Vol. 9, No. 2 0095-1137/79/02-0189/05$02.00/0
Evaluation of Sulfamethoxazole-Trimethoprim
Blood Agar
Plates
for Recovery of Group
A
Streptococci from Throat
Cultures
TERRENCE A. KURZYNSKI* AND CHERYLL K. MEISE
State Laboratory of Hygiene,Madison, Wisconsin53706 Received forpublication20November 1978
Wecompared the selective bloodagarmedium of Gunnetal.(J.Clin.Microbiol.
5:650-655, 1977) which containssulfamethoxazole plus trimethoprim (SXT-BA)
to the conventional bloodagarsurface plate (SBA) and a modified blood agar
pourplate plus brothmethod for therecoveryofgroupAstreptococcifromthroat
swabs. The influence of CO2 and ambient air incubation of the SXT-BA and SBA
plateswasalso evaluated. A total of696throatswabsfrom symptomatic children
werecultured simultaneously by the five methods and observed after overnight
incubation; 204 positive cultureswere detected overall. Recoveryratesof each
individual methodwere:SXT-BA (CO2),90.7%; SXT-BA (air), 87.7%;pourplate
plus broth, 83.3%; SBA(CO2),79.4%; and SBA (air) 77%. Approximatelyone-half
ofthe false-negative cultures in the SXT-BA (C02) and SXT-BA (air) methods
had colonycountsof -10to100coloniesperplate. Incontrast,for theSBA(CO2),
SBA (air), andpourplate plus broth methods, approximately 70% of the
false-negative cultures had colony countsof 210to100/plate. False-positive cultures
obtained by the SXT-BA (002) and SXT-BA (air) methodswere11 and 12.7%,
respectively-one-halfashigh astheratesobtained by the remainingmethods.
Beta-hemolyticstreptococci, groups C, F, and G, areinhibited onthe SXT-BA
plates andweretheprimarycause ofthe higher false-positiveratesonSBA and
pourplate plus broth methods. An additional 3% positive cultureswereobtained
by incubatingSXT-BA (002) platesupto48 hbefore discardingasnegative. We
recommend either the SXT-BA (002)ortheSXT-BA(air)method withupto48
h of incubationfor routineusein throat cultures.
Recently, Gunn et al. (6) introduced a new
selective surface
plating
medium, which containssulfamethoxazole plus trimethoprim (SXT-BA),
for the enhanced recovery ofgroup A
strepto-cocci from throat cultures. We have compared
the SXT-BA
plate
to the conventional sheepbloodagar(SBA) method andto amodification
ofapourplate plus broth (PP+B) method.We
also found the SXT-BA
plate
tobesuperior
tothe SBA
plate
in this and in a previous study(8a). TheSXT-BAwasalsomoresensitive than
themodifiedPP+Bmethod for the isolationof
group Astreptococci.
MATERIALS AND METHODS
SXT-BA and SBAplatesweremadeinour labora-toryby usingTrypticase soy agar(Baltimore
Biologi-cal Laboratories [BBL]).Thestock solution of sulfa-methoxazoleplustrimethoprim (BurroughsWellcome
Co.)wasmadeasoriginallydescribed(6),exceptthat
thefinal volumewasmadeupto 1literinsteadofthe recommended100ml.This greaterdilutionwasused toavoidworkingwith thesuspensionof thetwodrugs
that occurs with lesser dilution. Quantities of
sulfa-methoxazole-plus-trimethoprim stock solution were storedat-20°C until the day of use. For the SXT-BA plates, 10ml of stock solution was added per liter of agaraftermelting but before autoclaving. After
auto-claving and cooling to 55°C, 7% sheep blood was
added, and 15 ml of SXT-BA was dispensed into plastic petriplates (100by15mm). SBAplateswere prepared in thesamemanner,butwithout the sulfa-methoxazole plus trimethoprim. Plates were stored under refrigeration for up to 1 week, but we have found themtobe stable foratleast2weeksifkeptin aplastic bag.
A total of696throat swabs (cotton tipped) from symptomatic children between the ages of 2 and 15 yearsoldwerecultured.Specimensweresubmittedin amodified Stuarts transport medium by physicians
locatedthroughoutWisconsin. Most specimenswere intransit only 1 day, butsome were2days old upon culture.
