Introduction Post-transplant cyclophosphamide (PTCy) reduces acute graft versus host disease (GVHD) but alters vascular biomarkers, with an increase in angiopoietin -2 at day+28 post hematopoietic stem cell transplant (HSCT) in adult transplant recipients (Newell et al, Blood VTH 2024). Abatacept has been shown to reduce endothelial injury by reducing neutrophil extracellular traps at day+14 post HSCT. Objectives Compare the clinical incidence of transplant associated thrombotic microangiopathy (TA-TMA) and vascular biomarkers in PT-Cy to three different acute GVHD prophylaxis regimens in children and young adults. Methods Patients were categorized into four GVHD prophylaxis cohorts 1) Cyclosporine (CSA) and mycophenolate mofetil (MMF), 2) CSA, MMF and abatacept, 3) post-transplant cyclophosphamide (PTCy) and 4) T cell depleted (TCD) graft. ELISA for suppression of tumorigenicity 2 (ST2), vascular endothelial growth factor receptor 1 (VEGF-1) and angiopoietin-2 (ANG2) were performed at baseline and 30 days post- HSCT in all four groups. Clinical incidences of acute GVHD and TA-TMA by day+100 were compared between all 4 cohorts. Results Eighty-one children were included as follows: CSA/MMF (n=19), CSA/MMF/abatacept (n=24), PTCy (n=19) and TCD (n=19). Patient demographics are shown in (Figure 1). ST2, ANG-2, and FLT-1 were not different between PTCy and CSA/MMF/abatacept at baseline. (Figure 2). At day 30 post HSCT, ST2, VEGF-1 and ANG2 were lower in the CSA/MMF/abatacept group compared the PTCy cohort (Figure 2). Incidence of TA-TMA in the PTCy group was 84% (n=16/19) and 47% (n=9/19) in the CSA, MMF group, 0%(n=0/24) in abatacept cohort and 5%(n=1/19) in TCD (p<0.001). The incidence of acute GVHD was 31% (n=6/19) in the CSA/MMF group, 16% (n=4/24) in the CSA/MMF/ABA cohort, 52% (n=10/19) in the PTCy cohort and none in the TCD cohort(p=0.007) (Figure 3), likely driven by HLA disparities between the PT CY and CSA/MMF/abatacept cohort. Conclusion Vascular biomarkers of endothelial injury and clinical TA-TMA were higher in the PTCy group compared to the patients that received abatacept as part of their GVHD prophylaxis. The incidence of aGVHD was higher in the PTCy group. Further studies of the use of abatacept as part of GVHD prophylaxis regimens in children and specific vascular biomarkers should be studied further.

Background One third of patients still die from transplant associated thrombotic microangiopathy (TA-TMA) despite prompt treatment with eculizumab. We hypothesize that cellular and soluble complement modifiers contribute to eculizumab resistance in TA-TMA and knowledge of these mechanisms may guide novel biomarkers and therapies. Objectives We sought to perform a multiomic analysis of TA-TMA patient samples to determine mechanisms of eculizumab resistance. Methods A pharmacokinetic (PK)/pharmacodynamic (PD) analysis of eculizumab levels and sC5b-9 in 60 TA-TMA patients was performed to identify eculizumab responders and non-responders (described in Abstract #28077). Peripheral blood mononuclear cells (PBMCs) from responders (n=4) and non-responders (n=4) underwent single cell RNAseq at 2 timepoints: 1) TA-TMA diagnosis (prior to eculizumab initiation) and 2) at the time of eculizumab “escape” in non-responders (or matched timepoint in responders). Plasma from the same patients and timepoints underwent proteomic analysis (Somascan 11k). Results Non-classical monocytes (NCMs) were markedly different in responders and non-responders at diagnosis (Fig.1) and eculizumab escape (Fig.2). Non-responder NCMs at TA-TMA diagnosis were enriched for interferon alpha/beta (p=7.3e-9) and interferon signaling (p=3.9e-5). This aligned with the plasma proteome at TA-TMA diagnosis, which was enriched for interferon signaling (p=1.5e-3) and interferon alpha/beta signaling (p=7.9e-3) in non-responders. NCMs were also enriched for DNA damage/telomere stress induced senescence (p=2.8e-9), senescence-associated secretory phenotype (p=3e-4) and oxidative stress induced senescence (p=4.3e-3). Cell senescence is induced by interferons and impedes the crucial role of NCMs in vascular maintenance and repair.The role of senescence in TA-TMA is also seen at the time of eculizumab escape. NCMs in non-responders showed inhibited eukaryotic translation elongation (p=2.9e-33), EIF2 signaling (p=1.5e-30), eukaryotic translation initiation (p=1.6e-27) and RNA processing pathways (p=2.2e-25). These pathways are crucial for normal protein synthesis, and the strong inhibitory signal supports ongoing cell dysfunction from interferon-induced senescence. The plasma proteome at the time of eculizumab escape was also enriched for DNA damage/telomere stress induced senescence (p=4.6e-6), senescence pathway (p=8.8e-5) and senescence-associated secretory phenotype (p=3.1e-4) in non-responders. Conclusions We performed the first PK/PD guided multiomics analysis of PBMCs and plasma from eculizumab responders and non-responders. This study identified the novel role of interferon-induced cell senescence of NCMs in the biology of eculizumab resistance in TA-TMA. Multiple available therapies aim to prevent cell senescence and are therefore of significant interest in TA-TMA.

