J. Med. Microbiol. - Vol. 33 (1990), 121-126
0022-261 5/90/0033-O121/%
10.00
01990 The Pathological Society of Great Britain and Ireland
The use of selective media for the isolation of
Pseudomonas pseudomallei i n cl i nicaI practice
V. WUTHIEKANUN*, D. A. 6. DANCE*t, Y. WATTANAGOON", Y. SUPPUTTAMONGKOL",
W. CHAOWAGULS and N. J. WHITE*t
Bangkok Hospitalfor TropicalDiseases, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand,
t Nufield Department of Clinical Medicine, University of Oxford, Oxford and $Department of Medicine,
SappasitprasongHospital, Ubon Ratchatani, Thailand
Summary. Ashdown's selective-differential agar medium, with or without preenrichment in selective broth, was evaluated for the isolation of Pseudomonas
pseudomallei from 1972 clinical specimens obtained from 643 subjects in Northeast
Thailand; 226 patients proved to have melioidosis. The use of Ashdown's medium
significantly increased the frequency of recovery of P . pseudomallei from sites or
specimens with an extensive normal flora (throat, rectum, wounds and sputum) as
compared to the recovery on blood and MacConkey agars (p <0-01). The isolation
frequency from throat, rectal and wound swabs was further increased by the use of
the broth pre-enrichment. The colonial morphology of P . pseudomallei on Ashdown's
medium was sufficiently characteristic to allow presumptive identification. With the
use of these selective media it was possible to culture P . pseudomallei from throat
swabs taken from 87% of the patients from whom the organism could also be isolated
from corresponding tracheal aspirates or sputum specimens. P . pseudomallei was
isolated from rectal swabs taken from 51 patients, the first time that faecal excretion
of the organism has been demonstrated in man. The diagnosis of melioidosis would
not have been confirmed bacteriologically in eight patients (3.5%) without the use of
the selective media. It is suggested that, in areas endemic for melioidosis, all sputum
specimens should be cultured on selective media, such as Ashdown's. For the
investigation of clinically suspected cases of melioidosis, and for follow-up during
treatment of the disease, the use of broth pre-enrichment is recommended for
specimens obtained from sites with an extensive normal flora.
selective and differential agar medium first described by A ~ h d o w nand
, ~ on a broth pre-enrichment technique, for the isolation of P . pseudomallei
from clinical specimens other than blood and bone
marrow. These studies were conducted in conjunction with a prospective clinical study of melioidosis
in Ubon Ratchatani, Northeast Thailand, conducted from September 1986 to November 1989.'
Introduction
During the past decade, melioidosis has been
recognised as an important cause of morbidity and
mortality in Northeast Thailand.'. Isolation of the
causative organism, Pseudomonas pseudomallei,
remains the cornerstone of diagnosis.2Although P .
pseudomallei grows on most routine laboratory
media,3. overgrowth by more-rapidly-growing
commensal bacteria and a lack of familiarity with
the cultural features of the organism both contribute Materials and methods
to underdiagnosis of the d i ~ e a s eSeveral
.~
different
selective and differential media have been devel- Pa tients
oped to facilitate the isolation and identification of
Clinical specimens were obtained from three groups of
P.p~eudomallei.~-~
We report here on the use of the individuals in Ubon Ratchatani Province, 600 km northReceived 5 Feb. 1990; accepted 18 April 1990.
Correspondence and requests for offprints should be sent to Dr
D. A. B. Dance, (present address) Department of Clinical
Sciences, London School of Hygiene and Tropical Medicine,
Keppel Street, London WClE 7HT.
east of Bangkok: (1) 464 patients in Sappasitprasong
Hospital (the provincial hospital) in whom melioidosis
was suspected on clinical grounds, or from whom P .
pseudomallei had been isolated in the hospital laboratory ;
(2) 103 in-patients at Sappasitprasong Hospital who had
indirect haemagglutination (IHA) titres to P.pseudomaZlei
Downloaded from www.microbiologyresearch.org
by
121
IP: 88.99.165.207
On: Sat, 17 Jun 2017 06:10:51
122
V. WUTHIEKANUN ET AL.
antigens of 2640, but from whom P . pseudomallei had
not been isolated previously; (3) 76 healthy controls from
rice-farming communities in Northeast Thailand-these
individuals were either relatives of culture-positive
melioidosis patients (19) or were blood-donors.
