Vet Clin Equine 24 (2008) 261–283

Bone Marrow and Lymph

Node Evaluation
Susan J. Tornquist, DVM, PhD
Department of Biomedical Sciences, 200 Magruder Hall, College of Veterinary Medicine,
Oregon State University, Corvallis, OR 97331, USA

Bone marrow
Bone marrow sampling in horses is not a routine diagnostic procedure,
but it is a useful technique in the work-up of abnormal findings from other
procedures, such as the complete blood count (CBC). Because of the unique
tendency of horses to retain maturing erythroid cells in the bone marrow,
even in the face of blood loss or hemolytic anemia, bone marrow examina-
tion may be used more often in the horse than in other species for determin-
ing the cause of anemia. Although performing serial CBC can often
determine whether an anemia is regenerative or not, interpretation may be
complicated by splenic contraction, fluid therapy, and other factors, thus
making bone marrow sampling necessary to classify an anemia. Bone
marrow examination may also be important in diagnosis of neoplasia and
in immune-mediated diseases.

Indications for bone marrow evaluation

Bone marrow evaluation is indicated in cases of persistent nonregenera-
tive anemia, persistent leukopenia, or persistent thrombocytopenia that
cannot be explained based on history, clinical signs, or results from other
diagnostic modalities, including examination of the peripheral blood. In
particular, bone marrow examination should be considered when pancyto-
penia or bicytopenia are present, or abnormal blood cells are found in the
peripheral blood. Other conditions that might warrant bone marrow evalu-
ation include fever of unknown origin, persistent hyperproteinemia, persis-
tent hypercalcemia, a search for infectious agents, and staging of neoplasia
[1,2].

E-mail address: susan.tornquist@oregonstate.edu

0749-0739/08/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.cveq.2008.04.001 vetequine.theclinics.com
262 TORNQUIST

There are few contraindications for bone marrow sampling, but the pres-
ence of hemostatic disorders, such as disseminated intravascular coagulation
or marked thrombocytopenia, warrant special care to prevent hemorrhage
into the thorax or pericardium if the sternum or ribs are sampled. Careful
preparation and use of aseptic technique during bone marrow sampling
should always be used to prevent infection.

Collection and examination of bone marrow

Collection
Bone marrow aspirates or core biopsies, or both, may be collected. The
advantages of aspirates are that they may be more quickly processed and
evaluated, are slightly less invasive than core biopsies, and cellular morphol-
ogy is more easily evaluated. Infectious agents may also be seen more read-
ily. The disadvantages of aspirates are that they may be hemodiluted with
peripheral blood, samples may yield few cells if myelofibrosis or hypocellu-
larity is present, and samples may not contain representative cell types if the
marrow contains only small foci of certain cell types.
Bone marrow may be collected from the sternum, dorsal ribs, or ileum, all
of which contain hematopoietic tissue throughout the life of the horse. The
sternum is the most common site sampled in adult horses, while the tuber
coxae of the ileum is a good site to sample in younger horses [1,2]. Working
at the latter site is easier and generally safer for the person taking the sample.
The horse may be restrained in stocks or sedated for the procedure.
Types of needles that may be used for bone marrow collection include an 18-
gauge spinal needle or an Illinois sternal/iliac bone marrow aspirate needle for
aspirates and a Jamshidi bone marrow biopsy needle for core biopsies [1–3].
The sternum may be palpated or sampled at about the level of the elbows
and at the ventral midline. For collection from the ribs, the dorsal aspect of
the rib has a wider marrow cavity than the ventral aspect and is preferred.
Samples from the ileum may be taken from the middle or the top of the
tuber coxae. Hair is clipped and scrubbed with a disinfectant. The skin,
subcutis, muscle, and periosteum are injected with a local anesthetic, fol-
lowed by another surgical scrub of the site. A small stab incision is made
through the skin and the bone marrow needle is inserted to the level of
the bone. Advancing the needle through cortical bone is accomplished by
steady pressure and back and forth rotation of the needle through the
bone. There is often not an obvious change in pressure when the needle
enters the marrow cavity, but it is generally about 1 cm to 2 cm into bone.
When obtaining a marrow aspirate, the stylet is removed from the needle
at this point and a 12-mL syringe containing a small amount of ethylenedia-
minetetraacetic acid (EDTA) is attached. Aspiration with rapid, strong suc-
tion is necessary to obtain marrow. The latter resembles blood containing
bubbles or small particles. If no material is apparent in the syringe, the stylet
can be replaced and the needle repositioned for another try.
 BONE MARROW AND LYMPH NODE EVALUATION 263

After the syringe is disconnected from the bone marrow needle, the
material may be placed into a Petri dish or a few drops ejected onto a glass
slide. The sample may be examined grossly to look for evidence that marrow
particles and fat are present. If so, pressure should be held over the aspirate
site until bleeding has stopped. If the sample only appears to contain blood,
another aspiration can be performed.
Slides of bone marrow aspirates may be made by transferring particles
from material in the Petri dish to a glass slide and making squash prepara-
tions or by putting a drop of material from the syringe onto a glass slide,
placing another slide perpendicular to the first and gently pulling them apart
[3]. Regardless of which technique is used, multiple slides should be made.
After air-drying, at least one slide should be stained with a Romanowksy-
type stain, such as Wright’s-Giemsa. Longer fixing and staining times are
needed than those used for blood smears. One slide should be examined
at collection, to determine if there is adequate cellularity. If slides are to
be submitted to a diagnostic laboratory for evaluation, most slides should
just be air-dried and sent unstained. If any unclotted fluid marrow sample
remains after making 6 to 10 slides, it can be placed in an anticoagulant
tube (EDTA is preferable) and sent to the laboratory along with the slides.
This sample may be used in analysis by an automated hematology analyzer
or for other additional studies if needed [4].
For core biopsies, the preparation is the same and the biopsy needle is
also inserted until it contacts bone. At this point, the screw cap and stylet
are removed. As with the aspirate, steady pressure and back and forth rota-
tion of the needle is required to get through cortical bone to the marrow cav-
ity. When the appropriate depth has been reached, the needle should be
rotated several times in place to ensure the core detaches from the marrow.
The needle is withdrawn and the core removed. A good core biopsy should
have at least a few centimeters of dark red material, with white cortex usu-
ally also present. The core may be rolled onto a glass slide or two to make
cytologic preparations, and then put into 10% formalin for fixation. Forma-
lin bottles and cytology smears should be transported separately, as forma-
lin fumes can interfere with the staining of cytology specimens.

