RF-TB-1.18.pdf

Media

extracted text
ht cCt
icteriologiulosis and
Bull. Int.

epidemioducted in
.7-71,
wski S. et
uberculoi. Int. J.
IUATLD
■sis Proung Dis.,

.’acional de
le Salud,
eventiva.
.esurgent
m immuand the
ims. Am.

A. et al.
risk of
Jg users
is infec-

own and
oir. Dis.,
988) The
berculo■ammes.
21-4.
91) The
innode\dvances
s D.M.
iburgh.
A.B. et
nmunoIntern.

I. (1991)
due to
oatients
27-30.
iak, A.
k

the microbiology of
tuberculosis

At <■ <A,

3

P.A. Jenkins

3.1 INTRODUCTION

AAFB. This is generally shortened to 'acid­
fast bacilli' or AFB.
I he mycobacteria that cause classical tuber­
Acid-fastness is thought to be related to the
culosis are
uniquely thick cell wall, which is composed
of an interlacing layer of lipids, peptidogly­
• Mycobacterium tuberculosis
cans and arabinomannans. The aniline dye
• Asian variant
forms a complex with this layer and is held
• African I
fast despite the action of the acid-alcohol.
• African II
This
allows the detection of AFB in speci­
• Mycobacterium bovis
mens with a simple staining technique - the
M. tuberculosis and M. bovis are probably the Ziehl-Neelsen (ZN) stain, which has been in
same species and are more accurately use for over 60 years.
The uniquely thick cell wall also gives
referred to as M. tuberculosis var. hominis and
mycobacteria
another useful characteristic,
var. bovinus respectively. However it is more
which
is
their
resistance to the lethal effects
convenient to refer to them as above. M.
of
acids,
alkalis
and detergents. This is
ufricanum[l] and the Asian variants differ
fortunate
because
most
specimens are heav­
from the typical A4. tuberculosis in minor
ily
contaminated
with
other
bacteria, which
cultural and biochemical ways only. These
grow more quickly than mycobacteria and
distinctions are of purely epidemiological
would thus swamp the culture. It is therefore
interest and have no bearing on management
possible to 'treat' the specimen by adding an
or prognosis. In the UK, M. bovis disease is
acid or alkali, thus killing the other bacteria
largely reactivation of infection acquired
leaving the mycobacteria alive. Ultimately the
many years earlier[2]. There is normally little
mycobacteria will also be killed and a balance
need for contact tracing and patients are
has to be struck. Ideally between 2% and 3%
rarely smear-positive on direct microscopy. of cultures should be lost because of contami­
However M. bovis is naturally resistant to nation, indicating that the treatment pro­
pyrazinamide and this drug should not be cedure is neither too harsh nor too moderate.
used.
The laboratory diagnosis of tuberculosis
The unique character of the mycobacteria is thus still relies on direct microscopy and
that they are acid-fast. Once stained by an culture using techniques devised many years
aniline dye such as carbol fuchsin they resist ago. There is no 100% reliable serological test
decolorization with acid and alcohol and are for tuberculosis (Chapter 17a, p. 367 et seq.}.
thus termed 'acid- and alcohol-fast bacilli' or Many have been tried but all lack specificity
< linical Tuberculosis. Edited by P.D.O. Davies. Published in 1994 by Chapman & Hall, London. ISBN 0 412 48630 X