Introduction Diarrhea in pediatric transplant recipients is common but difficult to distinguish as infectious and gastrointestinal graft versus host disease (GI GVHD), which has major therapeutic implications. Amphiregulin (AREG), an epidermal growth factor ligand, is secreted by gut effector cells and alloreactive T cells when stimulated by IL-33. Detection of AREG in peripheral circulation reflects gut damage or direct secretion from immune cells. AREG levels are elevated in adult HSCT recipients with gut graft versus host disease (GI GVHD) at diagnosis. Objective We hypothesized that serum amphiregulin levels would be higher in acute GI GVHD compared to adenoviral colitis. Methods We selected two cohorts of patients on whom we had cryopreserved samples in our institutional biorepository as follows: 1) Adenoviral colitis and no acute GI GVHD at time of diagnosis (n=13) 2) Acute GI GVHD and no adenoviremia, at diagnosis(n=14). Serum amphiregulin levels were determined by ELISA. Regulatory islet 3-α (Reg3-α) levels and intestinal fatty acid binding protein (IFABP) were assessed by ELISA at same timepoints to study patterns of gut injury. Results Patient demographics are shown in Figure 1. Median age of adenoviremia cohort was 2.3 years (range 0.7-4.9 years) and 14.9 years (range 4-34 years) in the acute GI GVHD cohort. Median AREG levels were more than 2-fold higher in acute GI GVHD compared to adenoviral colitis (median 39.3 vs 15.6 pg/mL, p=0.02). This pattern was consistent with Reg3-α (higher in GI GVHD p=0.04 ; Figure 2B) but differed from IFABP (lower in GVHD p=0.008 Figure 2C), suggesting distinct mechanisms of gut injury and likely reflecting lower enterocyte mass in acute GI GVHD from mucosal damage. Conclusion Serum amphiregulin may serve as a clinically useful biomarker to distinguish GI GVHD from infectious colitis in children, enabling earlier and more accurate treatment decisions. Patterns of Reg3-α and IFABP further support mechanistic differences in gut injury.

Introduction Transplant associated thrombotic microangiopathy (TA-TMA) is a well described disease of small vessel thrombosis, vasculopathy and intravascular hemolytic anemia diagnosed in over one-third of hematopoietic stem cell transplant (HSCT) recipients that can cause multiorgan system injury via hyperactive complement activation, resulting in significant morbidity and mortality. We hypothesized that kidney fibrosis biomarkers could identify subclinical kidney fibrosis in children with TA-TMA, facilitating early intervention and identification of those who could benefit from prophylactic strategies. Objective To investigate candidate biomarkers of kidney fibrosis in pediatric and young adult allogeneic HSCT recipients with and without TA-TMA. Methods All participants were consented and prospectively enrolled in our IRB-approved HSCT registry and biorepository. Enzyme-Linked Immunosorbent Assays (ELISAs) were performed on cryopreserved urine samples from a cohort of 90 allogeneic HSCT recipients (45 with TA-TMA and 45 without TA-TMA) to quantify 3 candidate soluble factors (monocyte chemoattractant protein-1 (MCP-1), transforming growth factor beta (TGF-β) and Serpin E1) at 3 timepoints: baseline (prior to the start of the chemotherapeutic preparative regimen), day 14 and day 100 after HSCT. Additional clinical data was collective by retrospective chart review. Results As shown in Figure 1, Serpin E1 was detected significantly more frequently in the urine of patients with TA-TMA both at baseline (p=0.04) and day 14 (p=0.04). The concentration of urine MCP-1 in participants with and without TA-TMA was not significantly different at any timepoint (Fig. 2A). Urine TGF-β was significantly lower at day 14 in patients with TA-TMA (Fig. 2B, p=0.02). Participants with detectable urine Serpin E1 levels at baseline had significantly decreased median Cystatin C GFRs at baseline (p=0.045) and day 60 (p=0.027) (Fig. 3A). Participants with detectable urine Serpin E1 levels at baseline also had significantly higher median urine protein/creatinine ratios at baseline (p=0.0005), day 14 (p=0.0024), day 30 (p=0.0037) and day 60 (p=0.0086) after HSCT (Fig. 3B). As shown in Figure 3C, participants with detectable urine Serpin E1 levels at baseline were significantly more likely to require multi-agent anti-hypertensive therapy during their HSCT course (p=0.0012). Conclusion Patients with detectable urine Serpin E1 (also known as plasminogen activator inhibitor-1 (PAI-1)) are more likely to develop TA-TMA and have reduced kidney function during their HSCT course, making Serpin E1 a promising urine biomarker for early screening of subclinical renal fibrosis. Pre-transplant urine Serpin E1 differences indicate this may be due to a novel predisposition to kidney injury and/or TA-TMA, and merits further study.

