CASE REPORT Beware of the Pet Dog: A Case of Staphylococcus intermedius Infection Russell Kempker, MD, Deepa Mangalat, MD, Thiphasone Kongphet-Tran, MS and Molly Eaton, MD Abstract: Staphylococcus intermedius is a common commensal and pathogen in dogs and cats and only rarely has been identified as causing human infection. We report a human case of postoperative sinus infection caused by methicillin-resistant S. intermedius. A 28-year-old woman with a history of endoscopic pituitary adenoma resection presented with 3 weeks of foul smelling nasal discharge. Nasal endos- copy revealed purulent sinus drainage. Cultures, initially misidentified as coagulase-negative Staphylococcus and then as Staphylococcus aureus, revealed the presence of S. intermedius. Cultures from the patient���s pet dog also grew S. intermedius strains that were confirmed to be identical to those of the patient���s by pulse field gel electrophoresis analysis. The patient was successfully treated with endoscopic debride- ment and a prolonged antibiotic regimen with vancomycin and lin- ezolid. Our case illustrates the possibility of transmission of antibiotic- resistant bacteria causing infection from pets to humans. Key Indexing Terms: Staphylococcus intermedius postoperative si- nus infection. [Am J Med Sci 2009 338(5):425���427.] Useparate ntil Hajek1 identified Staphylococcus intermedius as a species in 1976, all tube coagulase-positive staph- ylococci were reported as Staphylococcus aureus. The species was named to reflect its shared phenotypic properties with both Staphylococcus epidermidis and S. aureus. S. intermedius soon became recognized as a common constituent of the normal skin and mucosal flora of a variety of animal species, including foxes, pigeons, minks, horses, cats, and dogs.1,2 In dogs, S. intermedius was also identified as a major pathogen responsible for the common canine skin condition of pyoderma.3 The molecular understanding of S. intermedius continues to evolve. Given the observed differences in phenotypic and genotypic diversity among isolates, recent studies were performed to examine if a S. intermedius group consisting of different species existed. Both Sasaki et al4 and Bannoehr et al5 dem- onstrated through molecular phylogenetic analyses that there does seem to be a S. intermedius group composed of 4 different species: S. intermedius, Staphylococcus pseudintermedius, and Staphylococcus delphini group A and B. Although there has been a recent trend in the literature to refer to the canine pathogen as S. pseudintermedius, no official change in the current nomenclature has been made. Although it is commonly found in animals, S. interme- dius is only rarely found associated with humans either as a commensal or as a zoonotic pathogenic organism. In 1989, Talan et al6 reported the first 3 cases of human infections due to dog bite���induced cellulitis. Subsequent studies showed S. intermedius to be a relatively infrequent cause of infections from dog bites.7,8 Non���canine-induced infections are rarer and only limited reports exist in the literature. Other infections that have been described include cases of endocarditis, 9 bacteremia,10 pneumonia, otitis externa,11 mastoiditis,12 and brain abscess.13 Only 2 of the cases confirmed the transmission of S. intermedius from canine pets to infected human, 1 using an rRNA-based molecular phylogenetic technique and the other using pulse field gel electrophoresis (PFGE) analysis. We were therefore interested in reporting a case of human S. intermedius infection with docu- mented transmission of the bacteria from a pet dog. CASE REPORT A 28-year-old woman room with 3 weeks of foul smell- ing nasal discharge and a mild headache presented to the emergency. A week earlier she was evaluated by an ENT physician who performed an endoscopic nasal culture and prescribed levofloxacin. The culture was reported as coagulase- negative Staphylococcus resistant to levofloxacin, and she had no improvement in her symptoms. She was subsequently re- ferred to our hospital. Eight months earlier, the patient under- went a microscopic surgical removal of a nonfunctioning mac- roadenoma. Residual tumor was noted on follow-up imaging, and she underwent transsphenoidal endoscopic tumor removal 5 weeks before the current admission. The procedure was complicated by a cerebrospinal fluid leak for which she had a fat graft and porex plate placed. The patient also had a history of diabetes insipidus and menorrhagia-related anemia. She denied alcohol, tobacco, or illicit drug use. She lived with her fiance �� and their pet bulldog and worked as an office manager. The findings from physical examination were remark- able for normal vital signs, absence of nasal discharge, lack of meningeal signs, and intact cranial nerves II���XI. On laboratory analysis, the leukocyte count was 6000/ L with 61% seg- mented neutrophils, and the hemoglobin was 8.3 g/dL. An extended chemistry panel was normal. After a computed to- mography of the head and sinuses revealed only postsurgical changes, a lumbar puncture was performed, which showed no evidence of meningitis. A nasal endoscopy demonstrated mild purulent drainage coming from the left sphenoid os. The patient was admitted to the hospital for further workup. Cultures sent during initial endoscopy revealed heavy growth of rapid coagulase test���negative but tube coagulase- positive staphylococci that were reported as methicillin-resis- tant S. aureus (MRSA). The patient was started on vancomycin. Five days into her hospital stay, the patient underwent bilateral