Streptococcal serology, the measurement of antibodies produced against the bacterium Streptococcus pyogenes [group A streptococcus (GAS)], is a useful diagnostic tool when used in the right circumstance. Although GAS causes a broad range of human disease, including pharyngitis, impetigo, scarlet fever and invasive infections, it is only the poststreptococcal sequelae that necessitate streptococcal serology for their diagnosis. The 2 most important postinfectious diseases are acute rheumatic fever (ARF), which leads to rheumatic heart disease in at least half of cases, and poststreptococcal glomerulonephritis (PSGN).1 Streptococcal serology helps in the diagnosis of these 2 conditions because they occur 2–3 weeks after the acute infection that triggers them, when culture of the causative bacterium is usually no longer possible. In contrast, streptococcal serology at the onset of acute infection is usually negative; at this time, culture and rapid antigen detection tests are more useful for diagnosis.
TESTS FOR STREPTOCOCCAL SEROLOGY
During an infection with GAS, there is a broad immune response to multiple extracellular and cellular antigens. However, the antibody response to streptolysin O and/or deoxyribonuclease B is used most in clinical practice. Although antibody responses to other streptococcal antigens such as hyaluronidase and streptokinase have been studied, there are insufficient data to support their use in routine clinical practice.2
Antistreptolysin O Assays
Assays to measure antistreptolysin O (ASO) antibodies are the most standardized and widely used today. Streptolysin O, a secreted toxin, is responsible for the zone of β-hemolysis around colonies of GAS seen on blood agar plates caused by lysis of erythrocyte membranes. The original method described for quantifying ASO antibodies is a dilutional assay that measures the ability of decreasing concentrations of human serum to block the hemolytic activity of streptolysin O. The ASO titer of a serum sample is equal to the inverse of the highest serum dilution that shows no hemolysis, that is, the serum dilution that inhibits the action of streptolysin O. An international standard for the streptolysin O reagent used in this assay was established by the World Health Organization in 1961, so that ASO values are reported as international units per milliliter.3 However, for most diagnostic laboratories today, the classical assay method is time consuming and has been superseded initially by latex agglutination kits and increasingly by nephelometric or turbidimetric assays. Results for these assays are also reported in international units per milliliter, with reference values largely based on the classical method.
There are a number of important caveats to the use of ASO. Stronger ASO responses tend to occur with throat infection than with skin infection, possibly because free cholesterol present in the skin binds to streptolysin O thereby decreasing its immunogenicity. False-positive results can occur with infection by nongroup A streptococci, most notably Streptococcus dysgalactiae subspecies equisimilis (Lancefield group C and G streptococci), because these streptococci also produce streptolysin O.2 False-positive results can also occur if serum is contaminated or old because lipoproteins present in serum can inhibit streptolysin O. Further, false-positive results may occur because of cross-reactivity in patients with myeloma, hypergammaglobulinemia, liver disease and autoimmune disease with increased rheumatoid factor.2 Titers may also remain increased in the setting of GAS carriage.4 False negatives may result from differential expression of streptolysin O in different GAS strains or because of host factors such as hyperlipidemia. Moreover, clinical studies suggest that approximately 20% of patients with bona fide streptococcal pharyngitis do not mount an ASO response.4
Antideoxyribonuclease B Assays
Assays for antideoxyribonuclease B (ADB) antibodies provide additional information to the ASO test.2,4 Deoxyribonuclease B, also called streptodornase or DNAse B, is a secreted enzyme that degrades extracellular DNA.5 It plays an important role in skin and soft tissue infections by allowing GAS to spread in the extracellular matrix. There are many variants of this toxin, most encoded by mobile genetic elements (bacteriophages), and so the presence of different DNAses varies between strains. Although not present in all strains, the isoenzyme B is the most widely distributed among GAS. Importantly, DNAse B seems to be more specific to GAS compared with group C and G streptococci. The classic ADB assay measures the ability of serum to neutralize the activity of deoxyribonuclease B in digesting a DNA substrate.6 The DNA substrate is complexed to a colored dye, and the ADB titer is equal to the inverse of the highest serum dilution that blocks the loss of color expressed by the digested substrate. There is no international reference serum standard for ADB. Both throat and skin infections lead to marked ADB responses, and so the ADB test may be particularly useful for PSGN in settings where the triggering infection is usually impetigo, such as in high incidence tropical settings.2,7 False-negative ADB results can occur because of infection with a strain that does not contain the DNAse B gene or expresses the toxin at a very low level.
