Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
445
Original Article
SMJ
Siripen Kanchanasuwan, M.D., Narongdet Kositpantawong, M.D.
Division of Infectious Diseases, Department of Internal Medicine, Faculty of Medicine, Prince of Songkla University, Songkhla 90110, ailand.
The Performance of Real-Time Polymerase Chain
Reaction in Patients with Scanty Positive Acid-Fast
Bacilli Sputum Smear in Diagnosis of Pulmonary
Tuberculosis: 5-Year Retrospective Study
ABSTRACT
Objective: To assess the performance of real-time polymerase chain reaction (RT-PCR) to diagnosis pulmonary
tuberculosis in patients with scanty positive acid-fast bacilli sputum smears, in a single hospital.
Materials and Methods: All patients, who had scanty positive AFB sputum smears in Songklanagarind Hospital;
between 2015 and 2019 were included. Demographic data, clinical data, radiographic ndings, RT-PCR and
mycobacterial culture results were reviewed.
Results: From a total of 269 patients reporting scanty AFB smears, 116 patients (43.1%) had cultures conrmed
as M. tuberculosis. From overall, samples from 92 patients with scanty AFB smear were processed for RT-PCR.
ere were 26 (28.3%) isolates having positive RT-PCR test results. Of these 26 isolates that RT-PCR positive, 25
(96.2%) were culture positive, while only 1 (3.8%) were culture negative. A remaining 66 samples that RT-PCR
negative, 15 (22.7%) were culture positive for tuberculosis. Using mycobacterial cultures as the gold standard, the
sensitivity, specicity, positive predictive value and negative predictive value of RT-PCR were 62.5%, 98.1%, 96.2%,
and 77.3%, respectively. Pulmonary consolidation and cavity on chest radiograph were associated with the growth
of M. tuberculosis, with an OR of 2.3 (95% C.I. 0.26-0.73) and 3.4 (95% C.I. 1.2-9.9), respectively.
Conclusion: Less than half of the patients with scanty smears had culture-conrmed tuberculosis; RT-PCR also
has low sensitivity. Consequently, a negative RT-PCR does not exclude tuberculosis; especially in cases of a high
index for clinical suspicion. Radiographic ndings; including pulmonary consolidation and cavities, are helpful
predictors for supporting thisdiagnosis.
Keywords: Performance; scanty acid fast bacilli; pulmonary tuberculosis; polymerase chain reaction (Siriraj Med
J 2021; 73: 445-450)
Corresponding author: Siripen Kanchanasuwan
E-mail: kaymed29@yahoo.com
Received 19 January 2021 Revised 11 May 2021 Accepted 17 May 2021
ORCID ID: https://orcid.org/0000-0002-6311-7463
http://dx.doi.org/10.33192/Smj.2021.58
INTRODUCTION
Pulmonary tuberculosis remains a serious, worldwide
public health problem.
1
ailand remains on the list of the
world’s 14
th
highest burden country of tuberculosis.
2-3
e
high number of tuberculosis cases and deaths indicates
that actions are urgently needed to reduce tuberculosis
incidence. Rapid identication and treatment of new
cases is the keystone of tuberculosis control.
4-5
Acid Fast
Bacilli (AFB) sputum smear microscopy is widely used
as the diagnostic test for mycobacterial disease; as it is a
simple, rapidand cost-eective method for diagnosing
tuberculosis.
6
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
446
Ziehl-Neelsen (ZN) stained smear is a conventional
method; however, auramine-staining has reported to have
10 percent more sensitivity with similar specicity.
7-12
Since 2006, the central laboratory of Songklanagarind
Hospital has applied the auramine uorescence microscopic
technique as its screening test for all requested sputum
acid fast smears, before being conrmed with ZN-stained.
With this more sensitive staining technique, we found
increasing reports of scanty AFB in sputum smears.
e scanty AFB results were dened as: the presence
of less than 10 bacilli in the 100 oil-eld on microscopy.
The International Union Against Tuberculosis and
Lung Disease (IUATLD), along with the World Health
Organization (WHO), previously recommended that
additional sputum AFB smears should be repeated in
this degree of positivity, when reported as doubtful.
