Introduction

The use of cryopreserved autologous as well as fresh PBSC allogeneic grafts has been well established for decades. Recent data from the Centre for International Bone Marrow Transplant Research and the National Marrow Donor Program suggest that the practice of using frozen allogeneic PBSC is becoming increasingly common among transplant centres.¹ However, until now the few clinical studies published on this topic did only compare fresh vs frozen BM-derived allografts.^{2, 3, 4, 5, 6, 7} There are no reported outcomes for recipients of cryopreserved allogeneic PBSC grafts. Based on reviewed literature, Frey et al.¹ suggest that the available data do not yet sufficiently justify the dogmatic use of fresh over frozen allografts.

In our last report⁸ we confirmed the independent prognostic values of transplanted colony-forming units (CFU) but not CD34⁺ numbers for both leucocyte and platelet engraftment. Those data underline the importance of using functional assays rather than phenotypic markers for assessing the quality of stem cell grafts. We have shown earlier that aldehyde dehydrogenase (ALDH) expression is a very useful marker for stem and progenitor cell activity (SSC^loALDH^br) in haematopoietic SCT correlating well with both engraftment data and CFU assays.⁹

For the present retrospective analysis we measured SSC^loALDH^br and CD34⁺ numbers, membrane integrity (MI) as well as colony-forming potential of HSC grafts that had been cryopreserved between 2000 and 2007 for allogeneic SCT. All parameters were assessed before as well as after freezing and thawing of HSC using reference samples frozen in parallel to the clinical grafts. Obtained results were set into relation to the individual clinical engraftment data of the respective patients.

Our data indicate that the sensitivity of PBSC towards possible deleterious effects of cryopreservation increases after prolonged transportation and storage. We show, moreover, that functional analysis of HSC grafts is highly predictive of later engraftment failure.

Patients and methods

Since cryopreservation is not a standard procedure with allogeneic transplants,^{1, 2, 3, 4, 5, 6, 7} the number of allografts analysed in this study was limited. However, between 2000 and 2007, 36 allogeneic PBSC and 13 allogeneic BM grafts had to be frozen due to medical and/or logistic reasons in our hospital. All 13 patients for whom BM had been stored were transplanted, whereas only 33 out of the 36 patients received their cryopreserved PBSC (3 patients died before transplantation). Relevant data are summarized in Table 1 (for BMT) and Table 2 (for PBSCT). A total of 38 of the transplanted grafts came from HLA-matched and 8 were from HLA-mismatched donors; 11 HSC grafts were from related and 35 were from unrelated donors. Matched donors were required to match the recipients for HLA-A, HLA-B, DRB1 and DQB1. All recipients were nursed at the Clinic for Stem Cell Transplantation in laminar air-flow isolation rooms.

Table 1 - In vitro characteristics of allogeneic BM grafts before and after cryopreservation in relation to clinical transplantation data.

Full table

Table 2 - In vitro characteristics of allogeneic PBSC grafts before and after cryopreservation in relation to clinical transplantation data.

Full table

Protocol of cryopreservation and thawing

All BM grafts were cryopreserved as described.⁴ Briefly, BM harvested from donors was collected in marrow collection bags and filtered by gravity passage through filters provided by the manufacturer (Baxter, Deerfield, IL, USA). 'In-house' and 'externally' harvested BM grafts were frozen within 24 h after the donation. After Ficoll-Paque separation, the separated marrow cells were suspended in plasma. Before freezing, dimethyl sulfoxide (DMSO) was added to a final concentration of 10%, BM was injected into freezing bags (Baxter, Cryocyte 4R 9957) and cryopreserved with the following chamber settings of the programmable freezer (Messer Griesheim, Sulzbach, Germany): -1 °C per min until -6 °C, then -25 °C per min until -50 °C, then +15 °C per min until -18 °C, then -1 °C per min until -45 °C and then -10 °C per min until -90 °C. Then BM was transferred to and stored in liquid nitrogen (-196 °C).

