Source plasma collection in the United States: Toward a more personalized approach

REFERENCES 1. Monteferrario D, Bolar NA, Marneth AE, et al. A dominant-negative GFI1B mutation in the gray platelet syndrome.NEngl JMed. 2014;370(3):245-253. 2. Marneth AE, Van Heerde WL, Hebeda KM, et al. Platelet CD34 expression and α/δ-granule abnormalities in GFI1B-and RUNX1-related familial bleeding disorders. Blood. 2017;129(12):1733-1736. 3. Saes JL, Simons A, de Munnik SA, et al. Whole exome sequencing in the diagnostic workup of patients with a bleeding diathesis. Haemophilia. 2019;25(1):127-135. 4. Wang CQ, Chin DW, Chooi JY, et al. Cbfb deficiency results in differentiation blocks and stem/progenitor cell expansion in hematopoiesis. Leukemia. 2015;29(3):753-757. 5. Khan A, Hyde RK, Dutra A, Mohide P, Liu P. Core binding factor beta (CBFB) haploinsufficiency due to an interstitial deletion at 16q21q22 resulting in delayed cranial ossification, cleft palate, congenital heart anomalies, and feeding difficulties but favorable outcome. Am J Med Genet A. 2006;140(21):2349-2354. 6. van Oorschot R, Marneth AE, Bergevoet SM, et al. Inherited missense variants that affect GFI1B function do not necessarily cause bleeding diatheses. Haematologica. 2019;104(6):e260-e264. 7. Rabbolini DJ, Morel-Kopp MC, Chen Q, et al. Thrombocytopenia and CD34 expression is decoupled from alpha-granule deficiency with mutation of the first growth factor-independent 1B zinc finger. J Thromb Haemost. 2017;15(11):2245-2258. 8. Thambyrajah R, Patel R, Mazan M, et al. New insights into the regulation by RUNX1 and GFI1(s) proteins of the endothelial to hematopoietic transition generating primordial hematopoietic cells. Cell Cycle. 2016;15(16):2108-2114.