Of these individuals, 212 in Group 1, 14 in Group 2
and none in Group 3 were diagnosed as culture-positive
cases of melioidosis during the study period.
Specimens
Wherever possible, the following clinical specimens
were collected from subjects in Groups 1 and 2: throat
and rectal swabs, urine, sputum or tracheal aspirates,
swabs from wounds or skin lesions, pus or aspirates from
abscesses and blood. Throat swabs only were taken from
subjects in Group 3. Other sites or lesions were sampled
as indicated clinically. Sites from which P . pseudomallei
was isolated were re-cultured weekly or at out-patient
follow-up visits.
Specimens were transported to the laboratory and
cultured immediately, or stored in a refrigerator for up to
12 h when immediate culture was not possible. Swabs
were held in transport medium (‘Transwabs’, Medical
Wire and Equipment Co. Ltd, Corsham) before inoculation.
previously.* The first isolate of P . pseudomallei from any
site for each patient was confirmed by the API 20NE kit
system of biochemical tests.*
Media
COL, HBA and MAC were prepared according to the
manufacturers’ instructions. To prepare ASH the following ingredients were added to each litre of distilled water:
Trypticase Soy Broth (BBL, Cockeysville, USA) 10 g;
agar (Eiken Chemical Co. Ltd, Japan, or Oxoid) 15 g;
glycerol 40 ml; aqueous crystal violet 0.1% w/v, 5 ml; and
aqueous neutral red 1% w/v, 5 ml. This mixture was
autoclaved at 121°C for 15 min and gentamicin was
added aseptically to a final concentration of 5 mg/l. The
medium was distributed in 20-25-m1 amounts into 9-cm
diameter petri dishes. SB contained (per litre of distilled
water): Trypticase Soy Broth 10 g; glycerol 40 ml; and
aqueous crystal violet 0.1% w/v, 5 ml. This mixture was
autoclaved at 121°C for 15 min. Colistin sulphomethate
was added aseptically to a final concentration of 15 000
U/L before 10-mlvolumes were dispensed into sterile 27ml containers. Agar plates were stored in a refrigerator
and used within 2 weeks of pouring. SB was stored at 4°C
for up to 1 month before inoculation.
Stat ist ical analyses
Laboratory methods
The study took place in three stages during the rainy
season months (June-November) of successive years.
During stage 1, lasting from September 1986 until
September 1987, the specimens were cultured on Columbia Agar (Oxoid) containing horse blood 5% (HBA), and
on Ashdown’s Medium (ASH). During stage 2, from
October 1987 to November 1988, the specimens were
cultured on HBA, ASH, and, except for urine samples,
inoculated into Selective Broth (SB). During stage 3, from
June until November 1989, the specimens were cultured
on HBA, ASH, MacConkey Agar (Oxoid, CM7) (MAC)
and, except for urine samples, inoculated into SB.
Occasional deviations from this protocol occurred owing
to the unavailability of media. When HBA was not
available, unsupplemented Columbia Agar (COL) was
used. Several of the specimens from Group 3 were
cultured on ASH and SB only. Most rectal swabs were
not cultured on HBA or COL, when MAC was being
used. For statistical analyses, results from the three stages
were combined.
Culture media were inoculated in the order: HBA (or
COL), MAC, ASH, SB. Agar plates were incubated in
air at 37°C for 4 days and were inspected daily. SB was
incubated at 37°C for 48 h, and then subcultured on to
quarter plates of ASH, which were incubated for a further
48 h.
The identity of suspected P . pseudomallei isolates was
confirmed by screening for morphology in gram-stained
smears, oxidase reaction, gentamicin- and colistanresistance, and colonial morphologyon ASH, as described
The sensitivity of each medium was defined as the
number of specimens yielding growth of P.pseudomallei
in that medium divided by the total number of culturepositive specimens cultured on that medium. Culturepositive, in this context, was defined as yielding growth
of P . pseudomallei on any medium. The results of culture
for P . pseudomallei from culture-positive specimens on
pairs of media were compared by the McNemar test or
the binomial test as cited by Siegel’, depending on the
number of pairs with dissimilar results.
