Choi, Sung Won

ACCESS is a prospective, multicenter phase II trial evaluating peripheral blood stem cell (PBSC) transplantation using 4-7/8 HLA-mismatched unrelated donors (MMUD) with post-transplant cyclophosphamide (PTCy), tacrolimus, and mycophenolate. Adults undergoing first HCT with myeloablative (MAC) or reduced-intensity/nonmyeloablative (RIC/NMA) conditioning were enrolled, with donors aged 18-35 years and matched at 4/8-7/8 HLA loci by high-resolution typing. This expanded analysis includes all enrolled adult PBSC recipients, including the protocol-specified RIC/NMA cohort expansion, to improve precision of outcome estimates and provide descriptive analyses by HLA match level. Among 268 adults, 183 received 7/8 and 85 received <7/8 MMUD grafts; 82.4% of the <7/8 cohort received 6/8 grafts. Median follow-up among survivors was 11.9 months. At 1-year, overall survival was 78.6% (95% CI, 71.9-83.9) in the 7/8 cohort and 85.6% (95% CI, 76.0-91.5) in the <7/8 cohort. One-year relapse, non-relapse mortality, and GVHD-free relapse-free survival were 17.1%, 13.7%, and 51.1% versus 22.8%, 8.4%, and 54.6%, respectively. The cumulative incidence of grade II-IV and grade III-IV acute GVHD by day 100 was 36.6% and 7.7% in the 7/8 cohort versus 28.3% and 5.9% in the <7/8 cohort. Moderate-to-severe chronic GVHD at 1 year was 11.3% versus 7.7%, respectively. Among adult ACCESS recipients treated on a uniform young-donor PBSC/PTCy platform, short-term outcomes with predominantly 6/8 MMUD transplantation were encouraging. Because donor match level was not prospectively assigned and between-group comparisons were exploratory, these findings are descriptive and should not be interpreted as establishing equivalence with 7/8 MMUD transplantation. (Registered as NCT04904588 at CT.gov)

Abstract Introduction Distinct temperature patterns have long been recognized to correlate with fevers of differing etiologies. While the use of wearable sensors for high-frequency temperature monitoring (HFTM) on a near minute-by-minute basis has been shown to detect fevers earlier than standard-of-care nursing vital sign assessments in hospitalized patients, leveraging these high-resolution datasets to computationally identify unique digital signatures for real-time diagnosis of underlying fever etiology has not been widely explored. Diagnostic uncertainty is common in patients undergoing hematopoietic stem cell transplantation (HCT), with only 20–30% of febrile neutropenic episodes being microbiologically documented. We hypothesized that unique temperature patterns extracted from HFTM data collected during episodes of febrile neutropenia could be used to develop a supervised machine learning classifier capable of accurately predicting underlying fever etiology in HCT patients. Methods We analyzed 68 clinically independent fever episodes recorded in HCT patients (n=90) outfitted with an FDA-cleared wireless temperature sensor (TempTraq®, BlueSpark Technologies) that measured axillary temperature every 2 minutes throughout hospitalization. Time-series features were extracted from temperature traces spanning 1 hour before to 3 hours after fever onset and used to train a suite of machine-learning models to distinguish engraftment fevers from other fever etiologies. Model training and evaluation were performed using repeated stratified 5-fold patient-level cross-validation, yielding 100 train-test evaluations. Results Among all classification models, the logistic regression classifier provided the best overall performance and interpretability, achieving 94% specificity (95% CI, 0.84–1.0) for identifying engraftment fevers with a mean AUROC of 0.88 ± 0.10. Feature importance analysis demonstrated that both clinical variables and HFTM-derived temperature dynamics contributed to model performance, with a strong reliance on time-series features captured within the first 4 hours of fever onset. Conclusion Our study provides a demonstration that continuous temperature data collected from patients outfitted with wearable sensors can be leveraged not only for early fever detection but also for machine learning–based diagnosis of fever etiology. These findings suggest that dynamic temperature patterns contain clinically meaningful physiologic information that with further studies could support real-time diagnostic decision-making and guide safe de-escalation of empiric antibiotics during febrile neutropenia in patients undergoing intensive cancer therapy.

