Tonsillitis is a common disease of childhood and adolescence. The diagnosis of tonsillitis generally requires the consideration of Group A beta-hemolytic Streptococcus (GABHS) infection. However, numerous other bacteria alone or in combinations, viruses and other infections and non-infectious causes should be considered. Recognition of the cause and choice of appropriate therapy are of utmost importance in assuring rapid recovery and preventing complications.

Penicillin is currently the first-choice treatment for GABHS pharyngotonsillitis. However, the growing failure of penicillin to eradicate GABHS is of concern. This website discusses the potential causes of penicillin failure ( i.e. the presence of beta-lactamse producing bacteria that can “protect” GABHS from penicillins) and methods to overcome them. It also discusses the role of anaerobic bacteria in tonsillitis and its complications.


Penicillin Failure in Treatment of Streptococcal Pharyngo-Tonsillitis

Despite its excellent in vitro efficacy, the frequently reported inability of penicillin to eradicate Group A beta-hemolytic streptococci (GABHS) from patients with acute and relapsing pharyngo-tonsillitis (PT) is cause for concern.  Over the past 50 years, the rate of penicillin failure has consistently increased from about 7% in 1950 to almost 40% in 2000.1


                                   Dynamic of tonsillitis illustrating all stages of the infection





Various explanations exist for the failure of penicillin to eradicate GABHS PT (Table 1).  One possibility is the poor penetration of penicillin into the tonsillar tissues as well as into the epithelial cells.2 Other explanations relate to the bacterial interactions between GABHS and other members of the pharyngo-tonsillar bacterial flora.  For example, it is hypothesized that beta-lactamase secreted by beta-lactamase-producing bacteria (BLPB), which colonize the pharynx and tonsils, may “shield” GABHS from penicillins.3 Another possibility is the coaggregation between Moraxella catarrhalis and GABHS, which can facilitate colonization by GABHS.4 Normal bacterial flora can interfere with the growth of GABHS,5 and the absence of such competitive bacteria makes it easier for GABHS to colonize and invade the pharyngo-tonsillar area.

Table 1.  Possible Causes for Antibiotic Failure in Therapy of GABHS Tonsillitis

  • Bacterial Interactions
       The presence of beta-lactamase–producing organisms that “protect” GABHS from penicillins
       Coaggregation between GABHS and M. catarrhalis
       Absence of members of the oral bacterial flora capable of interfering with the growth of GABHS (through production of bacteriocins and/or competition on nutrients)
  • Poor penetration of penicillin into the tonsillar cells and tonsillar surface fluid ( allowing intracellular survival of GABHS)
  • Resistance (ie, erythromycin) or tolerance (ie, penicillin) to the antibiotic used
  • Inappropriate dose, duration of therapy, or choice of antibiotic
  • Poor compliance
  • Reacquisition of GABHS from a contact or an object (ie, toothbrush or dental braces)
  • Carrier state, not disease
  • Younger age 
  • Early initiation of treatment


The bacterial flora that colonize the tonsils and pharynx, in health as well as in illness, contain over 600 different strains of aerobic and anaerobic bacteria, and more than 108-11 bacteria/mL of secretion.  Anaerobic bacteria outnumber their aerobic counterparts by a ratio of 10:1 to 100:1.6 



                                         Oropharyngeal flora where anaerobes predominate





GABHS interacts with these other organisms in a synergistic or antagonistic fashion.7 These interactions can involve sharing of metabolites, exchange of genetic material, and influence of extra-cellular enzymes and other compounds produced by some bacteria on their partners.

The pharyngo-tonsillar area is also repeatedly infected by viral agents that can act synergistically with potential bacterial pathogens and normal flora.  Such interactions have been demonstrated in other respiratory tract infections.8 Pharyngo-tonsillitis occurs in a heavily colonized location, and should be considered as a potential polymicrobial infection.

This page describes the mechanisms by which bacterial interactions contribute to the failure of penicillins to eradicate GABHS from the acutely infected pharyngo-tonsillar area, possibly resulting in clinical and bacterial failure.  Also discussed are therapeutic modalities that can overcome these causes of failure.




