Microbiology of Obstructive Tonsillar

Hypertrophy and Recurrent Tonsillitis

Izak H. Kielmovitch, MD; Georg Keleti, PhD; Charles D. Bluestone, MD; Ellen R. Wald, MD; Carlos Gonzalez, MD

\s=b\ A qualitative and quantitative analy-
sis of the tonsillar surface and core of
children with recurrent streptococcal ton-
sillitis and children with obstructive tonsil-
lar hypertrophy was performed. No quali-
tative difference was found within the two
population groups. Haemophilus influen-
zae and Bacteroides melaninogenicus
were the most prevalent \g=b\-lactamase\p=n-\
producing isolates in both groups. Staph-
ylococcus aureus had the highest rate of
\g=b\-lactamaseproduction on the tonsillar
surface of children with recurrent tonsilli-
A 20% to 40%
acute group
rate of
recurrence
A /3-hemolytic
streptococcal tonsillitis has been
reported after a standard course of
penicillin therapy and unchanged
organism susceptibility.13 The patho-
genesis of this recurrence was
explained in several ways. Household
contacts may harbor streptococci of
Accepted for publication December 20, 1988.
From the Department of Pediatric Otolaryn-
gology, Park Central Institute, St Louis, Mo (Dr
Kielmovitch); University of Pittsburgh (Pa)
Graduate School of Public Health (Dr Keleti);
Departments of Otolaryngology (Dr Bluestone)
and Pediatrics (Dr Wald), University of Pitts-
burgh School of Medicine; Department of Pediat-
ric Otolaryngology (Dr Bluestone) and Ambula-
tory Care Center (Dr Wald), Children's Hospital
of Pittsburgh; and Departments of Surgery and
Pediatrics, Uniformed Services University and
Otolaryngology Head and Neck Surgery Service,
Walter Reed Army Medical Center, Washington,
DC (Dr Gonzalez).
Reprint requests to Park Central Institute,
6125 Clayton Ave, Suite 430, St Louis, MO 63139
(Dr Kielmovitch).
tis, while Streptococcus pyogenes was
more prevalent in the tonsillar surface
cultures of children with obstructive ton-
sillar hypertrophy. The bacterial density
was high but not significantly different in
both groups of children. The similar
microbial composition and density of both
groups and the higher rate of S pyogenes
recovery may signify a subclinical disease
or normal flora in children with obstruc-
tive tonsillar hypertrophy.
(Arch Otolaryngol Head Neck Surg.
1989;115:721-724)
the
source
samestrain and constitute a
of reinfection.4 Streptococcus
pyogenes may be protected from peni¬
cillin by a penicillinase-producing
Staphylococcus aureus; however, peni-
cillinase-resistant antimicrobial
agents were tried, still without a sig¬
nificant reduction in the recurrence
rate of acute tonsillitis.56 Other aero¬
bic or anaerobic /3-lactamase-produc-
ing bacteria, not susceptible to
cillin or the staphylococcal penicillin-
ase-resistant antibiotics, may in
effect be the infecting agent.7 Recur-
Table 1.—Patient Age According to Population
Patient
Age, y Obstructive Tonsils
1-5 (42.3)
11
12 (46.2)
1 1-16 _3 (11.5)
Total 26 (100)
rent tonsillitiswas also suggested to
result from a mixed aerobic and
anaerobic polymicrobial infection
that in laboratory animals showed an
enhanced virulence.8·9
Therapy for recurrent acute strep-
tococcal tonsillitis is usually based on
tonsillar surface cultures. Certain
microorganisms confined to the ton¬
sillar core could therefore remain
untreated and cause the perpetuation
or the recurrence of this disease. A
discrepancy between the commonly
used tonsillar surface cultures and the
actually infecting tonsillar core
organisms was found in 30% to 40%
of the cases.1012 In contrast, a qualita¬
tive difference between the tonsillar
surface and tonsillar core flora was
not found in other reports.7
This study focused on three poten¬
tial factors in the etiology of recur¬
rent acute streptococcal tonsillitis and
obstructive tonsillar hypertrophy: (1)
peni¬ the aerobic and anaerobic microor¬
ganisms responsible for the tonsillar
inflammatory process; (2) the poten¬
tial discrepancy between tonsillar
Group
No. (%) of Patients
Recurrent Tonsillitis Total
_8 (32.0)_ 19
(44.0)
11 23
_6 (24.0)
25 (100) 51