Each swabwasfirst inoculated upontwoSBAand then upon two SXT-BA plates which were then streaked for isolation.Several stabsweremade in both theprimary andsecondaryareasofstreakingto
pro-mote betahemolysis. Onesetofplates consisting of
anSBA andanSXT-BAplatewasincubatedat37°C
under 5 to 10% CO2 [SBA (CO2) and SXT-BA 189
on February 7, 2020 by guest
http://jcm.asm.org/
(CO2)1; the secondsetwasincubated at370C under ambient atmosphere [SBA (air) and SXT-BA (air)]. Thiswasdonetodetermine whether theC02 incuba-tion originally used (6) was necessary. Swabs were
thenused for the PP+B culture method.
It has been determined that by swabbingasingle
plate directly with the swab, approximately3to5% of theorganismspresentareremoved from the swab(2).
Since fourplateswereswabbed, approximately 12to
20% of the organisms had been removed before the swab was used in the PP+B method. Practically
speaking, performing thePP+B method last should have had a negligible effect on the detection and
quantitation ofgroup A streptococci by the PP+B method.
The PP+B method usedwas a modification
rec-ommendedtoourlaboratory byElaineUpdyke (then atthe Center forDisease Control, Atlanta, Ga.).We didnotuseeither preincubationof theswabs in broth asisnowrecommended forgreatersensitivity (4),or
surfacestreakingofplates.Swabswereplacedin 3 ml
of Todd Hewitt broth and mixed vigorously on a
Vortex mixer for5s.Next, asterile 6-inch (ca.
15.2-cm) applicator stick wasinserted into thebroth,
re-moved, and reinserted into a tube of melted veal
infusionagar (15ml/tube; BBL). Theagarhad been cooledto 550C and contained 7% defibrinated sheep blood.Approximately0.001 mlof inoculumwas trans-ferred from the broth to the melted agar.The agar tubewasthen mixedby inversion, pouredintoplastic petri plates, and allowedtosolidify.Pourplatesand correspondingbroth cultureswereincubated at37°C
overnightinroomair.
Allthroat swabs were cultured before nooneach
dayand observed forbeta-hemolytic streptococci-like coloniesby 8:30 a.m.thefollowing day. Plates were observedby usinga fluorescentlight sourcelocated behind theupheld plate.
All cultures positive for beta-hemolytic strepto-cocci-like colonieswerescreened forgroupA strepto-cocciby using astandardfluorescent-antibody (FA) technique (9).Smearsweremade fromthe broth cul-ture used in the PP+B method, asis usual for that
method, or by making a directsmear from isolated
colonieson oneof the surface streak plates (3). The
LancefieldprecipitintestwasusedtoconfirmourFA resultswhenever differenceswerenoted in the ability ofthe various methods to recover groupA
strepto-cocci.
In general, for all methods group A streptococci
colonieswerequantitatedonthe basisof thefollowing
categories: negative; <10;10to100;and >100 colonies perplate. However,if <10 coloniesperplatewereonly
visibleon asurface streaked plate but they occurred
in the area of the secondary streaking, the colony countwas recorded asbeing 10to 100 colonies per plate. Additionally, if1to100colonieswerevisibleon
the streakplatebuttheyoccurred in theareaof the
third or fourth cross streak, the colony count was
recordedasbeing>100perplate.Thispolicywasfelt tobe themostconsistent andrepresentativebut it did result in reporting higher colony counts than were actuallyobserved with the SBA(002) and SBA(air) methods.
The quantitativerecoveryratesarepresentedas a
rough comparison byusinganonstandardized
inocu-lum for each of the five media. However, preliminary studies in our laboratory have shown that when a
throat swabisplatedtothe surface of sixplates,there was no detectable difference in either the types of bacteria isolatedorintheir relative numbers.
RESULTS
Of the 696 throat swabscultured, 204 (29.3%)
werepositive forgroupAstreptococci byatleast
oneof the five methods used for eachspecimen.
Nosingle method recoveredallof the 204group
Astreptococci recovered by all five methodsas
awhole(Table 1). The SXT-BA (CO2) method
yielded the highest recovery rate (90.7%)
fol-lowedby the SXT-BA (air) (87.7%), the PP+B
(83.3%), the SBA (CO2) (79.4%), and the SBA
(air) (77%) methods.
