In this study, we report the simultaneous detection of B. vinsonii subsp. berkhoffii infection in two family members who were experiencing neurological dysfunction. Bartonella vinsonii subsp. berkhoffii is an important emerging intravascular pathogen that has been isolated from patients with endocarditis, arthritis, neurological disease and vasoproliferative neoplasia [5, 6, 14, 15]. In the current case report, both the father and daughter were infected with B. vinsonii subsp. berkhoffii genotype II strains and both were either co-infected or sequentially infected with B. henselae. Considering the sequential serological test results obtained for the father, it seems likely that he was co-infected, when the first blood culture was obtained. Initial B. henselae antibody titers for the father were 8 to 12 fold higher than the titers obtained using B. vinsonii subsp. berkhoffii genotype II antigens, whereas only a 6 fold difference was found in the initial two serum samples from the daughter. As there is a less convincing serological association supporting co-infection in the daughter and as three of four cats in the household were B. henselae bacteremic during the period in which the father and daughter were being treated for their B. vinsonii subsp. berkhoffii infections, it is possible that sequential infection with two different Bartonella spp. was documented in the daughter during the course of this study. As the isolation and molecular detection of these bacteria from patient samples remains microbiologically challenging, we were unable to clearly establish whether either patient was co-infected or sequentially infected at various time points.
Previously, we described the preferential amplification of one Bartonella sp. when two or more species are present in the extracted sample . The mechanism(s) responsible for preferential amplification of one bacteria when DNA of two species is present in a patient sample is unclear, but mechanisms could include the relative concentrations of DNA of the respective organisms in the sample at the time of DNA extraction for PCR amplification, or selective amplification of one DNA sequence over the other one, when comparable template concentrations are present. Targeting multiple Bartonella genes can enhance the possibility of detecting co-infection in patient samples [17–19]. Potentially, PCR primers targeting different gene fragments preferentially amplify different Bartonella spp. in co-infected individuals. It is also possible that BAPGM enrichment culture preferentially selects for the growth of one Bartonella spp. in a co-infected individual. Therefore both individuals may have been co-infected with B. vinsonii subsp. berkhoffii and B. henselae at the outset of this study.
As the sequences of the B. henselae strains obtained from the father and daughter's blood cultures were identical (B. henselae ITS strain Houston1) to the blood culture strains obtained from the three cats, the family cats were the presumed source of this infection.
Although the mother and two sons were healthy and blood culture negative when tested on a single occasion, all three had serological evidence supporting prior exposure to B. henselae. The two newly acquired young dogs were not seroreactive to Bartonella antigens and both were blood culture negative; therefore it seems unlikely that these dogs played a role in transmission of B. henselae or B. vinsonii subsp. berkhoffii in the family. Although cats are the primary reservoir host, B. henselae has been isolated by blood culture from dogs and sequenced from dog saliva [8, 9]. It is possible that an arthropods or the two older dogs that had died prior to initiation of this study were the source of B. vinsonii subsp. berkhoffii infection, as historically this organism has been isolated from domestic and wild canines and humans [1, 2]. Infection with B. vinsonii subsp. berkhoffii was recently described in a cat with recurrent osteomyelitis that was bacteremic over a 15-month time period . Therefore, cats may be able to maintain a persistent B. vinsonii subsp. berkhoffii bacteremia and potentially serve as a source of bacterial transmission to humans . Efforts to amplify B. vinsonii subsp. berkhoffii from the cats in this study using subspecies-specific primers were not successful. Over 100 years ago, B. quintana was transmitted to human volunteers, when saliva from a febrile patient was applied to escharified skin . As saliva obtained from soldiers with trench fever apparently contains viable, infectious Bartonella quintana, oral transmission of B. vinsonii subsp. berkhoffii from the father or a pet cat to the daughter through close family contact cannot be ruled out. Bartonella spp. DNA has now been reported in the saliva of cats, dogs and humans [22, 23]. Clearly, additional data is needed to define the risk factors for Bartonella spp. transmission to humans and to their pets, but it seems prudent to recommend hygienic measures after contacting pet and perhaps human saliva.
The father in this report is the second patient in which enrichment culture enhanced the molecular detection of a Bartonella spp. in cerebrospinal fluid. In a previous study, B. henselae was repeatedly detected by blood or cerebrospinal fluid culture in a 23 year-old girl who developed progressively severe seizures following a history of cat scratch disease . In both patients, cerebrospinal fluid analyses were reported to be within normal limits; however, inadvertent contamination of the sample with blood cells cannot be ruled out. As Bartonella spp. appear to target vascular endothelial cells, it is possible that B. vinsonii subsp. berkhoffii and B. henselae contributed to the nonspecific area of vascular injury reported on the father's MRI.
