LABORATORY DIAGNOSIS OF PULMONARY TUBERCULOSIS : CONVENTIONAL AND NEWER APPROACHES+ pdf

CONTINUING MEDICAL EDUCATION
Ind. J. Tub., 1996,43,107

LABORATORY DIAGNOSIS OF PULMONARY
TUBERCULOSIS : CONVENTIONAL AND NEWER APPROACHES
+

N.K. Jain*
Inspite of the discovery of the causative
agent more than a century back, the bacteriological
diagnosis of tuberculosis has been a major hurdle
in the treatment and control of the disease. Due
to its global prevalence, the methods available
for diagnosis are under rapid development. Briefly
reviewed, below are the present and the possible
future laboratory tools for die diagnosis of
pulmonary tubercolosis.

MICROSCOPY
The diagnosis of pulmonary tuberculosis
can be made by the detection of acid fast bacilli
by direct microscopy, using carbol fuchsin stain
and/or fluorochrome stain. Microscopy is a rapid
method but lacks sufficient sensitivity and does
not distinguish between different species of
mycobacteria. The sensitivity of microscopy is
often not more than 25-40% as compared to
culture, but under ideal conditions, it is possible
to achieve a rate of 60-70%. The average rate
obtained in India is 12-15%, but microscopy is
still the mainstay of the tuberculosis control

programme of the developing countries, including
India. The fluorochrome stain technique is better
then carbol fuchsin stain method, but is very
expensive and less specific, and is not suitable
for adoption as a routine method in our country.

CULTURE
Culture is considered as the reference method
for the detection of tubercle bacilli, but
mycobacterial culture is laborious, expensive and
slow. It is worth noting, however, that not more
than 50% of clinically diagnosed patients are
confirmed by culture. Traditional culture on
Lowenstein Jensen egg medium takes 2-6 weeks.
The introduction of other media for culture of
tubercle bacilli, such as 7H10 and 7H11 agar,
ave resulted in a faster & higher rate of

h


I

detection. These media an commonly used in
developed countries and are costly. The radiometric
respirometry technique (BACTEC) is more sensitive
and can drastically reduce the time needed for
detection, the median time being around 4 weeks
both for detection and sensitivity testing. This
technique is very commonly used in the United

States, but due to its limitaiions in terms of cost,
specificity and biohazards, it is not suitable for
developing countries. The other rapid culture
methods are biphasic broth agar system (Septi-
Chek AFB, Becton-Dickinson) and slide culture
method. The slide culture method seems very
good in principle, as it may help reduce the time
needed for culture and sensitivity, but it needs
standardization before it can be accepted as a
routine method.

SEROLOGICAL TESTS
The tubercle bacillus, like all other bacteria,
is rich in antigens that stimulate antibody
production. Several assays have been developed
in the last one century to detect specific antibody
response in suspected patiets. These assays use
either whole bacteria or fragments of AFB,
culture filtrates, partially purified antigens and
purified antigens both by chromatography and by
recombinant DNA technology. Various techniques
are used to carry out the tests, such as ELISA,
radio-immuno-assay and immunoblot. Moreover,
instead of looking for specific antibodies, attempts
have been made to detect antigens in clinical
samples using specific polyclonal and monoclonal
antibodies. There are several inherent difficulties
in hunting for antibodies to diagnose tuberculosis.
Crude antigen preparations which are likely to
share epitopes with most, if not all, wild strains

of M. tuberculosis are also likely to share some
epitopes with non-pathogenic environmental
mycobacteria. Purified antigens, on the other
hand, may not be expressed by every patient

* Bacteriologist, New Delhi Tuberculosis Centre, J.L. Nehiu Marg, New Delhi - 110 002.

+
Interested readers could contact the author for a comprehensive list of books/articles on the subject.

108
N.K. JAIN


infected with tubercle bacilli. Antibody responses
may persist for years particularly if one accepts
the notion of a dormant population of bacilli that
occasionally surface to reproduce before becoming
dormant again. In high prevalence areas, it may
be difficult to distinguish current from old disease.
But, to date, no assay is acceptable for the
diagnosis of tuberculosis.

The tuberculin test is also used by many
physicians for the purpose of diagnosis. It is not
a test for diagnosis as it tells only whether or
not a person is infected with M.tuberculosis, but
cannot differentiate between infection and disease.
This test is of no use at all in countries with
endemic tuberculosis where a large part of the

population is already infected with the bacilli. It
may only be of use in children where other tests
are not of much help.

recent years, the diagnosis of tuberculosis
has ueen facilitated by the detection of
mycobacterial cell wall components
(Tuberculostearic acid or TSA) and increased
production of host enzymes (Adenosine Deaminase)
directly in clinical samples. TSA is present not
only in M.tuberculosis, but also in other
mycobacteria as well as other microorganisms
which constitute the normal microflora in various
body sites. Thus, this method is not suitable for
diagnosis. The sensitivity of ADA test is good
but specificity is low. Consequently, this test is
not very popular.