Colony counts obtained by surface streak
plate methodsweregenerallyhigher than those
obtainedbythe PP+B method (Table 1). This
was to beexpected, however, from the
consid-erably larger inoculum size used in the former
methods and is an advantage of the surface
streakplate methods over the PP+B method.
It isimportantto notethat the colony
count-ing procedure described in Materials and
Meth-ods resulted in our reporting higher colony
countsby the SBA methods thanwereactually
visible. If, however, the actual number of
colo-nies presentare tabulated, the enhanced
sensi-tivity of the SXT-BA method is even more
TABLE 1. Recovery rates ofgroup Astreptococci from 696 throat cultures by five culture methods
Colony count of No. of positive throat culturesfor group A streptococci as detected by each culturemethod' groupA
strepto-cocci SXT-BA
(002)
SBA(002)
SXT-BA (air) SBA (air) PP+B<10 12 27 10 11 42
lOto100 28 12 30 14 27
>100 145 123 139 132 101
Totalpositive 185 (90.7) 162 (79.4) 179 (87.7) 157 (77) 170 (83.3)
cultures
aValues in parentheses indicate percent based on atotal of 204 cultures positive for group A streptococci as detectedby all five methods as a whole.
on February 7, 2020 by guest
http://jcm.asm.org/
EVALUATION OF SXT-BA PLATE 191
apparent:therewere77instances noted in which
the SXT-BA (002) method raised the colony
count by one or more quantitation categories
over that obtained by the SBA (002) method.
Similarly, therewere52such instances noted in
which theSXT-BA (air) method raised the
col-ony count by one or more quantitation
cate-gories over that obtained by the SBA (air)
method.
Table 2 shows thelevel of colonycountswhich
wentundetected byeach method (false-negative
cultures). Significantly, only about one-half of
the false-negative cultures which occurred for
theSXT-BA (002) andSXT-BA (air) methods
showed high colony counts (i.e., 10 to 100 or
>100 colonies perplate). In contrast,
approxi-mately 70% of the false-negative cultures which
occurred for the SBA (002), SBA (air), and
PP+Bmethodswereinthe highrange.Thatis,
theSXT-BA (002) and SXT-BA (air) methods
werethe least likely methodstoyield
false-neg-ativeresults when the colonycountswerehigh.
False-negative cultures where the colonycount
was <10 colonies perplate probably represent
chance variation expected when the swab
con-tains lownumbers ofgroup Astreptococci; we
have, therefore, excludedthem from
considera-tion.
Throat cultures whichyielded beta-hemolytic
streptococci-like coloniesthatprovednot to be
groupAstreptococciweredesignated
false-pos-itive cultures. The SXT-BA(002)andSXT-BA
(air) methods yieldedthe lowest rates of
false-positive cultures, 11% and 12.7% respectively
(Table 3). In contrast, for theSBA (CO2), SBA
(air), and PP+B methods, the rates of
false-positive cultures were 27%, 23.8%, and 22.7%,
respectively. Both the SXT-BA (002) and the
SXT-BA(air)methods reducedbyone-half the
number of unnecessary FA staining tests
per-formedonnon-groupAbeta-hemolytic
strepto-cocci-like colonies.
Beta-hemolyticstreptococci,groups
C,
F,andG,areinhibitedontheSXT-BAplatesandwere
the primary cause of the higher false-positive
rates on SBA and PP+B plates. In addition,
beta-hemolytic gram-negativerods and
beta-he-molytic staphylococci also wereresponsible for
significantnumbers offalse-positivecultures by
the PP+Bmethod.
Inaseparate study using the SXT-BA
(GO2)
method alone, wefound that the isolation rate
of group A streptococci was significantly
in-creased byincubating plates up to 48 h before
discarding asnegative. Of 2,536 SXT-BA
(GO2)
plates which were negative after overnight
in-cubation, an additional 77 isolates (3%) were
positive afteranadditional day of incubation. In
three-fourths of the cases, the colony counts
were >10 per plate.
