Machine learning–based scoring models to predict hematopoietic stem cell mobilization in allogeneic donors

Abstract

Mobilized peripheral blood has become the primary source of hematopoietic stem cells for both autologous and allogeneic stem cell transplantation. Granulocyte Colony-Stimulating Factor (G-CSF) is currently the standard agent used in the allogeneic setting. Despite the high mobilization efficacy in most donors, G-CSF requires 4-5 days of daily administration, and a small percentage of the donors fail to mobilize an optimal number of stem cells necessary for a safe allogeneic stem cell transplant. In this study, we retrospectively reviewed 1361 related allogeneic donors who underwent stem cell mobilization at Washington University. We compared the standard mobilization agent G-CSF with five alternative mobilization regimens, including GM-CSF, G-CSF+GM-CSF, GM-CSF + Plerixafor, Plerixafor and BL-8040. Cytokine-based mobilization strategies (G-CSF or in combination with GM-CSF) induce higher CD34 cell yield after 4-5 consecutive days of treatment, while CXCR4 antagonists (plerixafor and BL-8040) induce significantly less but rapid mobilization on the same day. Next, using a large dataset containing the demographic and baseline laboratory data from G-CSF-mobilized donors, we established machine learning (ML)-based scoring models that can be used to predict patients who may have less than optimal stem cell yields after a single leukapheresis session. To our knowledge, this is the first prediction model at the early donor screening stage, which may help identify allogeneic stem cell donors who may benefit from alternative approaches to enhance stem cell yields thus insuring safe and effective stem cell transplantation.

Accurate Machine Learning Model to Diagnose Chronic Autoimmune Diseases Utilizing Information From B Cells and Monocytes

“…Based on our finding of powerful information that characterizes diseases, we tried to establish a machine learning model to distinguish chronic autoimmune diseases. Several reports have proven that the random forest (RF) machine learning method would give a high accuracy in disease classification when abundant features were included ( 71 , 72 ), and another reason for the random forest model was its interpretability—each gene contribution in the RF machine learning model was visible. Our area under curve (AUC) score for SLE indicates that our machine learning model has the potential to become an efficient tool for accurate diagnosis of SLE at the single-cell RNA level.…”

Updates on the mobilization pipeline for hematopoietic stem cell collection

“…A large retrospective study of donors mobilized with standard G‐CSF (n = 1025) and various other mobilization strategies available as Phase I/II trials (GM‐CSF, n = 40; G‐CSF + GM‐CSF, n = 167; GM‐CSF + plerixafor; plerixafor, n = 88; and BL‐8040, n = 18) was conducted by Xiang et al 27 The authors found that among allogeneic donors, cytokine‐based mobilization strategies (G‐CSF or in combination with GM‐CSF) induce higher CD34 cell yield after 4‐5 consecutive days of treatment, while CXCR4 antagonists (plerixafor and BL‐8040) induce significantly less but more rapid, same‐day mobilization 27 …”

“…15 MGTA-145 Single arm trial; Grade 1 pain (bone pain) most common, seen in 11/25 (44%) with 9/25 (38%) experiencing acute-onset bone pain with MGTA-145 (duration 7 min, range 3-28). 35 At last, follow up, 18/25 completed transplant with MGTA-145 mobilized graft with receipt of a median (range) of 3.5 (2.2-8.1) Â 10 6 CD34/kg and median (range) days to neutrophil engraftment 15 (11)(12)(13)(14)(15) and platelet engraftment 17.5 (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33). 35 mobilization.…”

“…According to the National Institutes of Health, it is estimated that fewer than 1000 people in the United States have this WHIM 26 in contrast to the over 20 000 annual transplants in the United States, many of which may benefit from additional mobilization optionsparticularly an oral agent that might save patients from injections! A large retrospective study of donors mobilized with standard G-CSF (n = 1025) and various other mobilization strategies available as Phase I/II trials (GM-CSF, n = 40; G-CSF + GM-CSF, n = 167; GM-CSF + plerixafor; plerixafor, n = 88; and BL-8040, n = 18) was conducted by Xiang et al 27 The authors found that among allogeneic donors, cytokine-based mobilization strategies (G-CSF or in combination with GM-CSF) induce higher CD34 cell yield after 4-5 consecutive days of treatment, while CXCR4 antagonists (plerixafor and BL-8040) induce significantly less but more rapid, same-day mobilization. 27…”

Future stem cell analysis: progress and challenges towards state-of-the art approaches in automated cells analysis

“…Both supervised ML approaches have achieved 89% and 97% accuracies, respectively, for the hPSCs classification system. Recently, the ML approach has also been implemented to predict allogeneic stem cell donors from the granulocyte colony-stimulating factor (G-CSF) of HSC ( Xiang et al, 2022 ). The established prediction methods were performed by cross-validation (10-fold) with appropriate hyper-parameter using a grid search technique.…”