Case 20
Case 20
of thickening of the gallbladder wall, and a cyst in the Urea nitrogen (mg/dl) 33
left kidney. A malarial smear was negative. The fever Creatinine (mg/dl) 2.1
recurred, and the administration of ciprofloxacin for Glucose (mg/dl) 139
six days and additional treatment with metronida- Sodium (mmol/liter) 137
zole (400 mg twice a day) did not relieve it. A cul- Potassium (mmol/liter) 4.7
ture of a blood specimen, obtained three days after Chloride (mmol/liter) 98
Carbon dioxide (mmol/liter) 29.8
the initiation of treatment with ciprofloxacin, yielded
Magnesium Normal
gram-negative bacilli, which were identified as b-lac- Calcium Normal
tamase–negative Moraxella phenylpyruvica. The micro- Phosphorus Normal
organisms were sensitive to amoxicillin, cefuroxime, Bilirubin
ciprofloxacin, gentamicin, and tetracycline and were Total Normal
resistant to trimethoprim. The patient was admitted Conjugated Normal
Enzymes Normal
to this hospital.
The patient’s weight had dropped from 93 to 78 kg *To convert the value for urea nitrogen to milli-
during the year before admission. He reported that moles per liter, multiply by 0.357. To convert the val-
ue for creatinine to micromoles per liter, multiply by
he had not ingested unpasteurized milk and had con- 88.4. To convert the value for glucose to millimoles
sumed little meat. per liter, multiply by 0.05551.
N Engl J Med, Vol. 344, No. 26 · June 28, 2001 · www.nejm.org · 2009
diogram showed nonspecific ST-segment and T-wave material, revealed multiple, calcified left-sided hilar and
abnormalities; the findings were otherwise normal. subcarinal lymph nodes; the lungs appeared normal.
Thoracic radiographs showed no abnormalities. A A CT scan of the abdomen and pelvis, obtained after
transrectal ultrasonographic study of the prostate the oral administration of contrast material, showed
gland showed no evidence of an abscess. a gallstone.
Urine and three samples of blood were obtained for On the fifth hospital day, a radionuclide bone scan
culture. Cefuroxime (750 mg every eight hours) was showed small foci of increased uptake in the cervical
administered. During the next several days, the peak spine, right foot, and right knee; these findings were
temperature each day ranged from 36.9° to 37.6°C. most consistent with the presence of degenerative
The urine culture and an assay for Clostridium diffi- changes. There was no evidence of osteomyelitis. On
cile toxin were negative. the sixth hospital day, one of the three blood-culture
On the second hospital day, an ultrasonographic specimens obtained on admission yielded gram-neg-
examination of the kidneys revealed bilateral cysts and ative rods.
a hyperechoic stone on the left side, without evidence A diagnostic procedure was performed.
of hydronephrosis. A transthoracic cardiac ultrason-
ographic study showed thickening of both mitral leaf- DIFFERENTIAL DIAGNOSIS
lets, calcification of the posterior mitral annulus, left DR. DIMITRIOS P. KONTOYIANNIS*: May we re-
atrial dilatation, thickened aortic leaflets, symmetric view the pertinent radiographic findings?
left ventricular hypertrophy, and segmental left ven- DR. CONSTANTINO S. PEÑA: A CT examination
tricular dysfunction. The estimated ejection fraction of the thorax (Fig. 1), performed after the oral admin-
was 60 percent. A transesophageal cardiac ultrason- istration of contrast material, showed small, calcified
ographic examination showed no evidence of vegeta- subcarinal and left-sided hilar lymph nodes that were
tions or paravalvular abscesses; there was moderate consistent with prior granulomatous disease.
