A Case of Septic Arthritis from Rat-Bite Fever in Hawai‘i
Jonathan Dworkin MD; Matthew J. Bankowski PhD; Stella M. Wenceslao MD; and Royden Young MD
pp. 65-67
Abstract
Background: Infection associated with a rat bite has been known for centuries. Streptobacillus moniliformis is a zoonotic organism identified in the 20th century as the causative agent of most cases of rat bite fever outside of mainland Asia. There are no previously published cases of this pathogen in Hawai‘i.
Case Presentation: The authors present a case of Streptobacillus moniliformis causing septic polyarthritis in a 59-year-old Hawaiian man with a history of alcohol abuse and recurrent exposure to rodents in his apartment. Blood cultures from the patient were negative. The organism was isolated after three days only in thioglycolate broth from a synovial fluid culture. 16S rRNA sequencing of extracted and purified DNA confirmed the organism as Streptobacillus moniliformis.
Conclusion: Diagnosis of infection from Streptobacillus moniliformis is difficult to make because of the fastidious nature of the organism’s growth, as well as inhibitors present in standard blood culture bottles. The use of bacterial 16S rRNA sequencing may aid in an earlier diagnosis for this disease. More research is required to identify host and virulence risk factors for involvement of specific organ systems.
Background
Illness following a rat bite has been known for at least two thousand years.1 The term rat-bite fever encompasses two distinct disease syndromes, which correspond with the two causative organisms found in the oropharyngeal flora of rodents.2 In the United States Wilcox reported the first case of rat bite fever in 1829,3 but it was not until 1914 that Schottmuller described the specific pathogen that became later known as Streptobacillus moniliformis.4 This organism is responsible for the majority of rat-bite fever cases in the United States, and it can cause a variety of clinical presentations and end-organ complications.1
In addition to Streptobacillus moniliformis, a second organism known as Spirillum minus is responsible for a majority of rat-bite fever cases in Asia. This organism is also a chronic colonizer of rodents, but produces a different clinical syndrome known as sodoku. This is characterized by rash and relapsing fever, but unlike S. moniliformis there is commonly ulceration at the bite site, and there is far less arthritis than is common in streptobacillary infections.1,2 Sodoku is not the subject of this case report, but it should be considered in the differential of febrile illnesses following a rodent bite, particularly if ulceration is present at the wound site.
Case Report
A 59-year-old Hawaiian man with history of hypertension, schizophrenia, depression, and moderate-to-heavy alcohol abuse presented to the emergency room with one month of progressively worsening joint pain. The patient first noticed pain and stiffness in his right knee, but pain and swelling subsequently progressed to involve the ankles, knees, and wrists bilaterally, as well as the right elbow. The symptoms worsened slowly over weeks, and at first the patient did not seek medical attention. On further questioning he reported subjective fevers for several weeks, as well as poor appetite, diarrhea, malaise, and subjective weight loss. He also reported seeing rats around his apartment, but denied any direct contact with them. He denied rashes, mucocutaneous ulcerations, nausea, vomiting, chest pain, cough, and shortness of breath. He also denied headache, stiff neck, and alterations in consciousness.
On admission the patient had a fever of 39.1 C. His blood pressure was 138/96, heart rate was 88, and he was saturating 95% on room air, with a respiratory rate of 16. On exam the patient was frail-appearing and bed-ridden, but in no acute distress. Joint range of motion was severely limited by pain and stiffness. The left knee, right wrist, and right elbow were warm and tender along the joint lines, with moderate sized effusions. There were no rashes or skin changes over the joints. The abdominal exam was also notable for voluntary guarding and mild tenderness to palpation without rebound over the right upper quadrant. The rest of the physical exam was normal.
