Temporal dynamics of substance P and its association with cytokine release syndrome after CD19 CAR-T therapy in pediatric B-ALL Authors: Águeda Molinos-Quintana, Alfonso Rodríguez-Gil, Teresa Caballero-Velázquez, Estefanía García-Guerrero, Patricia Alcalde-Mellado, Paola Hernández-Díaz, Javier Delgado-Serrano, Victoria Ruiz-Maldonado, Raquel Muñoz-García, Juan Luis Reguera-Ortega, Elia Sánchez-Valderrábanos, José Antonio Pérez-Simón Published: 2026 ABSTRACT BackgroundCytokine Release Syndrome (CRS) remains a major toxicity associated with chimeric antigen receptor T (CAR-T) cell therapy, particularly in pediatric patients. Although neuroimmune mediators have been implicated in systemic inflammation, the role of neuropeptides such as Substance P (SP) in CRS has not yet been explored in this setting.MethodsPlasma SP levels were measured using an enzyme-linked immunosorbent assay (ELISA) and analyzed longitudinally at predefined time points (day − 1, +7, + 14, and + 28) and during CRS when additional samples were available in 18 pediatric and young adult patients with relapsed/refractory B-cell acute lymphoblastic leukemia (R/R B-ALL) treated with CD19 CAR-T therapy (tisagenlecleucel) in a real-world clinical setting. SP dynamics were correlated with severity of CRS, inflammatory biomarkers, and CD19 CAR-T expansion kinetics.ResultsBaseline SP levels were similar between patients who developed severe CRS (grade ≥ 3) and those with non-severe CRS. In contrast, the increase in SP levels from baseline (ΔSP) was significantly higher in patients who developed severe CRS (1523 pg/mL vs. 189 pg/mL, p = 0.01). A peak in SP levels was observed at the onset of severe CRS in all three patients, coinciding with the beginning of clinical toxicity and occurring shortly before the elevation in ferritin levels. In these three cases, IL-6 levels increased in close temporal proximity to the rise in SP at the time of maximum CRS severity. Overall, SP levels declined by day + 14 post-infusion, shortly after the peak of CAR-T expansion. This decline coincided temporally with the administration of anti-CRS treatment in the subset of patients with severe CRS. SP levels later increased toward the end of the observation period (around day + 28), approaching baseline values.ConclusionsThese findings suggest that SP dynamics may be linked to the early inflammatory response during CRS and support further investigation of the SP–NK1 receptor axis as a potential therapeutic target to modulate CD19 CAR-T–related toxicity.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13287-026-05088-0. Manuscript Substance P (SP) is a neuropeptide member of the tachykinin family widely distributed in the nervous system [1]. It is involved in the modulation of inflammation through the communication and regulation of both immunity and hematopoiesis forming the hematological neuroimmune axis [2]. In the bone marrow, C fibers release SP in response to inflammatory stimuli. SP exerts potent pro-inflammatory effects through activation of the neurokinin-1 receptor (NK1R), leading to nuclear factor-κB (NF-κB) signaling, cytokine production, and recruitment of immune cells [3–6]. Moreover, SP may also enhance the expression of chemokines, thereby facilitating leukocyte trafficking and the recruitment of additional inflammatory cells, macrophage activation, endothelial permeability and exacerbating the cycle of inflammation [6–9]. Chimeric antigen receptor T-cell therapy targeting CD19 (CD19 CAR-T) is rapidly emerging as one of the most promising therapies for hematologic malignancies, achieving high rates of durable remission [10, 11]. Despite these advances, CD19 CAR-T is frequently associated with immune-mediated toxicities, most notably cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS) [12, 13]. CRS represents a systemic inflammatory response triggered by rapid immune activation following CAR-T cell recognition of tumor antigens. Mechanistic studies have demonstrated that although CAR-T cells produce early pro-inflammatory mediators such as interferon-γ and GM-CSF, the majority of the cytokine cascade driving CRS originates from activated monocytes and macrophages, which release high levels of interleukin-1 (IL-1) and interleukin-6 (IL-6) [14, 15]. This inflammatory amplification results in a systemic cytokine surge that promotes endothelial activation, vascular permeability, and multi-organ dysfunction in severe cases [16]. Therefore, SP may represent an additional regulatory pathway capable of amplifying inflammatory signaling during CD19 CAR-T–associated immune activation. Exploring the interaction between CD19 CAR-T–induced immune responses and SP-mediated inflammatory pathways could provide further insight into the mechanisms underlying CRS and identify potential therapeutic targets using NK1R antagonist drugs for mitigating severe immune-related toxicity. Our objective was to explore the relationship between the dynamics of serum SP levels after CD19 CAR-T therapy and its role in the cytokine-mediated inflammatory response implicated in acute toxicity among patients with relapsed-refractory B-cell acute lymphoblastic leukemia (R/R B-ALL). All patients met the inclusion criteria of the Spanish state-funded access program that includes patients with relapsed or refractory B-ALL under 25 years old in second or subsequent relapse, or patients with relapse at any time after allogeneic hematopoietic stem cell transplantation. All clinical investigation was conducted according to the principles of the Declaration of Helsinki and was approved by the relevant local institutional ethics Committee Hospitales Universitarios Virgen Macarena y Virgen del Rocío (code: 1458-N-21, date 20/07/2021). Local biobank informed consent and CAR-T therapy informed consent were obtained from all subjects. CRS was graded according to the American Society for Transplantation and Cellular Therapy (ASTCT) consensus criteria to guide treatment [12]. Severe CRS was defined as grade ≥ 3. Peripheral blood samples were determined to measure SP levels during CD19 CAR-T cell therapy on days − 1 (pre-infusion), + 7, +14 and + 28 after infusion. In a subgroup of patients, it was also determined during severe CRS and pre-lymphodepletion. The blood sample was collected in a serum tube containing aprotinin (Vacuette® EDTA aprotinin 4 ml). It was centrifuged at 1000 rpm for 15 min and the supernatant was immediately stored in aliquots at −80° until analysis. SP levels were determined using an enzyme-linked immunosorbent assay (ELISA) kit according to the manufacturer’s instructions (R&D Systems, Abingdon, UK). To evaluate pro-inflammatory cytokine secretion, plasma samples stored at temperatures below − 70 °C were analyzed using commercial ELISA kits for IL-6 (Human IL-6 ELISA MAX™ Deluxe Set, BioLegend, cat# 430504), following the manufacturer’s instructions. The increase in SP levels relative to baseline (ΔSP) was defined as the difference between baseline SP concentration measured on day − 1 (pre-infusion) and the highest SP value observed during the first 14 days after CD19 CAR-T therapy or during the CRS period when occurring later. Statistical analyses were conducted using R software, version 4.4.3. The distribution of the variable SP was assessed through statistical tests and graphical evaluation to examine the assumption of normality. Differences between independent groups were analyzed using the non-parametric Mann–Whitney U test for continuous variables. Generalized linear mixed models (GLMMs) were used for statistical analysis for variables measured repeatedly over time. Models were primarily fitted using a Gamma distribution, which is appropriate for continuous, positive, and right-skewed outcome variables such as SP concentrations. Individual subjects were included as random effects to account for within-subject variability and appropriately model the longitudinal structure of the data. Model assumptions, including residual diagnostics and distributional fit, were thoroughly evaluated using diagnostic tools from the DHARMa package. Post hoc intra- and intergroup comparisons were