OA2‐AM23‐TU‐03 | Development of a Predictive Algorithm for Optimizing Apheresis Duration for Hematopoietic Cell Collection
Year: 2023
Type: article
Abstract: CD34+ cell enumeration is necessary to determine if a hematopoietic progenitor cell (HPC) product has met the goal dose required for hematopoietic cell transplantation (HCT). To optimize collection times and avoid exceeding transplant dose goals, an algorithm was developed to predict apheresis CD34+ cell yields. While avoiding the collection of CD34+ cells that significantly exceeds the goal dose is desirable, the algorithm must also be sufficiently accurate to avoid under-collection of the goal dose. Optimization may reduce collection times, apheresis procedures, additional doses of granulocyte-colony stimulating factor (G-CSF), and strain to laboratory storage capacities. The algorithm was established utilizing minimal baseline data consisting of pre-collection CD34+ cell count, peripheral white blood cell count, and patient weight (kg). To determine the optimal duration of collection to meet the prescribed goal, the algorithm was run in intervals of 30 min. One hundred twenty-five collections (94 autologous; 32 allogeneic) were analyzed. Eighty-seven autologous collections were from patients with hematologic malignancies (multiple myeloma, n = 66; lymphoma, n = 15; other, n = 6) and seven were from patients with non-hematologic malignancies. Allogeneic collections were from healthy donors. Based on an analysis of the CD34+ cell data and apheresis duration, a linear regression model was developed to predict CD34+ cell collection rate and transformed to estimate time required to achieve 2 × 106 CD34+ cells/kg. The predicted duration of collection was compared in minutes using the predictive model to the current institutional practice. The institutional standard predictive calculation utilizes similar pre-collection results and assumes a collection efficiency ≥30%. A strong correlation was established between both predictive methods with the final observed value and predicted value of R2 = >.9. The correlation between the predictive model and the institutional practice was also strong (R2 = >.9). A strong correlation of CD34+ cell yield was proven between the experimental model and the final observed result. The experimental model was also compared to the institutional practice and showed a similar correlation. These two correlations enforced the validity of the experimental algorithm. This algorithm is advantageous as it uses minimal information that can be obtained prior to the collection, can be used on a real-time basis as it is calculated in minutes, and provides flexibility to the user when applying a desired interval.
Source: Transfusion
DOI: 10.1111/trf.10_17554
Authors: K. Su, R. Gill, S. Kim, D. Neuberg, P. Brunker, T. Spitzer
Institutions: Massachusetts General Hospital, Dana-Farber Cancer Institute
Language: English
Subfield: Hematology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4387572096