ORIGINAL ARTICLE Effect of umbilical cord blood prefreeze variables on postthaw viability Belinda Pope,1,2,3 Bevan Hokin,1,4 and Ross Grant 2,3,4
BACKGROUND: Assessment of the overall postthaw (PT) viability of an umbilical cord blood (UCB) unit is an important criterion for determining the quality of the unit for transplantation. Overall PT viability is a measure of cellular damage that can occur to the UCB during collection, storage, processing, freezing, and thawing. STUDY DESIGN AND METHODS: This study investigated factors measured before freezing of the UCB that could affect overall PT viability of the stem cell unit from 257 collected cord blood samples. The analysis included hematologic variables, cord blood collection characteristics, and stem cell separation and preservation factors. RESULTS: Each of the variables, postprocess (PP) neutrophils (%), PP hematocrit, overall PP viability (%), freeze rate (°C/min), and time from collection to freezing (hr) were shown to contribute to overall PT viability. Each UCB sample was given a calculated “viability prediction” (VP) score based on the influence or impact of each of these variables. This score was compared to the measured PT viability. Variables with a low VP score had correspondingly low PT viability, indicating more overall damage to the cells. The results showed that the higher the VP score, the higher the PT viability. CONCLUSION: These findings provide a framework for identifying those units that are most likely to have a high overall PT viability and hence an increased likelihood of successful engraftment of the CB-sourced stem cells. The VP score could aid in the selection of a donor cord blood unit for transplantation.
U
mbilical cord blood (UCB) is collected as a source of hematopoietic progenitor cells for use in transplantation.1 The cord blood unit (CBU) can be transplanted immediately into a recipient. More commonly, the CBU (whole unit or selected component) is banked for future use as either a directed autologous unit or as an allogeneic unit. Transplantation of banked units may occur at the same or at a different institution and up to many years after collection. Therefore, units used for transplantation may have been processed, frozen, and stored by different institutional methods before clinical use. Interlaboratory quality control (QC) programs have found high variability in viability and CD34+ enumeration postthaw (PT) due to differing methods of thawing and testing.2 UCB is commonly collected into CPD or a similar anticoagulant. Usually, the CBUs are minimally processed by reducing the volume of plasma and red blood cells (RBCs) before freezing. This process minimizes the overall volume of cord blood frozen as well as limiting the amount of dimethyl sulfoxide (DMSO) preservative required to be added to the product.3 CBU quality is monitored throughout processing by assessing cell viability, CD34 recovery,
ABBREVIATIONS: 7-AAD = 7-aminoactinomycin D; CBA = cord blood “Apgar”; CBU = cord blood unit; PP = postprocess; PT = postthaw; SCU = stem cell unit; TTF = time to freeze; UCB = umbilical cord blood; VP = viability prediction. From the 1Pathology Department and the 2Australasian Research Institute, Sydney Adventist Hospital, Wahroonga, NSW, Australia; and the 3Faculty of Medicine, University of New South Wales; and the 4Faculty of Medicine, University of Sydney, Sydney, NSW, Australia. Address correspondence to: Belinda Pope, Pathology Department, Sydney Adventist Hospital, 185 Fox Valley Road, Wahroonga, NSW 2076, Australia; e-mail: [email protected]. Received for publication May 11, 2014; revision received August 5, 2014, and accepted August 7, 2014. doi: 10.1111/trf.12873 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
POPE ET AL.
and RBC content (hematocrit [Hct]) before and after cryopreservation. CBUs have a mixed population of cell types that respond differently to freezing and thawing protocols. Freezing protocols optimized for stem cells may cause damage to other cell types including RBCs and neutrophils.4 It has been found that the proportion of 7-aminoactinomycin D (7-AAD)-positive dead cells in a unit reflects the degree of damage to the entire unit.5 Viability of the unit may have already been compromised before cryopreservation. In a previous study we reported that time from collection to freezing (time to freeze [TTF]) and total CB volume are factors that affect postprocess (PP) viability.6 While PP viability affects PT viability, there may be other factors that contribute to decreased PT viability and poor CD34+ recoveries. The aim of this study was to determine if overall PT viability can be predicted by assessing the individual and cumulative effects of various PP variables.