Results
A total of 1972 clinical specimens was examined
during the study. The numbers of each type of
specimen and the media on which these were
cultured are shown in table I. P. pseudomallei was
isolated from 489 (24.8%) of the specimens. P .
pseudomallei was the only pathogen isolated from
389 specimens; the remainder yielded additional
organisms considered to be potentially pathogenic.
The other pathogens most frequently isolated were
Staphylococcus aureus, enterobacteria and other
pseudomonads.
The frequencies with which P . pseudomallei was
isolated from each specimen type on the various
media are compared in table 11. ASH was significantly more likely to yield a positive culture than
HBA or MAC for throat swabs (sensitivities 64%,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 06:10:51
123
SELECTIVE MEDIA FOR P. PSEUDOMALLEI
Table I. Specimens cultured for P . pseudomallei
I
,
~
Specimen type
Throat swab
Urine
Rectal swab
Sputum
Wound swab
Pus or aspirate
Tracheal aspirate
Pleural fluid
Eye swab
Synovial fluid
Peritoneal fluid
CSF
Vaginal swab
Pericardial fluid
Bile
Number of specimens cultured for
P.pseudomallei on
Total
(each
COL* HBA MAC ASH SB site)
1
33
32
30
5
13
2
6
0
0
0
0
0
426
377
134
216
177
72
42
21
13
5
4
3
214
186
193
153
122
35
27
12
13
4
4
3
f
Total
122
1495
524
416
403
219
187
74
47
20
13
5
4
3
f
970
450
10
346
204
170
62
49
19
13
6
3
4
531
421
413
227
196
76
54
21
13
6
4
4
f
1921 1341 1972
* COL,
Columbia agar; HBA, Horse-blood agar; MAC,
MacConkey agar; ASH, Ashdown’s agar; SB, Selective broth.
8% and 23% respectively; p < 0.001), rectal swabs
(sensitivities 49%, 0% and 0% respectively; p <
0-OOl), sputum (sensitivities 97%, 49% and 67%
respectively;p c 0-OOl),and wound swabs (sensitivities 73%, 53% and 54%; p<O-OOl and 0-01
respectively). Pre-incubation in the SB followed by
subculture on ASH produced a further significant
increase in the isolation rate as compared to direct
culture on ASH for throat swabs, rectal swabs and
wound swabs (sensitivities 99%, 98% and 100%
respectively with pre-enrichment ; p < 0.001 for
each specimen type) but not for sputum (sensitivity
9973, and was more likely than each solid medium
to yield P . pseudomallei from eye swabs (cf. HBA,
p=0.016; cf. MAC and ASH, p=0-031). Other
differences in isolation rates were not significant,
although the numbers of several specimen types
were too small to allow valid comparison.
Three specimens (one throat swab, one rectal
swab and one sputum) yielded P.pseudomallei after
direct plating on ASH but gave negative cultures
after SB enrichment. In two of these cases, heavy
growth of other gram-negative bacilli occurred.
This probably suppressed the growth of P.pseudomallei in SB. In the third, only one colony of P .
pseudomallei grew on ASH, and the discrepancy
was probably accounted for by sampling error.
From five specimens (two urines, two wound swabs
and one eye swab) P . pseudomallei failed to grow on
ASH, but was isolated on HBA (three specimens)
or MAC (two specimens). In all these cases, less
than 10 colonies were present on the plates yielding
growth and the discrepancy was again attributed to
sampling error.
