Bone Marrow Culture Vs Blood Culture in FUO

Results: Total 57 cases of FUO were included in the study. Male female ratio was 1.22:1. Age range was fi ve to 83 years (median 30). Duration of fever was 21 to 365 days. Bacterial growth was seen in nine cases (15.78%) of BMCs and in three cases (5.26%) of corresponding BCs. Fungal or myocbacterial growth was not seen. Salmonella typhi was the commonest organism isolated in BMCs (three cases) followed by Staphylococcus aureus (two cases), Escherichia coli, Non fermenting Gram negative bacilli, Enterococcus species and Salmonella paratyphi–A (one case each). Two cases of Salmonella typhi and one case of Salmonella paratyphi–A were isolated in BCs.


INTRODUCTION
Fever of unknown origin (FUO) is a common clinical problem and an enormous number of conditions, either uncommon diseases or unusual presentations of common diseases are implicated as cause. 1 Diagnostic spectrum is changing over time because of advent of sophisticated tests; however, it still poses diagnostic challenge to the physicians.The three major categories of diseases giving rise to FUO are infections, tumors, and connective tissue diseases. 2Bone marrow cultures (BMCs) are performed in the evaluation of FUO, usually in conjunction with blood cultures (BCs), cultures of body fl uids and tissues. 3BMCs are often positive when BCs are negative, especially in cases of enteric fever. 4n cases of salmonella infection, BMCs increase the diagnostic yield by about one-third when compared with BCs. 5 In the present study we compared the usefulness of BMCs with BCs in an attempt to detect systemic infection in patient with FUO.

METHODS
This was a one year prospective cross sectional study.All the cases of FUO referred to department of pathology of Manipal Teaching Hospital, were included in the study.FUO was defi ned by the criteria of Petersdorf and Beeson: temperature of 38.3 o C (101 o F) or above persisting or recurring during a period of two weeks and seven days investigation in hospital, or if three out patient visits fail to result in a diagnosis. 6Patients not fulfi lling the above criteria were excluded from the study.
In all the cases clinical fi ndings were recorded and Patients were informed about the procedure.Venous blood is drawn for bacterial culture.Bone marrow aspirations were performed under local anesthesia from posterior superior iliac crest using Salah needle attached to 20 cc syringe, after aseptic precaution.About 10 ml of marrow was aspirated in each case for bacterial, myocbacterial and fungal cultures.For bacterial culture aspirate were inoculated into biphasic brain heart infusion medium and biphasic MacConkey medium and incubated at 37°C for seven days with regular subculture on third and sixth days till the growth was observed.Subcultures were done and incubated at 37°C for 18-24 hours.The growths in the slant were Gram stained and the isolates were identifi ed by biochemical and serological tests.For Mycobacteria, aspirates were inoculated into Lowenstain Jensen medium and incubated at 37 0 C till the appearance of the growth, maximum for eight weeks.For fungi aspirates were inoculated into Sabourad dextrose agar medium and incubated at room temperature till the appearance of the growth, maximum for 4 weeks.
The obtained data were collated and results were tabulated.The data were analyzed by measures of central tendency.

RESULTS
Total numbers of patients with fever, sent for BMCs were 74, but only 57 patients were included in the study.17 cases were excluded from the study.HIV test was negative in all the cases.30 cases were male and 27 were female.Male to female ratio was 1.22:1.Youngest patient was of fi ve year and oldest patient was 83 year old.Median age was 30 years.Six cases (10.5%) were children (<15 years) and 10 cases (17.5%) were elderly (>65 years).Table 1 shows the age distribution of 57 patients.Shortest duration of fever was 21 days and longest was 365 days with mean of 30 days (Table 2).
BMCs in nine (15.78%) out of 57 patients exhibited bacterial growth and in three cases (5.26%) corresponding BCs also exhibited bacterial growth (Table 3 and 4).None of the BMCs showed mycobacterial or fungal growth.Frequencies of various bacteria isolated in BMCs are shown in table 4; the commonest was Salmonella typhi (33.5%).In two cases bacterial growth in BMCs and BCs were comparable (Table 3).