Background Chronic graft-versus-host disease (cGVHD) affects ∼40% of allogeneic transplant recipients and is a major cause of morbidity and mortality. Although NIH consensus criteria standardize definitions, their application in routine practice is inconsistent. Early recognition is essential, yet screening and documentation vary widely. The Engraft Learning Network conducted a multicenter assessment to characterize this variation and identify opportunities for standardization across pediatric and adult transplant programs. Methods 30 structured interviews were completed with clinician representatives from 10 Engraft Learning Network centers (2 adult, 8 pediatric). Interviews explored timing and frequency of screening, organ-specific review of systems (ROS) and physical examinations, documentation approaches, and interpretation of findings. Each interview was transcribed, summarized for provider validation, and visualized in heat maps to illustrate cross-center variability. Results Marked heterogeneity was observed across all domains of cGVHD screening (Figure). Assessment schedules ranged from weekly to as-needed; 53% performed monthly evaluations in the first six months post-transplant, and only 37% continued quarterly assessments beyond one year. Documentation methods varied, 20% used structured flowsheets, 33% used dot phrases, and 47% relied on free-text notes. Only 57% documented NIH organ-based cGVHD scoring, and 10% incorporated the Lee Symptom Scale. Skin was universally assessed, though only 13% performed comprehensive exams. Oral evaluations were common, but comfort identifying oral cGVHD was modest (median 6/10). Ocular screening was often limited (17% not routinely assessed). Pulmonary evaluations were standard, but PFT intervals differed. Gastrointestinal and genitourinary reviews showed the greatest inconsistency, fewer than half routinely screened for dysphagia or GU symptoms. Musculoskeletal assessments were incomplete at many centers, and rare manifestations (e.g., neuropathy) were seldom addressed (<10%). Conclusions cGVHD screening practices vary widely in timing, documentation, and completeness across centers, contributing to delayed recognition and inconsistent management. These findings highlight the need for a standardized, evidence-informed approach with defined domains, assessment intervals, and structured documentation. Engraft, a collaborative of pediatric and adult transplant programs advancing outcomes through shared learning, data-driven research, and co-designed quality improvement, is developing this bundle to reduce variability, enable earlier detection, and foster cross-center learning. By translating guidelines into practical, patient-centered workflows, this initiative aims to reduce morbidity and improve quality of life and long-term outcomes for transplant recipients.

There are no biomarkers or therapies for graft rejection after hematopoietic stem cell transplant (HSCT) due to a limited mechanistic understanding of this event. We hypothesized that interferon activation from recipient cells drives donor cell elimination during graft rejection. To test this hypothesis, we studied plasma proteomes of HSCT recipients with graft rejection in comparison to febrile HSCT recipients who engrafted. Leading differentially expressed proteins (DEPs) included CXCL11 (q=2e-5) and CXCL10 (q=3.4e-3). The most highly enriched pathway was interferon gamma response (q= 8.59e-10). Interferon gamma, tumor necrosis factor (TNF) and interferon alpha were also the leading upstream regulators of DEPs. Drug network interactomes for interferon gamma and TNF inhibitors confirmed these therapies may inhibit activated pathways. We then performed single cell RNAseq on peripheral blood mononuclear cells (PBMCs) from a graft rejection patient and compared PBMCs at the time of rejection to a prior timepoint after initial engraftment. PBMC transcriptomes were separated by donor (female) and recipient (male) origin, revealing a marked increase in recipient NK cells (3.5% versus 83.7%) and recipient effector memory T-cells (8% versus 91.8%) during rejection. Donor classical monocytes during rejection showed ferroptosis pathway enrichment, an interferon-mediated form of cell death. To our knowledge, this is the first single cell RNAseq study in humans to confirm recipient NK cell and T-cell resurgence during graft rejection after HSCT. We additionally identified promising biomarkers (CXCL11, CXCL10, LAG3), a potential mechanism of donor cell death (ferroptosis) and targetable immunologic pathways (interferon gamma, TNF) meriting further study.

Introduction Transplant-associated thrombotic microangiopathy (TA-TMA) is a life-threatening complication after hematopoietic stem cell transplantation (HSCT), driven by endothelial injury and uncontrolled complement activation. Eculizumab, a terminal complement (C5) inhibitor, prevents membrane attack complex formation and reduces mortality in high-risk disease. Since 2014, our institution has applied a standardized pharmacokinetic/pharmacodynamic (PK/PD) algorithm for eculizumab, targeting trough concentrations ≥100 µg/mL with undetectable CH50. This precision strategy enables early therapeutic optimization, timely discontinuation, and avoidance of prolonged complement blockade. Objectives To evaluate eculizumab utilization, dosing efficiency, and survival outcomes in patients with high-risk TA-TMA treated under a uniform PK/PD-guided protocol developed collaboratively by the clinical pharmacology, BMT and pharmacy teams. We aimed to demonstrate the value of precision dosing, adherence, and stewardship across a decade of practice. Methods We retrospectively reviewed all patients with high-risk TA-TMA treated with eculizumab between 2012-2025 using published Jodele diagnostic and risk-stratification criteria (Jodele, Blood 2014 and 2024). The pharmacology team directed therapy through serial eculizumab levels and complement activity monitoring to guide initiation, escalation, and discontinuation (Mizuno, Blood Adv 2022). Annual dosing trends and outcomes were compared across pre- and post-implementation eras. Results Among 161 patients analyzed, a total of 2,331 eculizumab doses were administered, with a median of 8 doses per patient (IQR 7-9) - a remarkably brief course for a condition historically associated with high mortality. Following PK/PD implementation in 2014, dosing variability decreased and practice adherence improved, reflecting seamless coordination between pharmacy and medical teams. Survival of high-risk TA-TMA improved to >70%, compared to 16% in historical controls treated before PK/PD guided therapy (Jodele, Blood 2024). These findings confirm that precision dosing achieves optimal biologic effect with fewer doses, minimizing cost and exposure while sustaining complement control. Conclusion Over a decade, our PK/PD-guided eculizumab program jointly led by clinical pharmacology, BMT, and pharmacy transformed TA-TMA management. Achieving survival gains exceeding four-fold improvement with a median of only eight doses per patient exemplifies how collaborative, data-driven dosing can “do more with less.” This model highlights pharmacists as essential partners in precision biologic therapy, integrating science, stewardship, and patient-centered care in complex transplant populations.