conducted using estimated marginal means obtained with the “emmeans” package, with p-values adjusted for multiple comparisons using Tukey’s method. Statistical significance was defined as p < 0.05. A total of 18 pediatric and young adult patients with R/R B-ALL received a single infusion of tisagenlecleucel at our institution from 2020 to 2023, ranging in age from 4 to 24 years. The median follow-up time was 52 months (range 31–67 months). Table 1 describes the baseline clinical characteristics and post-CD19 CAR-T toxicity profile of the cohort. CRS occurred in 89% (16/18) of patients, with grade 1–2 CRS observed in 13 patients (72%) and grade ≥ 3 in 3 patients (16.7%). The median time to onset of CRS was day + 2 post-infusion and the median time to resolution was day + 7. Management of CRS included tocilizumab in 50% (9/18) of cases, while 22% (4/18) required corticosteroid therapy. No toxicity-related mortality was documented. Table 1Baseline characteristics, prior treatments, ALL-Hematotox score, and clinical outcomes and toxicity after CD19 CAR-T therapyCharacteristicsN=18Median age, years (range)10.5 (4-24)Female, n (%)7 (38.9) Citogenetic analyses, n(%) ETV6::RUNX1 3 (16.7) Low hypodiplody2 (11.1) IKZF1 plus*1 (5.5) BCR-ABL1 (5.5) B-other10 (55.6) Non available1 (5.5) Previous Allogeneic HSCT, n (%)15 (83.3) Mean % baseline bone marrow blast by MFC (95% CI)20.7 (5.13-36.27) Disease burden prior-infusion, n (%) Low-disease burden (<5% BM blast)10 (55.5) High-disease burden (≥5% BM blast)8 (45.5) Cytokine release syndrome (CRS) after infusion, n (%) CRS grade 02 (11.1) CRS grade 17 (38.9) CRS grade 26 (33.3) CRS grade 33 16.7) Immune effector cell-associated HLH-like syndrome3 (16.7)ICANS any grade3 (16.7)Severe ICANS ≥ 31 (5.5) Anti-cytokine therapy, n (%) Tocilizumab9 (50) Corticosteroids4 (22.2) Anakinra1 (5.5) Response evaluation D28 CR, n(%)3 (16.7) CRi**, n(%)14 (77.8) No response, n(%)1 (5.5) MRD negativity D28, n (%)17 (94.4)Relapse after CAR-T, n/n (%)9/17 (53) ALL Hematotox (Nair M et al), n (%) Low-risk < 4 points10 (55.5) High risk ≥ 4 points8 (45.5)*IKZF1 plus defined by the co-occurrence of IKZF1 gene deletion with specific additions (deletions in CDKN2A/B, PAX5, or PAR1), while excluding ERG.**CRi: Complete remission with incomplete hematologic recovery was defined by CR ≥1 of the following: neutrophils ≤1.0 x 109/L, platelets ≤100 x 109/L, or platelet/blood transfusions within 7 days.HSCT: hematopoietic stem cell transplant; BM: Bone marrow; ANC, absolute neutrophil count; EM: extramedullary disease; CI, confidence interval. MFC: multiparametric flow cytometry. Table 1Baseline characteristics, prior treatments, ALL-Hematotox score, and clinical outcomes and toxicity after CD19 CAR-T therapyCharacteristicsN=18Median age, years (range)10.5 (4-24)Female, n (%)7 (38.9) Citogenetic analyses, n(%) ETV6::RUNX1 3 (16.7) Low hypodiplody2 (11.1) IKZF1 plus*1 (5.5) BCR-ABL1 (5.5) B-other10 (55.6) Non available1 (5.5) Previous Allogeneic HSCT, n (%)15 (83.3) Mean % baseline bone marrow blast by MFC (95% CI)20.7 (5.13-36.27) Disease burden prior-infusion, n (%) Low-disease burden (<5% BM blast)10 (55.5) High-disease burden (≥5% BM blast)8 (45.5) Cytokine release syndrome (CRS) after infusion, n (%) CRS grade 02 (11.1) CRS grade 17 (38.9) CRS grade 26 (33.3) CRS grade 33 16.7) Immune effector cell-associated HLH-like syndrome3 (16.7)ICANS any grade3 (16.7)Severe ICANS ≥ 31 (5.5) Anti-cytokine therapy, n (%) Tocilizumab9 (50) Corticosteroids4 (22.2) Anakinra1 (5.5) Response evaluation D28 CR, n(%)3 (16.7) CRi**, n(%)14 (77.8) No response, n(%)1 (5.5) MRD negativity D28, n (%)17 (94.4)Relapse after CAR-T, n/n (%)9/17 (53) ALL Hematotox (Nair M et al), n (%) Low-risk < 4 points10 (55.5) High risk ≥ 4 points8 (45.5)*IKZF1 plus defined by the co-occurrence of IKZF1 gene deletion with specific additions (deletions in CDKN2A/B, PAX5, or PAR1), while excluding ERG.