- Email: [email protected]
SCID mouse model
American Journal of Obstetrics and Gynecology (2005) 193, 990–4
www.ajog.org
Determination of engraftment potential of human cord blood stem-progenitor cells as a function of donor cell dosage and gestational age in the NOD/SCID mouse model Serdar H. Ural, MD,a Mary D. Sammel, ScD,b Karin J. Blakemore, MDc Department of Obstetrics and Gynecology,a Department of Biostatistics and Epidemiology, Center for Clinical Epidemiology and Biostatistics,b University of Pennsylvania, Philadelphia, PA; Department of Gynecology and Obstetrics,c The Johns Hopkins University School of Medicine, Baltimore, MD Received for publication February 21, 2005; revised May 9, 2005; accepted May 10, 2005
KEY WORDS Human stem cell Gestational age NOD/SCID mouse
Objective: The purpose of this study was to determine cell dosage parameters for successful engraftment of human cord blood hematopoietic stem cells (HSC) using an in vivo assay system, and to determine if there are differences with donor gestational age. Study design: HSCs were transplanted into nonobese diabetic-severe combined immunodeficient (NOD/SCID) mice. Donor cell dosage and gestational age ranges were 1 to 40!106 CD34C cells per mouse, and 23 to 40 weeks, respectively. Recipient bone marrow was assessed for engraftment capacity of the HSCs. Results: There was increasing engraftment levels with increasing dosages of transplanted HSCs. When controlled for donor HSC dosage, engraftment levels using donor cord blood from earlier gestational ages were not different from that seen using later gestational ages. Conclusion: Similar dose responses are seen using HSCs derived from the late second trimester until term in engraftment potential in the NOD/SCID mouse model. Results from this study may be applicable to human postnatal and in utero transplantation studies. Ó 2005 Mosby, Inc. All rights reserved.
The primary goal of this study was to determine cell dosage parameters that contribute to successful engraftment of fetal cord blood stem cells using an in vivo assay system, and to assess differences in engraftment potential as a function of donor gestational age. Xenotransplantation in the nonobese diabetic-severe combined immunodeficient (NOD/SCID) mouse is a useful model for in utero hematopoietic stem-progenitor Presented at the Twenty-Fifth Annual Meeting of the Society for Maternal Fetal Medicine, February 7-12, 2005, Reno, Nev. Reprints not available from the authors. 0002-9378/$ - see front matter Ó 2005 Mosby, Inc. All rights reserved. doi:10.1016/j.ajog.2005.05.051
cell (HSC) transplantation.1 The donor cell dosage necessary to achieve substantial engraftment after transplantation in this model with human cord blood HSCs has not been systematically studied as a function of donor gestational age. Cord blood stem cells are used widely as an alternative source for HSC transplantation.2 Clinical information is limited on the long-term engraftment capacity of HSC from this source. The amount of total CD (cluster designation) 34C (hematopoietic stem-progenitor) cells available from a single donor is thought to be one limiting factor.3 Compared with postnatal donor sources, cord
Ural, Sammel, and Blakemore blood HSCs may have a higher proliferative capacity because of a greater potency of stem cells derived from the developing fetus.4 It is possible, furthermore, that the engraftment potential of fetal HSCs increases at earlier gestational ages.5,6 However, in vitro hematopoietic assays measuring long-term repopulating cells may not accurately reflect in vivo engraftment capacity. Sublethally irradiated NOD/SCID mice have been shown to support transplanted human fetal cord HSCs.7 Using this in vivo model, we proposed to determine the engraftment potential of fetal cord blood HSCs at a range of cell dosages and donor gestational ages. We hypothesized that fetal HSC engraftment potential is directly correlated with donor cell dosage but may be inversely correlated with donor gestational age.
