Vox Sanguinis (2015) 109, 375–386 © 2015 International Society of Blood Transfusion DOI: 10.1111/vox.12289
ORIGINAL PAPER
T-cell subsets in autologous and allogeneic peripheral blood stem cell concentrates J. Strobel, I. Moellmer, J. Zingsem, B. Hauck-Dlimi, R. Eckstein & E. Strasser Department of Transfusion Medicine and Hemostaseology, Friedrich-Alexander-University, Erlangen, Germany
Background and Objectives Regulatory T cells (Tregs) and other T-cell subsets are of importance in the setting of autologous and allogeneic stem cell transplantations. We conducted a study to assess the content of peripheral blood stem cell concentrates and related apheresis parameters in the autologous and allogeneic setting. Material and Methods We characterized 53 donors, patients and peripheral blood stem cell concentrates (PBSC) regarding the content of CD45+ cells, lymphocytes, CD3+ cells, CD3+ CD4+ T cells, CD3+ CD4+ CD25+ T cells, CD3+ CD4+ CD25+ CD127low/negative Tregs and CD34+ cells and calculated cell yields, recruitment factors and collection efficiency for all cell types. We compared allogeneic data with autologous data. Results Autologous PBSC show significantly lower concentrations of T-cell subsets compared to allogeneic PBSC (17 112/ll CD4+, 14 858/ll CD4+ CD25+ and 1579/ll CD3+ CD4+ CD25+ CD127low/negative Tregs in autologous compared to 65 539/ll CD4+, 44 208+/ll CD4+ CD25+ and 5040/ll CD3+ CD4+ CD25+ CD127low/negative Tregs in allogeneic PBSC, respectively), in contrast to CD34+ concentrations (5342/ ll CD34+ in autologous compared to 2367/ll CD34+ in allogeneic PBSC, respectively). Accordantly, all T-cell yields are lower in the autologous setting compared to allogeneic PBSC. However, recruitment factor and collection efficiency of all cell types are higher in autologous compared to allogeneic PBSC, but not all parameters differ significantly when groups are compared.
Received: 20 October 2014, revised 24 February 2015, accepted 27 March 2015, published online 3 June 2015
Conclusion T-cell subsets and especially Tregs are a substantial part of PBSC transplantation, as considerable recruitment during apheresis occurs. In large volume apheresis, the collection efficiency of Treg is comparable to that of CD34+ cells, while recruitment factors are even higher. Key words: apheresis, peripheral blood stem cells, regulatory T cells, T-cell subsets.
Introduction T lymphocytes and especially regulatory T cells (Tregs) are of importance regarding graft-versus-host disease (GvHD) [1], graft-versus-leukaemia effect [2] and a condition called ‘autologous graft-versus-host disease’ [3, 4] in the setting of stem cell transplantation (SCT). It has been Correspondence: Julian Strobel, Transfusionsmedizinische und Haemostaseologische Abteilung, Universit€atsklinikum Erlangen, Krankenhausstrasse 12, D-91054 Erlangen, Germany E-mail: [email protected]
demonstrated that Tregs are capable of suppressing acute GvHD in mice [5, 6]. Furthermore, Tregs seem to be potent inhibitors of an antitumour immune response [7]. Allogeneic GvHD is caused by tissue damage due to the conditioning regimen, the resulting release of inflammatory cytokines, and activation of antigen presenting cells and consecutively activated donor T cells [8]. It has been shown that T-cell depletion helps preventing GvHD, but this procedure might increase the risk of relapse, probably by reducing the GvM effect [9, 10]. High doses of Treg might be able to prevent or stop allogeneic GvHD [5]. Autologous GvHD is caused by the combination of
