1
2
Title: MICROSCOPIC-OBSERVATION DRUG-SUSCEPTIBILITY (MODS) FOR RAPID DIAGNOSIS
3
OF LYMPH NODE TUBERCULOSIS AND DETECTION OF DRUG RESISTANCE
4
5
Daniela E. Kirwana,b#, Cesar Ugarte-Gilb,c, Robert H. Gilmand,e, Luz Caviedese*, Syed M. Hasan
6
Rizvif, Eduardo Ticonag,h,i, Gonzalo Chavezg, José Luis Cabreraj, Eduardo D. Matosh,k, Carlton A.
7
Evansl,m, David A. J. Mooree,n, and Jon S. Friedlandb.
8
9
Department of Medical Microbiology, St. George’s Hospital, London, UKa; Infectious Diseases and
10
Immunity, Imperial College London, UKb; Instituto de Medicina Tropical Alexander von Humboldt,
11
Universidad Peruana Cayetano Heredia, Lima, Peruc; Department of International Health, Johns
12
Hopkins University, Baltimore USAd; Laboratorio de Investigación en Enfermedades Infecciosas,
13
Universidad Peruana Cayetano Heredia, Lima, Perue; Department of Cellular Pathology, Barts Health
14
NHS Trust, London, UKf; Infectious Diseases and Tropical Medicine Unit, Hospital Nacional Dos de
15
Mayo, Lima, Perug; Universidad Nacional Mayor de San Marcos, Lima, Peruh; Universidad de San
16
Martin de Porres, Lima, Perui; Department of Pulmonology, Hospital Daniel Alcides Carrión, Callao,
17
Perúj; Infectious Diseases Unit, Hospital Nacional Arzobispo Loayza, Lima, Perúk; Infectious
18
Diseases and Immunity, and Wellcome Trust Centre for Global Health Research, Imperial College
19
London, UKl; IFHAD: Innovation for Health and Development, Universidad Peruana Cayetano
20
Heredia, Lima, Perum; TB Centre, London School of Hygiene & Tropical Medicine, London, UKn.
21
22
# Address correspondence to: Dr. Daniela E. Kirwan, Department of Medical Microbiology, St.
23
George’s Hospital, Blackshaw Road, Tooting, London, United Kingdom, SW17 0QT.
24
(0044) 7794937328 dannikirwan@yahoo.com
25
26
* Deceased
1
27
28
29
Running Title:
MODS DIAGNOSIS OF TB IN SOLID TISSUES
Keywords:
MODS
30
31
32
33
Tuberculosis
34
Lymphadenitis
35
Drug susceptibility testing
36
Diagnosis
37
Resistance
38
.
39
40
Abstract
41
In the study, 132 patients with lymphadenopathy were investigated. 52 (39.4%) were diagnosed with
42
TB. MODS provided rapid (13 days), accurate diagnosis (sensitivity 65.4%), and reliable DST.
43
Despite lower sensitivity than other methods, faster results and simultaneous DST are advantageous
44
in resource-poor settings, supporting incorporation of MODS into diagnostic algorithms for
45
extrapulmonary TB.
(Words: 53)
46
2
47
In 2013 14.5% of new tuberculosis (TB) notifications worldwide were extra-pulmonary(1), and in
48
certain regions this percentage is much higher(2). Non-specific disease manifestations and
49
paucibacillary infection make diagnosing extra-pulmonary TB challenging(3, 4). Culture, the
50
diagnostic gold standard, allows species identification and drug-susceptibility testing (DST)(5), but
51
generating results takes several weeks; automated liquid culture systems are relatively faster(6) but
52
financial constraints limit their use.
53
54
The microscopic-observation drug-susceptibility (MODS) assay is a low-cost, liquid culture-based
55
diagnostic assay for TB(7, 8). With comparable accuracy as other culture techniques(7, 9, 10), MODS
56
is faster(7), provides simultaneous DST(10-13), and has World Health Organization (WHO) approval
57
for direct testing of sputum specimens in low-resource settings(14, 15). MODS accurately diagnoses
58
TB from cerebrospinal fluid(16) and pleural specimens(17), but its role in the diagnosis of solid tissue
59
TB remains unknown. This prospective cross-sectional study was designed to investigate the use of
60
MODS for culture of lymph node tissue in an operational setting.