                  The tonsillar crypts show GABHS mixed with other bacteria leading to bacterial interactions





The role of intracellular survival of GABHS and poor penetration of penicillin into the tonsils

Strains of GABHS have been demonstrated to internalize within epithelial cells both in vitro and in vivo in recent studies.9 The internalization-associated gene, prtF1/sfbI, has been identified more from patients with eradication failure of GABHS than has been recovered from patients with successful eradication.10 Since penicillins penetrate mammalian cells poorly, intracellular survival of GABHS possibly allows the pathogens to persist despite antibiotic treatment.11

The intracellular niche may therefore protect GABHS strains from penicillin that is not in high intracellular concentration.  In support of this hypothesis, GABHS strains were shown to survive 4-7 days within cultured epithelial cells.  Thus, internalization and intracellular survival represent a novel explanation for penicillin eradication failure.

Marouni et al.,12 compared the survival of GABHS strains from cases of eradication failure and eradication success, using an epithelial cell culture model.  “Eradication failure” strains showed significantly increased intracellular survival, compared to the 'eradication success' strains.  These results demonstrate how an intracellular reservoir of GABHS may play a role in the etiology of antibiotic eradication failure.
Kaplan et al.,2 recently examined the viability of intracellular GABHS in a human laryngeal epithelial cell line (HEp-2epithelial cell) after exposure to antibiotics (penicillin V, erythromycin, azithromycin, cephalothin, and clindamycin) that are commonly recommended for GABHS therapy.  Three techniques were used to study antibiotic killing of ingested GABHS: 

1) electron microscopy examination of ultrathin sections of internalized GABHS 
2) qualitative determination of intra–epithelial cell antibiotic
3) special stain evaluation of intracellular GABHS viability within antibiotic-treated epithelial cells.  

Group A beta-hemolytic streptococci survived intracellularly despite exposure of the GABHS-infected epithelial cells to penicillin.  Cephalothin (a cephalosporin) and clindamycin were more effective than penicillin in killing ingested GABHS.  However, erythromycin and azithromycin, agents known to accumulate to high levels in cells, were more effective than cephalothin and clindamycin in killing ingested GABHS.  These observations strongly suggest that the upper respiratory tract carrier state of GABHS results from intra-epithelial cell survival, and the failure of penicillin to kill internalized GABHS.

Penicillin’s failure to eradicate GABHS from pharyngo-tonsillar tissue may be impacted by its inability to eradicate intracellular GABHS as well as its inability to maintain sufficient concentration within the tonsillar fluid.

The stage of the GABHS PT inflammation determines the concentration of penicillin in tonsillar surface fluid.13  Stjernquist-Desatnik et al., investigated the concentration of penicillin in serum, as well as penetration to tonsillar surface fluid and saliva.  Among the nine healthy subjects, despite high serum penicillin concentrations (mean, 2.04 µg/ml), there was no penetration to tonsillar surface fluid or to saliva.  Of the nine patients with acute GABHS tonsillitis, eight manifested high concentrations of penicillin in tonsillar surface fluid (mean, 0.34 µg/ml) on the first day of treatment, but only two patients had penetration to the saliva.  On the tenth day of treatment, penicillin was present in the tonsillar surface fluid of only one patient and was not present in the saliva of any patients.  Furthermore, Orrling et al., demonstrated that the cephalosporins loracarbef and clindamycin maintained higher concentration in tonsillar surface fluid for longer duration than penicillin.14


The role of Beta-lactamase-producing bacteria

Penicillin therapy has resulted in a shift in the oral microflora over time by selecting for beta-lactamase-producing strains of Haemophilus spp, Staphylococcus aureus, M. catarrhalis, and anaerobic Gram-negative bacilli (eg, pigmented Prevotella, Porphyromonas and Fusobacterium spp.).15 These organisms are typically isolated from individuals who were recently treated with beta-lactam antibiotics. 3,15
Penicillin can be inactivated by BLPB, thereby protecting GABHS.3 Beta-lactam antibiotic therapy can select for BLPB that emerge in the oro-pharyngeal flora,15 and is especially common after repeated courses of beta-lactams given therapeutically or prophylactically.16  Antibiotic-treated individuals can also spread BLPB to other individuals.17



















                 Protection of penicillin susceptible bacteria by beta-lactamase producing bacteria





In-vitro demonstration of the "protection" phenomena
When grown with Staphylococcus aureus GABHS can resist penicillin

 