MATERIALS AND METHODS Collection UCB was collected for autologous use only by an Australian private cord blood bank. After birth of the baby, UCB was collected into a triple collection cord blood collection bag (MacoPharma, Mouvaux, France) containing 29 mL of CPD. The collected cord blood was transported to the stem cell unit (SCU) sandwiched within temperaturestabilizing blocks to maintain a temperature between 15 and 25°C. The kits were temperature monitored from dispatch to receipt at the SCU using a calibrated downloadable data logger. The CBUs were couriered to the SCU within a time frame to allow for the processing and freezing to occur within 48 hours of collection.
Volume reduction The CBUs were centrifuged at 3300 × g for 12 minutes at 10°C, in oval buckets with inserts to keep the blood bags upright. The volume of cord blood was reduced using the automated blood component separator (Optipress II, Baxter-Fenwal, Deerfield, IL). Plasma was separated into the top transfer pack and RBCs into the bottom transfer pack of the cord blood collection bags system. The Optipress program was set at a buffy volume of 25 mL (mean volume, 25.6 mL; range, 13.5-35.9 mL), buffy coat level of 6.5 unit, and force of 30, using a standard back plate for buffy coat preparation as previously described by Armitage and colleagues.7 A PP sample aliquot was used for the performance of total cell count, CD34+ cell count, and viability studies.
throughout multiple tube docking procedures allowing for the addition of DMSO and plasma to the final product and the joining of the cryopreservation bags to the final product bag. Aliquots of cord blood for testing were taken either by aseptically sealing off a length of tubing attached to the collection bag and dispensing it into a sample tube or by accessing the CB using a sterile needle and syringe via a sterilized rubber injection port. All manipulation was done within a Biohazard Class II cabinet. Samples were aliquoted aseptically for testing for microbial contamination at both the postcollection and prefreezing stages of the process and tested for aerobic and anaerobic bacteria and fungi.
Freezing procedure Autologous plasma collected using the Optipress procedure was used to dilute DMSO (CryoSure, WAK Chemie, Steinbach, Germany) to a final concentration of 10%. The amount of plasma required was calculated by subtracting the PP Optipress volume (25.27 ± 1.80 mL) plus 4.5 mL of DMSO from the final desired volume of 45 mL. DMSO was added to the calculated amount of cooled plasma and placed in a refrigerator (2-8°C) to cool for an additional 10 minutes. This cooled plasma/DMSO mixture was added slowly and aseptically to the PP Optipress stem cell product using a gentle rocking motion to ensure even mixing. Samples (0.5 mL aliquots) were taken for sample pilot vials and microbial testing. The CBU was split evenly into two 20-mL ethyl vinyl acetate freezer bags (MacoPharma). CBUs and pilot samples were placed in the control rate freezer and frozen at a programmed rate of 1°C/min between 4 and −40°C and then 10°C/min between −40 and −90°C before placing directly into a vapor-phase liquid nitrogen tank (MVE, Chart, OH). All CBUs and pilot samples were stored in vapor phase at a temperature of below −150°C.
Regulation The SCU was licensed by Australia’s Therapeutic Goods Administration. The Therapeutic Goods Administration is Australia’s regulatory agency for medical drugs and devices and regulates the production and storage of cord blood–derived stem cells. The SCU unit was assessed against the Australian Code for Good Manufacturing Practice and the NETCORD Fact Standard.
Thawing procedure For QC purposes, 257 samples were tested for PT viability. Pilot samples were plunged into a 37°C water bath until thawed. DMSO was not removed.8
Maintenance of a closed system
Viability and CD34+ enumeration
The use of a sterile tubing welder (Terumo, Ashland, MA) ensured that the cord blood remained in a closed system
CD34+ enumeration on UCB samples was performed with a commercially available CD34+ enumeration kit
2
TRANSFUSION Volume **, ** **
POSTTHAW VIABILITY OF UCB
(STEMKIT, Beckman Coulter, Brea, CA) in accordance with the manufacturer’s instructions. Viability testing was performed at the same time with the fluorescent dye 7-AAD. Briefly, testing was performed in duplicate with 100 μL of cord blood sample incubated with 20 μL of CD45FITC/CD34PE/7-AAD for 20 minutes at room temperature. An isoclonic control/7-AAD sample tube was also prepared. The RBCs were then lysed in 2 mL of NH4Cl and incubated for 10 minutes at room temperature. Before analysis, 100 μL of StemCount was added to the tubes and analyzed on a flow cytometer (Cytomics FC500, Beckman Coulter). Overall PT viable cells were defined as CD45+/7AAD+. The data were analyzed using the software (STEMONE, Beckman Coulter). The instrument was checked for alignment daily using FlowCheck (Beckman Coulter).