61
62
Patients aged ≥18 years with lymphadenopathy requiring diagnostic tissue sampling were recruited
63
consecutively from three public hospitals in Lima, Peru over 14 months. Ethical approval was
64
obtained from the Institutional Ethics Committee of the Universidad Peruana Cayetano Heredia,
65
Asociación Benéfica PRISMA, and each hospital’s ethics approval committee. All patients provided
66
written informed consent. For each patient, clinical and demographic data were collected and the
67
treating physician was asked to give the most likely diagnosis. Patients with unknown HIV status were
68
offered testing.
69
70
Tissue sampling was performed routinely, and samples were immediately divided into three equal
71
parts and processed as outlined in Figure 1. MODS was performed in accordance with published
72
Standard Operating Procedures(18). Microbiological criteria for TB were positivity on at least one of
3
73
Auramine microscopy, MODS, or Löwenstein-Jensen (LJ) culture. Strains obtained by LJ culture
74
underwent phenotypic DST using the proportions method, performed at the national tuberculosis
75
reference laboratory, and the in-house Tetrazolium Microplate Assay (TEMA). Samples for
76
histological evaluation were sealed in paraffin blocks and reported routinely. Once recruitment was
77
ended the blocks were retrieved, and slides fixed and stained with hematoxylin-eosin, Ziehl-Neelsen,
78
and Periodic acid-Schiff. 3 independent pathologists blinded to clinical data recorded presence of
79
Acid-fast bacilli (AFB), granulomas, and caseating necrosis, and gave an overall diagnosis. A
80
histological definition of TB required concordance between two or more pathologists; where retrieval of
81
paraffin blocks had not been possible (n=11) the hospital pathology report was obtained, and if TB had
82
been identified, this was utilized. TB was diagnosed when microbiological and/or histological criteria
83
were met.
84
85
Data were entered into Excel and analyzed using Stata version 12 (StataCorp). Nominal demographic
86
data and test characteristics were compared using Fisher’s exact test. Time to results was compared
87
using the Mann-Whitney U test. A p<0.05 was considered significant. Agreement between
88
pathologists was assessed using Cohen’s Kappa coefficient for multiple ratings; Kappa ≥0.81 was
89
taken to indicate substantial agreement(19).
90
91
144 specimens from 132 patients were tested. Patient demographics are presented in Table 1. 52
92
patients (39.4%) were diagnosed with TB (Table 1, Figure 2a): 19 were positive by Auramine
93
microscopy, 34 by MODS, 40 by LJ culture, and 43 by histology; sensitivities were 36.5%, 65.4%,
94
76.9%, and 82.7%, and negative predictive values (NPV) were 70.8%, 81.6%, 87.0%, and 89.9%
95
respectively. HIV positive patients were more likely to have positive Auramine microscopy results than
96
HIV negative patients (65% and 18.8% respectively, p=0.001). TB was detected in 5 of 12 patients
97
already taking TB therapy (5 by microbiological methods and 3 by histological analysis; all 5 were HIV-
98
positive with CD4+ counts <250 cells/mm3). Physicians suspected TB in 48 TB-positive and 53 TB-
4
99
100
negative patients, thus sensitivity, specificity, positive predictive value (PPV) and NPV of the
physician’s presumptive TB diagnosis were 92.3%, 33.8%, 47.5% and 87.1% respectively.
101
102
42 patients had one or more positive microbiological tests (Figure 2b). MODS, TEMA, and the
103
proportions method detected multi-drug resistant TB (MDR-TB) in two patients and isoniazid mono-
104
resistance in one patient. MDR-TB was identified by TEMA and the proportions method in one patient
105
whose isolate grew on LJ culture but not MODS. The proportions method reported low-level isoniazid
106
resistance in five additional samples; all others were fully drug-susceptible according to all methods.
