Dissertation committee:
Κωνσταντίνος Κωνσταντόπουλος, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Παναγιώτης Παναγιωτίδης, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Μαρία Αγγελοπούλου, Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Θεόδωρος Βασιλακόπουλος,Αναπληρωτής Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Μαρίνα Σιακαντάρη, Αναπληρώτρια Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Φλώρα Κοντοπίδου, Αναπληρώτρια Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Στυλιανή Κοκόρη, Επίκουρη Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Summary:
Background: Rapid and complete hematologic reconstitution, following autologous hematopoietic stem cell (HSC) transplantation (Auto-HSCT), serves as an in vivo quality control for the clinical and laboratory Department of a Transplantation center. The decline of the cryopreserved products’ viability necessitates a standardized viability assessment procedure with results that consistently correlate with patients’ reconstitution kinetics.
Purpose: The purpose of this study was a) the investigation of the effect of cryopreservation on viability and apoptosis of CD34+ cells and their effect on the kinetics of hematologic reconstitution, β) the discrimination of the safest CD34+ viability and “true” viability assessment method between the two most widely used methods, and γ) the definition of the viable and viable non apoptotic CD34+ cells threshold dose for an ideal hematological reconstitution following auto-HSCT.
Material and Methods: Cryopreserved peripheral blood stem cell (PBSC) samples of 61 patients with hematologic malignancies - multiple myeloma (MM), Hodgkin lymphoma (HL), and non-Hodgkin lymphoma )NHL) - who had already undergone auto-HSCT, were retrospectively studied.
As far as cell viability is concerned, the samples were analyzed by two methods, namely: a) the Trypan Blue exclusion method (TB) and b) flow cytometry (FACS). The concurrent use of 7-AAD and Annexin-V allowed the discrimination between viable (7-AAD-/Annexin-V-), early apoptotic (7-AAD-/Annexin-V+) and necrotic cells (7-AAD+/Annexin-V+). Viable cells were defined as those not stained with Trypan Blue (method a) and those that were negative for 7-AAD (method b). True viable (viable non-apoptotic) cells were defined as those which were negative both for 7-ADD and Annexin-V.
As far as the effect on viability and “true” viability of CD34+ cells is concerned, patients’ characteristics, mobilization parameters and cryopreserved products’ characteristics, were investigated with univariate and multivariate analysis (Table 1). As far as hematologic reconstitution kinetics are concerned, the following characteristics were investigated by univariate analysis: gender, age, disease, disease status at HSC collection, type of high dose therapy, mobilization kinetics, total volume infused product, total number of thawed bags, CD34+/Kg per bag, cryopreservation time-length, total number of CD34+/kg (as measured in the fresh product) and total number of viable and viable non apoptotic (true viable) CD34+/Kg infused. The viability of PB CD34+ cells on collection day, the viability of CD34+ cells in the fresh product and platelet (PLT) final concentration in the cryopreservation bag were also includes as possible prognostic factors for ideal hematologic reconstitution in univariate and multivariate analysis.
“Planned” dose was defined as the dose of CD34+/Kg (as measured in the fresh product) that was infused. The day of short-term reconstitution was the first day following infusion, on which neutrophils exceeded 500/μl and the first day after three consecutive days (without transfusion) on which PLT exceeded 20×103/μl. Long-term reconstitution was achieved when neutrophils exceeded 1500/μl (the latest on D90) and 2000/μl (the latest on D180) and PLT exceed ≥140×103/μl (the latest on D180). Patients achieving ideal hematological reconstitution were defined as those having neutrophil counts ≥500/μl until D12 and ≥1500/μl until D30, and simultaneously, reaching PLT counts ≥20x103/μl until D14 and ≥140x103/μl the latest until D90.
Results: Among the 61 samples analyzed, the median value of: a) viable nucleated cells (TNC) was 55,3% (range 10-73%, measured by TB), b) viable CD45+ cells was 61,7% (range 7,5-81,4%, measured by FACS), c) viable CD34+ cells was 45,9% (range 0,6-88%), d) apoptotic CD45+ cells was 8,4% (range 0,5-34,3%), e) apoptotic CD34+ cells was 4,3% (range 0-44,9%), f) viable non apoptotic CD45+ cells was 52,9% (range 5,4-80,9%) and g) viable non apoptotic CD34+ cells was 32,2% (range 0,5-79,5%). The cryopreserved CD34+ cells showed a statistical significant lower viability than CD45+ cells (median difference -12,6%, p<0,0005). ΤΒ method overestimated the % CD34+ cell viability in 70,3% of the samples.
As far as viability and “true” viability of CD34+ cells is concerned, 9 variables were found statistically significant in univariate analysis (table 1). In multivariate analysis only the following 3 remained statistically significant: a) PB CD34+ cell viability on collection day (p=0,020 for viability and p=0,036, for “true” viability), b) CD34+ cell viability in the fresh product (p=0,024 και p=0,006, respectively) and c) PLT final concentration in the cryopreservation bag ( 0,001 και p=0,004, respectively).