**CRi: Complete remission with incomplete hematologic recovery was defined by CR ≥1 of the following: neutrophils ≤1.0 x 109/L, platelets ≤100 x 109/L, or platelet/blood transfusions within 7 days.HSCT: hematopoietic stem cell transplant; BM: Bone marrow; ANC, absolute neutrophil count; EM: extramedullary disease; CI, confidence interval. MFC: multiparametric flow cytometry. Baseline characteristics, prior treatments, ALL-Hematotox score, and clinical outcomes and toxicity after CD19 CAR-T therapy *IKZF1 plus defined by the co-occurrence of IKZF1 gene deletion with specific additions (deletions in CDKN2A/B, PAX5, or PAR1), while excluding ERG.**CRi: Complete remission with incomplete hematologic recovery was defined by CR ≥1 of the following: neutrophils ≤1.0 x 109/L, platelets ≤100 x 109/L, or platelet/blood transfusions within 7 days. HSCT: hematopoietic stem cell transplant; BM: Bone marrow; ANC, absolute neutrophil count; EM: extramedullary disease; CI, confidence interval. MFC: multiparametric flow cytometry. Baselines SP showed considerable inter-patient variability across the cohort (range 428–7300 pg/ml). These pre-infusion SP levels were similar between patients who subsequently developed severe CRS (grade ≥ 3) and those with non-severe CRS (Mann–Whitney test, p = 0.74), whereas the increase in SP levels (ΔSP) was markedly higher in patients who developed severe CRS. No significant association was observed between baseline SP levels and the Hematotox score (Spearman correlation, p = 0.62). Baseline SP, ΔSP values, LDH and clinical risk parameters according to CRS severity are summarized in Supplementary Table 1. Given this variability in baseline SP concentrations, we analyzed changes in SP levels for each patient relative to their individual baseline values, defining the increase in SP during immune-related toxicity relative to baseline as ΔSP. Patients who did not develop CRS did not show a substantial increase in SP levels after CD19 CAR-T infusion. In contrast, patients who developed severe CRS exhibited a marked ΔSP, whereas those with non-severe CRS showed intermediate increases (Fig. 1A). Consistent with these findings, the mean ΔSP after CAR-T infusion was significantly higher in patients with severe CRS compared with those with non-severe CRS (1523 pg/ml vs. 189 pg/ml, Mann–Whitney test, p = 0.01) (Fig. 1B). Importantly, a peak in SP levels was observed at the onset of severe CRS in all three patients, coinciding with the beginning of clinical toxicity (Fig. 1C). In these cases that developed severe CRS, increases in C-reactive protein (CRP) were observed shortly after the rise in SP levels, whereas elevations in ferritin and, in some patients, lactate dehydrogenase (LDH) occurred later during the course of CRS suggesting that changes in SP levels may occur early during the inflammatory phase of severe CRS. Notably, later decrease in SP levels was observed to coincide temporally with the administration of anti-CRS treatment (Fig. 1C–D). In addition, the increase in SP appeared to precede the rise in IL-6, with both peaks temporally aligning with the point of maximum CRS severity (Supplementary Fig. 1). Fig. 1Dynamics of SP and association with CRS severity after CAR-T cell therapy in R/R B-ALL patients: Increase in SP levels for each individual patient from baseline (ΔSP) depending on the grade of CRS (1A). Comparison of ΔSP between patients with severe CRS (grade ≥3) and non-severe CRS (1B). Temporal dynamics of SP and inflammatory biomarkers expressed as z-scores in the three patients who developed severe CRS (patients 2, 4, and 6). The shaded areas indicate the duration of CRS, and vertical dashed lines indicate administration of tocilizumab. Blue shading represents corticosteroid treatment and darker shading indicates combined corticosteroids and anakinra (1C). Individual plasma SP concentrations (absolute values) in the same patients, illustrating the relationship between SP kinetics, CRS onset and resolution, and therapeutic interventions (1D). Dynamics of SP and association with CRS severity after CAR-T cell therapy in R/R B-ALL patients: Increase in SP levels for each individual patient from baseline (ΔSP) depending on the