107
108
Positive results were communicated after a median interval of 13 days (IQR 11-18 days, n=57) for
109
MODS and 22 days (IQR 17-28 days, n=78) for LJ culture (p<0.001), and negative results after 41
110
days (IQR 38-41 days, n=186) for MODS and 63 days (IQR 59-64 days, n=163) for LJ culture
111
(p<0.001). Median time to positivity for patients on TB treatment was 9.5 days (IQR 6.5-21 days, n=6)
112
for MODS and 15 days (IQR 15-17 days, n=8) for LJ culture. Auramine microscopy results were
113
communicated after 1 day for both positive (IQR 1-2 days, n=42) and negative (IQR 1-3 days, n=145)
114
samples. As DST was performed in batches, time to results is not available; however, laboratory data
115
indicate assay times of 40-45 days for the proportions method and 7-10 days for TEMA.
116
117
Contamination was reported for 38 (26.6%, n=143) direct MODS assays, and for none performed
118
following sample decontamination (n=144). For LJ culture, 42 (29.4%, n=143) direct and 4 (2.8%,
119
n=144) decontaminated cultures were reported as contaminated. Sensitivity and NPV were higher
120
following decontamination than for direct processing for all microbiological methods (Table 3). This
121
finding persisted when contaminated specimens were excluded from analysis. There was no
122
difference in time to positivity between pre-decontaminated and directly processed specimens.
123
5
124
Paraffin blocks from 121 patients were obtained for full histopathological assessment, and a
125
consensus diagnosis was reached for 117 patients (Table 2). The most frequent histological
126
diagnosis was TB (34.2%). AFB were observed in 8 specimens on ZN staining; all were culture-
127
positive, and 6 were positive on Auramine microscopy. Inter-observer agreement was high for the
128
presence of granulomas (κ=0.83) and caseous necrosis (κ=0.90), but not AFB (κ=0.18).
129
130
This study is the first prospective evaluation of MODS culture for solid tissue specimens. MODS
131
accurately detected TB from lymph node tissue specimens. In contrast with data from respiratory
132
specimens(20), MODS was less sensitive than LJ culture (65.4% vs 76.9%), although almost twice as
133
sensitive as Auramine microscopy (36.5%). Comparable to other studies of TB lymphadenitis,
134
sensitivities of all microbiological assays were lower than sensitivities reported using sputum
135
specimens(3, 6, 21, 22), possibly due to low bacterial load in tissue and/or clumping of pathogens(6).
136
Accordingly, time to results was longer than has been reported from sputum specimens (13 vs 7 days
137
for MODS, 26 vs 22 days for LJ culture)(7). For both culture methods, contamination rates on direct
138
testing were high. Specimen pre-decontamination increased sensitivity of all diagnostic tests.
139
140
MODS provided accurate data on isoniazid and rifampicin resistance simultaneously with diagnosis,
141
which facilitates timely initiation of appropriate regimens and may prevent further development of
142
resistance. The MDR-TB rate was similar to rates previously documented in Lima (7.7% vs.
143
8.6%)(24), although numbers in this study were small; further prospective testing on greater sample
144
numbers is needed to confirm accuracy of resistance testing. Universal DST is recommended in
145
regions where primary MDR-TB rates exceed 3%(23), and MODS may be particularly valuable in such
146
settings.
147
148
Physicians overestimated rates of TB, which leads to over-treatment and delays in obtaining correct
149
diagnoses. Universal access to MODS has similar potential to improve outcomes in extrapulmonary
6
150
TB, as in respiratory disease(25). Performing MODS and LJ culture in parallel would combine the
151
benefits of rapidity and simultaneous drug-resistance data afforded by MODS and the higher
152
sensitivity of LJ culture.
153
154
19 patients had discrepant microbiological and histological results. Similar findings have been
155
reported elsewhere(6). Possible explanations include failure of the host to generate a typical
156
histopathological response, visualization of killed bacilli in patients undergoing TB treatment, or recent
157
use of antibiotics with some antimycobacterial effect(26). Different test modalities may be beneficial in
158
these different patient groups. Concordance between the pathologists’ findings and diagnoses was
159
high, although agreement with respect to AFB visualization was poor. The pathologists were blinded
160
to clinical information whereas in practice findings are interpreted within a clinical context, and
161
accuracy may be greater than that observed in this study.