34

The microbiology of tuberculosis

and sensitivity. This is due to the ubiquitous
may be produced during the handling of the
nature of mycobacteria. They are widely
specimen or culture is unlikely to be a hazard
distributed in the environment and exposure
to the laboratory worker[5].
is unavoidable. As a result antibodies devel­
op to the cross-reacting antigens that are
ves 3.2 SPECIMENS
present in most mycobacteria and this gi._
rise to false-positive reactions. On the other
.^r Three specimens of sputum collected on
hand, individuals respond to different epi­
successive days should be provided. Saliva is
topes when infected and the use of a single
not adequate. A single specimen of sputum
antigen to detect antibodies gives false­ ’will
miss about
<
■"I! ~:s3
25% of microspically positive
negative results. Finally the widespread use and about 50% of culture-positive
cases. If
of BCG vaccination ma;■y make
' 1interpretation
r
there- ------is stilljaproblem
problemwith
withdiagnosis
diagnosisthen
thenan
an
difficult (Chapter 15d, p. 345 et seq.).
unlimited number of specimens should be
New approaches to the laboratory diagno
diag] ­ sent.
sis of tuberculosis include the detection of
When sputum is not available, for example
specific substances that can only be present in children the
because of the cconcomitant
■■
', —2 gastric contents may be
presence of collected. Ideally
the patient should cough
mycobacteria[3], the detection of mycobacter­
and swallow for 10 minutes before collection,
ial antigens by an ELISA technique[4] and the
which should be done early in the morning
polymerase chain reaction (Chapter 17b).
before food is taken. Laryngeal swabs are less
However, microscopy and culture remain the
efficient than sputum or gastric contents.
primary methods of laboratory diagnosis and
Broncho-alveolar lavage using a fibre-optic
are likely to do so for many years to come.
bronchoscope provides excellent specimens
Accurate and reliable identification and
from specific sites in the lung. However there
drug sensitivity tests are inecessapz
j for the
‘
is an increasing problem of contamination of
successful treatment of tuberculosis and
such specimens because tap water is being
other mycobacterial diseases. The Public
used in the machines for rinsing the bronchoHealth Laboratory Service (PHLS) has six
o ol.
,
scoPes instead of sterile distilled water
laboratories designated as E_
Regional1 Centres Unfortunately tap water contains environfor Tuberculosis Bacteriology. These work in
..
mental mycobacteria, which lodge in the
concert with the Mycobacterium Reference
servim f™- p i .
u
tubin8 that connects such machines to
Unit to provide a ‘
’ice for England and the mams supply. M. chelonei is the organism
Wales and currently^™
process
CLSS LT 95% °fm°St ,ikely to aPPe- - a contaminant
requests for identification and sensitivity
can be very difficult to eradicate^]. However
tests. Standard techniques are used through­
any one of the whole range of environmental
out the centres and sensitivity test results are
mycobacteria can cause this problem, ranging
subject to an on-going quality control system.
from A4. avium to M. kansasii.
Most of the mycobacteria that cause dis­
Aspirates such as pleural fluids, CSFs etc.
ease in man are classed as ^!.
c ‘ e8OZlPnthO? shouId be taken into sterile containers and
gens and therefore require specialized
handling in designated laboratories. Speci­ transported to the laboratory as rapidly as
possible. Actual pus is a better specimen than
mens suspected of containing mycobacteria
and cultures thought to be mycobacteria a pus swab but if possible a piece of tissue
from the site should be sent. It is often the
therefore have to be processed in microbio­
case that tissue is sent for histology and is
logical safety cabinets (class I or II) in a
placed directly into formalin, thus negating
laboratory having a class 3 containment
any chance of culture.
facility. This ensures that any aerosol that
Genitourinary tuberculosis is relatively rare

3I -

Culture

ng of the
a hazard

cted on
Saliva is
sputum
positive
-ases. If
then an
auld be

■xample
nay be
cough
lection,
lorning
are less
nts.
e-optic
cimens
■r there
tion of
being
□nchowater.
ivironin the
nes to
anism
and it
wever
nental
nging
's etc.
s and
lly as
i than

issue
n the
nd is
a ting

' rare

r

i

m England and Wales and most laboratories
|.H-k the expertise to detect and interpret the
significance of AFB in urine. However, when
gnnilourinary tuberculosis is considered a
pnssibilitv, then 50 ml of early morning urine
should be sent on three successive days;
24 h specimens are not recommended.
I he examination of faeces used to be rarely
necessary. Most cases of tuberculous enteritis
have lung lesions and tubercle bacilli in
faeces come from swallowed sputum. Howrver. with the advent of the acquired iminunodeliciecy syndrome (AIDS) the situa­
tion has changed. A significant number of
All )S patients have disseminated disease due
to Al. uoiuni and AFB are often demonstrable
m both faeces and biopsies of the intestinal
mucosa and can be cultured from these
spm miens.