grade of CRS (1A). Comparison of ΔSP between patients with severe CRS (grade ≥3) and non-severe CRS (1B). Temporal dynamics of SP and inflammatory biomarkers expressed as z-scores in the three patients who developed severe CRS (patients 2, 4, and 6). The shaded areas indicate the duration of CRS, and vertical dashed lines indicate administration of tocilizumab. Blue shading represents corticosteroid treatment and darker shading indicates combined corticosteroids and anakinra (1C). Individual plasma SP concentrations (absolute values) in the same patients, illustrating the relationship between SP kinetics, CRS onset and resolution, and therapeutic interventions (1D). When considering the entire cohort of 18 patients, longitudinal analysis revealed a characteristic pattern of SP kinetics following CD19 CAR-T infusion. The estimated marginal means showed that after the peak of SP during CRS (median 2–7 days post-infusion), plasma SP levels declined significantly by day + 14 after therapy (p ≤ 0.05) (Fig. 2A). This decrease occurred shortly after the peak of CAR-T cell expansion, which was observed at a median of day + 11 post-infusion (Fig. 2B). At later time points, SP levels increased again toward the end of the observation period (day + 28), approaching baseline values. Fig. 2Longitudinal plasma SP concentrations and CAR-T expansion kinetics in the full cohort. A Plasma SP concentrations over time following CD19 CAR-T infusion. Values represent estimated marginal means derived from generalized linear mixed models (GLMMs), accounting for repeated measures and unequal numbers of observations across time points. Statistical comparisons between selected time points are indicated. The number of patients contributing to each time point was as follows: Day −1 (n=18), Day +1 (n=7), Day +3–5 (n=10), Day +7 (n=13), Day +14 (n=16), and Day +28 (n=15). B CD19 CAR-T expansion kinetics in peripheral blood following infusion. There is an increased during the early post-infusion phase, reaching a peak around day +11, followed by a contraction phase. Values represent longitudinal measurements across the subset of patients with available monitoring (n=11). Statistical comparisons are indicated (*p < 0.001). Longitudinal plasma SP concentrations and CAR-T expansion kinetics in the full cohort. A Plasma SP concentrations over time following CD19 CAR-T infusion. Values represent estimated marginal means derived from generalized linear mixed models (GLMMs), accounting for repeated measures and unequal numbers of observations across time points. Statistical comparisons between selected time points are indicated. The number of patients contributing to each time point was as follows: Day −1 (n=18), Day +1 (n=7), Day +3–5 (n=10), Day +7 (n=13), Day +14 (n=16), and Day +28 (n=15). B CD19 CAR-T expansion kinetics in peripheral blood following infusion. There is an increased during the early post-infusion phase, reaching a peak around day +11, followed by a contraction phase. Values represent longitudinal measurements across the subset of patients with available monitoring (n=11). Statistical comparisons are indicated (*p < 0.001). To further contextualize these findings at the patient level, a swimmer plot summarizing the clinical course of all patients was generated (Fig. 3). This figure illustrates the onset and duration of CRS and the timing of therapeutic interventions including tocilizumab, corticosteroids, and anakinra. Inspection of individual patient timelines suggested heterogeneous SP dynamics across patients, with higher SP concentrations observed in some cases around the period of CRS onset and during the early post-infusion phase, although this pattern was not uniform across all patients. Notably, in patients who developed more severe inflammatory toxicity, larger ΔSP changes appeared to be more evident during CRS. Baseline clinical characteristics corresponding to each patient included in the swimmer plot and the occurrence of ICANS and IEC-HS are provided in Supplementary Table 2. In addition, raw plasma SP measurements