DISCUSSION
It is clear that a positive throat culture for
group Astreptococciis not adefinite indication
of streptococcal pharyngitis even in
sympto-maticpatients(12). Inmost cases of
streptococ-cal pharyngitis, a positive throat culture
will
show colonycounts of >100 colonies per plate
by the direct surface streak plate technique.
TABLE 3. False-positive cultures obtained by each culture method
Method Total no. of false-positive culturesa
SXT-BA(002) 23 (11)
SBA(002) 60(27)
SXT-BA (air) 26(12.7)
SBA(air) 49(23.8)
PP+B 50(22.7)
aAfalse-positive culture is a culture which yielded
beta-hemolyticstreptococci-likecolonies thatproved
not tobe group Astreptococci.Valuesinparentheses indicatepercent,determinedby dividing the number of instances in whichagivenmethoddetected beta-hemolytic streptococci-likecolonies into the number of instances in which the isolates were foundto be otherthan group Astreptococci,and thenmultiplying
by100.
TABLE 2. Magnitude of colonycountofgroup Astreptococciwhichwereundetectedbyeach culture method Corresponding highest No. offalse-negativeculturesa
colonycountofgroup Astreptococci
ob-servedbyapositive SXT-BA(GO2) SBA(002) SXT-BA(air) SBA (air) PP+B method
<10 10 12 11 12 10
lOto100 5 17 8 21 20
>100 4 13 6 14 4
Total no. of false- 19(9.3) 42 (20.6) 25(12.3) 47 (23) 34 (16.7)
negative cultures
aAfalse-negative culture is onewhichisnegative for group A
streptococci
by
the methodindicated butwhichwaspositivebyatleastoneoftheremaining4methods. Values inparenthesesindicatepercent.
9,
on February 7, 2020 by guest
http://jcm.asm.org/
192
However, occasionally the same colony counts
maybe found incultures from carriers of group
A streptococci (12). Practically speaking,
how-ever, ifapositivecultureis fromasymptomatic
patient,mostphysicianswill treat,regardless of
theamountofgroupAstreptococcifound.
The use of sulfamethoxazole plus
trimetho-prim in throat culture media to reduce the
in-hibitoryeffect of normal throat florabacteriaon
the growth and/or beta hemolysis ofgroup A streptococci wasexcellently discussed byGunn
et al. (6). This selectivity of SXT-BA platesis
the reason for their superiority over the SBA
plate. The superiority of the SXT-BA method
over the PP+B method that we found in this
studyisdue both to theselectivityofthe
SXT-BA method and to the larger inoculum size it
employs. The minimal inoculum size
recom-mended for the PP+B method is based upon
two assumptions. First, that use ofa small
in-oculum size will prevent normal flora bacteria
fromovergrowing and,thereby,frompreventing
detection of group A streptococci which can
occurbyusingaheavier inoculum size. Second,
that colonycountswillalways be highincases
of streptococcal pharyngitis. From ourown
ex-perience with the PP+B method, the first
as-sumptionappearstobe correct.Inmanythroat
culturesthere is averyheavygrowth of
alpha-hemolytic streptococci, facultative
gram-nega-tive rods, and Staphylococcus aureus. These
organisms oftenovergrowandpreventdetection
of the group A streptococci unless a smaller
inoculum size is used, asin the PP+B method
(10). The second assumption, however,
contra-dictsthe previousreportofKaplanetal.(7)who
found that one-third of thepatients with
phar-yngitis had lessthan 10colonies perplate, but
hadaserologicalresponseindicative of infection.
Ourdata indicate that the PP+B method with
its lowinoculum size isobviouslytooinsensitive
to detect these cases. In fact, Pike (10) has
calculated that there would have to be 4,000
colony-forming units ofapure culture of
strep-tococci permlofbroth before therewas a 98%
probability that a 0.001-ml portion contained
streptococci. Preincubation of swabs for 2 h in
broth followed by plating has been shown to
yieldup to 17%morepositive culturesthancan
be obtained by immediate plating (9). We did
notusethisenrichmenttechnique, sincewefeel
thatit istoo expensiveandimpracticalfor
rou-tine useandthat itmayyield misleading
infor-mation by failing to differentiate between the carrier and the infected state (11). This differ-entiation might ultimately be made by direct,
selectivequantitation ofgroupAstreptococciin
throat cultures, such as by using the SXT-BA
(GO2)
method. In this manner, it may bepossibletoshowacloser correlationbetween
high
colonycountsanda positive serologicalresponse than
observed
by Kaplan
etal.,
who usedanonselec-tive
plating
method.However,Gordisetal.(5)found in their
study
that one-third of thepatientswhodevelop
rheu-maticfever had beenrecently cultured and had
negative results. The negative cultures were
probablyduetovariables suchaspoorcollection
technique,
inadequate survival and/orover-growthintransportmedia,andpoor culture and
isolation techniques.Thisindicates that
micro-biologists should do all they can to recover as
many group A streptococcias
possible.