i

3.3 MICROSCOPY

I he patient whose sputum is positive on
direct microscopy (Plate 1) is most likely to
mlect his/her close contacts[7]. Such a patient
will have at least 5000 organisms/ml of
sputum and may have up to ten times that
number. Below 5000 organisms/ml it is un­
likely that ZN or fluorescence staining will
detect APB. in England and Wales approxi­
mately 53% of new cases of pulmonary
tuberculosis are positive on direct smear[8].
Direct microscopy of specimens other than
sputum is of doubtful utility. Few laborator­
ies have the expertise to interpret ZN-stained
smears of urine and most no longer undertake this examination. Very careful scrutiny
ot smears from aspirates and tissues is
necessary to detect the small number that are
positive. These are most likely to be biopsy
specimens from lymph nodes.
I he main role of microscopy is therefore to
identify the truly infectious patient so that
the chain of infection can be halted by the
appropriate treatment of the index patient. It
is accepted that 2-3 weeks' chemotherapy
\\ith a modern
modern antifrubpmilncic
antituberculosis rocimon
regimen will

Y ▼▼ ’

▼

35

make most patients non-infectious. Such a
regimen will consist of isoniazid, rifampicin
and pyrazinamide given for 2 months
followed by isoniazid and rifampicin given
for an additional 4 months[9] (Chapter 8b,
p. 141).
3.4 CULTURE

The definitive diagnosis is the isolation of
A4. tuberculosis in pure culture. However this
is only achieved in about 50% of cases and
the diagnosis then has to rely on the clinical
and/or radiological features, sometimes with
histological evidence.
In the UK the most commonly used media
for the isolation of tubercle bacilli are
Lowenstein-Jensen (LJ) egg medium or
Kirchner broth containing an antibiotic mix­
ture. LJ is a simple egg-based medium that
may be produced in a buffered form or as an
acid egg medium[10]. Two tubes should be
used, one containing glycerol and the other
sodium pyruvate. Glycerol tends to inhibit
the growth of M. bovis whereas sodium
pyruvate encourages it.
From a sputum specimen positive by direct
microscopy, LJ will show growth within 2-3
weeks at 37 °C. The colonies are rough and a
beige-to-brown colour and show up well on
the green background, which is due to the
presence in the medium of the dye malachite
green (Plate 2). Smear-negative sputa and
specimens from other sites may not give a
positive growth until they have been incu­
bated for up to 6 or 8 weeks. Cultures
showing no growth are reported negative at
this time.
Kirchner being a liquid medium is more
difficult to interpret and it is necessary to
centrifuge the medium and make a ZNstained smear of the deposit. If AFB are
detected then subcultures are made on LJ but
this inevitably entails a delay in reporting.
There are two other methods used in the
UK for the culture of mycobacteria. These are
the Roche MB check system and the BACTEC