for each patient at all sampling time points are provided in Supplementary Table 3. Fig. 3Swimmer plot showing individual patient timelines, CRS events, treatments, and longitudinal SP concentrations following CAR-T infusion. Each horizontal bar represents an individual patient. Day 0 corresponds to the time of CD19 CAR-T cell infusion. Bar colors indicate CRS severity according to ASTCT criteria: blue indicates CRS grade 1–2, red indicates CRS grade ≥3, and grey indicates absence of CRS. Red diamonds mark the onset of CRS. Circles represent measured plasma SP concentrations at each sampling time point. Circle size reflects SP concentration categories (<500, 500–999, 1000–1999, and ≥2000 pg/mL). Swimmer plot showing individual patient timelines, CRS events, treatments, and longitudinal SP concentrations following CAR-T infusion. Each horizontal bar represents an individual patient. Day 0 corresponds to the time of CD19 CAR-T cell infusion. Bar colors indicate CRS severity according to ASTCT criteria: blue indicates CRS grade 1–2, red indicates CRS grade ≥3, and grey indicates absence of CRS. Red diamonds mark the onset of CRS. Circles represent measured plasma SP concentrations at each sampling time point. Circle size reflects SP concentration categories (<500, 500–999, 1000–1999, and ≥2000 pg/mL). Exploratory analyses evaluating the relationship between peak changes in SP (ΔSP) and conventional inflammatory biomarkers showed a trend toward higher ferritin and CRP levels in patients with larger ΔSP values, although these associations did not reach statistical significance (Spearman r = 0.33, p = 0.182 for ferritin; r = 0.30, p = 0.229 for CRP) (Supplementary Fig. 2). Taken together, these results suggest that early increases in SP levels (ΔSP) following CD19 CAR-T infusion, rather than baseline concentrations, are associated with the severity of CRS and may reflect early inflammatory activation during CD19 CAR-T–related toxicity. SP is a neuropeptide with potent immunomodulatory properties mediated through the NK1R, which is expressed on T cells, monocytes, macrophages, and endothelial cells [5]. Engagement of NK1R by SP activates NF-κB and MAPK pathways, driving the release of proinflammatory cytokines such as IL-6, TNF-α, and IL-1β, while also increasing vascular permeability [17–20]. These mechanisms position SP as a potential contributor to the pathogenesis or severity of CRS during CD19 CAR-T cell therapy [21]. In our study, we observed an increase in SP levels at CRS onset particularly evident among patients who developed severe manifestations. This observation suggests that SP dynamics may be temporally associated with the early inflammatory phase of CRS. Notably, SP levels declined by approximately day + 14 after CAR-T infusion, shortly after the peak of CAR-T expansion. Several factors may contribute to this observation. On one hand, intrinsic regulatory mechanisms potentially related to SP–NK1 receptor interactions, such as ligand binding and receptor internalization, may influence circulating SP levels during the inflammatory response [22]. On the other hand, this decrease may also reflect the impact of anti-CRS therapies administered during this period, particularly tocilizumab and corticosteroids, which are known to attenuate systemic inflammation. Consistent with this interpretation, in the three patients who developed severe CRS, the decline in SP levels appeared to temporally coincide with the administration of anti-CRS therapies, as illustrated in Fig. 1C. In these cases, the increase in SP levels was observed earlier than the rise in conventional inflammatory biomarkers such as ferritin and C-reactive protein, suggesting that SP dynamics may precede other markers of systemic inflammation during severe CRS. Therefore, the mechanisms underlying these temporal changes remain speculative and should be interpreted cautiously within the observational design of this study. The temporal dynamics, with SP peaking at CRS onset and declining thereafter, are consistent with a model in which SP may