Standardization of platelet releasate products for clinical applications in cell therapy: a mathematical approach

Francesco Agostini, Jerry Polesel, Monica Battiston, Elisabetta Lombardi, Stefania Zanolin, Alessandro Da Ponte, Giuseppe Astori, Cristina Durante, Mario Mazzucato

Standardized animal-free components are required for manufacturing cell-based medicinal products. Human platelet concentrates are sources of growth factors for cell expansion but such products are characterized by undesired variability. Pooling together single-donor products improves consistency, but the minimal pool sample size was never determined. Supernatant rich in growth factors (SRGF) derived from n = 44 single-donor platelet-apheresis was obtained by CaCl2 addition. n = 10 growth factor concentrations were measured. The data matrix was analyzed by a novel statistical algorithm programmed to create 500 groups of random data from single-donor SRGF and to repeat this task increasing group statistical sample size from n = 2 to n = 20. Thereafter, in created groups (n = 9500), the software calculated means for each growth factor and, matching groups with the same sample size, the software retrieved the percent coefficient of variation (CV) between calculated means. A 20% CV was defined as threshold. For validation, we assessed the CV of concentrations measured in n = 10 pools manufactured according to algorithm results. Finally, we compared growth rate and differentiation potential of adipose-derived stromal/stem cells (ASC) expanded by separate SRGF pools. Growth factor concentrations in single-donor SRGF were characterized by high variability (mean (pg/ml)-CV); VEGF: 950-81.4; FGF-b: 27-74.6; PDGF-AA: 7883-28.8; PDGF-AB: 107834-32.5; PDGF-BB: 11142-48.4; Endostatin: 305034-16.2; Angiostatin: 197284-32.9; TGF-β1: 68382-53.7; IGF-I: 70876-38.3; EGF: 2411-30.2). In silico performed analysis suggested that pooling n = 16 single-donor SRGF reduced CV below 20%. Concentrations measured in 10 pools of n = 16 single SRGF were not different from mean values measured in single SRGF, but the CV was reduced to or below the threshold. Separate SRGF pools failed to differently affect ASC growth rate (slope pool A = 0.6; R(2) = 0.99; slope pool B = 0.7; R(2) 0.99) or differentiation potential. Results deriving from our algorithm and from validation utilizing real SRGF pools demonstrated that pooling n = 16 single-donor SRGF products can ameliorate variability of final growth factor concentrations. Different pools of n = 16 single donor SRGF displayed consitent capability to modulate growth and differentiation potential of expanded ASC. Increasing the pool size should not further improve product composition.