Hematologic cell counts A complete blood count and differential were performed on all samples using an automated hematology analyzer (either DxH 800, Beckman Coulter; or CellDyn 3200, Abbott Diagnostics, Abbott Park, IL).
Statistical analysis Results are expressed as mean ± 1 standard deviation (SD). After the data were confirmed to be normally distributed, the t test (paired and independent) was used to assess differences between two discrete populations. Associations between variables were assessed using Spearman’s correlation test. Differences were considered significant when p values were 0.05 or less. The statistical analysis was performed using a commercial computer package (SPSS/PC, IBM, Armonk, NY). One-way analysis of variance (ANOVA) with Tukey’s HSD was used to assess significance between groups. Multivariate analysis was used to assess the contribution each factor made to overall significance and the viability prediction score.
RESULTS Cohort characteristics PT viability studies were performed on the entire cell population in accordance with the International Society of Hematotherapy and Graft Engineering gating strategy, modified to maximize the detection of live and dead cells9 in 257 samples. The age of the mothers in the study was 33.0 ± 4.6 years (range, 18-53 years) and the gestational age of the babies was 39.2 ± 1.3 weeks (range, 33-41 weeks). The mean UCB volume collected was 80.7 ± 32.1 mL (range, 21.4-199.5 mL), which was subsequently mixed into a standard volume of 29 mL of CPD anticoagulant, diluting the UCB volume on average by 28.8 mL ± 8.7% (range, 13%-58%). The mean time from
collection to freezing (described as TTF) was 27:35 ± 9:14 hours (range, 5:28-54:03 hr). The PP total nucleated cell count was 9.1 × 108 ± 4.6 × 108 (range, 1.4 × 108-26.2 × 108) and the total CD34 count was 4.1 × 106 ± 4.3 × 106 (range, 0.3 × 106-31.9 × 106) The mean freeze volume containing 10% DMSO was 42.3 ± 1.6 mL (range, 35-45.7 mL). The mean volume of plasma added to the freeze mix was 15.2 ± 1.8 mL (range, 35-45.7 mL). The freeze mix containing 10% DMSO was in contact with the cells before freezing for 12.7 ± 6.4 minutes (range, 1-33 min) before commencing the programmed controlled-rate freeze. The processed cells were frozen between −4 and −40°C at a mean rate of −0.85 ± 0.10°C/min (range, −0.7 to −1.4) in the controlled-rate freezer. The mean prefreeze (PP) viability of the cohort was 93.18 ± 5.68% (range, 66.6%99.0%). The mean PT viability of the cohort was 65.4 ± 12.7% with a range of 29.7% to 92%. CD34 recovery after thawing was 44.97 ± 15.50% (range, 7.9%-99.63%). PT total CD34 was 2.1 × 106 ± 2.5 × 106 (range, 0.1 × 10618.6 × 106).
Relationship between PT viability, PP viability, and PT CD34 recovery PT viability was consistently lower than PP viability. PT viability positively correlated with PP viability (r = 0.218, p = 0.001). However, other factors are also likely to influence PT viability as some samples exhibited a large difference between PP viability and PT viability even when the PP viability was more than 95%. This difference between the two variables is highly variable (Fig. 1). PT viability positively correlated with percentage of CD34 recovery PT (r = 0.358, p > 0.001).
Factors affecting PT viability All available variables were investigated for correlations with PT viability. Correlations were found between PT percent viability and each of the following variables: PP overall percentage of neutrophils (%) (r = −0.266, p < 0.001), total number of PP RBCs (r = −0.319, p ≤ 0.001), PP Hb (r = −0.231, p < 0.001), PP Hct (r = −0.268, p < 0.001), PP RBC distribution width (r = 0.216, p = 0.001), PP percentage of viability (r = 0.218, p = 0.001), volume of plasma in freeze mix (mL; r = 0.181, p = 0.004), total freeze volume (mL; r = 0.181, p = 0.004), freeze rate (°C/min; r = −0.174, p = 0.007), TTF (hr; r = −0.290, p ≤ 0.001), and PT percentage of CD34 recovery (r = 0.358, p < 0.001). Importantly, each of these variables, with the exception of PT CD34 recovery, was able to be assessed before thawing of the samples. The variables, volume of plasma in the freeze mix and the freeze volume are specific to the individual processing laboratory and have not been included in any further analysis. Volume **, ** **
TRANSFUSION
3
POPE ET AL.
viability were divided into quintiles. Freeze rate was divided into groups of 0.1°C/min increments. TTF was divided into time intervals of 12 hours. Table 1 shows the descriptions for all the groups (G). Table 2 shows the mean PT viability of each group. Lower PP neutrophil (%), Hct, freeze rate, and TTF and higher PP viability are associated with higher PT viability.