13-14
In 2013, WHO revised the case denitions of pulmonary
tuberculosis.
15
Cases with scanty AFB on microscopy will
be considered as sputum smear positive tuberculosis.
is lowering diagnostic threshold thereby increases
case detection rates and treatment initiation. However,
this degree of positive acid fast smear might not well
correlate with cultures, as it could reect non-tuberculous
mycobacteria, or contamination by environmental
mycobacteria.
16-17
e nucleic acid amplication test
(NAAT) is useful as it can rapidly detect M. tuberculosis
bacteria in specimens within hours, while the turnaround
time of mycobacterial cultures requires 2 to 6 weeks for
reporting. However, the NAAT performance to detect
M. tuberculosis in scanty positive AFB sputum smears
remains a challenge, because of low bacillary loads.
20-22
Hence, the aim of this study was to explore the yield
of scanty positive acid fast sputum smears, and the
multiplex Real Time-Polymerase Chain Reaction (RT-
PCR) Anyplex
™
II MTB/MDR Detection technique to
diagnose pulmonary tuberculosis.
Objectives
e purposes of this study were:
1. To determine the performance of real-time
polymerase chain reaction in patients with scanty positive
acid-fast bacilli sputum smear in diagnosis of pulmonary
tuberculosis
2. To explore the factors that can predict the growth
of M. tuberculosis in patients with scanty positive AFB
sputum smears
MATERIALS AND METHODS
Study design and population
A retrospective review of medical and microbiological
records of all suspected respiratory tuberculosis patients,
whose sputum were sent for AFB staining, from January,
2015 to December, 2019 in Songklanagarind Hospital,
an 800-bed, teaching-based, tertiary care hospital in
Songkhla province, Southern ailand. e investigational
protocol was approved by the Institutional Review Boards
of Faculty of Medicine, Prince of Songkla University:
REC. 63-287-14-1. Enrollment criteria of the patients
were: (1) aged > 15 years, (2) had at least one specimen
showing a scanty positive acid fast smear, (3) having
at least one sputum specimen sent for mycobacterial
culture. Patients were excluded if they: (1) had recent
or ongoing treatment with anti-tuberculous agents,
(2) had no obtained mycobacterial culture, (3) had not
undergone chest radiograph during sputum collection.
Data collection
e patient’s demographic data, presenting symptoms,
underlying medical illness (es), results of mycobacterial
culture and multiplex RT-PCR, chest radiographic ndings,
rate of anti-tuberculosis treatment and adverse drug
events were all recorded.
As the laboratory routine works, sputum specimens
are processed and performed for decontamination and
concentration via a standard method, preparation of
slides and uorochrome staining with auramine O. With
auramine O staining, mycobacteria appear as bright,
yellow uorescent rods under a uorescent microscopy.
Slide with a presence of uorescent rods are ZN stained,
and evaluated by using a conventional light microscope.
Performing a ZN stain, aer initial auramine O staining,
is accepted as standard practice in Songklanagarind
Hospital. e number of acid-fast bacilli observed has
been quantied according to the IUALD and the WHO
scales. Mycobacterial cultures are the gold standard for
bacteriological conrmed diagnosis of tuberculosis, and
growing bacteria is required to perform drug-susceptibility
testing. Sputum specimens were cultured in liquid (7H9
broth) medium, with use of the automated Mycobacterial
Growth Indicator Tube system and solid Löewenstein-
Jensen medium.
Eligible patients were patients who had scanty
positive acid fast bacilli of sputum smear, which was
dened as: having presence of at least one of two, or
three slides reported to have less than 10 acid fast bacilli
(AFB) found in 100 oil elds; according to the WHO
scale. Regarding the results of sputum mycobacterial
cultures, they were assigned as either a positive culture for
M. tuberculosis (positive C/S MTB) or negative culture
for M. tuberculosis (negative C/S MTB). Negative C/S
for MTB was dened by there being at least one or
more sputum specimens sent for mycobacterial cultures
Kanchanasuwan et al.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
447
Original Article
SMJ
reported as growing non-tuberculous mycobacterial
(NTM), or no mycobacterial growth.