DISCUSSION
Diagnosis of FUO requires multidisciplinary approach including clinical history.History of travel in recent past and drug history is of great importance.Bone marrow examination in FUO is an important diagnostic modality.
In present study BMCs alone yielded accurate diagnosis in 15.78% of cases.15.78% (nine cases) of BMCs and only 5.26% (three cases) of BCs showed bacterial growth (Table 3).BCs did not reveal bacterial growth in isolation without BMCs revealing bacterial growth.Except in one case out of nine cases, BMCs alone were suffi cient for the diagnosis without any role of BCs.In that one case BC revealed growth of Salmonella typhi; BMC in this case revealed growth of Enterococcus species the clinical signifi cance of which is doubtful as patient responded to treatment for salmonella infection, although Enterococcus species is known to cause septicemia and infective endocarditis, and the ubiquitous nature of this organism requires caution in establishing the clinical signifi cance. 7In one case BMC exhibited growth of non-fermenting gram negative bacilli which was not further classifi able.In this study salmonella was the commonest isolate in the BMCs, Salmonella typhi and Salmonella paratyphi-A together constituted 44.44 % of total positive bone marrow bacterial culture (Table 4).If one case of positive blood culture for Salmonella typhi is considered, enteric fever alone constituted 55.55% of total infections and 8.77% of cases of FUO.There was no mycobacterial and fungal growth in any of the cases.
In a study by Haq et al, infectious disease accounted for 63.21% of cases of FUO.Tuberculosis was the commonest accounting for 24.53%, followed by enteric fever (12.74%) and visceral leishmaniasis (9.43%). 8n a study by Jung et al, commonest cause of FUO was infections (46.4%) and enteric fever was the commonest (29.6%) among that followed by malaria (9%), and tuberculosis (5.2%). 9But these studies were not based only on BMCs and BCs, as in the present study.Nonetheless in this study enteric fever was the commonest cause of infection related FUO. 9 In a study by Farooqui and colleague organism was isolated from the bone marrow in all the cases of salmonella infection and from blood in only 66% cases.
In the same study further subtyping of the Salmonella species showed Salmonella typhi 83%, Salmonella paratyphi-A 11.8%, Salmonella paratyphi-B 4.1% and Salmonella paratyphi-C 0.9%. 4In present study Salmonella typhi accounted for 80% and Salmonella paratyphi-A accounted for 20% of total salmonella infections (Table 3 and 4).Out of fi ve cases of salmonella infection four cases (80%) showed positive BMCs while two cases (40%) showed growth in BCs as well.In one case (20%) growth was seen only in BC.Findings in this study are in accordance to study of Farooqui and colleague.BMCs could confi rm a diagnosis of enteric fever in patients with negative BCs. 4 One of the reasons for BCs being negative and BMCs being positive in enteric fever may be antibiotic treatment before collection of blood for culture.An injudicious antibiotic regimen may diminish or eliminate the organism from blood, but salmonella being an intracellular organism, persists in the reticulo-endothelial system including bone marrow. 10The intracellular location of bacteria protects them from conventional chemotherapeutic measures. 11,12Enteric fever is the only bacterial infection in humans for which bone marrow examination is recommended. 5In a study by John et al, the number of bacteria in blood but not bone marrow was correlated inversely with the duration of preceding fever. 5Thus, with increasing duration of illness the ratio of bone marrow to blood bacterial concentrations increased; the median ratio was 4.8 during the fi rst week compared with 158 during the third week.This study provided proof that the concentrations of Salmonella typhi in bone marrow are higher than in blood.Bone marrow contained over 10 times more bacteria than blood. 5ffective antibiotic pre-treatment had greater effect in reducing blood counts of bacteria compared to bone marrow count. 5How salmonella survives in the bone marrow despite the presence of high concentrations of bactericidal antibiotic which reaches high intracellular concentration is unclear.
In a study by Volk et al only 0.5% of BMCs and corresponding BCs in the patients with FUO showed clinically signifi cant growth in an immunocompromised patient and the organism isolated was Mycobacterium avium intracellularae. 3In contrast present study showed BMCs nor the BCs will detect the causes of all systemic infections; an algorithm that conserves resources should be followed in which BMCs are performed if the BCs and other easily obtained tissues have been cultured, and the results are repeatedly inconclusive.

CONCLUSIONS
A bone marrow examination including BMCs is important part of investigation of FUO, but this alone is insuffi cient to trace specifi c etiology in all the cases.A good clinical history, radiological and hematological evaluation including bone marrow coupled with BMC and serology would yield diagnosis in a high proportion of cases.Yield of BMCs is more than BCs especially in cases of suspected salmonella infection and in patients treated with antibiotics as was the case in present study; however, its role is insignifi cant in isolating mycobacteria or fungi in immuno-competent patients.

Table 5 . Comparison of 5 studies evaluating detection of organism by bone marrow culture in patients with FUO
3,[13][14][15]th in BMCs.A summary of diagnostic yield of BMCs in fi ve studies is shown in table 5.A population characteristic of Volk et al was non-restricted while a population characteristic of Marsh et al, Riley et al and Nicholas et al was HIV patients.3,[13][14][15]Inpresentstudytheisolates were bacteria but in other four studies the isolate was Mycobacterium avium intracellularae.Yield of BMCs in present study was 32 times better (15.78% versus 0.5%) than the other study(Volk et  al).In absence of hematological justifi cation for bone marrow evaluation, BMCs and BCs in a work up of patient signifi cantly increase the cost.It is diffi cult to justify performing BMCs on all patients with FUO.Neither