II

36

if

The microbiology of tuberculosis

460 radiometric system. Both these rely on disease, extrapulmonary disease and tuber­
the growth of the bacilli in a Middlebrook culosis in children. All the methods that have
broth. This is a liquid medium that readily been tried have lacked specificity and sensiti­
supports the growth of mycobacteriafl 1]. The vity due to the diversity of the antigenic
Roche MB check system is a biphasic system determinants of the tubercle bacillus and
in which the broth is inoculated with the to exposure to environmental mycobacteria
specimen that has been decontaminated in that share antigens with the tubercle bacil­
the usual way. By inverting the system lus.
during incubation, this broth culture is
ELISA tests have been described for the
flooded over a slope containing, on the one diagnosis of tuberculosis meningitis[13,14].
side, Middlebrook 7H-10 agar and, on the However, no ELISA system has yet proved to
other side, a Middlebrook 7H-10 agar be acceptable in the UK. A murine monoclo­
containing p-nitro-a-acetyIamino-3-hydroxy nal antibody to the 38 kDa antigen of M.
propiophenone (NAP) and a chocolate agar. tuberculosis has been used in a modified
Mycobacteria growing in the broth will form serological competition assay but unfortunavisihle colonies on the Middlebrook agar. If tely this is not'generally available[15J. Th’e>
they do not belong to the tuberculosis com- application
” "
of the polymerase chain reaction
plex they will also grow on the Middlebrook for the direct detection of M. tuberculosis,
agar containing NAP. If they are inot• __.
y __
mycowhile having enormous potential, is not in
bacteria at all they will grow on the chocolate routine use at present[16-18] (Chapter 17b
agarfl 2],
p. 381).
The BACTEC 460 radiometric system also
Rather than look for antibodies, attempts
uses a Middlebrook broth but one that have been made to detect either antigens
incorporates a C14-labelled substrate. During themselves or substances specific for myco­
metabolism C14-labelled CO2 is produced bacteria. Rabbit antisera to BCG have been
and this is automatically monitored by the used to detect antigens in CSF[19,20] and
machinefl 1], Using this system M. tuber­ gas-liquid chromatography linked with
culosis will grow, on average, 7-10 days mass spectrometry adapted for selected ion
more quickly than in the conventional LJ monitoring has been used to detect tubersystem.
culostearic acid (10-methyloctadecanoic acid)
The Middlebrook broth described above in both CSF and sputum[3,21]. Both tech­
will not accommodate blood or bone marrow niques need further evaluation and the GLCand is therefore of no use in disseminated MS technique requires very expensive equip­
disease. There is now available another ment.
medium designated Middlebrook 13A, which
will take up to 5 ml of blood or a bone
marrow. This is not so important from the 3.6 IDENTIFICATION OF TUBERCLE
point of view of classical tuberculosis but is BACILLI
very important in considering^
<
'’ '
the proWems It is possible to identify most mycobacteria to
of M. avium infections in patients with AIDS. the species level with tests based on simple
cultural and biochemical properties[22]. It is
also possible to identify sub-specific variants
3.5 OTHER METHODS OF DIAGNOSIS
by further biochemical tests, by phage-typing
Despite considerable effort there is still no or by serotyping but this is of epidemiological
acceptable serological test for tuberculosis value only and has little clinical relevance. It
(Chapter 17a, p. 367). Such a test would be is thus important to identify a strain only as
very useful for smear-negative pulmonary far as is clinically necessary. The recent

■

I

I

i

I*

Drug resistance

Table
Table 3.1
3.1 Cultural characters of tubercle bacilli
.,ltrr 2 weeks' incubation
PNB
THIA
37°C
45°C

tuberthat have
d sensitiantigenic
Hus and
obacteria
de bacil-

r\n
iiiia

for the
s[13,14J.
roved to
nonoclon of M.
nodified
fortuna15]. The
reaction
rculosis,
i not in
ter 17b,

p-nitro benzoic acid
thi.uetazone

incubated at 37°C.

development of DNA fingerprinting using
H-stnction fragment length polymorphism
(KI I T) allows a very precise identification of
a strain and its possible involvement in an
outbreak situation (Chapter 17c, p. 391 et seq.).
l ubercle bacilli have a characteristic mor­
phology when a ZN-stained smear is exammed bv microscopy. However it requires a
considerable degree of expertise to interpret
this and even then one can be misled. It is
therelore necessary to set up a small number
ol cultural and biochemical tests to confirm
the identity of a strain.
A suspension of the strain is inoculated
onto I.J slopes for incubation at 25°C, 37°C
and 45°C and also on to LJ slopes containing
p-nitro-benzoic acid (PNB 500 p2g/ml) and
thiacetazone (10 pig/ml). These are incubated
lor 2-3 weeks and a tubercle bacillus will
show the results illustrated in Table 3.1.
Mycobacteria other than tubercle bacilli
(MOTE bacilli), also called opportunist or
non-tuberculous mycobacteria and even
atypical mycobacteria, will show more rapid
growth and/or growth at 25°C or 45°C and/
or growth on the PNB slope and/or the
thiacetazone slope. Some of them will also
produce pigment either in the light only
(photochromogen) such as M. kansasii or in
both the light and dark (scotochromogens)
such as M. gordonae (Chapter 13, p. 265).
Strains belonging to the tuberculosis com­
plex can be divided into a number of variants
as described in the introduction. The cultural
characteristics for these variants are shown in
I able 3.2 but, as said earlier, these are of little
clinical importance except in the identifica­
tion of M. bovis, which is naturally resistant to