Presumptive identification of P . pseudomallei on
the basis of colonialmorphology was rarely possible
during the first 24 h of incubation, unless the
organism was present in heavy, pure culture. By
48 h, the typical dry, wrinkled appearance and
violet-to-purple colour of P . pseudomallei were
usually apparent on ASH (fig. l), and in most cases
Table 11. Specimens positive for P . pseudomallei
Number of specimens positive/
number of positive specimens cultured on
Specimen type
COL*
HBA
MAC
ASH
15/66
18/20
0138
35/52
39/72
14/16
7/10
616
419
213
111
111
1421294
Throat swab
Urine
Rectal swab
Sputum
Wound swab
Pus or aspirate
Tracheal aspirate
Pleural fluid
Eye swab
Synovial fluid
CSF
Pericardial fluid
414
111
111
010
010
010
010
010
9/ 109
37/44
0114
39/79
531100
26/31
9/15
8111
319
314
111
111
Total
9/13
1891418
215
112
010
010
Total
SB
site)
741115
47/49
2515 1
75/77
741102
3 1/32
12/16
11111
419
314
111
111
108/109
010
52/53
7 1/72
1001100
29/29
19/19
10/10
919
515
111
111
118
50
54
80
106
33
21
11
9
5
1
1
3581468
4051408
489
~~~~~
* Cultures of peritoneal fluid, vaginal swabs and bile all gave negative results.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 06:10:51
124
V. WUTHIEKANUN ET AL.
Fig. 1. Appearance of P. pseudomallei on ASH, showing the
characteristic wrinkled colonies. The colonies take up dyes from
the medium to assume various shades of violet-purple.
a tough, dry pellicle with a matt surface was present
on the top of positive SB cultures (fig. 2). On the
basis of these findings the laboratory was able to
alert the physicians to the likely diagnosis.
Throat swabs were collected within 24 h of
positive sputum or tracheal aspirates on 46 occasions; P . pseudomallei was isolated from 40 (87%)
of these swabs. P . pseudomallei was also isolated
from 54 rectal swabs from 51 patients-36 with
septicaemic melioidosis, 14 with localised pulmonary disease and one with a parotid abscess; eight
of these patients, six of whom were septicaemic
had a history of diarrhoea. Although 25 rectal swab
cultures (49%) yielded P . pseudomallei by dirrect
plating on ASH, the growth was always confined to
the presence of a few colonies in the area of primary
inoculum. In no patient was a rectal swab the only
positive specimen. During the study seven patients
with culture-positive pulmonary melioidosis and
one with a soft tissue infection were found who
would not have been diagnosed bacteriologically
without the use of the selective media. These
represented 3.5% of all the culture-positive cases
examined. In one case, the scanty growth of P .
pseudomallei on ASH amongst a heavy, mixed flora
from urine was regarded as probable contamination, since it could not be repeated and the patient's
symptoms were referable only to the respiratory
tract.
Discussion
The selective media described for the isolation
of P . pseudomallei from clinical specimens rely
either on the organism's intrinsic antibiotic
resistance4' or its ability to use a specific carbon
and energy source.6 Ashdown's medium (ASH) is
particularly useful since, in addition to being
selective, it is also differential. The glycerol and
dyes in the medium give rise to highly characteristic
dry, wrinkled, purple colonies. In an earlier report,
we exploited this distinctive colonial morphology
in a rapid screening system for the presumptive
identification of P . pseudomallei.* Extended experience with over 1000 isolates of the organism has
confirmed that the morphology of the organism on
ASH alone always allows reliable identification
(data not shown), although experience is necessary
to distinguish occasional wrinkled isolates of P .
cepacia.8 The ingredients of the medium are
inexpensive and widely available.
The results of this study showed that use of ASH
significantly increased the rate of isolation of P .
pseudomallei from specimens with a normal bacteha1 flora, and that the isolation rate could be further
Fig. 2. P . pseudomallei growing in SB, with a typical cohesive,
dry wrinkled pellicle. This appearance is not invariable, but the increased by broth pre-enrichment' This increased
matt surface of the pellicle is a useful early indication of the diagnostic sensitivity was evident particularly with
presence of P.pseudomallei in a sample.