162
163
In conclusion, although MODS is less sensitive than LJ culture, it is able to accurately diagnose TB
164
from lymph node tissue significantly faster. It can also correctly detect resistance to rifampicin and
165
isoniazid simultaneous to diagnosis, enabling prompt initiation of targeted treatment. No single
166
diagnostic test for TB fulfils all properties of an ideal test, and multiple methods should be used in the
167
diagnostic work-up of patients with lymphadenopathy. MODS may have an important role to play in
168
the diagnosis of TB in resource-limited settings. These data support the expansion of MODS to solid
169
tissue specimens within programmatic guidelines.
170
171
172
173
7
174
Acknowledgements
175
Other members of the Lymph Node TB Working Group in Peru include: Johnny Cárdenas Núñez
176
(Department of Head and Neck Surgery, Hospital Daniel Alcides Carrión, Callao, Perú), Gustavo
177
Cerrillo (Department of Pathology, Hospital Nacional Dos de Mayo, Lima, Perú), Jaime Cok
178
(Department of Pathology, Hospital Nacional Cayetano Heredia, Lima, Peru), Romulo Escobedo
179
(Department of General Surgery, Hospital Nacional Dos de Mayo, Lima, Perú), Margarita Marchino
180
(Department of Head and Neck Surgery, Hospital Nacional Arzobispo Loayza, Lima, Perú), Ernesto
181
Nava (Department of Pathology, Hospital Nacional Arzobispo Loayza, Lima, Perú), and José Luis
182
Saavedra Leveau (Universidad Nacional Mayor de San Marcos, Lima, Peru and Department of Head
183
and Neck Surgery, Hospital Nacional Dos de Mayo, Lima, Perú). We are grateful to all clinical staff
184
and patients who participated in the study for their collaboration.
185
186
This work was supported by the Sir Halley Stewart Foundation (DK) and with an ISID Small Grant
187
from the International Society for Infectious Diseases (CUG). JSF thanks the Imperial College
188
Biomedical Research Centre for financial support. The views expressed within this report are those of
189
the authors and not necessarily of the funders. The funders had no role in study design, data
190
collection and analysis, or preparation of the manuscript.
191
192
193
Conflict of Interests
194
None.
8
195
References:
196
1.
197
198
Switzerland.
2.
199
200
WHO. 2014. Global Tuberculosis Report 2014. World Health Organization, Geneva,
Pedrazzoli D, Fulton N, Anderson L, Lalor M, Abubakar I, Zenner D. 2012. Tuberculosis
in the UK: 2012 report. Health Protection Agency,
3.
Hillemann D, Rusch-Gerdes S, Boehme C, Richter E. 2011. Rapid Molecular Detection of
201
Extrapulmonary Tuberculosis by the Automated GeneXpert MTB/RIF System. J Clin Microbiol
202
49:1202-1205.
203
4.
Chakravorty S, Tyagi J. 2005. Novel Multipurpose Methodology for Detection of
204
Mycobacteria in Pulmonary and Extrapulmonary Specimens by Smear Microscopy, Culture,
205
and PCR. Journal of Clinical Microbiology 43:2697-2702.
206
5.
Hillemann D, Richter E, Rüsch-Gerdes S. 2006. Use of the BACTEC Mycobacteria Growth
207
Indicator Tube 960 Automated System for Recovery of Mycobacteria from 9,558
208
Extrapulmonary Specimens, Including Urine Samples. J Clin Microbiol 44:4014-4017.
209
6.
Verma J, Dhavan I, Nair D, Manzoor N, Kasana D. 2012. Rapid culture diagnosis of
210
tuberculous lymphadenitis from a tertiary care centre in an endemic nation: Potential and
211
pitfalls. Indian Journal of Medical Microbiology 30:342-345.
212
7.
Moore DAJ, Evans CAW, Gilman RH, Caviedes L, Coronel J, Vivar A, Sanchez E, Pinedo
213
Y, Saravia JC, Salazar C, Oberhelman R, Hollm-Delgado M-G, LaChira D, Escombe AR,
214
Friedland JS. 2006. Microscopic-Observation Drug-Susceptibility Assay for the Diagnosis of
215
TB N Engl J Med 355:1539-1550.
216
217
8.
Caviedes L, Lee T-S, Gilman RH, Sheen P, Spellman E, Lee EH, Berg DE, MontenegroJames S, The Tuberculosis Working Group in Peru. 2000. Rapid, Efficient Detection and
9
218
Drug Susceptibility Testing of Mycobacterium tuberculosis in Sputum by Microscopic
219
Observation of Broth Cultures. J Clin Microbiol 38:1203-1208.