Background LDL-C lowering reduces cardiovascular risk, but very low levels may paradoxically increase all-cause mortality and impair endothelial function. The role of LDL-C in HSCT has not been defined. PCSK9 regulates LDL receptor turnover and is implicated in vascular injury. We evaluated the prognostic impact of baseline LDL-C and the dynamics of PCSK9 and oxidized LDL (oxLDL) in allogeneic HSCT recipients. Methods Lipid profiles (HDL-C, LDL-C, total cholesterol, triglycerides), oxLDL, and PCSK9 were measured in 86 pediatric and young adult allogeneic HSCT recipients. Cumulative incidence (CI) of transplant-associated thrombotic microangiopathy (TA-TMA) and 1-year overall survival (OS) were analyzed by stratifying baseline and day +14 lipoproteins above or below cohort medians. Results Profound shifts in lipid metabolism were observed between baseline and day +14 (Fig. 1). Median LDL-C increased from 72 mg/dL (range 15–206) to 125 mg/dL (22–289, p<0.0001). HDL-C fell sharply from 42 mg/dL (5–91) to 13 mg/dL (5–65, p<0.0001), and triglycerides decreased from 114 mg/dL (36–700) to 87 mg/dL (5–979, p=0.03). Baseline LDL-C was the most powerful predictor of outcomes: patients with LDL-C <72 mg/dL had a 1-year TA-TMA CI of 50% compared with 24% in those ≥72 mg/dL (p=0.01, Fig. 2A). Low baseline LDL-C was also associated with significantly inferior OS (88% vs 98% at 1 year, p=0.05, Fig. 2C). Neither HDL-C nor triglycerides stratified risk of TA-TMA or OS. OxLDL rose from 1,992 pg/mL (500–5,960) to 2,267 pg/mL (500–9,868, p=0.02), and PCSK9 increased from 503,751 pg/mL (175,433–1,390,133) to 663,125 pg/mL (383,725–2,243,403, p<0.0001) (Fig. 3A–D). Change–change analyses demonstrated ΔLDL-C correlated with ΔoxLDL (r=0.29, p=0.01), whereas ΔPCSK9 showed a modest inverse correlation with ΔoxLDL (r=−0.22, p=0.04) (Fig. 3E–F), supporting coordinated but distinct regulatory pathways. Baseline PCSK9 did not stratify 1-year TA-TMA risk (≥510,155 pg/mL vs <510,155: CI 0.42 vs 0.37; p=0.71), and baseline oxLDL likewise showed no discrimination (≥1,992 pg/mL vs <1,992: CI 0.36 vs 0.43; p=0.54) (Fig. 3G–H). Conclusions Low baseline LDL-C is a simple, clinically available biomarker that identifies patients at risk for TA-TMA and inferior survival following HSCT. In contrast, PCSK9 and oxLDL—despite rising post-transplant—did not predict TA-TMA when dichotomized, indicating their role as dynamic correlates of injury rather than baseline risk. Integrating LDL-C into pre-HSCT risk assessment may enable earlier identification of vulnerable patients and guide preventive strategies. Modeling continuous and time-varying PCSK9 and oxLDL will refine mechanistic understanding and inform selection of patients for lipid-targeted interventions.