Material and methods This study was approved by the Maternal Fetal Tissue Committee of the Institutional Review Board as it pertained to tissues being discarded, and the Institution Animal Care and Use Committee (IACUC). Umbilical cord venous blood was collected into heparinized sterile syringes from human pregnancies between 23 and 40 weeks of gestation after term or preterm labor. Samples were anonymous (all patient identifiers removed). Exclusion criteria included maternal fever, chorioamnionitis, known maternal infectious diseases, and fetal congenital anomalies. We utilized purified CD34C cells from individual (nonpooled) human cord blood specimens for these experiments. CD34C cell purification was performed on isolated mononuclear cells using immunomagnetic beads.1,8 Cells were stained with phycoerythrin-conjugated mouse antihuman monoclonal antibodies, and fluorescence isothiocyanate-conjugated rat antimouse CD45. Two-color flow cytometric analysis was performed using appropriate controls. Average purity of the CD34C preparations was O90%. Six-week-old NOD/SCID mice (Jackson Labs, Bar Harbor, Me) received a sublethal dose of irradiation. Transplants were performed by intravenous injection of CD34C purified cells into the mouse lateral tail vein. Control mice were injected with sterile 0.9% normal saline. Varying dosages of donor cells from 1!106 to 40!106 CD34C cells per mouse were used at each of 10 donor gestational ages. Animals were closely monitored for signs of marrow failure (and other radiation toxicity), such as sluggishness, loss of weight, and ruffled fur. Animals were necropsied at 60 days after transplantation. Allowing mice to survive longer (5 months) may result in thymomas, complicating final assessment.9 Single cell suspensions were prepared from bone marrow by flushing both femurs and tibias.7 Because these 4 bones constitute about 25% of total bone marrow in
991 mice, the total number of bone marrow cells per mouse was estimated by multiplying by 4.10 Engraftment of human cells in the transplanted mice was quantified by flow cytometry using monoclonal antibody for human CD45.11 This is a well-established assay with a sensitivity for engraftment detection of 0.5% human cells.12 Nucleated cells showing expression of human CD45 at a level above 0.5% were considered significant. The primary statistical objective of this study was to estimate the degree of engraftment from different donor gestational ages using varying doses of CD34C donor cells. Stem cell dose was computed by multiplying the stem cell count (in units of 106) by the proportion of viable cells. Various transformations of the percent engraftment were investigated because of the skewed distribution. All statistical analyses were performed on the transformed value of engraftment. The natural log transformation of the percent engrafted best met the necessary assumptions for the regression model, and a value of 1 was added to each response before the transformation, so that those with zero engraftment could be included in the analysis. To account for potential correlation among mice that received stem cells from the same donor, we furthermore utilized generalized linear models, assuming that the correlation among the mice within a donor, are equi-correlated. Statistical testing was done assuming the potential for misclassification of this correlation using generalized estimating equation methods.13 All analyses were done using Stata, version 8 (College Station, Tex).
Results A total of 73 mice underwent transplantation using 1!106 to 40!106 HSCs per mouse (Table I). Nineteen additional control mice were dispersed among the donor gestational age groups. Eleven mice died within 7 days of transplantation because of septicemia, and were excluded from further analysis. Of the 62 recipients analyzed, 2 died at posttransplant week 4, possibly from overengraftment. These 2 mice received 40!106 and 30!106 HSCs; engraftment levels were 89% and 69%, respectively. One mouse was sacrificed at week 4 because it appeared ill, with no evidence of engraftment. Engraftment by human HSCs was seen in all but 5 of the 59 other recipients, ranging from 0.50% to 44.38%. For analytical purposes, donor gestational age was described in 3 groups; extreme prematurity (23-28 weeks), prematurity (29-36 weeks), and term (37-40 weeks). Dose of HSC was considered a continuous covariate. We explored the assumption of a linear relationship between engraftment and dose, and evaluated the potential of differing slopes among the gestational age groups by incorporating appropriate interaction terms within the
992
Ural, Sammel, and Blakemore
Table I Distribution of donor gestational age, donors, transplanted mice, stem cell doses, and controls
Table II Engraftment analyzing gestational age and donor cell dosage
Donor Mice Donor HSC Control gestational Donors transplanted dosage age (wk) (n) (n) (!106/mouse) (n) 40 39 38 37 36 32 30 28 24 23
4 3 1 1 1 1 1 1 1 1
21 17 9 5 6 3 3 2 3 4
1-25 5-40 2-30 5-30 5-15 5-10 5-10 10 2-5 5-6
1 2 2 2 2 2 2 2 2 2
model. A quadratic dose relationship was also considered, which was statistically significant (P value for quadratic dose by gestational age interaction = .025). Therefore, a quadratic model was fit for the relationship between dose and engraftment, allowing flexibility for different relationships in each of the gestational age groups. Table II describes the amount of engraftment by gestational age of the donor, and also according to the dose received. Higher donor cell dosage yielded higher engraftment levels in recipient bone marrow. No correlation was seen between donor gestational age and level of engraftment. The relationship between dose and log-transformed engraftment is presented in the Figure. The slopes were analyzed across the 3 gestational age groups (interaction P value = .025). Generally the response, ie, level of engraftment, increased with increasing dose in all 3 donor gestational age groups. For the extreme prematurity group (23-28 weeks), the engraftment potential by cell dosage relationship had a steep initial rise, but then tapered off for doses greater than 5!106. The statistical significance for incorporating this quadratic effect was P = .017. This indicated that in the extreme prematurity group (23-28 weeks), the engraftment potential did not increase in a linear fashion across all donor cell dosages. A statistically significant difference in this gestational age group versus the later 2 gestational age groups for the engraftment potential by cell dose relationship was not apparent. Transformed engraftment increased linearly for the premature group (29-36 weeks) and term group (37-40 weeks), with P values for the quadratic component of P = .261 and P = .501, respectively. For this model, the correlation among the mice receiving cells from a single donor was estimated to be .61. Additional analyses were conducted using the premature (29-36 weeks) and term (37-40 weeks) groups only, assuming a linear relationship between dose and transformed engraftment. Both groups had statistically sig-