Although
the 11.4% greater recovery rate ofgroup Astreptococciobtained with the SXT-BA
(C02) method as compared to the SBA (GO2)
wassignificant, itwas not asgreat as the28%to
42% valuespreviously reported (6). The reason
forthese differences may be related to the fact
that ourthroat cultures wereall fromchildren,
whereas60% of thepatients studied by Gunnet
al. were adults. It ispossible there may be
dif-ferences between the concentrations of normal
flora and group A streptococci
colony-forming
units found in infected children and infected
adults. Morerecently,Gunn hasfound that the
differences in his isolation rates, using SXT-BA
(GO2) and SBA (GO2) plates, are now running
between 11 to 20%-much closer to what we
encountered in this study (personal
communi-cation).
Anapparentparadox of the SXT-BA method
isthat92% ofgrou,pAstreptococci, when tested
in thymidine-free media, cannot grow in the
presence of the 25 ,ig of
sulfamethoxazole-tri-methoprimper ml used intheSXT-BAmedium
(1). However, the Trypticase soy agar used for
the SXT-BA plates contains thymidine, which
group A streptococci are able to utilize,
circum-venting the metabolic inhibitory effect of 25 ig
ofsulfamethoxazole-trimethoprim per ml.
Viri-dansstreptococci and other normal throat flora
bacteria are less able to use the thymidine and
are thusgenerally inhibited (S. R. M. Bushby,
personalcommunication).
Unfortunately, the precise amount of
thymi-dine required to achieve the optimal selective
effect is unknown. Various commercial media
havebeen shown to varymarkedly in thymidine
content (8), and only Trypticase soy agar has
beenproven to be reliable for theSXT-BA
(O2)
method (6). Of some concern is the finding that
various lots of Trypticase soy broth also vary
considerably in thymidine content (S. R. M.
Bushby, personal communication). Our data
showed that whenincubation was limited to 24
on February 7, 2020 by guest
http://jcm.asm.org/
EVALUATION OF SXT-BA PLATE 193
h, in 9 instances the SXT-BA (CO2) method
failedtodetect group A streptococci in
quanti-ties of -10 to 100 colonies per plate, a
false-negative rate of 4.4%.Gunn et al. (6) reported a
false-negativerate of 1.4%(7/508) for the
SXT-BA(CO2)method. Hisvalue is considerably less
than what we observed and may indeed be due
to differences in thymidinecontent among
dif-ferent lots ofcommercial Trypticase soy agar
used.
Wefound,however, that incubating our
SXT-BA (CO2) plates for up to 48 h resulted in an
additional 3% recovery of group A streptococci. The 1.4% false-negative rate we could expect,
then, after 2 days of incubation is identical to
that reported by Gunn et al. using 24 h of
incubation of SXT-BA (CO2) plates.Thus,
pro-longing the incubation of SXT-BA (CO2) plates
to 48h mayreducethe effects of variable
thy-midinecontentin lots of Trypticasesoy agar. In
viewof thealready discussedimportance of
us-ing as sensitive a culture method as
possible,
microbiologists may wish to include an SBA
(CO2) platetodetectanadditional 1.4% positive
cultures.
In our experience, use of Streptococcus
py-ogenes ATCC 19615
(Bactrol,
DifcoLaborato-ries)to serve as apositive control for growthand
for beta
hemolysis,
ashas beensuggested (6),
issatisfactory. It may be
possible
anddesirable,
however,to
accurately
establish theamount ofthymidinenecessary toaddto a
thymidine-free
basal mediumtomaximize inhibition of normal
throat floraby
sulfamethoxazole-trimethoprim,
withoutinhibitinggrowth ofany group A
strep-tococci.