Special diagnostic procedures

Cytochemical staining of blood and bone marrow cells has been used to
identify their origin, based on enzymes present in these cells, and their reac-
tions with various stains. Stains that are most commonly used to identify
cells of myeloid origin include Sudan black B, choloracetate esterase, alka-
line phosphatase, a-naphthyl acetate esterase, and a-napththyl butyrate
esterase. Nonspecific esterases are generally monocyte markers. There is
some variability in these staining patterns [5].
The limitations of cytochemistry include the fact that immature, neoplas-
tic, or undifferentiated cells may stain differently than normal mature cells,
making interpretation of cytochemical staining problematic at times. In
264 TORNQUIST

addition, cytochemistry cannot be used to identify lymphocyte populations,

though positive staining with certain patterns suggests cells are not lym-
phoid [5,6].
Immunophenotyping of cells, using monoclonal antibodies to specific
cell-surface antigens to identify the cell lineage, is a more definitive diagnos-
tic technique and is being used more frequently in veterinary medicine.
Immunophenotyping may be performed on cell preparations, such as blood
and bone marrow smears, on fixed and paraffin-embedded histologic sec-
tions, and on individual cells using flow cytometry [6–8].
Monoclonal antibodies to equine antigens, including CD3, CD4, CD8,
CD79, and myeloperoxidase, have been developed and validated and are
currently being used to identify undifferentiated cell types in equine bone
marrow as well other tissues (W. Vernau, personal communication, 2007).
In addition to determining whether a neoplastic process in the blood or
bone marrow is lymphoid or myeloid, immunophenotyping can indicate
whether lymphocytes are of B cell or T cell origin. As this diagnostic tech-
nique becomes more commonly used, information on the expected behavior
of equine hematopoietic neoplasia will be generated.
Another technique that may be used in characterizing equine cells is the
polymerase chain reaction to detect immunoglobin or T cell receptor gene
rearrangement, to identify clonal populations of lymphocytes [9,10]. Using
this technique, a benign reactive proliferation of lymphocytes can be distin-
guished from a neoplastic population. It is currently being performed on
equine samples at a few locations (W. Vernau, personal communication,
2007). Along with immunophenotyping, this type of analysis should prove
invaluable in accurately classifying, and thus clarifying, the nature and be-
havior of equine lymphoid neoplasms.

Interpretation
Normal marrow
Bone marrow samples are most often evaluated by an experienced veteri-
nary clinical pathologist, and are thus usually submitted to a diagnostic labo-
ratory. They should always be interpreted within the context of current CBC
findings, so current results or a peripheral blood sample should always be
included with a bone marrow sample submission to a laboratory. Evaluation
of a bone marrow sample includes: judging its overall cellularity and preserva-
tion of cell morphology; assessing the three major cell lines, erythroid, mye-
loid, and megakaryocytes; and appraising other cell types that are present.
Review of overall cellularity begins using the low power (4–20) objec-
tives. A good bone marrow sample from a normal horse contains approxi-
mately equal amounts of hematopoietic cells and fat, which appears as
variably-sized clear vacuoles (Fig. 1). Samples with fewer hematopoietic
cells are subjectively hypocellular and those with greater proportions of
hematopoietic cells are subjectively hypercellular.
 BONE MARROW AND LYMPH NODE EVALUATION 265

Fig. 1. Normocellular bone marrow from a horse. Adipose tissue and hematopoietic cells are
both present (Wright’s-Giemsa stain, original magnification  200).

Megakaryocytes are assessed on low power as they are large cells (aver-
aging about 100 mm in diameter), present in relatively low numbers in
health, and often are not evenly distributed across a slide (Figs. 2–4). There
are most often one to four per average 10 field [3,4].
Erythroid, myeloid, and other cell types are evaluated at higher power
(40–100 objectives). In some cases, a myeloid-to-erythroid ratio (M:E)
is objectively calculated by classifying each of 500 nucleated cells as myeloid
or erythroid. In other cases, the ratio is determined more subjectively. The
range for the myeloid-to-erythroid ratio reported in normal horses is quite
wide (0.5–2.4), which makes it difficult to detect small changes in the ratio
that could be associated with disease [3,11]. See Figs. 5 and 6 for examples.
Both cell lines are then assessed for evidence of normal, orderly matura-
tion. For the erythroid series, this is indicated by the presence of

Fig. 2. Megakaryocyte, equine bone marrow (Wright’s-Giemsa stain, original magnification  1,000).
266 TORNQUIST

Fig. 3. Osteoclast, equine bone marrow. These are not commonly seen in bone marrow
samples, but may be present when bone remodeling is occurring. They are similar in size to
megkaryocytes and also multi-nucleated, but nuclei are separate and discrete in contrast to
megakaryocytes, which have nuclei that are multilobed (Wright’s-Giemsa stain, original
magnification  1,000).

progressively greater numbers of the mature erythroid cells, including rubri-

cytes and metarubricytes, with generally fewer than 5% rubriblasts and
prorubricytes (Fig. 7). Similarly, in the myeloid line, the great majority of
granulocytes, particularly neutrophils, are the more mature cells, including
progressively greater numbers of metamyelocytes, bands, and segmented
cells (Fig. 8).
Other cell types present in bone marrow from healthy horses include
small lymphocytes, which may comprise from about 2% to 10% of the nu-
cleated cells, monocytes and macrophages, and mitotic figures (Fig. 9) [3,5].

