Brucellosis, also known as “undulant fever,” “Malta fever,” “Mediterranean fever,” or “Gibraltar fever,” is a widespread global zoonotic disease [1,2,3], which is endemic in Latin America, Southern Europe, Africa, Mediterranean basins, and the Middle East. It also has a high incidence rate among central Asian countries such as Kyrgyzstan and Azerbaijan [1, 3,4,5]. A recent systematic review conducted in Iran showed that the annual incidence rate of brucellosis in the country is still high (0.001%, 95% confidence interval [CI] = 0.0005–0.0015%) [6], and up to 25% of infected cases are children [1, 7].

The disease is caused by Brucella species, which are small, pleomorphic, fastidious, non-motile, gram-negative coccobacilli [1, 4, 8,9,10,11,12]. It is the primary infection of domestic and wild animals, including goats, sheep, camels, cows, pigs, and dogs, that can be transmitted to humans [4, 11, 13, 14]. Transmission occurs via ingestion of infected meat, raw unpasteurized milk, and dairy products; or through direct contact with infected animals and their secretions; or through the breathing of infectious aerosols [4, 5, 10,11,12, 14, 15]. Mother-to-child transmission is an infrequent but probable route of disease transmission [9].

Brucellosis involves multiple systems of the body and has various clinical manifestations such as fever, fatigue, arthralgia, muscle pain, rash, respiratory and cardiac complications, and orchitis/epididymitis [2, 11, 16]. The differences in manifestations among children and adults are not considerable. The disease affects children of any age group or gender [3], with its acute and sub-acute forms being more common in children [3, 12].

The purpose of this study was to determine the main clinical, laboratory, and epidemiologic characteristics of pediatric patients with brucellosis who were hospitalized at the Children’s Medical Center (CMC) Hospital, a referral educational hospital in Tehran, Iran, during a 6-year period.

Patients and methods

This study was conducted at the CMC Hospital, Tehran, Iran, between May 2011 and December 2016. The medical records of all children with a definitive diagnosis of brucellosis admitted to the CMC Hospital were reviewed retrospectively. For each patient, a questionnaire was provided that contained demographic characteristics such as sex, age, nationality, date of admission, city of residence, history of ingestion of unpasteurized dairy products, history of brucellosis in the family, and history of exposure to infected animals, as well as clinical information such as signs and symptoms of the patient, laboratory findings, history of disease relapse, and applied treatment regimens.

The disease was diagnosed based on the presence of clinical signs and symptoms suggestive of brucellosis (such as fever, weight loss, anorexia, malaise and fatigue, arthralgia, sweating, nausea and vomiting, cough, etc.), together with positive serum agglutination tests (titers 1:160 or greater in the Wright test and titers 1:40 or greater in Coombs Wright and 2‑mercaptoethanol [2 ME] tests) or isolation of Brucella spp. from patients’ blood or bone marrow cultures. Blood cultures were performed using the BACTEC 9120 Blood Culture System (Becton Dickinson & 48 Co., Franklin Lakes, NJ, USA). If no growth was seen after 7 days, incubation continued for 21 days, whereafter subcultures were inoculated on chocolate agar plates. Identification of Brucella spp. isolates was performed by Gram staining and conventional biochemical tests such as oxidase, catalase, urease production, methyl red/Voges–Proskauer (MR-VP) test, citrate utilization, indole test, motility (at both 37 °C and 20 °C), hydrogen sulfide production, etc. [17].

Laboratory workups such as complete blood cell (CBC) and differential counts, hemoglobin levels, erythrocyte sedimentation rate (ESR), C‑reactive protein (CRP), and liver function tests were performed for all patients and interpreted based on the reference values specific for infancy and childhood [18].

Neurobrucellosis was diagnosed according to the following principles: the presence of clinical symptoms (such as insomnia, amenorrhea, incontinence, neck stiffness, confusion, depression, etc.) in combination with the isolation of Brucella spp. from cerebrospinal fluid (CSF); or detection of anti-Brucella antibodies in the CSF; or detection of lymphocytosis, elevated levels of protein, and diminished levels of glucose in the CSF; and outcomes of cranial magnetic resonance imaging (MRI) or computed tomography (CT) scan [19].

