1 Introduction

Selection of stem cell source is an important consideration for any physician planning an allogeneic haematopoietic cell transplant (HCT) and has evolved considerably since bone marrow (BM) was used as the stem cell source in the first successful allogeneic HCT in 1968 (Gatti et al. 1968). BM remained the only source of stem cells for the two decades that followed until experimental work demonstrating that peripheral blood (PB) stem cells can be enriched by pre-treatment with certain chemotherapy agents and haematopoietic growth factors (Richman et al. 1976; Socinski et al. 1988; Duhrsen et al. 1988) resulted in the first peripheral blood stem cell transplant in 1986 (Korbling and Freireich 2011). Alongside this, the recognition of cord blood (CB) as a rich source of stem cells (Prindull et al. 1978) led to the successful use of cord blood as a third stem cell source in allogeneic HCT in the late 80s (Gluckman et al. 1989).

Following the introduction of PB as a stem cell source, its use evolved significantly, and, by the early 2000s, it had overtaken BM as the predominant stem cell source in adult HCT (Giebel et al. 2019). This trend has continued to increase, with PB being by far the most commonly employed stem cell source, now used in more than 80% of adult allogeneic HCT across the globe (CIBMTR n.d.; Passweg et al. 2023). In contrast, BM remains the stem cell source of choice in the paediatric setting (CIBMTR n.d.), and, despite its appeal as an intuitively attractive alternate donor source in adults and paediatrics (Wynn et al. 2022), cord blood use is declining (Passweg et al. 2021).

2 Factors Influencing the Selection of the Source

There are a number of factors that must be considered when selecting the most appropriate stem cell source for HCT and the advantages and disadvantages for both the patient and donor must be considered as well as specific disease-related considerations.

2.1 Bone Marrow

The advantages with using BM as a stem cell source include the fact that it can be harvested from a paediatric donor, which is particularly relevant in the paediatric matched-related donor setting, and there is no need for donor mobilization using granulocyte colony-stimulating factor (G-CSF). There are also fewer T cells in bone marrow grafts compared with peripheral blood, which results in less graft-versus-host disease (GVHD).

The reduced donor T-cell contamination of the graft may be disadvantageous in the malignant disease setting since T cells are involved in the graft-versus-leukaemia (GVL) effect of allogeneic HCT (Sweeney and Vyas 2019). In addition to this, the cell dose is limited by the volume of the donor marrow and is therefore generally lower than the stem cell dose obtained with mobilized peripheral blood stem cells. Red blood cell (RBC) contamination may be significant, and grafts may therefore require RBC depletion when the donor and recipient are ABO-incompatible, which may further reduce the stem cell dose.

Furthermore, BM harvesting is an invasive procedure requiring a general anaesthetic operation for the donor and, in some circumstances, red blood cell transfusion in those cases in which high cellularity is needed.

2.2 Peripheral Blood

The advantages of peripheral blood stem cells include the large donor CD34 stem cell dose, which is generally collected. This helps in more rapid engraftment and immune reconstitution, leading to lower rates of graft failure, reduced risk of infectious complications and faster hospital discharge. Furthermore, the higher number of T cells in the graft may drive engraftment, particularly in reduced intensity conditioning, and may enhance the GVL (graft-versus-leukaemia) effect in malignant diseases. Apheresis of peripheral blood normally includes only very few red blood cells, so depletion is not necessary in most cases.

The disadvantages of peripheral blood stem cells include the increased risk of chronic GVHD due to the higher amount of T-cell content of the product. This may be associated with a prolonged need for immunosuppression therapy in the long-term, post-transplant period and may adversely affect the quality of life. Allogeneic peripheral blood stem cell donation is avoided in the paediatric setting due to ethical considerations with the use of paediatric donor G-CSF mobilization. Additionally, the procedure requires adequate venous access, which can be highly challenging in children.

For donors, PB stem cell harvests are also much less invasive and more tolerable, as no operation is required. However, they need to receive G-CSF to mobilize progenitors (although the safety profile has been demonstrated in large series with long-term follow-up) in some cases, a central vein catheter may be needed, or a second day of collection may be required for some poor mobilizers.

2.3 Cord Blood

The advantages of using cord blood include the fact that it is readily available since it is already tissue-typed and cryopreserved. Additionally, the less stringent human leukocyte antigen (HLA) matching required makes it easier to find a donor across HLA barriers (Gragert et al. 2014), which is particularly important for ethnic minorities. Cord blood transplant is associated with rapid immune reconstitution where no serotherapy is used (Chiesa et al. 2012) and higher rates of chimerism in some diseases (Wynn et al. 2022). Recent evidence has also demonstrated that cord blood transplant is associated with reduced relapse and therefore a greater GVL effect (Horgan et al. 2023; Milano et al. 2016).

