J Clin Pathol: first published as 10.1136/jcp.34.5.542 on 1 May 1981. Downloaded from J Clin Pathol 1981 ;34:542-547 Possible role of the anaerobe in tonsillitis SHEENA REILLY,* P TIMMIS,t AG BEEDEN,t AT WILLIS* From the Departments of *Microbiology and tENT Surgery, Luton and Dunstable Hospital, Luton, Bedfordshire SUMMARY Anaerobic bacteria were isolated from all tonsils removed from children at routine tonsillectomy; 75-6 % of specimens yielded moderate to heavy growth and 80 % of tonsils contained more than one anaerobic species. This recovery rate fell to 56 % after a 10-day course ofmetronidazole before tonsillectomy-in only 14-6 % of cases were anaerobes isolated in significant numbers. Surface swabbing of the tonsils permitted recovery of a similar spectrum of anaerobic bacteria but resulted in an overall loss of both aerobic and anaerobic pathogens. A comparison was made between the flora ofacutely inflamed tonsils and "healthy" tonsils: over 90 ofboth groups yielded anaerobic bacteria, but they were present in significant numbers in 56-2 % of swabs taken from acutely inflamed tonsils compared with 24 % of swabs from "healthy" children. The isolation rate for aerobic patho- gens was 37-5 % and 16 0% respectively. In many cases of acute andrecurrentacute tonsillitis, and throat department with histories of recurrent a microbiological diagnosis is not made. One possible acute tonsillitis, was selected for tonsillar swabbing.copyright. explanation is that the bacteria sampled by the sur- With the aid of the light of a head-lamp, sterile face swabbing technique are not an accurate reflec- cotton wool swabs were rubbed over the tonsillar tion of the flora of the tonsillar tissue.' Alterna- surfaces, care being taken to avoid oral contamina- tively laboratory isolation techniques may be tion. The swabs were transported to the laboratory incriminated. It has been suggested that hitherto in Stuart's transport medium. unrecognised penicillin-sensitive bacteria cause non- Tonsillar swabs were similarly taken by AGB and http://jcp.bmj.com/ streptococcal tonsillitis.2 The aetiological role of PT from 25 children, aged 2-17 years, who were anaerobic bacteria has received little attention. The referred to the ENT department for problems aim of this study was to establish the detailed unrelated to their tonsils (controls), and from 16 anaerobic bacteriology ofchildren's tonsils at various children, aged 2-18 years, with a clinical diagnosis pathological stages, in an attempt to delineate a of acute tonsillitis. pattern of causal relationship. BACTERIOLOGICAL CULTURE MEDIA on September 25, 2021 by guest. Protected Patients and methods Prepoured horse blood agar and chocolate agar plates (Lab M, Salford, Lancs) were chosen as the PATIENTS aerobic culture media. A selective medium for Whole tonsils removed at routine tonsillectomy by Corynebacterium diphtheriae was not used. The basal two of us (AGB and PT) were submitted for bacteri- medium for the anaerobic blood agar plates com- ological examination. A total of 41 specimens was prised Columbia agar base plus yeast extract 0-5 % examined, all from children with a past history of wt/vol to which were added the following: haemin uncomplicated recurrent tonsillitis. At the time of (BDH Ltd) 0-005 g/l; menadione (Sigma Ltd) removal the tonsils were not clinically inflamed. 0-0005 g/l; L-cysteine (BDH Ltd) 0 3 g/l; horse With their parents' consent, 41 more children were blood (Oxoid Ltd) 5% v/v. The selective medium given oral metronidazole (7 5 mg/kg three times initially chosen included vancomycin 0 0075 g/l and a day) for 10 days before tonsillectomy, after which kanamycin 0 075 g/l. Neomycin 0-075 g/l was later the excised tonsils were cultured in an identical substituted for kanamycin after the results of manner. A comparable group of 37 children aged preliminary experiments which indicated a low 3-12 years, who had been referred to the ear, nose, recovery rate of Fusobacterium spp with the vanco- mysin-kanamycin combination. The plates for Accepted for publication 8 October 1980 anaerobic culture were freshly poured as required, 542 J Clin Pathol: first published as 10.1136/jcp.34.5.542 on 1 May 1981. Downloaded from Possible role of the anaerobe in tonsillitis 543 sealed in plastic bags, and their surfaces dried before taining antibiotic concentrations ranging from use. 0 003-32 0 ,ug/ml. The lowest concentration of penicillin and ampicillin preventing colony forma- VIRUS CULTURE tion after incubation in an anaerobic jar at 370C for It was not possible to perform virus culture on the 48 hours was taken as the MIC for that strain. A majority of specimens. However, 20 of the tonsils strain of Staphylococcus aureus of known MIC was from children with recurrent tonsillitis were inocu- included in each set of tests. lated on to three cell lines: monkey kidney, HEp2, The production of penicillinase by a few of these and MRC5. strains was assessed using Intralactam