attempts
intigens
r myco/e been
20] and
I with
ted ion
tuber•ic acid)
h techeGLCequip-

teria to
simple
’]. It is
ariants
typing
-logical
nee. It
>nly as
recent

37

pyrazinamide. The inexpert administration
of vaccine strain Bacillus Calmette-Guerin
(BCG) can give rise to local abscesses and
even regional lymphadenitis and it is not
unusual to isolate BCG from such sites
(Chapter 14b, p. 297). The characteristics of
the strain have been included in Table 3.2 for
convenience.

3.7 IDENTIFICATION OF OPPORTUNIST
MYCOBACTERIA
The opportunist mycobacteria that most com­
monly cause pulmonary disease are:

• M. kansasii
• M. avium
• M. intracellulare
• M. scrofulaceum
• M. malmoense
• M. xenopi

MAIS complex

The identification of these organisms is best
undertaken by specialist laboratories. The
tests described for the identification of M.
tuberculosis are supplemented by a test for the
hydrolysis of Tween 80 and if necessary by
thin-layer chromatography of an extract con­
taining the superficial lipids from the cell
wall[22,23]. Many species of mycobacteria
have lipid patterns that are absolutely charac­
teristic and not shared with other species.
Sub-specific variants can also be identified
using this technique.
Most if not all species of opportunist
mycobacteria are widespread in the environ­
ment and they can gain access to specimens.
It can therefore be difficult to decide which
isolates are clinically significant. Multiple
isolates in the absence of any other pathogen
and with consistent clinical and/or radiologi­
cal features will usually indicate significance.

3.8 DRUG RESISTANCE IN TUBERCLE
BACILLI
Bacteria change their characters either: (i) in
response to a stimulus - that is, by induced

I

I ’

38

The microbiology of tuberculosis
Table 3.2 Variants of the tuberculosis complex
Variant
M. tuberculosis
Asian
African I
African II
M. bovis
BCG

Oxygen
preference"
A
A
M
M
M
A

TCHb

Nitratase

Pyrazinamide

Cycloserine

R
S
S

+
+

S
S

+

S
S
S
S

s
s
s
s
s

S

R
R

I

i

R

‘ A, Aerobic; M, microaerophilic; R, resistant; S, sensitive.
Thiophen-2-carboxylic acid hydrazide, 5 mg/1.