Fig. 4. Equine bone marrow. Three megakaryocytes showing that they may be unevenly
distributed in the marrow (Wright’s-Giemsa stain, original magnification  200).
 BONE MARROW AND LYMPH NODE EVALUATION 267

Fig. 5. Equine bone marrow with normal M:E ratio (Wright’s-Giemsa, original magnifica-
tion  500).

Red blood cell abnormalities

Regenerative anemia is more difficult to assess in the peripheral blood of
horses than in other species, because equine erythrocytes are generally not
released from the bone marrow until they are fully mature. A regenerative
response may include an increase in the peripheral blood mean corpuscular
volume (MCV) of about 10 fL to 15 fL, but this may not appear significant
unless the base-line MCV is known. Increased red cell distribution width, an
indicator of variability in red blood cell size, is also seen in regeneration
[4,11].
Peripheral blood reticulocyte counts, performed by some automated he-
matology analyzers, may be more sensitive than those performed by manual
methods. This is because of both the greater number of erythrocytes counted
(O40,000 by automated methods versus 1,000 by manual methods) and the

Fig. 6. Equine bone marrow with normal M:E ratio (Wright’s-Giemsa stain, original magnifi-
cation  1,000).
268 TORNQUIST

Fig. 7. Erythroid maturation, equine bone marrow (Wright’s-Giemsa stain, original magnifica-
tion  1,000).

methodology used in automated methods that identifies reticulocytes, which

is based on light absorbance and scatter characteristics. The reticulocyte
count by an automated hematology analyzer was reported to be more
than two times greater than the manual count in a report of a case of im-
mune-mediated hemolytic anemia in a horse [12].
Reticulocytes in the bone marrow may also be assessed by manual or
automated methods, although neither is currently reported often in the lit-
erature. Reticulocytes in bone marrow may be demonstrated by new meth-
ylene blue staining and are reported to comprise up to 3% of erythrocytes in
normal horses, with as many as 60% in an intensely regenerative response
[4,11].
A study of experimentally-induced marked blood loss anemia that in-
volved removal of about 42% of blood volume in six horses showed

Fig. 8. Active equine bone marrow, showing normal erythroid and myeloid cell maturation
(Wright’s-Giemsa stain, original magnification  500).
 BONE MARROW AND LYMPH NODE EVALUATION 269

Fig. 9. Active equine bone marrow, with two large mitotic figures (Wright’s-Giemsa stain,
original magnification  1,000).

a marked increase in bone marrow reticulocytes at 7-days after blood loss,

with a return to base-line values on day 40, but there was significant individ-
ual variation in this response [4]. The myeloid-to-erythroid ratio dropped
from a base-line mean of 0.75 to 0.48 at 7 days, illustrating that changes
in this ratio may be difficult to interpret because of significant base-line var-
iability between individual horses (Fig. 10).
In another study of bone marrow response to experimentally induced
blood loss in five horses, the lowest average myeloid-to-erythroid ratio of
0.20 was observed at 9 days after blood loss and remained below baseline
until the end of the study at 31 days [11].
In general, a regenerative response in the bone marrow has been charac-
terized by a mild increase in the number of early erythroid cells (rubriblasts

Fig. 10. Active erythropoiesis in bone marrow in regenerative anemia, showing predominance
of erythroid cells and decreased M:E ratio (Wright’s-Giemsa stain, original magnification 
500).
270 TORNQUIST

and prorubricytes), with a relatively greater increase seen in rubricytes and

metarubricytes [3]. In the study cited above [11], this pattern was character-
istic of the time period 14 to 31 days after blood loss, while the earlier
erythroid cells were greatly increased 3 to 9 days after blood loss. There is
a paucity of studies showing the bone marrow regenerative response in
immune-mediated anemia in horses, but it is generally thought to be greater
than that seen in blood loss anemia [12]. Bone marrow evidence of immune-
mediated anemia can include increased erythrophagocytosis and increased
macrophage iron, which can be demonstrated by Prussian blue staining [3].
In contrast, evidence of a nonregenerative anemia on bone marrow aspi-
rate smears includes a decrease in or lack of erythroid precursors, relative
scarcity of earlier cells such as prorubricytes, and an increased myeloid-to-
erythroid ratio in the bone marrow. Finding neoplastic populations of cells,
or fibrosis, or necrosis replacing erythroid cells, can provide further clarifi-
cation of the reason for a nonregenerative anemia. Equine infectious anemia
virus infection is most often associated with decreased erythroid production,
although immune-mediated destruction of peripheral erythrocytes and
increased erythrophagocytosis in the bone marrow may also be observed
[13].