Disadvantages of cord blood use include a higher transplant-related mortality (Eapen et al. 2007) compared with other donor sources and a high cost, which makes it challenging in resource-limited settings. The cell dose is also limited, especially in adults or larger patients. This can be overcome using a double cord, although this adds to the complexity of the HCT and further increases the cost and transplant-related mortality (TRM). In addition to this, engraftment may be slower. Recent advances have shown the possibility of expanding progenitor cells from CB, achieving an amount of CD34 near to that obtained in PB transplants (Horwitz et al. 2019). Another disadvantage is the impossibility of obtaining additional cells that can be used for donor lymphocyte infusion (DLI) in the event of disease relapse or the need to improve immune reconstitution.

2.3.1 Disease-Related Considerations

Disease-related factors must also be considered when deciding the most appropriate stem cell source for an individual patient. In non-malignant diseases, particularly severe aplastic anaemia, BM is still the preferred stem cell choice due to the lower risk of GVHD (Schrezenmeier et al. 2007). BM is also generally considered to be the stem cell source of choice in the paediatric setting, particularly when paediatric sibling donors are used for the reasons outlined previously. The higher cell dose and more rapid engraftment result in PB being generally preferred as a stem cell source in the adult malignancy setting, particularly where reduced intensity conditioning is used.

3 Differences in the Composition of BM vs. PB

BM and PB differ in terms of cellular content, with a higher number 2 to 3 times more CD34+ cells in PB than in BM and a higher number of lymphocytes (in the range of 1 log higher) than in BM. The next table shows the differences in the composition of the different sources.

 

Peripheral blood (×106/kg)

Bone marrow (×106/kg)

Ratio of PB to BM

CD34

7.3

2.4

3

CD3

701

49

14

TCR α/β

663

42

14

TCR γ/δ

26

2

13

CD3+CD4+

393

26

15

CD4+CD45RA+

188

11

17

CD4+CD45RO+

169

10

17

CD3+CD8+

236

18

13

CD3+CD4−CD8−

45

5

9

CD19+

93

15

6

CD16+CD56+CD3-

77

6

13

CD45+CD14+

599

25

24

  1. Adapted from Körbling and Ottinger (Körbling and Anderlini 2001)

4 Main Results of Studies Comparing BM and PB

There are several randomized trials, retrospective studies and meta-analyses comparing PB and BM as a source of cells in the allogeneic stem cell transplant setting.

Globally, the comparisons suggest that PB transplantation results in faster engraftment and lower graft failure compared to bone marrow (BM) transplantation. However, there is a higher incidence of acute and chronic GVHD associated with PB transplantation.

The risk of relapse appears to be slightly lower when using PB, possibly due to the higher incidence of GVHD. This effect is particularly evident in cases involving HLA identical donors and advanced stage diseases, resulting in a benefit in disease-free survival without significant differences in the overall survival across most studies. The main results of some of these studies are presented in the following table.

 

Peripheral blood

Bone marrow

P-value

SCTCG meta-an, 1:1. HLA id sib (S.C.T.C Group 2005)

N = 538

N = 554

 

Neutrophil engraftment (days)

Platelet engraftment (days)

Acute GVHD (%)

Grade III/IV acute GVHD (%)

Chronic GVHD (%)

Extensive GVHD (%)

3-Year NRM (%)

3-Year relapse (%)

3-Year DFS (%)

3-Year OS (%)

14

14

41

26

73

51

16

21

59

62

21

22

38

21

56

35

25

27

53

59

<0.001

<0.001

0.5

0.03

0.01

0.01

0.04

0.01

0.02

0.17

Holtick meta-an, Rel and URD (Holtick et al. 2015)

N = 653

N = 677

 

Neutrophil engraftment (days)

Platelet engraftment (days)

Acute GVHD (HR)

Grade III/IV acute GVHD (HR)

Chronic GVHD (HR)

Extensive GVHD (HR)

NRM (HR)

Relapse (HR)

DFS (HR)

OS (HR)

16

13

Reference

Reference

Reference

Reference

Reference

Reference

Reference

Reference

20

19

1.03

0.75

0.72

0.69

0.98

1.3

1.04

1.07

<0.001

<0.001

0.67

0.07

<0.001

0.006

0.91

0.07

0.6

0.43

Anasetti. RCT. MA. UD (Anasetti et al. 2012)