Ourlaboratoryprocessesanaverageof
37,000
throatculturesperyear.By
using
theSXT-BA(CO2) method instead of thePP+B method in
ourlaboratory, we have calculated aminimum
savings of $5,000peryear. This
savings
isbasedupon elimination of the
accompanying
brothculture andextramaterials neededinthe PP+B
methodandupon
saving
anaverage ofatleast8 sof worktimeperculture.TheSXT-BA
(CO2)
method is also
considerably
less tedious andrequires less hand
manipulations
than the PP+Bmethod.
Additionally,
wefindthatsmearsmadedirectly
from colonies onSXT-BA(CO2) plates
for FA
staining
generally
showgreater numbersoffluorescing
cells,
and individual cell walls stainmore
uniformly
positive
than with the PP+Bmethod (3). Based upon the substantial
eco-nomic and logistic advantages and upon the
superior sensitivity for detection of group A
streptococci,we recommendthe SXT-BA(GO2)
throat culturemethod of Gunn et al. (6).
Hope-fully, commercial media producers will offer
SXT-BA plates for laboratories unable to
pre-paretheirown.
ACKNOWLEDGMENTS
We thank Phil Wand for usefulsuggestions, Dennis Maki and ArlanHelstad for reviewing the manuscript, and Marcia Kirkegaard for technical assistance.
LITERATURE CITED
1. Bushby, S. R. M. 1973. Trimethoprim-sulfamethoxazole: in vitro microbiological aspects. J. Infect. Dis. 128(Suppl.):S442-S462.
2. Collee, J. G., B. Watt, R. Brown, and S. Johnstone. 1974.The recovery of anaerobic bacteria from swabs. J. Hyg. 72:339-347.
3. Ederer, G. M., and S. S. Chapman. 1972. Simplified fluorescent-antibodystaining method for primary plate isolates of group Astreptococci. Appl. Microbiol. 24: 160-161.
4. Facklam, R. R. 1974. Streptococci, p. 96. In E. H. Len-nette, E. H.Spaulding, and J. P. Truant (ed.), Manual ofclinicalmicrobiology,2nd ed. AmericanSocietyfor Microbiology, Washington, D.C.
5. Gordis, L.,A.Lilienfeld,and R.Rodriguez. 1969. An evaluation of the Maryland rheumatic fever registry. Public HealthRep. 84:333-339.
6. Gunn,B.A., D. K. Ohashi, C. A. Gaydos, and E. S. Holt. 1977. Selective and enhanced recovery of group Aand B streptococci from throat cultures with sheep blood agarcontaining sulfamethoxazole and trimetho-prim. J. Clin. Microbiol. 5:650-655.
7. Kaplan, E. L, F. H. Top, Jr., B. A. Dudding, and L. W. Wannamaker. 1971. Diagnosis of streptococcal pharyngitis: differentiation of active infection from the carrierstate inthesymptomatic child. J. Infect. Dis. 123:490-501.
8. Koch,A.E.,and J. J. Burchall.1971.Reversal of the antimicrobialactivity of trimethoprim by thymidine in commerciallypreparedmedia.Appl. Microbiol. 22:812-817.
8a.Kurzynski, T.,C.Meise,R.Daggs,and A.Helstad. 1979.Improved reliabilityof theprimary plate bacitra-cintest on throatcultures with sulfamethoxazole-tri-methoprimbloodagarplates.J.Clin. Microbiol. 9:144-146.
9. Moody, M. D., A. C. Siegel, B. Pittman, and C. C. Winter. 1963. Fluorescent-antibody identification of group Astreptococci from throat swabs.Am.J. Public Health53:1083-1092.
10. Pike, R.M. 1945.The isolation ofhemolytic streptococci
fromthroat swabs:experimentswith sodium azideand crystalvioletinenrichment broth.Am.J.Hyg. 41:211-220.
11. Stollerman, G.H. 1962. Therole of the selectivethroat culture for betahemolyticstreptococciinthediagnosis
ofacutepharyngitis.Am. J. Clin. Pathol. 37:36-40. 12. Wannamaker, L. W. 1972.Perplexity andprecisionin
thediagnosisofstreptococcalpharyngitis.Am.J. Dis. Child. 124:352-358.