Abstract Allogeneic hematopoietic cell transplantation (allo-HCT) is a potentially curative treatment for many hematological malignancies, but graft-versus-host disease (GVHD) is a common complication. Low gut microbiome diversity is associated with higher GVHD risk and shorter survival in multiple studies. Recently, the BMT CTN 1703 clinical trial demonstrated superiority of a GVHD-prophylaxis regimen including post-transplant cyclophosphamide (PTCy) compared to the standard prophylaxis (tacrolimus and methotrexate, Tac/MTX) in terms of GVHD-free, relapse-free survival at one year among reduced intensity conditioning allo-HCT recipients. However, the effect of PTCy on the gut microbiome and its association with clinical outcome have not been described. Here, we report on a companion randomized clinical controlled trial (BMT CTN 1801), which collected 2575 longitudinal stool samples from 304 study participants. Samples were obtained up to weekly up to day 84 post allo-HCT and at less frequent intervals thereafter, up to 2 years. Microbiome diversity and absolute microbial load were lower in the PTCy group compared to the Tac/MTX group on days 14–28 post-HCT. However, diversity at the timepoint closest to neutrophil engraftment was not significantly associated with non-relapse mortality after one year or other clinical outcomes, contrary to expectations from previous studies. Microbial domination events, when a single species exceeds 30% relative abundance, were comparable across treatment arms and reflected both pathogen blooms as well as less severe disruptions of the microbial community. Clostridium scindens and secondary bile acid metabolism pathways were less prevalent in the PTCy arm than in the Tac/MTX arm post-HCT, yet presence of secondary bile acid metabolism pathways was associated with a lower risk of chronic GVHD. Given that PTCy was associated with a greater disruption of the microbiome as measured by diversity, absolute microbial abundance, and bile acid metabolism capability, but better clinical outcomes overall, these data suggest that the importance of the microbiome in modulating the host immune systems after allo-HCT is specific to different types of GVHD prophylaxis.

592 Background: While not all fevers occurring during hematopoietic stem cell transplantation (HCT) are dangerous, episodes of febrile neutropenia tend to be managed aggressively given the significant mortality and morbidity associated with infectious complications of transplantation. This can potentially lead to unnecessary diagnostic testing, antibiotic treatment, and prolonged hospital stays for non-infectious fever etiologies such as engraftment. Our group has previously demonstrated that high-frequency body temperature monitoring (HFTM) using a wearable device detects fevers earlier in transplant recipients (Flora et al, Cancer Cell, 2021). We hypothesized that applying supervised machine learning to HFTM data collected from inpatient allogeneic HCT patients would allow for accurately distinguishing between non-infectious and infectious etiologies of fever and could potentially guide safe de-escalation of antibiotics. Methods: We analyzed 74 independent fever events detected in HCT patients (n=42) outfitted with an FDA-cleared, wireless sensor measuring axillary temperatures every 2 minutes for the duration of their respective hospitalizations. After data filtering, independent fever events were then retrospectively grouped by fever etiology using physician-determined annotation as previously described (Flora et al., 2021). A split ratio of 80:20 was used to create training and test sets from the parent data set, consisting of 59 and 15 fever events for each split, respectively. Training sets were used to create a classification algorithm identifying bacterial infections requiring antibiotics using the machine learning model XGBoost. The performance of this model was then evaluated using test sets of the remaining fevers (n=15). Results: Averaging over 1000 test sets of 15 independent fevers comprising 5 bacterial infection-related fevers and 10 non-infectious fevers, our model correctly identified 82% (mean recall) of cases requiring antibiotics with a mean precision of 0.41. The model also correctly identified 42% of non-infectious fevers (mean true negative rate, 0.42) that would otherwise have been treated with unnecessary antibiotics. The mean AUCPR was 0.58, with the best performing individual model achieving an AUCPR of 0.97. Conclusions: The preliminary results of this study demonstrate that our classification model shows promise for differentiating fevers requiring antibiotics due to bacterial infections from other non-infectious fever etiologies. Prospective studies further developing supervised machine learning models for subcategorization of febrile neutropenia could pave the way for safely de-escalating antibiotics in HCT and other immunocompromised patients undergoing cancer treatment monitored using HFTM.

