| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Laboratorio de Microbiología Clínica, Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru
2 Laboratorio de Micobacteriología, Oficina General de Epidemiología, Ministerio de Salud, Lima, Peru
3 Programa Nacional para el Control de la Tuberculosis, Hospital Nacional Cayetano Heredia, Lima, Peru
4 Mycobacteriology Unit, Prince Leopold Institute of Tropical Medicine, Antwerp, Belgium
Correspondence
Christian A. Ganoza
03896{at}upch.edu.pe
Received 15 February 2008
Accepted 12 May 2008
Abbreviations: AFB, acid-fast bacilli; DC, decontamination and concentration; HS–SH, hypertonic saline–sodium hydroxide; NALC-NaOH, N-acetyl-L-cysteine–sodium citrate–sodium hydroxide; NTP, National Tuberculosis Control Program; TB, tuberculosis.
Present address: Department of Immunology, Max Planck Institute for Infection Biology, Charitéplatz 1, 10117 Berlin, Germany. 
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
The gold standard for pulmonary TB diagnosis remains culture, but in developing countries diagnosis relies on microscopy for acid-fast bacilli (AFB) in sputum smears because it is simple, inexpensive and provides rapid results. Unfortunately, this technique has a low sensitivity [22–43 % for a single smear (Toman, 2004) and up to 60 % under optimal conditions (Apers et al., 2003; Siddiqi et al., 2003)] when compared with that of cultures. This sensitivity is even lower in paediatric and human immunodeficiency virus (HIV)/AIDS patients (Karstaedt et al., 1998; Khan & Starke, 1995). The threshold for detection of AFB in sputum samples under optimal conditions is between 104 and 105 bacilli ml–1; TB programme conditions and technical constraints further decrease this yield (IUATLD, 2005).
Decontamination and concentration (DC) methods, such as using N-acetyl-L-cysteine–sodium citrate–sodium hydroxide (NALC-NaOH), improve the yield of AFB microscopy and are routinely used in industrialized countries. The high cost of the reagents and the need for additional specialized training prevent their widespread use in developing countries; the same considerations also hinder the adoption of currently available immunological and molecular diagnostic techniques (Palomino, 2006; Perkins & Kritski, 2002). Therefore, an inexpensive, straightforward DC method would improve case detection in a setting such as Peru.
We have developed a novel DC method that uses hypertonic saline (HS) for mucolysis and sodium hydroxide (SH) for decontamination (Ricaldi & Guerra, 2008), based on the observation that hypertonic saline has been used successfully for the symptomatic treatment of cystic fibrosis (King et al., 1997). This new HS–SH method allowed better recovery of AFB in sputum than direct smears, could detect at least 104 bacilli (ml sputum)–1 and showed no significant difference when compared with NALC-NaOH in vitro in assays using AFB-spiked sputa (Ricaldi & Guerra, 2008). The present study evaluated the diagnostic accuracy and performance of the proposed DC technique in sputum and compared it with the standard NALC-NaOH method and with routine direct (unconcentrated) AFB smears used by the National Tuberculosis Control Program (NTP), in a hospital-based setting.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
14 days according to NTP guidelines), with or without previous diagnosis of pulmonary TB, and before starting appropriate anti-tuberculosis treatment. The NTP Microbiology Laboratory performed AFB detection using the direct technique of Ziehl–Neelsen staining according to national and Pan American Health Organization guidelines under routine conditions; the results were recorded for this study. The other procedures were performed in the Biological Safety Level 3 facility at the Microbiology Laboratory of the Instituto de Medicina Tropical Alexander von Humboldt of the Universidad Peruana Cayetano Heredia, Peru.
Samples were either processed immediately on arrival at the laboratory or stored for less than 24 h at 4 °C before processing. Two 1 ml portions were taken from each homogenized sample. One portion was used for each method. Samples with a volume of less than 2 ml were excluded.
DC methods
HS–SH procedure.
A 1 ml portion of sputum was mixed with 1 ml 7 % (w/v) NaCl and 1 ml 4 % (w/v) NaOH in a sterile 15 ml centrifuge tube (BD Falcon) and homogenized for 15–20 s using a vortex mixer [final concentrations (w/v) in 3 ml: 2.33 % NaCl, 1.33 % NaOH]. The tubes were then incubated at 37 °C for 30 min. After incubation, the mixture was neutralized with sterile PBS (pH 6.8), bringing the total volume to 15 ml. The mixture was vortexed for 5 s and then centrifuged at 3400 g for 15 min at 15 °C, using aerosol-proof shields (Thermo IEC Centra CL3-R, rotor 243; Thermo Scientific). The supernatant was discarded into a splash-proof container with a tuberculocidal solution. The pellet was then resuspended in 200 µl sterile PBS, homogenized for 5 s with a vortex mixer and inoculated into Löwenstein–Jensen (LJ) culture media; a smear was prepared for Ziehl–Neelsen staining.