N = 273

N = 278

 

Neutrophil engraftment (days)

Platelet engraftment (days)

Engraftment failure (%)

Acute GVHD (%)

Chronic GVHD (%)

Extensive GVHD (%)

2-Year NRM (%)

2-Year relapse (%)

2-Year DFS (%)

2-Year OS (%)

Reference

Reference

3

≈50

53

48

≈25

≈30

≈45

≈50

5 days later

7 days later

9

≈50

41

32

≈25

≈30

≈45

≈50

0.001

0.001

0.002

NS

0.01

0.001

0.66

0.74

0.38

0.33

Savani. Registry. RIC (Savani et al. 2016)

N = 9011

N = 837

 

Neutrophil engraftment (days)

Acute GVHD (%)

Chronic GVHD (%)

2-Year NRM (%)

2-Year relapse (%)

2-Year DFS (%)

2-Year OS (%)

16

24

36

20

35

44

50

20

17

29

18

43

39

47

<0.001

0.005

<0.001

0.19

<0.001

0.009

0.05

Ruggeri. Registry. Haplo donor (Ruggeri et al. 2018)

N = 191

N = 260

 

Neutrophil engraftment (days)

Acute GVHD (%)

Grade III/IV acute GVHD (HR)

Chronic GVHD (%)

2-Year NRM (%)

2-Year relapse (%)

2-Year DFS (%)

2-Year OS (%)

2-Year GRFS

17

38

14

32

23

22

54

56

43

18

21

4

36

23

26

49

55

44

0.001

0.01

0.01

0.28

0.61

0.38

0.39

0.57

0.82

Simonin. Registry. Paediatrics. (Simonin et al. 2017)

N = 719

N = 1743

 

Neutrophil engraftment (days)

Acute GVHD (odd ratio)

Chronic GVHD (%)

3-Year NRM (%)

3-Year relapse (%)

3-Year DFS (%)

3-Year OS (%)

16

Reference

44

20

26

54

62

19

1.07

21

12

29

59

67

0.001

0.55

<0.001

0.002

0.29

<0.001

<0.001

  1. Meta-an Meta-analysis, Sib Sibling, Rel Related, UR Unrelated, NS Not significant, MA Myeloablative, RIC Reduced intensity conditioning

5 The Role of Cord Blood

Cord blood is an intuitively attractive stem cell source owing to its greater permissiveness for HLA mismatch and subsequent increased donor pool (Schrezenmeier et al. 2007), low rates of chronic GVHD (Keating et al. 2019; Sharma et al. 2020; Barker et al. 2020) and ready availability, yet its recent decline has been mirrored by a rise in haploidentical transplantation (Passweg et al. 2021). This is largely due to the fact that cord blood transplants are more challenging, with a higher TRM (Eapen et al. 2007) that may in part be attributed to lower stem cell doses and T-replete strategies. Despite this, cord remains an important stem cell source and has specific utility in both malignant and non-malignant disease settings. Indeed, T-replete cord blood transplant is associated with reduced relapse and improved event-free survival in paediatric patients with high-risk myeloid malignancy (Horgan et al. 2023), and a similar reduction in relapse with cord has been observed in adults (Milano et al. 2016; Sharma et al. 2020). In non-malignant disease, the superior chimerism associated with cord blood (Church et al. 2007) results in improved outcomes in patients with metabolic disease (Lum et al. 2017; Boelens et al. 2009). Strategies to reduce the high TRM associated with cord blood transplant may serve to broaden its applicability as a stem cell source and safeguard the future of cord blood transplant, particularly in settings where the use of alternate strategies may be associated with an adverse outcome. In this line, a recent randomized study has shown that ex vivo expansion of CB is associated with faster engraftment and less early transplant complications in comparison with standard CB transplantation.

Key Points

  • The use of PB compared to BM is associated with earlier haematological recovery, a higher risk of GVHD and, possibly, a greater GVL effect. Overall, there are no significant differences observed in terms of overall survival or disease-free survival.

  • In patients at a high risk of relapse, the use of PB may be associated with a GVL effect, especially in the context of reduced-intensity transplantation.

  • In general, for patients with non-neoplastic diseases, particularly aplastic anaemia, the use of BM is preferred.

  • Since the overall results are similar and there is a lower incidence of chronic GVHD, in patients with standard-risk diseases, the consideration of BM transplantation is important as it may be associated with better long-term quality of life.