As part of a diagnostic investigation, direct molecular
testing was performed on 92 scanty sputum smears,
submitted by physician request using the Anyplex
™
II
MTB/MDR Detection kit. is assay is a semi-automated
system, and provides rapid results within 3-4 hours of
sample receipt. is technique also has a lower rate of
error and contamination [11], compared to the Line
probe assay. Before performing the multiplex RT-PCR
technique, DNA ofM. tuberculosis was extract by using
DNA-extraction solution, provided in the kits. Anyplex
™
II
MTB/MDR Real-time Detection was performed following
the directions provided by the manufacturer.Amplication
and detection were performed on a Rotor-Gene 3000
instrument, for all sample extracts. M. tuberculosis
detection targeted the IS6110 and MPB64 genes. Result
interpretation was performed automatically, using the
instrument’s soware according to threshold and cuto
values outlined by the manufacturer (16). Overall, the
Anyplex MTB/NTM assay demonstrated sensitivity,
specicity, PPV, and NPV of 86%, 99% 96%, and 95%, for
M. tuberculosis detection compared with mycobacterial
cultures.
Statistical analysis
To describe the variables characteristics, these were
expressed as mean with standard deviation, median with
range, proportion in percentage and ratio. To examine
dierences between 2 groups of variables these were
analyzed by Fisher’s exact or χ
2
test. e categorical variables
were analyzed by Student’s t-test or Mann-Whitney U
test for the continuous variables, according to types of
its distribution. Statistical analysis was performed with
SPSS soware version 23 (SPSS Inc., Chicago, IL, USA).
P-values <0.05 were considered statistically signicant.
RESULTS
During the ve-year study period, 416 patients met
criteria for inclusion in the study. From this, 147 patients
(35.3%) were excluded for the following reasons: 126 patients
had recent or ongoing treatment with anti-tuberculous
agents, 2 patients underwent no chest radiograph during
sputum collection, and 19 had no obtained mycobacterial
culture. In total, 269 patients were enrolled in this study
(Fig 1).
Of the 269 patients with scanty positive AFB smears,
116 patients (43.1%) had positive culture for M. tuberculosis,
while the remaining 153 patients (56.9%) had negative
Fig 1. Flow diagram for patient
enrollment.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
448
cultures. ere were no statistically signicant dierences
between these 2 groups; in terms of: patient characteristics,
presenting symptoms or comorbidities, as shown in
Table 1. Means age of positive and negative mycobacterial
culture groups were 55.6±18.7 years vs. 61.7±17.1 years,
respectively (p=0.134). About two-thirds of patients in
both groups were male. Pulmonary consolidation and
cavity on chest radiograph were frequently detected in
the culture positive group (44.8% vs. 26.1%, 10.3% vs.
3.2%, respectively, P<0.05), while normal chest radiograph
favored the culture negative group (1.7% vs. 24.8%,
P<0.05). Reticular lesions were commonly found in
active respiratory tuberculosis; however, there was no
statistically signicant dierence between positive and
negative mycobacterial culture groups (P=0.676). From
this, 93.1% of patients in the culture positive group
received anti-tuberculosis treatment, compared to 56.9%
of patients in the culture negative group (P<0.05). e
percentage of adverse drug events were 26.7% in the
positive culture group and 9.8% in the negative culture
group (P =0.135)
ere were 92 sputum scanty AFB smears sent for
multiplex RT-PCR, and overall there were 26 (28.3%)
cases having positive RT-PCR test results. Of these,
25 (96.2%) were culture positive, while only 1 (3.8%)
were culture negative. A remaining 66 samples that
RT-PCR negative, 15 (22.7%) were culture positive for
tuberculosis. Using mycobacterial cultures as the gold
standard, the sensitivity, specicity, positive predictive
value and negative predictive value of RT-PCR were 62.5%,
98.1%, 96.2%, and 77.3%, respectively (Table 2). Overall,
the diagnostic yield of multiplex RT-PCR in scanty AFB
was relatively low. However, anti-tuberculous treatment
should be initiated without delay in patients with positive
RT-PCR, as it has a low, false positivity (3.8%).