TIMING OF TESTING
Classical descriptions suggest that the ASO titer increases a week after infection, peaks after 3–5 weeks, begins to decrease at 8 weeks and returns to preinfection levels at around 8 months.8 In contrast, the ADB titer peaks at 6–8 weeks, begins to decrease at 12 weeks and returns to preinfection levels by 12 months.9 However, a study of 160 children over a 2-year period found that this classic response does not always occur.4 In this study, among the 36 children who acquired pharyngeal GAS, 68% and 56% of the children with a documented increase in ASO and ADB, respectively, maintained titers above preinfection levels for >1 year.4
SINGLE VERSUS REPEATED MEASUREMENT
ASO and ADB titers may be interpreted either by comparison of acute with convalescent titers or against a reference upper limit of normal (ULN). An increase in titer from acute to convalescent (at least 2 weeks and preferably 4 weeks apart) is considered the best evidence of antecedent GAS infection.4,7 Documenting an increase in titer is preferred for a number of reasons. First, an increasing titer may peak below the ULN value. In the study of 160 children mentioned earlier, there were 54 and 51 occasions, respectively, when a significant increase in ASO and ADB was observed. Of these, the peak was below the recommended ULN in 59% and 61% cases, respectively. Second, titers may remain increased for many months without demonstrable infection, including in the pharyngeal carrier state.4
However, because of the kinetics of ASO (as outlined earlier) and the timing of presentation of patients with ARF and PSGN (ie, 2–3 weeks after the causative infection), the ASO may reach its peak at the time of presentation and may not increase any further when the convalescent sample is tested. In this case, this may mistakenly be interpreted as negative serology. Therefore, as described later, the ULN for streptococcal serology can be useful because an increased ASO titer is frequently observed at presentation. Furthermore, the increase in titer that is clinically significant has not been clearly defined, especially for ADB. For the classic hemolysis neutralization ASO assay, the World Health Organization suggest an increase of ≥0.2 log10 between acute and convalescent-phase sera assayed in parallel should be considered significant (eg, log10 100 to log10 400).6,10 Application of this change in titer to newer assay methods for ASO or to any assay for ADB has not been well studied. However, a 4-fold increase is generally accepted as being significant in clinical practice (eg, 100–400 IU). There are no comparable data for ADB, and expert opinion varies.2
The ULN for streptococcal serology is generally defined by the 80th percentile. In a study in Australia, diagnosis of poststreptococcal syndromes using an ULN value for ASO had a sensitivity of 72.7% and specificity of 97.5%.11 This suggests that >20% of bona fide cases would be missed by using an ULN value in isolation. However, in this study, the addition of ADB to ASO increased the sensitivity to >95%.11 These data confirm older studies that observed that >85% of patients presenting with ARF have an ASO and/or ADB titer above the ULN value.9
INFLUENCE OF AGE AND GEOGRAPHY
Average and ULN values for both ASO and ADB vary with age, reflecting age-specific incidence of GAS disease. Many package inserts in commercially available kits use a single cutoff derived from adult data that are inappropriate for use in children. When possible, age-stratified and population-specific ULN values should be used for interpretation.
Streptococcal impetigo is very common in many tropical settings, and, as a result, average streptococcal antibody titers are generally higher across all age groups. However, data from the Pacific demonstrate that if children with recent streptococcal infections are meticulously excluded from analysis of normative data, ULN values are similar in temperate and tropical settings (Fig. 1).7 In temperate settings, where streptococcal impetigo is much less common, both ASO and ADB titers primarily reflect the incidence of streptococcal pharyngitis with a clear peak in children aged 5–15 years (Fig. 1).12
INFLUENCE OF BACKGROUND DISEASE BURDEN
In patients presenting with clinical features consistent with ARF or PSGN in endemic settings, where the pretest probability of these diseases is high, we recommend interpreting ASO and ADB titers against age-stratified ULN values. An initial ASO titer above the ULN for age is sufficient for diagnosis if obtaining a second convalescent sample is not possible. If the initial ASO titer is below the ULN for age, every effort should be made to (1) measure ADB if possible and (2) obtain a convalescent sample for testing of ASO (and ADB if possible). Where no other data are available, particularly in the Asia-Pacific region, age-stratified ULN values derived in Fiji can be applied (Fig. 1).7
Indiscriminate use of streptococcal serology may be misleading in industrialized settings where ARF and PSGN are uncommon, particularly when used as a screen for rheumatic disease.2,10 In these settings, where the pretest probability of a poststreptococcal syndrome is low, a normal initial ASO titer is reassuring, and if both ASO and ADB are negative, then a poststreptococcal syndrome is highly unlikely. If the initial ASO is increased above the ULN, then a convalescent sample should be obtained to establish whether there is an increase rise in titer.
