Journal of Infection and Public Health (2010) 3, 124—129

Diagnostic yield of bone marrow examination inHIV associated FUO in ART naïve patients

Arindam Pandea,∗, Maitreyee Bhattacharyyab, Shantasil Paina,Anirban Ghosha, Ajanta Samantac

a Department of General Medicine, Medical College & Hospital, Kolkata, West Bengal, Indiab Department of Haematology, NRS Medical College & Hospital, Kolkata, Indiac North Bengal Medical College & Hospital, Sushrutanagar, Darjeeling, West Bengal, India

Received 1 January 2010; received in revised form 12 July 2010; accepted 15 July 2010

KEYWORDSFUO;AIDS;HIV;Tuberculosis;Bone marrow

Summary A bone marrow (BM) aspiration and biopsy is often believed to be amuch needed diagnostic procedure in the work up of patients with fever of unknownorigin (FUO), especially in the setting of AIDS. Is it worthwhile to proceed with thisinvasive diagnostic method? The usefulness of a BM aspiration or biopsy to assist inthe diagnosis of FUO or prolonged fever in AIDS patients has been reported previouslyto range from 4% to 40%. The purpose of this study was to assess the usefulness ofa BM aspiration and biopsy in diagnosing the cause of FUO in patients with AIDSand to identify the utility of the procedure for the diagnosis of malignancies/otherhematological disorders resulting in the FUO. In this study, comprising of 30 patients,we have tried to find the diagnostic yield of bone marrow examination in finding theetiology of ‘‘FUO associated with HIV infection’’. Though similar studies have beenreported in the literature but it is lacking from eastern India.

The majority of BM examination in this series revealed infections followed by

hematological disorders. Our study showed the diagnostic yield of bone marrowexamination in ‘‘HIV associated FUO’’ to be 26.7%. It was found to be positive in33.3% of the patients, who had the final diagnosis of an infective etiology and 100%of the patients, who had a final diagnosis of an underlying hematological etiology.© 2010 King Saud Bin Abdulaziz University for Health Sciences. Published by Elsevier

.

Ltd. All rights reserved

∗ Corresponding author at: Department of General Medicine, MedicWest Bengal,India. Tel.: +91 9830686081.

E-mail address: drapande@gmail.com (A. Pande).

1876-0341/$ — see front matter © 2010 King Saud Bin Abdulaziz University for Health

doi:10.1016/j.jiph.2010.07.006

al College & Hospital, 88, College Street, Kolkata 700073,

Sciences. Published by Elsevier Ltd. All rights reserved.

D

I

AHlhmlonoca≥pmiaipetoewquomoh1tt

M

TvFpHHiafInnotibs

HCuiutVmatsvrosbFacdupmoc

(uMuafl

rndMsdKmSit

R

Ds

iagnostic yield of bone marrow examination

ntroduction

n estimated 33 million people are living withIV/AIDS worldwide with approximately 2.7 mil-

ion getting newly infected every year [1]. Indiaas an estimated 2.5 million infections [1]. Clinicalanifestations of HIV disease include hemato-

ogic abnormalities (isolated anemia, leucopeniar thrombocytopenia, bicytopenia or pancytope-ia), opportunistic infections, unexplained feverr a fever of unknown origin (FUO), and, lessommonly, certain types of neoplasms [2,3]. ‘‘HIVssociated FUO’’ is defined by a temperature of38.3®C (≥101®F) on several occasions over aeriod of more than 4 weeks for outpatients orore than 3 days for hospitalized patients with HIV

nfection [4]. This diagnosis is invoked if appropri-te investigations over 3 days, including 2 days’ncubation of cultures, reveal no source [4]. Theatients were recruited in this study on strict adher-nce to the above-mentioned criteria. In additiono the recommended investigations for diagnosisf FUO, bone marrow examination was done inach and every patient, and its diagnostic yieldas calculated. Bone marrow examination is fre-uently undertaken, especially in the evaluation ofnexplained or persistent fever [1], in suspectedpportunistic infection with bone marrow involve-ent [5], or for staging in malignancy [6]. Findings

f bone marrow examination in patients with AIDSave been described in the literature since the980s [7]. We have tried to estimate its utility inhe diagnosis of FUO in eastern Indian patients inhis study.