Downloaded From: http://archotol.jamanetwork.com/ by a Cleveland Health Sciences Library User on 06/10/2015

 surface and tonsillar core microflora;
and (3) the importance of pMacta-
Table 2.—Bacterial Isolates Recovered per Patient According to Population Group
mase-producing aerobic and anaero¬ No. of Isolates/Patient
bic microorganisms in the cause and (% /i-Lactamase Producing)
pathogenesis of both entities. These Obstructive Tonsils Recurrent Tonsillitis
aims were accomplished by a qualita¬ Bacteria Location (N =
26) (N -
25)
tive and quantitative analysis of the Aerobic Surface 5 (35) 6 (100)
Core 4(38) 3 (60)
tonsillar surface and the tonsillar core
of tonsils obtained from two groups of Anaerobic Surface 4(96) 3(76)
Core 2 (27) 2(40)
children. One was a group of children
with recurrent acute streptococcal
tonsillitis and the other was a group
of children with obstructive tonsillar Table 3.—Bacteriology of the Tonsillar Surface and Core According to Population Group
hypertrophy.
% of Patients (% /3-Lactamaso-
PATIENTS AND METHODS Producing Bacteria)
Population Obstructive Recurrent
Tonsils Tonsillitis
Between June 1984 and July 1985, two Organism Location (N =
26) (N =
25)
groups of children who underwent elective Aerobic Bacteria
tonsillectomy at the Children's Hospital of Streptococcus viridans Surface 100 100
Pittsburgh (Pa) were selected for the Core 65 66
study. The first was a group of 25 children, Streptococcus B, C, F, G Surface 58 48
11 girls and 14 boys aged 2 to 16 years Core
(mean, 6 years), who had a history of three Streptococcus D Surface 88
or more episodes of recurrent acute group Core 19 40
A 0-hemolytic streptococcal tonsillitis dur¬ Streptococcus pyogenes Surface 46
ing the immediate 12 months preceding Core 27 8
surgery. The second was a group of 26 Haemophilus ¡nfluenzae Surface 69 (12) 76 (28)
children, 13 girlsand 13 boys aged 2 to 14 Core 58(12) 40 (8)
years (mean, 6 years), who had no history Staphylococcus aureus Surface 15(4) 44 (8)
of recurrent tonsillitis but required sur¬ Core 27 (15) 44 (40)
gery for signs and symptoms of upper Staphylococcus epidermidis Surface 15(0) 40(4)
aerodigestive tract obstruction due to Core 27 (0) 28(4)
obstructive tonsils (Table 1). Children Haemophilus parainñuenzae Surface 12(4) 40(12)
receiving any antimicrobial therapy dur¬ Core 4 (0,1 12(0)
ing the 4 weeks preceding surgery were Haemophilus parahaemolyticus Surface 15(0) 16(4)
excluded from this study. Core 4(0) K0)
Branhamella catarrhalls Surface 4(4) 8(0)
Microbiology Core 0(0) 0(0)
Immediately after being excised, one of Corynebacterium species Surface
each patient's randomly chosen tonsils was Core 12
transported in a sterile Petri dish to the Eikenella corrodens Surface
microbiology laboratory, where aerobic Core
and anaerobic cultures of the tonsillar Moraxella species Surface 12
surface and the tonsillar core were Core
obtained. Anaerobic Bacteria
The tonsillar surface of the freshly Bacteroides melaninogenlcus Surface 69 (54) 52 (36)
excised tonsil was thoroughly rubbed for Core 8(8) 40 (28)
30 s with a sterile cotton-tipped applicator, Fusobacterium necrophorum Surface 31 (0) 36 (0)
then put into a tube of 10 mL of freshly Core 31 (0) 16(0)
Veillonella párvula Surface 38 (0) 32 (0)
prepared thioglycollate medium and vor-
texed for 1 minute. Next, the tonsillar Core 15(0) 24(0)
surface was sterilized with a heated spatu¬ Peptostreptococcus micros Surface 35 (0) 28(0)
la. Through a stab incision in the sterilized Core 19(0) 16(0)
surface area, a tissue fragment from the Fusobacterium russii Surface 12(0) 12(0)
tonsillar core was removed by means of Core 12 (0) 12(0)
sterile technique. This tissue fragment was
then placed into 6 mL of freshly prepared
thioglycollate medium and homogenized in inoculated into tryptic soy broth and onto agar and the chocolate agar were incu¬
a Potter-El vehj em apparatus. 5% sheep blood agar, McConkey agar, and bated under 5% carbon dioxide at 35 °C for
For the detection of aerobic and faculta¬ chocolate agar. The McConkey agar and 48 hours. The isolated aerobic microorga¬
tively anaerobic bacteria, the tonsillar sur¬ tryptic soy broth were incubated aerobical- nisms were identified according to Len-
face and the tonsillar core specimens were ly at 35°C for 48 hours. The 5% sheep blood nette et al.13