375
376 J. Strobe et al.
peripheral Treg depletion due to chemotherapy and CD8+ T cells autoreactive for a major histocompatibility complex class II invariant chain peptide [3]. It has also been proposed that Treg might be responsible for a reduced rate of autologous GvHD [11] and that this might be an explanation for reduced GvM effects in autologous transplantation. However, scientists are confronted with the problem of Treg identification, quantification and maybe selection while, at the same time, maintaining all functional properties of Treg. It has been shown that Foxp3 expression is a good marker in CD4+ CD25+ T cells for Treg identification [12], but this marker cannot be used to isolate viable Treg as cells need to be fixed for Foxp3 staining. Furthermore, not all cells expressing Foxp3 are Tregs [13]. It was recently shown that CD4+ CD25hi CD127low/negative T cells characterize a majority of Treg [14, 15]. Although not all Treg subsets might be identified using those markers in flow cytometry [16], this could be a feasible method of maintaining functional properties of Treg and, at the same time, collect a majority of Treg subpopulations in cell sorting [17]. Methods in accordance with good manufacturing practice for Treg isolation are currently still based on selection of CD25+ CD127low/negative cells [18]. Therefore, it is of interest to know the content of T cells, subsets of T cells and Tregs in stem cell concentrates, especially if T-cell depletion or Treg expansion is to be performed. In 2013, a vast majority of transplants listed by the NMDP was produced via stem cell apheresis from peripheral blood [19]. We characterized peripheral blood stem cell concentrates (PBSC) regarding their cell contents and compared the apheresis parameters cell yield, recruitment factor and collection efficiency between healthy donors in allogeneic and patients in autologous SCT. To our knowledge, this is the first study presenting such data.
Material and methods Peripheral blood stem cell were collected with the Cobe Spectra (Terumo BCT, Lakewood, CO). Patients were to be collected as soon as they had a minimum of 10 CD34+ cells per ll. Only one patient did not reach 10 cells/ll. All patients received G-CSF prior to collection. Only two patients additionally received plerixafor. The patients’ chemotherapy regiments prior to collection are presented in Table 1. The allogeneic donors were mobilized with G-CSF only and were collected on day 5 of G-CSF collection regardless of the CD34+ cell count. All apheresis procedures were planned to last for 300 min with a target of approximately 35–4 total blood volumes processed, the last being dependent on the venous flow possible. We characterized the PBSC regarding the content of CD45+
Table 1 Mobilization chemotherapy of patient groups Mobilization chemotherapy
Auto
Cyclophosphamide Cyclophosphamide and epirubicin Cyclophosphamide and etoposide Cytarabine and thiotepa G-CSF Ifosfamid, epirubicin and etoposide Dasatinib, dexamethason, methotrexat and cytarabine Rituximab, cyclophosphamide, hydroxydaunorubicin, vincristin and prednisolon Rituximab, dexamethasone, high-dose cytarabine and cisplatin Rituximab, ifosfamid, carboplatin and etoposide Rituximab, ifosfamid, episubicin and etoposide Vincristin, ifosfamid, hydroxydaunorubicin and etoposide Dexamethasone, high-dose cytarabine and cisplatin
1 1
1
6
6 (19 Plerixafor)
1 1 5
Lymphoma
Myeloma
1 5
1 2
2 (19 Plerixafor)
6
6
7
7
1
1
6
2
2
cells, lymphocytes, CD3+ T cells, CD3+ CD4+ T cells, CD4+ CD25+ T cells, CD4+ CD25+ CD127low/negative T cells and CD34+ cells using flow cytometry (FACS Calibur, BD, Heidelberg, Germany). These 53 PBSC were gathered from 13 allogeneic donors and 40 patients. In the patient group, 18 patients suffered from lymphoma, 13 from myeloma, eight from solid tumours (six Ewing sarcoma, one mamma carcinoma and one hepatoblastoma) and one patient from leukaemia. We compared the group of healthy donors (‘allo’) with the group of all patients combined (‘auto’), the lymphoma (‘lymphoma’) and the myeloma (‘myeloma’) group, respectively. We also compared the lymphoma with the myeloma group. We deemed these two latter groups to be of interest due to the controversial roles of Treg reported in those diseases: In most lymphoma disease entities, high numbers of Treg found in the tumour environment are considered to be of positive effect. In myeloma, Tregs have been described to be either reduced or elevated. Overall, Treg might be relevant for myeloma development by preventing immunity of the host [20]. Prior to apheresis, all patients and donors had given informed consent to these quality control measurements from their blood samples. No clinical information except for the main diagnosis, mobilizing chemotherapy, © 2015 International Society of Blood Transfusion Vox Sanguinis (2015) 109, 375–386