The use of an optimized insect-based cell culture growth medium can facilitate the isolation or enhanced molecular detection of Bartonella spp. following culture-enrichment of patient samples prior to performing PCR [5, 6, 8, 14, 18]. The enhanced diagnostic utility of the enrichment approach is best illustrated in Table 2, where B. vinsonii subsp. berkhoffii and B. henselae DNA were never directly amplified from extracted blood samples obtained from the daughter and were only detectable by PCR following enrichment culture. In our experience, at least a 7 to 14-day incubation period is required before the enriched sample is obtained from liquid BAPGM for DNA extraction. At no time during this study was Bartonella spp. DNA amplified from a DNA extraction control, a BAPGM un-inoculated enrichment culture control or a subculture agar plate control. Although not detailed in the table, it is not unusual for 7-day enrichment cultures and subcultures to be PCR negative for Bartonella spp. DNA, whereas the respective organism can be amplified and sequenced after a 14-day incubation period from the enriched liquid culture, the agar plate isolate, or both. Due to the high level of B. henselae bacteremia generally found in cats, the same strain could be detected in each cat following direct extraction of the blood sample, enrichment culture and the agar plate isolates (Table 3). Unfortunately, despite the enhanced utility of enrichment culture for the molecular microbiological diagnosis of Bartonella infection, obtaining viable agar plate isolates after subculture from liquid BAPGM at 7 or 14 days post-incubation remains technically difficult. In this study, only one B. vinsonii subsp. berkhoffii isolate was obtained from the daughter's post-antibiotic blood culture after a 14-day enrichment period, whereas the 7-day BAPGM enrichment culture and subculture were both PCR negative. An isolate was never obtained from the father; however as described above, using the same BAPGM enrichment platform, B. henselae isolates were obtained from three of the four cats in the household. As overtly healthy cats can maintain a high-level of bacteria in systemic circulation for months to years, isolation of B. henselae form cat blood samples is comparatively easy to achieve, as compared to isolation using the same approaches from dog or human blood samples [11, 12]. Failure to obtain stable Bartonella isolates is a major patient management limitation, as it prevents routine testing for antibiotic sensitivity and resistance of specific isolates at time points prior to and following antibiotic administration. This was of particular concern for these two patients, as B. henselae DNA was still sequenced from an enrichment culture of the father's blood following a 3-week course of doxycycline and rifampin and B. vinsonii subsp. berkhoffii was detected in the girl's blood following a 6-week course of azithromycin. Based upon negative post-antibiotic blood cultures and a decremental decrease in antibody titers to non-detectable levels, antibiotic treatment appeared to correlate with microbiological elimination of B. vinsonii subsp. berkhoffii and B. henselae infections, cessation of antibody production, and with the eventual clinical resolution of illness in both patients.
The father reported a recent history of severe periodontal disease. Two previous molecular microbiological studies identified Bartonella spp. DNA in subgingival samples from patients with periodontitis. [24, 25]. In addition, other investigators have detected B. henselae DNA and B. quintana DNA in the parotid salivary glands of an immunocompetent woman and man, respectively and B. quintana in the dental pulp of a homeless man [26–28]. After which B. vinsonii subsp. berkhoffii genotype II was successfully sequenced from the periodontal swab, but attempts to sequence the amplicon obtained from the salivary swab were not successful (Table 1). As identical techniques were used, this result might be explained by a higher concentration of B. vinsonii subsp. berkhoffii DNA at the periodontal surface, as compared to dilution of bacterial DNA targets floating in saliva in the oral cavity. In our laboratory, a B. henselae SA2 strain (2.5 copies/reaction) is used as the source of positive control DNA for the ITS PCR reaction; therefore, contamination with positive control DNA could not explain any PCR results obtained in these two patients. As over 500 species of bacteria have been estimated to inhabit the oral cavity, BAPGM enrichment culture was not attempted because rapidly growing organisms would negate efforts to increase Bartonella numbers via the enrichment process [29, 30]. It is also possible that detection of B. vinsonii subsp. berkhoffii in close approximation of the periodontal surface reflects passive leakage of bacteria through inflamed and compromised vascular tissues or alternatively the establishment of an active foci of infection that contributed to the recent history of periodontitis. Detection of DNA in the oral cavity does not confirm the presence of viable bacteria; however, caution should be exercised by dentists and physicians when examining the oral cavity of an individual with chronic Bartonella spp. bacteremia.