A brucellosis relapse was considered as the re-emergence of brucellosis clinical manifestations together with the positive laboratory results (blood culture or serological tests) 2 months to 1 year after finishing the treatment [20].

Descriptive statistics were used to characterize patients’ features. P values were determined using the chi-square test. A P value ≤ 0.05 was considered as significant. Analyses were performed using SPSS (version 16; SPSS Inc., Chicago, IL, USA) software.

Results

Of the total 8018 patients admitted to the infectious diseases ward of the CMC Hospital between May 2011 and December 2016, 43 patients were diagnosed as Brucella-infected cases (0.54%). Among them, 26 (60.5%) patients were male and 17 (39.5%) were female (Table 1). Patientsʼ age ranged from 1 to 13 years (mean ± standard deviation [SD]: 7.02 ± 3.5 years). There was no statistically significant difference between the age and sex of patients (P > 0.05). Three cases (7%) had a history of brucellosis in their families and 88% (N = 38) reported a history of ingestion of raw or unpasteurized dairy products. Eleven patients (26%) had contact with suspicious animals. There was no statistically relevant association between the patients’ gender and route of disease transmission (P > 0.05). All patients were Iranian, and the total number of reported patients in Tehran was higher than in other cities (N = 12, 28%).

Table 1 Demographic data of patients with brucellosis (N = 43)
Full size table

The lowest frequency of brucellosis was recorded in 2011 (N = 2, 5%) and the highest rates were observed in 2013 (N = 10, 23%) and 2015 (N = 11, 26%), (Table 2). Most cases were admitted in summer (N = 14, 33%) and spring (N = 12, 28%), and the lowest rate of cases was detected in winter (N = 7, 16%; Table 3).

Table 2 Frequency of brucellosis cases in different years of the study period
Full size table
Table 3 Frequency of brucellosis cases in different seasons of the year
Full size table

Fever (N = 39, 91%), arthralgia (N = 33, 77%), and malaise (N = 33, 77%) were the main complaints and symptoms of patients, while only one patient (2%) had night sweating (Table 4). Thirty patients (70%) had the classic triad of undulating fever, sweating, and arthralgia. Other clinical findings included arthritis (N = 22, 51%), anorexia (N = 19, 44%), weight loss (N = 16, 37%), abdominal pain (N = 11, 26%), headache (N = 7, 16%), diarrhea (N = 7, 16%), splenomegaly (N = 6, 14%), vomiting (N = 6, 14%), cough (N = 5, 12%), and hepatomegaly (N = 4, 9%). Neurobrucellosis was seen in 2 patients (5%). One of these patients was a 9-year-old girl who was admitted to our hospital with a history of headaches and back pain existing for 1 year. Following a decrease of alertness, neuroimaging studies including head CT and MRI were performed. Tuberculosis was suspected, but the biopsy was negative for acid-fast bacilli. With suspicion of brucellosis, Wright, Coombs Wright, and 2 ME tests were performed. The 2 ME titer was 1:320, and Wright and Coombs Wright tests were 1:320 and 1:160, respectively.

Table 4 Clinical manifestations of patients with brucellosis (N = 43)
Full size table

The other case was a 13-year-old boy who was admitted with headache, vomiting, mild to moderate grade fever, and abdominal pain existing for 8 months. Paraclinical findings included hypereosinophilia in CBC (40%); 2 ME titer was 1:160 and Wright test was 1:320. With suspicion of brucellosis, treatment was initiated with rifampin, doxycycline, and gentamycin. In both cases, the blood and CSF cultures were negative and patients had a history of ingestion of raw or unpasteurized dairy products and contact with suspicious animals.

Table 5 shows hematological findings, the results of blood and bone marrow cultures, and results of serum agglutination tests of 43 children suffering from brucellosis. Leukocyte counts were normal in 33 patients (77%), 10 patients (23%) had leukocytosis, and no patients (0%) had leukopenia. Of all patients, 51% (N = 22) had lymphocytosis. The level of hemoglobin was normal in 15 patients (35%), whereas the remaining patients had anemia (N = 28, 65%). Thrombocytopenia was observed in 4 cases (9%). An elevated ESR was seen in 37 patients (86%) and elevated CRP values were recorded in 29 patients (67%). Increased levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were observed in 15 (35%) and 10 (23%) patients, respectively.