NALC-NaOH procedure. The procedure used has been described previously (Kent & Kubica, 1985) and is recommended by the Centers for Disease Control and Prevention and the WHO/IUATLD (WHO, 1998). Briefly, 1 ml sputum was added to a 50 ml BD Falcon centrifuge tube with 1 ml of solution containing 0.5 % (w/v) NALC, 2.67 % (w/v) NaOH and 1.45 % (w/v) sodium citrate and mixed well (final concentrations (w/v) in 2 ml: 0.25 % NALC, 1.34 % NaOH, 0.73 % sodium citrate). The tubes were incubated at room temperature for 15 min. After incubation, the mixture was neutralized with PBS, bringing the total volume to 50 ml. The rest of the procedure was as described above.
Culture. Immediately after the DC process, 100 µl each resuspended pellet was used to inoculate a tube of LJ medium. The culture tubes were incubated at 37 °C for up to 8 weeks before being reported as negative. Mycobacterium sp. isolates were identified using the following criteria: time to visible growth, growth at 37 °C, colony pigmentation and morphology, AFB smear and niacin strip test (Remel) (Kent & Kubica, 1985).
AFB smears. One drop of each suspended pellet was used to prepare slides for AFB microscopy using the Ziehl–Neelsen stain. Each slide was coded, read blindly by a qualified technician and reported according to the NTP and WHO/IUATLD standards (WHO, 1998). Smears were reported as follows: <1+, 1–9 AFB in 100 microscopic fields (few bacilli); 1+, 10–99 AFB in 100 fields; 2+, 1–10 AFB per field in at least 50 fields; 3+, more than 10 AFB per field in at least 20 fields.
Statistical analysis.
The values for sensitivity, specificity and efficiency were calculated using a 2x2 table; the degree of association (using the kappa statistic) and difference of proportions and their significance were calculated using McNemar's 
2 and the Z test, using SPSS 11.0.3 for Macintosh (SPSS). When comparing the results of AFB smears, only their negativity or positivity for AFB was used because of the expected difference in quantitative results between the concentrated and unconcentrated portions of the same sample.
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
. From the 106 samples included in the study, 12 were excluded because of contamination of their cultures [11 cultures contaminated with the NALC-NaOH method, one contaminated with both methods; 13/212 (6.13 %) LJ tubes contaminated in total]. The remaining 94 samples were evaluated.
|
). Agreement and reproducibility between the two DC methods in cultures was excellent (kappa index=0.794, 95 % CI=0.638–0.951). DC using the HS–SH method yielded 4.3 % more positive cultures than the NALC-NaOH method; this difference was not significant (95 % CI=–0.8 % to +9.3 %; McNemar's 2=2.67, P=0.1) and did not produce a significant increase in sensitivity (Table 2).
|
). The AFB counts on the smears of these samples had a total of 1–9 AFB in 100 high-power fields or over the whole slide, suggesting that the bacterial load in these specimens was at or near the limit of sensitivity of the culture (<1+ in the NTP format; data not shown). Previous observations have indicated that, in samples with bacillary counts of less than 104 bacilli ml–1, the sensitivity of culture for mycobacteria using LJ medium is suboptimal, ranging from 67 % to a maximum of 95 % (Levy et al., 1989; Long, 2001). This may be due, for example, to the high toxicity of NaOH, the loss of bacilli when decanting the supernatant of centrifuged samples or the presence of damaged bacteria, all reducing the final recovery of bacilli, which has been estimated to be up to 60 % of the initial total number under optimal conditions (Allen & Mitchison, 1992; Kent & Kubica, 1985). There were 10.4 % more contaminated cultures by NALC-NaOH than by HS–SH (95 % CI=4.6–16.2, McNemar's 2=11, P<0.05); this statistically significant difference favoured the HS–SH method. The overall contamination rates (6.13 %, total LJ cultures) are similar to those observed by others, ranging from 1.5 to 13.3 % using LJ medium (Mirovic & Lepsanovic, 2002; Somoskovi & Magyar, 1999).
Sensitivity of AFB microscopy is improved by both DC methods
Of the 21 culture-positive samples, AFB were detected in 15 (71.4 %) of the slides processed by the HS–SH DC method and 14 (66.7 %) by the NALC-NaOH method (Table 2). In contrast, the direct method performed at the NTP Microbiology Laboratory detected AFB in only six (28.6 %) of the culture-positive samples. The sensitivity, specificity and efficiency of AFB microscopy after processing by the HS–SH and NALC-NaOH DC methods and for the direct AFB microscopy processed at the NTP Microbiology Laboratory were calculated by comparing their results with the total number of positive cases as determined by culture. The sensitivity of AFB microscopy was increased by the two DC methods compared with the direct microscopy method; this increase was statistically significant (Table 2).