DISCUSSION
Scanty positive AFB reporting accounted for one-third
from all sputum smear positives in our institute. Of these,
41.3% grew M. tuberculosis. Similarly, previous studies
showed that fewer than half of the smears, with scanty
AFB, yielded positive cultures.
23-25
e chance of scanty
positive AFB to be a positive culture for M. tuberculosis
is limited, and might be due to low bacillary loads. False
positive scanty sputum smears are possibly from non-
tuberculous mycobacteria
26
; particularly asymptomatic
patients without chest radiological evidence suggesting
active tuberculosis.
As for the results of this study; 195 patients (72.5%)
who had scanty positive AFB smears were being treated
with anti-tuberculosis treatment. Interestingly, 12 out
of 38 patients (31.6%), who had scanty AFB smears and
negative mycobacterial cultures, were receiving anti-
tuberculosis treatment; despite normal chest radiograph.
Our study also revealed that RT-PCR had low sensitivity
for patients who had scanty positive AFB smears. e
results are supported by IUATLD/WHO recommendations,
in that a single negative NAA test could not exclude the
diagnosis of pulmonary tuberculosis; especially in cases
of moderate to high index of clinical suspicion. Chest
radiographic ndings are helpful in terms of prediction the
following growth of M. tuberculosis. Chest radiographic
ndings indicate active respiratory tuberculosis; including,
consolidation opacities, and cavitation, which are related
to the growth of M. tuberculosis. Only 2 of our 116 patients
(1.7%) having normal chest radiographic ndings had
M. tuberculosis in their sputum. Our ndings suggest
empiric anti-tuberculous treatment for all scanty positive
AFB patients with radiographic ndings suggestive of
active tuberculosis, as this might be appropriate; whereas,
the follow up approach is considered in patients who
have normal chest radiography. is strategy is to avoid
overtreatment and drug related complications.
There are some limitations in this study. First,
tuberculosis diagnosis is based entirely on the detection
of M. tuberculosis via culture techniques. Failure to isolate
M. tuberculosis cannot denitely exclude a diagnosis of
active tuberculosis. Second, our retrospective study may
have introduced a selection bias, because approximately
two-thirds half of the patients with scanty positive AFB
smears did not performed RT-PCR tests. us, the outcome
might not reect the true performance of RT-PCR in all
patients with scanty positive results.
CONCLUSION
Presence of scanty positive AFB in sputum smear
was common, while only 43.1% of them were nally
conrmed as tuberculosis by mycobacterial culture.
Also, the RT-PCR sensitivity in scanty acid fast sputum
smears is very low. Hence, if clinical suspicion is high,
tuberculosis should not be ruled out based solely on
negative RT-PCR results. Chest radiographic ndings
are helpful in determining empirical anti-tuberculosis
treatment.
REFERENCES
1. Global tuberculosis report. Geneva: World Health Organization;
2020. Available from: https://www.who.int/tb/publications/
global_report/en [cited 2020 Dec 9].