Score calculation
Fig. 1. PP viability compared to PT viability.
PP RBC variables all correlated with PT viability, Hct (r = −0.231, p < 0.001), Hb (r = −0.231, p < 0.001), total number of RBCs (r = −0.319, p ≤ 0.001), and RBC distribution width (r = 0.216, p = 0.001), and were also highly correlated with each other. For this reason only Hct, traditionally used in blood banking as a measure of RBC volume, was chosen for further analysis. As expected, Hct positively correlated with total number of RBCs (r = 0.751 p ≤ 0.001), Hb (r = 0.812, p ≤ 0.001), and RBC distribution width (r = −0.177, p = 0.004). In this study, cell concentration at the time of freezing was not found to impact PT viability. This study found that PP white blood cell (WBC) concentration correlated with total PP neutrophil cell count (r = 0.696, p > 0.001). However, increasing PP WBC concentration had only a weak correlation with the percentage of neutrophils PP (r = 0.131, p = 0.037). RBC count and Hct were found to be independent of PP cell concentration. The variables PP neutrophils, Hct, PP viability, freeze rate, and TTF were subjected to further analysis. Multiple regression analysis revealed that these five variables explained 30% (R2 = 0.300, p < 0.001) of the variance in PT viability.
Variable analysis Five variables were found to have an impact on PT viability. To further analyze the impact of these variables on PT viability, the data were analyzed as percentiles or divided into periods of time or freezing rate. Based on significance, PP neutrophils (%) were divided into quartiles, Hct and PP 4
TRANSFUSION Volume **, ** **
The five variables that were found to impact PT viability were PP percentage of neutrophils, PP Hct (%), PP viability (%), freeze rate (°C/min), and TTF (hr). For each of these variables a score of 1 was assigned to each group that was associated with a positive shift in PT viability (Table 3). Therefore, an overall PT VP score could be calculated for each UCB sample, with a maximum possible score of 5, by summing the scores for each of the five variables relative to their effect on PT viability (Table 3). One variable had a result of 0 and was combined with the score category 1. The total score data were compared against PT viability (%). Figure 2 shows a boxplot of each of the final scores. The higher the final score the higher the PT viability. A one way, between-groups ANOVA was conducted to explore the relationship between the final VP score of combined variables and PT viability. There was a significant overall difference between the final VP scores (F = 3, 24.336; p < 0.001). Post hoc comparisons using the Tukey’s HSD test found that score categories 1, 2 were different from all other scores (p < 0.001). Scores 3 and 4 were not significantly different from each other. The mean PT viability for Score 1 (n = 16) was 48.48 ± 8.80%; Score 2 (n = 79), 61.37 ± 11.74%; Score 3 (n = 119), 67.40 ± 11.49%; Score 4 (n = 38), 72.37 ± 10.00%; and Score 5 (n = 5), 82.65 ± 8.42%. While Group 5 had a low sample number it was included to show the continuing trend of increased PT viability with increasing VP score. These CBUs were collected in a private blood bank setting and were frozen regardless of PP viability. Nine CBUs had a low PP viability of less than 80%; eight of the nine of these samples had a lowVP score of 2 predicting that PT viability will also be low. One of the nine CBUs had an intermediate VP score of 3. Conversely those CBUs with a high PP viability could have a low VP score predicting a low PT viability when all factors were taken into consideration. The data were also analyzed for a smaller cohort of CBUs that had the desirable CB banking characteristics
POSTTHAW VIABILITY OF UCB
TABLE 1. Characteristics of groupings Group statistics Group 1 Group 2 Group 3 Group 4 Group 5
Neutrophils (%) ≤41.20 41.21-49.35 49.36-55.82 ≥55.83
Hct (%) ≤0.590 0.591-0.620 0.621-0.650 0.651-0.690 ≥0.691
PP viability (%) ≤90.2 90.3-94.0 94.1-95.8 95.9-97.3 ≥97.4
Freeze rate (°C/min) ≤0.8 0.8-0.9 >0.9
TTF (hr) ≤12 12-24 24-36 >36
TABLE 2. Mean PT viability for groupings* Group statistics ANOVA F (p) Group 1 Group 2 Group 3 Group 4 Group 5
Neutrophils (%) 7.983 (