White blood cell abnormalities

Bone marrow examination is rarely performed to investigate leukopenia
alone, unless abnormal leukocytes are seen in peripheral blood smears. Bone
marrow typically responds to an increased demand for neutrophils by mark-
edly increasing production of these cells, and this increase may be seen in the
bone marrow several days before it is reflected in the peripheral blood. Toxic
changes, such as Döhle bodies, toxic granulation, and cytoplasmic baso-
philia or vacuolation may be present in neutrophils exposed to inflamma-
tory mediators during their development in the bone marrow [3,5].
There is one report of a family of eight related Standardbreds that were
intermittently neutropenic, with seven of them also intermittently thrombo-
cytopenic [14]. Examination of bone marrow showed these horses to have
myeloid hypoplasia with decreased myeloid-to-erythroid ratios. Some also
had erythroid hypoplasia, though maturation of the erythroid line appeared
orderly. Megakaryocytes were decreased or absent in the bone marrow of
seven of the eight horses. A familial basis for this disorder was suspected
and bone marrow culture results suggested a bone marrow microenviron-
ment or growth factor defect might be present.

Platelet abnormalities
In the face of thrombocytopenia, increased numbers of bone marrow
megakaryocytes are associated with peripheral platelet destruction or con-
sumption, while decreased numbers are evidence of a production problem.
Though a moderate to marked thrombocytopenia is a consistent finding
early in the course of infection with equine infectious anemia virus,
 BONE MARROW AND LYMPH NODE EVALUATION 271

megakaryocyte density in the bone marrow is neither increased nor de-

creased, suggesting that ineffective platelet production may be occurring [15].

Bone marrow neoplasia

Neoplasia in the bone marrow can arise from lymphoid or nonlymphoid
cells and may be classified as a lymphoproliferative or myeloproliferative
disorder. Bone marrow neoplasia is most often, although not always, also
associated with circulating neoplastic cells in the peripheral blood [3].

Lymphoproliferative disorders
The bone marrow may contain neoplastic lymphoid cells that represent
primary leukemia if the bone marrow is the primary site, or secondary
leukemia if neoplastic lymphocytes have infiltrated the bone marrow in an
advanced stage of lymphoma. Primary lymphocytic leukemia is character-
ized by a lack of neoplastic lymphocytes in other tissues, as well as a diffuse
distribution of these cells in the bone marrow [6,16]. In secondary lympho-
cytic leukemia, there is usually extensive involvement of other tissues and
more focal aggregates of neoplastic cells in the marrow [16]. Primary lym-
phoid leukemias may also be classified as acute lymphocytic leukemia or
chronic lymphocytic leukemia. These are differentiated both by the appear-
ance of the neoplastic lymphocytes and by the typical clinical course of the
disease.
Reports of primary acute lymphoid leukemia in horses indicate that pan-
cytopenia is a very common finding and that circulating lymphoblasts are
sometimes, but not consistently, found (Figs. 11 and 12) [16–19]. Bone mar-
row examination is usually diagnostic, with large numbers of hematopoietic
blast cells and few erythroid cells, myeloid cells, and megakaryocytes pres-
ent. In some cases, a battery of cytochemical stains on bone marrow have
been used to identify the blast cells as nonmyeloid in origin [16,19].

Fig. 11. Circulating lymphoblasts, lymphoid leukemia. (Wright’s-Giemsa stain, original magni-
fication  200).
272 TORNQUIST

Fig. 12. Circulating lymphoblasts, lymphoid leukemia. (Wright’s-Giemsa stain, original magni-
fication  1,000).

Chronic lymphocytic leukemia has been rarely reported in horses, with

only one report of two cases [20]. Both horses had marked lymphocytosis
with circulating lymphocytes that were essentially small and mature in
appearance. Both had relatively large numbers of lymphocytes in the bone
marrow, with otherwise normal numbers of other hematopoietic cells pres-
ent in one horse, and decreased numbers of erythroid, myeloid, and mega-
karyocytic cells in the other. Based on flow cytometric analysis of the
peripheral blood, one case was determined to be of B cell and the other
of T cell origin.
Plasma cell myeloma is a neoplastic proliferation of plasma cells that
usually involves the bone marrow. There are several reports of plasma cell
myeloma in horses [21–23], with either diffuse or multifocal involvement
of the bone marrow described. Anemia is the most consistent clinical finding
and pancytopenia was present in some cases. Hyperglobulinemia is also
common. The neoplastic plasma cells in the bone marrow may appear
morphologically normal or abnormal. Because plasma cells may also be in
the bone marrow in reaction to antigenic stimulation, demonstration that
a population is clonal would be very helpful in establishing a diagnosis of
plasma cell myeloma [21].

Myeloproliferative disorders
Myeloproliferative disorders include myeloid leukemia and malignant
histiocytosis. Myeloid leukemias may affect the granulocytic, monocytic,
erythroid, or megakaryocytic cell lines, or more than one cell line, such as
in myelomonocytic leukemia. Reports of megakaryocytic or erythrocytic
leukemia (absolute polycythemia) in horses are not found in the literature.
There are scattered reports of acute myelocytic, chronic myelocytic, acute
monocytic, acute myelomonocytic, and myeloblastic leukemias in horses
in the literature, but these few reports do not necessarily reflect the number
 BONE MARROW AND LYMPH NODE EVALUATION 273

