ORIGINAL ARTICLE Fewer Symptoms Occur in Same-Serotype Recurrent Streptococcal Tonsillopharyngitis

Lucia H. Lee, MD; Elia Ayoub, MD; Michael E. Pichichero, MD

Background: Most patients with acute rheumatic tial GABHS infection and was associated with a 0.2- or fever report no antecedent pharyngitis. greater log rise in either antistreptolysin O or anti– deoxyribonuclease B titer, indicating bona fide recur- Objective: To determine the clinical and microbio- rent infection; these recurrences all occurred within 55 logical characteristics of recurrent group A ␤-hemolytic days. Fewer patients with recurrent GABHS pharyngitis streptococcal (GABHS) tonsillopharyngitis. of the same serotype had headache (P=.02), sore throat (P=.006), fever (P=.008), pharyngeal erythema (PϽ.001), Design: Prospective randomized trial. pharyngeal edema (PϽ.001), pharyngeal exudate (P=.04), and adenitis (P=.03) compared with the initial episode. Subjects: Symptoms were recorded and throat cul- Chills, stomachache, scarlatina, tonsillar enlargement, and tures were obtained at 4 to 6, 18 to 21, and 32 to 35 days palatal petechiae were similar for both episodes. following the start of treatment. A subset of 60 patients with subsequent GABHS episodes occurring were evalu- Conclusions: Fewer symptoms occur during recurrent ated for a 0.2-or greater log rise in either antistreptoly- GABHS pharyngitis of the same serotype compared with sin O or anti–deoxyribonuclease B titer to confirm a bona the initial infection. These patients may be less likely to fide recurrence. seek physician attention, yet their infections put them at risk for sequelae. Results: Sixteen (27%) of 60 patients had recurrent GABHS tonsillopharyngitis of the same serotype that occurred 21 days or longer following the onset of the ini- Arch Otolaryngol Head Neck Surg. 2000;126:1359-1362

ROUP A ␤-hemolyticstrep- study period; 19 patients (6%) had the tococcal (GABHS) phar- same streptococcal serotype isolated yngitis accounts for sig- during the recurrences (Table 1). nificant childhood mor- These recurrences were detected at 21 bidity, but little is known to 34 days and 35 days or longer follow- aboutG the morbidity associated with recur- ing the initial infection in 74% and 26% rent GABHS pharyngitis. An increase in the of cases, respectively (mean, 30 days; cases of acute rheumatic fever (ARF) in chil- range, 21-55 days). Children with same- drenwasnotedinthelate1980sintheUnited serotype recurrences had significantly States.1-10 During this resurgence, many pa- fewer symptoms (headache, sore throat, tients with documented ARF had no recol- or fever) and signs (pharyngeal ery- lection of an episode of pharyngitis in the thema, pharyngeal edema, pharyngeal preceding months.3-5,7,11 During routine exudate, or adenitis) compared with the follow-up of patients in a comparative an- initial infections (Table 2). The fre- tibiotic study protocol, we serendipitously quency of chills, stomachache, scarla- notedthatsomepatientsreturnedwithfewer tina, tonsillar enlargement, and palatal and/or milder symptoms and signs of ton- petechiae were similar on both occa- From the Food and Drug sillopharyngitis, and had GABHS-positive sions; however, our statistical power to Administration, Rockville, Md throat cultures. We sought to determine the detect differences was limited by sample (Dr Lee); the Department of clinical characteristics and significance of size. Pediatrics, University of these recurrences. Ten patients (3%) had new acquisi- Florida, Gainesville (Dr Ayoub); and the tions of GABHS infection, which oc- Department of Microbiology RESULTS curred at a mean of 20 days (range, 11-33 and Immunology, University days) following the initial infection. These of Rochester Medical Center, Sixty (20%) of 295 children had a sub- patients did not show statistically signifi- Rochester, NY (Dr Pichichero). sequent GABHS pharyngitis during the cant differences for any sign or symptom

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, NOV 2000 WWW.ARCHOTO.COM 1359