adaptation; or (ii) spontaneously (genetically) are ICT6 for rifampicin, 1(T5 for strepto­
by mutation. Mutations usually die out but if mycin and isoniazid, 10-3 for cycloserine and
they have survival value
. , , they
- will multiply 10-2 for pyrazinamide. Therefore the incimore than or iinstead’ of' the
‘ original popula- dence of doubly-resistant mutants would be
tion of bacilli and will eventually replace it. for example IO-10
for streptomycin plus
All populations of tubercle bacilli generate isoniazid.
small numbers that are resistant to any
particular drug. If the patient is given the
drug, a resistant mutant will not be affected 3-9 SENSITIVITY TESTS TO TUBERCLE
and if it is not killed by the body's defences or BACILLI
another drug, it will multiply and replace the There are opposing views of the need for
sensitive population. When a single drug is sensitivity tests. Most would agree thaTthev
given the outcome is finely balanced - the —
>•
....
.6
y
are helpful in patients who relapse or do not
defences may win over the very small initial convert and that if the technique is poor then
number of mutants but they may fail. This is they are not worth the effort. The disagree­
why single-drug therapy is often unsuccess­ ment arises over the advisability or need for
ful, resulting in drug resistance.
sensitivity tests in new cases and the need to
Excluding chemically related drugs, a do tests for drugs other than streptomycin,
mutant resistant to one drug is not especially isoniazid, rifampicin and ethambutol.
likely to be resistant to a different drug. Thus,
To plan treatment properly it is necessary
if two effective unrelated drugs are g
given to know the incidence of primary resistance
together there is only a small chance of there in the country as a whole, and to determine
being any mutant present that is resistant to this it is necessary to test a good sample of
both drugs at the same time. However, with strains not just once but regularly. It seems
very
. t severe disease, immunity
/ is low and the
-- reasonable therefore to carry out sensitivity
bacterial population is very large so doubly- tests on new cases if the facilities are available
resistant mutants may be present and there and the technique is adequate. Such tests also
can be failure to eliminate them. This is why provide useful supporting evidence for the
triple therapy is advisable initially - the identity
' ’
' organisms.
of' the
chance of a triply-resistant mutant can be
There is a further disagreement on how
ignored. When the bacterial population is sensitivity tests should be performed. It is
reduced (e.g. on conversion to culture­ simple to determine the minimum inhibitory
negative) it is safe to drop one of the drugs. concentration (MIC) of a drug. A control
Examples of resistant-mutant incidence slope and others containing increasing con-

I

i

Sensitivity tests

Table 3.3

39

Example of a modal resistance control set
Drug concentrations

'e

Strain A
B
C
D
E
Mode

Control

1

2

CG
IC
CG
CG
IC

CG
IC
CG
CG
IC
CG

CG
IC
IC

4

8

16

+

IC

CG, confluent growth; IC, innumerable discrete colonies; +, 20-100 colonies;
—, less than 20 colonies.

streptoerine and
the incivould be,
cin plus

:le
need for
that they
)r do not
■oor then
iisagreeneed for
.• need to
omycin,
1.
ecessary
■sistance
?termine
mple of
t seems
nsitivity
ivailable
ests also
for the

)n how
d. It is
hibitory
control
ig con-

(entrations of the drug are inoculated equally
and incubated. However, the results differ
considerably in different laboratories owing
tn v.inations in medium or technique and to
M>mc extent at different times in any one
l.iboratory. Even if the MIC is accurate - what
does it mean? The MIC can be related to
concentrations of drug attainable in the blood
or tissues or lesions but these concentrations
are not known for all drugs in all situations.
Also, in the body the concentrations are
always changing - furthermore there are
sometimes antagonistic substances present
and variations in pH or oxygen tension and
other conditions that effect the potency of
each drug in an unpredictable way. The MIC
is an in vitro test under special in vitro
conditions and there cannot be exact quanti­
tative equivalence (i.e. in |ig/ml).
I he only valid method of assessing sensiti­
vity tests is to relate the laboratory findings to
the results of treatment, initially in animal
experiments but then, more importantly, in
clinical trials. Such an approach can establish
the general outline of sensitivity technique
and interpretation but it has only been used
in man for a few drugs and then not
thoroughly. The chief difficulty is that drugs
are not used singly in therapy trials.
I he
problem
has
therefore
been
approached as follows:

1 • Establish the behaviour of normal (wild)
strains of TB.

2. Determine the response of the patient's
strain.
3. Assume that any difference (loss) of
sensitivity is harmful.
The normal disc-diffusion methods used to
determine the sensitivity of most bacteria are
not suitable for mycobacteria. The long incu­
bation period means that plates dry out and
the dynamics of the diffusion of the drug into
the agar become very complicated. It is
necessary therefore to test strains by a titra­
tion method. Tubes of medium containing a
range of concentrations of the drug (usually
doubling) are inoculated and incubated. The
results can be expressed in three ways:

1. MIC - the minimal inhibitory concentra­
tion.
2. RR - the resistance ratio to a standard
strain, usually H37Rv, (i.e. ratio of
MICs).
3. MR - modal resistance; this compares the
MIC of the test strain with the mode or
most common MIC of a group of normal
strains.

