BMC Cancer BioMed Central
ssResearch article Open Acce
The bone marrow aspirate and biopsy in the diagnosis of 
unsuspected nonhematologic malignancy: A clinical study of 19 
cases
Fahir Ozkalemkas*1, Rıdvan Ali1, Vildan Ozkocaman1, Tulay Ozcelik1, 
Ulku Ozan1, Hulya Ozturk2, Ender Kurt3, Turkkan Evrensel3, Omer Yerci2 
and Ahmet Tunali1
Address: 1Division of Hematology, Department of Internal Medicine Uludag University School of Medicine, Uludag University Hospital, Bursa, 
Turkey, 2Department of Pathology, Uludag University School of Medicine, Uludag University Hospital, Bursa, Turkey and 3Division of Medical 
Oncology, Department of Internal Medicine Uludag University School of Medicine, Uludag University Hospital, Bursa, Turkey
Email: Fahir Ozkalemkas* - fahir@uludag.edu.tr; Rıdvan Ali - ridvanali@uludag.edu.tr; Vildan Ozkocaman - vildanoz@uludag.edu.tr; 
Tulay Ozcelik - tulayoz@uludag.edu.tr; Ulku Ozan - ulkuoz@uludag.edu.tr; Hulya Ozturk - hulyaozturk@uludag.edu.tr; 
Ender Kurt - ekurt@uludag.edu.tr; Turkkan Evrensel - evrensel@uludag.edu.tr; Omer Yerci - yerci@uludag.edu.tr; 
Ahmet Tunali - atunali@uludag.edu.tr
* Corresponding author    
Published: 01 November 2005 Received: 18 May 2005
Accepted: 01 November 2005
BMC Cancer 2005, 5:144 doi:10.1186/1471-2407-5-144
This article is available from: http://www.biomedcentral.com/1471-2407/5/144
© 2005 Ozkalemkas et al; licensee BioMed Central Ltd. 
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), 
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background: Although bone marrow metastases can be found commonly in some malignant tumors,
diagnosing a nonhematologic malignancy from marrow is not a usual event.
Methods: To underscore the value of bone marrow aspiration and biopsy as a short cut in establishing a
diagnosis for disseminated tumors, we reviewed 19 patients with nonhematologic malignancies who
initially had diagnosis from bone marrow.
Results: The main indications for bone marrow examination were microangiopathic hemolytic anemia
(MAHA), leukoerythroblastosis (LEB) and unexplained cytopenias. Bone marrow aspiration was not
diagnostic due to dry tap or inadequate material in 6 cases. Biopsy results were parallel to the cytological
ones in all cases except one; however a meticulous second examination of the biopsy confirmed the
cytologic diagnosis in this patient too. The most common histologic subtype was adenocarcinoma, and
after all the clinical and laboratory evaluations, the primary focus was disclosed definitively in ten patients
(5 stomach, 3 prostate, 1 lung, 1 muscle) and probably in four patients (3 gastrointestinal tract, 1 lung). All
work up failed in five patients and these cases were classified as tumor of unknown origin (TUO).
Conclusion: Our series showed that anemia, thrombocytopenia, elevated red cell distribution width
(RDW) and hypoproteinemia formed a uniform tetrad in patients with disseminated tumors that were
diagnosed via bone marrow examination. The prognosis of patients was very poor and survivals were only
a few days or weeks (except for 4 patients whose survivals were longer). We concluded that MAHA, LEB
and unexplained cytopenias are strong indicators of the necessity of bone marrow examination. Because
of the very short survival of many patients, all investigational procedures should be judged in view of their
rationality, and should be focused on treatable primary tumors.Page 1 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144Background parafin blocks, and 0.3 micrometer sections were cut. In
Diagnosis and management of many hematologic dis- some cases the sternum was sampled using a Rosenthal
eases depends on examination of the bone marrow, which needle for aspiration. Touch preparations were done if the
usually involves two separate specimens: a cytologic and a aspiration resulted in a "dry tap" or if aspiration material
histologic preparation. While cytologic preparation of was considered to be technically inadequate for evalua-
bone marrow, obtained by aspiration, allows excellent tion, or if it was hemodilute. Bone marrow aspiration,
visualization of cell morphology, the second one, usually touch preparations and peripheral blood smears which
obtained with a Jamshidi needle, allows optimal evalua- were obtained at the same time by biopsy were stained by
tion of cellularity, fibrosis or infiltrative disease. In addi- May Grunwald-Giemsa. Hematoxylene-eosin, Giemsa
tion to hematologic malignancies, bone marrow and reticulin stains were routinely performed in biopsy
examination has been increasingly useful in documenting sections. If the nonhematologic properties of the tumor
metastatic involvement of tumors. During the past four could be identified with routine stains, and if the primary
decades, prospective evaluation of bone marrow aspirates of the metastatic tumor could easily be diagnosed mor-
and biopsy specimens has come into widespread use for phologically with routine hematoxylene-eosin stains, as
accurate staging of many malignant diseases. Recognition in the cases of signet ring cell carcinoma; no immunohis-
of metastasis in random biopsies presents challenges to tochemical study was held. The pathologist's approach to
hematologists and pathologists when diagnosing the pri- definitive diagnosis of the patient with metastasis of
mary focus [1,2] unknown primary effectively followed a few sequential