atheromatous disease of the descending thoracic aorta DR. KONTOYIANNIS: A traveler returning from the
with several protruding plaques, including one that Middle East who presents with a nonspecific, pro-
appeared ulcerated. tracted febrile illness may have any one of a variety of
On the third hospital day, a tuberculin skin test bacterial, parasitic, rickettsial, fungal, viral, or non-
(purified protein derivative, 5 TU) was positive, with infectious diseases (Table 3).1-15
an induration that was 16 mm in diameter. (The pa-
tient had received bacille Calmette–Guérin vaccine in Bacterial Infections
the past and a booster dose more recently; a tubercu- Enteric fevers, especially typhoid fever, often have
lin skin test with purified protein derivative, performed no distinctive clinical and laboratory features and
after the booster dose, was negative.) Additional blood should therefore be considered in all febrile patients
and urine specimens were obtained for culture, and who have traveled to areas where enteric fevers are
a single dose of vancomycin was given. A computed endemic, such as the Middle East, even if there are no
tomographic (CT) examination of the thorax (Fig. 1), gastrointestinal symptoms.2 The ineffectiveness of cip-
performed after the oral administration of contrast rofloxacin in this patient, however, is evidence against
infection with various salmonella species that cause
enteric fever.3 Also, this patient had no gastrointesti-
nal symptoms suggestive of an enteric infection caused
by nontyphoidal salmonella species, Shigella flexneri,
enteroinvasive strains of Escherichia coli, or various
campylobacter or noncholera vibrio species.2 Finally,
none of the aforementioned gram-negative bacteria
have ever been misidentified as M. phenylpyruvica.
Another bacterial possibility is leptospirosis, a cosmo-
politan zoonosis,4 but it is very unlikely in the ab-
sence of a history of exposure to the causative organ-
ism and a biphasic presentation with an influenza-like
illness and conjunctivitis, followed by meningitis, uvei-
tis, rash, or hepatorenal dysfunction. I shall return to
two other bacterial diseases that are more likely to
have caused this patient’s illness and choose one of
them as the almost certain diagnosis.
2010 · N Engl J Med, Vol. 344, No. 26 · June 28, 2001 · www.nejm.org
FINDINGS DISORDERS
Parasitic and Rickettsial Infections East, where tuberculosis is a serious health problem.10
Visceral or viscerotropic leishmaniasis, acute schis- In view of the patient’s age, a reactivated infection is
tosomiasis, malaria, and Mediterranean fever due to more likely. Extrapulmonary tuberculosis in an older
Rickettsia conorii, all of which are seen in the Middle patient can easily be overlooked, since the clinical
East, are ruled out in this case by the absence of clin- findings are frequently nonspecific and insidious, and
ical or laboratory evidence of these diseases.5,6 In a pa- they may accompany coexisting diseases, masking the
tient with unexplained, recurrent gram-negative bac- tuberculosis.11,12
teremia, the possibility of a hyperinfection syndrome This patient probably does not have isolated skel-
due to Strongyloides stercoralis, a cosmopolitan intes- etal or genitourinary tuberculosis, since in such cases
tinal nematode, should be considered. In that disor- the overall clinical picture tends to reflect the main
der, however, the bacteremia is always due to common focus of infection.13,14 In addition, he does not have
enteric gram-negative bacteria, the clinical manifes- any evidence of miliary tuberculosis. Nevertheless, the
tations are severe, and risk factors (e.g., corticoster- available data do not entirely rule out the possibility
oid use) are usually present.7 of tuberculosis.
Although endemic fungal infections are uncommon The distinctive constellation of findings in this case
in the Middle East,8 chronic pulmonary histoplasmo- suggests the diagnosis of infection with Brucella me-
sis should be considered in this case because of the litensis. There are several important clues to this di-
finding of calcified intrathoracic lymph nodes.8 How- agnosis: the patient’s recent travel to the Middle
ever, the patient had no risk factors for the develop- East, an area where B. melitensis, an enzootic disease,
ment of histoplasmosis, and his clinical presentation is highly endemic1; a clinical picture that is very sug-
was not characteristic of it. gestive of brucellosis15-19; the initial identification of
the gram-negative bacillus as M. phenylpyruvica, which
Mycobacterial Infection is often confused with B. melitensis by laboratories
In a patient with an involuntary weight loss of 15 kg, using rapid, automated commercial methods 20-22; and
a subacute, nonspecific febrile illness, a newly posi- the fact that moraxella species other than M. catar-
tive tuberculin test, a mild monocytosis, and intratho- rhalis rarely infect humans.22 Since I have found no re-
racic lymphadenopathy, a diagnosis that merits serious ports of cases of systemic infections caused by M. phen-
consideration is concomitant or antecedent myco- ylpyruvica, I shall now discuss brucellosis and possible
bacterial infection in the form of subacute extrapul- coexisting diseases as the explanation for the clinical
monary tuberculosis with possible pulmonary, renal, picture in this case.