Background Increased oxidative stress in the early hematopoietic cell transplant (HCT) period can lead to endothelial injury and subsequent sinusoidal obstructive syndrome (SOS) and/or transplant-associated thrombotic microangiopathy (TA-TMA). Vitamin C is an important antioxidant that is critical for preventing reactive oxygen species (ROS) mediated endothelial dysfunction in sepsis and ischemia/reperfusion models. Vitamin C has been shown to reduce C reactive protein (CRP), a marker of inflammation. Vitamin C kinetics are poorly described in children undergoing HCT and in the context of post-HCT endothelial injury syndromes. Objectives Low peri-HCT vitamin C levels are associated with increased risk of endothelial injury syndromes. Methods We prospectively analyzed vitamin C levels and CRP in 61 pediatric and young adult patients undergoing first allogeneic HCT at our center from August 2024 to July 2025. Levels were assessed at baseline (pre-conditioning), days 0, 14, and 28. Continuous variables were assessed via the Mann-Whitney U test and correlation with Spearman's rank correlation test. Clinical outcomes were assessed in 51 patients who had adequate follow-up through at least day +100. Results Patient characteristics and outcomes are shown in Figure 1. Median age at HCT was 5.8 years (range 0.12-28.6). Baseline vitamin C was similar between underlying diagnoses. Vitamin C decreased significantly during the early post-HCT period, despite dietician-managed supplementation in most (N=46/61, 75%) and inversely correlated with CRP (Figure 2). Vitamin C was lower at baseline (median 43 vs. 68 μmol/L, p=0.04) and day +14 (median 23 vs. 52 μmol/L, p=0.006) in patients who developed SOS and/or high-risk TA-TMA compared to patients without SOS and/or TMA (Figure 3A). Additionally, median vitamin C level at day +14 was 24 μmol/L in all risk TMA compared to 50 μmol/L in no TMA (p=0.05). Median vitamin C level at day +28 was 34 μmol/L in SOS compared to 60 μmol/L in patients without SOS (p=0.03). Lower vitamin C levels at day +28 (median 13 μmol/L vs. 51 μmol/L) was associated with higher day +100 mortality (p=0.003, Figure 3B), with endothelial injury syndromes contributing to causes of death. There was no correlation of vitamin C levels with acute GVHD at the assessed timepoints. Conclusion Low vitamin C post-HCT correlates with increased incidence of endothelial injury-related HCT complications and mortality in children, despite dietician-managed supplementation. We have previously shown that lysis of hematopoiesis during conditioning leads to a surge of oxidant stress in the early period after transplant (Cook et al, TCT 2025). Our data suggest that vitamin C may be consumed during this time, and that conventional supplementation is not adequate. Moreover, CRP might be a useful surrogate biomarker of adequacy of supplementation.

Background Kidney injury is a hallmark of transplant-associated thrombotic microangiopathy (TA-TMA) and cystatin C estimated glomerular filtration rate (GFR), serum creatinine, and random urine protein to creatinine ratios (rUPCR) are common clinical measures of kidney function. We hypothesized that these widely available tests may differentiate TA-TMA patients from those without TA-TMA and may identify innate differences in kidney function prior to transplant and long term. Objectives We longitudinally analyzed clinically obtained cystatin C GFR, serum creatinine, and rUPCR in a large, single-center, cohort of pediatric allogeneic hematopoietic stem cell transplant (HSCT) recipients. Methods A total of 261 allogeneic HSCT recipients were prospectively screened for TA-TMA. Cystatin C GFR, serum creatinine, and rUPCRs were compared between TA-TMA patients and patients without TA-TMA across 11 timepoints: baseline (pre-HSCT), day 0, 7, 14, 21, 28, 35, 42, 100, 180 and 365 after HSCT. rUPCR elevations were categorized as mild (0.2-0.99 mg/mg), moderate (1.0-1.99 mg/mg), or severe (≥2 mg/mg) proteinuria. Results Cystatin C GFR was significantly lower in TA-TMA patients at all 11 timepoints studied (Fig.1A). Eight out of 11 timepoints had a p-value ≤0.001 including baseline (median 107 vs 121 mL/min/1.73m2, p=0.001). GFR decline from baseline occurred by day 21 in both TA-TMA and no TA-TMA cohorts, after which the median GFR remained stable through 1 year after HSCT. Serum creatinine, another marker to estimate GFR, was not different between TA-TMA patients and patients without TA-TMA at any timepoint (Fig.1B).rUPCRs were significantly higher in TA-TMA patients at all points studied. Nine out of 11 timepoints studied had a p-value ≤0.0001. TA-TMA patients had significantly more total days with proteinuria (median 126.5 vs 42 days, p<0.0001) and a higher maximum rUPCR (median 6.6 vs 0.8 mg/mg, p<0.0001) during transplant (Fig.2A/C). Days with mild (median 59 vs 35 days, p<0.0001), moderate (median 14 vs 0 days, p< 0.0001) and severe (median 19 vs 0 days, p< 0.0001) proteinuria were also significantly higher in TA-TMA patients. Conclusions Cystatin C GFR and rUPCR are widely available clinical tests of kidney injury and our study found that both are consistently different in TA-TMA patients through 1-year post-transplant. We observed that TA-TMA patients have lower GFRs prior to transplant and an early decline in GFR that later stabilizes without acute worsening during active TA-TMA. In contrast, the rUPCR increased acutely during TA-TMA but gradually increased by 1 year. Cystatin C GFR and rUPCR are therefore valuable, inexpensive tools for identifying and monitoring TA-TMA patients, whereas serum creatinine is less informative in pediatric TA-TMA.