The protocol for cryopreservation of PBSC grafts has been described elsewhere.^{10, 11} Briefly, 'in-house' apheresis products were always frozen within 24 h. In total, 19 out of 25 external PBSC were cryopreserved within 24 h after the donation, whereas 6 out of 25 grafts were frozen in the period between 24 and 48 h after donation. No adjustment for storage of the pre-freeze cell concentration was done at the time of this study. The storage temperature prior to freezing varied between +4 and +8 °C, but transport temperatures had not been continually documented at that time. All PBSCs were cryopreserved after adding an additional 15 ml ACD-A (Baxter) as an anti-coagulant to the final apheresis product as recommended.¹⁰ The apheresis products (approximately 200–400 ml) were centrifuged for volume adjustment and cryopreserved in 1–3 (on an average 2) aliquots of 105 ml in freezing bags in the presence of 10% DMSO (Cryo Sure, WAK-Chemie, Steinbach, Germany). For freezing we used the same protocol as for BM, which resulted in a cooling rate of 1.4 K per min in the critical temperature range between the end of the plateau phase (approximately -8 °C) and -45 °C.¹¹ From -45 to -90 °C cooling was performed at a rate of approximately 10 °C per min, thereafter the samples were either transferred to the vapour phase over liquid nitrogen or stored directly therein. The temperature inside the samples stored in the vapour phase over liquid nitrogen was always below -150 °C. This is below the glass transition temperature and therefore considered to be sufficiently low to prevent any time-dependent deterioration.

The thawing procedures for BM and PBSC were also identical. All frozen grafts were thawed rapidly at the time of HSCT in a warm water bath at 37 °C and transfused unfiltered through a central venous catheter.

The number of available frozen reference samples for analysis was limited to 31 consecutive PBSC and 8 BM allografts. They were thawed under the same conditions as grafts for HSCT.

Cell counting and flow cytometry

After thawing numbers of erythrocytes (RBC), platelets (PLT) and leucocytes (WBC) were measured on the micro cell counter Sysmex F-800 (Sysmex Deutschland GmbH, Norderstedt, Germany). Cell probes were then analysed using flow cytometry essentially as previously described.¹² Briefly, CD34 expression was examined according to the International Society of Hematotherapy and Graft Engineering (ISHAGE) guidelines. SSC^loALDH^br were identified using AldeFluor (Stem Cell Technologies, St. Katharinen, Germany), and CD34⁺ cell MI was determined with 7-amino actinomycin D (7-AAD) dye¹³ (Beckman Coulter, Krefeld, Germany). FACS data were acquired and analysed on an FACSCalibur (BD Biosciences, Heidelberg, Germany).

Determination of CFU-GM

In parallel to the cell counting and FACS analyses, CFU-GM assays were carried out before freezing and after thawing essentially as described.¹⁴ Briefly, 1 10⁵ cells were seeded in 1 ml methyl cellulose supplemented with growth factors (Methocult, Stem Cell Technologies), the numbers of CFU-GM were counted after 11–14 days. CFU-GM determinations for unfrozen and frozen/thawed BM and unfrozen PBSC were carried out in the laboratories of the Clinic for Stem Cell Transplantation, values for PBSC after thawing were assessed by the Department for Transfusion Medicine of our University Medical Centre.

Statistical analysis

Biostatistical analysis was carried out with WinSTAT3.1 statistical programme using the Student's t-test and Spearman's range correlation tests. A P-value below 0.05 was considered as indicative for statistical significance.

