0International Journal of Infectious Diseases (2007) 11, 52—57
CI CC\/ICD
http://intl.elsevierhealth.com/journals/ijid
Acute brucellosis: presentation, diagnosis, and treatment of 144 cases
Panos Andriopoulosa *, Maria Tsironib, Spiros Deftereosc, Athanassios Aessoposc, Giorgos Assimakopoulosb
a Gytheio Health Care Centre, Gytheio Lakonia 23200, Greece b Internal Medicine, Sparta General Hospital, Sparta Lakonia, Greece
c 1st Department of Internal Medicine, Laikon Hospital, University of Athens Medical School, Athens, Greece
Received 22 May 2005; received in revised form 21 August 2005; accepted 12 October 2005 Corresponding Editor: Raymond A. Smego, Sohar, Oman
Summary
Objectives: Brucellosis, whether in an endemic region or not, remains a diagnostic puzzle due to occasional misleading unusual presentations and non-specific symptoms. Presented herein is our 14-year experience with acute brucellosis at Sparta General Hospital, Lakonia, Greece. Methods: A case series of 144 patients admitted to the internal medicine, pediatrics, and urology departments, through evaluation of history, occupational data, serological tests, cultures of blood and other body fluids, and imaging studies. Patients were treated with a 21-day course of intramuscular streptomycin and a prolonged two-month course of doxycycline with a six-month follow-up.
Results: Infected patients had a relevant occupational history in fewer than 20% of cases. Clinical manifestations included non-specific symptoms (fever, malaise, sweats, arthralgias, lower back pain, headache), findings such as splenomegaly (51%), osteoarticular involvement (42%), cervical lymphadenitis (31%), hepatomegaly (25%), genitourinary involvement (13% of men), cholecystitis (2%), breast abscess (0.7%), and acute abdomen (0.7%). Ninety-five percent of the patients had a serological titer >1/160 with culture-proven brucellosis. Overall, 82% of blood cultures and 100% of other body fluid cultures (synovial, bile) were positive. Ninety-seven percent of the patients were cured. Relapse in the follow-up period was observed in four patients who had not complied with treatment.
Conclusions: Brucellosis is an infection with multiple presentations, and whether in an endemic region or not, a thorough history of exposure and clinical suspicion are required since thresholds in serological evaluation may lead to misdiagnosis and withholding of adequate treatment. © 2006 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
KEYWORDS
Brucella bacteremia; Acute brucellosis; Brucella serology; Brucella localized infection
* Corresponding author. Tel.: +302733022001; fax: +302733029010. E-mail address: ermanitari@hotmail.com (P. Andriopoulos).
1201-9712/$30.00 © 2006 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijid.2005.10.011
Introduction
Since its first description by Marstan in I860,1 brucellosis remains a debilitating illness of global distribution. The Mediterranean basin and Greece in particular, are two of the most endemic regions worldwide.2 For its size, Lakonia, a small and mostly rural county of approximately 80 000 inhabitants, has an unusually high annual incidence of new infections due to poor preventive measures and inadequate public health policies. Since brucellosis presents with a wide clinical spectrum3"5 and often with vague and nonspecific symptoms, the diagnostic work-up is always a challenge for the clinician. In this paper, we report our 14-year experience with acute brucellosis at Sparta General Hospital, Lakonia, Greece, focusing on the clinical findings, occupational history, laboratory evaluation, and treatment of hospitalized patients.
Materials and methods
Sparta General Hospital is a public institution that provides secondary care to people living in the county of Lakonia. The population, comprising both Greeks and immigrants (about 4% from eastern European countries), lives in rural areas (of less than 500 inhabitants) with regular exposure to animals, or they consume local animal products. It is not uncommon for residents to come into contact with Brucella-infected sheep and goats or to acquire a food-borne infection.
One hundred and forty-four patients discharged with a diagnosis of acute brucellosis from 1990 to 2003 were identified via a review of the hospital medical records. Extracted information included demographic data, clinical presentation, serologic and blood culture results, treatment, and clinical course.