or osseous involvement.9 In this case, active tuber- The clinical findings are compatible with the pres-
culosis could have been either a reactivation of latent ence of B. melitensis infection for several reasons. First,
disease or a new infection acquired in the Middle the protean, frequently elusive, nonspecific manifes-
N Engl J Med, Vol. 344, No. 26 · June 28, 2001 · www.nejm.org · 2011
tations of this disease, including a nonspecific febrile seminated intravascular coagulation.30-32 Most patients
syndrome without localizing signs and symptoms, with brucella endocarditis have underlying valvular
have long been recognized.15-19 Neither a normal heart disease.30-33
white-cell count nor the absence of neutrophilia rules Transesophageal echocardiography has a very high
out the diagnosis.16 Second, the absence of a history sensitivity for the diagnosis of infective endocarditis.34
of ingestion of unpasteurized dairy products in this In this case, the transesophageal echocardiographic
case does not rule out brucellosis, since the patient study showed only thickening of the mitral and aortic
might have been exposed to the organism in another leaflets, a common incidental finding in elderly pa-
manner.1 Third, even though other pathogenic bru- tients but not a major risk factor for infective endo-
cella species, such as B. abortus, B. suis, and B. canis, carditis.35 The fever and anemia in combination with
can cause a clinical picture identical to that caused sclerotic leaflets may account for the patient’s systol-
by B. melitensis, these other species have never been ic murmur.36
misidentified as M. phenylpyruvica, as in this case, nor In the rare cases in which brucella species cause in-
are they prevalent in the Middle East.1,15-19 fective aortitis, it is always associated with backache
and an expanding abdominal aortic aneurysm — find-
Identification of the Organism ings that were not present in this case.37 Moreover,
Misidentification of B. melitensis is not uncommon I found no reports of aortitis caused by brucellosis
in clinical microbiology laboratories in both Europe in patients with atherosclerotic plaques alone.
and the United States that routinely use a rapid, au-
tomated commercial method (API [Analytical Pro- Explanation of Organ Abnormalities
file Index] 20 NE system, BioMerieux) to identify Could a multisystem disease due to brucella account
gram-negative bacilli not belonging to the Entero- for this patient’s renal, pulmonary, and osseous ab-
bacteriaceae family.22 Brucella species may grow quick- normalities? Renal abnormalities in cases of brucella
ly with the use of newer blood-culture systems.23,24 infection are rare,18,19 and the absence of any associ-
The description of the gram-negative bacterium as a ated symptoms, the negative urine cultures, and the
bacillus (in the first positive blood culture) and as a nonspecific appearance of the renal cysts in the present
rod (in the second positive blood culture), the absence case make the possibility of renal involvement by bru-
of b-lactamase production, and the susceptibility pro- cella very unlikely. In rare cases, brucellosis has caused
file of the organism are all consistent with infection pneumonia after occupational inhalation of the or-
by a brucella species, which can have a coccoid or ganism.38 Although the yield of sputum cultures for
rod-like appearance.22,25,26 brucella species is very low,19,38 attempts to culture
the organism as well as tubercle bacilli would be advis-
Explanation of the Relapse of Disease able in this case, even though the patient has no pul-
If this patient had brucellosis, why did his disease monary symptoms. The possibility of early spondylitis
relapse? The most likely explanation is inappropriate due to brucella merits consideration in this patient
monotherapy with a quinolone (ciprofloxacin) and a because it is very common in brucellosis.15-19,39 Final-
cephalosporin (cefuroxime).26,27 Drug-resistant bru- ly, the absence of sacroiliac- or sternoclavicular-joint
cellosis is very unlikely, since resistant brucella strains involvement makes the diagnosis of brucella arthritis
are rarely found in patients who remain infected after unlikely.