The white blood cell count was 12,400 with 79% neutrophils, no bands, 5% lymphocytes, 16% monocytes, and no eosinophils or basophils. The hemoglobin was 14.9, hematocrit was 42.9, and platelets were 144. Chemistry was significant for a sodium of 129, potassium 3.5, chloride 93, bicarb 27,
BUN 24, creatinine 1.1, and glucose 128. ESR was 34, uric acid was 3.5, AST was 229, ALT was 166, and alkaline phosphatase was 154. The total bilirubin was 5.7, and direct bilirubin was 2.1. Albumin was 3.2 and lipase was 31. An ultrasound of the RUQ demonstrated no stones, wall thickening, or common bile duct abnormality. There was a 0.71 cm right hepatic lobe lesion that was echogenic and consistent with a hemangioma. The hepatitis C antibody screen was positive, but the HCV RNA viral load and cryoglobulins were undetectable. The patient admitted to moderate-to-heavy alcohol use at home. Over subsequent days his LFTs normalized.
As a result of the initial findings, the clinical team made a preliminary diagnosis of polyarthritis, and they began a further work-up to determine whether the etiology was infectious or collagen vascular in nature. Peripheral blood cultures were collected on hospital day 1, 3, 8, and 12, as well as aspirates of synovial fluid from the left knee on day 3 and day 11. Multiple serologic studies in addition to the cultures were ordered. Anti-nuclear antibody was elevated at 1:160. Rheumatoid factor, complement levels, and ANCA titers were normal range. Serologic markers for dengue, typhus, leptospira, syphilis, and HIV were all normal range. An initial trial of NSAIDs and oral steroids was unsuccessful in alleviating the patient’s symptoms, and his functional status continued to deteriorate during the first week of his hospitalization.
Additional imaging consisted of a transesophogeal echocardiogram to rule out culture negative endocarditis. This study showed a “whispy” mitral valve abnormality consistent with chordae tendinae, but vegetations could not be excluded. As a result of this finding, as well as the patient’s tenuous clinical status, further work-up was initiated to exclude culture negative endocarditis. This included Bartonella and rickettsia serologies, as well as a duodenal biopsy for Whipple’s disease. During this initial phase of the work-up, the patient did not receive antibiotic therapy.
Appearance of the left knee of the patient on hospital day 5. (Figure 1)
Several days following the first synovial fluid culture, a fastidious gram negative rod was identified in the thioglycolate broth (Figure 2 and Figure 3). A gram stain of the growth in the thioglycollate broth showed “pleomorphic” gram negative bacilli, consistent with Streptobacillus moniliformis. Growth occurred only in the thioglycolate broth, not on the culture media plates, and culture identification using phenotypic methods (i.e. culture growth characteristics and biochemical testing) was attempted. Phenotyping was unsuccessful due to the fastidious nature of the microorganism. Therefore, a genotyping approach was attempted by use of bacterial sequencing of the 16S rRNA region. This test was referred to an external laboratory where the sequencing was perfomed with interpretation and identification by our laboratory. Sequencing consisted of nucleic acid extraction followed by PCR amplification and sequencing of the 16S rRNA region (approximately 1400 bp). Sequencing was performed using two primer sets with a BigDye Terminator cycle sequencing kit (version 3.1) on an ABI 3730XL DNA analyzer (Applied BioSystems, Foster City, CA). The full 16S rRNA gene sequences were then assembled by use of the Seqman program (DNAStar). Sequence analysis was performed using ChromasPro program (version 1.33; Technelysium Pty. Ltd.). A final search with the BLAST program (www.ncbi.nlm.nih.gov/BLAST) identified the bacteria as Streptobacillus moniliformis.
Bacterial sequencing revealed a base pair match of 1408/1410 (99%) for S. moniliformis compared to Sneathia sp. (1119/1198) and Leptotrichia sp. (1126/1210), and both of the latter are at a much further distance on the phylogenic scale. The same isolate was also found in the second culture of the aspirated joint fluid. However, it should be noted that all blood cultures were negative. This result was not unexpected due to the known inhibition of Streptobacillus moniliformis by the anticoagulant, sodium polyanethol sulfonate (SPS), included as an essential component of the blood culture media (Bactec, BD).