A clear association has been established in GABHS PT therapy between the failure of patients to respond to penicillin and the pre-existence of BLPB in pharyngo-tonsillar flora.18  Over 75% of tonsils removed are a result of recurrent tonsillitis harboring BLPB 8, 19-24 (Table 2).  Free beta-lactamase was detected in the core of most of those tonsils.23   Antibiotics that are effective against GABHS and are also resistant to the enzyme beta-lactamase attain higher success rates in eradication of acute and recurrent GABHS PT (Table 1).  These antibiotics included cephalosporins, lincomycin, clindamycin, macrolides, and amoxicillin-clavulanate.26-31

Table 2.          Prevalence of Beta-lactamase–Producing Bacteria in Excised Tonsils
Investigators
(country, year)
No. of Patients
% Beta-lactamase–producing bacteria
Brook et al.
 (USA, 1981)19
50
74
Reilly et al.
 (UK, 1981)20
41
78
Tuner and Nord
 (Sweden, 1982)21
167
73
Chagollan et al.
 (Mexico, 1984)22
10
80
Kielmovitch et al.
 (USA, 1989)23
25
100
Brook et al.
 (USA, 1995)24
50
94


A correlation was noted between the rate of recovery of BLPB in healthy children and the rate of amoxicillin failure to eradicate GABHS. Brook & Gober obtained pharyngo-tonsillar cultures from 228 children with GABHS pharyngo-tonsillitis, treated with amoxicillin for 10 days, and 663 healthy children.31a Amoxicillin failed to eradicate Group A streptococci from 48 of the 228 treated children (21%). Amoxicillin failure rate varied from month to month; it was high between October and May (22-32%), with the exception of April (11%); and low between June and September (8% to 12%). BLPB were recovered from 226 of 663 (34%) well children. The rate of recovery of BLPB varied; it was also high between October and May (40-52%), with exception of April (23%), and the lowest between June and September (10-12%). Prior to their treatment, BLPB were recovered from 26 of the 48 (54%) children who eventually failed amoxicillin therapy, and from 28 of the 180 (16%) who did not fail (p < 0.001). A high failure rate of penicillins in eradication of GABHS in PT can serve as sensitive indicator for a high prevalence rate of beta-lactamase-producing bacteria in the community.
Arecent study of 44 children who had undergone elective tonsillectomy reported the isolation of MRSA in the cores of in 7 (16%) of the surgically excised tonsils. 31b  Since most of the MRSA ( 5 of 7) were also beta-lactamase producers their presence could potentially interfere with the eradication of GABHS by penicillin.  MRSA that is also able to produce beta-lactamase can survive treatment with beta-lactam antibiotics and continue to “shield” GABHS from penicillins through the production of the enzyme beta-lactamase. Most of the S. aureus isolated from the tonsilar cores of the patients in the study ( 19 of 26 or 73% ) were, however, beta-lactamase producers and not MRSA. These organisms are susceptible to beta-lactamase resistant penicillins as well as most cephalosporins. 



 

The role of bacterial interference

An intricate microbial balance exists in the oro-pharynx, which includes the pharyngo-tonsillar area, between potential pathogens, (eg, GABHS, Haemophilus spp, Moraxella spp, and Streptococcus pneumoniae) and the normal oro-pharyngeal flora.6, 32-34




                               Bacterial interference as demonstrated on an blood agar plate






Illustration of bacterial interference as shown on the blood agar plate picture above



        

Prevention of upper respiratory tract bacterial infections is partially achieved because of bacterial interference.5  The normal oro-pharyngeal flora utilize various mechanisms to interfere with colonization and subsequent infection by pathogens, including competition for nutritional substances and the production of antibiotic-like substances called “bacteriocins” that kill other bacteria.5  The oro-pharyngeal flora of over 85% of individuals who are not otitis media-, sinusitis-, or tonsillitis-prone contains numerous organisms that are capable of interfering with the in vitro growth of potential pathogens.  In contrast, only 25% to 30% of children who suffer from recurrent upper respiratory tract bacterial infections harbor interfering organisms.32-34


      Bacterial interference: normal flora bacteria interfere with colonization by pathogens





Balance between normal and interfering bacteria maintains a healthy homeostasis




Only 30% of patients who suffer from recurrent GABHS PT harbor organisms that are capable of interfering with GABHS.34  On the other hand, 85% of individuals who stay GABHS PT-free are colonized by those protective organisms.  The major aerobic protective organisms are gamma- and alpha-hemolytic streptococci (AHS), and the predominant anaerobic bacteria are Peptostreptococcus spp. and Prevotella spp.  These organisms play a homeostatic role by colonizing the pharyngo-tonsillar area in sufficient numbers to prevent colonization and infection by GABHS.