Background: BMT CTN 1703 was a Phase 3 trial comparing Tac/MTX (n = 217) to post-transplant cyclophosphamide (PT-Cy)/Tac/MMF (n = 214) GVHD prophylaxis after reduced intensity conditioning unrelated donor hematopoietic cell transplant (HCT). PT-Cy significantly improved GVHD-free Relapse-Free Survival (Bolanos-Meade et al NEJM 2023). However overall survival was not significantly different and the # of Grade 2 infections was higher with PT-Cy. To investigate its biologic underpinnings 1703 was linked to a mechanistic study, BMT CTN 1801, which co-enrolled 324 pts (159 CNI/MTX 165 PT-Cy). The resulting dataset enabled an analysis of post-HCT T cell reconstitution at an unprecedented level of detail. Methods: Samples were collected from the HCT infusion and on D+7, 14, 28, 63, 98, 180, 270, 1- and 2-yr. We extracted DNA (2,225 samples) and performed Adaptive Biotech β-strand T Cell Receptor (TCR) Immunosequencing (TCR-Seq). We profiled 44,077 (median) TCRs/sample resulting in 215,155,719 T cells analyzed. We evaluated multiple TCR diversity measures longitudinally and before/after cGVHD and infections. Each measure evaluates a specific aspect of TCR diversity, including Simpsons Clonality (focuses on the most expanded clones) Diversity Slope (mid-range clones) Richness (both mid-range and less-expanded clones) and Singletons (the TCR-Seq equivalent of naïve T cells). This analysis was robust to the full range (1000-50,000) of down-sampled TCRs sequenced. Results: We identified major differences in TCR reconstitution in Tac/MTX vs PT-Cy that begin early and persist through 2 yr post-HCT. These resulted in a significantly less diverse TCR repertoire with PT-Cy with substantial implications for its efficacy and toxicity. The most salient differences are: (1) PT-Cy caused significant in-vivo T cell depletion early post-HCT (% T cells/total WBC at D+28: 8.2% (Tac/MTX) vs 2.2% (PT-Cy) p=2.2x10 -23). (2) The remaining donor T cells in PT-Cy undergo substantial clonal expansion ( Fig 1) resulting in significantly constrained TCR diversity measured by Clonality, Slope, Richness, Singletons (not shown) and Singletons/Richness ( Fig 2) which began early (D+14) and persisted for 2 yr. (3) Linked T cell depletion and clonal expansion resulted in a TCR diversity bottleneck after PT-Cy ( Fig 1). Due to the bottleneck, while the fraction of T cells originating from the transplant infusion at =/> 1 yr was similar between arms (18% vs 24%, p=0.7) there were significantly fewer Singleton TCRs in PT-Cy, suggesting a reduction in naïve T cells ( Fig 2), an outcome that has critical consequences for both cGVHD and infection. Thus while cGVHD onset with Tac/MTX was linked to a significant drop in TCR diversity due to singleton clonal expansion (median log-2 fold change in singletons with Tac/MTX = -0.25 p = 0.01) this drop was not observed with PT-Cy (log-2 fold change = -0.05, p = 0.1). Along with the significantly decreased # singletons at all time-points in PT-Cy ( Fig 2) this suggests that while cGVHD after Tac/MTX is linked to substantial naïve T cell expansion, the same is not true with PT-Cy, identifying a novel mechanism for cGVHD control with PT-Cy. The substantially fewer singletons in PT-Cy vs Tac/MTX may also increase the risk of Grade 2+ infections. T cell responses to new infections rely on antigen recognition/effector maturation of naïve (singleton TCR) T cells. After infection, the TCR repertoire became less diverse for both Tac/MTX and PT-Cy (log-2-fold singleton change -0.17 and -1.2 p = 0.0006 and 0.01) indicating that singletons in both cohorts were capable of expanding to pathogens. However the markedly decreased # singletons in PT-Cy at all timepoints was associated with significantly fewer singletons prior to infection (mean 3979 at 5K read-depth for Tac/MTX vs 2820 for PT-Cy p<0.001). This singleton deficit would place PT-Cy patients at higher risk of missing the critical mass of TCRs required to respond effectively to infections, consistent with their increased rate of Gr 2 infections. Conclusions: We performed longitudinal TCR-Seq on >200 million T cells from 324 HCT patients, representing the largest post-HCT TCR-Seq analysis in the history of the field, and enabling novel insights into TCR dynamics. These data identify a previously unappreciated global TCR diversity bottleneck with PT-Cy that simultaneously protects patients from cGVHD and places them at higher risk of infectious complications.