steps: First of all we tried to determine the cell line of dif-
Although marrow metastases can be found commonly in ferentiation e.g. carcinoma, lymphoma, melanoma, sar-
some tumors, especially when newer sensitive methods coma, or germ cell, with the help of morphological
are applied for the detection of tumor cells, diagnosing a findings and if needed, immunohistochemical stains. Our
nonhematologic malignancy from marrow is a rare event. panel of antibodies contained pancytokeratin, HMB45,
We reviewed 19 patients with nonhematologic malignan- Leukocyte Common Antigen (LCA), Vimentin and Pla-
cies who were diagnosed initially from bone marrow, to cental Alkalen Phosphatase (PLAP). If it was an epithelial
underscore the value of bone marrow aspiration and tumor we tried to determine the cytokeratin (CK) type or
biopsy as a short cut in establishing a diagnosis for dis- types of distribution in the tumor cells, since some subsets
seminated tumors. Additionally we reported the details of of cytokeratins are uniqe to certain tumor types. Our
the management and survival of the cases to offer a prac- panel contained AE1/AE3, CAM5.2, CK7, CK20, and 34
tical suggestion about work up of these patients to find a beta E12. In our further studies we tried to determine if
primary focus. there was expression of supplemental antigens of epithe-
lial or germ cell derivation, that was Carcinoembryonic
Methods Antigen (CEA), Epithelial membrane Antigen (EMA) or
This study is based on our retrospective analysis of 19 PLAP. Last step was to determine if there was expression
patients with solid tumors whose diagnoses were made of cell-specific structures or receptors that are unique iden-
from bone marrow seen at the Department of Hematol- tifiers of cell types, for example neuroendocrine granules,
ogy, Uludag University Medical Faculty, over a period of 9 peptide hormones, thyroglobulin, Prostate Specific Anti-
years. Patients with non-Hodgkin's lymphomas and gen (PSA), Gross Cystic Disease Fluid Protein (GCDFP) or
Hodgkin's diseases were not included in this study and Thyroid Transcription Factor-1 (TTF-1). Our panel of anti-
patients with known neoplastic disease were also bodies contained Synaptophysin, Chromogranin, Neuron
excluded. Specific Enolase (NSE), Thyroglobuline, PSA, GCDFP,
and TTF-1. Cases with sarcomatous properties were
The standard technique was employed in obtaining the immunostained with Desmin, Smooth Muscle Actin,
samples from posterior iliac crest using a Jamshidi needle S100, Vimentin and Myoglobulin. After the confirmation
(Regular/Adult, 11-gauge). All of the trephine biopsies of original peripheral blood and bone marrow cytological
were performed unilaterally, because clinically none of findings and histopathological diagnoses by a senior
the diseases that focally involve the bone marrow were hematologist and an expert pathologist the patient charts
included in the differential diagnosis prior to the biopsy were reviewed. Cases with a history of malignancy at the
procedure. Length of the biopsy cores ranged between 1.2 time of presentation were excluded.
cm and 2.2 cm (mean 1.7 cm). Trephine biopsies were
fixed in 10% neutral buffered formaline for at least 24 Patient characteristics were recorded in each case, includ-
hours, and then decalcified overnight in a decalcifying ing: presenting symptoms, onset of symptoms, physical
solution which is a mixture of 8% HCl and 10% formic examination findings, peripheral blood counts, periph-
acid at equal amounts of volume. Following the auto- eral blood morphology, diagnostic evaluation, manage-
mated tissue processing, biopsies were embedded in ment, and survival.Page 2 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144Table 1: Clinical and cytopathological characteristics of patients
No Age/sex Presenting symptoms/ Performance status*/ Cytological examination Cytological examination Pathological 
Onset of symptoms/ Physical findings of peripheral blood of bone marrow examination of bone 
Presence of marrow
constitutional 
symptoms
1 50/M Back and chest pain/3 3/Pallor, subicterus MAHA, LEB Dilute, not optimal, not Adenocarcinoma with 
weeks/WL, F, NS diagnostic signet ring cell features 
(Suggestion: primary 
focus should be 
searched in GI tract)
2 40/M Abdominal pain, failure 3/Ecchymoses, MAHA, LEB Dilute, not optimal, not Adenocarcinoma
to passage gas and stool tenderness in diagnostic
by rectum, epigastrium and lower-
hematemesis/2 weeks/ right quadrant
NS
3 41/F Lumbar and extremity 3/Pain with deep LEB, MAHA Dilute, not optimal, not Adenocarcinoma
pain, lack of appetite, palpation of whole diagnostic
nausea/4 weeks/WL, F abdomen
4 48/F Lack of appetite, fatigue, 3/A few ecchymoses LEB Dry tap; touch Indifferentiated 
nausea, vomiting, fever/ preparation is not carcinoma (only CK 
2 months/F, WL optimal for evaluation positive strongly)
5 49/F Dyspepsia, weakness/1 3/Pallor, multiple LEB Dense foreign cell Adenocarcinoma
month/F ecchymoses, axillary infiltration forming 
single microLAP groups 
(adenocarcinoma)
6 71/M Lumbar and leg pain, 4/Pallor, impaired LEB Dry tap; imprint: highly Atypical epithelial cells 
somnolence/20 days/- consciousness, dense atypical in clusters (round cells 
dysorientation, somewhat large round infiltration) (Suggestion: 
dyscooperation, or oval cells infiltration Primary focus should be 
agitation in clusters investigated in lungs)