Recovery of interfering organism is lower in patient with recurrent otitis, sinusitis & tonsillitis





Therapeutic colonization of the nasopharynx with interfering AHS, in an attempt to prevent relapsing GABHS PT, was evaluated in a series of studies performed in Göteborg, Sweden.35-38 These studies demonstrated the efficacy of this approach in reducing the clinical and bacteriological recurrence rate of GABHS PT in the AHS‑treated children as compared to those who received a placebo.  However, the use of AHS as a probiotic agent is not yet accepted as a therapeutic modality and is currently only experimental.

                           Efficacy of alpha-streptococci in preventing GABHS infection






Although artificial reconstitution of the interfering flora may be useful, preservation of the normal interfering flora is even more advantageous.  Since administration of antimicrobial agents can affect the composition of the nasopharyngeal bacterial flora, including the reduction of interfering organisms, judicious use of antimicrobial agents is important in the preservation of the normal interfering flora.39  Oral flora organisms with interfering capabilities are generally susceptible to amoxicillin, and include aerobic- and anaerobic‑streptococci, as well as penicillin-susceptible Prevotella spp.  Amoxicillin‑clavulanate is also effective against potentially interfering beta-lactamase-producing Gram-negative bacilli (ie, Prevotella spp.).  In contrast, these organisms are relatively resistant to the extended spectrum and second- and third-generation cephalosporins.40
Brook and Gober39 compared the effects of amoxicillin-clavulanate and cefdinir on the nasopharyngeal flora in children with acute otitis media.  While both agents are effective against potential pathogens (S. pneumoniae, H. influenzae and M. catarrhalis), they have selective activity against members of the nasopharyngeal bacterial flora.  At the end of amoxicillin-clavulanate therapy, the oral flora were more depleted of aerobic and anaerobic organisms with interfering capability than observed following cefdinir therapy. The differences between the two therapy groups persisted through two other specimens taken a month and two months later, and correlated with the faster reacquisition of potential pathogenic bacteria that took place in those treated with amoxicillin-clavulanate.
This study suggests a potential beneficial effect of using a narrow-spectrum antimicrobial that selectively spares the interfering organisms while eliminating pathogenic bacteria.  The benefit of such therapy is the prevention of reacquisition of pathogenic bacteria in the oro-pharynx.  In contrast, utilization of a broad-spectrum antimicrobial is associated with prolonged absence of inhibitory organisms, and rapid recolonization with potential pathogens.

The role of coaggregation between M. catarrhalis and GABHS

Several studies suggest that the colonization of the tonsils by GABHS and certain other aerobic and anaerobic bacterial species may contribute to the inflammatory process and the failure of penicillin therapy.41  A mutual symbiotic enhancement of growth of GABHS in the presence of other aerobic and anaerobic bacteria was demonstrated in an animal model.42  Such a synergistic relationship may also exist between these organisms and GABHS in patients with PT.  An example of a potential synergistic relationship is the ability of M. catarrhalis to increase the adherence of GABHS to human epithelial cells through species-specific coaggregation.4



     Illustration of the ability of GABHS to adhere to the epithelial cells assisted by M. catarrhalis





