Due to the immaturity of the hematopoietic system at birth and different oxygenation and immune response needs of the growing organism, the bone marrow composition at birth and early infancy differs as compared to older children and adults. These age-related differences, while generally recognized, are not well known to the world of hematopathology. The purpose of this article is to address the current limitation of the literature by reviewing the bone marrow ontology, its composition at birth, and the changes occurring during early infancy, and to compare these findings to adults. The review also provides a useful framework for bone marrow examination in children.
The bone marrow (BM) is the last blood-forming tissue that develops in ontogenesis and from birth, and thereafter, it is the major hematopoietic site. It is a functionally dynamic organ, and its composition depends highly on the needs for oxygenation, fighting infections, and proper hemostasis. As such needs vary drastically during different stages of development as well as early childhood and later in life, the BM composition also changes to meet those needs. Therefore, it is important when evaluating BM of a child to distinguish between the findings due to the normal development and those that result from various diseases.
The hematopoiesis in the bone marrow begins in the long bones at 6–8.5 weeks of gestation and is completed by 16 weeks of gestation with final organization of the bones into areas of dense hematopoiesis surrounded by areas of fully calcified bone [1, 2]. Between the 11th and 24th weeks of gestation, both the liver and BM are hematopoietic organs concomitantly, yet each supports a different set of hematopoietic lineages: the erythropoiesis occurs mostly in the fetal liver and the granulopoiesis and lymphopoiesis mostly in the BM. The total marrow volume markedly increases during the second trimester, and after the 24th week of gestation, the hematopoiesis shifts from the fetal liver to the BM.
From birth onward, the BM is the major hematopoietic site. At birth, all cavities of the skeleton contain red hematopoietic BM and almost no fat. In the first year of life, the hematopoiesis occurs in both the axial and appendicular skeleton, and thereafter, there is a gradual decrease in the hematopoiesis in the long bones until about age 15. At that age, active hematopoiesis is confined to the proximal quarters of the shafts of the femur, humerus, and the axial skeleton .
The BM is a functionally dynamic structure, and if the needs for erythrocyte, leukocyte, or platelet production increase, the hematopoiesis expands, and the fat is replaced by bone marrow. In young children, however, an increase in hematopoiesis is accommodated by a reduction in the proportion of marrow sinusoids, and in severe congenital anemia, the marrow cavities expand leading to bone deformity.
The BM is located between the bone trabeculae and has a highly complex three-dimensional structure composed of extracellular matrix, stromal cells including osteoblasts, and capillary venous sinuses. The localization of the various hematopoietic elements is nonrandom, and in histologic sections, the cells with proliferative activity are preferentially located near the bone trabeculae, and the differentiated elements are observed in the central, intertrabecular spaces .
The early myeloid progenitors are localized in the paratrabecular areas close to the adventitia of the small arteries. Normally, the layer of immature granulocytes does not exceed 2–3 rows of maturing cells. With maturation, the cells migrate to the intertrabecular spaces.
The erythroid progenitors mature and differentiate in erythroblastic islands that consist of a central macrophage – a key component of the erythroid differentiation – surrounded by developing erythroblasts. As the erythroblasts become more differentiated, the erythroid islands migrate toward sinusoids because they are a mobile structure. The erythroid islands are not readily observable on histologic section, but can be seen on bone marrow aspirates, particularly in patients with erythroid hyperplasia.
Megakaryocytes reside adjacent to marrow sinusoids, which allow easy shedding of platelets directly into the circulation.