sphenoidotomy with the removal of infected fat grafts. Opera- tive cultures were again initially reported as heavy growth of MRSA. At the day of discharge, the microbiology laboratory amended their initial culture results, changing from MRSA to S. intermedius. The patient clinically improved and was dis- charged home with a plan for 6 weeks of vancomycin. On further questioning, the patient reported having close physical contact with her bulldog, including frequent licking of From the Division of Infectious Diseases (RK, ME) and Cardiology (DM), Emory University School of Medicine, Atlanta, Georgia and Division of Epidemiology (TK), Grady Memorial Hospital, Atlanta, Georgia. Submitted March 15, 2009 accepted in revised form May 7, 2009. Correspondence: Russell Kempker, MD, 905 Summit Pointe Way NE, Atlanta, GA 30329 (E-mail: rkempke@emory.edu). The American Journal of the Medical Sciences ��� Volume 338, Number 5, November 2009 425

her face, and that the dog had recent bouts of pyoderma requiring antibiotics. To evaluate if the S. intermedius infection may have been transmitted from her dog, we cultured 5 samples from the gingival, nasal, ear, and tail areas of the dog. Cultures from the tail and ear site were identified as S. intermedius by biochemical profiles. S. intermedius isolates from the patient���s and dog���s ear had matching antimicrobial susceptibility pat- terns (Table 1). All human and canine isolates were typed by using PFGE. (Figure 1). Performing bionumeric analysis on the PFGE patterns, the patient���s sinus and nasal isolates were determined to be the same strain as the isolate from the dog���s ear. These results confirm that the source of the patients S. intermedius was her pet dog. The patient remained on vanco- mycin, and later, she was switched to oral linezolid and made a complete recovery. DISCUSSION The transsphenoidal endoscopic approach to pituitary adenoma removal has been refined and popularized over the last decade and offers the potential advantages of superior visualization, more complete tumor resection, and shorter re- covery times. Case series have shown low rates of postopera- tive infection between 0% and 2.05%.14,15 Our patient���s infec- tion was very likely caused by the transmission of S. intermedius from close contact with her pet dog. Because her dog was inflicted with pyoderma, our patient may have been at higher risk of acquisition. Guardabassi et al16 showed that 46% of owners of dogs affected by pyoderma carried identical strains of S. intermedius in their nasal cavity. In contrast, nasal carriage of S. intermedius in veterinary staff members with daily contact with canines was found to be only 0.7%.17 In most situations, the largely clonal population structure of methicillin- resistant S. intermedius species makes for a difficult epidemio- logical investigation4,5 however, in our case, the patients close canine contact and the presence of active skin infection in her pet dog make the pet the probable source of S. intermedius. As our case illustrates, S. intermedius can often be misdiagnosed as other staphylococci species. Laboratories that use a rapid coagulase test will often misidentify the bacteria as a coagulase-negative Staphylococcus. The rapid test detects for the presence of protein A, which is absent in the cell wall of S. intermedius or clumping factor, which is present in only 14% of isolates.1 In contrast, all S. intermedius strains should have a positive tube coagulase test, which detects the presence of free coagulase. However, given that human infection with coagu- lase-positive staphylococci other than S. aureus is extremely rare, some laboratories use a positive coagulase test as criterion to identify S. aureus. Talan et al6 reanalyzed 14 isolates of coagulase-positive staphylococci from dog bite wounds and reclassified 3 of the isolates as S. intermedius. Our laboratory initially misidentified the isolate from the patient as MRSA based on a positive tube coagulase test. The isolate was iden- tified correctly as S. intermedius after analysis with the ID32 Staph system using the API database. This system uses specific biochemical reactions to differentiate between the various co- agulase-positive staphylococci and is the primary and most reliable method used in routine medical practice for this pur- pose. To clearly distinguish between S. intermedius and S. aureus, differences in biochemical reactions must be used. In a comparison of the 2 organisms, Talan et al6 found that 0% of S. intermedius and 100% of S. aureus had a positive acetoin reaction, whereas 100% of S. intermedius and 0% of S. aureus had a positive ��-galactosidase reaction. The isolate of S. intermedius from our patient exhibited high-level drug resistance with in vitro resistance to tetracy- cline, levofloxacin, cefazolin, co-trimoxazole, and oxacillin. This drug resistance profile highlights the increasing antimi- crobial resistance being reported in S. intermedius isolates, TABLE 1. Antimicrobial resistant profiles and site of isolation of selected human and dog Staphylococcus intermedius isolates Antimicrobial agent Patient sinus Dog ear Dog tail Cefazolin R R S Levofloxacin R R S Gentamycin I I S Oxacillin R R S Rifampin S S S Co-trimoxazole R R R Tetracycline R R R Vancomycin S S S Linezolid S S S R, resistant S, susceptible I, intermediate. FIGURE 1. Pulse field gel electrophoresis (PFGE) patterns of se- lected human (H) and dog (D) isolates. The human isolates and the isolate from the dog���s ear are the same strain. Kempker et al 426 Volume 338, Number 5, November 2009