T-cell subsets in stem cell apheresis 377
which contains Trucount beads for quantification. In an additional experiment (Fig. 1a–d), lymphocytes expressing CD3 were considered to be T cells (TC) and T-helper cells (THC) in case of expressing CD3 and CD4. Although CD4+ T cells with high expression of CD25 have formerly been described as being Treg [21], we only considered CD25+ cells to be Tregs if CD127 expression was low while CD3, CD4 and CD25 expression was high, due to the inverse correlation of Foxp3 and CD127 expression described recently [14, 15]. Such phenotyping strategy will not
age, weight and sex was included in this study, and no information on Treg content of the samples was issued to the wards. No exceptional phlebotomies were necessary for all analyses, as we used the same samples routinely needed for measuring the content of CD34+ cells in a sample pre- and postapheresis and the PBSC. CellQuest software (V 6.0, BD, Heidelberg, Germany) was used for data acquisition and analysis. CD45+ cells, CD34+ cells and lymphocytes were quantified according to the ISHAGE protocol with the stem cell enumeration kit (BD),
(a)
(c)
(b)
(d)
Fig. 1 Gating strategy of T-cell subsets. (a) Gating of lymphocytes. (b) Gating of CD3+ and CD3+ CD4+ double-positive cells out of lymphocyte gate. (c) Gating of CD3+ and CD3+ CD25+ double-positive cells out of lymphocyte gate. (d) Gating of CD25+ CD127 low cells out of CD3+ CD4+ double-positive gate (upper right quadrant, Fig. 1b).
© 2015 International Society of Blood Transfusion Vox Sanguinis (2015) 109, 375–386
378 J. Strobe et al.
include all Treg populations and is a limitation to the data set discussed later on. The antibodies used where mouse anti-CD3 labelled with the fluorochromes peridinin chlorophyll (PerCP), mouse anti-CD4 labelled with fluorescein isothiocyanate (FITC), mouse anti-CD25 labelled with R-phycoerythrin (PE) and mouse anti-CD127 labelled with allophycocyanin (APC) (all BD Pharmingen, Heidelberg, Germany). The samples were diluted with PBS to a concentration of approximately 10000 CD45+ cells per ll to analyse samples with a comparable cell to antibody ratio. We calculated collection efficiency (CE), cell yield and recruitment factor (RF) as follows: Yield = Cell concentration in concentrate [per ml] volume of concentrate (ml) Recruitment factor ¼
Postapheresis cells þ Yield Preapheresis cells
*Pre- and postapheresis cells: absolute number of cells in peripheral blood preand postapheresis = preor postapheresis blood cell count [9109/l] 9 total blood volume of subject. Collectionefficiency¼ 100
Yield : Total blood volume (ml)Preapheresis cell count (cells/ml)
For the calculation of yield per kg body weight in the allogeneic group, we used the weight of both the recipients and the donors. This is due to the fact that one aim of this pilot study was to find out whether healthy donors’ mobilization of T-cell subsets differs from that of patients. The recipients’ weight cannot be used for the calculation of recruitment, as they are not involved in the apheresis procedure. However, as yield/kg is of great clinical interest regarding the recipient’s body weight, we included those values as well. Group comparisons were also calculated with the values for recipients’ body weight. As in stem cell apheresis, the blood volume processed is exceedingly larger than in platelet or granulocyte apheresis (3–5 9 TBV vs. 06–1 9 TBV), the formula as used for calculating the CE in plateletpheresis [CE apheresis = Yield 9 100%/(processed blood volume 9 05 9 (pre- + postdonation cell count))] is not suitable. As soon as more blood volume is processed, the above-named formula underestimates the collection efficiency. This is due to the fact that by this formula the donor is seen as ‘open system’, in which much more cells are assumed to circulate than is really the case, due to the usage of the blood volume processed. However, in large volume apheresis, cells may be mobilized from
extravasal resources (e.g. bone marrow). As described previously [22, 23], we therefore referred only to the preapheresis cell counts for calculating the CE. Thus, the collection efficiency has been defined as the percentage of harvested cells in relation to the circulating cells before apheresis.