Abstract Abstract 739 The combination of a calcineurin inhibitor and methotrexate has been the standard of care in graft-vs.-host disease (GVHD) prophylaxis for over 25 years, with resultant rates of grade II-IV acute GVHD between 30–50%. The mTOR inhibitor, sirolimus, has demonstrated promise in a number of Phase II trials as an immunosuppressant used for GVHD prophylaxis. The BMT CTN, sponsored by the NHLBI and NCI, conducted a multicenter, randomized controlled trial comparing the combination of tacrolimus and sirolimus (Tac/Sir) with tacrolimus and methotrexate (Tac/Mtx) as GVHD prophylaxis after matched, related donor (MRD) hematopoietic stem cell transplantation (HSCT). Methods: Eligible patients were between ages 2 – 60 years, and had acute leukemia in remission, myelodysplasia or chronic myeloid leukemia in chronic or accelerated phase. All had adequate organ function, and a 6/6 HLA-A, B, DRB1 matched sibling donor. 304 patients were randomly assigned to either Tac/Sir (n = 151) or Tac/Mtx (n = 153) as GVHD prophylaxis after TBI-based conditioning and MRD HSCT. An intent-to-treat analysis was performed on the primary endpoint of Grade II-IV GVHD-free survival 114 days from randomization. Ten subjects who received busulfan-based conditioning and were previously reported were excluded from analysis. Three subjects who did not undergo HSCT are included in the primary analysis, but not secondary analyses. Results: Treatment groups were well balanced. The median age of participants was 44 years (range 13 – 59) and 83% had acute leukemia. Neutrophil and platelet engraftment were both faster in the Tac/Sir group (14 vs. 16 days, p < 0.001; 16 vs. 19 days, p = 0.03, respectively), but this did not affect the time to first hospital discharge (20 vs. 21 days, p = 0.37). The incidence of grade II-IV and grade III-IV acute GVHD at 100 days were lower in the Tac/Sir group (26 vs. 34%, p = 0.17; 8 vs. 15%, p = 0.05). Day 100 treatment-related mortality was no different between groups (7 vs. 7%, p = 0.43). The primary endpoint of 114-day acute GVHD-free survival was not statistically different between groups (67 vs. 62%, p = 0.38, Figure). The cumulative incidence of relapse at 2 years from transplantation was not different between groups (27 vs. 30%, p = 0.81). The competing-risk cumulative incidence of chronic GVHD was higher in the Tac/Sir arm (54 vs. 43%, p =0.044). Overall toxicities were not different between groups, with two notable exceptions. The peak and average OMAS oral mucositis scores were lower in the Tac/Sir arm (peak 0.70 vs. 0.96, p < 0.001; average 0.31 vs. 0.47, p < 0.001), however, there was an increased rate of the endothelial injury syndromes, veno-occlusive disease (11 vs. 4%, p = 0.03), and thrombotic microangiopathy (5 vs. 1%, p = 0.05) in the Tac/Sir arm. Causes of death were not different between groups. At 2 years from transplantation, disease-free (DFS) and overall survival (OS) were not different between study arms (DFS 53 vs. 53%, p = 0.76; OS 60 vs. 61%, p = 0.44). Conclusions: No difference in 114-day acute GVHD-free survival was noted between treatment arms. Compared with Tac/Mtx in MRD HSCT, Tac/Sir is associated with more rapid engraftment, less severe acute GVHD and oral mucositis, excess chronic GVHD and endothelial injury syndromes, and similar long-term outcomes. Understanding the trade-offs between regimens, Tac/Sir can be used as an alternative to Tac/Mtx in MRD HSCT. Disclosures: Cutler: Pfizer, inc: Research Funding; Astellas, Inc: Consultancy, Research Funding. Off Label Use: Sirolimus - Prevention of GVHD Tacrolimus - Prevention of GVHD. Waller:Outsuka: Research Funding.

Key Points Tac/Sir prophylaxis provides equivalent GVHD-free survival when compared with Tac/Mtx in MRD transplantation. Tac/Sir is associated with more rapid engraftment and reduced oropharyngeal mucositis after MRD transplantation.