Distribution Percent of mice engraftment percent (n) Avg (SD) Donor gestational agey %28 weeks 29-36 weeks 37-40 weeks Stem cells injectedz (0) Control (1-5)!106 (6-10)!106 (11-20)!106 O20!106
14.5% (9) 16.2% (10) 69.3% (43)
23.4% 32.0% 25.9% 6.2% 12.5%
Transformed engraftment* Avg (SD)
0.82 (0.83) 0.49 (0.46) 0.64 (0.85) 0.39 (0.44) 8.28 (19.81) 1.21 (1.21)
(19) 0 (26) 2.28 (21) 3.90 (5) 3.88 (10) 29.71
(3.29) (9.53) (2.03) (38.19)
0.87 1.02 1.52 2.41
(0.74) (0.84) (0.42) (1.61)
* Transformation log(engraftmentC1). Log = natural log base e. y P value = .003. z P value ! .001.
nificant increasing linear associations between cell dosage and engraftment (P ! .001 for both). For the premature group (29-36 weeks), an increase in dose corresponded to an increase in transformed engraftment of .062 with 95% CI (0.047–0.077). In the term (37-40 weeks) group, this same incremental dose was reflected in an increase in engraftment of .085, with 95% CI (0.056– 0.115). Although the point estimates appear discrepant, there was no statistically significant difference when comparing the slope for the premature group (29-36 weeks) to that of the term group (37-40 weeks), P = .173.
Comment HSC transplantation using cells from adult bone marrow is an established treatment modality for immunohematopoietic disorders and malignancies. Cord blood stem cells represent a useful alternate donor source for human transplantation. The number of total HSCs available from a single cord blood donor is a major limiting factor.3 However, cord blood HSCs may have higher engraftment capacity than postnatally derived marrow HSCs.4,14 Fetally derived HSCs possess several biological and therapeutic properties, which seem to be ideal for successful stem cell engraftment and for reconstitution of genetic deficiency diseases. These include their ability to rapidly survive, self-renew, proliferate, and differentiate.6 Fetal HSCs at increasingly earlier gestational ages may have progressively more potential for long-term engraftment.5,6 Although a smaller total cell product can be obtained from early fetuses, a greater engraftment potential may compensate for this drawback, thereby rendering this a promising source for donor HSCs.
Ural, Sammel, and Blakemore
993
Figure Associations between dose of stem cells (in units of 106) and transformed engraftment by gestational age groups. Engraftment transformation was y = log(engraftment C 1). Fitted regression equations are as follows: (a) y = 0.010 C 0.290 ! (dose) 0.022 ! (dose)2; (b) y = 0.068 C 0.062 ! (dose); (c) y = 0.200 C 0.085 ! (dose).
A linear regression model allowing comparison of levels of engraftment in the NOD/SCID mouse model has previously been described.10 Using this in vivo model, we sought to study the effect of donor gestational age on engrafting potential of fetal blood stem cells, and to also evaluate the cord blood HSC dose/engraftment relationship. Our results employing human cord blood HSCs indicate that engraftment levels increased as the donor cell dosage increased, using an assay which assesses in vivo engrafting capacity. Thus, our study confirms that ex vivo purified cord blood CD34C cells can generate engraftment in NOD/SCID mice in a dose response manner.1 These results may have relevance to postnatal as well as in utero HSC transplantation. The results of our study did not support the concept of a gestational age–related difference for engraftment potential using second- and third-trimester cord blood as the donor source. While our statistical methods of
analysis were appropriate for the number of animals studied, a larger sample size would be desirable to confirm our findings. It is certainly possible that cord blood–derived HSCs from even earlier donor gestational ages may have higher engraftment potential. Fetal bone marrow or fetal liver would provide the highest number of CD34C cells, and would appear to yield the population of choice for HSC reconstitution in gene therapy.15,16 Because obtaining fetal bone marrow or fetal liver is logistically difficult, obtaining cord blood at delivery is currently a more practical approach to procuring fetally derived donor HSCs for human transplantation purposes. It is therefore reassuring to confirm that cord blood stem cells had a high engraftment potential in NOD/SCID mice. The NOD/SCID mouse has, furthermore, been demonstrated to be a useful model for human in utero HSC transplantation. Using human bone marrow–derived HSC doses equivalent to those used in in utero bone
994 marrow transplantation clinically, Leung et al were able to demonstrate comparable levels of engraftment in the NOD/SCID mouse model. These authors, furthermore, found the pathologic sequelae of overengraftment and the timing posttransplant of these sequelae to be remarkably similar to that seen in the human.1 A similar result was seen in 2 mice in our study using human cord blood at a similar HSC dosage. Our results on donor cell dosing using the NOD/SCID mouse model further the initial observations of Leung et al, and may have direct relevance to cell dosing for future trials of human in utero HSC transplantation. This promising therapy has the potential to lead to safe and effective treatment of a variety of human congenital disorders. In summary, we were able to demonstrate the high potential for successful stem cell transplantation in the NOD/SCID mouse model using human HSCs derived from fetal blood. Engraftment levels increased with increasing doses of transplanted HSCs. There was no statistically significant correlation between engraftment levels and donor gestational age across the late second and the third trimesters. Our model is a useful one to further study the relationship between donor gestational age and engraftment potential using increased numbers of cord blood samples in earlier gestational ages.