they have received antibiotic therapy.26 Two prospec-
tive studies in Spain28,29 identified several independent Possibility of Concomitant Tuberculosis
risk factors for a relapse of brucellosis: suboptimal Extrapulmonary tuberculosis can simulate brucel-
antibiotic therapy, male sex, and various clinical mark- losis, especially when the spine is involved.13 A diag-
ers indicative of severe infection (positive blood cul- nosis of brucellosis does not exclude the possibility
tures during the initial infection, illness for less than of associated tuberculosis, and a partial response to
10 days before the start of treatment, thrombocyto- treatment for brucellosis with drugs that have activity
penia, and a high initial temperature). This patient had against mycobacteria, such as rifampin, streptomycin,
almost all these risk factors. and quinolones, may make it difficult to establish a
This patient’s relapsing bacteremia may be due to diagnosis of concomitant tuberculosis. Plain radio-
a persistent visceral nidus of infection, valvular veg- graphs and CT or magnetic resonance imaging studies
etation, or an infected atheromatous plaque. An oc- occasionally help differentiate between osseous tuber-
cult splenic or prostatic abscess was ruled out, how- culosis and brucellosis.15 Paraspinal abscesses and ver-
ever, by the imaging studies. Brucella endocarditis, a tebral collapse are more common in the former dis-
rare but potentially lethal complication of brucello- order, whereas a combination of reactive sclerosis and
sis,30-32 has a predilection for the aortic valve. It is usu- lytic changes are more likely in the latter.15,19 If the
ally manifested as a destructive process causing ab- spinal findings in this case suggest the presence of
scesses in the myocardium and aortic root, rapidly spondylitis or diskitis, a fine-needle aspiration biopsy
evolving heart failure, major thromboemboli, and dis- of the lesion may be helpful. I would also rule out
2012 · N Engl J Med, Vol. 344, No. 26 · June 28, 2001 · www.nejm.org
active pulmonary and renal tuberculosis by obtain- of serum, detects antibodies to the common brucella
ing sputum and urine samples, respectively, for acid- pathogens, B. abortus, B. suis, and B. melitensis, but
fast bacilli smears and cultures. not to B. canis. The titer of IgM and IgG antibodies
In summary, the most likely diagnosis in this case is in this case was high (1:640). Typically, elevated im-
subacute, partially treated brucellosis caused by B. me- munoglobulin titers become detectable one to two
litensis. Confirmatory studies could include a serolog- weeks after infection and may persist for weeks or
ic test for brucella species and identification of the months after effective therapy has been administered.
gram-negative bacillus in the patient’s blood as B. me- Moraxella organisms and brucella have similar ap-
litensis. pearances on Gram’s staining, but the former have
the spherical shape of cocci, whereas the latter have
CLINICAL DIAGNOSIS the elongated shape of coccobacilli. The two organ-
Subacute brucellosis. isms have similar biochemical profiles, including de-
tectable catalase, oxidase, and urease activity. Brucella
DR. DIMITRIOS P. KONTOYIANNIS’S may also resemble other gram-negative coccobacilli
DIAGNOSIS such as haemophilus species.
Brucellosis due to Brucella melitensis. DR. TIMOTHY E. GUINEY: After the diagnosis
had been established, the patient was treated with
PATHOLOGICAL DISCUSSION doxycycline and rifampin, and the fever disappeared
DR. JAMES VERSALOVIC: Blood was obtained for rapidly. When seen four months after discharge, the
culture, and serologic studies were performed. The patient felt well. On questioning him, we learned that
blood specimen was inoculated into our routine during his travels in Lebanon, he had handled raw
blood-culture system (the BACTEC 9240 system), sheep and goat meat and had regularly consumed
and both aerobic and anaerobic bottles were used. unprocessed cheese.