7 63/M Cough, dysphagia, 2/Scleral icterus, 5 cm LEB Epithelioid cells in solid Small cell carcinoma
abdominal swelling, hepatomegaly, melana in clusters (small cell 
weakness, prominent rectal digital palpation carcinoma infiltration)
loss in weight/WL, F, 
NS/1 month
8 57/F Back, lumbar and leg 3/Scleral icter, left MAHA, LEB Infiltration with signet Metastatic carcinoma 
pain, weakness, lack of axillary 2 cm LAP ring cells (signet ring cell 
appetite/3 months/WL adenocarcinoma)
9 45/M Lumbar and leg pain, 3/Pallor; LEB Infiltration with atypical Metastatic carcinoma 
prominent loss in large epithelioid cells (Suggestion: Primary 
weight, generalized tumor should be 
body pain/2 months/WL investigated in prostate)
10 25/F Weakness, hip pain/6 2/Pallor, right inguinal 2 Rare blastic cells Infiltration with blastic Higly dense atypical 
months/ WL cm LAP, 1 cm cells with vacuolated cells in alveolar 
hepatomegaly, 2 cm cytoplasm (MPO structure-actin, desmin 
splenomegaly negative; flow: B and T and vimentin positive, 
cell markers negative) LCA and CK negative-
(metastatic alveolar 
rhabdomyosarcoma)
11 35/M Nausea, vomiting, 4/Pallor, cachexia MAHA Adenocarcinoma cell Metastatic 
prominent loss of forming groups adenocarcinoma
weight, fatigue (alcohol, 
hashish and heroin 
dependence)/1 month/
F, WL
12 83/F Prominent loss in 3/Pallor; multipl MAHA, LEB Infiltration with signet Metastatic 
weight, nausea, ecchymoses ring cell carcinoma cells adenocarcinoma (signet 
vomiting; backpain/6 ring cell carcinoma 
months/WL metastasis)
13 61/F Headache, sore throat, 2/Pallor and scleral LEB Dry tap in first 2 Metastatic carcinoma 
abdominal pain, icterus, pain with attempts and very dilute (CK positive atypical 
constipation, nausea, palpation of right without particle in 3rd epithelioid cell 
vomiting (hematemesis hypochondrium attempt. No atypical infiltration, some of 
and melana history), cells; imprint: technically them in signet ring cell 
weakness/2 months/- inadequate shape)Page 3 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144Table 1: Clinical and cytopathological characteristics of patients (Continued)
14 75/M Multiple ecchymoses on 1/Ecchymoses and Only minimal shift to Non-hematopoetic cell Metastatic 
body and on purpura; 1.5 cm left; no erythroblast or infiltration forming adenocarcinoma (CK+ 
extremities, purpura on supraclavicular LAP poikilocytosis groups and some with cells forming glandular 
lower extremities; mucineous character and tubular structures
hematuria/2 days/-
15 73/M Confusion, adynamia/20 4/Hypotension, Slight shift to left, toxic Dilute and not optimal Nondiagnostic in first 
days/WL hypothermia, granulation, slight but there are report but 
dehydration, scleral poikilocytosis, no nonhematopoietic cells adenocarcinoma 
icterus, pallor, cachexia, erythroblast, no in small groups like metastasis reported 
2 cm hepatomegaly fragmentation adenocarcinoma cells after meticulous 
examination of new 
further sections of 
blocks
16 75/F Dyspnea, abdominal 4/Rales bilaterally, 2 cm LEB, no poikilositosis or Atypical non- ND
swelling/20 days/- hepatomegaly, pretibial fragmentation haematopoetic cells 
edema, petechia in (adenocarcinoma 
lower extremities metastasis)
17 68/M Weakness, lack of 1/Enlargement and LEB, slight Infiltration with Metastatic carcinoma 
appetite, prominent loss nodulation in prostate poikilocytosis, no adenocarcinoma cells compatible with 
of weight, lumbar pain/3 in digital examination fragmentation showing acinar and prostate carcinoma 
months/WL tubular structures (PSA +)
18 56/M Pain in hips and legs, 2/Pallor, cachexia, LEB, MAHA Infiltration with atypical CK + PAS- 
weakness/2 months/F, multiple microLAPs in epithelioid cells forming adenocarcinoma 
WL, SN servical, axillary and papillar and acinar metastasis (Suggestion: 
inguinal regions, a few structures Primary focus should be 
petechiae and (adenocarcinoma investigated in prostate)
eccyhymoses metastasis)
19 45/M Neck and hip pain, 3/restriction in physical LEB, MAHA Infiltration with Metastatic 
abdominal pain, activity adenocarcinoma cells adenocarcinoma; CK+ 
weakness/ 1 month/WL epithelioid cells, some 
of them mucinous and 
in shape of signet cell 
(Suggestion: Primary 
focus should be 
investigated in stomach)
*According to WHO/ECOG; WL: weight loss; F: fever; NS: night sweats; LAP: lymphadenopathy; MAHA: microangiopathic hemolytic anemia; LEB: 
leukoerythroblastosis; MPO: myeloperoxidase; CK: Cytokeratin LCA: leukocyte common antigen; ND: not done
Results and in 8 of them at least one anomaly was detected at
Between 1995 and 2004, bone marrow metastasis was presentation; additionally, in two patients who had nor-
diagnosed in 19 samples among 5420 bone marrow aspi- mal test results initially subsequent tests were found
rations and 856 bone marrow trephine biopsies. The ages abnormal. All patients had hypoproteinemia; the second
of the patients were between 25 and 83 years (median age: most common anomaly was elevated serum lactate dehy-
56), eight of them were female. drogenase (LDH) level, which was found normal in only
one patient (patient # 13) in biochemical analyses; some
The main indications for bone marrow examination were hepatic and renal parameters were abnormal in some
microangiopathic hemolytic anemia (MAHA), leukoery- patients but none of them was a constant finding.