BACKGROUND: Assessment of the overall postthaw (PT) viability of an umbilical cord blood (UCB) unit is an important criterion for determining the quality of the unit for transplantation. Overall PT viability is a measure of cellular damage that can occur to the UCB during collection, storage, processing, freezing, and thawing. STUDY DESIGN AND METHODS: This study investigated factors measured before freezing of the UCB that could affect overall PT viability of the stem cell unit from 257 collected cord blood samples. The analysis included hematologic variables, cord blood collection characteristics, and stem cell separation and preservation factors. RESULTS: Each of the variables, postprocess (PP) neutrophils (%), PP hematocrit, overall PP viability (%), freeze rate (°C/min), and time from collection to freezing (hr) were shown to contribute to overall PT viability. Each UCB sample was given a calculated “viability prediction” (VP) score based on the influence or impact of each of these variables. This score was compared to the measured PT viability. Variables with a low VP score had correspondingly low PT viability, indicating more overall damage to the cells. The results showed that the higher the VP score, the higher the PT viability. CONCLUSION: These findings provide a framework for identifying those units that are most likely to have a high overall PT viability and hence an increased likelihood of successful engraftment of the CB-sourced stem cells. The VP score could aid in the selection of a donor cord blood unit for transplantation.
U
mbilical cord blood (UCB) is collected as a source of hematopoietic progenitor cells for use in transplantation.1 The cord blood unit (CBU) can be transplanted immediately into a recipient. More commonly, the CBU (whole unit or selected component) is banked for future use as either a directed autologous unit or as an allogeneic unit. Transplantation of banked units may occur at the same or at a different institution and up to many years after collection. Therefore, units used for transplantation may have been processed, frozen, and stored by different institutional methods before clinical use. Interlaboratory quality control (QC) programs have found high variability in viability and CD34+ enumeration postthaw (PT) due to differing methods of thawing and testing.2 UCB is commonly collected into CPD or a similar anticoagulant. Usually, the CBUs are minimally processed by reducing the volume of plasma and red blood cells (RBCs) before freezing. This process minimizes the overall volume of cord blood frozen as well as limiting the amount of dimethyl sulfoxide (DMSO) preservative required to be added to the product.3 CBU quality is monitored throughout processing by assessing cell viability, CD34 recovery,
ABBREVIATIONS: 7-AAD = 7-aminoactinomycin D; CBA = cord blood “Apgar”; CBU = cord blood unit; PP = postprocess; PT = postthaw; SCU = stem cell unit; TTF = time to freeze; UCB = umbilical cord blood; VP = viability prediction. From the 1Pathology Department and the 2Australasian Research Institute, Sydney Adventist Hospital, Wahroonga, NSW, Australia; and the 3Faculty of Medicine, University of New South Wales; and the 4Faculty of Medicine, University of Sydney, Sydney, NSW, Australia. Address correspondence to: Belinda Pope, Pathology Department, Sydney Adventist Hospital, 185 Fox Valley Road, Wahroonga, NSW 2076, Australia; e-mail: [email protected]. Received for publication May 11, 2014; revision received August 5, 2014, and accepted August 7, 2014. doi: 10.1111/trf.12873 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
POPE ET AL.
and RBC content (hematocrit [Hct]) before and after cryopreservation. CBUs have a mixed population of cell types that respond differently to freezing and thawing protocols. Freezing protocols optimized for stem cells may cause damage to other cell types including RBCs and neutrophils.4 It has been found that the proportion of 7-aminoactinomycin D (7-AAD)-positive dead cells in a unit reflects the degree of damage to the entire unit.5 Viability of the unit may have already been compromised before cryopreservation. In a previous study we reported that time from collection to freezing (time to freeze [TTF]) and total CB volume are factors that affect postprocess (PP) viability.6 While PP viability affects PT viability, there may be other factors that contribute to decreased PT viability and poor CD34+ recoveries. The aim of this study was to determine if overall PT viability can be predicted by assessing the individual and cumulative effects of various PP variables.
MATERIALS AND METHODS Collection UCB was collected for autologous use only by an Australian private cord blood bank. After birth of the baby, UCB was collected into a triple collection cord blood collection bag (MacoPharma, Mouvaux, France) containing 29 mL of CPD. The collected cord blood was transported to the stem cell unit (SCU) sandwiched within temperaturestabilizing blocks to maintain a temperature between 15 and 25°C. The kits were temperature monitored from dispatch to receipt at the SCU using a calibrated downloadable data logger. The CBUs were couriered to the SCU within a time frame to allow for the processing and freezing to occur within 48 hours of collection.