B. melitensis was cultured from the aerobic bottle six
days after inoculation, signaling the presence of a ANATOMICAL DIAGNOSIS
fastidious organism. Gram’s staining showed the small, Brucellosis masquerading as moraxella infection.
gram-negative coccobacilli that are characteristic of
B. melitensis (Fig. 2). REFERENCES
The isolate was subcultured on horse-blood agar 1. Corbel MJ. Brucellosis: epidemiology and prevalence worldwide. In:
Young EJ, Corbel MJ, eds. Brucellosis: clinical and laboratory aspects. Boca
incubated in an atmosphere containing 5 percent car- Raton, Fla.: CRC Press, 1989:25-40.
bon dioxide. The isolate was nonmotile and was pos- 2. DuPont HL. Gastrointestinal infections. In: Stein JH, ed. Internal med-
itive for catalase, oxidase, and urease — findings that icine. 4th ed. St. Louis: Mosby–Year Book, 1994:1915-25.
3. Magill AJ. Fever in the returned traveler. Infect Dis Clin North Am
are consistent with the diagnosis of brucellosis. The 1998;12:445-69.
identification of B. melitensis was confirmed by exten- 4. Torten M. Leptospirosis. In: Stoenner H, Kaplan W, Torten M, eds.
sive biochemical testing and fatty-acid analysis at the CRC Handbook series in zoonoses. Section A. Bacterial, rickettsial, and
mycotic diseases. Vol. 1. Boca Raton, Fla.: CRC Press, 1979:363-421.
Centers for Disease Control and Prevention. 5. Oster CN, Tramont EC. Fever in a recent visitor to the Middle East.
Serum agglutination, or tube-based agglutination Curr Clin Top Infect Dis 1993;13:57-73.
6. Maguire JH, Gantz NM, Moschella S, Pan SC. Leishmanial infections:
a consideration in travelers returning from abroad. Am J Med Sci 1983;
285:32-40.
7. Igra-Siegman Y, Kapila R, Sen P, Kaminski ZC, Louria DB. Syndrome
of hyperinfection with Strongyloides stercoralis. Rev Infect Dis 1981;3:
397-407.
8. Wheat J. Histoplasma. In: Gorbach SL, Bartlett JG, Blacklow NR, eds.
Infectious diseases. 2nd ed. Philadelphia: W.B. Saunders, 1998:2335-44.
9. Menzies D. Interpretation of repeated tuberculin tests: boosting, con-
version, and reversion. Am J Respir Crit Care Med 1999;159:15-21.
10. Raviglione MC, Snider DE Jr, Kochi A. Global epidemiology of tu-
berculosis: morbidity and mortality of a worldwide epidemic. JAMA 1995;
273:220-6.
11. Alvarez S, Shell C, Berk SL. Pulmonary tuberculosis in elderly men.
Am J Med 1987;82:602-6.
12. King D, Davies PDO. Disseminated tuberculosis in the elderly: still a
diagnosis overlooked. J R Soc Med 1992;85:48-50.
13. Gorse GJ, Pais MJ, Kusske JA, Cesario TC. Tuberculosis spondylitis:
a report of six cases and a review of the literature. Medicine (Baltimore)
1983;62:178-93.
14. Simon HB, Weinstein AJ, Pasternak MS, Swartz MN, Kunz LJ. Geni-
tourinary tuberculosis: clinical features in a general hospital population.
Am J Med 1977;63:410-20.
15. Gotuzzo E, Carrillo C. Brucella. In: Gorbach SL, Bartlett JG, Black-
low NR, eds. Infectious diseases. 2nd ed. Philadelphia: W.B. Saunders,
1998:1837-4.
Figure 2. Blood-Culture Specimen Showing Faint Staining of 16. Young EJ. Brucella species. In: Mandell GL, Bennett JE, Dolin R, eds.
Small Coccobacilli Characteristic of Brucella melitensis (Gram’s Mandell, Douglas, and Bennett’s principles and practice of infectious dis-
Stain, ¬500). eases. 5th ed. Vol. 2. Philadelphia: Churchill Livingstone, 2000:2386-93.
N Engl J Med, Vol. 344, No. 26 · June 28, 2001 · www.nejm.org · 2013
17. Idem. Brucellosis: current epidemiology, diagnosis, and management. factors for relapse of brucellosis in humans. Clin Infect Dis 1995;20:1241-
Curr Clin Top Infect Dis 1995;15:115-28. 9.