220
9.
Arias M, Mello Fernanda CQ, Pavon A, Marsico Anna G, Alvarado-Galvez C, Rosales S,
221
Pessoa Carlos Leonardo C, Perez M, Andrade Monica K, Kritski Afranio L, Fonseca
222
Leila S, Chaisson Richard E, Kimerling Michael E, Dorman Susan E. 2007. Clinical
223
Evaluation of the Microscopic-Observation Drug-Susceptibility Assay for Detection of
224
Tuberculosis Clinical Infectious Diseases 44:674-680.
225
10.
Rasslan O, Hafez S, Hashem M, Ahmed O, Faramawy M, Khater W, Saleh D, Mohamed
226
M, Khalifa M, Shoukry F, El-Moghazy E. 2012. Microscopic observation drug susceptibility
227
assay in the diagnosis of multidrug-resistant tuberculosis. International Journal of Tuberculosis
228
& Lung Disease 16:941-946.
229
11.
Dixit P, Singh U, Sharma P, Jain A. 2012. Evaluation of nitrate reduction assay, resazurin
230
microtiter assay and microscopic observation drug susceptibility assay for first line
231
antitubercular drug susceptibility testing of clinical isolates of M. tuberculosis. Journal of
232
Microbiological Methods 88:122-126.
233
12.
Makamure B, Mhaka J, Makumbirofa S, Mutetwa RB, C. Dimairo, M. Bandason, T.
234
Munyati, S S. Mangwanya, D. Mungofa, S. Butterworth, A E. Mason, P R. Corbett, E L.,
235
Mfupfumi L, Mason P, Metcalfe J. 2013. Microscopic-Observation Drug-Susceptibility
236
Assay for the Diagnosis of Drug-Resistant Tuberculosis in Harare, Zimbabwe. PLoS One
237
8:e55872.
238
239
13.
Bwanga F, Hoffner S, Haile M, Joloba M. 2009. Direct susceptibility testing for multi drug
resistant tuberculosis: a meta-analysis. BMC Infectious Diseases 9:67.
10
240
14.
WHO. 2011. Noncommercial Culture and Drug-Susceptiblity Testing Methods for Screening
241
Patients at Risk for Multidrug-Resistant Tuberculosis: Policy Statement. World Health
242
Organization, Geneva.
243
15.
244
245
WHO. 2009. Strategic and technical advisory group for tuberculosis (STAG-TB) Report of the
ninth meeting. World Health Organization,
16.
Caws M, Dang T, Torok E, Campbell J, Do D, Tran T, Nguyen V, Nguyen T, Farrar J.
246
2007. Evaluation of the MODS culture technique for the diagnosis of tuberculous meningitis.
247
PLoS ONE 2:e1173.
248
17.
Tovar M, Siedner M, Gilman R, Santillan C, Caviedes L, Valencia T, Jave O, Escombe A,
249
Moore D, Evans C. 2008. Improved diagnosis of pleural tuberculosis using the microscopic-
250
observation drug-susceptibility technique. Clinical Infectious Diseases 46:909-912.
251
18.
Moore DAJ. 25th November 2008 2008. Diagnosing extrapulmonary tuberculosis with the
252
MODS assay. Standard operating procedure for sample preparation and inoculation for TB
253
detection: Lymph node aspirate.
254
http://www.modsperu.org/sops/SOP_LN_aspirates_v5_Nov_2008_English.pdf. Accessed 18
255
July 2015.
256
19.
257
258
Landis J, Koch G. 1977. The measurement of observer agreement for categorical data.
Biometrics 33:159-174.
20.
Moore DAJ, Mendoza D, Gilman RH, Evans CAW, Hollm Delgado M-G, Guerra J,
259
Caviedes L, Vargas D, Ticona E, Ortiz J, Soto G, Serpa J, the Tuberculosis Working
260
Group in Peru. 2004. Microscopic Observation Drug Susceptibility Assay, a Rapid, Reliable
261
Diagnostic Test for Multidrug-Resistant Tuberculosis Suitable for Use in Resource-Poor
262
Settings J Clin Microbiol 42:4432-4437.
11
263
21.