of cases that occur [24–28]. In these case reports, the horses were nearly
always anemic and often thrombocytopenic and leukopenic as well. Abnor-
mal circulating blast cells were seen in some cases, but not others. Bone
marrow samples in these cases were usually hypercellular, with a high mye-
loid-to-erythroid ratio and large numbers of immature blast cells present
[24,25,27].
In some of the reported cases, cytochemical stains including Sudan black
B, alkaline phosphatase, chloroacetate esterase, and a-naphthyl butyrate
esterase were used to confirm the identity of the blast cells as myeloid
[24,25,27].
Malignant histiocytosis, a proliferation of atypical histiocytes, has been
reported in one horse [29], a 16-month-old Arabian filly. She presented
with pancytopenia, with normal lymphocyte and monocyte numbers in
the peripheral blood. Aspirates of both enlarged superficial nodes and
bone marrow showed similar populations of very large, atypical histiocytic
cells, with evidence of phagocytosis of erythrocytes. These features are
similar to those seen in malignant histiocytosis of human beings and dogs.
Cytochemical staining confirmed the cells to be of monocytic lineage.
Myelodysplastic syndrome (MDS) refers to a group of disorders in which
there are peripheral blood cytopenias, as well as atypical morphologic fea-
tures of any of the hematopoietic cell lines. This group of disorders arise
from mutations in hematopoietic stem cells that lead to clonal proliferation
[26,30]. The morphologic atypias may include abnormal nuclear shapes
(Fig. 13), asynchrony between maturity of the nucleus and cytoplasm
(Fig. 14), and fragmented nuclei. Marked regenerative anemia is a very com-
mon feature. Bone marrow, however, is often hypercellular with immature
hematopoietic cells present that are not predominantly (O30%) blast cells.
In human beings, progression of MDS to acute myeloid leukemia is com-
mon, and this may occur in other species [5]. There is one case report of

Fig. 13. Equine bone marrow, myelodysplastic syndrome. Note abnormal nuclear shape
(arrow) (Wright’s-Giemsa stain, original magnification  1,000).
274 TORNQUIST

Fig. 14. Equine bone marrow, myelodysplastic syndrome, showing nuclear to cytoplasmic
asynchrony (arrows) (Wright’s-Giemsa stain, original magnification  500).

MDS in a horse [31]. The horse had a marked pancytopenia that included
a profound anemia with some macrocytes present. The bone marrow had
low normal cellularity with erythroid precursor cell morphologic abnormal-
ities typical of MDS. The horse was euthanized so it is not known whether
the disease would have progressed to leukemia.

Aplastic pancytopenia
Pancytopenia accompanied by a bone marrow that does not contain
developing hematopoietic cells is known as aplastic pancytopenia or aplastic
anemia. Although underlying causes, such as infection, toxins, and immune
mechanisms have been identified in some aplastic pancytopenia cases in
other species, this condition is often idiopathic. There are three cases
reported in horses [17–19]. In one of these, the condition was responsive
to steroids and an immune-mediated cause was suspected [19]; in the other
two, no etiology was identified [17,18]. In general, bone marrow examina-
tion shows a hypoplastic marrow, which may contain low numbers of lym-
phocytes and stromal cells but lacks evidence of developing erythroid,
myeloid, and megakaryocytic cells (Fig. 15).

Lymph nodes
Lymph nodes may be enlarged because of hyperplasia or reaction to an
antigenic stimulus, by inflammation, or by neoplasia, either primary or met-
astatic. Because lymphocytes usually exfoliate well to needle aspiration,
cytology may be a quick, minimally invasive, and useful diagnostic tech-
nique to distinguish between causes of equine lymphadenomegaly. In partic-
ular, lymph node cytology can most often distinguish between a primarily
inflammatory process and neoplasia. Some features of equine lymphoma,
 BONE MARROW AND LYMPH NODE EVALUATION 275

Fig. 15. Hypoplastic bone marrow, core biopsy (Hematoxylin-eosin stain, original magnifica-
tion  200).

however, can complicate the interpretation of cytologic findings and

histopathologic examination of an enlarged node is sometimes necessary
for a definitive diagnosis [7,32].

Indications for lymph node cytology

For the most part, lymph nodes are sampled only when they are palpably
enlarged. In horses, the submandibular, cervical, retropharyngeal, and su-
perficial inguinal lymph nodes are among the more common peripheral
lymph nodes examined cytologically [33]. Internal nodes may be sampled
during surgery. Clinical signs that may also be an indication for lymph
node sampling include focal swelling, localized edema, anorexia or dyspha-
gia (if retropharyngeal or submandibular lymph nodes are enlarged), fever,
respiratory distress, pleural effusion, weight loss, and straining to defecate
(if anorectal lymph nodes are enlarged). In addition, lymph nodes may be
sampled to further characterize neoplasia that has been diagnosed at a differ-
ent site.

Collection and processing

Samples are most often obtained by an aspiration technique or a nonaspi-
ration or stab technique. Sedation or stocks are not typically used, as the
procedure is similar to giving an injection. No special skin preparation is
needed. The node to be sampled is stabilized between the thumb and fore-
finger and a 21- to 25-gauge needle on a 5-mL to 12-mL syringe as inserted
into the node.
For the aspiration technique, the plunger is then quickly withdrawn to
pull cells out of the node. The needle may be redirected without pulling it
out of the node and negative pressure applied again to obtain sample
276 TORNQUIST