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 SUBJECTS AND METHODS hours, and then examined for ␤-hemolysis and colony mor- phology. All available strains of GABHS were typed by the STUDY DESIGN capillary precipitation method for M protein and the ag- glutination technique for T protein according to standard methods.13,14 Antistreptolysin O (ASO) and anti– Of the studies undertaken at the Elmwood Pediatric Group deoxyribonuclease B (anti–DNase B) titers were deter- in Rochester, NY, one comparative antibiotic trial provided the mined simultaneously on acute and on convalescent se- necessary design to assess same-serotype GABHS recurrences rum samples, collected at a mean of 19 days later (range, occurring shortly after documented GABHS tonsillopharyn- 9-29 days) for the initial and repeated episodes. Antibody gitis. This study prospectively compared the efficacy of cepha- titer assays were performed using previously described meth- lexin with penicillin treatment and was conducted between ods.15 A rise in either an ASO titer or an anti–DNase B titer June 1, 1981, and June 1, 1984.12 The protocol was approved of 0.2 log or greater was considered significant.16-18 Previ- by the local institutional review boards and informed consent ous studies showed that the peak of the antibody response wasobtained.Twohundredninety-fivepreviouslyhealthychil- to these streptococcal antigens occurs 2 to 4 weeks after dren aged 4 to 17 years were enrolled based on clinical symp- the acute pharyngitis and is followed by persistence of the toms of acute pharyngitis and a GABHS-positive throat cul- peak titer at the same level.19-21 Because an elevated anti- ture. Patients were excluded if they had 2 or more sore throats body level from serum obtained within this period could in the preceding 6 months, if they were treated with antibi- represent persistence of an elevated antibody from the pre- otics in the previous 2 weeks, or if they had a sibling concur- vious episode, titers obtained on serum samples during sub- rently enrolled in the study. The aspects of the original study sequent episodes of pharyngitis were considered signifi- design relevant to the purpose of this study included a record cantly increased if their value exceeded the peak titer of of all signs and symptoms on case report forms, serotyping the initial serum by 0.2 log or greater. of all available GABHS isolates, and confirmation of bona fide Drug compliance during the initial episode was con- initial and recurrent infection by analysis of acute and conva- firmed if a throat culture yielded no growth of GABHS dur- lescent serum samples for streptococcal antibodies. ing treatment. If adequate compliance was demonstrated and a 10-day antibiotic regimen was completed as deter- DATA COLLECTION mined by a dosing diary and return of empty antibiotic bottles, patients with a throat culture yielding GABHS dur- Patients were evaluated by their regular pediatrician at 4 to 6, ing a subsequent episode were evaluated for a bona fide 18 to 21, and 32 to 35 days following the start of treatment. GABHS recurrence using additional criteria. A bona fide Fever was defined by a temperature of 38°C rectally, 37.5°C recurrence included GABHS tonsillopharyngitis of the same orally, or 37°C axillary. Symptoms of sore throat, headache, serotype that occurred 20 days or longer following the on- chills,andstomachacheandsignsofpharyngealerythema,pha- set of the initial GABHS infection and was associated with ryngeal edema, pharyngeal exudate, scarlatina, and palatal pe- a 0.2-log or greater rise in either an ASO titer or an anti– techiae were recorded as present or absent. Cervical adenitis DNase B titer.20,22 Carriers were defined as patients with the and tonsillar size were recorded on a semiquantitative basis same serotype isolated during the initial visit and subse- usinga0to4+scale with 0 indicating absent; 1+, mild; 2+, mod- quent follow-up evaluation, but with less than a 0.2-log rise erate; 3+, severe; and 4+, extremely severe adenitis. Absent ton- in either an ASO titer or an anti–DNase B titer. Although sils, small, average, moderately enlarged, and severely enlarged antibiotics may suppress a rise in antibody titer,20 the 0.2- tonsillar size were assigned a score of 0, 1+, 2+, 3+, or 4+, re- log or greater rise in either an ASO titer or an anti–DNase spectively. For our analysis, tonsillar size was classified as pres- B titer was deemed a necessary criterion to distinguish same- ent if greater than 2+, and absent if 2+ or less. A throat culture serotype recurrences from carriers. Cases in which the iso- and blood sample for antibody determination were obtained late from the recurrent GABHS infection differed from the at each evaluation and sent to the laboratory of the late Hugh initial infection were considered new acquisitions. Since a Dillon, MD, in Birmingham, Ala, for analysis. Data on signs shift to a different serotype within a few weeks was highly and symptoms of GABHS were collected from multiple cen- unlikely to occur in carriers, serotype alone was believed ters on a standardized case report form. The dichotomous rat- to be sufficient to distinguish new acquisitions of GABHS ing scale we used to evaluate signs and symptoms reduced the from carriers. influence of subjective assessments by different observers. STATISTICAL ANALYSIS LABORATORY ANALYSIS The McNemar test was used for analysis of significant dif- All throat cultures obtained were immediately inoculated ferences in symptoms and signs of initial and recurrent on 5% sheep blood agar plates, incubated at 37°C for 24 GABHS infections.