htroblastosis (LEB) and peripheral cytopenias. Constitu-
tional symptoms and pain were the most prominent In 6 cases bone marrow aspiration was not diagnostic due
presenting symptoms of the patients. Clinical findings to dry tap or inadequate material. Histopathological
were highly variable according to the underlying disease. examinations confirmed the nonlymphohematopoietic
cell infiltration in first evaluation except one (patient #
In the laboratory, the most common findings were ane- 15). However, a meticulous second evaluation of biopsy
mia and thrombocytopenia, which were found in all confirmed the cytologic diagnosis in this patient too. The
patients. White blood cell counts (WBC) were between most common histologic subtype was adenocarcinoma.
3.3 × 109/l and 17.1 × 109/l (median: 8.1 × 109/l), red cell After the all clinical and laboratory evaluations, the pri-
distribution width (RDW) was increased in all cases, mary focus was disclosed in ten patients definitively (5
whereas mean corpuscular volume (MCV), mean corpus- stomach, 3 prostate, 1 lung, 1 muscle) and in four patients
cular hemoglobin (MCH), reticulocyte percentage (RET) probably (3 gastrointestinal tract, 1 lung). In five patients
and erythrocyte sedimentation rate (ESR) were highly var- all work up failed and these cases were classified as tumor
iable. Coagulation tests were carried out in 17 patients of unknown origin (TUO).Page 4 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144Table 2: Evaluation, clinical course and survival of patients
Evaluation for primary focus Final decision for Clinical course/Management Survival
primary focus
1 Chest XR, abdominopelvic USG and transrectal USG: GI tractus General condition deteriorated rapidly; stupor and 6 days
N; CEA and CA15.3 are high slightly and Ca19.9 is coma developed/Supportive care; 2 courses of TPE
high more than 10 folds, among CEA, α FTP, PAP, 
CA-125, CA19.9, CA15.3
2 Chest XR, abdominopelvic USG and CT: N; Stomach DIC, subdural hematoma developed/Multiple 11 days
gastroscopy: malign ulcus erythrocyte, platelet and plasma transfusions; 2 
courses of TPE
3 Abdominopelvic USG and CT: thickness on the Stomach DIC diagnosed at admission. Hematemesis and 20 days
antrum wall, gastrohepatic and portahepatic epistaxis developed later/Despite full transfusion 
microLAPs; only α-FTP is in normal limits, among the support; died due to intracerebral bleeding
CEA, α-FTP, CA 125, CA15.5; CA19.9 is high more 
than 10 folds; gastroscopy: infiltration in corpus and 
antrum (linitis plastica)
4 Chest XR, mammography: N; abdominopelvic USG TUO Fever resisted despite AB; general condition 12 days
and CT: N except minimal free pelvic fluid; upper GI deteriorated gradually, generalized seizures 
tract endoscopy: unremarkable developed without abnormal cranial CT finding; 
hypoxemia developed due to secretions/Supportive 
care only
5 Chest XR: N; abdominopelvic CT: not optimal; Stomach AB resistant fever (FUO) and GI tract bleeding 37 days
suspected thickness on the stomach wall, suspected developed later/Despite full transfusion support her 
metastatic lesions in columna vertabralis; upper GI general condition deteriorated rapidly. Died in 
tract endoscopy: malign ulcus in cardia; CA125 is high MODS picture
slightly and CA-19-9 is high approximately 6 times, 
among CEA, α-FTP, Ca 125, CA 19-9, CA 15-3
6 Cranial CT: N; thorax, abdomen and pelvis MR: Lung? His consciousness impaired progressively; refractory 5 days
multiple mediastinal LAPs in conglemeration with fever and hypotension developed
suspected parenchimal infiltration, benign prostate 
hyperthrophy; lumbar MR: multiple pathologic signal 
in backbone and degenerative alterations; Skeleton 
scintigraphy: multiple thoracal and lumbar uptake 
(degeneration, metastasis, trauma?)