Ural, Sammel, and Blakemore
4.
5.
6.
7.
8.
9.
10. 11.
12.
References 13. 1. Leung W, Blakemore K, Jones RJ, Moser HW, Mukherjee G, Griffin CA, et al. The human-murine chimera model for in utero human hematopoietic stem cell transplantation. Biol Blood Marrow Transplant 1999;5:1-7. 2. Cairo MS, Wagner JE. Placental and/or umbilical cord blood: an alternative source of hematopoietic stem cells for transplantation. Blood 1997;90:4665-78. 3. Gluckman E, Rocha V, Boyer-Chammard A, Locatelli F, Arcese W, Pasquini R, et al. Outcome of cord-blood transplantation from related and unrelated donors. Eurocord Transplant Group and the
14.
15. 16.
European Blood and Marrow Transplantation Group. N Engl J Med 1997;337:373-81. Thilaganathan B, Nicolaides KH, Morgan G. Subpopulations of CD34-positive haematopoietic progenitors in fetal blood. Br J Haematol 1994;87:634-6. Shields LE, Andrews RG. Gestational age changes in circulating CD34 supC hematopoietic stem/progenitor cells in fetal cord blood. Am J Obstet Gynecol 1998;178:931-7. Wyrsch A, dalle Carbonare V, Jansen W, Chklovskaia E, Nissen C, Surbek D, et al. Umbilical cord blood from preterm human fetuses is rich in committed and primitive hematopoietic progenitors with high proliferative and self-renewal capacity. Exp Hematol 1999;27:1338-45. Pflumio F, Izac B, Katz A, Shultz LD, Vainchenker W, Coulombel L. Phenotype and function of human hematopoietic cells engrafting immune-deficient CB17-severe combined immunodeficiency mice and NOD/SCID mice after transplantation of human cord blood mononuclear cells. Blood 1996;88:3731-40. Civin CI, Almeida-Porada G, Lee MJ, Olweus J, Terstappen LW, Zanjani ED. Sustained, retransplantable, multilineage engraftment of highly purified adult human bone marrow stem cells in vivo. Blood 1996;88:4102-9. Shultz LD, Schweitzer PA, Christianson SW, Gott B, Schweitzer IB, Tennent B, et al. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J Immunol 1995;154:180-91. Boggs DR. The total marrow mass of the mouse: a simplified method of measurement. Am J Hematol 1984;16:277-86. Leung W, Ramirez M, Novelli EM, Civin CI. In vivo engraftment potential of clinical hematopoietic grafts. J Invest Med 1998; 46:303-11. Holyoake TL, Horrocks C, Thomas T, Eaves CJ, Eaves AC. Cell separation improves the sensitivity of detecting rare human normal and leukemic hematopoietic cells in vivo in NOD/SCID mice. Cytotherapy 2000;2:411-21. Liang KY, Zeger S. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. Tanavde V, Malehorn M, Lumkul R, Gao Z, Wingard J, Garett E, et al. Human stem-progenitor cells from neonatal cord blood have greater hematopoietic expansion capacity than those from mobilized adult blood. Exp Hematol 2002;30:816-23. Michejda M. Quo vadis? Fetal tissue tranplantation. J Hematother 1996;5:185-8. Surbek DV, Gratwohl A, Holzgreve W. In utero hematopoeitic stem cell transfer: current status and future strategies. Eur J Obstet Gynaecol Reprod Biol 1999;85:109-15.
