FALSE POSITIVE DETECTION OF GROUP A STREPTOCOCCUS ANTIGEN RESULTING FROM CROSS-REACTING STREPTOCOCCUS INTERMEDIUS (STREPTOCOCCUS MILLERI GROUP)

Rapid tests for detection of Streptococcus pyogenes (group A, beta-hemolytic streptococci, GAS) directly from throat swabs are frequently used for the presumptive diagnosis of GAS pharyngitis. Most of these tests are based on detection of the Lancefield group A-specific carbohydrate. The sensitivities of these tests vary, and it is generally recommended that throat cultures be performed on patients with negative rapid tests.1, 2 In contrast the specificity of these tests is excellent, ranging from 95 to 99% in most studies.3-7 Thus a positive antigen test is considered diagnostic for GAS obviating the performance of a throat culture.1, 2 We report a child repeatedly treated with antibiotics for GAS pharyngeal infection based on detection of GAS antigen in throat swabs. These tests were falsely positive as a result of pharyngeal carriage with an alpha-hemolytic Streptococcus that cross-reacted in the antigen detection test. Case report. A previously healthy 9-year-old girl was seen by her pediatrician with a complaint of sore throat and nasal congestion without fever. Physical examination revealed pharyngitis without exudate and enlarged cervical lymph nodes. In the office a throat swab tested positive for GAS antigen (Abbott Testpack Plus® Strep A; Abbott Laboratories, Abbott Park, IL) and a 10-day course of cefadroxil was prescribed. Six weeks later the patient complained of sore throat and physical examination revealed pharyngitis. A throat swab again tested positive for GAS antigen and a 10-day course of amoxicillin was prescribed. Two days after the antibiotic course was completed, testing of a throat swab was positive for GAS antigen. A 10-day course of cephalexin was prescribed; GAS antigen was again detected at the end of this therapy. Ten days later she complained of a sore throat. Physical examination revealed rhinitis and anterior cervical adenopathy, but the pharynx was not inflamed. A throat swab tested positive for GAS antigen and a 10-day course of loracarbef was prescribed. At the end of therapy a throat swab was persistently positive for GAS antigen, and a 10-day course of clindamycin coupled with a 4-day course of rifampin was prescribed. Six weeks later she complained of sore throat and congestion. Physical examination was negative, GAS antigen was again detected in a throat swab (Abbott Testpack Plus® Strep A) and a throat culture was performed with a group A Streptococcus selective medium (BBL SSA®; Becton Dickinson Microbiology Systems, Cockeysville, MD). After 24 h of aerobic incubation at 37°C, the culture yielded normal respiratory flora and no beta-hemolytic colonies. A complete blood count was normal and an anti-streptolysin O titer and a test for antibodies to several streptococcal antigens (Streptozyme®; Wampole Laboratories, Stamford, CT) were negative. One year later her 5-year old sister developed exudative tonsillitis. A throat swab tested positive for GAS antigen (Abbott Testpack Plus® Strep A) and she was treated with a 10-day course of cephalexin. At the end of therapy a throat swab again tested positive for GAS antigen, and a throat culture was negative for beta-hemolytic colonies. Methods. The commercial GAS antigen detection kits used to test the colonies recovered from the throat culture were Abbott Testpack Plus® Strep A, Clearview Strep A® (Wampole Laboratories, Cranbury, NJ) and Strep A OIA® (Biostar, Inc., Boulder, CO). These tests were performed by swabbing the surface of the blood agar plate from the primary culture and testing the swab according to the manufacturers' instructions for testing throat swabs. Isolated colonies subcultured from the primary plate were tested by preparing a suspension of colonies swabbed from the surface of the blood agar plate and testing a swab moistened with this suspension according to the manufacturers' instructions for testing throat swabs. The alpha-hemolytic colonies were identified by the Vitek Auto Microbic System® with The Gram Positive Identification Card (bioMerieux-Vitek, Inc., Hazelwood, MO) and confirmed with the API 20Strep kit® (bioMerieux-Vitek). Serologic group determination was performed with Streptex® latex typing reagents (Murex Diagnostics, Norcross, GA). In vitro antibiotic susceptibility testing was performed by agar disk diffusion on Mueller-Hinton agar containing 5% sheep blood. Results. A swab of the mixed bacterial colonies from