Age-Specific Differences in the BM
Due to the immaturity of the hematopoietic system at birth and the nature of hematopoiesis with its dynamic response to the oxygenation needs and immune response of the growing organism, several important differences between the BM cellularity and composition in children and adults exist (Table 1) [5-8]. At birth, the bone marrow is almost devoid of fat (100%) and contains only hematopoietic elements, and with advance of age, the red hematopoietic marrow is replaced by fat. The BM cellularity in normal infants – 3 months of age or younger – is 80% or more. Particles obtained from the iliac crest and the sternum have compatible cellularity and, in normal children aged 18 months to 11 years, ranged between 50% and 70% .
|Age||Cellularity||Cellular composition||Bone trabeculae|
|Newborn||90–100%||Myeloid hyperplasia with a shift to immaturity. Less than 5% blasts|| |
Very active bone remodeling
Prominent osteoblastic rimming
|Neonate (birth to 28 days)||90%|| |
Decrease in the number of myeloid cells in the first 2 weeks of life
Decrease in the number of erythroid progenitors
Monolobated small megakaryocytes
Gradual increase in the number of lymphoid cells, mostly B cells
|Infant (1 month to 1 year)||80–90%|| |
M:E ratio 5–12 : 1a
Reaches a steady state level ~30–35%
After the initial drop, a transient peak in the 2nd month, there is a second decrease at 3–4 months with subsequent stabilization of the total erythroblast count of 7–9%; Relative erythroid hypoplasia is most pronounced during the physiologic nadir
Monolobated small megakaryocytes
Diffuse interstitial lymphocytosis
Lymphoid cells, the major population after the 1st month (47.2 ± 9.2%)
Predominance of normal B cell progenitors (hematogones)
T-cells and NK-cells minor components
Lymphoid aggregates not normally present
Plasma cells – rare, polytypic, frequently associated with various diseases
L:M:E ratio ~ 6 : 5 : 1a
Iron stores – absent in young children
|2–5 years||60–80%|| |
M:E ratio 3–4 : 1
Increase in the myeloid and erythroid component and decrease in the number of B cells and hematogones and slight increase in the number of T cells. Note, an increased number of hematogones can be seen in infections or children with congenital cytopenias due to primary bone marrow failure.
Detectable stainable iron after age 4–5 years
|Prominent bone remodeling|
|6–12 years||50–70%|| |
M:E ratio 3–4 : 1
Less than 20% lymphocytes
T cells exceed B cells
Iron stores reach adult level
|Bone remodeling may be evident, particularly boys|
|Older than 12 years and adults||40–60%|| |
M:E ratio 3–4 : 1
Less than 20% lymphocytes
T cells exceed B cells
Lymphogranuloma and lymphoid aggregates may be present, their number increases with age
Inconspicuous osteoblasts and osteoclasts
No bone remodeling
In a study of 448 healthy BM donors and children with non-neoplastic hematologic disorders or nonhematologic malignancies, Friebert et al.  found that children younger than 2 years have the highest BM cellularity (79.8 ± 15.7%) and with age the cellularity declined to 68.6% (±16.5%) for children aged 2–4 years, to 59.1% (±20.1%) for children aged 5–9 years, and to 60% (±17.9%) for children aged 10–14 years. Thus, BM cellularity of 60% is achieved during the first 5 years of life and remains relatively constant compatible with other age groups after that. Similar findings are obtained using imaging studies. Ogawa et al.  found 60.0 ± 20.0% cellularity in the BM of children aged 0–9 years declining to 56.5 ± 4.4% for ages 10–19 years. These findings clearly demonstrate the inaccuracy of the concept of BM cellularity determined by the formula 100% minus the age of the patient. Using such an approach will significantly overestimate the BM cellularity in young children and underestimate the cellularity in older adults. Thus, it should not be applied.
Similarly to the BM cellularity, the composition of the marrow is also age dependent [5, 10-13]. In a prospective study of BM composition of 88 clinically healthy children with normal peripheral blood counts, serum proteins, and transferrin saturation of at least 16%, Sturgeon found that at birth, the BM has a predominance of myeloid progenitors and relatively low number of erythroid progenitors and lymphocytes. The most significant changes take place during the first month of life and are manifested by a decrease in the percentage of myeloid progenitors and erythroblasts and an increase in the number of lymphocytes (Figure 1). The total myeloid component initially decreases during the first 2 weeks of life followed by a sharp drop around the 3rd week to reach a steady level around 30–35% after the 1st month. The marrow eosinophils range between 2 and 3%, the monocytes remain below 2%, and the basophils remain below 1% at all times. The plasma cells and other marrow cells comprise only a small fraction of the cellularity.