www.ajog.org
Determination of engraftment potential of human cord blood stem-progenitor cells as a function of donor cell dosage and gestational age in the NOD/SCID mouse model Serdar H. Ural, MD,a Mary D. Sammel, ScD,b Karin J. Blakemore, MDc Department of Obstetrics and Gynecology,a Department of Biostatistics and Epidemiology, Center for Clinical Epidemiology and Biostatistics,b University of Pennsylvania, Philadelphia, PA; Department of Gynecology and Obstetrics,c The Johns Hopkins University School of Medicine, Baltimore, MD Received for publication February 21, 2005; revised May 9, 2005; accepted May 10, 2005
KEY WORDS Human stem cell Gestational age NOD/SCID mouse
Objective: The purpose of this study was to determine cell dosage parameters for successful engraftment of human cord blood hematopoietic stem cells (HSC) using an in vivo assay system, and to determine if there are differences with donor gestational age. Study design: HSCs were transplanted into nonobese diabetic-severe combined immunodeficient (NOD/SCID) mice. Donor cell dosage and gestational age ranges were 1 to 40!106 CD34C cells per mouse, and 23 to 40 weeks, respectively. Recipient bone marrow was assessed for engraftment capacity of the HSCs. Results: There was increasing engraftment levels with increasing dosages of transplanted HSCs. When controlled for donor HSC dosage, engraftment levels using donor cord blood from earlier gestational ages were not different from that seen using later gestational ages. Conclusion: Similar dose responses are seen using HSCs derived from the late second trimester until term in engraftment potential in the NOD/SCID mouse model. Results from this study may be applicable to human postnatal and in utero transplantation studies. Ó 2005 Mosby, Inc. All rights reserved.
The primary goal of this study was to determine cell dosage parameters that contribute to successful engraftment of fetal cord blood stem cells using an in vivo assay system, and to assess differences in engraftment potential as a function of donor gestational age. Xenotransplantation in the nonobese diabetic-severe combined immunodeficient (NOD/SCID) mouse is a useful model for in utero hematopoietic stem-progenitor Presented at the Twenty-Fifth Annual Meeting of the Society for Maternal Fetal Medicine, February 7-12, 2005, Reno, Nev. Reprints not available from the authors. 0002-9378/$ - see front matter Ó 2005 Mosby, Inc. All rights reserved. doi:10.1016/j.ajog.2005.05.051
cell (HSC) transplantation.1 The donor cell dosage necessary to achieve substantial engraftment after transplantation in this model with human cord blood HSCs has not been systematically studied as a function of donor gestational age. Cord blood stem cells are used widely as an alternative source for HSC transplantation.2 Clinical information is limited on the long-term engraftment capacity of HSC from this source. The amount of total CD (cluster designation) 34C (hematopoietic stem-progenitor) cells available from a single donor is thought to be one limiting factor.3 Compared with postnatal donor sources, cord
Ural, Sammel, and Blakemore blood HSCs may have a higher proliferative capacity because of a greater potency of stem cells derived from the developing fetus.4 It is possible, furthermore, that the engraftment potential of fetal HSCs increases at earlier gestational ages.5,6 However, in vitro hematopoietic assays measuring long-term repopulating cells may not accurately reflect in vivo engraftment capacity. Sublethally irradiated NOD/SCID mice have been shown to support transplanted human fetal cord HSCs.7 Using this in vivo model, we proposed to determine the engraftment potential of fetal cord blood HSCs at a range of cell dosages and donor gestational ages. We hypothesized that fetal HSC engraftment potential is directly correlated with donor cell dosage but may be inversely correlated with donor gestational age.