compared with the initial episode (Table 2). Six pa- COMMENT tients (2%) with recurrent GABHS infections had an an- tistreptococcal antibody rise less than 0.2 log, and thus During the recent resurgence of ARF in the United States, were categorized as carriers. Twenty-eight patients (9%) a low frequency of antecedent pharyngitis was noted in had incomplete clinical or serotyping data available; the most cases.3,5,7,8 The number of children reporting phar- absence of these data precluded classification of these epi- yngitis prior to the onset of ARF ranged from 23% to 58%. sodes as recurrent or newly acquired infections. While 33% to 52% of the patients gave no history of an

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, NOV 2000 WWW.ARCHOTO.COM 1360

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 Table 1. Same-Serotype, Bona Fide Recurrent Streptococcal Infections Tonsillopharyngitis*

Initial Episode Recurrent Episode

Delta Log Delta Delta Log Delta Log Anti–DNase Delta Recurrent Delta Log Anti–DNase Delta Total ID No. M Type T Type ASO† B Titer‡ AC§ Episodeሻ M Type T Type ASO† B Titer‡ AC§ 199ND110.301 02223ND1100.301 22 2 102 ND 12 1.2 02023ND120.301 018 3 133 ND 6 0 0.602 16 21 ND 6 0.477 −0.6 29 4 147 ND 11 0.451 01955ND1100.451 23 5 195 12 12 0.6 0.301 21 25 12 12 0 0.301 24 6 198 12 12 0 0.453 27 34 12 12 0 0.301 23 7 199 4 4 0.301 0.152 20 40 4 4 0.452 0.15 21 8 208 12 12 0 0.602 19 46 12 12 0.451 −0.3 17 9 216 12 12 0.45 0 212312120.301 021 10 227 12 12 0.301 −0.301 23 32 12 12 −0.15 0.452 18 11 228 ND NT 0.301 02021NDNT00.477 21 12 403 ND 28 0.301 0 21 23 ND 28 −0.151 0.452 9 13 452 NT 3 0.301 02036NT30.301 018 14 477 12 12 0.15 0.602 17 30 12 12 −0.451 0.602 19 15 481 12 12 0.151 0.453 20 21 12 12 0.752 021 16 514 3 3 0.602 0.603 17 22 3 3 0.301 0.301 20 17 519 NT 25 0.452 0.752 18 36 NT 25 0.752 0.151 19 18 618 NT 12 0.151 0.301 19 35 NT 12 0 0.301 20 19 1028 3 3 −0.15 0.302 18 26 3 3 0.602 021

*ND indicates M serotyping not done; if the T type for the strain was the same in the first and second episode, it was assumed to be the same strain including if it was nontypeable (NT). Boldfaced numbers indicate log values of 0.2 or greater. †Delta log ASO indicates log antistreptolysin O (ASO) postinfection Ϫ log ASO preinfection. ‡Delta log DNase B indicates log anti-deoxyribonuclease B (Anti–DNase B) postinfection Ϫ log anti–DNase B preinfection. §Delta AC indicates the days between acute and convalescence titer. ࿣Delta recurrent episode indicates the days between the onset of the initial infection and the onset of the recurrent group A ␤-hemolytic streptococcal tonsillopharyngitis.

Table 2. Frequency of Sign or Symptom by Episode in Same-Serotype and Newly Acquired Streptococcal Tonsillopharyngitis

Same-Serotype Recurrence New Acquisition

No. (%) of No. (%) No. (%) of No. (%) of Initial Episodes Subsequent Episodes Initial Episodes Subsequent Episodes Symptom or Sign (n = 19) (n = 19) P* (n = 10) (n = 10) P* Fever 8 (42) 0 (0) .008 7 2 .06 Sore throat 11 (58) 1 (5) .006 8 6 .63 Headache 9 (47) 1 (5) .02 7 4 .37 Chills 5 (26) 2 (11) .453 4 2 .63 Stomachache 5 (26) 1 (5) .22 2 1 .99 Pharyngeal erythema 11 (58) 0 (0) Ͻ.001 7 4 .37 Pharyngeal edema 11 (58) 0 (0) Ͻ.001 6 2 .22 Pharyngeal exudate 10 (53) 2 (11) .04 6 2 .22 Adenitis 6 (32) 0 (0) .03 3 0 .25 Enlarged tonsils 4 (21) 1 (5) .37 7 4 .25 Scarletina 3 (16) 0 (0) .25 2 0 .5 Palatal petechiae 5 (26) 1 (5) .22 1 0 .99