Serologic evaluation was performed using Brucella agglutination tests (the Rose Bengal slide agglutination test and the Wright standard tube agglutination (STA) test). All sera were routinely tested both in low and high dilutions (from 1/ 40 to 1/3200) in order to avoid negative results due to the prozone phenomenon. A titer of 1/160 or higher was deemed strongly suggestive of acute infection, and titers of 1/40 to 1/ 80 were considered suspicious in view of the endemicity in our region (no single titer was dismissed without being analyzed in light of historical, epidemiologic, and clinical findings). Since all patients presented with fever, blood cultures were obtained independently of STA titers. Prior to 1993, blood and other body fluids were cultured in classic biphasic Castaneda flasks. Subsequently, the microbiology department switched to an automated blood culture system (Vital Aer-BioMerieux, Marcy L'Etoile, France) that shortened the time needed to detect Gram-negative cocco-bacilli expressing urease activity and requiring CO2. Isolates were identified as Brucella species with conventional biochemical tests and agglutination tests. In view of our diagnostic limitations, further identification of Brucella strains was not performed routinely. If the clinical localization of infection was unusual, the samples were sent for further analysis and identification. Patients aged 14 years and older were treated with intramuscular streptomycin for 21 days plus a two-month course of 100 mg of doxycycline twice daily. Patients under 14 years of age were treated with oral rifampin plus co-
trimoxazole for a period of six weeks. Susceptibility to streptomycin, doxycycline, co-trimoxazole, rifampin, ciprofloxacin, and tetracycline was assessed using a Kirby—Bauer agar diffusion test. Isolates remained fully susceptible during the study period.
The mean follow-up period was six months; given the rural character and lack of adequate primary care network in our region, longer-term follow-up for most patients was not feasible. A relapse was defined as improvement after initiation of treatment and reappearance of the symptoms within a month after the end of the prescribed regimen. Cure was defined as no recurrence of symptoms and signs during the six-month follow-up.
Results
Demographic and occupational data
Included in the analysis were 144 patients. One hundred and nineteen (83%) patients were admitted to the internal medicine department, 13 to pediatrics, and 12 to urology. Overall, this represented 2—10 per 1000 patients discharged from these three departments during the study years. The annual patient distribution is shown in Figure 1, and Table 1 summarizes the demographic data of study patients. An occupational history relevant for Brucella exposure was present in only 24% of the 125 adults (<20% for the entire study cohort), while the remaining subjects had probably consumed contaminated animal products.
Clinical manifestations
Fever, malaise, sweating, lower back pain, and arthralgias were the most common presenting symptoms, each with a duration of less than seven days. All patients were febrile but none presented with sepsis. The severity of symptoms varied from mild illness to severe painful localized disease.
Table 1 Demographic and occupational data of 144 patients with acute brucellosis
Feature No. of patients Percent
Gender Male Female
Children Male Female
Adults Male Female
Rural residence Urban residence Farmer Office worker Abattoir worker Animal breeder Housewife
125 87
Cases per Year
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 21)00 2001 2002 2003 2004
Figure 1 Annual distribution of hospitalized patients with brucellosis. The 1999 peak is concordant with the recorded increase of infection rates in our region.5
Table 2 summarizes symptoms and clinical findings. Localized single organ involvement accounted for 64% of cases, while 32% of subjects presented with multiple osteoarticular, genitourinary, gastrointestinal, or hematopoietic localization (Table 3).
Diagnosis
Patients were initially evaluated with serologic tests and blood cultures. Of the 144 patients, 137 (95%) had an initial STA titer >1/160 (Figure 2) and a positive Rose Bengal test. Most of these patients had a subsequent increase in titer, although 11 (8.5% of 129, since 15 patients had a titer on presentation 1/3200) did not have such an increase. All these
STA Tilers
1/40 1/80 1/160 1/320 1/400 1/640 1/800 1/1280 1/1600 1/2560 1/3200
Figure 2 Standard tube agglutination (STA) test results on presentation.
patients were considered Brucella-infected and treatment was begun immediately.
Overall, 82% of blood cultures were positive. Negative results were associated with serologic titers higher than 1/ 1600 and localized infection affecting only a single system (e.g., monoarthritis of shoulder, spondylitis, etc.). All cultures of body fluids, such as synovial fluid and bile, and abscess aspirates grew Brucella spp. Such cultures were
Table 2 Clinical manifestations in 144 patients with acute brucellosis
No. of patients Percent
Symptoms
Fever 144 100
Malaise 140 97
Sweats 138 96
Arthralgias 125 87
Lower back pain 104 72
Headache 95 66
Findings
Splenomegaly 74 51
Osteoarticular involvementa 60 42
Cervical lymphadenitis 45 31
Hepatomegaly 36 25
Genitourinary involvementb 12 13
Cholecystitisc 3 2
Breast abscess 1 0.7
Acute abdomen 1 0.7
a Details on osteoarticular and genitourinary involvement are displayed in Table 3. b Genitourinary involvement accounts for 12 of 93 male patients.
c Cholecystitis was diagnosed in patients with fever, abdominal pain with positive Murphy's sign, ultrasound evidence of gallbladder involvement, and serologic and culture results confirming Brucella infection. Cholecystitis was both calculous (one patient) and acalculous (two patients).
