Abstract
Brucellosis has been reported mainly among pregnant women, and it may lead to spontaneous abortion, intrauterine fetal death, or delivery of an infected neonate. Transmission through breast milk has also been described, but congenital cases are not commonly reported. We present the clinical findings, laboratory studies, treatment, and final outcome of a late prenatal transmission from a mother to her term infant of Brucella melitensis biovar 1. Because the maternal disease was undetected due to lack of clinical suspicion, diagnosis was made possible only by the results of infant blood cultures. Differential diagnosis of fever of unknown origin (FUO) misdiagnosed could result, as in our case, in the administration of inappropriate antimicrobial therapy. Primary health care physicians should be alerted to the clinical and laboratory findings of this infection, and pregnant women should routinely be tested serologically in areas where brucellosis is still a problem.
Introduction
Materials and Methods
Case
A 20-year-old second-pregnancy patient gave birth in a complicated vaginal delivery to a 40-week 2.750-g female infant who presented signs of fetal suffering, meconium aspiration syndrome, and an Apgar score of 6–7. The neonate was resuscitated with oxygen, the Apgar score increased to 7–8, and the infant was referred to the intensive care unit of another hospital. At that time the infant was cyanotic with bradycardia and hypothermic symptoms. Evaluation showed preaxial polydactylya in the left foot, negative serological diagnosis for syphilis, and a normal electrocardiogram, whereas the chest radiograph showed atelectasis affecting the inferior right lobe of the lung and cranial sonography demonstrated brain hypoxic ischemic encephalopathy grade 2. Initial management included conventional mechanical ventilation for 1 day, supplementary oxygen therapy for 6 days, and nasal cannula for 2 days. Because an infection was suspected, blood cultures were made, and the infant was treated empirically with ampicillin-gentamicin for 4 days. On the third day after intensive care unit admission, she presented convulsions and was placed on a loading dose of phenobarbital 15–20 mg/kg intravenously (IV) over 15–20 min and a maintenance dose of 3–5 mg/kg IV, intramuscularly (IM), by mouth every 12–24 h (first dose administered 12–24 h after loading) for 6 days.
On the fifth day, small Gram-negative coccobacilli that grow slowly on chocolate agar medium were isolated from the blood culture, whose corresponding antibiogram was sensitive to ampicillin and cefotaxime. Because of morphology of the colonies and patient clinical findings, it was presumptively identified as Haemophilus spp. Fresh samples of urine, blood, and cerebrospinal fluid (CSF) were taken for culture, and the antibiotics were changed to cefotaxime-amikacin for 4 days. Cultures were negative in all these samples, and 15 days after admission the infant weighing 2.870 g and presenting well-being was discharged to her home and followed up by hospital outpatient services. Thinking in terms of respiratory distress, the strain isolated had been sent to the Instituto Nacional de Enfermedades Infecciosas (INEI)-ANLIS-Clinic Bacteriology Service for identification; it was nonfermentative and showed colistin resistance and rapid urea hydrolysis, was slow-growing, and suspected via microscopy to be a Brucella sp. and therefore was referred to the INEI-ANLIS-Brucellosis Service.
Microbiological methods
The strain was identified and typed as B. melitensis biovar 1 by classical methods, including CO2 requirement, agglutination pattern with monospecific anti-A and anti-M sera, urease test, production of hydrogen sulfide (H2S), growth on dyes, erythritol, streptomycin, and penicillin sensitivity, and lysis by Tb and R/C phages, following procedures described previously (Corbel and Banai 2005).
Serological methods
Sera were assayed by buffered plate agglutination test (BPAT), Rose Bengal test (RBT), serum agglutination (SAT), 2-mercaptoethanol test (2MET), complement fixation (CFT), and competitive enzyme-linked immunosorbent assay (cELISA) following previously described methods (Lucero et al. 2007).