Linasmita P, Srisangkaew S, Wongsuk T, Bhongmakapat T, Watcharananan S. 2012.
264
Evaluation of real-time polymerase chain reaction for detection of the 16S ribosomal RNA gene
265
of Mycobacterium tuberculosis and the diagnosis of cervical tuberculous lymphadenitis in a
266
country with a high tuberculosis incidence. Clin Infect Dis 55:313-321.
267
22.
Hailu E, Assefa T, Bekele Y, Mihret A, Aseffa A, Hussien J, Ali I, Abebe M. 2012.
268
Comparison of PCR with standard culture of fine needle aspiration samples in the diagnosis of
269
tuberculosis lymphadenitis. J Infect Dev Ctries 6:53-57.
270
23.
271
272
WHO. 2009. Treatment of Tuberculosis: guidelines for national programmes. World Health
Organization, Geneva, Switzerland.
24.
Asencios L, Quispe N, Mendoza-Ticona A, Leo E, Vásquez l, Jave O, Bonilla C. 2009.
273
Vigilancia Nacional de la Resistencia a Medicamentos Antituberculosis, Perú 2005-2006. Rev
274
Peru Med Exp Salud Publica 26:278-287.
275
25.
Mendoza-Ticona A, Alarcón E, Alarcón V, Bissell K, Castillo E, Sabogal I, Mora J, Moore
276
D, Harries A. 2014. Effect of universal MODS access on pulmonary tuberculosis treatment
277
outcomes in new patient in Peru. Public Health Action 2:162-167.
278
26.
279
280
TR. S. 2004. The WHO/IUATD diagnostic algorithm for tuberculosis & empiric
fluoroquinolone use: Potential pitfalls. . Int J Tuberc Lung Dis 8:1396-1400.
27.
WHO. 2007. Definition of a new sputum smear-positive TB case, 2007.
281
http://www.who.int/tb/laboratory/policy_sputum_smearpositive_tb_case/en/index.html.
282
Accessed 27th March 2015
283
284
12
Table 1: Demographic information and results for all study participants.
Auramine
TB positive
(n=52)
Age, median years (IQR) b
positive
positive
(n=19)
(n=34)
(n=80)
Total
diagnosis
(n=40)
(n=132)
of TB (n=43)
23 (44.2)
34 (42.5)
0.86
3 (15.8)
13 (38.2)
16 (40.0)
22 (51.2)
57 (43.2)
39 (26-46)
41 (27-55)
0.38
32 (29-40)
32 (25-48)
32.5 (25-46)
35 (24-47)
40 (25-52)
20 (38.5)
34 (42.5)
0.72
13 (68.4)
12 (35.3)
14 (35.0)
15 (34.9)
54 (40.9)
HIV positive (%)a
CD4+ count, median cells/mm3 (IQR) b
Histological
LJ positive
P value
Females (%) a
MODS
TB negative
75 (27-218) 163 (121-271)
0.18
87 (25-140) 75 (22-163) 87 (25-140)
63 (25-219) 156 (41-234)
Positive sputum smear (%)a
4 (7.7)
0 (0)
0.02
4 (2.1)
3 (8.8)
4 (10.0)
3 (7.0)
4 (3.0)
On TB treatment >1week (%)a
5 (9.6)
7 (8.8)
1.00
5 (26.3)
4 (11.8)
4 (10.0)
3 (7.0)
12 (9.7)
Previous TB (%)a
10 (19.2)
13 (16.3)
0.65
3 (15.8)
2 (5.9)
4 (10.0)
7 (16.3)
23 (17.4)
CXR normal (%)a
27 (51.9)
38 (47.5)
0.375
8 (42.1)
15 (48.4)
18 (45.0)
22 (51.2)
65 (49.2)
CXR abnormal* (%)a
18 (34.6)
28 (35.0)
0.558
10 (52.6)
14 (44.1)
16 (40.0)
14 (32.6)
46 (34.8)
a
Compared using Fisher’s exact test
b
Compared using Wilcoxon rank-sum test
*Where specified, abnormalities included pulmonary infiltrates and/or consolidation (n=14), pleural effusion(s) (n=14), hilar and/or paratracheal
adenopathy (n=9), cavitation (n=1), and miliary TB (n=1).