2. World Health Organization. Regional Oce for South-East
Asia.Tuberculosis control in the South-East Asia Region:
Annual Report 2016.Available from: https://apps.who.int/
iris/handle/10665/205286 [cited 2020 Dec 9]
Kanchanasuwan et al.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
449
Original Article
SMJ
Characteristics Patients with scanty positive AFB (n=269)
Positive C/S for MTB Negative C/S for MTB p-value
n=116 (43.1%) n=153 (56.9%)
Age in years (mean ±S.D.) 55.64 ±18.68 61.66 ±17.09 0.134
Male gender (%) 65.5% 57.5% 0.183
Presenting symptoms*
Cough 63 (54.31%) 66 (43.14%) 0.069
Fever 22 (18.97%) 20 (13.07%) 0.187
Weight loss 20 (17.24%) 26 (16.99%) 0.957
Hemoptysis 8 (6.89%) 12 (7.84%) 0.769
Pleurisy 3 (2.58%) 2 (1.30%) 0.442
Dyspnea 13 (11.2%) 15 (9.8%) 0.709
Underlying diseases*
HIV/AIDS 6 (5.17%) 9 (5.88%) 0.802
Diabetes 14 (12.07%) 23 (15.03%) 0.485
Malignancy 23 (19.83%) 43 (28.1%) 0.118
Renal failure 3 (2.59%) 5 (3.27%) 0.744
Immuno-suppressive therapy 6 (5.17%) 7 (4.58%) 0.821
Chest radiograph nding*
Normal 2 (1.72%) 38 (24.84%) <0.05
Consolidation 52 (44.82%) 40 (26.14%) <0.05
Reticular 22 (18.97%) 26 (16.99%) 0.676
Cavity 12 (10.34%) 5 (3.27%) 0.018
Miliary 6 (5.17%) 1 (0.65%) 0.021
Pleural effusion 15 (12.93%) 12 (7.84%) 0.169
Nodule/mass 28 (24.14%) 51 (33.33%) 0.101
Received Anti-TB treatment 108 (93.1%) 87 (56.86%) <0.05
TABLE 1. Demographic, clinical manifestations and laboratory results of patients with scanty AFB, and results of
their mycobacterial culture.
Plus–minus values are means ±SD
Abbreviations: AFB=acid fast bacilli, C/S= Culture, MTB= Mycobacterium tuberculosis, HIV/AIDS = Human immunodeciency virus/
Acquired Immune Deciency Syndrome
*Patients may have had >1 manifestation.
TABLE 2. e results of multiplex real-time polymerase chain reaction of patients with scanty positive AFB sputum
smears and mycobacterial culture.
Positive C/S for MTB Negative C/S for MTB Total
(n=40) (n =52) (n=92)
Positive PCR for MTB 25 1 26
(n= 26)
Negative PCR for MTB 15 51 66
(n= 66)
40 52 92
Abbreviations: AFB=acid fast bacilli, PCR = Polymerase Chain Reaction, C/S = Culture, MTB=Mycobacterium tuberculosis, scanty AFB=
positive acid fast staining but < 10 AFB/100 OF
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
450
3. Glaziou P, Falzon D, Floyd K, Raviglione M. Global epidemiology
of tuberculosis. Semin Respir Crit Care Med 2013;34(1):3-16.
4. Cattamanchi A,Davis JL,Worodria W,Boon S, Yoo S, Matovu
J, et al. Sensitivity and Specicity of Fluorescence Microscopy
for Diagnosing Pulmonary Tuberculosis in a High HIV Prevalence
Setting. Int J Tuberc Lung Dis 2009;13(9):1130-6.
5. Young DB, Perkins MD, Duncan K, Barry CE. Confronting
the scientic obstacles to global control of tuberculosis. J Clin
Invest 2008;118(4):1255-65.
6. Monkongdee P, McCarthy KD, Cain KP, Tasaneeyapan T,
Nguyen HD, Nguyen TNL, et al. Yield of Acid-fast Smear
and Mycobacterial Culture for Tuberculosis Diagnosis in People
with Human Immunodeciency Virus. Am J Respir Crit Care
Med 2009;180(9):903-8.
7. Githui W, Kitui F, Juma ES, Obwana DO, Mwai J, Kwamanga
D. A comparative study on the reliability of uorescence
microscopy and Ziehl-Neelsen method in the diagnosis of
pulmonary tuberculosis. East Afr Med J 1993;70(5):263-6.
8. Prasanthi K, Kumari AR. Ecacy of uorochrome stain in the
diagnosis of pulmonary tuberculosis co-infected with HIV.
Indian J Med Microbiol 2005; 23(3):179–85.
9. Kivihya-Ndugga LE, van Claa MR, Githui WA, Nganga LW,
Kibuga DK, Odhiambo JA, et al. A comprehensive comparison
of Ziehl-Neelsen and uorescence microscopy for the diagnosis
of tuberculosis in a resource poor urban setting. Int J Tuberc
Lung Dis 2003;7(12):1163-71.
10. Singh NP, Parija SC. e value of uorescence microscopy of
auramine stained sputum smears for the diagnosis of pulmonary
tuberculosis. Southeast Asian J Trop Med Public Health 1998;
29(4):860-3.