Statistical analysis Statistical analysis was performed with SPSS for Windows (IBM, Chicago, IL, USA, Version 20.0) and R 3.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Normal distribution of variables was tested with the Shapiro–Wilks test. Group comparisons consider the comparison between donors (‘Allo’-group) and patients (‘Auto’-group) and between donors and the two groups ‘lymphoma’ and ‘myeloma’. For group comparisons, the U-test was used for data not distributed normally, and the t-test was performed for normally distributed data. Differences were considered significant for P-values
ORIGINAL PAPER
T-cell subsets in autologous and allogeneic peripheral blood stem cell concentrates J. Strobel, I. Moellmer, J. Zingsem, B. Hauck-Dlimi, R. Eckstein & E. Strasser Department of Transfusion Medicine and Hemostaseology, Friedrich-Alexander-University, Erlangen, Germany
Background and Objectives Regulatory T cells (Tregs) and other T-cell subsets are of importance in the setting of autologous and allogeneic stem cell transplantations. We conducted a study to assess the content of peripheral blood stem cell concentrates and related apheresis parameters in the autologous and allogeneic setting. Material and Methods We characterized 53 donors, patients and peripheral blood stem cell concentrates (PBSC) regarding the content of CD45+ cells, lymphocytes, CD3+ cells, CD3+ CD4+ T cells, CD3+ CD4+ CD25+ T cells, CD3+ CD4+ CD25+ CD127low/negative Tregs and CD34+ cells and calculated cell yields, recruitment factors and collection efficiency for all cell types. We compared allogeneic data with autologous data. Results Autologous PBSC show significantly lower concentrations of T-cell subsets compared to allogeneic PBSC (17 112/ll CD4+, 14 858/ll CD4+ CD25+ and 1579/ll CD3+ CD4+ CD25+ CD127low/negative Tregs in autologous compared to 65 539/ll CD4+, 44 208+/ll CD4+ CD25+ and 5040/ll CD3+ CD4+ CD25+ CD127low/negative Tregs in allogeneic PBSC, respectively), in contrast to CD34+ concentrations (5342/ ll CD34+ in autologous compared to 2367/ll CD34+ in allogeneic PBSC, respectively). Accordantly, all T-cell yields are lower in the autologous setting compared to allogeneic PBSC. However, recruitment factor and collection efficiency of all cell types are higher in autologous compared to allogeneic PBSC, but not all parameters differ significantly when groups are compared.
Received: 20 October 2014, revised 24 February 2015, accepted 27 March 2015, published online 3 June 2015
Conclusion T-cell subsets and especially Tregs are a substantial part of PBSC transplantation, as considerable recruitment during apheresis occurs. In large volume apheresis, the collection efficiency of Treg is comparable to that of CD34+ cells, while recruitment factors are even higher. Key words: apheresis, peripheral blood stem cells, regulatory T cells, T-cell subsets.
Introduction T lymphocytes and especially regulatory T cells (Tregs) are of importance regarding graft-versus-host disease (GvHD) [1], graft-versus-leukaemia effect [2] and a condition called ‘autologous graft-versus-host disease’ [3, 4] in the setting of stem cell transplantation (SCT). It has been Correspondence: Julian Strobel, Transfusionsmedizinische und Haemostaseologische Abteilung, Universit€atsklinikum Erlangen, Krankenhausstrasse 12, D-91054 Erlangen, Germany E-mail: [email protected]
demonstrated that Tregs are capable of suppressing acute GvHD in mice [5, 6]. Furthermore, Tregs seem to be potent inhibitors of an antitumour immune response [7]. Allogeneic GvHD is caused by tissue damage due to the conditioning regimen, the resulting release of inflammatory cytokines, and activation of antigen presenting cells and consecutively activated donor T cells [8]. It has been shown that T-cell depletion helps preventing GvHD, but this procedure might increase the risk of relapse, probably by reducing the GvM effect [9, 10]. High doses of Treg might be able to prevent or stop allogeneic GvHD [5]. Autologous GvHD is caused by the combination of