Results

BM before and after cryopreservation

We did not find any statistically significant changes in CD34⁺ contents of BM samples after one freezing/thawing cycle (Table 1). There were also no significant differences (P=0.77) with regard to MI of CD34⁺ cells after cryopreservation (n=8) compared to values obtained from unfrozen allografts in a different control group (n=14): 89.33.5%, range 87–93 and 97.42.6, 92–100%, respectively (Figure 1a). The amount of SSC^loALDH^br cells after cryopreservation correlated well with the CD34⁺ cell content in frozen (mean ALDH^br/CD34⁺ ratio: 1.010.06) as well as in unfrozen samples (0.940.07; Figure 1b). The in vitro growth potential of HSC and progenitors (as measured as CFU activity) was also not influenced by the freezing/thawing cycle (5.544.46 10⁵ per kg body weight (BW) before cryopreservation and 6.465.03 10⁵ per kg BW after thawing respectively, P=0.68) (Figure 1c). All available patients who received cryopreserved BM had a well-timed leucocyte engraftment (at day 17.232.80, n=13) and reached full donor chimaerism (n=8). Two of these 13 patients were conditioned with a reduced-intensity conditioning (RIC) regimen. There were also no differences in GVHD between patients who received fresh vs frozen products.

Figure 1.

Influence of cryopreservation on cell properties in BM vs PBSC allografts. Cryopreservation is associated with a significant decrease of (a) membrane integrity, (b) ALDH^br/CD34⁺ ratio and (c) colony-forming unit (CFU) activity in PBSC, but not BM allografts.

Full figure and legend (104K)

PBSC before and after cryopreservation

In contrast to the results obtained with BM, MI of CD34⁺ cells from available frozen/thawed PBSC samples (n=31; Table 2) was significantly (P<0.01) impaired as compared to unfrozen allografts (control group, n=32): 69.69.4%, range 19–99 and 96.45.6, 91–100%, respectively (Figure 1a). Accordingly, the numbers of viable (7-AAD-negative) CD34⁺ cells per kg patient BW were significantly (P<0.001) lower after thawing (5.52.9 10⁶ per kg) as compared to the number of CD34⁺ cells per kg BW calculated before freezing (10.97.6 10⁶ per kg). Also, the SSC^loALDH^br/CD34⁺ ratio was significantly (P<0.001) decreased in frozen/thawed PBSC samples (0.710.27) as compared to this parameter in unfrozen probes (0.930.17) (Figure 1b). In line, the CFU potential after cryopreservation turned out to be also significantly (P<0.0001) lower (6.575.64 10⁵ per kg BW) than before the freezing/thawing cycle (26.5212.56 10⁵ per kg BW) (Figure 1c).

Nine out of the thirty-three patients who received cryopreserved PBSC allografts did not achieve engraftment. Six of them had a primary graft failure; three died without engraftment correspondingly at days 19, 26 and 27 after HSCT. Eight of the nine patients who showed no engraftment had been treated with an RIC regimen.

Indications for impaired engraftment with decreased numbers of SSC^loALDH^br cells

As shown above, cryopreserved PBSC samples revealed very variable ranges for MI, functionality and in vitro growth activity. Therefore, for further analysis we combined the 22 samples with comparatively high (>0.5) SSC^loALDH^br/CD34⁺ ratios (mean 0.87, range 0.59–1.01) into one group (group I), whereas those 9 of the available 31 frozen probes with particularly low (<0.5) SSC^loALDH^br/CD34⁺ ratios (mean 0.32, range 0.26–0.50) comprised group II (Figure 2a). Interestingly, the mean CD34⁺ percentages in the two groups did not differ significantly (P=0.17) from each other (Figure 2b) but the 9 samples united in group II showed significantly (P<0.0001) lower rates of SSC^loALDH^br cells (0.150.03%) compared to the other 22 cryopreserved transplants from group I (0.54%0.23%) (Figure 2b). Notably, only one out of nine grafts from group II was frozen later than 24 h, but within 48 h after donation.

Figure 2.