from different areas of the node. If blood appears in the syringe, the proce-
dure should be stopped and a new aspirate performed, as blood contamina-
tion of a lymph node sample complicates interpretation. Before the needle is
withdrawn from the node, negative pressure is released. The needle should
then be taken off the syringe and air drawn into the syringe before the needle
is replaced. The aspirated material is ejected onto a glass slide with some
force [33]. It is often possible to make several slides from the material in
one aspirate.
The cells then must be spread out to form a monolayer that will allow
evaluation of individual cells. As lymphoid cells tend to rupture easily,
this spreading procedure should be done gently. For relatively ‘‘thin’’ spec-
imens, it can be performed similarly to making a blood smear: that is, by
holding a second slide at an angle to the first slide, pulling it into the aspi-
rate, and spreading it forward rapidly. For ‘‘thicker’’ specimens, a squash
preparation, using a second slide placed over the aspirate and spreading
it, may also be used.
The nonaspiration technique often results in less peripheral blood con-
tamination and cell rupture. With this technique, a 3-mL to 12-mL syringe
is filled with air and a 20- to 22-gauge needle is attached. The node is stabi-
lized and the syringe and needle are used to make multiple, quick stabbing
actions into the lymph node without pulling out of the node. [33]. Cells are
collected by being packed into the needle, and no aspiration is performed.
Cells are ejected onto a slide and the slide preparation is the same as for
the aspirated sample.
If lymph nodes are surgically excised, impressions can be made after
blood is blotted off. The cut surface is gently pressed against a clean glass
slide without smearing the sample.
A technique that is not often used, but can be very useful if immunocy-
tochemistry or flow cytometric analysis for immunophenotyping may be
needed, is to collect an aspirate into buffered saline to make a cell suspen-
sion. Multiple cytocentrifuged preparations can be made from this cell
suspension, allowing for immunostaining with a panel of different monoclo-
nal antibodies, as well as performing positive and negative controls on slides
of equivalent cellularity.
Staining for cytologic examination is performed on air-dried slides.
Wright’s stain, Wright’s-Giemsa, Diff-Quik, or other Romanowksy stains
are all suitable for staining lymph node samples. If the sample is thick, longer
staining times should be used. If slides are to be submitted to a diagnostic lab-
oratory for evaluation, at least some unstained slides should be submitted.

Interpretation
It is important to only evaluate intact lymphocytes in a cytology sample.
Ruptured cells cannot be accurately assessed for size and when the cells rup-
ture, nucleoli become visible [33]. This could lead to over-counting of large
 BONE MARROW AND LYMPH NODE EVALUATION 277

lymphoblasts. It is thought that lymphoblasts are more fragile and

easily ruptured than small lymphocytes. For this reason, evaluating
a lymph node sample with large numbers of ruptured cells can lead to
misinterpretation.

Normal lymph node

Cells present in normal lymph nodes include small, medium, and large
lymphocytes. Small lymphocytes are round cells that are smaller than neu-
trophils. They have a relatively large round nucleus, often with an indented
or cleft shape, dense chromatin, no nucleolus, and a thin rim of blue cyto-
plasm. In normal lymph nodes, small lymphocytes comprise roughly 75% to
95% of the cells. Medium lymphocytes are similar in size to neutrophils.
They have more stippled chromatin than the small lymphocytes and have
slightly more cytoplasm that is also a moderate blue. A nucleolus may or
may not be apparent. These cells comprise about 5% to 15% of the cells
in a normal lymph node. Large lymphocytes or lymphoblasts are larger
than neutrophils and have a nucleus with stippled or granular chromatin.
One to several nucleoli are present. Large lymphocytes make up about
5% to 10% of the cells in a normal lymph node (Fig. 16).
Normal lymph nodes may also have few plasma cells, macrophages, neu-
trophils, and eosinophils. There are also usually fragments of lymphocyte
cytoplasm, called lymphoglandular bodies, in lymph node smears [33].
These are small, round to irregular basophilic, sometimes granular struc-
tures (Fig. 17).

Reactive lymph node

When lymph nodes are undergoing an immune response to an antigen,
the cell types are usually similar to those seen in the normal lymph node,
but there are often increased numbers of plasma cells as well as increased
proportions of medium and large-sized lymphocytes. The percentage of

Fig. 16. Normal equine lymph node (Wright’s-Giemsa stain, original magnification  1,000).
278 TORNQUIST

Fig. 17. Normal equine lymph node (Wright’s-Giemsa stain, original magnification  500).

small lymphocytes is relatively reduced, but still greater than 50% of the
cells. This is a very nonspecific response and may be associated with local-
ized or systemic lymphadenomaegaly (Fig. 18).

Lymphadenitis
Increased numbers of inflammatory cells are seen in a lymph node that is
itself inflamed or is draining an area of inflammation. The type of inflamma-
tion is determined by the predominant inflammatory cell type. A lymph
node with a predominance of neutrophils is said to have neutrophilic, sup-
purative, or purulent inflammation. If bacteria are seen, the inflammation is
also termed ‘‘septic.’’ This is common in cases of strangles, caused by Strep-
tococcus equi, that infect the lymph node, causing abscessation (Fig. 19).
Increased numbers of macrophages in a lymph node sample are typical of
chronic inflammatory processes, as well as infection with some fungi,

Fig. 18. Reactive equine lymph node. Increased numbers of plasma cells and medium-sized
lymphocytes are present (Wright’s-Giemsa stain, original magnification  500).
 BONE MARROW AND LYMPH NODE EVALUATION 279

Fig. 19. Equine lymph node with normal cell population replaced by degenerative neutrophils
and Streptococcus equi bacteria, often in chains (Wright’s-Giemsa stain, original magnifica-
tion  1,000).

mycobacteria, and protozoans. In some of these cases, there are also in-
creased neutrophil numbers, resulting in pyogranulomatous inflammation.
Granulomatous inflammation, typified by the presence of large, multinucle-
ated cells, is also seen in some fungal infections.
It is also not uncommon to have increased macrophages or giant cells in
equine lymphoma. In one study, 8 of 34 equine lymphomas contained signif-
icant numbers of macrophages and giant cells [8]. These cells may be present
because of tumor-induced inflammation or necrosis or may be Langhans’
cells that are responding to tumor-induced cytokines. This is a situation
in which lymph node cytology might be misleading, as granulomatous
inflammation would be diagnosed rather than lymphoma.
Eosinophilic lymphadenitis, with or without mast cells, may occur in
hypersensitivity responses and parasitic and some fungal infections
(Fig. 20) [33].