11. Hanscheid T, Ribeiro CM, Shapiro HM, Perlmutter NG.
Fluorescence microscopy for tuberculosis diagnosis. Lancet
Infect Dis 2007;7(4):236-7.
12. Steingart KR, Henry M, Ng V, Hopewell PC, Ramsay A,
Cunningham J, et al. Fluorescence versus conventional sputum
smear microscopy for tuberculosis: a systematic review. Lancet
Infect Dis 2006;6(9):570-81.
13. Enarson DA, Rieder HL, Arnadottir T, Trébucq A. Management
of tuberculosis: a guide for low income countries. 5
th
ed. Paris,
France: International Union Against Tuberculosis and Lung
Disease; 2000.
14. World Health Organization. Laboratory services in tuberculosis
control. WHO/TB/98.258. Geneva: WHO; 1998.
15. World Health Organization. Denitions and reporting framework
for tuberculosis-2013 revision. World Health Organization;
2013.
16. Lee JS, Kim EC, Joo SI, Lee SM, Yoo CG, Kim YW, et al. e
incidence and clinical implication of sputum with positive acid-
fast bacilli smear but negative in mycobacterial culture in a tertiary
referral hospital in South Korea. J Korean Med Sci 2008;23(5):
767-71.
17. Van Deun A, Salim AH, Cooreman E, Daru P, Das AP, Aung
KJ, et al. Scanty AFB smears: what’s in name. Int J Tuberc
Lung Dis 2004;8(7):816-23.
18. Wei Z, Zhang X, Wei C, Yao L, Li Y, Zhang X, et al. Diagnostic
accuracy of in-house real-time PCR assay for Mycobacterium
tuberculosis: a systematic review and meta-analysis. BMC
Infect Dis 2019;19(1):701.
19. Shinu P, Nair A, Singh V, Kumar S, Bareja R. Evaluation of
rapid techniques for the detection of mycobacteria in sputum
with scanty bacilli or clinically evident, smear negative cases
of pulmonary and extra-pulmonary tuberculosis. Memórias
do Instituto Oswaldo Cruz 2011;106(5):620-4.
20. Chumpa N, Kawkitinarong K, Rotcheewaphan S, Sawatpanich
A, Petsong S, Tumwasorn S, et al. Evaluation of Anyplex™
II MTB/MDR kit’s performance to rapidly detect isoniazid
and rifampicin resistant Mycobacterium tuberculosis from
various clinical specimens. Mol Biol Rep 2020;47(4):2501-8.
21. Perry MD, White PL, Ruddy M. Potential for use of the Seegene
Anyplex MTB/NTM real-time detection assay in a regional
reference laboratory. J Clin Microbiol 2014;52(5):1708-10.
22. Igarashi Y, Chikamatsu K, Aono A, Yi L, Yamada H, Takaki
A, et al. Laboratory evaluation of the Anyplex™ II MTB/MDR
and MTB/XDR tests based on multiplex real-time PCR and
melting-temperature analysis to identify Mycobacterium tuberculosis
and drug resistance. Diagn Microbiol Infect Dis 2017;89(4):276-
81.
23. Kubica G P. Correlation of acid fast staining methods with
culture results for mycobacteria. Bull Int Union Tuberc 1980;
55:117-24.
24. Mokhtari Z, Larbaoui D. Correlation of results of slightly-
positive direct diagnosis with those obtained from culture.
Bull Union Int Tuberc 1973;48:88-9.
25. Odubanjo MO, Dada-Adegbola HO. e microbiological
diagnosis of tuberculosis in a resource-limited setting: is
acid-fast bacilli microscopy alone sucient? Ann Ib Postgrad
Med 2011;9(1):24-9.
26. Lawson L, Yassin MA, Ramsay A, Emenyonu NE, Squire SB,
Cuevas LE, et al. Comparison of scanty AFB smears against
culture in an area with high HIV prevalence.Int J Tuberc Lung
Dis 2005;9(8):933-5.
Kanchanasuwan et al.