participate in pathways associated with the early amplification of inflammatory signaling during CRS and subsequently modulated (Fig. 4). Fig. 4Schematic representation of the temporal pattern of SP levels during CRS after CD19 CAR-T infusion. The diagram illustrates the observed dynamics of SP levels in relation to the onset of CRS, the subsequent inflammatory phase, and the later recovery period during follow-up. The figure is intended as a graphical summary of the clinical observations reported in this study and does not derive from experimental data Schematic representation of the temporal pattern of SP levels during CRS after CD19 CAR-T infusion. The diagram illustrates the observed dynamics of SP levels in relation to the onset of CRS, the subsequent inflammatory phase, and the later recovery period during follow-up. The figure is intended as a graphical summary of the clinical observations reported in this study and does not derive from experimental data From a clinical perspective, NK1R antagonists such as fosaprepitant, currently approved for chemotherapy-induced nausea and vomiting at a standard single dose of 150 mg intravenous infusion and half-life of ≈ 9–13 h, may represent promising candidates for therapeutic repurposing [23, 24]. Preclinical and clinical studies have shown that blockade of the SP/NK1R axis reduces IL-6, TNF-α, IL-1β production, and NF-κB activation in inflammatory settings [25, 26]. Although NK1R antagonists have not yet been evaluated in the CD19 CAR-T context, NK1R inhibition has demonstrated biological effects in other cytokine storm syndromes, including severe infections and COVID-19, by lowering IL-6 and TNF-α levels [27]. Short-term NK1 receptor blockade with fosaprepitant/aprepitant could potentially mitigate the evolving inflammatory response and reduce CRS severity after CAR-T infusion. A limitation of this study is the relatively small sample size and its observational design, which limit the ability to draw definitive conclusions regarding the mechanistic role of SP in CRS pathophysiology. In addition, SP measurements were obtained in a real-world clinical setting, resulting in some missing values and variability in sampling time points. More frequent serial determinations of SP during the CRS period would have allowed a more detailed characterization of its temporal dynamics. Although circulating cytokines such as IL-1 and IL-6 were not systematically measured across the entire cohort, analysis of the three patients who developed severe CRS revealed a consistent pattern in which the rise in SP preceded, or closely paralleled, the increase in IL-6 at the onset of peak CRS severity. These findings suggest a possible temporal association between SP and IL-6 dynamics, although this observation should be interpreted with caution given the limited sample size. However, in routine clinical practice, CRS diagnosis and grading rely primarily on clinical criteria—including fever, hypotension, and hypoxemia—rather than on specific inflammatory biomarkers, and cytokine measurements are therefore not consistently available in many real-world CAR-T cohorts. Despite these limitations, previous experimental studies have demonstrated that SP can stimulate the production of pro-inflammatory cytokines such as IL-1β, IL-6, and TNF-α in human immune cells, supporting the biological plausibility of the SP–cytokine interaction [28]. A strength of our study is that it reflects real-world clinical practice, including biomarker monitoring and therapeutic interventions guided by routine clinical decision-making. Larger prospective studies with comprehensive biomarker profiling will be required to further define the role of SP in CRS-associated inflammation. Taken together, while additional mechanistic investigations and clinical trials are warranted, our findings provide preliminary clinical observations supporting further investigation of NK1R blockade as a potential therapeutic target in CD19 CAR-T–related toxicity. Supplementary Information Supplementary Material 1. Supplementary Material 1.