Downloaded From: http://archotol.jamanetwork.com/ by a Cleveland Health Sciences Library User on 06/10/2015

 from the tonsillar surface and the
Table 4.—Mean Concentration of the Most Prevalent Bacteria
tonsillar core of both population
According to Population Group
groups was Haemophilus influenzae
Concentration (Table 3). Staphylococcus aureus
Obstructive Tonsils Recurrent Tonsillitis
showed a significantly higher rate of
Organism (N =
26) (N =
25) /3-lactamase production in the tonsil¬
Aerobic Bacteria lar surface (P .05) of the recurrent
=

Streptococcus viridans Surface 9.7 X 104/mL 1.1 X 105/mL tonsillitis group. The remaining iso¬
Core 2.5 X 106/g 3.7 X 10°/g lates did not show a significant differ¬
Neisseria species Surface 1.2 X 105/mL 5.3 X 104/mL
ence in the colonization and /3-lacta¬
Core 1.3 X 107/g 3.0 X 106/g
Haemophilus influenzae Surface 1.3 X 105/mL 1.5 X 105/mL mase production rates between the
Core 3.4 X 107/g 1.3 X 1C two population groups studied.
Staphylococcus aureus Surface 3.3 X 104/mL 2.2 X 104/mL The most prevalent aerobic non-
Core 2.1 X 107/g 5.2 X 10b/g ß-lactamase-producing isolates are
Staphylococcus epidermidis Surface 1.0 X 105/mL 1.1 X 106/mL also shown in Table 3. Significant was
Core 2.4 X 10d/g 5.2 X 10=/g the higher rate of S pyogenes coloniza¬
Streptococcus A Surface 1.4 X 105/mL 7.0 X 103/mL tion of the tonsillar surface in the
Core 1.7 X 10r/c 5.9 X 107/g obstructing tonsil group (P .05). No
=