13
1
Table 2: Histopathological findings and histological diagnoses for patients in whom a consensus diagnosis was reached by
2
two or more pathologists.
Agreement between ≥ 2
Kappa
pathologists (%) (n=117)
AFB on ZN stain
2 (1.7)
0.18
Granuloma
42 (35.9)
0.83
Caseous material
34 (29.1)
0.90
TB1
40 (34.2)
0.85
Lymphoma
15 (12.8)
0.66
3 (2.6)
0.87
Other malignancy
16 (13.7)
0.88
Hyperplasia
31 (26.5)
0.64
Histoplasmosis
1 (0.85)
0.24
Other
11 (9.4)
0.33
Findings on histological
examination
KS
Histological diagnosis
1
Where agreement was between two pathologists only, diagnosis given by the third pathologist
was: reactive changes (n=3), lymphoma (n=1), non-lymph node tissue (n=2), other/unspecified
(n=2). Both patients with non-lymph node tissue were culture-positive for TB.
3
14
Table 3: Comparison of results and test characteristics for specimen processing following decontamination and direct
inoculation of the specimen without prior decontamination.
Auramine microscopy
MODS
LJ Culture
Decontaminated
Direct
Decontaminated
Direct
Decontaminated
Direct
(n=144)
(n=143)
(n=144)
(n=143)
(n=144)
(n=143)
Positive
22 (15.3%)
20 (14.0%)
39 (27.1%)
18 (12.6%)
46 (31.9%)
32 (22.4%)
Negative
122 (84.7%)
123 (86.0%)
101 (70.1%)
85 (59.4%)
94 (65.3%)
69 (48.3%)
Indeterminate
0
0
4 (2.8%)
2 (1.4%)
0
0
Contaminated
0
0
0
38 (26.6%)
4 (2.8%)
42 (29.4%)
Sensitivity (%)
44.9
40.8
79.6
37.5
93.9
66.7
NPV (%)
77.9
77.2
90.5
76.0
97.0
88.6
Sensitivity (%)
44.9
40.8
79.6
54.5
93.9
91.4
77.9
77.2
90.5
82.8
96.8
95.7
All specimens
Contaminated
specimens excluded NPV (%)
Sensitivity (%)
44.9
81.6
93.9
NPV (%)
77.9
91.3
96.9
Overall
15
Figure legends
Figure 1: Flow diagram indicating procedures following patient enrollment.
For Auramine microscopy, ≥1 acid-fast bacilli (AFBs) visualized per 100 fields was considered
positive(27). *Drug susceptibility testing (DST) performed for all patients positive on LJ culture.
One strain tested per patient.
Figure 2: Venn diagrams showing numbers of patients positive according to different test
modalities. a) Positivity for TB according to Auramine microscopy, TB culture, and
histological evaluation (n=132). 12 patients were positive by all three methods, 26 were
positive by two methods, and 14 by one method only. Two patients were positive on Auramine
microscopy only. b) Positivity according to microbiological methods of detection of TB
(n=132). All patients positive by MODS were also positive by LJ culture. 8 patients were
positive by Auramine microscopy (n=3) and/or LJ culture (n=6), but not MODS.
16
Figure 1
Patients enrolled
(n=132)
Tissue sampling:
biopsy/aspiration
Specimen in formaldehyde
Hospital pathology
laboratory
(n=132)
Paraffin blocks not
retrieved
(n=11)
Specimen in NaCl, 4°C
TB laboratory, UPCH
(n=132)
Specimen in NaCl,
4°C. Hospital microbiology
laboratory
(n=132)
Fungal culture
Tissue homogenization; dilution
in 0.85% NaCl to volume of
5mls
Bacterial
culture
Specimen
decontamination, NALCNaOH
Direct
inoculation
Slides for triple
evaluation
(n=121)
Auramine
microscopy
ZN stain
H&E stain
LJ culture
MODS
culture
Auramine
microscopy
LJ culture
PAS stain
DST: TEMA and
Proportions
method *
17
DST: TEMA and
Proportions
method *
MODS
culture
Figure 2
Auramine
2
Culture
5
2
12
21
10
80
Negative
Histology
a)
MODS culture
Auramine
2
16
18
1
90
Negative
5
LJ Culture
b)
18