The number of erythroid progenitors also decreases significantly during the first month of life and after a transient peak at the end of the second month is followed by more gradual but significant secondary decrease through months 3 and 4 to stabilize and account for 7–9% of total erythroblasts at various stages of maturation. These changes are broadly followed by the peripheral blood reticulocyte count.
The number of megakaryocytes in children is compatible with adults. However, there is a difference in their size and lobulation. Small megakaryocytes with uniform size and monolobated nucleus, occasionally forming clusters, are characteristic for young children and should not be mistaken for abnormal megakaryopoiesis. In a study of the size of megakaryocyte of 61 young children, Fuchs et al.  found a single peak signifying uniform small size at the youngest ages, which diverges into separate clusters of smaller and larger cells beginning at 2 years that is followed by an overall shift toward larger megakaryocytes at age 4 years.
Another significant difference between the BM composition of infants and older children and adults is the presence of a high number of lymphocytes that increase significantly in the immediate neonatal period to become the largest population in the marrow (47.2 ± 9.2%) by the end of the first month . During the next 17 months, the number of lymphocytes is relatively stable whereupon it begins to decrease gradually. During the first 4 years of life, the B cells comprise 65% of the lymphocytes in contrast to adult marrow where T lymphocytes predominate . Most of the B cells, 80%, are normal B-cell progenitors, hematogones. These cells comprise a heterogeneous population of immature, surface immunoglobulin-negative B cells that include early, intermediate, and late forms (Figure 1). The early, stage I, hematogones express CD34 and TdT and are mostly CD20 negative. The stage II and stage III hematogones loose CD34 and TdT and gradually acquire CD20. Furthermore, stage III hematogones acquire cytoplasmic μ and loose CD10. During the first 2 years, the CD20 and surface immunoglobulin-positive naïve B cells comprise a minor population, and with age, their number gradually increases, whereas the number of hematogones decreases. After the age of 4 years, the overall number of B cells gradually decreases, and this decline is accompanied by an increase in the number of CD3+ T lymphocytes. These T cells are mature and express either CD4 or CD8, and the percent of CD8+ cells is at least twice the percentage of CD4+ cells. The number of NK cells in the marrow is low and does not depend on age.
The proliferative capacity of BM in children and young adults determined by Ki-67 appears to be higher and the apoptotic rate lower as compared to older people . Active bone remodeling, prominent osteoblastic rimming, osteoid seams, and incomplete ossification are frequent in children and do not signify pathology.
Bone Marrow Examination in Children
The bone marrow diagnosis in children as well as in adults is based on the integration of data from various diagnostic studies, including peripheral blood count and film evaluation, BM aspirate smears, particle clot sections, BM trephine biopsy, and imprint morphology along with the results of cytochemistry, immunophenotypic analysis, cytogenetics, and molecular studies . The most frequent indication for bone marrow examination in children includes investigation of abnormal blood counts suggestive of BM pathology; initial workup for a child with peripheral cytopenias and suspected primary bone marrow failure or occult malignancy; unexplained organomegaly in children with mass lesions inaccessible for biopsy; following response to therapy for acute leukemia and detection of minimal residual leukemia; to determine BM engraftment following a stem cell transplant; and staging for Hodgkin or non-Hodgkin lymphoma, neuroblastoma, or rhabdomyosarcoma. Of note, unlike neuroblastoma and rhabdomyosarcoma that metastasize to the BM frequently, other small blue cell tumors such as the Ewing sarcoma family of tumors, Wilms tumor, and nonrhabdomyosarcoma soft tissue sarcomas rarely involve the BM; thus, BM examination is not part of the routine staging for those tumors.
In young children, BM studies are not indicated for the determination of iron stores, as stainable iron is absent from the marrow during the first year of life and storage iron progressively increases and reaches adult levels only by the fifth to 6th year . Similarly, BM studies are not required in the initial evaluation of children with thrombocytopenia prior to initiation of steroid therapy as such studies have been shown to contribute little and not to change significantly the quality-of-life years of such children.