DNA PROBES
Recent developments in the field of molecular
genetics of mycobacteria have made it possible
to identify particular sequences of DNA that are
specific for individual mycobacterial species.
These unique DNA sequences can be detected
using labelled oligonucleotides that are exactly
complementary to the nucleotide sequence in the
mycobacterial genomic DNA. Such DNA probes
can identify genus with high specificity, or
species specific bacterial DNA sequences. Although
such probes have been shown to be highly

sensitive and specific, when used in research
laboratories, these lose specificity and sensitivity
when used directly in clinical samples.

PCR
Currently, much attention is being focussed
on the use of polymerase chain reaction (PCR).
The principle of the PCR technique is based on
the amplification of a given DNA sequence to
a large number of copies that can be identified
by separation on gel electrophoresis and,
subsequently, either with or without probing with
a labelled oligonucleotide specific for the
amplified DNA fragment. During the last decade,
several such unique sequences have been reported
for the M.tuberculosis complex. The usefulness
of PCR for detection of tubercle bacilli in clinical
specimens has been confirmed in several recent
studies, with sensitivities and specificities
ranging from 60% to 100%. Contamination is a
problem that all PCR laboratories must face,
since the product of the reaction contains
millions of suitable templates that can be carried
back to the next assay on finger tips, pipettes,
clothing or in aerosols. With mycobacteria there
is the additional difficulty of extracting DNA
from within the cells and this is one of the
important limiting factors in determining the
sensitivity of PCR assays for M.tuberculosis in
clinical material. There have been many reports

in recent years and several groups have published
encouraging results involving considerable number
of samples, but routine mycobacteriology
laboratories are not yet convinced of its practical
utility. The results of a recent study organised
by the WHO (Noordhoek et al, 1993) must be
an eye opener for those of us who are carried
away by certain reports coming from research
publications. Batches of 200 samples containing
varying numbers of M.bovis BCG suspended in
water, saliva or sputum were sent, coded, to
seven laboratories that had established PCR
assays for M.tuberculosis. One of these was
unable to provide results within 6 months. Three
had results with specificities of less then 80%
and the three laboratories, in which false positives
were rare, only picked up 60% of samples with
10
3
organisms in 0.2 ml, just a little improvement
on a good microscopy service. At present, there
are no commercial PCR kits that have been
evaluated and licensed by the Food and Drug
Administration (U.S.A) for detection of
M.tuberculosis in clinical laboratories.

LABORATORY DIAGNOSIS OF PULMONARY TUBERCULOSIS

109

RFLP
Molecular genetic methods are also useful
in studying the epidemiology of tuberculosis.
Recent outbreaks of multiple drug resistant
tuberculosis _have been traced by DNA finger
printing of tubercle bacilli isolates, using
restriction fragment length polymorphism (RFLP).
The principle of the method is to extract the
mycobacterial DNA from cultured organisms,
digest it with chosen DNA cleaving restrictive
enzyme(s), separate the DNA fragments produced
by gel electrophoresis, whereby certain repetitively
occurring DNA sequences (insertion sequences)
are identified by specific probes. This results in
a fingerprint highly specific for each individual
mycobacteral strain. This molecular genetic
identification on the sub-species level now makes
it possible to trace the spread of specific strains
in the community. This technique may, in the
near future, drastically change the morphology of
tuberculosis laboratories, especially in
industrialized countries. But what about developing
countries, which have 2/3rds of the world’s
population and almost 90% of patients? Do we
foresee that the fruits of these newer technologies
can help the developing countries control the
problem of tuberculosis in the near future? I
would like to suggest that we should stick to the
original established tools for the diagnosis of
Tuberculosis and try to improve upon our own

deficiencies for making the best use of the old
technique of microscopy instead of criticizing the
inherent deficiencies of this technique. If we can
do so, this old but time tested technique can
match the newer technologies as far as tuber-
culosis control is concerned. The saying that
OLD is GOLD stands true still. Lipsky, in a
review of factors affecting the clinical’ value of
microscopy for acid fast bacilli, concludes that
when the results of all specimens from each
patient are considered in total, the acid fast smear
has a predictive value of >96% and remains one
of the most rapidly performed tests in the
detection of pulmonary tuberculosis. All other
techniques need to be compared critically with
microscopy and their diagnostic role assessed.
The traditional model of history, examination and
sputum microscopy has proved itself capable of
detecting the majority of infectious cases and, in
conjuction with adequate case-finding and

treatment, remains the backbone of tuberculosis
control programmes.

DRUG SUSCEPTIBILITY
The role of drug sensitivity testing should
not be underestimated in the treatment of difficult
cases of tuberculosis. This is another very important
area where the laboratory plays an important role.
Anti-tuberculosis therapy depends on the

susceptibility of the tubercle bacilli. The newer
regimens include four drugs because of increase
in initial drug resistance levels in the high
prevalence countries.

