Infection with Campylobacter spp. most often presents as an acute, self-limited gastrointestinal disease. Extraintestinal manifestations of Campylobacter infection have been described in rare cases, and include meningitis, endocarditis, septic arthritis, osteomyelitis, and neonatal sepsis [7].
Only one case of subdural space infection caused by C. jejuni has been described before—in a child with underlying brain abnormality and VP shunt. The child presented with anorexia, lethargy and vomiting, and C. jejuni was cultivated from subdural fluid sample [15]. Other reported C. jejuni CNS infections were neonatal meningitis [5, 12,13,14]. Clinical presentations were in the majority of patients described as nonspecific, such as convulsions, somnolence and lethargy [5, 14, 15]. Fever was the sole symptom in two neonates [12, 13]. None of the neonates with C. jejuni meningitis had other risk factors. As was described in our study, infection of fluid collection in the subdural space can present as fever of unknown origin (FUO) and produce no neurological symptoms. Prolonged fever with a history of prior head trauma in adults and children should raise suspicion of intracranial infection.
In the reviewed cases, C. jejuni was isolated from the CSF in three patients and in blood culture in four [5, 12, 13, 15]. In one patient C. jejuni was detected in CSF using matrix assisted laser desorption/ionization mass spectrometry with time-of-flight detector (MALDI-TOF MS) [14]. This is the first case report of pediatric CNS Campylobacter infection being diagnosed due to use of 16sDNA sequence analysis. Molecular detection of Campylobacter spp. by 16S rDNA PCR resulted with prompt reduction of empirical antimicrobial therapy and VP shunt removal. This method is rapid and non-specific, and can be especially useful for pathogens that are biochemically inert, fastidious, neglected or slow-growing [8, 16]. Drancourt et al. proposed guidelines for including 16S rDNA sequencing as a reverence method for bacterial identification, especially for slow-growing and fastidious organisms which can remain unidentifiable after the application of all available phenotypic tests [8]. For the genus Campylobacter, 16S DNA sequencing does however lack the sufficient discriminatory power to distinguish C. jejuni and C. coli so the final species identification was performed by hippurate hydrolysis test of the achieved isolate.
World Health Organization does not recommend chloramphenicol treatment for bacterial meningitis as the first-line therapy due to its toxicity and carcinogenic potential in humans [17]. High rates of resistance make cephalosporins poor choice for C. jejuni meningitis [7]. Although C. jejuni is susceptible to gentamicin, its nephrotoxicity and ototoxicity, as well as poor CNS penetration, make it an inappropriate choice for meningitis treatment [18,19,20]. Fluoroquinolones and tetracyclines have demonstrated favorable CNS penetration in adults, however data are limited due to their potential pediatric-specific toxicities [20,21,22,23]. Imipenem and meropenem are the only carbapenems with pediatric data. As imipenem is related to higher seizure risk, meropenem is considered to be a safer option in CNS infection management [20, 24]. Good CNS penetration, low rate of adverse reactions, and good bacterial susceptibility were in favor of meropenem as the best choice in our patient [20]. There are no high-quality evidence guidelines regarding the duration of treatment of bacterial meningitis. Recommended duration for uncomplicated meningitis depends on the causative agent. Three weeks of therapy or a minimum of two weeks beyond the first sterile CSF culture is recommended for gram-negative meningitis, whatever of above is longer [25]. There are only several case reports describing two to three weeks meropenem treatment of C. jejuni meningitis in adults [26,27,28]. We decided for 4-weeks meropenem treatment. Initial treatment failure in our patient could be explained by the size of the hygroma and development of chronic subdural infection. In this setting, the neurosurgical evacuation was necessary for successful treatment.
In conclusion, we report a case of C. jejuni subdural hygroma infection in a patient with holoprosencephaly and VP shunt. Although we didn’t cultivate Campylobacter from stool samples, parents reported regular poultry consumption. Therefore, bacteremia from gastrointestinal tract probably resulted in infection of the posttraumatic subdural hygroma. Congenital brain abnormalities together with the brain injury probably made our patient more prone to C. jejuni CNS invasion. Since C. jejuni meningitis is most commonly diagnosed in neonates, this pathogen is often neglected as a causative agent for CNS infection in older children. Regardless of unspecific clinical presentation, sudden onset of symptoms along with the history of prior closed head trauma, and exclusion of VP shunt meningitis should always raise suspicion of subdural space infection. Early adequate treatment of CNS infections is crucial for satisfying recovery and prompt detection of a pathogen has a significant role in the choice of antimicrobial therapy. Therefore, rapid broad-range molecular methods represent an irreplaceable tool for the detection of rare or unexpected pathogens. Localized subdural infection may require a prolonged antibiotic therapy. In cases of subdural space infection surgical treatment should be considered. In our case, 4-weeks of meropenem treatment combined with surgical evacuation lead to a good clinical outcome.