aterials and methods

his was a non-randomized cross-sectional obser-ational study undertaken from March’08 throughebruary’09. The study population included theatients attending ‘‘Apex referral clinic forIV/AIDS — Medical college, Kolkata — the oldestIV treatment centre in Eastern India, established

n 1995’’, and Medicine ward of Medical Collegend Hospital, Kolkata, who fulfilled the criteriaor ‘‘HIV associated FUO’’. All the patients werendian by origin, mostly from West Bengal and theeighboring state of Bihar. The duration of fever,umber of outpatient visits, steps taken at those

utpatient visits were analyzed for fulfillment ofhe criteria for FUO, from the patient’s past med-cal records. The following steps were followedefore the selection of patients: a comprehen-ive history taking, detailed physical examination,

aacmT

125

IV ELISA, CD4 count, complete hemogram, ESR,RP, blood biochemistries, 3 sets of blood culture,rinalysis & culture, stool routine & parasitolog-cal examination, sputum for AFB, chest X-ray,ltrasonography of abdomen, fundoscopy, hepati-is B surface antigen, anti-HCV antibody & serumDRL. All the patients were then subjected to bonearrow aspiration, trephine biopsy & examination

nd ZN stain, fungal stain followed by mycobac-erial and fungal culture if required. Also done inelected patients were serum iron, TIBC, ferritin,itamin B12 and folate studies, anti-nuclear factor,heumatoid factor, muscle enzymes, CT/MRI scanf brain, CSF study, ascitic fluid study, pleural fluidtudy, crypococcal antigen, anti-toxoplasma anti-ody, serum protein electrophoresis, lymph nodeNAC or biopsy, skin biopsy, CT scan of chest,bdomen and pelvis, upper GI endoscopy andolonoscopy. However radionuclide scanning proce-ures could not be used for diagnosis because ofnavailability in our institution. Patients on ART,regnant patients, patients with previously knownalignancy were excluded from study. The main

utcome measure was the final diagnosis at dis-harge or within 3 months follow up.

HIV was diagnosed by Rapid ELISA test kitHIV Comb, Tri-dot and Bioimmune) at the Vol-ntary Counseling and Testing Center (VCTC) atedical College, Kolkata. CD4 counts were doc-mented by a FACS Counter (Becton-Dickinson)t the School of Tropical Medicine, Kolkata, byuochrome-conjugated antibody to CD4.

Bone marrow aspirates were taken from poste-ior iliac crest with Salah’s bone marrow aspirationeedle and bone marrow biopsy when required wasone with Jamshidi bone marrow biopsy needle.arrow aspirate staining was done by Leishman

tain. ZN staining of bone marrow aspirate wasone in selected cases from DOTS, Medical College,olkata. Special fungal staining and culture of bonearrow aspirate was done in selected cases from

chool of Tropical Medicine, Kolkata. Bone marrowron staining was done by Perl’s stain. Bone marrowrephine biopsy was stained by H—E stain.

esults

uring the one year period of study, we randomlyelected 30 patients who met the criteria for ‘‘HIV

ssociated FUO’’. Twenty one patients were mennd 9 women. Twenty four out of the 30 had CD4ount < 200 cells/�L (range 6—303 cells/�l). Theean age was 40.6 years (range 19—63 years).he final diagnosis responsible for FUO consisted

126 A. Pande et al.

pmobifpidw(carIt

cbD—(

Figure 1 Break up of all the cases which were includedin our study, with final diagnosis.

of infectious diseases in 15 (50%) patients, hema-tologic/neoplastic disease in 3(10%) patients andundiagnosed in 12(40%) patients. Among infec-tious causes, there were 12 bacterial, 1 fungal, 1mixed (fungal + bacterial) and 1 viral etiologies (seeFigs. 1—3).