Introduction Chronic kidney disease (CKD) is a proinflammatory state associated with endothelial dysfunction. Decreased glomerular filtration rate (GFR) prior to hematopoietic stem cell transplant is associated with need for renal replacement therapy (RRT) post-HSCT and increased overall mortality in adults. Little is known about outcomes for pediatric patients who have decreased GFR prior to HSCT. Objective Evaluate if decreased GFR prior to pediatric HSCT is associated with more endothelial dysfunction syndromes, RRT, or mortality. Methods All pediatric BMT patients between September 2016 and August 2024 with a nuclear medicine GFR (NM GFR) less than 90 and were over 2 years old at the time of BMT were evaluated and grouped according to severity of kidney dysfunction: less than 60, between 60 and 79, and between 80 and 89 mL/min/1.73m2. Demographics, transplant characteristics, NM GFR, cystatin C, creatinine, and RRT needs at baseline and at 1-year post-HSCT were reviewed. Overall survival was compared across GFR groups using Kaplan–Meier curves and the log-rank (Mantel–Cox) test. Given the ordered nature of GFR groups, we also performed a log-rank test for trend. Fishers exact test was used to compare variables with a significance of p<0.05. Results There were patients 10 with GFR < 60, 16 with GFR 60 – 79, and 18 with GFR 80-89 mL/min/1.73m2. Transplant characteristics are shown in Figure 1. Patients in highest GFR group had better 1-year overall survival, with a difference among the three groups (p<0.05) and a significant ordered trend (p<0.01) (Figure 2). Need for RRT increased as GFR decreased (p<0.01) (Figure 3a), and need for RRT was associated with 1-year all-cause mortality (p < 0.0001) (Figure 3b); the odds of death were ∼37-fold higher in the RRT group (OR = 37.33, 95% CI 5.14–204.0). Of the patients in the GFR <60 group, 60% required RRT, with an average length of continuous RRT of 17 days and intermittent hemodialysis of 24 days. GFR was not associated with endothelial injury syndromes (TMA, VOD, GVHD) (Figure 3c). For survivors, change in estimated cystatin C GFR at 1 year was statistically significant (p = 0.0013) with an average decrease of 13.2 mL/min/1.73m2. No survivors required chronic RRT. Conclusion Decreased GFR prior to pediatric HSCT is associated with a higher risk of need for RRT and all-cause 1-year mortality. Survivors had a significant decrease in kidney function over the course of the transplant, but none required chronic RRT. Low GFR should not be prohibitive for HSCT. Additionally, for patients with baseline GFR <60, placement of a dialysis catheter upfront may be beneficial given the increased risk of RRT in patients in this group. Further studies including larger cohorts are warranted to understand the impact of decreased kidney function before HSCT in the pediatric population.

Background Pathologic complement activation is a key driver of transplant-associated thrombotic microangiopathy (TA-TMA) and is thought to occur following one or more endothelial injury events. In this study we hypothesized that endothelial cell specific changes are present in TA-TMA patients prior to transplant and influence the development and progression of TA-TMA during transplant. Objectives We sought to study the plasma proteome in hematopoietic stem cell transplant (HSCT) recipients with and without TA-TMA to identify influential endothelial cell processes. Methods The plasma proteome of 25 allogenic HSCT recipients (11 with TA-TMA, 14 without) was studied using the Somascan 7k platform across 6 timepoints: baseline, day 7, 14, 30, 60 and 90. To validate our findings, we reanalyzed published bulk RNAseq data from our novel TA-TMA in vitro model which involve culturing primary kidney endothelial cells with serum from patients with TA-TMA (n=5) and without (n=4, Sabulski et al 2023). Results The plasma proteome of TA-TMA patients was compared to patients without TA-TMA at each timepoint. Qiagen Ingenuity Pathway Analysis of pre-transplant differentially expressed proteins (DEPs) showed enrichment and predicted activation of acute phase response signaling (p=2.6e-6), pathogen induced cytokine storm (2.6e-5) and atherosclerosis signaling (p=1.2e-4), suggesting a pre-transplant heightened inflammatory state in TA-TMA patients.A total of 211 proteins were consistently different in TA-TMA patients at 4 or more timepoints and underwent Hallmark pathway analysis (Table 1). The top enriched pathways were the complement system (q=0.002), epithelial mesenchymal transition (EMT, q=0.002) and coagulation (q=0.001). Next, we longitudinally compared pathway enrichment in TA-TMA patients vs patients without (Fig.1). Interestingly, we observed a peak in EMT pathway enrichment coinciding with complement and coagulation pathways but a decline in angiogenesis pathway enrichment over time. This led us to hypothesize that the observed EMT signal may actually represent endothelial to mesenchymal transition (EndMT), which is not differentiated from EMT in Hallmark. To test this hypothesis, we re-analyzed differentially expressed genes (DEGs) from our in vitro TA-TMA model. Nine established EndMT markers were DEGs and the direction of gene expression changes was consistent with EndMT (Table 2). Conclusions We conclude that pre-transplant inflammation creates a toxic environment for endothelial cells and, in conjunction with TA-TMA, leads to EndMT. EndMT occurs in other inflammatory vascular disorders and results in vascular dysfunction, fibrosis and a chronic pro-inflammatory state. This novel observation suggests TA-TMA patients may be more susceptible to long term vascular complications and therapies that prevent EndMT may have an unrecognized role in TA-TMA.

