Natural History of and Dynamic Changes in Clinical Manifestation, Serology, and Treatment of Brucellosis, China

In China, 13.3% of human brucellosis cases progressed to chronic disease; serum agglutination test might cause treatment elongation.


B rucellosis is a zoonosis caused by the bacterium
Brucella that typically manifests in insidious onset of fever, malaise, arthralgias, and nonspecific physical findings, including hepatomegaly, splenomegaly, or lymphadenopathy (1). Accurate diagnosis and proper management of human brucellosis continues to challenge clinicians. Several studies have described the clinical characteristics of human brucellosis and evaluated diagnostic methods, but most of these studies are cross-sectional and focused on baseline manifestations or diagnostic accuracy (2)(3)(4)(5)(6). Much remains unclear about the dynamic changes of clinical manifestations, serology, and the tendency of brucellosis to persist and become chronic during development and treatment.

The Study
We conducted a retrospective, real-world cohort study at 8 hospitals in Liaoning and Xinjiang Provinces, 2 of the most brucellosis-endemic areas in China, to investigate the characteristics of brucellosis during natural history and treatment. We enrolled patients confirmed to have brucellosis during 2014-2020. We collected information on contact history, clinical manifestations, laboratory parameters, and antibiotic therapy from the hospital information system and treatment outcome by telephone (Appendix, https://wwwnc.cdc.gov/EID/ article/28/7/21-1766-App1.pdf). This research was carried out according to the principles of the Declaration of Helsinki. The study protocol was approved by the Ethics Committees of Huashan Hospital of Fudan University (KY2019-412). Informed consent was obtained from all patients before diagnosis, and patient data were anonymized.
We included 5,270 patients confirmed to have brucellosis during September 2014-December 2020. Three persons were excluded for positive HIV detection, 668 were excluded because they lacked positive culture or serologic results, and 1,191 were excluded for incomplete clinical information. We ultimately enrolled 3,411 persons; we performed follow-up for 1,676 persons at periods of 14, 28, 42, 90, 180, 360, or 720 days after diagnosis and treatment initiation (Appendix Figure 1).
We observed the natural history of brucellosis with symptom duration <180 days (early stage) or >180 days (late stage) before patients received antibiotic therapy. The 3 most common symptoms in early-stage disease were fatigue (72.3%), fever (64.0%), and sweating (34.6%). The most common symptoms in late-stage disease were fatigue (71.6%), fever (61.1%), and arthritis (36.6%) (Figure 1, panel A). Arthritis was more common in the late stage than the early stage (20.7%; p<0.0001). We observed neurobrucellosis in 9.9% of patients in the early stage and in 4.1% of patients in the late stage (p = 0.0020). After adjusting for confounding factors through propensity score-matching (PSM) (7), culture-diagnosed patients (compared with patients with SAT-diagnosed brucellosis) had higher incidence of fever ( Table 1).
Among 1,676 participants with whom we conducted follow-up, we observed further clinical characteristics after treatment initiation. Most newly developed manifestations were reported within the first 2 weeks, but most patients recovered with persistent treatment (Appendix Figure 3). Two weeks after treatment initiation, 107 patients had newly developed cardiac inflammation, 112 neurobrucellosis, 140 urogenital inflammation, and 146 arthritis. Overall, 1,453 (86.7%) persons with acute brucellosis Serum agglutination test plus exposure history were used to diagnose most cases of human brucellosis in 2 China provinces. After appropriate treatment, 13.3% of acute brucellosis cases progressed to chronic disease; arthritis was an early predictor. Seropositivity can persist after symptoms disappear, which might cause physicians to subjectively extend therapeutic regimens.
After conducting PSM for age, sex, nationality, and year of enrollment, we performed multivariate logistic regression to identify risk factors for chronic brucellosis in 148 acute cases and 148 chronic cases (Table 2). Fever, sweating, myalgia, arthritis, and C-reactive protein and erythrocyte sedimentation rates at baseline were possible predictors for chronic brucellosis in univariate analysis (p<0.10). Arthritis was the only risk factor after multivariate analysis (odds ratio 4.11 [95% CI 1.22-16.92]; p = 0.0318).
The first limitation of our study is that we failed to follow up culture results during treatment. Second, we failed to distinguish transient and persistent exposure history, which might play a role in persistent symptoms or serologic results. Finally, infection was diagnosed by heterogenous methods, including culture and a series of serologic tests, which might introduce bias in baseline and prognosis analysis.

Conclusions
Our study gives a thorough, dynamic description of clinical characteristics and serologic surveillance during the natural history and treatment of human brucellosis in a large population. Culture was 23.2% positive but SAT 99.1% positive in confirmed brucellosis. SAT plus exposure history remained the most effective diagnostic tool. Human brucellosis had variable manifestations at different disease  stages. Untreated cases mainly manifested as fatigue, fever, or sweating in the early stage, whereas fatigue, fever, and arthritis were the most common symptoms at the late stage. After appropriate treatment, 13.3% of acute brucellosis cases progressed to chronic disease. Arthritis can serve as an early predictor of chronic brucellosis. Seropositivity can persist after symptoms disappear, which might cause physicians to subjectively and unnecessarily extend therapeutic regimens. Emerging Infectious Diseases • www.cdc.gov/eid Vol. 28, No.7, July 2022 A gram-negative, facultatively anaerobic rod, Escherichia coli was named for Theodor Escherich, a German-Austrian pediatrician. Escherich isolated a variety of bacteria from infant fecal samples by using his own anaerobic culture methods and Hans Christian Gram's new staining technique. Escherich originally named the common colon bacillus Bacterium coli commune. Castellani and Chalmers proposed the name E. coli in 1919, but it was not officially recognized until 1958.
Sources: treatment initiation. This study had been registered on ClinicalTrials.gov Protocol Registration and Results System (NCT04020536).

Data Collection and Definitions
Electronic health records were collected with a medical record system designed for realworld studies provided by Huashan Hospital across 8 centers. Basic information on epidemiologic exposure history, symptoms and signs, complications and comorbidities, pathogenic diagnosis data, laboratory test data, and antibiotic therapy on the first medical visit were collected. Symptom, serology, examination, laboratory test, antibiotic therapy variation was pathogens. Early phase is defined as the natural history of brucellosis with symptom duration <180 days before patients received antibiotic therapy, while late phase is defined as duration >180 days. Acute cases are defined as patients who symptomatically recover within 6 months of treatment. Chronic brucellosis is defined as brucellosis-associated symptoms persisting over 6month treatment according to expert consensus in China (1). Recurrence is defined as the reappearance of symptoms, or new positive blood cultures after treatment completed and clinical manifestations disappeared (2).

Statistical Analysis
Continuous variables were expressed as median (interquartile range) and compared with the Mann-Whitney U test; categorical variables were expressed as number (%) and compared by χ 2 test or Fisher exact test. Symptoms, serology, and treatment variation was described with Kaplan-Meier curve and log-rank test. Propensity-score matching (PSM) was used to balance the difference between acute and chronic group, and risk factors were analyzed by logistic regression model. Two-tailed p<0.05 was considered statistically significant. Statistical analyses were performed using R software version 4.0.2. All figures were plotted by ggplot2 package.