the blood agar plate of the throat culture after overnight incubation tested positive for GAS antigen by three GAS antigen tests designed for testing throat swabs. When individual colonies from the plate were subcultured and tested for reactivity in the GAS antigen assay, one colony morphology tested positive. After growth on trypticase soy agar with 5% sheep blood, the colonies were small, white, slightly convex and alpha-hemolytic. The bacterium was catalase-negative and reacted strongly with group A typing serum and weakly with group D typing serum. It was identified as a member of the Streptococcus milleri group, specifically Streptococcus intermedius on the bioMerieux Vitek Auto Microbic System® and as S. milleri biotype 2 using the API 20Strep® kit. By disc diffusion the isolate was susceptible to penicillin, ampicillin, oxacillin, cephalothin, imipenem, vancomycin, gentamicin, rifampin, chloramphenicol and tetracycline and resistant to clindamycin, erythromycin and clarithromycin. A swab of bacterial colonies from the blood agar plate of the patient's sister's posttreatment throat culture tested positive for GAS antigen (Abbott Testpack Plus Strep A®). Isolated colonies, identified as S. intermedius, were reactive with group A streptococcal antiserum. This isolate was susceptible to all the antibiotics listed in the preceding paragraph. Discussion. It is likely that few or possibly none of the patient's upper respiratory tract infections were caused by S. pyogenes. Factors making it relatively unlikely that GAS caused her symptoms included: (1) clinical factors: the absence of fever, soft palatal petechiae and abdominal pain and the presence of nasal congestion during some of the episodes8, 9; (2) the failure of antibiotic regimens highly effective at eradicating GAS from the pharynx, specifically clindamycin or rifampin,10, 11 to render the antigen test negative; and (3) the failure to recover GAS from a throat swab (which simultaneously tested positive for GAS antigen). The specificity of tests for GAS antigens in throat swabs is high, generally 95% or greater.3-7 For example Schwabe et al.5 using TestPack Strep A® found a negative antigen test in 258 of 265 culture-negative specimens resulting in a specificity of 97.4%. They speculated that the seven false positive detections were in fact true positives explained by nonviable GAS (such as might occur after antibiotic therapy) or the presence of GAS for which beta-hemolysis was masked. They considered cross-reactivity with other streptococci as an explanation but found no positive results when testing colonies from 40 non-group A streptococcal isolates (9 group B, 10 group C, 9 group F and 12 group G). The reason for the false positive GAS antigen detections on the last patient visit and her sister's posttreatment swab was pharyngeal colonization with a S. intermedius strain that cross-reacted strongly with Lancefield group A antisera. It is likely but unproved that the previous detections of GAS antigen were also false positive results. It has previously been noted that certain S. milleri group strains reside in the pharynx as commensals, are generally not beta-hemolytic and may express group A antigen (or group B or G antigens).12-14 However, we are unaware of reports of false positive GAS antigen tests of throat swabs resulting from carriage of S. intermedius. To our knowledge this patient is the first in whom this cross-reactivity resulted in errors in clinical management. It is of interest that the cross-reacting strain may have persisted in the patient's pharynx for at least 4 months despite multiple courses of antibiotics. To our knowledge the prevalence of pharyngeal carriage with S. intermedius expressing Lancefield group A antigen is unknown. Patients with GAS pharyngitis diagnosed by GAS antigen detection who fail to improve, who relapse or whose infection recurs should be evaluated for GAS infection by culture. Colonies on culture plates from throat swabs that do not yield GAS can be screened for false positive reactions by testing them for GAS antigen reactivity with a GAS detection kit. Acknowledgment. We thank Kathryn Ruoff, Ph.D., for confirming the identification of S. intermedius. Lorry G. Rubin, M.D.; Ronald A. Kahn, M.D.; Ernestine M. Vellozzi, Ph.D.; Henry D. Isenberg, Ph.D. Division of Infectious Diseases (LGR) and Department of Pediatrics (LGR, RAK) Schneider Children's Hospital Division of Microbiology Department of Pathology (EMV, HDI) Long Island Jewish Medical Center New Hyde Park, NY