Volume reduction The CBUs were centrifuged at 3300 × g for 12 minutes at 10°C, in oval buckets with inserts to keep the blood bags upright. The volume of cord blood was reduced using the automated blood component separator (Optipress II, Baxter-Fenwal, Deerfield, IL). Plasma was separated into the top transfer pack and RBCs into the bottom transfer pack of the cord blood collection bags system. The Optipress program was set at a buffy volume of 25 mL (mean volume, 25.6 mL; range, 13.5-35.9 mL), buffy coat level of 6.5 unit, and force of 30, using a standard back plate for buffy coat preparation as previously described by Armitage and colleagues.7 A PP sample aliquot was used for the performance of total cell count, CD34+ cell count, and viability studies.
throughout multiple tube docking procedures allowing for the addition of DMSO and plasma to the final product and the joining of the cryopreservation bags to the final product bag. Aliquots of cord blood for testing were taken either by aseptically sealing off a length of tubing attached to the collection bag and dispensing it into a sample tube or by accessing the CB using a sterile needle and syringe via a sterilized rubber injection port. All manipulation was done within a Biohazard Class II cabinet. Samples were aliquoted aseptically for testing for microbial contamination at both the postcollection and prefreezing stages of the process and tested for aerobic and anaerobic bacteria and fungi.
Freezing procedure Autologous plasma collected using the Optipress procedure was used to dilute DMSO (CryoSure, WAK Chemie, Steinbach, Germany) to a final concentration of 10%. The amount of plasma required was calculated by subtracting the PP Optipress volume (25.27 ± 1.80 mL) plus 4.5 mL of DMSO from the final desired volume of 45 mL. DMSO was added to the calculated amount of cooled plasma and placed in a refrigerator (2-8°C) to cool for an additional 10 minutes. This cooled plasma/DMSO mixture was added slowly and aseptically to the PP Optipress stem cell product using a gentle rocking motion to ensure even mixing. Samples (0.5 mL aliquots) were taken for sample pilot vials and microbial testing. The CBU was split evenly into two 20-mL ethyl vinyl acetate freezer bags (MacoPharma). CBUs and pilot samples were placed in the control rate freezer and frozen at a programmed rate of 1°C/min between 4 and −40°C and then 10°C/min between −40 and −90°C before placing directly into a vapor-phase liquid nitrogen tank (MVE, Chart, OH). All CBUs and pilot samples were stored in vapor phase at a temperature of below −150°C.
Regulation The SCU was licensed by Australia’s Therapeutic Goods Administration. The Therapeutic Goods Administration is Australia’s regulatory agency for medical drugs and devices and regulates the production and storage of cord blood–derived stem cells. The SCU unit was assessed against the Australian Code for Good Manufacturing Practice and the NETCORD Fact Standard.
Thawing procedure For QC purposes, 257 samples were tested for PT viability. Pilot samples were plunged into a 37°C water bath until thawed. DMSO was not removed.8
Maintenance of a closed system
Viability and CD34+ enumeration
The use of a sterile tubing welder (Terumo, Ashland, MA) ensured that the cord blood remained in a closed system
CD34+ enumeration on UCB samples was performed with a commercially available CD34+ enumeration kit
2
TRANSFUSION Volume **, ** **
POSTTHAW VIABILITY OF UCB
(STEMKIT, Beckman Coulter, Brea, CA) in accordance with the manufacturer’s instructions. Viability testing was performed at the same time with the fluorescent dye 7-AAD. Briefly, testing was performed in duplicate with 100 μL of cord blood sample incubated with 20 μL of CD45FITC/CD34PE/7-AAD for 20 minutes at room temperature. An isoclonic control/7-AAD sample tube was also prepared. The RBCs were then lysed in 2 mL of NH4Cl and incubated for 10 minutes at room temperature. Before analysis, 100 μL of StemCount was added to the tubes and analyzed on a flow cytometer (Cytomics FC500, Beckman Coulter). Overall PT viable cells were defined as CD45+/7AAD+. The data were analyzed using the software (STEMONE, Beckman Coulter). The instrument was checked for alignment daily using FlowCheck (Beckman Coulter).