Table 3 Features of Brucella infection, according to organ system
No. of patients Percent
Osteoarticular involvement 60 42
Spondylitis 41 28
Spondylitis with 3 2
neurologic impairmenta
Hip 12 8
Sacroiliitis 9 6
Knee 8 6
Shoulder 3 2
Wrist 3 2
Ankle 1 0.7
Sternoclavicular 1 0.7
Hematologic complications 28 19
Normocytic anemia 24 17
Neutropenia 6 4
Thrombocytopenia 2 1
Genitourinary involvementb 12 13
Epididymoorchitis 11 12
Prostatitis 1 0.7
Gastrointestinal involvement 15 10
Hepatitis 9 6
Cholecystitis 3 2
Diarrhea 2 1
Peritonitis 1 0.7
a Neurological impairment consisted of pain, paresthesias, and diminished muscle strength to the affected neural root; L4—L5 was the localization in each case (one bilateral, two unilateral).
b Genitourinary involvement accounts for 12 of 93 male patients.
Figure 3 Magnetic resonance imaging (MRI) of a patient with Brucella spondylitis and epidural abscess formation. The patient was medically treated with a successful outcome, with full neurological recovery, and without requiring surgical decompression.
obtained in conjunction with blood cultures and thus their diagnostic utility was complementary. Diagnosis was via serology alone for only 18% of cases.
In patients with antibody titers <1/160, appropriate clinical suspicion and further evaluation led to a correct diagnosis in seven (4.8%) patients, typically via positive blood cultures.
Treatment and outcome
Patients over 14 years of age were treated with 1 g of intramuscular streptomycin (adjusted for age in patients over 70 years) for 21 days and a two-month course of oral doxycycline 100 mg twice daily. Subjects under 14 years of age were treated with oral rifampin and co-trimoxazole according to body weight for six weeks. Surgical intervention was required only in the patient with acute abdomen.6 Patients with spondylitis and neurologic impairment (Figure 3) were treated medically and all made a full recovery. At the six-month follow-up, 140 patients were cured and a relapse was diagnosed in only four (3%). Relapse was identified to have occurred as a result of non-adherence to treatment because of photosensitivity in one subject and early self-termination of treatment in three. For these relapsed patients, clinical re-presentation was generally milder and all were retreated with the same initial drug regimen and evaluated again after six more months. One was
lost to follow-up and the remaining three were then free of signs and symptoms.
Discussion
Brucellosis is a zoonosis transmitted to humans from infected animals; Brucella melitensis is the most common and virulent cause of the disease worldwide.4 Brucellae, small aerobic Gram-negative bacilli, are killed by boiling or pasteurization of milk and milk products; they may survive up to 8 weeks in unpasteurized white soft cheese and are not killed by freezing.1'6 Infection occurs either through direct inoculation or through consumption of contaminated food. The majority of cases (84%) in our series involved food-borne transmission and this may explain the male predilection seen since it is common for men to consume local animal products in village taverns. The highest number of cases reported in 1999 corresponds to a rise in the recorded infection rates in the broader area of Peloponnisos, in which Lakonia is located.7 The dramatic increase in brucellosis from 1997 to 2000 probably represents the insufficiency of veterinary control and inadequate education of farmers. Since then, public health measures designed to control disease transmission from infected to healthy animals have resulted in a minor drop in recorded cases.