Material and methods This study was approved by the Maternal Fetal Tissue Committee of the Institutional Review Board as it pertained to tissues being discarded, and the Institution Animal Care and Use Committee (IACUC). Umbilical cord venous blood was collected into heparinized sterile syringes from human pregnancies between 23 and 40 weeks of gestation after term or preterm labor. Samples were anonymous (all patient identifiers removed). Exclusion criteria included maternal fever, chorioamnionitis, known maternal infectious diseases, and fetal congenital anomalies. We utilized purified CD34C cells from individual (nonpooled) human cord blood specimens for these experiments. CD34C cell purification was performed on isolated mononuclear cells using immunomagnetic beads.1,8 Cells were stained with phycoerythrin-conjugated mouse antihuman monoclonal antibodies, and fluorescence isothiocyanate-conjugated rat antimouse CD45. Two-color flow cytometric analysis was performed using appropriate controls. Average purity of the CD34C preparations was O90%. Six-week-old NOD/SCID mice (Jackson Labs, Bar Harbor, Me) received a sublethal dose of irradiation. Transplants were performed by intravenous injection of CD34C purified cells into the mouse lateral tail vein. Control mice were injected with sterile 0.9% normal saline. Varying dosages of donor cells from 1!106 to 40!106 CD34C cells per mouse were used at each of 10 donor gestational ages. Animals were closely monitored for signs of marrow failure (and other radiation toxicity), such as sluggishness, loss of weight, and ruffled fur. Animals were necropsied at 60 days after transplantation. Allowing mice to survive longer (5 months) may result in thymomas, complicating final assessment.9 Single cell suspensions were prepared from bone marrow by flushing both femurs and tibias.7 Because these 4 bones constitute about 25% of total bone marrow in
991 mice, the total number of bone marrow cells per mouse was estimated by multiplying by 4.10 Engraftment of human cells in the transplanted mice was quantified by flow cytometry using monoclonal antibody for human CD45.11 This is a well-established assay with a sensitivity for engraftment detection of 0.5% human cells.12 Nucleated cells showing expression of human CD45 at a level above 0.5% were considered significant. The primary statistical objective of this study was to estimate the degree of engraftment from different donor gestational ages using varying doses of CD34C donor cells. Stem cell dose was computed by multiplying the stem cell count (in units of 106) by the proportion of viable cells. Various transformations of the percent engraftment were investigated because of the skewed distribution. All statistical analyses were performed on the transformed value of engraftment. The natural log transformation of the percent engrafted best met the necessary assumptions for the regression model, and a value of 1 was added to each response before the transformation, so that those with zero engraftment could be included in the analysis. To account for potential correlation among mice that received stem cells from the same donor, we furthermore utilized generalized linear models, assuming that the correlation among the mice within a donor, are equi-correlated. Statistical testing was done assuming the potential for misclassification of this correlation using generalized estimating equation methods.13 All analyses were done using Stata, version 8 (College Station, Tex).
Results A total of 73 mice underwent transplantation using 1!106 to 40!106 HSCs per mouse (Table I). Nineteen additional control mice were dispersed among the donor gestational age groups. Eleven mice died within 7 days of transplantation because of septicemia, and were excluded from further analysis. Of the 62 recipients analyzed, 2 died at posttransplant week 4, possibly from overengraftment. These 2 mice received 40!106 and 30!106 HSCs; engraftment levels were 89% and 69%, respectively. One mouse was sacrificed at week 4 because it appeared ill, with no evidence of engraftment. Engraftment by human HSCs was seen in all but 5 of the 59 other recipients, ranging from 0.50% to 44.38%. For analytical purposes, donor gestational age was described in 3 groups; extreme prematurity (23-28 weeks), prematurity (29-36 weeks), and term (37-40 weeks). Dose of HSC was considered a continuous covariate. We explored the assumption of a linear relationship between engraftment and dose, and evaluated the potential of differing slopes among the gestational age groups by incorporating appropriate interaction terms within the
992
Ural, Sammel, and Blakemore
Table I Distribution of donor gestational age, donors, transplanted mice, stem cell doses, and controls
Table II Engraftment analyzing gestational age and donor cell dosage
Donor Mice Donor HSC Control gestational Donors transplanted dosage age (wk) (n) (n) (!106/mouse) (n) 40 39 38 37 36 32 30 28 24 23
4 3 1 1 1 1 1 1 1 1
21 17 9 5 6 3 3 2 3 4
1-25 5-40 2-30 5-30 5-15 5-10 5-10 10 2-5 5-6
1 2 2 2 2 2 2 2 2 2
model. A quadratic dose relationship was also considered, which was statistically significant (P value for quadratic dose by gestational age interaction = .025). Therefore, a quadratic model was fit for the relationship between dose and engraftment, allowing flexibility for different relationships in each of the gestational age groups. Table II describes the amount of engraftment by gestational age of the donor, and also according to the dose received. Higher donor cell dosage yielded higher engraftment levels in recipient bone marrow. No correlation was seen between donor gestational age and level of engraftment. The relationship between dose and log-transformed engraftment is presented in the Figure. The slopes were analyzed across the 3 gestational age groups (interaction P value = .025). Generally the response, ie, level of engraftment, increased with increasing dose in all 3 donor gestational age groups. For the extreme prematurity group (23-28 weeks), the engraftment potential by cell dosage relationship had a steep initial rise, but then tapered off for doses greater than 5!106. The statistical significance for incorporating this quadratic effect was P = .017. This indicated that in the extreme prematurity group (23-28 weeks), the engraftment potential did not increase in a linear fashion across all donor cell dosages. A statistically significant difference in this gestational age group versus the later 2 gestational age groups for the engraftment potential by cell dose relationship was not apparent. Transformed engraftment increased linearly for the premature group (29-36 weeks) and term group (37-40 weeks), with P values for the quadratic component of P = .261 and P = .501, respectively. For this model, the correlation among the mice receiving cells from a single donor was estimated to be .61. Additional analyses were conducted using the premature (29-36 weeks) and term (37-40 weeks) groups only, assuming a linear relationship between dose and transformed engraftment. Both groups had statistically sig-