375
376 J. Strobe et al.
peripheral Treg depletion due to chemotherapy and CD8+ T cells autoreactive for a major histocompatibility complex class II invariant chain peptide [3]. It has also been proposed that Treg might be responsible for a reduced rate of autologous GvHD [11] and that this might be an explanation for reduced GvM effects in autologous transplantation. However, scientists are confronted with the problem of Treg identification, quantification and maybe selection while, at the same time, maintaining all functional properties of Treg. It has been shown that Foxp3 expression is a good marker in CD4+ CD25+ T cells for Treg identification [12], but this marker cannot be used to isolate viable Treg as cells need to be fixed for Foxp3 staining. Furthermore, not all cells expressing Foxp3 are Tregs [13]. It was recently shown that CD4+ CD25hi CD127low/negative T cells characterize a majority of Treg [14, 15]. Although not all Treg subsets might be identified using those markers in flow cytometry [16], this could be a feasible method of maintaining functional properties of Treg and, at the same time, collect a majority of Treg subpopulations in cell sorting [17]. Methods in accordance with good manufacturing practice for Treg isolation are currently still based on selection of CD25+ CD127low/negative cells [18]. Therefore, it is of interest to know the content of T cells, subsets of T cells and Tregs in stem cell concentrates, especially if T-cell depletion or Treg expansion is to be performed. In 2013, a vast majority of transplants listed by the NMDP was produced via stem cell apheresis from peripheral blood [19]. We characterized peripheral blood stem cell concentrates (PBSC) regarding their cell contents and compared the apheresis parameters cell yield, recruitment factor and collection efficiency between healthy donors in allogeneic and patients in autologous SCT. To our knowledge, this is the first study presenting such data.
Material and methods Peripheral blood stem cell were collected with the Cobe Spectra (Terumo BCT, Lakewood, CO). Patients were to be collected as soon as they had a minimum of 10 CD34+ cells per ll. Only one patient did not reach 10 cells/ll. All patients received G-CSF prior to collection. Only two patients additionally received plerixafor. The patients’ chemotherapy regiments prior to collection are presented in Table 1. The allogeneic donors were mobilized with G-CSF only and were collected on day 5 of G-CSF collection regardless of the CD34+ cell count. All apheresis procedures were planned to last for 300 min with a target of approximately 35–4 total blood volumes processed, the last being dependent on the venous flow possible. We characterized the PBSC regarding the content of CD45+
Table 1 Mobilization chemotherapy of patient groups Mobilization chemotherapy
Auto
Cyclophosphamide Cyclophosphamide and epirubicin Cyclophosphamide and etoposide Cytarabine and thiotepa G-CSF Ifosfamid, epirubicin and etoposide Dasatinib, dexamethason, methotrexat and cytarabine Rituximab, cyclophosphamide, hydroxydaunorubicin, vincristin and prednisolon Rituximab, dexamethasone, high-dose cytarabine and cisplatin Rituximab, ifosfamid, carboplatin and etoposide Rituximab, ifosfamid, episubicin and etoposide Vincristin, ifosfamid, hydroxydaunorubicin and etoposide Dexamethasone, high-dose cytarabine and cisplatin
1 1
1
6
6 (19 Plerixafor)
1 1 5
Lymphoma
Myeloma
1 5
1 2
2 (19 Plerixafor)
6
6
7
7
1
1
6
2
2
cells, lymphocytes, CD3+ T cells, CD3+ CD4+ T cells, CD4+ CD25+ T cells, CD4+ CD25+ CD127low/negative T cells and CD34+ cells using flow cytometry (FACS Calibur, BD, Heidelberg, Germany). These 53 PBSC were gathered from 13 allogeneic donors and 40 patients. In the patient group, 18 patients suffered from lymphoma, 13 from myeloma, eight from solid tumours (six Ewing sarcoma, one mamma carcinoma and one hepatoblastoma) and one patient from leukaemia. We compared the group of healthy donors (‘allo’) with the group of all patients combined (‘auto’), the lymphoma (‘lymphoma’) and the myeloma (‘myeloma’) group, respectively. We also compared the lymphoma with the myeloma group. We deemed these two latter groups to be of interest due to the controversial roles of Treg reported in those diseases: In most lymphoma disease entities, high numbers of Treg found in the tumour environment are considered to be of positive effect. In myeloma, Tregs have been described to be either reduced or elevated. Overall, Treg might be relevant for myeloma development by preventing immunity of the host [20]. Prior to apheresis, all patients and donors had given informed consent to these quality control measurements from their blood samples. No clinical information except for the main diagnosis, mobilizing chemotherapy, © 2015 International Society of Blood Transfusion Vox Sanguinis (2015) 109, 375–386
T-cell subsets in stem cell apheresis 377
which contains Trucount beads for quantification. In an additional experiment (Fig. 1a–d), lymphocytes expressing CD3 were considered to be T cells (TC) and T-helper cells (THC) in case of expressing CD3 and CD4. Although CD4+ T cells with high expression of CD25 have formerly been described as being Treg [21], we only considered CD25+ cells to be Tregs if CD127 expression was low while CD3, CD4 and CD25 expression was high, due to the inverse correlation of Foxp3 and CD127 expression described recently [14, 15]. Such phenotyping strategy will not
age, weight and sex was included in this study, and no information on Treg content of the samples was issued to the wards. No exceptional phlebotomies were necessary for all analyses, as we used the same samples routinely needed for measuring the content of CD34+ cells in a sample pre- and postapheresis and the PBSC. CellQuest software (V 6.0, BD, Heidelberg, Germany) was used for data acquisition and analysis. CD45+ cells, CD34+ cells and lymphocytes were quantified according to the ISHAGE protocol with the stem cell enumeration kit (BD),
(a)
(c)
(b)
(d)
Fig. 1 Gating strategy of T-cell subsets. (a) Gating of lymphocytes. (b) Gating of CD3+ and CD3+ CD4+ double-positive cells out of lymphocyte gate. (c) Gating of CD3+ and CD3+ CD25+ double-positive cells out of lymphocyte gate. (d) Gating of CD25+ CD127 low cells out of CD3+ CD4+ double-positive gate (upper right quadrant, Fig. 1b).