Prior to identification of the organism, the patient’s IgG titer showed a high positive titer for Bartonella henselae, the causative organism of cat scratch disease and a known cause of culture-negative endocarditis. Even though the Bartonella henselae IgM titer was negative, the patient was started on ceftriaxone, doxycycline, and gentanicin. After Streptobacillus moniliformis was identified by gene sequencing, the Bartonella titer was treated as prior exposure and not active infection. The patient’s antibiotics were then changed to penicillin, doxycycline, and gentanicin for presumed streptobacillary endocarditis. Although the patient never met formal modified Duke’s criteria for endocarditis, the echocardiographic findings, combined with the high mortality in the known case reports of streptobacillary endocarditis, prompted the team to treat for a full 6-week course. In addition the patient required orthopedic consultation, with arthroscopy and synovectomy of the left knee.
The patient’s response to antibiotic therapy was dramatic and immediate. Within 24 hours the patient reported a subjective improvement in pain symptoms and was able to demonstrate increased range of motion in all joints. Within a week the patient was making notable progress in physical therapy. Prior to discharge from the hospital the patient was ambulatory with a minimal assist device.
Discussion
Streptobacillus moniliformis is a gram negative bacillus that causes a syndrome characterized by fever, rash, and arthralgias, often following a brief incubation period of ten days.1 Complications include endocarditis,5 septic arthritis,6 meningitis,7 overwhelming sepsis,8 amnionitis,9 and abscesses in virtually any organ.10-12 Often symptoms will spontaneously resolve within two weeks. In patients with end-organ complications illness can persist longer, and some patients develop a relapsing and remitting fever of unknown origin.1 Untreated patients have a mortality rate as high as 13%, and in the pre-antibiotic era cases of endocarditis were often fatal.1
In addition to sporadic cases from rodent bites, Streptobacillus moniliformis has been associated with epidemic outbreaks of rat-bite fever. The first well-described outbreak occurred in Haverhill, MA in 1926.13 A second outbreak occurred in an English boarding school in 1987.14 Clinically the epidemic form of the disease closely resembles the sporadic form, but in the epidemic form the bacteria presumably invades through the GI mucosa following widespread ingestion of contaminated food products.15 In the English outbreak McEvoy and colleagues reported an independent association between ingestion of both water and milk, and development of subsequent disease.14
In a recent review of cases of septic arthritis caused by Streptobacillus moniliformis, Wang and Wong proposed that streptobacillary septic arthritis is a clinically distinct entity from rat-bite fever.16 Their assertion was based on a review of 12 cases, of which 10 had polyarticular involvement. All 12 patients had Streptobacillus moniliformis isolated from synovial fluid, but on
ly four had dermatologic findings, and only one had a positive blood culture. This limited experience led the authors to suggest that septic arthritis from Streptobacillus moniliformis may be the direct result of joint invasion by the infecting organisms. This would be in contrast to rat-bite fever, where patients experience more cutaneous symptoms, more bacteremia, and an autoimmune arthritis.
Given the myriad sites of involvement that have been reported in streptobacillary infections, it may not be necessary to categorize septic arthritis as a separate disease state from rat-bite fever. A simpler solution is to think of the disease as a systemic illness, with the pattern of organ involvement reflecting underlying host risk factors. Wang and Wong are likely correct in suggesting that “underlying joint abnormalities, such as Paget’s disease and osteoarthritis, increase the risk of developing localized streptobacillary septic arthritis,” though in this case report there were no such identified host risk factors. If we accept this appealing hypothesis, then an early step in recognition of the disease will be to identify host risk factors that may predispose the patient to endocarditis, soft tissue abscesses, septic arthritis, or any of the other known patterns of rat bite fever. More research will be required to firmly establish the host and virulence mechanisms that determine which pattern of disease a patient experiences after exposure to this ancient pathogen.