Bone marrow aspiration led to the diagnosisof underlying etiology in 6(20%) cases with bonemarrow biopsy contributing in additional 2(6.7%)cases. So the combined diagnostic yield of bone

marrow aspiration/biopsy in the diagnosis of ‘‘HIVassociated FUO’’, in our study, was found tobe 26.7%. Bone marrow was the only diagnostictool which could clinch the diagnosis in 4(13.3%)

Figure 2 Histoplasma in bone marrow.

—Tt(—

Figure 3 AFB in bone marrow.

atients. Among the diseases diagnosed by bonearrow examination, 5(62.5%) were found to be

f infectious origin and 3(37.5%) were found toe of hematologic/neoplastic origin. Among thenfectious causes, Mycobacterium tuberculosis wasound to be responsible in 3 cases (60%), Histo-lasma capsulatum was found to be responsiblen 2 cases (40%). Hematologic and malignant con-itions diagnosed from bone marrow examinationere Hodgkin’s disease (1 case), multiple myeloma

1 case) and hemophagocytic lymphohistiocytosis (1ase). Bone marrow staining showed AFB in 1 casend histoplasma in 2 cases; however culture wasequired to diagnose the remaining 2 cases of AFB.n 1 of those 2 cases, bone marrow biopsy showedubercular granuloma (see Tables 1 and 2).

Other diagnostic procedures whichlinched/aided in diagnosis were lymph nodeiopsy (3 cases: 1 — histoplasma, 1 — Hodgkin’ss, 1 — tuberculosis), lymph node FNAC (2 cases

tuberculosis), CT scan of chest and abdomen1 case — Hodgkin’s disease), CSF study (2 cases

TB meningitis), ascitic fluid study (1 case —

B peritonitis), pleural fluid study (1 case —ubercular pl effusion), sputum for AFB and CXR2 cases — pulmonary TB), skin biopsy (1 case

histoplasma), serum protein electrophoresis

Table 1 Distribution of etiologies diagnosed by bonemarrow examination.

Etiology diagnosed by BM exam No. of cases(n = 8)

InfectiousAFB 3Histoplasma 2

Malignant/hematologicHodgkin’s disease 1Multiple myeloma 1Hemophagocytic lymphohistiocytosis 1

Diagnostic yield of bone marrow examination 127

Table 2 Bone marrow changes among the studypopulation.

BM findings No. of cases

CellularityHypocellular 9Hypercellular 9Normocellular 12

ErythropoiesisNormal 7Increased 7Decreased 9Dysplasia 7

GranulopoiesisNormal 23Increased 0Decreased 4Dysplasia 3

MegakaryopoiesisNormal 20Increased 0Decreased 4Dysplasia 6

Plasmacytosis 3

Granuloma 1

StainingPositive 3Negative 27

Culture positive (in staining negative) 2

Iron storesIncreased 1Decreased 3Normal 26

(dt

D

Fitc4mF

Fc

shdmw

rtwmnolhe1nmeitHshmWoo21tKbcoAWira

Malignant involvement 3

1 case — multiple myeloma). Fig. 4 summarizesistribution of diagnostic tools required to clinchhe final diagnosis.

iscussion

UO still remains an intriguing problem in HIV-nfected patients. Despite all available diagnosticools, its cause remains elusive in up to 40% of

ases [8—14]. Our study shows similar results with0% cases not having any objective diagnosis. Theost common disease found to be associated with

UO in our study was tuberculosis, as in other

rn

d

igure 4 Distribution of diagnostic tools required tolinch the final diagnosis.