Results
Follow-up of the case
After these results, the infant at 1 month and 22 days of age and weighing 3.910 g, having been breast-fed only, was referred again to the hospital and evaluated at the developmental follow-up clinic. Her nutritional status was eutrophic with generalized pale skin, and she was doing well with normal neurodevelopment and physical growth. Laboratory tests showed red cells 2,420,000/mm3, white cells 6380/mm3 with 27.8% neutrophils, 58.3% lymphocytes, hemoglobin 6.8 g/dL, hematocrit 22.4%, and platelet count 151,000/mm3. Serological tests for Chagas disease, toxoplasmosis, syphilis, hepatitis B virus (HBV), and hepatitis C virus (HCV) were negative. The ultrasound image showed liver (61 mm), spleen (50 mm), right kidney (47 mm), and gallbladder to be normal, but the lobular hypoechoic left kidney (43 mm) demonstrated pathological morphology. Hospitalization was decided, 3 new blood cultures were taken, and the infant was placed on trimethoprim/sulfamethoxazole (TMP/SMX) 10 mg/k per day and rifampicin 5 mg/k per day. This combination of antimicrobial agents had demonstrated success for children under 8 years of age (Lubani et al. 1989). Four days later, B. melitensis biovar 1 was again recovered and treatment was changed to TMP/SMX 10 mg/k per day and rifampicin 20 mg/k per day; she was also given ranitidine 5 mg/k per day as gastrointestinal protection. Evolution was good, without fever; neurodevelopment was normal, and her appetite increased. She was supplemented with vitamins A, C, and D 0.6 mL, folic acid 1 mg, and ferrous sulfate 3 mg/day. Cardiovascular system and Doppler echocardiogram were normal. The infant was discharged to her home when she was 2 months and 6 days old and weighed 4.200 g, with instructions to continue the treatment. She was evaluated at 14 months at the developmental follow-up clinic and continued to be asymptomatic and clinically normal, with blood culture negative and serological tests as shown in Table 1.
After indicative case admission.
BPAT, buffered plate agglutination test; RBT, Rose Bengal test; SAT, serum agglutination test, reciprocal of titer; 2MET, 2-mercapto-ethanol test, reciprocal of titer; CFT, complement fixation test, reciprocal of titer; cELISA, complementary enzyme-linked immunosorbent assay cutoff %I>28; m, month; d, days; y, year; Pos, positive; Neg, negative; NS, not sample; A, anticomplementary serum.
Follow-up of the family
The parents were farmers breeding goats, sheep, and pigs in an area of northwestern Argentina where goat infection with Brucella spp. is common. Sera samples were taken from the infant's parents and brother for brucellosis serology (Table 1).
Discussion
Diagnosis of human brucellosis as reported worldwide in different countries is sometimes delayed because it is commonly misdiagnosed as other febrile diseases. A high index of clinical suspicion of this disease is needed in febrile patients with any history of contact with animals or their raw products. Although it is included in the differential diagnosis of fever of unknown origin (FUO), it is frequently misdiagnosed, resulting, as in our case, in the administration of inappropriate antimicrobial therapy that could potentially harm the infant's health. There were various reasons for this misdiagnosis: unusual presentation of the disease, social conditions in which the mother never seeks medical advice, and lack of serological tests. Its clinical appearance as respiratory distress and difficulties with the culture could be the reason for the strain to have been mistaken for H. influenzae. Most Brucella spp. behave as slow-growing, fastidious organisms on primary isolation. Fresh isolates have complex nutritional requirements and grow poorly on ordinary media unless they are supplemented with blood, serum, or tissue extracts (Corbel and Banai 2005).
This family lived on a farm and was involved in the care of animals; their socioeconomic conditions and standard of living were low, and the mother did not receive prenatal care during gestation in spite of a history of ingestion of nonpasteurized dairy products. The habitual use of unpasteurized foods by consumers is the main source of infection and proves that awareness of this disease in rural areas is still lacking. At INEI-ANLIS between 1994 and 2011, of the 512 strains isolated from humans, 42.2% were B. suis, 17.8% B. abortus, 1.7% B. canis, and 38.3% B. melitensis, the last coming from 17 of the 24 provinces of the country.
This infection continues to be a serious public health problem in northwestern Argentina, where people keep herds of unvaccinated goats and sheep. It is advisable to generate a high degree of awareness of brucellosis in pregnant women both in rural and urban areas, and screening programs should be set up (Elshamy and Ahmed 2008
It is advisable to generate a high degree of awareness of brucellosis in pregnant women in both rural and urban areas. In heavily infected areas, a screening program may be warranted. Public health educational efforts should be addressed to the general population concerning the source of infection, the dangers of contact with infected animals, the dangers of consumption of raw milk and milk products, clinical manifestations of brucellosis, and modes of treatment.
Footnotes
Acknowledgments
We are very grateful to Drs. Griselda Rangeón and Francisco Garcia Campos at the Department of Epidemiology, Salta, and Dr. Néstor R. Jacob at the Argerich Hospital, Buenos Aires, Argentina, for their valuable assistance with this case.
Author Disclosure Statement
No competing financial interests exist.