*Calculated using the McNemar test. Boldfaced numbers indicate statistically significant values.

antecedent illness, the remainder of the patients re- tures, serotyping of all strains, follow-up, and multiple ported having mild flulike symptoms 1 to 3 weeks be- blood samplings for antibody testing.12 Recurrent GABHS fore the onset of ARF. The occurrence of ARF in pa- tonsillopharyngitis occurring 21 days or longer follow- tients with no definite clinical signs of pharyngitis ing the onset of the initial GABHS episode was confirmed emphasizes the challenge faced in our continuing ef- by a throat culture showing the same M and T strain of forts at the primary prevention of ARF. These observa- GABHS during an initial and repeated episode, and a 0.2- tions prompted the present study to expand our under- log or greater rise in either an ASO titer or an anti–DNase standing of the clinical presentation of streptococcal B titer after each episode. A significant increase (Ն0.2 log) disease that may precede ARF.23 in streptococcal antibody after each episode differenti- We used the information from a study completed in ated a streptococcal infection from persistent throat car- the 1980s because its design included multiple throat cul- riage. Although our definition of same-serotype recur-

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, NOV 2000 WWW.ARCHOTO.COM 1361

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 rent infection is sound, we cannot exclude the possibility published as an abstract in the American Journal of Dis- that concurrent viral infection caused signs and/or symp- eases of Children (1991;145:389-390). toms of infection in either or both GABHS episodes. Nine- This article is the view of the authors and is not in- teen (6%) of 295 patients who were part of a prospective tended to represent the opinion of the Food and Drug Ad- study had recurrent GABHS pharyngitis with the same se- ministration. rotype during the study. Fewer symptoms were associ- Corresponding author: Lucia H. Lee, MD, Food and ated with these infections. No difference in the frequency Drug Administration, Division of Vaccines and Related Prod- of signs or symptoms was observed with newly acquired uct Applications, HFM 475, Suite 370N, 1401 Rockville Pike, GABHS infections. The presence of fewer symptoms dur- Rockville, MD 20852. ing same-serotype recurrent streptococcal tonsillophar- yngitis continues to be observed in the 1990s (M.E.P., un- published data, 1990-1997). REFERENCES The same-serotype recurrent GABHS infections we studied occurred at a mean of 30 days after the initial epi- 1. Kaplan EL. Recent epidemiology of group A streptococcal infections in North sode (Table 1), tended to be seen with fewer signs and America and abroad: an overview. Pediatrics. 1996;97:945-948. 2. Hefelfinger DC. Resurgence of acute rheumatic fever in west Alabama. South Med symptoms and could have been easily missed. These pa- J. 1992;82:261-265. tients did not seek medical attention during any interim 3. Westlake RM, Graham TP, Edwards KM. An outbreak of acute rheumatic fever in period prior to the prescheduled follow-up visit. The du- Tennessee. Pediatr Infect Dis J. 1990;9:97-100. 4. Veasy LG, Wiedmeier SE, Orsmond GS, et al. Resurgence of acute rheumatic ration of symptoms with recurrent infections could not fever in the intermountain area of the United States. N Engl J Med. 1987; 316: be assessed owing to the study design. 421-427. 5. Hosier DM, Craenen JM, Teske DW, Wheller JJ. Resurgence of acute rheumatic In our study, of the 17 children with same- fever. AJDC. 1987;141:730-733. serotype recurrent GABHS infection for whom age- 6. Kaplan EL, Hill HR. Return of rheumatic fever: consequences, implications, and adjusted ASO and anti–DNase B normal values were avail- needs. J Pediatr. 1987;111:244-246. 24 7. Wald ER, Dashefsky B, Feidt C, Chiponis D, Byers C. Acute rheumatic fever in able, 8 children had either elevated preexisting ASO or western Pennsylvania and the tristate area. Pediatrics. 1987;80:371-374. anti–DNase B titers and 1 child had elevated preexisting 8. Griffiths SP, Gersony WM. Acute rheumatic fever in New York City (1969 to 1988): titers of both ASO and anti–DNase B. For these 9 chil- a comparative study of two decades. J Pediatr. 