DIFFICULTY OF INTERPRETATION IN OTHER CONDITIONS
The role of GAS is well established in the pathogenesis of ARF and PSGN. Although streptococcal serology is routinely used in the diagnosis of poststreptococcal reactive arthritis there are fewer data to guide interpretation for this condition. Beyond these conditions, we recommend considerable caution when interpreting streptococcal serology because there are a number of other syndromes for which GAS is purported to play a role on the basis of streptococcal serology but for which there is no definitive proven causal link. These include anaphylactoid purpura, Kawasaki disease, pediatric autoimmune neuropsychiatric disorder associated with streptococcal infection, acute demyelinating encephalomyelitis, cutaneous polyarteritis nodosa and others.
ASO and ADB are important tests for the diagnosis of poststreptococcal sequelae. However, the clinician must be aware of the considerations outlined earlier when ordering and interpreting these tests. Additional clinical studies of ASO and ADB across the broad range of GAS disease in a variety of settings are needed, along with the development of new tests taking advantage of increasing knowledge of GAS genomes, virulence pathways and host immune response.
1. Carapetis JR, Steer AC, Mulholland EK, et al. The global burden of group A streptococcal diseases. Lancet Infect Dis. 2005;5:685–694
2. Parks T, Smeesters PR, Curtis N, et al. ASO titer or not? When to use streptococcal serology: a guide for clinicians. Eur J Clin Microbiol Infect Dis. 2015;34:845–849
3. Spaun J, Bentzon MW, Larsen SO, et al. International standard for antistreptolysin-O. Bull World Health Organ. 1961;24:271–279
4. Johnson DR, Kurlan R, Leckman J, et al. The human immune response to streptococcal extracellular antigens: clinical, diagnostic, and potential pathogenetic implications. Clin Infect Dis. 2010;50:481–490
5. Walker MJ, Barnett TC, McArthur JD, et al. Disease manifestations and pathogenic mechanisms of group a Streptococcus. Clin Microbiol Rev. 2014;27:264–301
6. Johnson DR, Kaplan EL, Sramek J, et al. Laboratory Diagnosis of Group A Streptococcal Infections. 1996 Geneva World Health Organization
7. Steer AC, Vidmar S, Ritika R, et al. Normal ranges of streptococcal antibody titers are similar whether streptococci are endemic to the setting or not. Clin Vaccine Immunol. 2009;16:172–175
8. Kaplan EL, Ferrieri P, Wannamaker LW. Comparison of the antibody response to streptococcal cellular and extracellular antigens in acute pharyngitis. J Pediatr. 1974;84:21–28
9. Ayoub EM, Wannamaker LW. Evaluation of the streptococcal deoxyribonuclease B and diphosphopyridine nucleotide antibody tests in acute rheumatic fever and acute glomerulonephritis. Pediatrics. 1962;29:527–38
10. Sen ES, Ramanan AV. How to use antistreptolysin O titre. Arch Dis Child Educ Pract Ed. 2014:99
11. Blyth CC, Robertson PW. Anti-streptococcal antibodies in the diagnosis of acute and post-streptococcal disease: streptokinase versus streptolysin O and deoxyribonuclease B. Pathology. 2006;38:152–156
12. Kaplan EL, Rothermel CD, Johnson DR. Antistreptolysin O and anti-deoxyribonuclease B titers: normal values for children ages 2 to 12 in the United States. Pediatrics. 1998;101(1 pt 1):86–88