Anaerobic Bacteria
Bacteroldes melaninogenlcus Surface
Core
Fusobacterium necrophorum Surface
Core
Veillonella párvula Surface
Core
Peptostreptococcus micros Surface
Core
Fusobacterium russii Surface
Core
For anaerobic growth, the tonsillar sur¬
face and tonsillar core specimens were
inoculated onto Centers for Disease Con¬
trol (Atlanta, Ga) anaerobic blood agar
with hemin, vitamin Ki (3-phydylmena-
dione), kanamycin-vancomycin laked blood
agar in Columbia colistine nalidixic acid.
The plates were then incubated in a gas
pack jar with a hydrogen-carbon dioxide
generator for 5 days at 35°C. The isolated
organisms were identified according to
Finegold et al.14 All organisms were tested
for /3-lactamase enzyme production by the
chromogenic cephalosporin method.
The surface and core of aerobic and
anaerobic bacteria from both groups of
children were quantified in colony-forming
units per milliliter and per gram, respec¬
tively.
Statistical Evaluation
The statistical significance of values
obtained from the microbiologie studies in
each of the population groups was deter¬
mined by means of Student's t test.
RESULTS
In the following analysis and tables,
children with a history of recurrent
acute streptococcal tonsillitis were
referred to as having "recurrent ton¬
sillitis." Children with a history of
6.0 X 104/mL 1.6 X 105/mL such difference was found among the
1.0 X 10b/ç 1.8 X 10'/g remaining bacterial strains isolated.
4.8 X 104/mL 1.2 X 104/mL Among the anaerobic bacteria (Ta¬
4.8 X 107/g 8.0 X 106/g ble 3), Bacteroides species were the
2.0 X 104/mL 1.1 X 106/mL most prevalent 0-lactamase-produc-
1.0 X 105/g 7.7 X 107/g ing bacteria. ß-Lactamase production
2.9 X 105/mL 7.4 X 104/mL within the other anaerobic isolates
3.4 X 107/g 8.0 X 107/g was uncommon. Bacteroides melani-
5.4 X 104/mL 2.7 X 104/mL
4.9 X 106/g 1.6 X 106/g
nogenicus was the most prevalent
anaerobic isolate and the most preva¬
lent anaerobic /3-lactamase-producing
obstructive tonsillar hypertrophy bacterium. The yield of anaerobic iso¬
were referred to as having "obstruc¬ lates, in general, tended to be higher
tive tonsils." from the tonsillar surface than from
All 51 subjects (51 tonsils) had one the tonsillar core and was similar in
or more bacterial strains isolated both population groups studied.
from the tonsillar core and the tonsil¬ Table 4 summarizes the aerobic and
lar surface. Table 2 summarizes the anaerobic bacterial density of the ton¬
mean number of bacterial isolates per sillar surface and the tonsillar core,
patient recovered and the prevalence respectively, in the two population
of /3-lactamäse-producing strains in groups. The bacterial concentration
each one of the patients studied. The was relatively high but not signifi¬
number of aerobic isolates was in gen¬ cantly different in both groups. The
eral somewhat higher than that of the aerobic bacteria ranged from
anaerobic isolates. The bacterial yield 2.2 X lOVmL to 1.5 X lOVmL (mean,
of the tonsillar surface was also 8.6 X lOVmL) in the tonsillar surface
slightly higher than the one of the and 1.2 X lOVg to 1.3 X lOVg (mean,
tonsillar core. No qualitative differ¬ 6.6 X lOVg) in the tonsillar core. The
ence was found between the tonsillar anaerobic bacteria ranged from
surface flora and the tonsillar core 1.2 X lOVmL to 1.1 X lOVmL (mean,
flora in either population group. The 5.6 X lOVmL) in the tonsillar surface
prevalence of aerobic and anaerobic and 1.0 X lOVg to 8.0 X lOVg (mean,
/3-lactamase-producing bacteria 4.0 X lOVg) in the tonsillar core.
tended to be higher in the recurrent
tonsillitis group than in the obstruct¬ COMMENT
ing tonsil group, especially among the In a qualitative and quantitative
aerobic isolates. However, these dif¬ analysis of the tonsillar surface and
ferences did not achieve statistical tonsillar core microflora, two popula¬
significance 'in either group. tions of children undergoing tonsillec-
The most prevalent aerobic ß-lacta- tomy were evaluated. The tonsillar
mase-producing bacteria recovered surface and tonsillar core of both pop-

Downloaded From: http://archotol.jamanetwork.com/ by a Cleveland Health Sciences Library User on 06/10/2015

 ulation groups were found to be colo¬
nized by an aerobic and anaerobic
polymicrobial flora. No qualitative
difference was found between the ton¬
sillar surface and the tonsillar core
flora within the two population
groups. A lower yield of isolates from
the tonsillar core was present in both
groups of tonsils. This might have
been due to a lower tonsillar core
colonization rate, consistent with the
fact that none of the children had an
acute illness at the time of surgery, or
to the sampling technique itself.
The prevalence of /3-lactamase-pro-
ducing organisms tended to be higher
in children with recurrent tonsillitis.
This was especially so within the aer¬
obic flora. Staphylococcus aureus
showed the highest /3-lactamase pro¬
duction rate among the aerobic orga¬
nisms in both population groups,
while H influenzae was the most
prevalent aerobic /3-lactamase-pro-
ducing isolate in both population
groups. The possible role of H influen¬
zae in the etiology of recurrent tonsil¬
litis and obstructive tonsillar hyper¬
trophy remains unknown, since this
organism may be simply a constituent
of normal tonsillar flora.
A special emphasis was recently
placed on the role of the anaerobic
flora in the pathogenesis of recurrent
tonsillitis.7·15·16 We also found mela-
ninogenicus to be the most prevalent
anaerobic bacterium. It also had the
highest rate of 0-lactamase produc¬
tion.
Not reported previously, to our
knowledge, was the composition of the
tonsillar surface and tonsillar core
microflora of obstructive tonsillar
hypertrophy. Its qualitative composi¬
tion and density were similar to those
in the group of children with recur¬
rent tonsillitis. The prevalence of ß-
lactamase-producing organisms
relatively high (Table 2), especially
within the surface of the anaerobic
flora. Streptococcus pyogenes coloni¬
zation within this group was also
higher than in the recurrent tonsilli¬
tis group, particularly in the tonsillar
surface cultures, where statistical sig¬
nificance was achieved. The signifi¬
cance of this is not clear, since these
findings could possibly mean subclini-
cal disease or normal flora. Certainly,
a trial with an antimicrobial agent
that is ß-lactamase resistant and has
a broad spectrum may be of benefit.
This type of antimicrobial trial may
also benefit patients who have had
recurrent acute streptococcal tonsilli¬
tis and in whom penicillin therapy has
failed. The number of potential tonsil-
lectomies, in both groups of children,
could thereby be decreased should this
therapy prove to be effective.
was

