Direct Detection of Indirect Transmission of Streptobacillus moniliformis Rat Bite Fever Infection

Abstract

We describe the evaluation of culture-negative synovial fluid from a 3-year-old boy by PCR and electrospray ionization followed by mass spectrometry (PCR/ESI-MS). Our patient developed a diffuse rash and fever with systemic signs and symptoms of sepsis, but four sets of blood cultures obtained prior to initiation of antibiotics were negative. After 1 week of illness, he developed right-knee swelling. Analysis of synovial fluid was consistent with infection, but cultures of specimens obtained following initiation of antimicrobial treatment were negative for growth. PCR/ESI-MS detected Streptobacillus moniliformis in the synovial fluid sample. Our patient completed an appropriate course of antibiotic treatment and remained completely asymptomatic in follow-up evaluation. This unique case suggests that PCR/ESI-MS may be a useful diagnostic tool for direct detection of unusual or unexpected pathogens directly from clinical specimens, particularly when samples have been obtained from patients following initiation of antibiotic therapy.

CASE REPORT

A previously healthy 3-year-old boy who was at 85th percentile in height and weight and was current on all Advisory Committee for Immunization Practices (ACIP) recommended immunizations presented for evaluation of fever and a rash. His mother reported a 3-day history of fever, irritability, poor intake, emesis, and a rash that began as a papulovesicular rash on the forearms. He became progressively irritable over the following 3 days and began to complain of left leg pain. On the day of admission, the rash had spread to the face, neck, legs, hands and feet, including palms and soles, and he was unable to tolerate walking or standing on his left leg due to pain.

Upon presentation, the patient was febrile and exhibited tachycardia. The exam was notable for several 1- to 2-mm vesicular lesions in the posterior pharynx, as well as erythematous papular 2- to 4-mm vesicles on the arms, legs, buttocks, palms, soles, ears, and cheeks. The left hip was tender to palpation and painful, with both an active and passive range of motion, but no erythema, swelling or edema was noted. Urinalysis, complete metabolic profile, and blood counts were all normal, but the C-reactive protein (CRP) was markedly elevated at 15.24 mg/dl.

After three sets of blood cultures were obtained in the emergency department (ED), the patient was hospitalized for further testing, which included urine culture, throat culture for Streptococcus pyogenes, a nasopharyngeal swab for respiratory viral cultures, and one additional blood culture (all cultures were ultimately negative), as well as Epstein-Barr virus (EBV), cytomegalovirus (CMV), and Leptospira serologies, which were also all negative.

During the patient's hospital course, his fever persisted, with daily core temperatures exceeding 39°C (102.2°F), and the erythematous papular rash migrated throughout the surface of his body (see Fig. 1). On hospital day four, he developed a right knee effusion and the CRP had increased to 19.31 mg/dl. Further questioning of the family revealed that the boy had suffered a minor laceration of his right index finger from the cage where his grandmother keeps her pet rat 1 week prior to the onset of symptoms. He had no direct contact with the rat. Ampicillin (800 mg given intravenously [i.v.] every 6 h) was started empirically for suspected rat bite fever (RBF). On hospital day eight, following two doses of i.v. ampicillin, a right-knee arthrocentesis was performed in standard fashion. Cloudy, yellow synovial fluid was obtained that was notable for a white blood cell (WBC) count of 62,415 cells/ml (91% polymorphonuclear leukocytes [PMNs]). Gram stain and cultures were all negative. The microbiology laboratory staff was notified of our concern regarding RBF, and the specimen was inoculated onto Trypticase soy agar; cultures were held for 30 days. In addition to septic arthritis, endocarditis is also a well-described complication of RBF; blood cultures were also held for 30 days.

FIG 1.

Representative maculopapular exanthema on right thigh and left thigh and buttocks, day 5.

The patient was treated with ampicillin (800 mg given i.v. every 6 h) for 6 days. He gradually responded to treatment. Rash, right-knee swelling, and left-hip pain all improved. On the morning of hospital day 10, the CRP had decreased to 2.92 mg/dl and the patient was discharged home with a plan to complete an additional 7 days of oral amoxicillin. A follow-up CRP test 1 week following completion of antibiotic treatment showed 0.20 mg/dl.