specimens taken during appropriate antibiotic
5 9 7
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 06:10:51
SELECTIVE MEDIA FOR P . PSEUDOMALLEI
125
treatment of melioidosis. Nevertheless, relatively bowel remains to be determined. Diarrhoea is a
few extra cases of melioidosis were diagnosed in well recognised feature of septicaemic melioidosis'
this study through the use of the selective media. and was present in eight of our rectal-swab-positive
This may be because the patients seen in a regional cases, six of whom were septicaemic and two of
referral centre such as Sappasitprasong Hospital whom had localised pulmonary infections. Diarare those with the most severe manifestations of rhoea may also be a non-specific feature of febrile
melioidosis. Approximately 60% of the cases seen illness, and in none of our patients was P.
were septicaemic, and, therefore, could be diag- pseudomallei recovered in large numbers from the
nosed easily by blood culture." It is likely that the rectal swabs. The presence of the organism in the
selective media discussed here would provide a gut may be of greater significance as a means of
greater advantage in the recognition of earlier and dissemination or by providing opportunities for the
milder cases of the disease. Furthermore, all the acquisition of antibiotic resistance.
In this study, the patients in whom the throatcultures studied here were examined by a team
specifically investigating melioidosis. Such focused and rectal-swab cultures were positive were sufferattention would be unusual in routine laboratory ing from either septicaemic or respiratory tract
practice. In several cases, a few colonies of P . melioidosis. We also saw several patients with
pseudomallei growing on 'non-selective' media melioidosis in deep visceral sites, including liver or
would probably have been overlooked had the splenic abscesses. Throat and rectal cultures freorganism not been identified in the corresponding quently were negative in these individuals, and two
patients in Group 2 were subsequently proved to
selective cultures.
Throat-swab cultures are not usually used in the have culture-positive melioidosis despite initially
diagnosis of melioidosis, except in the recently negative cultures on screening. Blood cultures, or
the aspiration and culture of pus remain the
described pharyngocervical form of the disease.
Of 91 throat-swab-positive patients in this study, diagnostic methods of choice in such patients.
Pre-incubation in SB followed by subculture on
only one was suffering from the latter condition.
The remainder had other forms of melioidosis ASH was the most sensitive laboratory technique
localised to the lungs (24 cases), parotid gland (five for isolation of P . pseudomallei examined in this
cases), eyes (two cases), or were septicaemic (59 study (> 99% sensitivity compared to all methods
cases). It is impossible to say whether recovery of combined). Nevertheless the method did fail to
the organism from a throat swab reflects colonisa- isolate P . pseudomallei from three specimens which
tion of the throat prior to clinical infection or yielded the organism on other media. Since oversecondary spread from the bloodstream or an growth by other gram-negative organisms appeared
adjacent site. Nevertheless, the strong concordance to contribute to this problem, it is possible that a
between a positive throat culture and a positive solution may lie in increasing the selectivity of the
sputum or tracheal aspirate suggeststhat the culture enrichment broth, for example by increasing the
of throat swabs on the selective media may be concentration of colistin or adding another broaduseful for the investigation of patients in whom spectrum agent such as amikacin. P . pseudomallei
lower respiratory tract melioidosis is suspected but is resistant to considerably higher concentrations
who are unable to produce sputum (e.g., children). of colistin than the 15 000 U/L present in SB."
It is also possible that P . pseudomallei, as a widely Although this concentration was effective in prelimdistributed environmental saprophyte, may form inary experiments with seeded stool and clinical
part of the transient or permanent normal upper specimens (data not shown), it may not be optimal.
Although melioidosis is attracting increasing
respiratory tract flora in some normal individuals
in endemic areas. Contamination of specimens in interest in countries in South East Asia, it is
clinical or laboratory areas is another possibility. undoubtedly still underdiagnosed. We believe that
Although isolation of the organism was considered selective media such as ASH should be used
compatible with a clinical diagnosis of melioidosis routinely for the bacteriological examination of
in all but one of our cases, further studies on the sputum specimens within endemic areas, as this
healthy population of endemic areas are necessary would help to compensate for clinical underto determine the specificity of positive throat recognition. Enrichment culture should also be
considered for the investigation of clinically-suscultures for the diagnosis of the clinical disease.