Distribution Percent of mice engraftment percent (n) Avg (SD) Donor gestational agey %28 weeks 29-36 weeks 37-40 weeks Stem cells injectedz (0) Control (1-5)!106 (6-10)!106 (11-20)!106 O20!106
14.5% (9) 16.2% (10) 69.3% (43)
23.4% 32.0% 25.9% 6.2% 12.5%
Transformed engraftment* Avg (SD)
0.82 (0.83) 0.49 (0.46) 0.64 (0.85) 0.39 (0.44) 8.28 (19.81) 1.21 (1.21)
(19) 0 (26) 2.28 (21) 3.90 (5) 3.88 (10) 29.71
(3.29) (9.53) (2.03) (38.19)
0.87 1.02 1.52 2.41
(0.74) (0.84) (0.42) (1.61)
* Transformation log(engraftmentC1). Log = natural log base e. y P value = .003. z P value ! .001.
nificant increasing linear associations between cell dosage and engraftment (P ! .001 for both). For the premature group (29-36 weeks), an increase in dose corresponded to an increase in transformed engraftment of .062 with 95% CI (0.047–0.077). In the term (37-40 weeks) group, this same incremental dose was reflected in an increase in engraftment of .085, with 95% CI (0.056– 0.115). Although the point estimates appear discrepant, there was no statistically significant difference when comparing the slope for the premature group (29-36 weeks) to that of the term group (37-40 weeks), P = .173.
Comment HSC transplantation using cells from adult bone marrow is an established treatment modality for immunohematopoietic disorders and malignancies. Cord blood stem cells represent a useful alternate donor source for human transplantation. The number of total HSCs available from a single cord blood donor is a major limiting factor.3 However, cord blood HSCs may have higher engraftment capacity than postnatally derived marrow HSCs.4,14 Fetally derived HSCs possess several biological and therapeutic properties, which seem to be ideal for successful stem cell engraftment and for reconstitution of genetic deficiency diseases. These include their ability to rapidly survive, self-renew, proliferate, and differentiate.6 Fetal HSCs at increasingly earlier gestational ages may have progressively more potential for long-term engraftment.5,6 Although a smaller total cell product can be obtained from early fetuses, a greater engraftment potential may compensate for this drawback, thereby rendering this a promising source for donor HSCs.
Ural, Sammel, and Blakemore
993
Figure Associations between dose of stem cells (in units of 106) and transformed engraftment by gestational age groups. Engraftment transformation was y = log(engraftment C 1). Fitted regression equations are as follows: (a) y = 0.010 C 0.290 ! (dose) 0.022 ! (dose)2; (b) y = 0.068 C 0.062 ! (dose); (c) y = 0.200 C 0.085 ! (dose).
A linear regression model allowing comparison of levels of engraftment in the NOD/SCID mouse model has previously been described.10 Using this in vivo model, we sought to study the effect of donor gestational age on engrafting potential of fetal blood stem cells, and to also evaluate the cord blood HSC dose/engraftment relationship. Our results employing human cord blood HSCs indicate that engraftment levels increased as the donor cell dosage increased, using an assay which assesses in vivo engrafting capacity. Thus, our study confirms that ex vivo purified cord blood CD34C cells can generate engraftment in NOD/SCID mice in a dose response manner.1 These results may have relevance to postnatal as well as in utero HSC transplantation. The results of our study did not support the concept of a gestational age–related difference for engraftment potential using second- and third-trimester cord blood as the donor source. While our statistical methods of
analysis were appropriate for the number of animals studied, a larger sample size would be desirable to confirm our findings. It is certainly possible that cord blood–derived HSCs from even earlier donor gestational ages may have higher engraftment potential. Fetal bone marrow or fetal liver would provide the highest number of CD34C cells, and would appear to yield the population of choice for HSC reconstitution in gene therapy.15,16 Because obtaining fetal bone marrow or fetal liver is logistically difficult, obtaining cord blood at delivery is currently a more practical approach to procuring fetally derived donor HSCs for human transplantation purposes. It is therefore reassuring to confirm that cord blood stem cells had a high engraftment potential in NOD/SCID mice. The NOD/SCID mouse has, furthermore, been demonstrated to be a useful model for human in utero HSC transplantation. Using human bone marrow–derived HSC doses equivalent to those used in in utero bone
994 marrow transplantation clinically, Leung et al were able to demonstrate comparable levels of engraftment in the NOD/SCID mouse model. These authors, furthermore, found the pathologic sequelae of overengraftment and the timing posttransplant of these sequelae to be remarkably similar to that seen in the human.1 A similar result was seen in 2 mice in our study using human cord blood at a similar HSC dosage. Our results on donor cell dosing using the NOD/SCID mouse model further the initial observations of Leung et al, and may have direct relevance to cell dosing for future trials of human in utero HSC transplantation. This promising therapy has the potential to lead to safe and effective treatment of a variety of human congenital disorders. In summary, we were able to demonstrate the high potential for successful stem cell transplantation in the NOD/SCID mouse model using human HSCs derived from fetal blood. Engraftment levels increased with increasing doses of transplanted HSCs. There was no statistically significant correlation between engraftment levels and donor gestational age across the late second and the third trimesters. Our model is a useful one to further study the relationship between donor gestational age and engraftment potential using increased numbers of cord blood samples in earlier gestational ages.