Table 5 Laboratory findings in patients with brucellosis (N = 43)
Full size table

Blood culture was accomplished in 37 patients and positive in 11 patients (30%). Bone marrow culture was performed for only 13 patients and was positive in 4 cases (31%). Wright tests were performed for all patients, and 29 cases (67%) had an agglutination titer of 1/160 or higher. Coombs Wright and 2 ME tests were performed in 37 and 40 cases, respectively. Positive titers were observed in 34 cases (92%) in Coombs Wright and 34 cases (85%) in 2 ME. The highest titers for Wright and Coombs Wright were 1:2560 and 1:5120, respectively, and were observed in 1 patient. In 9 patients (21%) agglutination titers < 1:160 were recorded in the Wright test, but the titers were positive in Coombs Wright.

Patients received various antibiotic combination regimens that consisted of two- or three-drug therapy. The median length of antibiotic therapy was 12 weeks (range 2–24 weeks). In patients ≤ 8 years (N = 27, 63%), the following drug combinations were prescribed: trimethoprim-sulfamethoxazole and rifampicin (N = 18, 42%); trimethoprim-sulfamethoxazole, rifampicin, and gentamicin (N = 6, 14%); rifampicin and ciprofloxacin (N = 2, 4%); trimethoprim-sulfamethoxazole, rifampicin, and doxycycline (N = 1, 2%); and streptomycin and clarithromycin (N = 1, 2%).

In patients older than 8 years, six drug combinations were used: the combination of trimethoprim-sulfamethoxazole and rifampicin (N = 6, 14%); doxycycline and rifampicin (N = 5, 12%); trimethoprim-sulfamethoxazole, rifampicin, and gentamicin (N = 1, 2%); doxycycline, rifampicin, and gentamicin (N = 1, 2%), doxycycline, rifampicin, and streptomycin (N = 1, 2%); and trimethoprim-sulfamethoxazole, doxycycline, and streptomycin (N = 1, 2%).

Relapse occurred in 9 patients (21%), other patients did not experience recurrence of signs and symptoms of the disease. Of the patients with relapse, 44% (4 out of 9) were treated by the combination of trimethoprim-sulfamethoxazole and rifampicin; however, the difference between the type of treatment and rate of relapse was not statistically significant (P > 0.05). No deaths were observed among infected cases.

Discussion

Although the course of brucellosis is benign [2], if left untreated, it can become chronic and cause severe complications including endocarditis, meningoencephalitis, granulomatous hepatitis, uveitis, optic neuritis, pancytopenia, chronic anemia, spondylodiscitis, and sacroiliitis [9]. Despite control of brucellosis in the developed world, the disease remains an important health problem in developing countries [11] such as Iran, which is one of the main endemic areas [6]. Pediatric brucellosis is very important in Iran [18], rendering necessary the study of its clinical and epidemiologic characteristics in different regions of the country. This study describes the principal epidemiologic, clinical, and laboratory characteristics of children with brucellosis referred to the CMC Hospital in Tehran, Iran, between 2011 and 2016.

In the present study, the prevalence of brucellosis was 0.54%. In the earlier study conducted at this center from 2002 to 2010 [5], 34 out of 10,864 patients admitted to the infectious diseases ward (0.31%) were diagnosed as Brucella-infected cases.

In the current study, there was a male predominance among patients (60.5%), which is in agreement with the results of previous studies performed at the same hospital [5, 22] as well as other studies [1, 7, 9, 10, 15, 18, 23, 24]. However, in a study performed in Chicago, USA, most of the brucellosis cases were female [25]. Possible explanations for higher rates of brucellosis among boys could be the higher risk of direct contact with animals and consumption of unsafe food products in boys [5, 15, 26].

In this survey, brucellosis was more frequent in school-aged children (children ≥ 7 years of age) and none of the patients were younger than 1 year. This is in agreement with findings of other studies [1, 2, 18, 21, 23, 27]. However, in some studies, the infection was diagnosed in patients < 1 year of age [3, 7, 15]. The lower rate of infection in children younger than 1 year may be due to the fact that brucellosis has non-specific presentations or a milder course in this age group, which results in non-diagnosis and under-reporting of infection in infants. In addition, breast milk may have anti-brucellosis antibodies that would prevent the occurrence of infection in infants [18].