In children with suspected acute leukemia, the diagnosis can be established on BM aspirate and peripheral blood alone, and BM biopsy may not be necessary. However, in children with suspected aplastic anemia, BM involvement by small blue cell tumor, or staging for Hodgkin or non-Hodgkin lymphoma, a BM biopsy will be necessary to establish the diagnosis. In such studies, sampling both, the BM aspirates and biopsy, will enhance the chances of arriving to the right diagnosis.
The iliac crest is the most frequent BM sampling site in children. In newborns and neonates, conventional BM biopsies may be difficult to perform, so alternative techniques, such as obtaining marrow clot sections, can be successfully utilized [17, 18]. The BM particles in clot sections contain preserved marrow architecture, so that they provide information on the overall BM cellularity and the number and morphology of megakaryocytes. Such sections can also be used for immunohistochemical stains or other special stains. While the adult requirement for the size of BM biopsy of at least 1.5–2 cm may not be achievable in very young children, an effort should be made so that sufficient material is provided for proper examination .
After chemotherapy, BM aspirates may be paucicellular and may not contain hematopoietic elements. In the absence of particles or megakaryocytes, or other hematopoietic precursors, the sample should be reported as ‘dry tap’ or peripheral blood . In such instances, fresh BM biopsy is suitable for flow cytometry that will provide valuable information on the presence or absence of minimal residual disease.
Examination of BM aspirate smears at low-power magnification is essential for determining the number and cellularity of marrow particles, megakaryocytes as well as presence of tumor clumps or abnormal cells that may be of low incidence. This is particularly important in paucicellular BM smears in children with suspected neuroblastoma or rhabdomyosarcoma.
Higher magnification provides valuable information on the composition of marrow particles, cytological details, and cell inclusions. As with adults, both quantitative and qualitative evaluation of all cell lineages is important in generating a differential diagnosis. The myeloid-to-erythroid ratio should also be documented. The BM trephine biopsies provide valuable information on the overall cellularity, abnormal localization of progenitors, fibrosis, and presence of extrinsic cells or lymphoid aggregates. In children, lymphoid aggregates are rare and often present in individuals with underlying immune deficiency, autoimmune disorders, or other systemic diseases.
The adult guidelines for BM reporting are applicable to children [15, 19]. The BM report should include BM cellularity described as acellular, reduced, normal, increased, or markedly increased along with quantitative and qualitative comments on all cell lineages – myeloid and erythroid progenitors with M:E ratio, megakaryocytes, lymphocytes, plasma cells – and any abnormal cells such as cells extrinsic to the bone marrow, leukemic blasts, increased number or abnormal mast cells, and the presence of histiocytes displaying hemophagocytosis. It is important to note that histiocytes displaying hemophagocytosis is a nonspecific finding, and while in proper clinical settings, it may be indicative of hemophagocytic lymphohistiocytosis, it may also be seen in a variety of disorders including but not limited to infections, after chemotherapy, and red cell transfusion. The presence of stromal damage, chemotherapy effect, or other stromal changes should be noted. Flow cytometric findings, cytogenetics, and molecular studies should be incorporated in the pathology report and should correlate with the BM morphologic findings. Lastly, the findings should be correlated with the previous results if the studies are carried out to monitor the disease progression or response to therapy. The final BM interpretation should be made in the context of the clinical and preliminary diagnostic findings. The BM diagnosis and/or differential diagnosis, when applicable, should be in accord with the international consensus guidelines .
In summary, the bone marrow is a dynamic organ, and its composition depends on the needs for oxygenation, immune response, and hemostasis. Because such needs vary significantly during fetal life and early childhood, the marrow composition will vary as well. Knowledge of the normal bone marrow findings at various ages is essential for the proper interpretation of bone marrow studies in children.
The author thanks Dr Deborah Fuchs for the critical reading and comments of the manuscript.