Lymphoma
Lymphoma has a low prevalence in horses, but is one of the most com-
mon equine malignant neoplasms [28,32]. Peripheral lymph nodes are often
not involved.
Lymphoma is easily recognized cytologically when there is a homogenous
population of large lymphoblasts (Fig. 21). However, in many cases, the
lymph node contains a heterogenous population of lymphocytes and the rel-
ative proportions of these different lymphocyte types becomes important in
interpretation of the sample (Figs. 22 and 23). In most cases of lymphoma,
more than 50% of the cells are medium-sized lymphocytes or large lympho-
blasts [33]. However, lymphoma of small lymphocytes can also occur and is
usually of a T cell type [7]. This form of lymphoma cannot be diagnosed
cytologically. In cases of suspected lymphoma, with nondiagnostic cytologic
280 TORNQUIST

Fig. 20. Equine lymph node with eosinophilic lymphadenitis and mast cells. This would be
a typical reaction to hypersensitivity (Wright’s-Giemsa stain, original magnification  1,000).

findings, excision and histopathologic examination of a lymph node are

indicated. Lymphomas typically efface the normal lymph node architecture
[8].
Use of immunophenotyping has revealed some interesting features of
equine lymphomas. A recent study of 37 equine lymphomas classified
70% as T cell and 19% as B cell, with 11% unclassified [8]. Of the B cell lym-
phomas, five of seven (70%) were classified as T cell-rich B cell lymphomas.
These findings contrast with an earlier study of 31 equine lymphomas that
classified 77% as B cell and 20% T cell in origin [7]. The B cell lymphomas
included 33% T cell-rich B cell tumors. In the latter tumors, the T lympho-
cytes were determined to be non-neoplastic based on normal morphologic
appearance, lack of reaction with markers of proliferation, and low mitotic
index. T cell-rich B cell tumors are considered to be neoplastic B cell tumors
with significant numbers (from 10%–90% in human beings) of reactive
T lymphocytes present [8]. Clonality assays of these tumors show that the

Fig. 21. Equine lymph node, lymphoblastic lymphoma. The majority of cells are large, with one
to several prominent nucleoli (Wright’s-Giemsa stain, original magnification  500).
 BONE MARROW AND LYMPH NODE EVALUATION 281

Fig. 22. Equine lymph node with a large granular lymphocytic lymphoma. These medium to
large-sized lymphocytes, with few small magenta granules, were found in multiple organs as
well as in peritoneal fluid (Wright’s-Giemsa stain, original magnification  500).

B lymphocytes are a clonal population and the T lymphocytes are a poly-

clonal population (W. Vernau, personal communication, 2007). Histiocytic
inflammation was commonly present in both studies [7,22].
These studies of equine lymphomas, while differing significantly in their
findings, illustrate the value of characterizing these tumors by lymphocyte
type, which is necessary for accurate prognosis. It is likely that immunophe-
notyping, both by immunocytochemistry and by flow cytometry, will
become more commonly used in evaluation of equine lymph nodes.

Metastatic neoplasia
Metastasis of nonlymphoid tumors can be diagnosed when there are cells
of a type that are not normally present in a lymph node, or when there are

Fig. 23. Equine lymph node with histiocytic lymphoma. Notice multiple large cells with irreg-
ularly-shaped nuclei and abundant, lightly basophilic cytoplasm (Wright’s-Giemsa stain, orig-
inal magnification  200).
282 TORNQUIST

larger than expected numbers of some cells. Malignant epithelial cell tumors
are probably the most commonly diagnosed type in lymph nodes [33]. The
epithelial cells tend to occur in clusters, but may also be single. They are
often large and angular and have features of malignancy, such as marked
anisocytosis, anisokaryosis, and prominent nucleoli. Malignant sarcomas
are rarely found in lymph nodes. Mast cell tumor metastasis to lymph
node has been reported in a single case [34]. Malignant melanoma is associ-
ated with metastasis to lymph nodes fairly frequently [35].

Summary
Evaluation of equine bone marrow and lymph node samples can provide
the definitive diagnosis in some cases, and may provide useful information
in other cases. With some newer techniques, including immunophenotyping
of cells and clonality assays, we have the capability to more precisely iden-
tify cells, both as to origin and malignancy. Use of these techniques on
equine bone marrow and lymph node samples, and compiling of the data,
will eventually provide invaluable information about equine neoplasia
that will greatly improve our ability to predict tumor behavior and response
to therapy.