Frozen/thawed PBSC allografts may functionally be divided into two groups—conserved (group I) vs decreased (group II) aldehyde dehydrogenase (ALDH) activity. All available cryopreserved PBSC samples were analysed regarding the remaining ALDH activity of putative HSC after the freezing/thawing cycle. (a) Based on those data, the 31 samples ('total') were divided into two groups: Those with normal (or close to normal) SSC^loALDH^br to CD34⁺ cell ratios (that is>0.5, group I) and those with decreased ratios (<0.5, group II) (see Table 2, in bold). (b) Notably, the numbers of CD34⁺ cells were very similar in both groups (left diagram), whereas the numbers of SSC^loALDH^br were significantly lower in group II as compared to group I (right diagram). (c) In line, there was no significant difference in both CD34⁺ numbers per kg body weight (BW) before (left diagram) and after thawing (middle). In contrast, the numbers of SSC^loALDH^br cells transplanted per kg BW were significantly lower in group II as compared to group I (right diagram). (d) As expected, the colony-forming unit (CFU) potential was also strongly decreased in group II.

Full figure and legend (125K)

We next compared the numbers of CD34⁺ vs ALDH^br cells actually transplanted per kg in these two groups. CD34⁺ counts had initially been measured before cryopreservation (Figure 2c)—there was almost no difference between the two groups (group I: 11.87.8 10⁶ per kg vs group II: 10.57.4 10⁶ per kg). Furthermore, no significant (P=0.82) difference in CD34⁺ numbers was observed after thawing (Figure 2c), even when dead cells were excluded using 7-AAD staining in accordance with the ISHAGE protocol (group I: 6.02.9 10⁶ per kg vs group II: 4.22.7 10⁶ per kg). In contrast, when we calculated numbers of SSC^loALDH^br cells transplanted per kg (Figure 2c), we found significantly lower numbers in group II as compared to group I (group I: 5.22.6 10⁶ per kg vs group II: 2.00.8 10⁶ per kg; P<0.01). Importantly, those lower SSC^loALDH^br cell numbers in the nine transplants from group II correlated well (r=0.66; P<0.05, Pearson's correlation) with the observed decrease of CFU-GM potential in that group (group I: 9.11.6 10⁵ per kg vs group II: 1.81.1 10⁵ per kg, P<0.05; Figure 2d). It is noteworthy that seven of the nine patients who received decreased SSC^loALDH^br cell numbers (group II, marked bold in Table 2) had no engraftment. Two engrafted patients in group II revealed only mixed chimaerism and suffered an early relapse. In contrast, all other 22 patients (group I, not bold in Table 2) showed leucocyte engraftment at day 15.93.4 and 19 of them reached full donor chimaerism.

Discussion

Allogeneic grafts cryopreserved due to medical and/or logistic reasons have successfully been used for transplantation.^{1, 2, 3, 4, 5, 6, 7} In a retrospective study we analysed functionality (ALDH activity and CFU) vs CD34 positivity and MI in 31 consecutive allogeneic PBSC and 8 allogeneic BM grafts frozen in our hospital between 2000 and 2007. Clinical data including leucocyte engraftment and chimaerism kinetics were available for 33 PBSC and 13 BM patients transplanted with those grafts.

No loss of stem cell potential was seen in frozen/thawed BM samples. In line, all patients showed early and sustained engraftment after transplantation of these grafts. In contrast, based on the ratios of SSC^loALDH^br to CD34 numbers frozen/thawed PBSC grafts could be divided into two groups: those grafts with a close to normal ratio (>0.5) and those with a decreased (<0.5) ratio. We found that in all cases (nine out of nine) a ratio below 0.5 (that is decreased SSC^loALDH^br as compared to CD34 numbers) was associated with engraftment problems: either engraftment failure (seven out of nine) or permanent absence (two out of nine) of full haematopoietic donor chimaerism after engraftment. Absolute numbers of transplanted ALDH-expressing cells as well as transplanted CFU numbers per kg for these 9 patients were significantly decreased as compared to the other 22 transplant recipients, who showed well-timed leucocyte engraftment (22 out of 22) with establishment of full haematopoietic donor chimaerism in 19 of 22 patients (86%). Importantly, no differences in mean CD34 numbers as determined before cryopreservation and after thawing were observed between these groups. These results are in line with our previous data that only assessment of HSC functionality such as CFU and ALDH activity, but not phenotypic markers (CD34) may help in predicting the engraftment kinetics and the risk of graft failure.⁸