eries [15]. It was followed by other infections likeistoplasmosis and hepatitis B. Hematologic disor-ers/malignancies like Hodgkin’s disease, multipleyeloma and hemophagocytic lymphohistiocytosisere found in the remaining cases.In contrast to many other investigators, who

eported lack of utility of bone marrow aspira-ion/biopsy in diagnosis of FUO in HIV patient [12],e found evidence of infective processes in bonearrow in 5/15(33.3%) cases having the final diag-

osis of some infection. The most common infectiverganism responsible was mycobacterium tubercu-osis, as already mentioned. Tuberculosis is a majorealth concern in India. According to 2009 WHOstimate, the national prevalence of TB is 185 per00,000 persons. India has the highest number ofew TB cases in the world with estimated 1.98illion detected in 2008. In the same year WHO

stimated the national prevalence of HIV to be 6.7%n TB patients [16]. There is no official data onhe magnitude of the TB in HIV infection in India.owever, the data garnered from studies on HIVeropositivity among adult TB patients in tertiaryealth care centers does provide an insight on theagnitude of the problem. Data from the state ofest Bengal, which has an estimated populationf around 90 million, shows that the total numberf registered TB patients undergoing treatment in008 was 107,225 [17]. In our study, we had total of3 patients who had been diagnosed with mycobac-erium tuberculosis infection. Similar studies byilby et al. and Santos et al. revealed compara-le results with mycobacterium being the mostommon infective organism. A major proportionf the infections were caused by Mycobacteriumvium Complex (MAC) in those two studies [18,19].e however did not have any patients with MAC

nfection and all the patients diagnosed with TBesponded well to standard anti-tubercular ther-py. It is a well known fact that MAC infection is very

are in India. The reason for this rare occurrence isot yet very clear.

Bone marrow examination was also found to beiagnostic in all 3/3(100%) cases of hematological

2mn[(

facmIie

C

Fepaictitadasrpwd

R

128

disorder/malignancy underlying FUO. It has beenfound to be a rapid diagnostic tool for detectinghematologic disorders as we found the evi-dence of Hodgkin’s disease, multiple myeloma andhemophagocytic lymphohistiocytosis on BM exami-nation in our work up of FUO. The total diagnosticyield of bone marrow examination in diagnosis ofFUO in HIV was found to be 8/30(26.7%), howeverit was found to be 8/18(44.4%) in whom a diag-nosis could be reached. Bone marrow examinationwas singularly helpful in clinching the diagnosis in4 cases (2 — AFB, 1 — histoplasma, 1 — hemophago-cytic lymphohistiocytosis).

In our study, we could diagnose all 2/2(100%)cases of histoplasmosis from bone marrow findings.Literature shows bone marrow yield in dissemi-nated histoplasmosis in AIDS patients to be >70%in some studies [9,10]. We also found higherincidence of Hodgkin’s disease (1 case) involv-ing bone marrow compared to NHL (0 case) inthese patients. It also suggests that AIDS asso-ciated HD have more advanced disease at thetime of diagnosis. Although HD does not representan AIDS-defining condition, recent evidence con-sistently indicates that HIV-infected people havea significantly increased risk of developing HD[20]. HIV-related HD is characterized by the pre-ponderance of aggressive histological subtypes,advanced stage at diagnosis, and malignant clin-ical course [20]. Though NHL is considered tobe an important etiology of FUO in HIV-infectedpatients, in some studies it has been shown thatBM biopsy is not helpful in clinching this diagno-sis [19]. Also worth mentioning is the occurrenceof multiple myeloma in 1 of our patients. Tothe best of our knowledge, about 50 cases ofAIDS complicated by multiple myeloma have beenreported in the medical literature [25]. Multi-ple myeloma is shown in some series to occurwith greater frequency in HIV-infected patients[21,22]. Although multiple myeloma is of rareoccurrence in HIV-infected individuals [25], bonemarrow plasmacytosis is a relatively common find-ing in such patients [23—26]. In our study, wefound the presence of plasmacytosis in 3 patients(10%). Another rare disorder which we found in 1 ofour patients was hemophagocytic lymphohistiocy-tosis (1 case). Hemophagocytic lymphohistiocytosishas been diagnosed with increasing frequency inpatients infected with HIV [27,28]. Hemophagocy-tosis is commonly found, to varying degrees, in the

bone marrow of HIV-infected individuals withoutan underlying diagnosis of HLH, which raises thepossibility of a sub clinical form of HLH [29]. Anautopsy study of 56 patients with acquired immun-odeficiency syndrome (AIDS) found an approximate

A. Pande et al.

0% prevalence of hemophagocytosis [30]. In fact,ost of the reported cases of HLH have been diag-

osed during the advanced stages of HIV infection27—29] as in our case who had a very low CD4 count88 cells/�L).