Hematologic cell counts A complete blood count and differential were performed on all samples using an automated hematology analyzer (either DxH 800, Beckman Coulter; or CellDyn 3200, Abbott Diagnostics, Abbott Park, IL).
Statistical analysis Results are expressed as mean ± 1 standard deviation (SD). After the data were confirmed to be normally distributed, the t test (paired and independent) was used to assess differences between two discrete populations. Associations between variables were assessed using Spearman’s correlation test. Differences were considered significant when p values were 0.05 or less. The statistical analysis was performed using a commercial computer package (SPSS/PC, IBM, Armonk, NY). One-way analysis of variance (ANOVA) with Tukey’s HSD was used to assess significance between groups. Multivariate analysis was used to assess the contribution each factor made to overall significance and the viability prediction score.
RESULTS Cohort characteristics PT viability studies were performed on the entire cell population in accordance with the International Society of Hematotherapy and Graft Engineering gating strategy, modified to maximize the detection of live and dead cells9 in 257 samples. The age of the mothers in the study was 33.0 ± 4.6 years (range, 18-53 years) and the gestational age of the babies was 39.2 ± 1.3 weeks (range, 33-41 weeks). The mean UCB volume collected was 80.7 ± 32.1 mL (range, 21.4-199.5 mL), which was subsequently mixed into a standard volume of 29 mL of CPD anticoagulant, diluting the UCB volume on average by 28.8 mL ± 8.7% (range, 13%-58%). The mean time from
collection to freezing (described as TTF) was 27:35 ± 9:14 hours (range, 5:28-54:03 hr). The PP total nucleated cell count was 9.1 × 108 ± 4.6 × 108 (range, 1.4 × 108-26.2 × 108) and the total CD34 count was 4.1 × 106 ± 4.3 × 106 (range, 0.3 × 106-31.9 × 106) The mean freeze volume containing 10% DMSO was 42.3 ± 1.6 mL (range, 35-45.7 mL). The mean volume of plasma added to the freeze mix was 15.2 ± 1.8 mL (range, 35-45.7 mL). The freeze mix containing 10% DMSO was in contact with the cells before freezing for 12.7 ± 6.4 minutes (range, 1-33 min) before commencing the programmed controlled-rate freeze. The processed cells were frozen between −4 and −40°C at a mean rate of −0.85 ± 0.10°C/min (range, −0.7 to −1.4) in the controlled-rate freezer. The mean prefreeze (PP) viability of the cohort was 93.18 ± 5.68% (range, 66.6%99.0%). The mean PT viability of the cohort was 65.4 ± 12.7% with a range of 29.7% to 92%. CD34 recovery after thawing was 44.97 ± 15.50% (range, 7.9%-99.63%). PT total CD34 was 2.1 × 106 ± 2.5 × 106 (range, 0.1 × 10618.6 × 106).
Relationship between PT viability, PP viability, and PT CD34 recovery PT viability was consistently lower than PP viability. PT viability positively correlated with PP viability (r = 0.218, p = 0.001). However, other factors are also likely to influence PT viability as some samples exhibited a large difference between PP viability and PT viability even when the PP viability was more than 95%. This difference between the two variables is highly variable (Fig. 1). PT viability positively correlated with percentage of CD34 recovery PT (r = 0.358, p > 0.001).
Factors affecting PT viability All available variables were investigated for correlations with PT viability. Correlations were found between PT percent viability and each of the following variables: PP overall percentage of neutrophils (%) (r = −0.266, p < 0.001), total number of PP RBCs (r = −0.319, p ≤ 0.001), PP Hb (r = −0.231, p < 0.001), PP Hct (r = −0.268, p < 0.001), PP RBC distribution width (r = 0.216, p = 0.001), PP percentage of viability (r = 0.218, p = 0.001), volume of plasma in freeze mix (mL; r = 0.181, p = 0.004), total freeze volume (mL; r = 0.181, p = 0.004), freeze rate (°C/min; r = −0.174, p = 0.007), TTF (hr; r = −0.290, p ≤ 0.001), and PT percentage of CD34 recovery (r = 0.358, p < 0.001). Importantly, each of these variables, with the exception of PT CD34 recovery, was able to be assessed before thawing of the samples. The variables, volume of plasma in the freeze mix and the freeze volume are specific to the individual processing laboratory and have not been included in any further analysis. Volume **, ** **
TRANSFUSION
3
POPE ET AL.
viability were divided into quintiles. Freeze rate was divided into groups of 0.1°C/min increments. TTF was divided into time intervals of 12 hours. Table 1 shows the descriptions for all the groups (G). Table 2 shows the mean PT viability of each group. Lower PP neutrophil (%), Hct, freeze rate, and TTF and higher PP viability are associated with higher PT viability.