Ural, Sammel, and Blakemore
4.
5.
6.
7.
8.
9.
10. 11.
12.
References 13. 1. Leung W, Blakemore K, Jones RJ, Moser HW, Mukherjee G, Griffin CA, et al. The human-murine chimera model for in utero human hematopoietic stem cell transplantation. Biol Blood Marrow Transplant 1999;5:1-7. 2. Cairo MS, Wagner JE. Placental and/or umbilical cord blood: an alternative source of hematopoietic stem cells for transplantation. Blood 1997;90:4665-78. 3. Gluckman E, Rocha V, Boyer-Chammard A, Locatelli F, Arcese W, Pasquini R, et al. Outcome of cord-blood transplantation from related and unrelated donors. Eurocord Transplant Group and the
14.
15. 16.
European Blood and Marrow Transplantation Group. N Engl J Med 1997;337:373-81. Thilaganathan B, Nicolaides KH, Morgan G. Subpopulations of CD34-positive haematopoietic progenitors in fetal blood. Br J Haematol 1994;87:634-6. Shields LE, Andrews RG. Gestational age changes in circulating CD34 supC hematopoietic stem/progenitor cells in fetal cord blood. Am J Obstet Gynecol 1998;178:931-7. Wyrsch A, dalle Carbonare V, Jansen W, Chklovskaia E, Nissen C, Surbek D, et al. Umbilical cord blood from preterm human fetuses is rich in committed and primitive hematopoietic progenitors with high proliferative and self-renewal capacity. Exp Hematol 1999;27:1338-45. Pflumio F, Izac B, Katz A, Shultz LD, Vainchenker W, Coulombel L. Phenotype and function of human hematopoietic cells engrafting immune-deficient CB17-severe combined immunodeficiency mice and NOD/SCID mice after transplantation of human cord blood mononuclear cells. Blood 1996;88:3731-40. Civin CI, Almeida-Porada G, Lee MJ, Olweus J, Terstappen LW, Zanjani ED. Sustained, retransplantable, multilineage engraftment of highly purified adult human bone marrow stem cells in vivo. Blood 1996;88:4102-9. Shultz LD, Schweitzer PA, Christianson SW, Gott B, Schweitzer IB, Tennent B, et al. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J Immunol 1995;154:180-91. Boggs DR. The total marrow mass of the mouse: a simplified method of measurement. Am J Hematol 1984;16:277-86. Leung W, Ramirez M, Novelli EM, Civin CI. In vivo engraftment potential of clinical hematopoietic grafts. J Invest Med 1998; 46:303-11. Holyoake TL, Horrocks C, Thomas T, Eaves CJ, Eaves AC. Cell separation improves the sensitivity of detecting rare human normal and leukemic hematopoietic cells in vivo in NOD/SCID mice. Cytotherapy 2000;2:411-21. Liang KY, Zeger S. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. Tanavde V, Malehorn M, Lumkul R, Gao Z, Wingard J, Garett E, et al. Human stem-progenitor cells from neonatal cord blood have greater hematopoietic expansion capacity than those from mobilized adult blood. Exp Hematol 2002;30:816-23. Michejda M. Quo vadis? Fetal tissue tranplantation. J Hematother 1996;5:185-8. Surbek DV, Gratwohl A, Holzgreve W. In utero hematopoeitic stem cell transfer: current status and future strategies. Eur J Obstet Gynaecol Reprod Biol 1999;85:109-15.