Other than these diagnostic findings, we alsoound dysplasia in erythropoiesis, megakaryopoiesisnd granulopoiesis, hypo & hyperplasia of 1 or moreell lines and megaloblastic changes in the bonearrow of these ART naïve HIV-infected patients.

n addition to this we also found both low and highron stores in these patients in concordance witharlier studies [7,31].

onclusion

rom the discussion, we can safely conclude thatxamining the bone marrow is a useful diagnosticrocedure that facilitates the process of achievingfaster diagnosis in patients with AIDS present-

ng with FUO, especially in infective cases and inases where the underlying etiology is a hema-ological disorder. A BM biopsy may be preferredn patients who are extremely ill in whom evenhe slight possibility of a positive biopsy may favorggressive action, possibly in patients whose con-ition warrants a faster diagnosis or those in whom

neoplastic or primary hematologic disease istrongly suspected. So a routine bone marrow aspi-ation/biopsy examination may be included in therimary work up of FUO in HIV patients, particularlyho are naïve to ART, for more rapid & enhancediagnosis of the varied etiologies underlying.

Funding: No funding sources.Competing interests: None declared.Ethical approval: Not required.

eferences

[1] UNAIDS/WHO. 2007 AIDS epidemic update; December 2007.[2] Brizzi MF, Porcu P, Porteri A, Pegoraro L. Haematologic

abnormalities in acquired immunodeficiency syndrome.Haematologica 1990;75:454—63.

[3] Mayo J, Collazos J, Martinez E. Fever of unknown origin inthe HIV infected patient: new scenario for an old problem.Scand J Infect Dis 1997;29:327—36.

[4] Gelfand JA, Callahan MV. Harrison’s principles of internalmedicine, 19, 17th ed. New Delhi: McGraw-Hill; 2008. p.130.

[5] Akpek G, Lee SM, Gagnon DR, Cooley TP, Wright D. Bonemarrow aspiration, biopsy, and culture in the evaluation of

HIV infected patients for invasive mycobacteria and Histo-plasma infections. Am J Hematol 2001;67:100—6.

[6] Seneviratne L, Espina BM, Nathwani BN, Chan JA, BrynesRK, Levine AM. Clinical, immunologic, and pathologic cor-relations of bone marrow involvement in 291 patients with

D

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

iagnostic yield of bone marrow examination

acquired immunodeficiency syndrome-related lymphoma.Blood 2001;98:2358—63.

[7] Castella A, Croxson TS, Mildvan D, Witt DH, Zalusky R. Thebone marrow in AIDS. Am J Clin Pathol 1985;84:425—32.

[8] Prego V, Glatt AE, Roy V, Thelmo W, Dincsoy H, Raufman JP.Comparative yield of blood culture for fungi and mycobac-teria, liver biopsy, and bone marrow biopsy in the diagnosisof fever of undetermined origin in HIV-infected patients.Arch Intern Med 1990;150:333—6.

[9] Nichols L, Florentine B, Lewis W, Sattler F, Rarick MU,Brynes RK. Bone marrow examination for the diag-nosis of mycobacterial and fungal infections in theacquired immunodeficiency syndrome. Arch Pathol Lab Med1991;115:1125—32.

10] Engels E, Marks PW, Kazanjian P. Usefulness of bone mar-row examination in the evaluation of unexplained fevers inpatients infected with human immunodeficiency virus. ClinInfect Dis 1995;21:427—8.

11] Rodriguez JN, Dieguez JC, Moreno MV, Aguayo DM, VegaMD, Conde J, et al. Usefulness of bone marrow examina-tion in patients with advanced HIV infection. Rev Clin Esp1996;4:213—6.