The clinical presentation of brucellosis is often nonspecific with predominating symptoms like fever, malaise, sweats, and headache. Dissemination via the bloodstream can result in involvement of almost any organ.3 For the experienced physician, localized disease is often easy to recognize, but attributing more subtle findings to brucellosis requires epidemiologic knowledge and clinical suspicion. In our series, the diagnostic procedure often required culture of body fluids in order to establish the diagnosis. In endemic areas serologic testing is considered to be diagnostic when titers of 1/320 to 1/640 are present,8 while titers of 1/160 are considered suspicious. The sensitivity and specificity of a serologic test are generally very high.8 However, 18 of our patients had an initial titer of 1/160 and 20 a highest titer of 1/160. Since all of these patients had culture-proven brucellosis, we emphasize that a threshold cannot be defined and all titers, whether in an endemic region or not, must be viewed as suspicious. Therefore, when low antibody titers are encountered, cultures of body fluids are mandatory in order to establish a diagnosis. The rate of positive blood cultures in brucellosis ranges from 15% to 80%.1 Acute brucellosis usually has a higher frequency of culture positivity, while negative results are more typically seen with the chronic form of the disease or with localization of infection in a single organ. The use of automated blood culture systems has shortened the time needed for detection. In our study, acute brucellosis (duration of symptoms <7 days) was associated with a positive blood culture rate of 82%, and 100% of cultured body fluids or abscess aspirates grew Brucella organisms. These results are similar to those previously reported in the literature.1,3,6,9
Osteoarticular involvement, especially spondylitis10 is the most common form of localized brucellosis.1,11 The infection may also affect prosthetic joints12 or become clinically significant after a latent period.13 When disease
is confined to a single joint, blood cultures may not be positive and serologic and culture tests of the synovial fluid may establish the diagnosis. Blood leukocyte counts usually indicate bacterial inflammation.14 Genital infection is also frequently seen and usually presents as epididymoorchitis or occasionally as prostatitis, and most cases have multiorgan involvement.15,16 Development of necrotizing orchitis may require surgical treatment.15 Brucellosis can affect virtually any tissue or organ, and various literature reports describe Brucella infection of the respiratory system,17 the skin,18 and the gastrointestinal tract,19,20 and involving children.21 The clinician should conduct a thorough history looking for animal or food-borne exposures when attempting to explain non-specific symptoms suggesting brucellosis, but he/she must bear in mind that many patients may not recall a relevant epidemiologic exposure.
Brucella infection commonly causes mild hematologic abnormalities2 such as anemia and leukopenia. Thrombocy-topenia is far less common (<2% of cases) but there are reports of thrombocytopenic purpura associated with brucellosis,22 with high mortality due to hemorrhage into the central nervous system.
For brucellosis the World Health Organization (WHO) recommends six weeks of combination treatment with oral doxycycline plus rifampin, or six weeks of doxycycline plus 21 days of streptomycin.23 Many studies have evaluated the use of fluoroquinolones24 or macrolides25 without demonstrating any in vivo superiority of these newer antibiotics. In vitro susceptibility to all of these agents remains very high. Monotherapy of brucellosis is associated with unacceptably high rates of clinical relapses and is not recommended.26 Drug treatment for all of our adult patients, independent of the site of infection, included streptomycin plus a prolonged two-month course of doxycycline. This regimen was chosen because of the rural character of our region, which made proper monitoring of liver function tests (given the prolonged course of rifampin) impossible, and also because of better compliance with a simpler, single drug oral regimen. All 144 patients with acute infection responded favorably to treatment without long-term complications.
Several limitations are present in our study. First, our laboratory could not identify the isolated Brucella strains because of diagnostic constraints. Newer diagnostic techniques such as IgG and IgM antibodies and PCRfor Brucella were not employed because these tests are not available at our institution. Imaging studies (e.g., magnetic resonance imaging of the spine) were not routinely used. Most importantly, our series represents only a fraction of brucellosis cases in our county over the 14 study years. Private practitioners rarely record such infections and we could not accurately estimate their true incidence.
In conclusion, we emphasize that clinicians should never forget brucellosis in the differential diagnosis of febrile illness, especially in geographic areas where the disease in endemic. The potential for unusual clinical presentation and the low titers of serologic reactivity serve to remind us that brucellosis remains always a diagnostic challenge requiring clinical suspicion and thorough evaluation. Furthermore, control and eradication of brucellosis depends on the education of farmers, veterinarians, and the general public and requires close epidemiologic surveillance of animal and human cases.
Acknowledgments
We are indebted to George Papadopoulos, Associate Professor of Pharmacology, University of Athens Medical School, for his comments on the paper.
Conflict of interest: No conflict of interest to declare.
References
1. DoganayM, Aygen B. Human brucellosis: an overview. Int J Infect Dis 2003;7:173-82.
2. Corbel MJ. Brucellosis: an overview. Emerg Infect Dis 1997;3: 213-21.
3. Young EJ. An overview of human brucellosis. Clin Infect Dis 1995;21:283-9.
4. Andriopoulos P, Tsironi M, Asimakopoulos G. Acute abdomen due to Brucella melitensis. Scand J Infect Dis 2003;35: 204-5.