strips (Mast Laboratories Ltd, Bootle, Merseyside). No attempt PREPARATION OF SPECIMENS was made to induce penicillinase by growing the The tonsils were transported in sterile universal strains of B melaninogenicus in the presence of containers with the minimum of delay from theatre penicillin before testing them by the strip method. to the laboratory. Transport medium was not used. Each tonsil was sectioned with a sterile scalpel and Results impression smears of the cut surface were stained by the methods of Gram and Leishman. The tissue was STAINED FILMS OF SPECIMENS then chopped with scissors to a very fine consistency In general, the Gram and Leishman stained smears and a standard loopful inoculated on to each of the were unhelpful: stainable material bore little correla- four bacteriological media. Preliminary investiga- tion to subsequent bacteriological findings. A notable tions had shown that washing the tonsils in sterile exception was a case of Vincent's stomatitis, present- saline was an unnecessary procedure. Smears were ing as a painful oropharynx, in which tonsillar likewise prepared from the tonsillar swabs, which swabs stained with carbol fuchsin showed an were then inoculated on to the plates in the routine abundance of fusobacteria and spirochaetal forms. manner. CULTURE OF WHOLE TONSILS copyright. INCUBATION OF CULTURES The aerobic blood agar plates were examined after Aerobic pathogens overnight incubation at 37°C, as were the chocolate A semiquantitative assessment of bacterial growth plates which were incubated in candle jars. All was made in terms of + + + (heavy) to ± (scanty). anaerobic plates were incubated in 10-litre anaerobic fl-haemolytic streptococci of Lancefield's groups A, jars (Don Whitley Scientific) fitted with freshly C, or G were isolated in significant numbers from http://jcp.bmj.com/ activated palladium catalyst, which were evacuated 15 out of 41 tonsils (36-6 %), and a heavy to moderate to an internal pressure of440 mm Hg and the vacuum growth of Haemophilus influenzae was recovered replaced with a gas mixture of 10% CO2, 90% H2. from 14 out of 41 tonsils (34-1 %). There were no Growth was first seen after 48 hours' incubation and isolates of Streptococcus pneumoniae. Staph aureus was inspected daily for a total of six days. was isolated in large numbers from seven specimens (17%). IDENTIFICATION OF BACTERIA Aerobic bacteria were identified according to the Anaerobic isolates (Table 1) on September 25, 2021 by guest. Protected methods of Cowan (1974).3 All anaerobic isolates Eighty per cent of the tonsils contained more than were identified to species level as far as possible using one anaerobic species. By using a 5 ,ug metronidazole the biochemical criteria and gas-liquid chromato- disc on the non-selective anaerobic plate, the heavy graphic patterns of Holdeman et al.4 growth of anaerobes was readily discernible from the large zone of inhibition. Simultaneous inspection MINIMUM INHIBITORY CONCENTRATION of the neomycin-vancomycin plates facilitated the (MIC) DETERMINATIONS recovery of individual anaerobic species. After six In view of the fact that the majority of strains of days' incubation, anaerobes were isolatedfrom 100 % Bacteroides melaninogenicus recovered from the of the tonsils. There was moderate to heavy anaerobic children's tonsils were resistant on disc testing to 1 growth in 75 6% of specimens. unit of benzylpenicillin, the MICs of penicillin and B melaninogenicus was the most frequently ampicillin were determined for 38 of the isolates. A isolated anaerobic bacterium, being present in 95 % suspension ofeach organism was prepared in nutrient of the tonsils examined. In 46 % of these specimens, broth to a concentration of 2 x 104 cells/ml, more than one colonial type of B melaninogenicus from which 1 ,lI was applied to the surface of was detectable. Full identification of some isolates well-dried, freshly prepared blood agar plates con- was impossible, but the majority belonged to the two J Clin Pathol: first published as 10.1136/jcp.34.5.542 on 1 May 1981. Downloaded from 544 Reilly, Timmis, Beeden, Willis Table 1 A summary of the anaerobic bacteria isolatedfrom the five main groups ofspecimens Anaerobe No. of specimens Tonsils (41) Tonsils after 10 Tonsil swabs Tonsil swabs Tonsil swabs days metroni- (recurrent tonsil- (acute tonsillitis) (controls) (25) dazole (41) litis) (37) (16) Bacteroides melaninogenicus ss melaninogenicus and ss intermedius 50 18 34 19 16 Bacteroides melaninogenicus ss unidentified 12 10 10 4 10 Bacteroides spp (unidentified) 28 7 19 7 7 B ruminicola ss ruminicola 3 0 0 0 0 Fusobacterium spp 22 6 10 5 16 Leptotrichia buccalis 2 0 0 0 0 Veillonella parvula 0 0 2 2 0 Peptococcus/peptostreptococcus spp 7 0 1 0 0 Eubacterium spp 5 0 9 2 6 Clostridium spp 0 0 0 0 0 Bifidobacterium spp 2 0 0 0 0 Total 131 41 85 39 55 Anaerobic bacteria isolated per specimen 3-2 1-0 2-3 2-4 2-2 subspecies melaninogenicus and intermedius.