© 2015 International Society of Blood Transfusion Vox Sanguinis (2015) 109, 375–386
378 J. Strobe et al.
include all Treg populations and is a limitation to the data set discussed later on. The antibodies used where mouse anti-CD3 labelled with the fluorochromes peridinin chlorophyll (PerCP), mouse anti-CD4 labelled with fluorescein isothiocyanate (FITC), mouse anti-CD25 labelled with R-phycoerythrin (PE) and mouse anti-CD127 labelled with allophycocyanin (APC) (all BD Pharmingen, Heidelberg, Germany). The samples were diluted with PBS to a concentration of approximately 10000 CD45+ cells per ll to analyse samples with a comparable cell to antibody ratio. We calculated collection efficiency (CE), cell yield and recruitment factor (RF) as follows: Yield = Cell concentration in concentrate [per ml] volume of concentrate (ml) Recruitment factor ¼
Postapheresis cells þ Yield Preapheresis cells
*Pre- and postapheresis cells: absolute number of cells in peripheral blood preand postapheresis = preor postapheresis blood cell count [9109/l] 9 total blood volume of subject. Collectionefficiency¼ 100
Yield : Total blood volume (ml)Preapheresis cell count (cells/ml)
For the calculation of yield per kg body weight in the allogeneic group, we used the weight of both the recipients and the donors. This is due to the fact that one aim of this pilot study was to find out whether healthy donors’ mobilization of T-cell subsets differs from that of patients. The recipients’ weight cannot be used for the calculation of recruitment, as they are not involved in the apheresis procedure. However, as yield/kg is of great clinical interest regarding the recipient’s body weight, we included those values as well. Group comparisons were also calculated with the values for recipients’ body weight. As in stem cell apheresis, the blood volume processed is exceedingly larger than in platelet or granulocyte apheresis (3–5 9 TBV vs. 06–1 9 TBV), the formula as used for calculating the CE in plateletpheresis [CE apheresis = Yield 9 100%/(processed blood volume 9 05 9 (pre- + postdonation cell count))] is not suitable. As soon as more blood volume is processed, the above-named formula underestimates the collection efficiency. This is due to the fact that by this formula the donor is seen as ‘open system’, in which much more cells are assumed to circulate than is really the case, due to the usage of the blood volume processed. However, in large volume apheresis, cells may be mobilized from
extravasal resources (e.g. bone marrow). As described previously [22, 23], we therefore referred only to the preapheresis cell counts for calculating the CE. Thus, the collection efficiency has been defined as the percentage of harvested cells in relation to the circulating cells before apheresis.
Statistical analysis Statistical analysis was performed with SPSS for Windows (IBM, Chicago, IL, USA, Version 20.0) and R 3.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Normal distribution of variables was tested with the Shapiro–Wilks test. Group comparisons consider the comparison between donors (‘Allo’-group) and patients (‘Auto’-group) and between donors and the two groups ‘lymphoma’ and ‘myeloma’. For group comparisons, the U-test was used for data not distributed normally, and the t-test was performed for normally distributed data. Differences were considered significant for P-values
