Intra-apheresis CD34+ cell count: A dynamic approach to predicting peripheral blood stem cell collection yield

Accurate prediction of peripheral blood stem cell (PBSC) yield is essential for optimizing hematopoietic stem cell collections. Traditional CD34+ prediction formulas often overlook the ongoing recruitment of CD34+ cells during the collection procedure. This study evaluated whether a single intra-apheresis CD34+ measurement could improve yield prediction compared to existing prediction models.

We retrospectively analyzed 150 PBSC collections from 123 autologous and allogenic donors. A single intraprocedural CD34⁺ measurement and final collection yield were obtained, and univariable and multivariable linear regressions were developed to assess predictive accuracy. Model stability and generalizability were tested via bootstrap resampling and cross-validation. Subgroup analyses by age, sex, weight, diagnosis, and sampling time were performed to confirm robustness. We also compared our results against published pre-apheresis formulas.

Intra-apheresis CD34⁺ counts showed a strong correlation with the final yield (r = 0.97; R² = 0.95). Subgroup analyses across donor demographics and clinical factors yielded similarly high R² values (0.91-0.99). Multivariable analysis identified intra-apheresis CD34+ measurement as the primary contributor (β = 0.67, p <.0001). Later sampling (≥1.20 of total blood volumes processed) reduced error metrics relative to earlier draws. Compared to pre-apheresis models, the intra-apheresis approach achieved higher R² and lower error metrics.

A single intra-apheresis CD34+ measurement, particularly at or beyond 1.20 processed blood volumes, offers a reliable, real-time predictor of final PBSC yield. This intraprocedural approach outperforms published pre-apheresis formulas, enabling more efficient collection and reduced donor burden across diverse clinical scenarios.