P . pseudomallei has not been isolated previously pected cases of the disease, and for follow-up
from faeces or rectal swabs from man, although it cultures during treatment. In the future, the use of
has been detected in the faeces of animals6 these selective techniques outside the areas generWhether the organism is directly pathogenic to the ally recognised as being endemic for P . pseudomallei
''
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 06:10:51
126
V. WUTHIEKANUN ET AL.
will be valuable in helping to define the worldwide
epidemiology of melioidosis. However, even with
selective media the time taken to make a bacteriological diagnosis of melioidosis is seldom less than
48 h and is a problem in life-threatening disease, as
this tends to delay urgent therapeutic decisions.
New, sensitive, specific and rapid diagnostic techniques, such as the direct detection of P.pseudomallei antigens or nucleic acids in clinical specimens, are, therefore, still urgently needed.
We thank the Directors of Sappasitprasong Hospital for their
support of this study; Drs T. M. E. Davis, S. Looareesuwan, W.
Supanaranond, S. Pukrittayakamee, B. Sakarin, and the medical
and nursing staff of the hospital for their help in collecting
specimens; Mrs P. Naigowit, Miss N. Teerawattanasook, Mrs
Sukim Piriyakitphaiboon and Mr Sayan Langla for technical
assistance; and Dr L. Ashdown for his helpful advice. The
support and encouragement of Dr J. B. Selkon and the Oxford
Regional Public Health Laboratory, Professor D. Bunnag and
Professor T. Harinasuta are also gratefully acknowledged. This
study was part of the Wellcome-Mahidol University, Oxford
Tropical Medicine Research Programme, funded by the Wellcome Trust.
REFERENCES
1. Chaowagul W, White NJ, Dance DAB et al. Melioidosis: a
major cause of community-acquired septicemia in
Northeastern Thailand. J Infect Dis 1989; 159: 890899.
2. Leelarasamee A, Bovornkitti S. Melioidosis : review and
update. Rev Infect Dis 1989; 11: 413-425.
3. Sanford JP. Pseudomonasspecies (including melioidosis and
glanders). In: Mandell GL, Douglas RG, Bennett JE
(eds) Principles and Practice of Infectious Diseases,
2nd edn. New York, John Wiley & Sons. 1985: 12501254.
4. Ashdown LR. An improved screening technique for
isolation of Pseudomonas pseudomallei from clinical
specimens. Pathology 1979;11: 293-297.
5. Farkas-Himsley H. Selection and rapid identification of
Pseudomonas pseudomallei from other gram-negative
bacteria. Am J Clin Patholl968; 49: 850-856.
6. Galimand M, Dodin A. Le point sur la mtlioidose dans le
monde. Bull Soc Pathol Exot Filiales 1982;75: 375-383.
7. Khanthawa B, Tharavichitkul P. Selective medium for
8.
9.
10.
11.
12.
Pseudomonas pseudomallei. In : Punyagupta S, Sirisanthana T, Stapatayavong B (eds) Melioidosis. Bangkok,
Bangkok Medical Publisher. 1989: 154-157.
Dance DAB, Wuthiekanun V, Naigowit P, White NJ.
Identification of Pseudomonas pseudomallei in clinical
practice: use of simple screening tests and the API
20NE. J Clin Patholl989; 42: 645-648.
Siege1S. The case of two related samples. In :Nonparametric
statistics for the behavioural sciences. New York,
McGraw-Hill. 1956: 61-94.
White NJ, Dance DAB. Clinical and laboratory studies of
malaria and melioidosis. Trans R Soc Trop Med Hyg
1988; 82: 15-20.
Silpapojakul K, Ratanachai S, Mutirakul P, Silpapojakul
K, Kalnaowakul S. Melioidosis in Southern Thailand :
111. Pharyngocervical melioidosis. In : Punyagupta S,
Sirisanthana T, Stapatayavong B (eds) Melioidosis.
Bangkok, Bangkok Medical Publisher. 1989 : 80-84.
Dance DAB, Wuthiekanun V, Chaowagul W, White NJ.
The antimicrobial susceptibility of Pseudomonas pseudomallei. Emergence of resistance in vitro and during
treatment. J Antimicrob Chemother 1989; 24: 295-309.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 06:10:51