Even with common pathogens, reliance on culture for identification of bacteria in synovial fluid is problematic. In clinical practice, the sensitivity of synovial fluid culture is roughly 80%, and it is significantly diminished by prior empirical antibiotic treatment and/or the presence of organisms not detected by standard culture techniques (1). Detection of common bacterial pathogens by PCR can be useful but requires correct anticipation of expected organisms (2). PCR and electrospray ionization followed by mass spectrometry (PCR/ESI-MS) has demonstrated utility in detection of pathogenic bacteria directly from specimens obtained following initiation of antimicrobial treatment and is not predicated on prior anticipation of the most likely pathogen (3).

Institutional Review Board (IRB) approval was obtained for submission of samples of our patient's serum and right-knee synovial fluid for testing by PCR/ESI-MS, as described by Kaleta et al. (4) Compared to clinical samples, the assay performs with 98.7% and 96.6% concordance at the genus and species levels, respectively. PCR/ESI-MS identified Streptobacillus moniliformis in synovial fluid at a high level of detection (255 genome equivalents per PCR well) but did not detect an organism in peripheral blood or BacTec blood culture bottles. Our failure to isolate this organism in culture despite inoculation of appropriate media is not surprising. Even with appropriate media and without prior antibiotic exposure, S. moniliformis does not reliably grow in culture. The PCR/ESI-MS identification was confirmed by partial 16S rRNA gene sequencing using Clinical and Laboratory Standards Institute MM18-A M13-tagged bacterial 16S rRNA gene primers with a single base modification in the reverse primer and M13 tags added to the 5′ end. The primer sequences were as follows: FDA-16S-4F, 5′-M13F-TTGGAGAGTTTGATCCTGGCTC-3′; FDA-16S-801R-2, 5′-M13R-GGCGTGGACTACCAGGGTATCT-3′; M13F, 5′-CCCAGTCACGACGTTGTAAAACG-3′; M13R, 5′-AGCGGATAACAATTTCACACAGG-3′.

An NCBI BLAST query of our 708-bp partial 16S sequence against GenBank reference bacterial and Archaea 16S rRNA gene sequences exhibited 100% identity to the Streptobacillus moniliformis DSM 12112 16S rRNA complete gene sequence, positions 19 to 726 in the Escherichia coli numbering system (GenBank sequence ID NR_074449.1).

S. moniliformis, the primary agent of RBF in North America, is a pleomorphic, filamentous, Gram-negative, nonmotile bacillus. It is extremely fastidious, requiring microaerophilic conditions to grow in culture. Confirmation of suspected RBF is of critical clinical importance because sequelae from undiagnosed cases include endocarditis, pericarditis, arteritis, volvulus, septicemia, bronchitis, and pneumonia, and untreated RFB has a mortality rate of 10% (5). Our patient had knee pain, a palpable knee effusion, and persistently elevated CRP with negative cultures and serologic testing, prompting further evaluation of right-knee fluid. Although our patient had no direct contact with mice or rats, the clinical symptoms and source of the finger laceration provoked suspicion of RBF. Yet antibiotic pretreatment may have been responsible for negative cultures, so we were concerned about common bacterial pathogens as well. Uncommon pathogens detected by PCR/ESI-MS in synovial fluid have been reported, but this is the first report of disseminated RBF detected by PCR/ESI-MS (6).

Clearly, the approach to culture-negative infections and our understanding of syndromes caused by pathogens that cause disseminated disease but are not readily diagnosed by conventional microbiologic methods require further study. Nonetheless, our experience in this case suggests that molecular methods capable of timely identification of unusual or unanticipated bacteria directly from synovial fluid may provide an opportunity to detect virulent pathogens and direct antibiotic treatment and that PCR/ESI-MS may have a role as an adjunct to conventional diagnostic microbiologic methods for culture-negative synovial fluid specimens.