7 Thorax, abdominal CT: subcarinal LAPs, a mass in Lung Pneumonia and atrial fibrillation developed/A course 47 days
right hilus, eosophagial compression, pulmonary of CT (Etoposide+Cisplatine) was given. Died duo to 
artery and pericardium invasion, a hipodens lesion in CHF
1 cm diameter in liver (USG in terminal period: 
multiple lesions compatible with metastasis); 
bronchoscopy: inoperable bronchial carcinoma; 
biopsy: small cell carcinoma); upper GI tract 
endoscopy: N. CEA, α-FTP, PSA, freePSA, CA125, 
CA 19.9 all: N
8 Axillary node FNAB: benign; cranial MR: compatible TUO Performance status deteriorated gradually. GI tract 38+ days
with bone metastasis and leptomeningeal bleeding developed/She refused colonoscopy and 
carcinomatosis; thorax CT: only bone metastasis; other supportive therapies and was discharged in 
abdominopelvic CT: 3 mm hipodens lesion in liver very bad condition
(metastasis?), backbone metastasis; transvaginal USG: 
N; bone scintigraphy: multiple metastasis; 
mammography: N; whole spine MR: generalized 
sclerotic and lytic lesions; upper GI tract endoscopy: 
erythemateous gastritis; only CA 125 is high 2 folds, 
among CEA, α-FTP, CA 125, CA 19-9, CA15-3, 
BHCG
9 Chest XR: N; transrectal USG: prostate carcinoma?; Prostate After his work up bisphosphonates therapy was 7+ months
prostate biopsy: adenocarcinoma; skelatal XR survey: initiated and was fallowed as outpatient; cranial 
multiple sclerotic metastasis and compression metastasis developed later; despite progressive 
fracture in L3; bone scintigraphy: generalized complaints he refused admission
metastatic involvement; tumor markers: PSA and free 
PSA are very high
10 Abdominopelvic CT: a solid mass in the pelvis Muscle VAC/IE (Vincristine, Adriamycine, 7 months
originated probably right gluteal muscle and Cyclophosphomide, Ifosfamide, etoposide) therapy 
homogeneous hepatosplenomegaly; biopsy from the resulted in partial response; died because of 
mass: rhabdomyosarcoma; skelatal XR survey: lytic progression later
lesions in only pelvic bones and proximal femur; bone 
scintigraphy: pathologic uptake in bilateral knee, 
pelvic area and, 5. and 7. ribsPage 5 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144Table 2: Evaluation, clinical course and survival of patients (Continued)
11 Chest XR: N; abdominopelvic US: N except GI system? His general condition deteriorated rapidly; GI tract 10 days
homogen minimal hepatomegaly bleeding and subdural hematoma developed, Died 
because of herniation/Supportive care only
12 Abdominopelvic CT: normal except suspected Stomach One course 5FU+FA was given; died as out patient 1 month
rigidity in stomach wall; gastroscopy: malign ulcus in 
junction of corpus and fundus; biopsy: 
Adenocarcinoma; Mammography: N; Backbone XR: 
loss of height in Th11 and Th12
13 Nasopharynx biopsy: N; pleural fluid cytology: GI system Transfusion support. Lost to follow up 45+ days
negative, biopsy: nonspecific chronic pleuritis; 
mammography:N; bone scintigraphy: multiple uptake; 
colonoscopy: N; gastroscopy: N; CEA and CA15.3: 
N, CA19.9 and CA125: very high
14 Chest XR and abdominopelvic USG: N TUO Nothing. Out of follow in 2nd week 14 days
15 Chest XR: nondiagnostic; CEA, α-FTP, PSA, free TUO Despite vigorous transfusion support and antibiotics 4 days
PSA, CA15.3, Ca19.9, Ca125 all: N his vital functions deteriorated progressively and died 
in MODS
16 Chest XR: nondiagnostic; previous available tests: TUO She died because of hypertensive crises and CHF 1 day
thorax CT: linear atelectasis and minimal right pleural after admission
fluid, 1–2 cm multiple mediastinal LAPs;.abdominal 
CT: homogeneous hepatomegaly and multiple cysts 
in 1.5 cm diameter in head of pancreas
17 Pelvic and transrectal USG: Prostatic hypertrophy; Prostate Flutamide (antiandrogen) and Goserelin asetat (LH- 15 months
prostate biopsy: Adenocarcinoma; thoracolumbar LR analogue) were given. Paraparesis and paraplegia 
MR and bone scintigraphy: multiple bone metastasis unresponsive to RT developed and died because of 
in backbone progressive disease and CHF
18 Neck and thorax CT: N; abdominopelvic CT: Prostate Gaserolin asetat+ Bikalutamid (LH-RH analogues) 3+ months
paraaortic 1.5 cm LAPs and heterogen prostatic were given; (he was in a good condition when 
hypertrophy, prostate biopsy: Adenocarcinoma; writing)
pelvis XR: multiple sclerotic lesions; bone 
scintigraphy; multiple + focuses in whole skeleton: 
PAP and PAS: very high; CEA, AFP, CA19.9: N
19 Bone scintigraphy: multiple + uptake; gastroscopy: Stomach Supportive care and palliative RT were given; he died 51 days
malign ulcus; biopsy: signet cell carcinoma (antrum); because of progressive disease
CEA, CA 19.9: very high, Ca125: high, AFP, PSA, F-
PSA:N; thoracal MR: loss of height in Th8; toraks CT: 
frosted glass appearance in lower and middle zones, 
minimal pleural effusion bilaterally; abdominopelvic 
CT:N; skelatal XR: Lumbar and pelvic sclerotic 
lesions
XR: direct radiography; CEA: carcinoembryonic antigen; α FTP: alpha fetoprotein; PAP: prostate specific antigen; BHCG: Beta human chorionic 
gonodotropin; USG: ultrasonography; CT: computed tomography; MR: magnetic resonance; Th: thoracal; LAP: lymphadenopathy; GI: 
gastrointestinal; DIC: disseminated intravascular coagulation; N: normal; FNAB: fine needle aspiration biopsy; TUO: tumor of unknown origin; TPE: 
therapeutic plasma exchange; AB: antibiotic; FUO: Febris of unknown origin; MODS: multiple organ deficiency syndrome; CHF: congestive hearth 
failure; RT: radiotherapy;
The prognosis of patients was very poor and survivals were and ii) bone marrow examination in evaluating patients
only a few days or weeks except for 4 patients whose sur- with suspected malignancy has a limited value unless sup-
vivals were longer (patients # 9, 10, 17, 18). To avoid ported by some other clinical finding, such as a leuko-
missing the individual features, detailed tables were pre- erythroblastic reaction [1,3,4]. Although there are a
pared: Clinical and cytopathologic characteristics are number of correlates that may be useful, there is no single
shown in Table 1, clinical courses and survivals are shown specific finding that is an indicator of marrow infiltration.
in Table 2, and some hematological and biochemical fea- Furthermore, in most patients with neoplastic infiltration
tures are shown in Table 3. of the marrow, the peripheral blood findings do not differ
significantly from those without marrow involvement.