Brook and Gober43 investigated whether the recovery of M. catarrhalis, H. influenzae, S. aureus, and S. pneumoniae is associated with the isolation GABHS.  Among 548 children with acute PT, GABHS was isolated from 112 (20.4%) children.  Of the 114 H. influenzae isolates, 32 isolates were associated with GABHS and 82 isolates were recovered without GABHS (P<0.05).  Of the 69 M. catarrhalis isolates, 25 isolates were associated with GABHS and 44 isolates were recovered without GABHS (P<0.05).  In contrast, there was no association between the isolation of GABHS and S. aureus or S. pneumoniae.  One hundred four isolates of GABHS were recovered from 548 healthy children.  Of the 69 M. catarrhalis isolates, 24 isolates were associated with GABHS (23% of all patients with GABHS) and 80 isolates were recovered without GABHS (10%) (P<0.05).  There was no association between the isolation of GABHS and the recovery of H. influenzae, S. aureus, or S. pneumoniae among healthy childrenThis study demonstrates an association between the recovery of GABHS and H. influenzae and M. catarrhalis from patients with PT, and between GABHS and M. catarrhalis from healthy children.
The increased isolation of H. influenzae (in PT only) and M. catarrhalis in association with GABHS may be due to a synergistic relationship between these organisms.5, 43  The ability of H. influenzae and M. catarrhalis to produce beta-lactamase, which can inactivate the penicillin in the tonsillar tissues,4 may also make these organisms, as well as protecting GABHS, more resistant to eradication by penicillin and contribute to the failure of penicillin therapy.

An indirect support for the potential synergistic relationship between GABHS and H. influenzae and M. catarrhalis is the better clinical efficacy in eradicating GABHS, as compared to penicillin, of antimicrobial effective against these organisms.  These antimicrobials include the second, extended-spectrum, and third-generation cephalosporins44, 45 as well as amoxicillin-clavulanate.28, 29  The superior efficacy of these agents compared to penicillin may be due to their direct antibacterial activity against GABHS as well as beta-lactamase producing H. influenzae and M. catarrhalis.



The implications of microbiologic interactions on therapy

Penicillin is still recommended by some guidelines as the antibiotic of choice, (46) although other antibiotics are more effective in the bacteriologic and clinical cure of acute and recurrent GABHS PT.27-31 Macrolides are more effective than penicillin in acute GABHS PT,44, 45 while lincomycin, clindamycin, and AC are more effective in relapsing GABHS PT.28, 29  Cephalosporins are superior to penicillins in both acute and relapsing GABHS PT. 44, 45
The goal of the treatment of PT in patients who failed penicillin therapy, is to eradicate the BLPB that protect GABHS from penicillin, while preserving whatever "protective" interfering organisms  (ie, AHS) may be present at the pharyngo-tonsillar area.  Cephalosporins have been successful in eradicating GABHS better,44 and in some instances much faster45 than penicillin.  Over 35 randomized studies showed cephalosporins of all classes to have a much higher success rate in eradicating GABHS than penicillin (a third of the failure rate of penicillin),44 and 12 studies illustrated equal or better success rate in 5 to 7 days of therapy as compared with 10 days of penicillin.47
A study by Brook & Gober48 demonstrated that GABHS was eradicated from children’s tonsils 48 hours earlier in those treated with cefdinir than in those treated with amoxicillin.  At the end of therapy GABHS was recovered from 5 (20%) amoxicillin‑treated patients and 2 (8%) cefdinir-treated patients.

The explanation for the ability of cephalosporins to perform so well is that even though they are as efficacious in vitro as penicillin in eradicating GABHS, penicillin also eliminates the aerobic- and anaerobic-interfering organisms.  Penicillin use may therefore deprive the patient of the potential beneficial effects of interfering organisms, which can reduce colonization with GABHS.  Cephalosporins, on the other hand, are less inhibitory of aerobic- and anaerobic-interfering organisms, thus these organisms are more likely to be preserved following therapy.43 The higher their generation, the less effective are the cephalosporins against both aerobic- and anaerobic-interfering organisms. In a paradoxical way, cephalosporins lesser efficacy against interfering organisms is their potential advantage.  Therefore the administration of a cephalosporin to patients has a selective effect: preserving both aerobic- and anaerobic-interfering organisms while eradicating GABHS (Table 3). The sparing effects of the cephalosporins persisted for at least 2 month after their use. 39


Table 3.          Antibacterial Activity of Penicillin Compared to Cephalosporins in the Management of Acute GABHS Tonsillitis
Antimicrobial Activity
Penicillins
Cephalosporins
Aerobic beta-lactamase–producing bacteria
No
Yes
Interfering organisms
Yes
No
GABHS
Yes
Yes



Another advantage of cephalosporins is that they are generally resistant to beta-lactamase.  Therefore, they are not hydrolyzed by extracellular beta-lactamases produced by BLPB.  However, cephalosporins antibacterial efficacy against BLPB themselves is generation-dependent (Table 4).  First-generation cephalosporins (eg, cephalexin, cefadroxil) are effective only against S. aureus; second-generation  (cefprozil, cefuroxime acetil) and extended-spectrum third generation (cefdinir and cefpodoxime axetile) cephalosporins are effective against S. aureus, Haemophilus spp. and Moraxella spp.; and third-generation (eg, cefixime, ceftibuten) cephalosporins are only effective against Haemophilus spp. and Moraxella spp.  However, as a group, the cephalosporins are capable of overcoming BLPB, including M. catarrhalis, which allows for microbial co‑aggregation; preserving interfering organisms; and eradicating GABHS.