Thus, in our study we observed significant losses of MI, functionality (low ratio ALDH^br/CD34⁺) and in vitro growth potential (CFU) after cryopreservation for PBSC but not BM allografts. This correlates well with our earlier finding that BM preparations are more resistant than PBSC grafts to physiological stress such as storage at room temperature for prolonged periods of time (at least 72 h).⁹

Overall, 9 of the 33 patients (27%) who received cryopreserved PBSC allografts did not reach engraftment. For comparison, primary graft failure was observed in our centre in only 7 out of 493 recipients (1.4%) of fresh allogeneic grafts. Moreover, we did not see any graft failure in either the 129 patients who received frozen autologous transplants or the 13 recipients of cryopreserved BM during the same period (2000–2007). In contrast, all these patients showed well-timed leucocyte engraftment. This striking difference may be due to several factors: (1) allogeneic PBSC preparations contain significantly (P<0.01) higher concentrations of RBC and PLT compared to both autologous PBSC aphaeresis products and allogeneic BM grafts as well as significantly (P<0.001) higher concentrations of WBC in comparison with allogeneic BM grafts (Figure 3). RBC and granulocyte contaminations may affect the stability of HSC during cryopreservation, whereas platelets may promote clotting after thawing resulting in a reduced haematopoietic potential of the grafts; (2) since all nine PBSC allografts with decreased numbers of SSC^loALDH^br were collected from unrelated donors at outside facilities, it cannot be excluded that transportation from the donation centre in combination with delayed cryopreservation involved too much 'stress' for HSC and progenitors; (3) in principle it might also be possible that our investigation has been biased due to a pre-selection of patients receiving frozen grafts. However, given our results with BM and in the PBSC group with a high ALDH/CD34 ratio this possibility seems not very likely; (4) it has to be noted that in eight out of nine cases engraftment failure was observed after RIC. Thus, impaired engraftment potential may be particularly deleterious in the context of RIC regimens; (5) of course, it must be emphasized that other parameters such as frequency of GVHD might be different for frozen vs unfrozen grafts and as well influence the period of haematological recovery after SCT. Therefore, further studies need to be performed to clarify the influence of all variations regarding pre-freeze and frozen storage conditions on the effectiveness of HSC grafts and (6) finally, this retrospective study compares limited numbers of heterogeneous patients in different cohorts with regard to disease type, disease stage, conditioning regimens and other factors, for example degree of HLA matching between the donors and recipients. Therefore, we suppose that the influence of all parameters needs to be investigated in a larger set of data.

Figure 3.

Quantitative comparison of different blood components in cryopreserved BM and autologous as well as allogeneic PBSC preparations. Allogeneic PBSC (PBSC-allo) preparations (n=31) contain significantly higher concentrations of (a) RBC and (b) platelets (PLT) as compared to both autologous PBSC (PBSC-auto, n=19) and allogeneic BM (BM-allo, n=8) grafts. (c) Also, the concentration of WBC is significantly higher in comparison with BM grafts.

Full figure and legend (84K)

In conclusion, we suggest that cryopreservation of BM influences neither the quality of grafts nor the engraftment kinetics after BMT, whereas the use of transported and frozen/thawed allogeneic PBSC might be associated with a higher risk of graft failure, in particular in the context of RIC regimens and should therefore be avoided until more valid data are available. If cryopreservation of PBSC allografts cannot be circumvented, the assessment of HSC functionality (for example AldeFluor) may help predicting the risk of a potential engraftment failure.