12] Fernandez-Aviles F, Ribera JM, Romeu J, Batlle M, NavarroJT, Manterola JM, et al. The usefulness of the bone mar-row examination in the etiological diagnosis of prolongedfever in patients with HIV infection. Med Clin (Barc)1999;112:641—5.

13] Natividad B, Nunez A, Gorgolas M, Esteban J, Calabuig T,Rivas MC, et al. Bone marrow biopsy in the diagnosis offever of unknown origin in patients with AIDS. Arch InternMed 1997;157:1577—80.

14] Volk EE, Miller ML, Kirkley BA, Washington JA. The diag-nostic usefulness of bone marrow cultures in patients withfever of unknown origin. Am J Clin Pathol 1998;110:150—3.

15] Armstrong WS, Katz JT, Kazanjian PH. Human immunodefi-ciency virus-associated fever of unknown origin: a study of70 patients in the United States and review. Clin Infect Dis1999;28:341—5.

16] TB INDIA 2010. RNTCP status report. Directorate Generalof Health Services, Ministry of Health and Family Welfare,Govt. of India.

17] Health on the March 2008—2009. Directorate of Health Ser-vices, Govt. of West Bengal.

[

Available online at www.

129

18] Kilby JM, Marques MB, Jaye DL, Tabereaux PB, Reddy VB,Waites KB. The yield of bone marrow biopsy and culturecompared with blood culture in the evaluation of HIV-infected patients for mycobacterial and fungal infections.Am J Med 1998;104((February) 2):123—8.

19] Santos ES, Raez LE, Eckardt P, DeCesare T, Whitcomb CC,Byrne Jr GE. The utility of a bone marrow biopsy in diagnos-ing the source of fever of unknown origin in patients withAIDS. J Acquir Immune Defic Syndr 2004;37((December 15)5):1599—603.

20] Dolcetti R, Boiocchi M, Gloghini A, Carbone A. Pathogeneticand histogenetic features of HIV-associated Hodgkin’s dis-ease. Eur J Cancer 2001;37:1276—87.

21] Goedert JJ, Cote TR, Virgo P. Spectrum of AIDS-associatedmalignant disorders. Lancet 1998;351:1833—9.

22] Fiorino AS, Atac B. Paraproteinemia, plasmacytoma,myeloma and HIV infection. Leukemia 1997;11:2150—6.

23] Aboulafia DM, Mitsuyasu RT. Hematologic abnormalities inAIDS. Hematol Oncol Clin North Am 1991;5:195—8.

24] Delacretaz F, Percy L, Schmidt PM. Histopathology of bonemarrow in HIV infection. Virchows Arch 1987;411:543—6.

25] Saif MW, Kevin S. Multiple myeloma and HIV infection:an association or a coincidence. J Appl Res 2005;5:318—24.

26] Heriot K, Hallquist AE, Tomar RH. Paraproteinemia inpatients with AIDS or LAS. Clin Chem 1985;31:1224—6.

27] Grateau G, Bachmeyer C, Blanche P, Jouanne M, Tulliez M,Galland C, et al. Haemophagocytic syndrome in patientsinfected with the human immunodeficiency virus: ninecases and a review. J Infect 1997;34:219—25.

28] Tiab M, Mechinaud Lacroix F, Hamidou M, Gaillard F,Raffi F. Reactive haemophagocytic syndrome in AIDS. AIDS1996;10:108—11.

29] Costello C. Haematological abnormalities in humanimmunodeficiency virus (HIV) disease. J Clin Pathol1988;41:711—5.

30] Niedt GW, Schinella RA. Acquired immunodeficiency syn-

drome. Clinicopathologic study of 56 autopsies. Arch PatholLab Med 1985;109:727—34.

31] Shenoy CM, Lin JH. Bone marrow findings in acquiredimmunodeficiency syndrome (AIDS). Am J Med Sci1986;292:372—5.

sciencedirect.com