Score calculation
Fig. 1. PP viability compared to PT viability.
PP RBC variables all correlated with PT viability, Hct (r = −0.231, p < 0.001), Hb (r = −0.231, p < 0.001), total number of RBCs (r = −0.319, p ≤ 0.001), and RBC distribution width (r = 0.216, p = 0.001), and were also highly correlated with each other. For this reason only Hct, traditionally used in blood banking as a measure of RBC volume, was chosen for further analysis. As expected, Hct positively correlated with total number of RBCs (r = 0.751 p ≤ 0.001), Hb (r = 0.812, p ≤ 0.001), and RBC distribution width (r = −0.177, p = 0.004). In this study, cell concentration at the time of freezing was not found to impact PT viability. This study found that PP white blood cell (WBC) concentration correlated with total PP neutrophil cell count (r = 0.696, p > 0.001). However, increasing PP WBC concentration had only a weak correlation with the percentage of neutrophils PP (r = 0.131, p = 0.037). RBC count and Hct were found to be independent of PP cell concentration. The variables PP neutrophils, Hct, PP viability, freeze rate, and TTF were subjected to further analysis. Multiple regression analysis revealed that these five variables explained 30% (R2 = 0.300, p < 0.001) of the variance in PT viability.
Variable analysis Five variables were found to have an impact on PT viability. To further analyze the impact of these variables on PT viability, the data were analyzed as percentiles or divided into periods of time or freezing rate. Based on significance, PP neutrophils (%) were divided into quartiles, Hct and PP 4
TRANSFUSION Volume **, ** **
The five variables that were found to impact PT viability were PP percentage of neutrophils, PP Hct (%), PP viability (%), freeze rate (°C/min), and TTF (hr). For each of these variables a score of 1 was assigned to each group that was associated with a positive shift in PT viability (Table 3). Therefore, an overall PT VP score could be calculated for each UCB sample, with a maximum possible score of 5, by summing the scores for each of the five variables relative to their effect on PT viability (Table 3). One variable had a result of 0 and was combined with the score category 1. The total score data were compared against PT viability (%). Figure 2 shows a boxplot of each of the final scores. The higher the final score the higher the PT viability. A one way, between-groups ANOVA was conducted to explore the relationship between the final VP score of combined variables and PT viability. There was a significant overall difference between the final VP scores (F = 3, 24.336; p < 0.001). Post hoc comparisons using the Tukey’s HSD test found that score categories 1, 2 were different from all other scores (p < 0.001). Scores 3 and 4 were not significantly different from each other. The mean PT viability for Score 1 (n = 16) was 48.48 ± 8.80%; Score 2 (n = 79), 61.37 ± 11.74%; Score 3 (n = 119), 67.40 ± 11.49%; Score 4 (n = 38), 72.37 ± 10.00%; and Score 5 (n = 5), 82.65 ± 8.42%. While Group 5 had a low sample number it was included to show the continuing trend of increased PT viability with increasing VP score. These CBUs were collected in a private blood bank setting and were frozen regardless of PP viability. Nine CBUs had a low PP viability of less than 80%; eight of the nine of these samples had a lowVP score of 2 predicting that PT viability will also be low. One of the nine CBUs had an intermediate VP score of 3. Conversely those CBUs with a high PP viability could have a low VP score predicting a low PT viability when all factors were taken into consideration. The data were also analyzed for a smaller cohort of CBUs that had the desirable CB banking characteristics
POSTTHAW VIABILITY OF UCB
TABLE 1. Characteristics of groupings Group statistics Group 1 Group 2 Group 3 Group 4 Group 5
Neutrophils (%) ≤41.20 41.21-49.35 49.36-55.82 ≥55.83
Hct (%) ≤0.590 0.591-0.620 0.621-0.650 0.651-0.690 ≥0.691
PP viability (%) ≤90.2 90.3-94.0 94.1-95.8 95.9-97.3 ≥97.4
Freeze rate (°C/min) ≤0.8 0.8-0.9 >0.9
TTF (hr) ≤12 12-24 24-36 >36
TABLE 2. Mean PT viability for groupings* Group statistics ANOVA F (p) Group 1 Group 2 Group 3 Group 4 Group 5
Neutrophils (%) 7.983 (