References

1. Frey NV, Lazarus HM, Goldstein SC. Has allogeneic stem cell cryopreservation been given the 'cold shoulder'? An analysis of the pros and cons of using frozen versus fresh stem cell products in allogeneic stem cell transplantation. Bone Marrow Transplant 2006; 38: 399–405. | Article | PubMed | ChemPort |
2. Lasky LC, Van Buren N, Weisdorf DJ, Filipovich A, McGlave P, Kersey JH et al. Successful allogeneic cryopreserved marrow transplantation. Transfusion 1989; 29: 182–184. | Article | PubMed | ChemPort |
3. Eckardt JR, Roodman GD, Boldt DH, Clark GM, Alvarez R, Page C et al. Comparison of engraftment and acute GVHD in patients undergoing cryopreserved or fresh allogeneic BMT. Bone Marrow Transplant 1993; 11: 125–131. | PubMed | ChemPort |
4. Stockschläder M, Krüger W, Kroschke G, Zeller W, Hoffknecht M, Löliger C et al. Use of cryopreserved bone marrow in allogeneic bone marrow transplantation. Bone Marrow Transplant 1995; 15: 569–572. | PubMed | ChemPort |
5. Stockschläder M, Krüger W, tom Dieck A, Horstmann M, Altnöder M, Löliger C et al. Use of cryopreserved bone marrow in unrelated allogeneic transplantation. Bone Marrow Transplant 1996; 17: 197–199. | PubMed | ChemPort |
6. Stockschläder M, Hassan HA, Krog C, Krüger W, Löliger C, Horstman M et al. Long-term follow-up of leukaemia patients after related cryopreserved allogeneic bone marrow transplantation. Br J Haematol 1997; 96: 382–386. | Article | PubMed | ChemPort |
7. Shinkoda Y, Ijichi O, Tanabe T, Ishikawa S, Kamitamari A, Nishikawa T et al. Identical reconstitution after bone marrow transplantation in twins who received fresh and cryopreserved grafts harvested at the same time from their older brother. Clin Transplant 2004; 18: 743–747. | Article | PubMed |
8. Lioznov M, Ikogho R, Fehse B, Bacher U, Kröger N, Zander AR. Factors predicting haematological reconstitution following haemopoietic stem cell transplantation. Bone Marrow Transplant 2008 (in press).
9. Lioznov M, Freiberger P, Kröger N, Zander AR, Fehse B. Aldehyde dehydrogenase activity as a marker for the quality of hematopoietic stem cell transplants. Bone Marrow Transplant 2005; 35: 909–914. | Article | PubMed | ISI | ChemPort |
10. Sputtek A, Sputtek R. Cryopreservation in transfusion medicine and hematology.In: Fuller BJ, Lane N, Benson EE (eds). Life in the Frozen State. CRC Press: Boca Raton, 2004, pp 483–504.
11. Sputtek A, Jetter S, Hummel K, Kühnl P. Cryopreservation of peripheral blood progenitor cells: characteristics of suitable techniques. Beitr Infusionsther Transfusionsmed 1997; 34: 79–83. | PubMed | ChemPort |
12. Fehse N, Fehse B, Kröger N, Zabelina T, Freiberger P, Krüger W et al. Influence of anti-thymocyte globulin as part of the conditioning regimen on immune reconstitution following matched related bone marrow transplantation. J Hematother Stem Cell Res 2003; 12: 237–242. | Article | PubMed | ChemPort |
13. Philpott NJ, Turner AJ, Scopes J, Westby M, Marsh JC, Gordon-Smith EC et al. The use of 7-amino actinomycin D in identifying apoptosis: simplicity of use and broad spectrum of application compared with other techniques. Blood 1996; 87: 2244–2251. | PubMed | ISI | ChemPort |
14. Fehse B, Uhde A, Fehse N, Eckert HG, Clausen J, Rüger R et al. Selective immunoaffinity-based enrichment of CD34⁺ cells transduced with retroviral vectors containing an intracytoplasmatically truncated version of the human low-affinity nerve growth factor receptor (LNGFR) gene. Human Gene Ther 1997; 8: 1815–1824. | Article | ChemPort |

Acknowledgements

We thank P Freiberger, U Fritzsche-Friedland, E Merle and M Reckhaus for the excellent technical assistance, and G Amtsfeld and J Hagelberg for their help with clinical data.