Introduction Transplant-associated thrombotic microangiopathy (TA-TMA) is a life-threatening complication of hematopoietic stem cell transplantation (HSCT), driven by endothelial injury and complement activation. Eculizumab, a terminal complement C5 inhibitor, improves survival in high-risk TA-TMA, yet one-third of patients fail to respond. Objective To determine mechanisms of eculizumab non-response and early predictors of outcome in TA-TMA that are critical for improving therapy. Methods We studied 90 children and young adults with high-risk TA-TMA, comparing eculizumab responders (n=60) and non-responders (n=30). All HSCT recipients were prospectively screened for TA-TMA. TA-TMA risk group and response to eculizumab was determined based on published criteria (Fig1). All patients with high-risk TA-TMA received PK/PD-guided eculizumab dosing. We performed novel PK/PD modeling to determine complete complement blockade (“equilibrium”). Complement blockade, PK/PD dynamics, biomarker trajectories for sC5b-9, CXCL9, ST2, proteinuria by random urine protein/creatine ratio (rUPCR), and outcomes were compared in responders and non-responders. Results Non-responders displayed greater systemic organ injury and worse survival despite optimized eculizumab dosing and PK/PD monitoring (Fig 2A). PK/PD modeling confirmed that complement blockade can be achieved with intensified dosing, but steady blockade state (equilibrium) requires approximately six weeks of therapy, underscoring that short-term biomarker normalization is not sufficient to declare response (Fig 2B). Importantly, non-responders exhibited persistently elevated rUPCR and rising ST2 despite suppression of sC5b-9, indicating ongoing complement-independent endothelial injury. Longitudinal biomarker profiling showed that non-responders had progressive CXCL9 elevation despite sC5b-9 control, and that adding IFNγ blockade with emapalumab improved eculizumab PK/PD, accelerated complement suppression, and enhanced survival, underscoring the pathogenic role of the complement-interferon loop (Fig3). We showed that elevated sC5b-9 is associated with TA-TMA risk, higher therapy needs and poor outcomes, whereas comprehensive profiling of other complement components did not differentiate responders, and normal C3/C4 values did not exclude TA-TMA highlighting the need for functional biomarkers. Conclusion These findings reveal that eculizumab non-response reflects complement independent persistent endothelial injury. Early monitoring of ST2, CXCL9, rUPCR, may identify patients at risk for suboptimal eculizumab response and guide precision therapies that can be integrated with complement blockade. Such strategies will be essential to improve survival and prevent irreversible vascular injury in high-risk TA-TMA.

Chronic graft-versus-host disease (cGVHD) affects up to 54% of allogeneic hematopoietic cell transplantation recipients and is the leading cause of late non-relapse morbidity and mortality. Despite the availability of National Institutes of Health consensus criteria, Foundation for the Accreditation of Cellular Therapy accreditation standards, and CIBMTR reporting requirements, standardized screening practices remain inconsistent. The Engraft Learning Health Network (LHN) conducted a multicenter assessment to characterize cGVHD screening practices across participating transplant centers. To characterize current cGVHD screening practices across participating transplant centers within the Engraft LHN, identify areas of practice variation, and inform development of a standardized screening approach. This multicenter qualitative study included 10 transplant centers (8 pediatric and 2 adult) within the Engraft LHN. Thirty-three structured interviews were conducted with clinician representatives using a standardized guide addressing eight screening domains: assessment schedules, documentation practices, and organ-specific screening of the skin, oral cavity, ocular, pulmonary, gastrointestinal, genitourinary, and musculoskeletal systems. Interview summaries were validated by participants and analyzed to quantify screening practices across centers using descriptive statistics and heat map visualization. Marked heterogeneity was observed across all screening domains, with no uniform approach identified across centers. Assessment schedules ranged from weekly to as-needed evaluations, with only 52% of respondents reporting monthly evaluations in the first 6 mo following day 100 post-transplant and 37% continuing quarterly assessments beyond 1 yr. Only 10% incorporated the Lee Symptom Scale into routine documentation. Skin was universally included in the review of systems, but comprehensive examination (12%) and structured scoring (21%) were infrequent. Pulmonary function test monitoring showed significant gaps at intermediate time points, with most centers performing PFTs at annual milestones but fewer than 25% reporting assessment at quarterly intervals recommended by current guidelines. Genitourinary screening demonstrated the greatest inconsistency, with 42% and 63% of respondents not routinely discussing symptoms with female and male patients, respectively. Rare manifestations including neuropathy and serositis were screened for by fewer than 10% of centers. cGVHD screening practices vary widely across transplant centers in timing, documentation, organ-specific assessment, and utilization of standardized tools. Existing assessment frameworks, while clinically detailed, pose practical challenges for routine use outside specialized GVHD clinics. These findings highlight the need for a streamlined, evidence-informed screening bundle with defined assessment intervals, guideline-recommended PFT surveillance, and integration of patient-reported outcome measures. The Engraft LHN is leveraging these findings to develop and implement such a bundle at the provider, patient, and center levels using rapid-cycle quality improvement methodology.