5. Tsironi M, Andriopoulos P, Kalkani M, Asimakopoulos G. Human mammary abscess caused by Brucella melitensis: a case report. Int J Infect Dis 2003;7:236.
6. MadkourMM, Kasper DL. Brucellosis. In: Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson JL, editors. Harrison's principles of internal medicine. 15th ed. New York: McGraw Hill; 2001. p. 1166-9.
7. Brucellosis in Peloponnesus. KEEL 2000. Athens, Greece: Hellenic center for infectious diseases control.
8. Mert A, Ozaras R, Tabak F, Bilir M, Yilmaz M, Kurt C, et al. The sensitivity and specificity of Brucella agglutination tests. Diagn Microbiol Infect Dis 2003;46:241-3.
9. Memish Z, Mah MW, Al Mahmoud S, Al Shaalan M, Khan Y. Brucella bacteremia: clinical and laboratory observation in 160 patients. J Infect 2000;40:59-63.
10. Solera J, Lozano E, Martinez-Alfaro E, Espinosa A, Castillejos M, Abad L. Brucellar spondylitis: review of 35 cases and literature survey. Clin Infect Dis 1999;29:1440-9.
11. Tasove Y, Saltoglou N, Sahin G, Aksu HSZ. Osteoarthricular involvement of brucellosis in Turkey. Clin Rheumatol 1999; 18:214-9.
12. Weil Y, Mattan Y, LiebergallM, RahavG. Brucella prosthetic joint infection: a report of 3 cases and a review of the literature. Clin Infect Dis 2003;36:e81-6.
13. Kubler PA, Klestov AC. Osteoarticular brucellosis with long latent period. Rheumatol 2001;20:444-6.
14. Press J, Peled N, Buskila D, Yagupsky P. Leukocyte count in synovial fluid of children with culture-proven brucellar arthritis. Clin Rheumatol 2002;21:191-3.
15. Navarro-Martinez A, Solera J, Corredoira J, Beato JL, Alfaro-Martinez E, Atienzar M, et al. Epididymoorchitis due to Brucella melitensis: a retrospective study of 59 patients. Clin Infect Dis 2001;33:2017-22.
16. Cesur S, Ciftci A, Sozen TH, Tekeli E. A case of epididymoorchitis and paravertebral abscess due to brucellosis. J Infect 2003;46: 251-3.
17. PappasG, Bosilkovski M, Akritidis N, Mastora M, Krteva L, Tsianos E. Brucellosis and the respiratory system. Clin Infect Dis 2003;37:e95-9.
18. Milionis H, Christou L, Elisaf M. Cutaneous manifestations in brucellosis: case report and review of the literature. Infection 2000;28:124-6.
19. Miranda RT, Gimeno AE, Rodriguez TF, de Arriba JJ, Olmo DG, Solera J. Acute cholecystitis caused by Brucella melitensis: case report and review. J Infect 2001;42:77-8.
20. Gencer S, Ozer S. Spontaneous bacterial peritonitis caused by Brucella melitensis. Scand J Infect Dis 2003;35:341-3.
21. Shaalan MA, Memish ZA, Mahmoud SA, Alomari A, Khan MY, Almuneef M, et al. Brucellosis in children: clinical observations in 115 cases. Int J Infect Dis 2002;6:182-6.
22. Young E, Tarry A, Genta R, Ayden N, Gotuzzo E. Thrombocytopenic purpura associated with brucellosis: report of 2 cases and literature review. Clin Infect Dis 2000;31:904-9.
23. Ariza J, Gudiol F, Pallares R, Viladrich PF, Rufi G, Corredoira J, et al. Treatment of human brucellosis with doxycycline plus rifampin or doxycycline plus streptomycin. A randomized, double blind study. Ann Intern Med 1992;117:25-30.
24. Agalar C, Usubutun S, Turkyilmaz R. Ciprofloxacin and rifampi-cin versus doxycycline and rifampicin in the treatment of brucellosis. Eur J Clin Microbiol Infect Dis 1999; 18: 535—8.
25. Solera J, Beato J, Martinez-Alfaro E, Segura J, de Tomas E. Azithromycin and gentamycin therapy for the treatment of humans with brucellosis. Clin Infect Dis 2001;32:506—9.
26. Rolain JM, Maurin M, Raoult D. Bactericidal effect of antibiotics on Bartonella and Brucella spp: clinical implications. J Antimi-crob Chemother 2000;46:811—4.
