should be added for refractory symptoms.
  • Grade 4: Requires aggressive, combined therapy with IL-6 blockade (tocilizumab) and high-dose corticosteroids in an ICU setting.

4. Differential Diagnosis

It is critical to distinguish CRS from its clinical mimics. The differential diagnosis is guided by the timing of symptom onset relative to infusion, specific laboratory profiles, and the overall clinical context.

A. Sepsis / Systemic Inflammatory Response Syndrome (SIRS)

  • Clinical Overlap: Fever, hypotension, tachycardia, and organ dysfunction are common to both.
  • Key Differentiators: CRS onset is temporally linked to cell infusion, often in the absence of a clear infectious source. The cytokine profile, if available, is distinct.
  • Initial Workup: Always includes blood cultures, procalcitonin, and lactate measurement. Empirical antibiotics should be administered until infection is reasonably excluded.

B. Hemophagocytic Lymphohistiocytosis (HLH) / Macrophage Activation Syndrome (MAS)

Editor’s Note

Insufficient source material was provided for a detailed algorithm on HLH/MAS. A complete section would require specific diagnostic criteria (e.g., HLH-2004), guidance on interpreting ferritin kinetics and sCD25 levels, the role of bone marrow findings, and a discussion of treatment overlaps and distinctions with CRS.

C. Tumor Lysis Syndrome (TLS) & Other Mimics

Editor’s Note

Insufficient source material was provided for this topic. A full section would detail the laboratory criteria for TLS (hyperuricemia, hyperphosphatemia, hypocalcemia), its typical timing relative to therapy, and management overlaps/distinctions with CRS.

5. Clinical Decision Algorithms

Algorithmic approaches that integrate clinical signs and biomarkers are being developed to improve risk stratification and optimize the timing of immunomodulatory interventions.

A. Integrated Scoring Systems

Predictive models, such as the one developed by Komanduri et al., assign points based on early clinical and laboratory data to generate a risk score. This approach aims to identify high-risk patients before they clinically deteriorate.

CRS Predictive Model Workflow A flowchart showing how early clinical and lab data (fever, hypotension, CRP/IL-6) are used to calculate a predictive score, which then guides the decision to either monitor or give early tocilizumab. Early Data Inputs Early Fever Magnitude Degree of Hypotension CRP / IL-6 Trends CalculatePredictive Score Score >Threshold? Yes Tocilizumab No Monitor
Figure 1: Conceptual Workflow of an Integrated CRS Scoring System. Early clinical and laboratory data are combined to generate a predictive score. If the score exceeds a predefined threshold, it triggers an alert for early intervention with tocilizumab, aiming to prevent progression to severe CRS.

B. Early vs. Delayed Classification

A key challenge is determining the optimal time to formally grade CRS. Early classification may lead to overtreatment of transient symptoms, while delayed classification risks life-threatening progression. Currently, no consensus exists on the optimal timing beyond the initial onset of fever.

C. Biomarkers in Clinical Workflows

The integration of biomarkers into real-time clinical workflows remains largely investigational. Some centers are piloting EHR-based alerts that combine vital sign changes with laboratory trends to prompt clinicians to perform a formal CRS grading. However, clinical grading remains the standard of care, as real-time cytokine monitoring has not yet been proven to improve outcomes.

Key Clinical Pearl

Implement institution-specific CRS pathways with clear triggers (e.g., fever onset) for formal grading. Utilize pharmacy-led, pre-built order sets for tocilizumab and corticosteroid dosing to streamline care, reduce cognitive load on providers, and avoid critical delays in therapy.

References

  1. Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus grading for CRS and neurologic toxicity associated with immune effector cells. Biol Blood Marrow Transplant. 2019;25(4):625–638.
  2. Teachey DT, Lacey SF, Shaw PA, et al. Identification of predictive biomarkers for CRS after CAR T-cell therapy for ALL. Cancer Discov. 2016;6(6):664–679.
  3. Shimabukuro-Vornhagen A, Gödel P, Subklewe M, et al. Cytokine release syndrome. J Immunother Cancer. 2018;6(1):56.
  4. Hay KA, Hanafi LA, Li D, et al. Kinetics and biomarkers of severe CRS after CD19 CAR-T cell therapy. Blood. 2017;130(21):2295–2306.
  5. Komanduri KV, Belousov A, Byrtek M, et al. Development of a predictive model for cytokine release syndrome to inform risk stratification. Blood. 2021;138(Supplement 1):1459.