Drug resistance may be defined as the
ability of strains of tubercle bacilli to survive
and grow despite exposure to concentrations of
the drug that inhibit or kill the parental bacilli,
and to transfer this characteristic to its progeny.
It has also been defined as a decrease in
sensitivity to a drug of sufficient degree to
make it reasonably certain that the strain
concerned is different from a sample of tubercle
bacilli that have never come into contact with
the drug. At present, the only known
mechanism for M. tuberculosis’s acquisition of
drug resistance is spontaneous, mutation. To
date, there is no clear evidence indicating that
tubercle bacilli acquire resistance through
resistance transfer factors or other genetic
mechanisms.

Untreated tuberculosis patients infected with
drug-resistant bacilli are referred to as having
‘primary’ drug resistance, to distinguish them
from those who had drug-sensitive organisms
orginally and later developed resistance because
of inadequate or inappropriate chemotherepy.
The latter are generally referred to as having

‘acquired’ drug resistance. A better term for
‘primary’ drug resistance might be ‘initial’ drug
resistance, because what is reported in the literature
is usually a mixture of primary drug resistance
with an unknown degree of acquired drug
resistance. Recently, another term - multi-drug
resistant tuberculosis (MDR-TB) - has been
added and is defined as “drug resistance against
two most potent drugs i.e., Isoniazid and Rifampicin
with or without resistance against any other
drugs”.

110
N.K. JAIN


Three widely used methods for testing
sensitivity are the proportion method, the resistance
ratio method amd the absolute concentration
method. Resistance ratio method is most
commonly used in developing countries, whereas
the proportion method is common in industrialized
countries. The more rapid method commonly
used in U.S.A. is the radiometric system, the
BACTEC, and is a modification of conventional
proportion method. The test uses Middlebrook
7H12 broth with a radio-labelled fatty acid
substrate. The amount of growth, indicated by
changes in the growth index in the media with
known drug concentrations as compared to that

in the control bottle, has been correlated to the
presence or absence of resistance in 1% of the
inoculum. Thus, if an isolate grows beyond a
specific growth index compared with the control,
it is considered resistant to the specific agent.
This method of testing has a very rapid turnaround
time, with results for sensitivity usually available
in less than 7 days. This method is very costly
both in establishment and recurring expenses,
besides being hazardous. It may not be possible
to use this as routine method in the developing
countries in the near future. The resistance ratio
or proportion method uses the routine LJ medium
and takes 4 weeks. The use of 7H10 or 7H11
medium instead of LJ for drug susceptibility can
reduce the time by one week. The use of slide
culture method can reduce the time considerably,
but needs standardization before being used as
a routine method.

Innovative methods for rapid drug testing
are currently being developed. In one such
method, mycobacteria are infected with specific
reporter phage expressing the firefly luciferase
gene. Light production is dependent on phage
infection, expression of the luciferase gene, and
the level of cellular ATP. Signals can be detected
within hours of infection of virulent M.tuberculosis
isolates with reporter phage. When organisms are
exposed to drugs to which they are susceptible,

the-light is extinguished unlike strains resistant
to that drug. The result may be available within

two days. The method is still in its developmental
stage.

Another approach for the rapid determination
of drug susceptibilities is based on potential
difference between the genetic material of a drug
resistant strain and that of a drug susceptible
strain. This difference may be applied for Isoniazid
resistance detection by detecting the presence or
absence of Kat G and INHA genes associated
with Isoniazid resistance to M.tuberculosis. There
is also a good possibility of detecting Rifampicin
resistance by a PCR based assay that detects
difference between Rifampicin susceptible and
Rifampicin resistant strains. The rpoB gene which
encodes the RNA polymerase subunit b, has been
cloned and ‘mutations in this gene have been
identified in rifampicin resistant strains but not
in rifampicin sensitive strains. The rapid screening
test for rpoB gene mutations has been successfully
carried out by using PCR single-strand confirmation
polymorphism (SSCP) technique. Similarly,
resistance to Streptomycin in M.tuberculosis is
due to mutations in rpsL gene which encodes the
ribosomal protein S12 and rrs gene which encodes
16S rRNA. With PCR-SSCP technique, the
mutations in rpsl and rrs genes have been

detected in clinical isolates of M.tuberculosis
resistant to Streptomycin. Similarly, different
PCR-SSCR patterns of mutations in codons of
gyrA have been found to be associated with
resistance to Ciprofloxacin. This novel technique,
PCR-SSCR, can detect drug resistance in 48-72
hours. However, rapid identification on molecular
genetic principles may be too intricate to be
practical, at least in the developing countries.

The diagnosis of tuberculosis and
antimycobacterial susceptibility testing are
important problems confronting laboratory,
physicians and scientists. However, until new
technologies and techniques are well established,
we should make good use of old and time tested
techniques to their maximum efficacies to treat,
manage and control tuberculosis as well as
expand the boundaries of our knowledge.