18. Idem. An overview of human brucellosis. Clin Infect Dis 1995;21: 29. Solera J, Martinez-Alfaro E, Espinosa A, Castillejos ML, Geijo P, Rod-
283-90. riguez-Zapata M. Multivariate model for predicting relapse in human bru-
19. Idem. Clinical manifestations of human brucellosis. In: Young EJ, Cor- cellosis. J Infect 1998;36:85-92.
bel MJ, eds. Brucellosis: clinical and laboratory aspects. Boca Raton, Fla.: 30. Berbari EF, Cockerill FR III, Steckelberg JM. Infective endocarditis
CRC Press, 1989:97-126. due to unusual or fastidious microorganisms. Mayo Clin Proc 1997;72:
20. Barham WB, Church P, Brown JE, Paparello S. Misidentification of 532-42.
Brucella species with use of rapid bacterial identification systems. Clin In- 31. Chan R, Hardiman RP. Endocarditis caused by Brucella melitensis.
fect Dis 1993;17:1068-9. Med J Aust 1993;158:631-2.
21. Batchelor BI, Brindle RJ, Gilks GF, Selkon JB. Biochemical mis-iden- 32. Fernandez-Guerrero ML. Zoonotic endocarditis. Infect Dis Clin
tification of Brucella melitensis and subsequent laboratory-acquired infec- North Am 1993;7:135-52.
tions. J Hosp Infect 1992;22:159-62. 33. Flugelman MY, Galun E, Ben-Chetrit E, Caraco J, Rubinow A. Bru-
22. Miscellaneous fastidious gram-negative bacilli. In: Koneman EW, Allen cellosis in patients with heart disease: when should endocarditis be diag-
SD, Janda WM, Schreckenberger PC, Winn WC Jr. Color atlas and text- nosed? Cardiology 1990;77:313-7.
book of diagnostic microbiology. 5th ed. Philadelphia: J.B. Lippincott, 34. Daniel WG, Mügge A. Transesophageal echocardiography. N Engl J
1997:395-472. Med 1995;332:1268-79.
23. Bannatyne RM, Jackson MC, Memish Z. Rapid diagnosis of Brucella 35. Durack DT. Prevention of infective endocarditis. N Engl J Med 1995;
bacteremia by using the BACTEC 9240 system. J Clin Microbiol 1997;35: 332:38-44.
2673-4. 36. Durack DT, Lukes AS, Bright DK. New criteria for diagnosis of infec-
24. Yagupsky P, Peled N, Press J, Abramson O, Abu-Rashid M. Compar- tive endocarditis: utilization of specific echocardiographic findings. Am J
ison of BACTEC 9240 Peds Plus medium and isolator 1.5 microbial tube Med 1994;96:200-9.
for detection of Brucella melitensis from blood cultures. J Clin Microbiol 37. Aguado JM, Barros C, Gomez Garces JL, Fernandez-Guerrero ML.
1997;35:1382-4. Infective aortitis due to Brucella melitensis. Scand J Infect Dis 1987;19:
25. Ariza J, Bosch J, Gudiol F, Linares J, Viladrich PF, Martin R. Rele- 483-4.
vance of in vitro antimicrobial susceptibility of Brucella melitensis to relapse 38. García-Rodriguez JA, García-Sánchez JE, Muñoz Bellido JL, Ortiz de la
rate in human brucellosis. Antimicrob Agents Chemother 1986;30:958- Tabla V, Bellido Barbero J. Review of pulmonary brucellosis: a case report on
60. brucellar pulmonary empyema. Diagn Microbiol Infect Dis 1988;11:53-60.
26. Hall WH. Modern chemotherapy for brucellosis in humans. Rev In- 39. Ariza J, Gudiol F, Valverde J, et al. Brucellar spondylitis: a detailed
fect Dis 1990;12:1060-99. analysis based on current findings. Rev Infect Dis 1985;7:656-64.
27. Lang R, Rubinstein E. Quinolones for the treatment of brucellosis.
J Antimicrob Chemother 1992;29:357-60.
28. Ariza J, Corredoira J, Pallares R, et al. Characteristics of and risk Copyright © 2001 Massachusetts Medical Society.
2014 · N Engl J Med, Vol. 344, No. 26 · June 28, 2001 · www.nejm.org