Table 4.          Antibacterial Activity of Cephalosporins Against Aerobic Beta‑lactamase–Producing Bacteria.
beta-lactamase–producing bacteria
First-generation (cephalothin)
Second- generation (cefprozil, cefuroxime)
Extended spectrum (cefdinir, cefpedoxime)
Third- generation (cefixime, cefibuten)
S. aureus
Yes
Yes
Yes
No
H. influenzae
No
Yes
Yes
Yes
M. catarrhalis
No
Yes
Yes
Yes


The above selective activity of cephalosporins has been demonstrated in an animal model study of mixed infection,40 as well as in patients who were treated with an extended-spectrum cephalosporin prior to elective tonsillectomy.48
Brook and Foote48 compared two modes of therapy of relapsing tonsillitis due to GABHS: one using penicillin and the other using an extended-spectrum cephalosporin (cefdinir).  Both agents were administered for 10 days prior to surgery to children who suffered from recurrent GABHS tonsillitis and were scheduled for elective tonsillectomy.  Of the two, cefdinir was more effective in eradicating GABHS, reducing the number of BLPB, and preserving AHS that are capable of inhibiting GABHS.  These data illustrate the ability of an oral extended-spectrum cephalosporin, in comparison to penicillin, to eradicate GABHS as well as BLPB, while preserving the interfering AHS.

References

1.          Pichichero ME, Casey JR.  The evidence base for cephalosporin superiority over penicillin in streptococcal pharyngitis.  Diagn Microbiol Infect Dis. 2007;57(Supplement):S39-S45.
2.          Kaplan EL, Chatwal GS, Rohde M.  Reduced ability of penicillin to eradicate ingested Group A streptococci from epithelial cells: clinical and pathogenetic implications. Clin Infect Dis. 2006;43:1398-406.
5.          Grahn E, Holm SE.  Bacterial interference in the throat flora during a streptococcal tonsillitis outbreak in an apartment house area.  Zbl Bakl Hyg A. 1983;256:72–79.
6.          Socransky SS, Manganiello SD.  The oral microflora of man from birth to Senility.  J Periodontol 42, 485-96, 1971.
8.          Tarsia P, Aliberti S, Pappalettera M, Blasi F.  Mixed community-acquired lower respiratory tract infections.  Curr. Infect. Dis. Rep. 9:14-20, 2007.
9.          Hagman MM, Dale JB, Stevens DL.  Comparison of adherence to and penetration of a human laryngeal epithelial cell line by GABHS of various M protein types.  FEMS Immunol Med Microbiol. 1999;23:195-204.
10.          Neeman R, Keller N, Barzilai A, Korenman Z, Sela S.  Prevalence of internalization‑associated gene, prtF1, among persisting group-A streptococcus strains isolated from asymptomatic carriers.  Lancet. 1998;352:1974-7).
11.          Stjernquist-Desatnik A, Samuelsson P, Walder M.  Penetration of penicillin V to tonsillar surface fluid in healthy individuals and in patients with acute tonsillitis.  J Laryngol Otol. 1993;107:309-12.
12.          Marouni MJ, Barzilai A, Keller N, Rubinstein E, Sela S.  Intracellular survival of persistent Group A streptococci in cultured epithelial cells.  Int J Med Microbiol. 2004;294:27-33.
13.          Stjernquist-Desatnik A, Samuelsson P, Walder M.  Penetration of penicillin V to tonsillar surface fluid in healthy individuals and in patients with acute tonsillitis.  J Laryngol Otol. 1993 ;107:309-12.
14.          Orrling A, Kamme C, Stjernquist-Desatnik A. Scand J Infect Dis.  Penicillin V, loracarbef and clindamycin in tonsillar surface fluid during acute Group A streptococcal pharyngotonsillitis.  2005;37:429-35
20.          Reilly S, Timmis P, Beeden AG, Willis AT.  Possible role of the anaerobe in tonsillitis.  J Clin Pathol. 1981;34:542–547.
21.          Tuner K, Nord CE.  Beta-lactamase–producing anaerobic bacteria in recurrent tonsillitis.  J Antimicrob Chemother. 1982;10(suppl A):153-156.
22.          Chagollan J, Macias JR, Gil JS.  Flora indigena de las amigdales.  Invest Med Int. 1984;11:36-39.
23.          Kielmovitch IH, Keleti G, Bluestone CD, Wald ER, Gonzales C.  Microbiology of obstructive tonsillar hypertrophy and recurrent tonsillitis.  Arch Otolaryngol Head Neck Surg. 1989;115:721-725.
26.          Brook I. Beta-lactamase-producing bacteria and their role in infection.  Rev Med Microbiol. 2005;16:91-99.
29.          Kaplan EL, Johnson DR.  Eradication of Group A streptococci from the upper respiratory tract by amoxicillin with clavulanate after oral penicillin V treatment failure.  J Pediatr. 1988;113:400-403.
30.          Holm S, Henning C, Grahn E, Lomberg H, Staley H.  Is penicillin the appropriate treatment for recurrent tonsillopharyngitis?  Results from a comparative randomized blind study of cefuroxime axetil and phenoxymethylpenicillin in children.  The Swedish Study Group.  Scand J Infect Dis. 1995;27:221-228.
31.          Holm SE, Roos K, Stromberg A.  A randomized study of treatment of streptococcal pharyngotonsillitis with cefadroxil or phenoxymethylpenicillin (penicillin V).  Pediatr Infect Dis J. 1991;10:S68-71.
31a. Brook I, Gober AE. Failure to eradicate streptococci and beta-lactamase producing bacteria. Acta Paediatrica. 2008; 96;193-195.
31b. Brook, I., Foote, P.A. Jr. Isolation of methicillin resistant Staphylococcus aureus from the surfaceand core of tonsils in children. Int J Pediatr Otorhinolaryngol. 2006;70:2099-2102.