In this study, 7% of subjects had a positive family history of brucellosis. This has also been reported in other studies [1, 7, 15, 18, 22]. The positive family history highlights the necessity of examining patientsʼ family members for the presence of brucellosis.

In the current study, ingestion of raw unpasteurized milk or milk products was the main route of disease transmission (88%) and contact with suspected animals was in second place (26%), which is in accordance with the results of other studies [2, 5, 15, 18, 25, 28,29,30]. This reflects the importance of increasing awareness among both citizens and villagers about brucellosis transmission routes and the importance of avoiding consumption of unpasteurized raw milk and dairy products and contact with suspicious animals.

All patients in the present study were Iranian; none were migrants from other countries. In the Iranian study performed by Sasan et al. [7], only one child with brucellosis was an Afghan refugee and the remaining 81 patients were Iranian. Logan and colleagues from the USA [25] detected 22 cases of childhood brucellosis, with 19 cases (86%) having non-American nationalities. In addition, 18 of these cases had recently traveled or resided in Mexico, another brucellosis endemic area. This indicates that in industrialized countries such as the USA, brucellosis is most commonly seen in migrants and travelers from endemic areas [7]. However, the abovementioned result confirms that brucellosis is endemic in the Iranian population and suggests the need for implementing preventive and control measures.

In the current study, brucellosis was more pronounced in summer and spring. A similar seasonal distribution was observed in previous studies performed at the CMC Hospital [5, 22] as well as in other studies from Iran [7] and Turkey [10, 24, 31]. The possible reason for this observation could be more livestock activities such as milking and slaughtering [15], increased access to fresh goat and sheep milk in local markets [7], a higher rate of travel to rural areas with consumption of fresh non-pasteurized dairy products in those areas [22], and consumption of traditional ice cream made from unpasteurized milk in spring and summer.

Brucellosis is a systemic disease, and Brucella can invade any organ. The variable symptoms are partly attributable to the variable pathogenicity of different strains of Brucella spp. [12]. Signs and symptoms of the patients in this study were consistent with those reported by other authors [3, 5, 18, 25, 28, 31], and fever, malaise, and arthralgia were the main complaints in brucellosis-infected children. Similar to other studies [28, 32], arthralgia (77%) was more common than arthritis (51%).

Liver function tests such as ALT and AST were elevated in 23 and 35% of cases, respectively, and hepatomegaly and splenomegaly were documented in 9% and 14% of patients, respectively. In the study conducted by Logan et al. [25], much higher rates of hepatomegaly (55%) and splenomegaly (60%) were noted. Various factors including duration of illness and quality of the examination can affect the reported rate of organomegaly in different studies [5].

Nonspecific gastrointestinal symptoms such as abdominal pain, diarrhea, and vomiting were observed in some patients. Mesenteric adenitis or inflammation and ulceration of the Peyer’s patches induced by Brucella infection are the probable causes of these symptoms [28].

In the present study, anemia (65%), lymphocytosis (51%), and elevated ESR (86%) and CRP (67%) were the most prominent anomalies observed in Brucella-infected cases. Leukopenia, thrombocytosis, and pancytopenia were not detected in any case. In the previous study conducted at this center [5], anemia (53%) and leukopenia (33%) were the main hematological findings of patients and, similar to the current study, none of the patients had pancytopenia. In a study from Saudi Arabia [11] among 133 Brucella-infected children, anemia occurred in 43% of patients, followed by leukopenia (38%), leukocytosis (20%), and pancytopenia (18%). In the study conducted by Liang et al. from China [33], elevated CRP (71.1%) and ESR (63.2%) were the most frequent abnormal findings observed in 38 children with brucellosis. Several mechanisms contribute to anemia and other blood abnormalities associated with brucellosis, including hypersplenism [24, 28], bone marrow suppression, alteration of iron metabolism, autoimmune hemolysis [11, 24], hemophagocytosis, and granulomatous lesions of the bone marrow [11].