Discussion Our series showed that MAHA, LEB and unexplained cyto-
Although bone marrow metastases are frequently encoun- penias are strong indicators of the necessity of bone mar-
tered in patients with disseminated solid tumors, it is not row examination. When we reviewed routine hematologic
a common event as a presenting sign. It stems from two and biochemical parameters we found that only four were
possible reasons: i) in general, bone marrow is seldom the present in all patients: anemia, thrombocytopenia, ele-
sole site of systemic involvement by malignant disease vated RDW and hypoproteinemia. However, our resultsPage 6 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144should be interpreted with caution, as the current study cal picture with TTP as presentation, we started plasma
was not designed to determine the predictive parameters exchange immediately in patient # 1 and # 2 but therapy
of marrow metastases. failed as expected. Hemolysis, in our CR-TM patients, was
so severe that several units of transfusion per day were
LEB is the term used to describe the combination of nucle- required to maintain a safe hemoglobin level. Reticulocy-
ated red cells (erythroblasts) and immature myeloid pre- tosis is expected but this finding could not be a reliable
cursors (e.g. myelocytes and myeloblasts) in the indicator of hemolysis, as seen in our patients, because of
peripheral blood film. The mechanism of leukoerythrob- marrow infiltration or chronic disease. In our series, seven
lastic reactions is not defined. The invasion of metastatic out of eight MAHA patients showed LEB and six showed
cancer cells may cause the early release of some cytokines, abnormality in some coagulation tests suggesting dissem-
leading to the development of a myelophthisic blood pic- inated intravascular coagulation. When we reviewed our
ture even before the marrow is completely replaced. This eight MAHA cases there were two TUO and one prostate
is a possible explanation for some instances of leukoeryth- carcinoma. All definitive stomach carcinomas and a prob-
roblastic changes in the blood of the patients with tumors able gastrointestinal carcinoma were in the MAHA group.
in whom marrow metastases are not documented by his- The survivals of our MAHA patients were between 11 and
tologic examination. Although metastatic foci can be 51 days except for one (patient # 18; prostate carcinoma).
found in a high percentage of some carcinomas, the devel- This suggests that once microangiopathic hemolysis is
opment of frank LEB occurs much less frequently, so its seen in a patient with disseminated carcinoma, the overall
absence should not exclude marrow involvement. If there prognosis is poor, especially in stomach adenocarcinoma
is LEB in a case of suspected malignancy, bone marrow presented with MAHA.
examination should be considered. Our series confirmed
this judgment, because fifteen of the patients were pre- "Dry tap" is a term used to describe failure to obtain bone
sented with LEB. On the other hand despite clear bone marrow on attempted marrow aspirations. Extensive mar-
marrow involvement in 4 patients there were no erythroid row fibrosis and hypercellularity have been proposed as
and myeloid precursors in their peripheral blood films mechanisms to account for the inability to withdraw mar-
(patients # 10, 11, 14, 15). LEB was reported at different row by aspiration [12,13]. Because it can be attributed to
rates in different series consisting of cancer patients with faulty technique it should only be used retrospectively
bone marrow metastases: 10/27 [5]; 19/25 [6]; 5/63 [7]; after review of the biopsy. We have two dry taps (patients
26/73 [2]. Our relatively high LEB ratio (15/19) is not # : 4, 6). If the definition of dry tap includes cases in which
comparable with other series because they consisted of material is obtained but no, or inadequate marrow cells
cancer patients who were diagnosed before bone marrow found in films we have four additional cases (patients # 1,
examination. Our high LEB ratio, when considered along 2, 3, 13). We concluded that bone marrow biopsy is cer-
with the bad outcomes, may reflect the late and advanced tainly indicated whenever aspiration results in an insuffi-
cases. In recent years there is a scarcity of publications on cient material especially in the presence of LEB, MAHA,
the association of myelophthisis with cancer. As men- cytopenias and an elevated serum LDH.