References
[1] Russell KE, Sellon DC, Grindem CB. Bone marrow in horses: indications, sample handling,
and complications. Compend Contin Educ Vet 1994;16:1359–65.
[2] Sellon DC. How to obtain a diagnostic bone marrow sample from the sternum of an adult
horse. Proceedings of the American Association of Equine Practitioners 2006;52:621–5.
[3] Latimer KS, Andreasen CB. Bone marrow. In: Cowell R, Tyler R, editors. Diagnostic
cytology and hematology of the horse. 2nd edition. St. Louis (MO): Mosby; 2002.
p. 217–26.
[4] Cooper C, Sears W, Bienzle D. Reticulocyte changes after experimental anemia and eryth-
ropoietin treatment of horses. J Appl Physiol 2005;99:915–21.
[5] Thrall MA, Weiser G, Jain N. Laboratory evaluation of bone marrow. In: Thrall MA,
editor. Veterinary hematology and clinical chemistry. Philadelphia: Lippincott Williams
& Wilkins; 2004. p. 149–78.
[6] McClure JT. Leukoproliferative disorders in horses. Vet Clin North Am Equine Pract 2000;
16:165–82.
[7] Kelley LC, Mahaffey EA. Equine malignant lymphomas: Morphologic and immunohisto-
chemical classification. Vet Pathol 1998;35:241–52.
[8] Meyer J, DeLay J, Bienzle D. Clinical, laboratory, and histopathologic features of equine
lymphoma. Vet Pathol 2006;43:914–24.
[9] Burnett RC, Vernau W, Modiano JF, et al. Diagnosis of canine lymphoid neoplasia using
clonal rearrangements of antigen receptor genes. Vet Pathol 2003;40:32–41.
[10] Keller RL, Avery AC, Burnett RC, et al. Detection of neoplastic lymphocytes in peripheral
blood of dogs with lymphoma by polymerase chain reaction for antigen receptor gene
rearrangement. Vet Clin Pathol 2004;33:145–9.
[11] Malikides N, Kessell A, Hodgson JL, et al. Bone marrow response to large volume blood
collection in the horse. Res Vet Sci 1999;67:285–93.
 BONE MARROW AND LYMPH NODE EVALUATION 283

[12] Weiss DJ, Moritz A. Equine immune-mediated hemolytic anemia associated with
Clostridium perfringens infection. Vet Clin Pathol 2003;32:22–6.
[13] Sellon DC, Fuller FJ, McGuire TC. The immunopathogenesis of equine infectious anemia.
Virus Res 1994;32:111–38.
[14] Kohn CW, Swardson C, Provost P, et al. Myeloid and megakaryocytic hypoplasia in related
Standardbreds. J Vet Intern Med 1995;9:315–23.
[15] Wardrop KJ, Baszler TV, Reilich E, et al. A morphometric study of bone marrow megakar-
yocytes in foals infected with equine infectious anemia virus. Vet Pathol 1996;33:222–7.
[16] Lester GD, Alleman AR, Raskin RE, et al. Pancytopenia secondary to lymphoid leukemia in
three horses. J Vet Intern Med 1993;7:360–3.
[17] Angel KL, Spano JS, Schumacher J, et al. Myelophthisic pancytopenia in a pony mare. J Am
Vet Med Assoc 1991;198:1039–42.
[18] Bergen PC. Aplastic anemia in a horse. J Am Vet Med Assoc 1981;179:1400–2.
[19] Lavoie JP, Morris DD, Zinkl JG, et al. Pancytopenia caused by bone marrow aplasia in
a horse. J Am Vet Med Assoc 1987;191:1462–4.
[20] Dascanio JJ, Zhang CH, Antczak DF, et al. Differentiation of chronic lymphocytic leukemia
in the horse. J Vet Intern Med 1992;6:225–9.
[21] Edwards DF, Parker JW, Wilkinson JE, et al. Plasma cell myeloma in the horse. J Vet Intern
Med 1993;7:169–76.
[22] Kim DY, Taylor HW, Eades SC, et al. Systemic AL amyloidosis associated with multiple
myeloma in a horse. Vet Pathol 2005;42:81–4.
[23] Pusterla N, Stacy BA, Vernau W, et al. Immunoglobulin A monoclonal gammopathy in two
horses with multiple myeloma. Vet Rec 2004;155:19–23.
[24] Buechner-Maxwell V, Zhang C, Robertson J, et al. Intravascular leukostasis and systemic
aspergillosis in a horse with subleukemic acute myelomonocytic leukemia. J Vet Intern
Med 1994;8:258–63.
[25] Clark P, Cornelisse CJ, Schott HC, et al. Myeloblastic leukaemia in a Morgan horse mare.
Equine Vet J 1999;31:446–8.
[26] McManus PM. Classification of myeloid neoplasms; a comparative review. Vet Clin Pathol
2005;34:189–212.
[27] Ringger NC, Edens L, Bain P, et al. Acute myelogenous leukaemia in a mare. Aust Vet J
1997;75:329–31.
[28] Savage CJ. Lymphoproliferative and myeloproliferative disorders. Vet Clin North Am
Equine Pract 1998;14:563–78.
[29] Lester GD, Alleman AR, Raskin RE, et al. Malignant histiocytosis in an Arabian filly.
Equine Vet J 1993;25:471–3.
[30] Weiss DJ. Recognition and classification of dysmyelopoiesis in the dog: a review. J Vet Intern
Med 2005;19:147–54.
[31] Durando MM, Alleman AR, Harvey JW. Myelodysplastic syndrome in a Quarter Horse
gelding. Equine Vet J 1994;26:83–5.
[32] Schneider D. Lymphoproliferative and myeloproliferative disorders. In: Robinson NE,
editor. Current therapy in equine medicine 5. Philadelphia: Saunders; 2003. p. 359–62.
[33] Cowell RL, Tyler RD, Dorsey KE, et al. Lymph nodes. In: Cowell R, Tyler R, editors.
Diagnostic cytology and hematology of the horse. 2nd edition. St. Louis (MO): Mosby;
2002. p. 99–106.
[34] Riley CB, Yovich JV, Howell JM. Malignant mast cell tumours in horses. Aust Vet J 1991;10:
346–7.
[35] MacGillivray KC, Sweeney RW, Del Piero F. Metastatic melanoma in horses. J Vet Intern
Med 2002;16:452–6.