1990;116:882-887. 9. Marcon MJ, Hribar MM, Hosier DM, et al. Occurrence of mucoid M-18 Strepto- dren, the possibility cannot be excluded that the initial coccal pyogenes in a central Ohio pediatric population. J Clin Microbiol. 1988; episode for enrollment in our study was itself a recur- 26:1539-1542. 10. Kaplan EL, Johns DR, Cleary PP. Group A streptococcal serotypes isolated from rent infection. Regardless of whether the first infection patients and sibling contacts during the resurgence of rheumatic fever in the United at enrollment in the study represented an initial episode States in the mid-1980s. J Infect Dis. 1989;59:101-103. or a recurrent infection, fewer signs and symptoms were 11. Congeni B, Rizzo C, Congeni J, Sreenivasan VV. Outbreak of acute rheumatic fever in northeast Ohio. J Pediatr. 1987;111:176-179. observed during the subsequent GABHS infection due to 12. Disney FA, Dillon H, Blumer JL, et al. Cephalexin and penicillin in the treatment the same serotype. of group A ␤-hemolytic streptococcal throat infections. AJDC. 1992;146:1324- 1327. 13. Swift HF, Wilson AT, Lancefield RC. Typing group A hemolytic streptococci by CONCLUSIONS M precipitin reactions in capillary pipettes. J Exp Med. 1943;78:127-133. 14. Moody MD, Padula J, Lizona D, Hall CT. Epidemiologic characterization of group A streptococcus by T-agglutination and M-precipitin tests in the public health Patients with same-serotype bona fide recurrent GABHS lab. Health Lab Sci. 1965;2:149-622. pharyngitis may be at risk of developing the complica- 15. Ayoub EM, Harden E. Immune response to streptococcal antigens: diagnostic tions of this streptococcal infection. Since fewer signs and methods. In: Rose WR, de Macario EC, Folds JD, Lane HC, Nakamura RM, eds. Manual of Clinical Laboratory Immunology. Washington, DC: American Society symptoms occur in these patients, they may not be di- for Microbiology; 1997:450-457. agnosed and treated for a bona fide GABHS pharyngitis. 16. Wannamaker L, Ayoub E. Antibody titers in acute rheumatic fever. Circulation. The lack of typical symptoms in same-serotype recur- 1960;21:598-614. 17. Kaplan EL, Top FH Jr, Dudding BA, Wannamaker LW. Diagnosis of streptococ- rent GABHS tonsillopharyngitis suggests the need to re- cal pharyngitis: differentiation of active infection from the carrier state in the symp- evaluate our current practice regarding the follow-up man- tomatic child. J Infect Dis. 1971;123:490-501. 18. Kaplan EL, Gastanaduy AS, Huwe BB. The role of the carrier in treatment failures agement of these infections. At a minimum, physicians after antibiotic therapy for group A streptococci in the upper respiratory tract. should be aware that this phenomenon can occur and ad- J Lab Clin Med. 1981;98:326-335. vise patients that a recurrence of a throat infection with 19. Ayoub EM. Streptococcal antibody tests in rheumatic fever. Clin Immunol News- letter. 1982;3:107-111. fewer symptoms within 6 weeks of the initial episode 20. Ayoub EM. Immune response to group A streptococcal infections. Pediatr Infect should not be ignored. If a subsequent throat infection Dis. 1991;10(suppl):S15-S19. with fewer symptoms does occur, we recommend a care- 21. Ayoub EM, Wannamaker LW. Evaluation of the streptococcal desoxyribonucle- ase B and diphosphopyridine nucleotidase antibody tests in acute rheumatic fe- ful follow-up medical history and physical examina- ver and acute glomerulonephritis. Pediatrics. 1962;29:527-537. tion. If observations justify it, a throat culture should be 22. Levin RM, Grossman M, Jordan C, Ticknor W, Barnett P, Pascoe D. Group A strep- tococcal infections in children younger than three years of age. Pediatr Infect obtained. Dis J. 1988;7:581-587. 23. Dillon HC Jr. Streptococcal pharyngitis in the 1980s. Pediatr Infect Dis J. 1987; Accepted for publication April 11, 2000. 6:123-130. 24. Kaplan EL, Rothermel CD, Johnson DR. Antistreptolysin O and anti- Presented as a poster at the 31st Ambulatory Pediat- deoxyribonuclease B titers: normal values for children ages 2 to 12 in the United ric Society Meeting, New Orleans, La, April 29, 1991, and States. Pediatrics. 1998;101:86-88.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 126, NOV 2000 WWW.ARCHOTO.COM 1362

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021