In the UK the modal resistance method is
most commonly used[24]. The mode is deter­
mined by testing a large number of strains
from new (untreated) cases of tuberculosis.
In the example shown in Table 3.3 only five
strains are used, but in practice five strains
are tested every week, i.e. 260 a year.

£;-X

/
40

The microbiology of tuberculosis

3.10 PYRAZINAMIDE

5. DHSS (1978) A Code of Practice for the Prevention
of Infection in Clincal Laboratories, Department
of Health and Social Security, HMSO, Lon­
don.
6. Nye, K.z Chadha, D.K., Hodgkin, P. et al.
(1990) Mycobacterium chelonei isolated from
broncho-alveolar lavage fluid and its practical
implications. J. Has. Infect., 16, 257-61.
7. British Thoracic and Tuberculosis Association
(1978) A study of standardized contact pro­
cedure in tuberculosis. Tubercle, 59, 245-59.
8. Medical Research Council Cardiothoracic Epi­
3.11 CONCLUSION
demiology Group (1992) National Survey of
notification of tuberculosis in England and
The decline in the incidence of tuberculosis
Wales. Thorax, 47, 770-5.
has seen a concomitant decline in expertise
9. British Thoracic Society (1984) A controlled
in general microbiology laboratories. Direct
trial of 6 months chemotherapy in pulmonary
microscopy and culture of specimens is still
tuberculosis. Final report: Results during the
undertaken in most hospital laboratories but
36 months after the end of chemotherapy and
as the number of positives has fallen there is
beyond. Br. J. Dis. Chest, 78, 330-6.
evidence that lack of experience has led to 10. Zaher, F. and Marks, ]. (1977) Methods and
false-positive results being issued. 'Pseudo'
medium for the culture of tubercle bacilli.
Tubercle, 58, 143-5.
outbreaks have aroused unnecessary alarm
[25]. False-negative results are much more 11. Middlebrook, G., Reggiardo, A. and Tigertt,
W.D. (1977) Automatable radiometric detec­
difficult to identify but it is reasonable to
tion of growth of Mycobacterium tuberculosis in
assume that they occur. It is to be hoped that
selective
media. Am. Rev. Respir. Dis 115
the increased incidence in tuberculosis aris­
1066-9.
ing from the problems in the USA and
12. Damato, R S., Isenberg, H.D., Hochstein, L. et
particularly that of multi-drug resistant
al. (1991) Evaluation of the Roche Septi Check
strains will rekindle interest at all levels in a
AFB system for the recovery of mycobacteria.
disease that was well on its way to being
/. Clin. Microbiol, 29, 2906-8.
eradicated in most developed countries.
13. Grange, J.M., Gibson, J., Nassau, E. and
Kardijito, T. (1980) Enzyme-linked immuno­
REFERENCES
sorbent assay (ELISA): a study of antibodies to
Mycobacterium tuberculosis in the IgG, IgA and
1. Castets, M., Rist, N. and Boisvert, H. (1969)
IgM classes in tuberculosis and sarcoidosis
La Varidte africaine du bacille tuberculeux
Tubercle, 61, 145-52.
humain. Med. Afr. Noire; 16, 321-2.
2. Hardy, R.M. and Watson, J. (1992) Mycobacter­ 14. Sada, E.D., Ferguson, L.E. and Daniel, T.M.
(1990) An ELISA for the serodiagnosis of
ium bovis in England and Wales: past present
and future. Epidemiol. Infect., 109, 23-33.
tuberculosis using a 30 000 Da native antigen
3. French, G.L., Teoh, R., Chan, C.Y. etal. (1987)
of Mycobacterium tuberculosis. I. Infect Dis
162, 928-31.
Diagnosis of tuberculous meningitis by detec­
tion of tuberculostearic acid in cerebrospinal 15. Wilkins, E.G.L. and Ivanyi, J. (1990) Potential
fluid. Lancet, ii, 117-19.
value of serology for diagnosis of extra pul­
4. Watt G., Zaraspe, G„ Bautista, S. and Laugh­
monary tuberculosis. Lancet, 336, 641-4.
lin, L. (1988) Rapid diagnosis of tuberculous 16. Bnsson-Noel, A., Gicquel, B., Lecossier D et
meningitis by using an enzyme-linked immu­
al. (1989) Rapid diagnosis of tuberculosis by
nosorbent assay to detect mycobacterial antiamplification of mycobacterial DNA in clinical
L)L ai58a681>^7 CerebrosPinal fluid- lnfectsamples. Lancet, ii, 1069-71.
17. Sjobnng, U., Meckleburg, M., Bengard, A.A.