In this survey, blood culture and bone marrow culture were positive in 11 (N = 11/37, 30%) and 4 cases (4/13, 31%), respectively. In the study conducted by El-Koumi et al. [11], blood culture was positive in 23 out of 133 infected children (17%), and bone marrow culture was positive for 3 out of 9 tested children (33%). The low in vitro growth rate of Brucella spp. results in long incubation times and, therefore, a low bacterial isolation rate [13]. In addition, the diagnostic yield of blood culture differs in the acute, sub-acute, and chronic phases of the disease [4].

A positive titer in the Wright test (titers ≥ 1:160) is indicative of active brucellosis [24, 28] and the Coombs Wright test is performed to diagnose the sub-acute and chronic forms [28]. In this study, the Wright test was positive in 67% of patients. In nine cases (21%) the Wright test was negative, but the Coombs Wright was positive. In five patients (12%) the Wright test was negative, and disease was confirmed by positive Coombs Wright, 2 ME, and blood or bone marrow cultures. The sensitivity of serological tests for diagnosis of brucellosis is between 65 and 95%; however, their specificity is lower. Factors influencing the specificity of these tests include the high prevalence of anti-brucellosis antibodies in the healthy population of endemic areas (a response to their continuous exposure to the infective sources [28]) and the occurrence of antigen cross-reactivity between various species of Brucella [4]. Therefore, in endemic areas, the serologic tests should be interpreted with regard to the clinical manifestations [12] and the diagnosis of brucellosis should not be based solely on these tests.

The main purpose of brucellosis treatment is to control the acute infection and inhibit the occurrence of disease relapse. Given the fact that Brucella is an intracellular pathogen, the selected drug regimens should have the ability to penetrate into cells [24, 28]. In addition, patient age is another factor influencing the selection of therapeutic regimens. According to World Health Organization (WHO) recommendations, the use of tetracyclines (such as doxycycline) in children < 8 years of age is prohibited, since they have side effects on childrenʼs teeth [3, 24]. Moreover, the use of quinolone-containing regimens is not recommended in children < 8 years of age [9]. The therapeutic regimens used in this study were highly variable, and the most frequent regimen was the combination of trimethoprim-sulfamethoxazole and rifampicin (N = 24, 56%). The median treatment duration was 12 weeks (range 2–24 weeks). Various drug combinations and treatment durations for brucellosis have been reported in pediatric studies [3, 10, 21, 24, 25, 28]. According to the literature, it seems that combination therapy with two or three drugs active against Brucella spp. for a period of at least 6 weeks would be an effective treatment protocol for childhood brucellosis [21, 25].

In the present study, a relapse rate of 21% was observed. Relapse did not occur in patients who were treated with the following combinations: doxycycline, rifampicin, and streptomycin; trimethoprim-sulfamethoxazole, doxycycline, and streptomycin; doxycycline, rifampicin, and gentamicin; streptomycin and clarithromycin. In the previous study performed at this center [5], relapse was noted in 2.9% (1 out of 34) of patients. In the study conducted by Yoldas et al. [10] from Turkey, two patients (2%) experienced a relapse. Some of the main causes of brucellosis relapse include the occurrence of bacterial drug resistance [15], an inappropriate treatment regimen or inadequate treatment duration, a patientʼs failure to adhere to and complete the course of treatment, re-use of unpasteurized contaminated dairy products, and contact with infected livestock, especially in rural areas [7].

There are some limitations to this study. Firstly, it was conducted at the CMC Hospital, which is a tertiary children’s referral hospital in Tehran, Iran, and patients from all over Iran refer to this hospital. Therefore, the patients did not comprise a representative sample of Tehran residence. Secondly, it was a single-center study and its study population was small. Thirdly, it was a retrospective study and results were extracted from medical records of the patients; where records were incomplete, we did not have access to the patients.

Conclusion

The findings of this study demonstrate that childhood brucellosis is an endemic disease in Iran. It is mostly related to consumption of unpasteurized dairy products and is more prevalent among children ≥ 7 years of age and male gender. Therefore, physicians should be aware of its manifestations, diagnosis, and treatment protocols. In addition, implementation of control programs and preventive measures by the government are highly recommended, such as regular examination of domestic animals, mass vaccination of livestock against Brucella, slaughter of infected animals, control of animal trade and migration, pasteurization of milk and milk products, as well as training and increasing public awareness about the dangers of consumption of unpasteurized dairy products.