The results (positivity or negativity) of the AFB smears prepared after the two DC methods had a high correlation: 81/94 (86.2 %) were concordant and all of the 13 samples that were discordant had an AFB count of <1+ (data not shown). The kappa value of agreement of the two AFB microscopy methods was 0.631 (95 % CI=0.445–0.817). DC using the HS–SH method for AFB microscopy yielded 1.1 % more positive results than the NALC-NaOH method; this difference was not significant (95 % CI=–6.5 % to +8.6 %; McNemar's 2=0.08, P=0.78).
Of the 21 culture-positive samples, 16 (76.2 %) had a positive AFB smear in one or both DC methods; the NTP's direct AFB smears missed 10 culture-positive samples. Microscopy using the NALC-NaOH and HS–SH DC methods detected 23 and 24 of the 30 samples that had AFB in at least one of the DC methods, respectively, whilst microscopy with the direct method used by the NTP detected only six AFB-positive samples; all smears were read by the same skilled technician. This indicated that 24 out of the 30 (80 %) smear-positive samples detected by at least one DC method were missed by the NTP's direct microscopy method. Twelve of these missed samples had a low AFB count (<1+), even after concentration (data not shown); the rate of false-negatives with the direct AFB smear technique was greater in samples with low bacillary counts, a common observation also made by others (Marei et al., 2003; Peterson et al., 1999; Rattan et al., 1994; Van Deun et al., 2000). It should be noted that the gain in sensitivity using DC methods may also indicate poor routine direct AFB smear processing by the NTP, as a 29 % yield for AFB microscopy is low in high-incidence settings.
Advantages of the HS–SH DC method
Accurate case detection is essential for the directly observed treatment short-course (DOTS) TB control strategy. All efforts to control TB depend on early detection to decrease transmission. The HS–SH technique substantially increased the sensitivity of AFB smears compared with the NTP's routine direct smear for AFB and correlated well with the NALC-NaOH DC method for sputum culture in suspected TB patients.
Many techniques have been developed to improve the sensitivity and timely detection of AFB in sputum. Among the procedures that one can find in the literature, most use concentration techniques and a chemical decontamination step. Some achieve sensitivities and specificities similar to those of the standard NALC-NaOH method, but they use expensive reagents and equipment. Methods using bleach (NaOCl) and xylol show similar results for microscopic detection of AFB, but they both kill mycobacteria, which is a disadvantage because in specific populations at risk for multidrug-resistant TB, culture is mandatory (Steingart et al., 2006b). An added advantage of the HS–SH method over the NALC-NaOH method is the much lower cost and the greater stability of NaCl solutions. The mixture containing NALC has to be prepared daily and must be discarded if unused (Kent & Kubica, 1985); in contrast, the NaCl solution can be kept at room temperature for several days without change. The proposed HS–SH method allows the use of 15 ml sterile centrifuge tubes, which are less expensive and more convenient to use in ordinary centrifuges than the 50 ml tubes recommended for the NALC-NaOH method. Another advantage of the proposed method is that the final pellet can also be stored for later use in drug-susceptibility testing.
The greatest limitation of the proposed method is the need for a high-speed centrifuge with aerosol-proof capped tube holders. This can be a problem in low-resource settings, but the considerable increase in patient detection may improve the efficiency of the overall control programme by reducing the number of undetected TB-positive carriers, who would otherwise continue spreading the disease.
This study was limited by a relatively small sample size and by a lack of information about the HIV status of the pulmonary TB suspects. Also, we only recovered M. tuberculosis, probably because of the low HIV prevalence (
2 per 100 000) (WHO, 2005) and a low non-TB mycobacterial prevalence in Peru (M. Cabello, personal communication). Despite these limitations, the advantages show that the HS–SH technique could improve diagnosis in difficult situations, such as in paediatric, HIV/AIDS and other paucibacillary pulmonary TB patients. Along with more sensitive techniques such as fluorescence microscopy (Steingart et al., 2006a) and rapid culture methods, this DC procedure could improve the care and prognosis of TB patients by providing a more accurate and timely diagnosis.
The HS–SH DC method offers major improvements in sensitivity of microscopy for AFB compared with the direct AFB smear method and its performance for culture is comparable to that of the recommended NALC-NaOH method with fewer contaminated cultures. The proposed HS–SH method is simpler and less expensive than the recommended NALC-NaOH DC method and is therefore a promising candidate for evaluation and adoption by national TB control programmes in developing countries.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