1. Rosentein BJ, Markowitz M, Goldstein E, et
al. Factors involved in treatment failures follow-
ing oral penicillin therapy of streptococcal phar-
yngitis. J Pediatr. 1968;73:513-520.
2. Brook I. The role of \g=b\-lactamaseproducing
bacteria in the persistence of streptococcal ton-
sillar infection. Rev Infect Dis. 1984;6:601-607.
3. Turner K, Nord CE. Impact of phenoxyme-
thylpenicillin and clindamycin on microflora in
recurrent tonsillitis. Ann Otol Rhinol Laryngol.
1985;94:278-280.
4. Stillerman M, Bernstein SH. Streptococcal
pharyngitis therapy. AJDC. 1964;107:35.
5. Brzezinska H, Brozik H, Mikucki J, Sobis M,
Szarapinska J. Concentrations of cloxacillin and
erythromycin in the tonsils of children after
administration of therapeutic doses. Int J Pe-
diatr Otorhinolaryngol. 1984;7:51-61.
6. Gastanadny AS, Kaplan EL, Huwe BB,
References
McKay C, Wannamaker LW. Failure of penicillin
to eradicate group A streptococci during an
outbreak of pharyngitis. Lancet. 1980;2:498.
7. Turner K, Nord CE. Beta-lactamase produc-
ing microorganisms in recurrent tonsillitis.
Scand J Infect Dis Suppl. 1983;39:83-85.
8. Socransky S, Gibbons R. Required role of
Bacteroides melaninogenicus in mixed anaerobic
infections. J Infect Dis. 1965;115:247-253.
9. Sundquist G, Eckerbom M, Larson A, Sjo-
gren T. Capacity of anaerobic bacteria from
necrotic dental pulps to induce purulent infec-
tions. Infect Immun. 1979;25:685-693.
10. Sprinkle P, Veltri R. The tonsil and ade-
noid dilemma: medical or surgical treatment.
Otolaryngol Clin North Am. 1974;7:909-925.
11. Rosen G, Samuel YL, Vered Y. Surface
tonsillar microflora vs deep tonsillar microflora
in recurrent acute tonsillitis. J Laryngol Otolar-
yngol. 1977;91:911-913.
12. Brook I, Yocum P, Shah K. Surface vs core
tonsillar aerobic and anaerobic flora in recurrent
tonsillitis. JAMA. 1980;244:1696-1698.
13. Lennette E, Balows A, Hansler YL, Truant
P. Manual of Clinical Microbiology. 3rd ed.
Washington, DC: American Society for Microbi-
ology; 1980:83-375.
14. Finegold S, Martin W, Scott E. Bailey and
Scott's Diagnostic Microbiology. 5th ed. St Louis,
Mo: CV Mosby Co; 1978:221-238.
15. Brook I, Grover A. Bacteroides melanino-
genicus: its recovery from tonsils of children
with acute tonsillitis. Arch Otolaryngol Head
Neck Surg 1983;109:818-820.
16. Brook I, Foot P. Comparison of the micro-
biology of recurrent tonsillitis between children
and adults. Laryngoscope. 1986;96:1385-1388.

Downloaded From: http://archotol.jamanetwork.com/ by a Cleveland Health Sciences Library User on 06/10/2015