tioned by the authors in a recent report [8], early diagnosis
and more effective therapies are possible explanations for Classically, marrow biopsy is unequivocally the best
this decrease. method of detecting lymphomatous involvement because
in approximately one-third of cases, the aspirate is unre-
MAHA or thrombotic microangiopathy (TM) describes markable and the biopsy shows tumor. In solid tumors
the association of hemolytic anemia with red cell frag- other than those of lymphomatous origin the data in the
mentation caused by microangiopathy mechanically. literature comparing the relative value of bone marrow
Cancer related thrombotic microangiopathy (CR-TM) is a aspiration and biopsy in detecting marrow involvement
rare and severe complication; it usually occurs in the late are not so conclusive. [2,14]. Naturally trephine biopsies
or terminal stage of cancer with a short-term life-threaten- have a definitive advantage over aspirates in cases of dry
ing prognosis [9,10]. CR-TM shares certain clinical simi- tap. Additionally histologic sections may allow classifica-
larities with thrombotic thrombocytopenic purpura such tion of the type of tumor cells, and this is of particular
as neurological and renal impairment; also both are char- value in the investigation of a patient with a malignancy
acterized by circulating platelet aggregates containing of unknown primary site. Aspirations were diagnostic in
ultra large multimers of Von Willebrand factor (VWF). A 12 patients in our series. The other attempts yielded no, or
recent report showed no VWF cleaving protease deficiency inadequate material. Cytological diagnoses could not be
in a patient with metastatic adenocarcinoma of the colon confirmed in one patient because of the patient's objec-
and microangiopathic hemolysis, who is refractory to tion to biopsy procedure. Only in one instance was cytol-
plasma exchange [11]. We do not have any data about ogy superior to biopsy in first examination (a
protease activity but, because of an almost identical clini- retrospective examination of biopsy confirmed the cyto-Page 7 of 9
(page number not for citation purposes)
Table 3: Hematologic parameters of peripheral blood and biochemistry of serum of patients*
No WBC Hb MCV MCH RDW PLT RET ESR Coagulation tests Total LDH Total/ AST ALT ALP Urea Creatinine
(× 109/l) (mg/dl) (fl) (pg) (× 109/l) (%) (mm/1 h) Protein/ Direct 
Albumin Bil
1 8.1 9.9 94.5 33.3 17.0 32 2.5 17 PT and FDP: high 6.3/3.4 508 3.1/1.4 49 38 525 70 0.8
2 6.8 7.0 93.2 28.7 15.6 19 3.0 20 Ne initially and all abnormal 5.6/3.0 1042 1.5/0.5 157 33 365 56 1.0
later
3 11.6 6.4 79.7 27.0 16.1 20 1.0 36 PT aPTT and FDP: abnormal 5.0/2.1 776 2.9/1.1 67 57 587 46 0.4
4 13.1 6.8 74.1 25.2 18.4 41 0.8 78 N 5.7/2.6 1669 0.9/0.5 104 48 282 148 2.1
5 15.8 6.6 82.0 26.3 16.7 18 3.5 34 N initially but abnormal later 5.8/2.7 1016 2.4/1.1 61 11 661 84 0.8
6 10.6 10.6 73.8 24.4 15.3 12 0.2 107 N 6.2/2.8 1255 0.5/0.3 52 22 240 89 1.2
7 4.2 9.3 85.7 29.2 15.2 21 1.4 80 N 5.8/3.6 1438 1.1/0.5 78 54 155 68 1.0
8 6.5 8.1 89.3 30.9 16.5 72 4.7 44 N 5.7/2.9 880 2.3/0.9 30 17 1548 33 0.5
9 6.4 9.5 61.9 20.5 16.3 45 1.0 65 N 6.4/3.4 665 0.5/0.4 41 14 433 43 0.3
10 7.7 5.3 79.3 27.5 15.8 11 0.0 140 N 5.7/2.2 1612 0.5/0.2 26 4 71 15 0.6
11 10.3 4.0 90.0 29.9 16.8 32 1.8 5 D-Dimer and PT: high 6.6/4.0 1899 1.3/0.7 78 22 262 48 0.8
12 11.5 7.0 84.7 26.4 17.0 37 3.2 30 PT, aPTT and D-Dimer: high 6.7/4.6 904 1.8/0.7 130 78 402 63 1.2
13 6.5 6.4 90.7 28.6 17.3 103 2.8 40 N 6.6/3.5 258 4.4/1.4 11 10 1097 26 0.5
14 9.1 12.6 87.9 30.0 14.6 48 0.8 10 ND 6.2/3.0 630 1.1/0.4 41 25 144 29 0.9
15 3.3 3.3 89.3 30.2 17.7 24 0.4 95 PT, aPTT and D-Dimer: high 5.4/2.6 777 1.8/1.6 43 18 82 161 1.6
16 16.2 8.7 92.1 30.3 17.8 23 ND 85 ND 5.9/2.3 708 3.3/1.6 41 18 156 114 2.3
17 37.1 6.6 80.5 26.2 21.7 76 3.6 94 PT and D-Dimer high, aPTT: N 7.3/3.3 1127 1.2/0.5 70 48 1225 23 1.0
18 37.4 8.5 91.8 28.9 16.0 18 2.2 30 PT, aPTT and D-Dimer: high 6.1/3.9 393 1.7/0.6 27 36 1507 17 0.5
19 17.1 8.4 91.0 31.5 16.9 54 4.2 47 PT high D-Dimer upper limit, 6.2/3.8 1290 0.7/0.3 68 75 638 37 0.5
aPTT:N
*Normal ranges of serum chemical parameters: Total protein: 6.4–8.8 g/dl; Albumin3.0–5.5 g/dl; LDH: 190–380 U/l; Total bilirubin: 0.2–1.1 mg/dl; Direct bilirubin:0.0–0.4 mg/dl; AST: 0–40 U/L; ALT: 0–
43 U/L; ALP: 27–147 U/L; Urea: 15–50 mg/dl; Creatinine: 0.5–1.6 mg/dl; PT: prothrombin time; aPTT: active partial thromboplastin time; FDP: fibrin degradation products ND: not done; N: normal
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144
Page 8 of 9
(page number not for citation purposes)
BMC Cancer 2005, 5:144 http://www.biomedcentral.com/1471-2407/5/144logical finding in this patient). Although biopsies have Authors' contributions
some advantages, tumor cells occasionally can be seen in FO, RA and VO participated in the conception and design
aspirate preparations when biopsy sections are normal as of study, acquisition of data, analysis and interpretation
mentioned in literature, these two procedures should of data. TO, UO, HO and EK participated in acquisition of
therefore be regarded as complementary. data and drafting the article. TE, OY and AT participated
in revising it critically for important intellectual content.