As far as hematologic reconstitution is concerned, the univariate analysis disclosed that, with a statistical significance: a) age >53 years was associated with slower restoration of neutropenia {D11 vs D10} and thrombocytopenia {D14 vs D13} and with delayed long-term PLT reconstitution { 66,7% of patients >53 years vs 90,3% of those ≤53 years}, b) in patients with ΜΜ, restoration of neutropenia was delayed {D11 (MM) vs D9 (LH) and D10 (NHL)}, while in patients with HL restoration of thrombocytopenia was faster {D11 (LH), vs D14 (MM), and D15 (NHL)}, c) very good mobilizers were associated with faster short-term reconstitution {D12 (very good) D14 (good) και D15 (poor)} and long-term reconstitution of platelets {96.6% of very good mobilizers vs 62.5% of the remaining}, c) the infusion of >360ml of cell product was associated with delayed short-term {D13 vs D15} as well as long-term PLT reconstitution { 56% vs 94,4% of those who received ≤360ml}, d) the infusion of >3 bags of cell product was associated with delayed long-term PLT reconstitution { 56,3% vs 86,7% of those who were infused with ≤3}, and e) the infusion ≤2,4×106/Kg CD34+ cells per bag was associated with delayed short-term PLT reconstitution {D15 vs D12} and with delayed long-term hematologic reconstitution { 62,1% vs 96,3% of those who were infused with>2,4×106/Kg per bag}, f) the “planned” dose of CD34+/Kg, was significantly associated only with short-term PLT reconstitution {D18 (if <3×106), D15 (if 3-6×106) και D13 (if >6×106}, g) the dose of viable and viable non apoptotic CD34+/Kg infused, was statistically significantly associated with short-term as well as long-term hematologic reconstitution.
Ideal hematologic reconstitution – as defined in methods - was achieved by 32 patients (52,5%). In univariate analysis the following parameters were found as statistically significant for ideal hematological reconstitution achievement: age, disease, mobilization kinetics, peripheral bloodCD34+ cell viability on collection day, fresh product CD34+ cell viability, total volume of thawed product, CD34+/Kg per bag, PLT final concentration in the cryopreservation bag and the dose of viable and viable non-apoptotic CD34+/Kg infused. In multivariate analysis, only higher doses of viable or viable non-apoptotic CD34+/Kg infused were associated with statistically and clinically significant higher probability for ideal hematological reconstitution. The threshold dose for ideal hematological reconstitution achievement was inferred as 6,9x106 CD34+/Kg (measured in fresh product), as 3,4x106 viable CD34+/Kg and as 2,9x106 viable non apoptotic CD34+/Kg.
The reduction of viable non-apoptotic CD34+ cells in the cryopreserved products had as a consequence the dramatic decline of “true” doses of CD34+ cells infused, compared to “planed” doses. Specifically, patients infused with “planed” doses of D34+/Kg: a) <3x106 were actually infused with 0,9x106 viable non-apoptotic CD34+/Kg (median value), b) 3-6x106 were actually infused with 1,1x106 viable non-apoptotic CD34+/Kg and c) >6x106 CD34+/Kg were actually infused with 3,1x106 viable non-apoptotic CD34+/Kg.
Conclusions: The significant association of patients’ and cryopreserved products’ characteristics with the viability of thawed CD34+ cells, largely explains the observed variation of viability among different samples. Thus, these parameters should be taken into account during the cryopreservation procedure (product dilution and apportionment in the cryopreservation bags). Moreover, “prognostic indicators of viability” known before, during and after the HSC collection procedure, may direct the decision to extend the collection, aiming to a higher dose of CD34+/Kg or to test the cell viability of the frozen product prior to HSCT. Rapid and complete hematologic reconstitution in HSCT is also significantly affected by patients’ and cryopreserved products’ characteristics. The number of CD34+/Kg, as counted in the fresh product, is only statistically significantly associated with platelets’ short-term reconstitution. On the contrary, the number of viable and viable non-apoptotic CD34+/Kg infused proved as the only statistically and clinically significant factor for ideal hematologic reconstitution. Therefore, it seems preferable to target the dose of CD34+/Kg that ensures an ideal hematologic reconstitution, which in turn is accurately represented by the viable non-apoptotic CD34+ cells, rather than the number of CD34+ cells counted from freshly collected product. Viability assessment of the cryopreserved product to be infused is not necessary under ideal conditions of cryopreservation and thawing. However, it is necessary, when the number of viable CD34+/Kg is expected lower than the threshold for ideal hematological reconstitution, based upon PB CD34+ cell viability, Cd34+ cell viability in the fresh product and PLT concentration in the cryopreservation bag. While cryopreserved CD34+ cells’ apoptosis leads to further CD34/Kg dose reduction, and it can’t be predicted, it should be co-evaluated during the product’s viability estimation. Trypan blue exclusion method is only recommended as a rapid alternative at the first step of viability assessment process, which is safer to be accompanied by FACS, especially for products with low number of CD34+/Kg.