35.          Roos K, Grahn E, Holm SE, Johansson H, Lind L.  Interfering alpha‑streptococci as a protection against recurrent streptococcal tonsillitis in children.  Int J Pediatr Otol. 1993;25:141‑148.
36.          Roos K, Holm SE, Grahn E, Lind L.  Alpha-strepharyngo-tonsillitisococci as supplementary treatment of recurrent streptococcal tonsillitis:  a randomized placebo-controlled study.  Scand J Infect Dis. 1993;25:31‑35.
37.          Roos K, Holm SE, Grahn-Hakansson E, Lagergren L.  Recolonization with selected alpha- streptococci for prophylaxis of recurrent streptococcal pharyngotonsillitis -- a randomized placebo-controlled multicentre study.  Scand J Infect Dis. 1996;28:459‑462.
38.          Falck G, Grahn-Håkansson E, Holm SE, Roos K, Lagergren L.  Tolerance and efficacy of interfering alpha-streptococci in recurrence of strepharyngo-tonsillitisococcal pharyngotonsillitis:  a placebo-controlled study.  Acta Otolaryngol (Stockh). 1999;119:944-948.
42.          Brook I, Gillmore JD.  Enhancement of growth of Group A beta-hemolytic streptococci in mixed infections with aerobic and anaerobic bacteria.  Clin Microbiol Infect. 1996;1:179-182.
44.          Casey JR, Pichichero ME.  Meta-analysis of cephalosporins versus penicillin for treatment of Group A  streptococcal tonsillopharyngitis in adults.  Clin Infect Dis. 2004;38:1526-1534.
45.          Pichichero ME.  Cephalosporins are superior to penicillin for treatment of  streptococcal  tonsillopharyngitis: is the difference worth it?  Pediatr Infect Dis J. 1993;12:268-274.
46.          Bisno AL, Gerber MA, Gwaltney JM, Jr, Kaplan EL, Schwartz RH.  Practice guidelines for the diagnosis and management of Group A streptococcal pharyngitis.  Clin Infect Dis. 2002;35:113-125.