The management of a patient, whose solid malignancy is All authors read and approved the final manuscript.
disclosed from bone marrow, depends on his/her primary
tumor. The pathologist can assist the clinician by thor- References
oughly examining the histologic specimen. In some cases 1. Papac RJ: Bone marrow metastases.  Cancer 1994, 74:2403-2413.
an immunoperoxidase staining for organ-specific antigen 2. Mohanty SK, Dash S: Bone marrow metastasis in solid tumors.Indian J Pathol Microbiol 2003, 46:613-616.
examination might be sufficient to bring out of the pri- 3. Burkhardt R, Frisch B, Bartl R, Kettner G, Schlag R, Hill W: Detec-
mary focus. As an example, an immunoperoxidase stain tion of haematologic and nonhaematologic cancer by bone
biopsy.  Cancer Detect Prev 1981, 4:619-627.
for prostatic acid phosphatase or prostate-specific antigen 4. Sar R, Aydogdu I, Ozen S, Sevinc A, Buyukberber S: Metastatic
may be helpful in establishing a diagnosis of metastatic bone marrow tumours: a report of six cases and review of
prostate cancer. On the opposite side of the spectrum in the literature.  Haematologia (Budap) 2001, 31(3):215-223.
5. Bezwoda WR, Lewis D, Livini N: Bone marrow involvement in
some cases the pathologist could not comment on the pri- anaplastic small cell lung cancer. Diagnosis, hematologic fea-
mary site because of a poorly differentiated tumor. Maxi- tures, and prognostic implications.  Cancer 1986, 58:1762-1765.
mum effort should be made to minimize the target. In 6. Tritz DB, Doll DC, Ringenberg QS, Anderson S, Madsen R, Perrt MC,Yarbro JW: Bone marrow involvement in small cell lung can-
these tumors as a first step leukocyte common antigen cer. Clinical significance and correlation with routine labora-
may be used to differentiate lymphohematopoietic neo- tory variables.  Cancer 1989, 63:763-766.
7. Campling B, Quirt I, DeBoer G, Feld R, Shepherd FA, Evans WK: Is
plasm from other cancers. Patient # 7 and 10, who were bone marrow examination in small cell lung cancer really
reported recently elsewhere separately [15,16], are good necessary?  Ann Intern Med 1986, 105:508-512.
examples for bringing out the primary focus by extra stain- 8. Makoni SN, Labor DA: Clinical spectrum of myelophthisis incancer patients.  Am J Hematol 2004, 76:92-93.
ings. The examination of patients with bone marrow 9. Lin YC, Chang HK, Sun CF, Shih LY: Microangiopathic hemolytic
metastases of unknown origin should focus on detecting anemia as an initial presentation of metastatic cancer of
treatable primary tumors [17]. This work up may result in unknown primary origin.  South Med J 1995, 88:683-687.10. Nordstrom B, Strang P: Microangiopathic hemolytic anemias
certain improvements in survivals of patients with pros- (MAHA) in cancer. A case report and review.  Anticancer Res
tate carcinoma. Bone marrow metastases commonly arise 1993, 13:1845-1849.
11. Forman RB, Benkel SA, Novik Y, Tsai HM: Presence of
from lung, breast and prostate cancers; therefore, in case ADAMTS13 activity in a patient with metastatic cancer and
of no clinical sign a reasonable work up should include a thrombotic microangiopathy.  Acta Haematol 2003, 109:150-152.
chest roentgenogram, a prostate examination with serum 12. Humpries JE: Dry tap bone marrow aspiration: Clinical signifi-cance.  Am J Hematol 1990, 35:247-250.
prostate specific acid phosphotase (PAP) in men; a breast 13. Hyun BH: Bone marrow examination: Adventures in diagnos-
examination and mammography in women; computed tic hematology.  Yonsei Med J 1986, 27:100-105.
14. Atac B, Lawrence C, Goldberg SN: Metastatic tumor: The com-
tomography scans also should be performed in suspected plementary role of the marrow aspirate and biopsy.  Am J Med
cases. Serum tumor markers are not so useful in many Sci 1991, 302:211-213.
cases except PAP; considering their less specificity it is no 15. Ali R, Ozkalemkas F, Ozcelik T, Ozan U, Ozkocaman V, Tunali A:Small cell lung cancer presenting as acute leukaemia.  Cytopa-
surprising. Because the expected survival of many patients thology 2005, 16:262-263.
is quite short, laboratory and imaging studies should be 16. Ali R, Ozkalemkas F, Ozan U, Ozcelik T, Ozkocaman V, Filiz G, Man-
considered in view of their rationality. avoglu O, Tunali A: Rhabdomyosarcoma of the perianal regionpresenting as acute leukemia.  Ann Hematol 2004, 83:729-730.
17. Ringenberg QS, Doll DC, Yarbro JW, Perry MC: Tumors of
Conclusion unknown origin in the bone marrow.  Arch Intern Med 1986,146:2027-2028.
In conclusion, MAHA, LEB and unexplained cytopenias
are strong indicators of the necessity of bone marrow Pre-publication history
examination. Anemia, thrombocytopenia, elevated RDW The pre-publication history for this paper can be accessed
and hypoproteinemia form a uniform tetrad in patients here:
with disseminated tumors that are diagnosed via bone
marrow aspiration and biopsy. Because of the very short http://www.biomedcentral.com/1471-2407/5/144/pre
survival of many patients, all investigational procedures pub
should be focused on treatable primary tumors.
Competing interests
The author(s) declare that they have no competing inter-
ests.Page 9 of 9
(page number not for citation purposes)