Background Central venous catheters (CVCs) are almost universally needed during transplant and cellular therapy (TCT) but are associated with complications including bloodstream infections (BSIs) and thrombosis. These complications lead to reduced quality of life and increased healthcare costs, adding burdens to patients and caregivers. Striking the right balance in keeping a CVC in place for the appropriate length of time is essential. The Engraft Learning Network is a unique quality improvement (QI) collaboration between 13 adult and pediatric transplant and cellular therapy (TCT) centers dedicated to QI and improving the lives of patients and their caregivers. Methods Participating centers co-designed the learning network’s inaugural initiative between January and July 2024 with the goal of promoting timely CVC removal. A key driver diagram identified interventions including standardized monthly assessments of line necessity for all post TCT patients, consistent post-transplant CVC evaluations, documentation of center-specific practices, and structured cross-center learning as means to reduce line days. The outcome measure was the number of CVC days (numerator) per 100 patient days (denominator) within the first 365 days post-transplant. Exclusion criteria included relapse, transfer, or death. Baseline data of allogenic TCT were displayed beginning January 2022 to emphasize pre-intervention stability and highlight the special-cause variation observed beginning in 2024. Statistical process control (SPC) charts were used to evaluate trends and, control limits, and variance over time. Results Across the Engraft Network, approximately 5,000–6,000 CVC days are recorded each month. Since project initiation, the mean rate has decreased from 68 to 58 CVC days per 100 patient days—a 15–20% improvement. This corresponds to an estimated 750–1,000 fewer CVC days each month and 9,000–12,000 fewer annually. Post-intervention SPC charts show a clear downward shift in the mean with narrower control limits, reflecting reduced variability and greater process reliability across centers (Figure 1). The most compelling special-cause variation emerged beginning in March 2025. As centers started to discuss and implement the assessment tool, there was a downward trend of CVC days seen and this trend continued after fully implementing the standardized monthly assessments. Discussion This multi-center QI initiative demonstrates that structured, shared learning can lead to meaningful and sustained reductions in CVC retention after TCT. The decline in both mean CVC days and data variability underscores a transition toward more standardized practice patterns and safer, more consistent, and financially aware post-transplant care that will improve the quality of life for patients and caregivers.

Background BK polyomavirus (BKPyV) reactivation is nearly universal after pediatric hematopoietic stem cell transplant (HSCT); yet, only a quarter of patients develop BKPyV cystitis. We hypothesize that BKPyV reactivation alone is not sufficient for cystitis and that differential host immune responses to BKPyV are major determinants of disease. Objectives We assessed differences in cellular and soluble host factors that contribute to BKPyV cystitis after HSCT. Methods Plasma from 25 HSCT recipients underwent proteomics analysis (Somascan 7k platform) at 6 timepoints: baseline (pre-HSCT), day 7, 14, 30, 60 and 90. Paired immunoglobulin-like type 2 receptor alpha (PILRA) genotyping was performed on 112 consecutive patients at baseline. Single cell RNAseq was performed on day 60 peripheral blood mononuclear cells (PBMCs) from 12 HSCT recipients. Results We compared the plasma proteomes of patients with BKPyV viremia and cystitis (n=5) versus no cystitis (n=6). PILRA was a leading differentially expressed protein (DEP) at baseline (p=2.64e-2), day 7 (p=4.28e-2), day 14 (p=3e-3), day 30 (p=8.04e-3) and day 60 (p=5.09-3, Fig.1A). PILRA genotyping tested whether a known functional polymorphism (rs1859788) affects cystitis risk (Fig. 1B). Baseline heterozygotes (A/G, 12.5% cystitis rate) were significantly less likely to develop cystitis compared to homozygous (G/G, 39% cystitis rate; or A/A, 31% cystitis rate) patients (p=0.004).PBMC transcriptomes were compared between patients with BKPyV viremia and cystitis (n=3) versus BKPyV viremia and no cystitis (n=3). PILRA is predominantly expressed on monocytes and was a differentially expressed gene (DEG) in both CD16+ (q=4.6e-6) and CD14+ (q=2.6e-57) monocytes (Fig.2B). Pathway and upstream regulator analysis of CD14+ and CD16+ monocytes showed increased antiviral response (p=4.6e-19 and p=7.8e-23), interferon gamma (p=4e-17 and p=2.7e-12), and interferon alpha/beta (p=3.5e-15 and p=2.1e-11) enrichment in cystitis patients (Fig.2A). The interferon stimulated genes, interferon induced protein 44-like (IFI44L) and interferon induced protein 44 (IFI44), were leading DEGs in 9/11 cell types studied (Fig. 2C). Conclusions PILRA is a negative regulator of inflammatory processes and exists both in a membrane bound and soluble form. Baseline PILRA genotype was associated with cystitis risk in this study and the significantly lower soluble PILRA levels coupled with increased cellular PILRA expression support an increase in cell surface PILRA in cystitis patients. We conclude that soluble and monocytic PILRA differences contribute to a heightened interferon response to BKPyV that may cause inflammation and cystitis through interferon stimulated genes such as IFI44L and IFI44. Ongoing work aims to identify the mechanisms of PILRA-associated interferon production, as well as multiomic data analysis from BKPyV negative patients.