sufficient for source plasma, a goal that many European countries still aim to achieve. However, there currently is a critical shortage of plasmaderived medicines, particularly of intravenous immunoglobulin (IVIG) 1 in the U.S. and abroad that is impacting patients.
Concerns for donor safety focus on well-understood short-term effects, most importantly, vasovagal hypotensive events and, less frequently, citrate reactions. Longer-term potential complications, including iron or protein depletion and osteoporosis, are less wellcharacterized. Previous studies have demonstrated a temporary reduction in serum protein levels but have also shown a rebound effect, even with intensive donation schedules. 2 Regular monitoring of protein levels in serial plasma donors has been incorporated into federal regulations as a safety measure (21CFR630.15 and 21CFR640.65). The hypothesized risks of osteoporosis due to citrate effects and iron depletion have not been confirmed and long-term observational studies have established the safety of regular donations. 3 A nomogram regulating the volume of source plasma that can be extracted per donation serves as a key instrument to ensure donor safety and, in particular, to reduce the likelihood of hypotensive events. The U.S. Food and Drug Administration (FDA) issued its current plasmapheresis nomogram in 1992. 4 To maximize ease of use and to minimize operator error, the nomogram consists simply of three allowable collection volume categories based on donor weight.
For nearly 30 years, this nomogram has proven to be effective, resulting in a strong safety record for donor plasmapheresis. Reaction rates are very low, with fewer than 0.03% severe reactions. 5 However, while the 1992 nomogram fulfilled the objective of simplicity, it has limitations. It does not account for height or body mass index (BMI), nor for hematocrit levels. These factors are known to influence the total plasma volume (TPV) of a donor and would be valuable components of a more personalized approach to calculate target collection volumes. 6 Moreover, the current weight-based step-wise approach leads to abrupt target changes of up to 20% between groups. In summary, while the safety profile has been good across the donor population, there could be subgroups of donors at increased, yet currently unquantified risk.
Here, we report a systematic retrospective analysis of a large real-world data set of source plasma collections following current U.S. standards to better understand the implications and potential The TPV was calculated by first estimating each donor's BMI-adjusted blood volume per kilogram, and then factoring in the donor weight and hematocrit level. 7 The PVC was then compared to the TPV for each donation. Mean plasma volume collected was 760 mL, with highest plasma yields in the high-weight group (>175 lb) and the lowest yields in the low-weight group (110-149 pounds). A substantial majority (73%) of the donors were in the highest weight category.
An analysis of the amount of PVC as a percentage of the donor's TPV (PVC/TPV), when plotted against the TPV, showed a distribution along three bands, representing the three weight categories ( Figure 1A). This pattern is in accordance with the expected theoretical patterns. However, the distribution along the bands offered additional insights into the heterogeneity of these values in a real-world data set.
The PVC/TPV values ranged from 15%-42% between all individuals. The ratio of PVC to TPV was inversely proportional to weight, Second, the efficiency of the source plasma collection system seems sub-optimal, wherein some donors are only permitted to contribute a relatively small percentage of their total plasma volume. This is particularly critical given the recent supply issues and the fact that the U.S. is the major contributor to the global supply of plasma-derived medicines.
Limitations of our analysis include the lack of clinical outcomes data, such as donor adverse events. Future studies using actual collection volumes should include analyses of donor outcomes. Our data was obtained from 86 plasma collection centers by one commercial plasma collector (Octapharma Plasma). The time period was randomly selected and the 86 centers are representative of the overall donor population of this plasma collector. However, it is worth considering that there may be slight differences between the populations of different plasma collectors and that there are seasonal effects on plasma collection frequencies and adverse event rates, although neither of these would influence target volumes.
The 1992 nomogram has served plasma donors well and kept them collectively safe; the event rates for moderate and severe hypotensive events are very low. However, the use of an exclusively weight-based approach along three weight categories has led to a skewed practice, where a disproportionately high volume of plasma is collected relative to available total plasma volume in smaller donors.
Conversely, an inefficiently low volume of source plasma is collected, relative to total plasma volume, in larger persons.
Further studies are needed to better understand the risks associated with high PVC/TPV values, and to explore the benefits of a more personalized approach. This might include the creation of a continu-

(C)
F I G U R E 1 A, Distribution of the percentage of the donor's plasma volume collected in the plasma product as a function of the donor's total plasma volume (TPV). Three discrete bands are described by the three different nomogram weight groups. B, Distribution of donors and the percent contribution of plasma across the total plasma volume. C, Plasma volume collected as a percentage of total plasma volume for adjusted weight categories White matter volume changes in adult beta-thalassemia: Negligible and unrelated to anemia and cognitive performances To the Editor: We read with great interest the study by Choi on brain involvement in anemic patients, showing white matter volume reduction according to anemia and cognitive impairment severity. 1 The findings that they reported with a quantitative analysis and showing the interplay among anemia, white matter volume changes and cognition, were both fascinating and reasonable. However, the miscellaneous recruitment of anemic patients did not help understand whether the phenomenon of white matter volume decrease was effectively a biomarker of any type of anemia. Indeed, white matter involvement has been widely shown in sickle cell disease, 2 while brain involvement is still uncertain or not fully studied in other forms of anemia. For example, recent studies with control groups have failed to detect significantly increased vascular-like white matter changes in beta-thalassemia patients. 3,4 Even very sensitive advanced MRI analyses (ie, magnetization transfer and diffusion tensor based techniques) failed to reveal structural white matter changes in beta thalassemia patients. 5 On these pre-  Quotient scores, that were significantly worse in beta thalassemia patients (see Tartaglione et al. 6 ).
Considering the striking contrast between the present study and the study by Choi, we also applied to our sample a tensorbased morphometry analysis (TBM; see Supplementary material), i.e. the less common white matter analysis performed by Choi. By means of TBM, we failed to find any significant difference by comparing either patients subgroups or healthy controls and E142 CORRESPONDENCE