Normal methods cannot be used as the drug
is only active in acid medium (pH 5.0 to 5.4)
and strains vary in their ability to grow at
such low pH levels. It is best to test at three
different pH levels, e.g. 5.0, 5.2 and 5.4.
Vigorous strains give a true result a pH 5.0,
feeble strains at 5.4 and others are inter­
mediate. To reduce the work, each test can be
limited to a control and a single drug tube.

w

i
L

References
HUi Miorner, H. (1990) Polymerase chain
reaction for detection of Mycobacterium tubercu/. Clin. Microbiol., 28, 2200-4.
is shankar, I’., Manjunath, N., Mohan, K.K. et
' (1991) Rapid diagnosis of tuberculosis
meningitis by polymerase chain reaction.
IiiiutI, 337, 5-7.
sada, IL, Ruiz-Palacios, G.M., Lopez-Vidal, Y.
and de Leon, S. (1983) Detection of mycobac­
terial antigens in cerebrospinal fluid of
patients with tuberculous meningitis by
en/\me-linked immunosorbent assay. Lancet,

■ Prevention
’epartment
4SO, Lon-

, P. et al.
ated from
ts practical
-61.
association
ntact pro245-59.
aracic Epi­
Survey of
;land and

ii, (151—2.
?0

controlled
ulmonary
uring the
?rapy and

hods and
le bacilli.

d Tigertt,
ric detecrculosis in
Dis. 115,

I

I

tein, L. et
pti Check
jbacteria.
E. and
immuno­
bodies to
IgA and
coidosis.

iel, T.M.
;nosis of
3 antigen
’Ct. Dis.,

Potential
xtra pulsier D et
ilosis by
n clinical
rd, A.A.

i

Kadhakrishan, V.V., Sehgal, S. and Mathai,
A (1990) Correlation between culture of Myco­
bacterium tuberculosis and detection of myco­
bacterial antigens in cerebrospinal fluid of
patients with tuberculous meningitis. J. Med.
Microbiol., 33, 223-6.

41

21. Oldham, G., Larsson, L. and Mardh, P. (1979)
Demonstration of tuberculostearic acid in
sputum from patients with pulmonary tuber­
culosis by selected ion monitoring. ]. Clin.
Invest., 63, 813-19.
22. Marks, J. (1976) A system for the examination
of tubercle bacilli and other mycobacteria.
Tubercle, 57, 207-52.
23. Marks, J. and Jenkins, P.A (1971) Thin layer
chromatography of mycobacterial lipids as an
aid to classification: technical improvements:
Mycobacterium avium, M. intracellulare (Battey
bacilli). Tubercle, 52, 219-25.
24. Marks, J. (1961) The design of sensitivity tests
of tubercle bacilli. Tubercle, 42, 314-16.
25. Davies, P.D.O. Williams, C.S.D., Shears, P. et
al. (1992) Pseudo-outbreak of tuberculosis in a
paediatric oncology ward [Abstract]. Thorax,
47, 220 P.

Position: 1867 (3 views)