The clinical application of bacterial DNA sequencing techniques, such as the one described in the present paper, may also aid in earlier diagnosis and treatment of rat-bite fever. 16S rRNA is a component of the 30S ribosomal subunit found in prokaryotes. It is useful in the molecular diagnosis of bacterial infections because it contains primer binding sites that are conserved across most bacterial species, while in addition containing hypervariable regions that can be identified as the signature of a particular species. One limitation of this technology includes the need for accurate databases containing complete nucleotide sequences, which can be used as references to correctly identify the organism of interest. In addition some bacterial species contain > 99% similarity to each other, limiting the ability of the gene sequencing technology to identify a particular isolate. These problems with DNA sequencing should improve as gene libraries become more complete and accurate, and as experience with sequencing technology in routine clinical practice continues to grow.
Acknowledgements
We would like to thank Terrie Koyamatsu, Clinical Microbiology Manager and the Microbiology Departmental staff at Diagnostic Laboratory Services (The Queen’s Medical Center) for their microbiological and molecular expertise. We would also like to thank Dr. Shaobin Hou from the University of Hawai‘i at Manoa for his molecular sequencing expertise.
Supporting Organization: Diagnostic Laboratory Services – Provided diagnostic services that aided in identification of the pathogen.
Authors’ Affiliation:
- Internal Medicine Residency Program; University of Hawai‘i John A Burns School of Medicine; Honolulu, HI
Correspondence to:
Jonathan Dworkin MD
3453a Pahoa Avenue
Honolulu, HI 96816
Ph: (808) 454-3949
Email: Jdworkin@hotmail.com
References
1. Washburn R.G. Mandell, Bennett, & Dolin: Principles and Practice of Infectious Diseases. 6th ed. Philadelphia, PA: Elsevier Churchill Livingstone; 2005: Chapter 228.
2. Elliott S. Rat Bite Fever and Streptobacillus moniliformis. Clinical Micro. Reviews. 2007,20:13-22.
3. Wilcox, W. Violent symptoms from bite of rat. Am. J. Med. Sci. 1839,26:245.
4. Schottmuller, H. Zur Atiologie und Klinik der Bisskrankheit. Dermatol. 1914,58:77.
5. Rupp ME. Streptobacillus moniliformis endocarditis: Case report and review. Clin. Infect. Dis. 1992,14:769-772.
6. Anderson D, Marrie TJ. Septic arthritis due to Streptobacillus moniliformis. Arthritis Rheum. 1987,30:229-230.
7. Sens, MA, Brown EW, Wilson LR, Crocker TP. Fatal Streptobacillus moniliformis infection in a two month old infant. Am. J. Clin. Pathol. 1989,91:612-616.
8. Centers for Disease Control and Prevention. Fatal rat-bite fever – Florida and Washington. Morb. Mortal. Weekly Rep. 2005,53:1198-1202.
9. Faro S, Walker C, Pierson RL. Amnionitis with intact amniotic membranes involving Streptobacillus moniliformis. Obstet. Gynecol. 1980,55S:9S-11S.
10. Dijkmans BA, Thomeer RT, Vielvoye GJ, Lampe AS, Mattie H. Brain abscesses due to Streptobacillus moniliformis and Actinobacterium meyerii. Infection. 1984,12:34-36.
11. Vasseur E, Joly P, Nouvellon M, Laplagne A, Lauret P. Cutaneous abscess: a rare complication of Streptobacillus moniliformis infection. Br. J. Dermatol. 1993,129:95-96.
12. Pins MR, Holden JM, Yang JM, Madoff S, Ferraro MJ. Isolation of presumptive Streptobacillus moniliformis from abscesses associated with the female genital tract. Clin. Inf. Dis. 1996,22:471-476.
13. Place EH, Sutton LE. Erythema arthriticum epidemicum. Arch. Intern. Med. 1934,54:659-684.
14. McEvoy MB, Noah ND, Pilsworth R. Outbreak of fever caused by Streptobacillus moniliformis. Lancet. 1987,2:1361-1363.
15. Parker JR, Hudson NP. The etiology of Haverhill fever (erythema arthriticum epidemicum). Am. J. Pathol. 1926,2:357-379.
16. Wang T, Wong S. Streptobacillus moniliformis septic arthritis: a clinical entity distinct from rat-bite fever? BMC Infectious Diseases. 2007,7:56.
