Thin-layer agar ( TL 7 H 11 ) for rapid isolation of Mycobacterium tuberculosis in sputum specimens

Background: Tuberculosis (TB) remains one of the major causes of death from a single infectious agent worldwide. The early detection of new cases of pulmonary tuberculosis is an important goal in tuberculosis control program. Objective: In this study, thin layer agar (TLA) culture was compaxed with Lowenstein-Jensen (LJ) culture for rapid detection of pulmonary tuberculosis. Methods: It was a cross sectional study conducted in National Tuberculosis Reference Laboratory (NTRL) of National Institute of Disease of Chest and Hospital (NIDCH), Dhaka, from July 2010 to June 201 1. A total of 100 sputum smear positive for acid fast bacilli (AFB) by Z-N staining, pulmonary tuberculosis patients were included in this study. Samples were processed by modified Petroff method and then cultured on thin layer 7H11(TL7H11) plates and L-J tubes .TLTHll plates were observed microscopically for microcolony growth once a week .for 6 weeks, and L-J tubes were observed once a week for 8 weeks. Results: The recovery rates of mycobacteria on only TLA, only LJ and on both media were g0o ,gTyoand 88% respectively. Overall positivity was 99% in both L-J and TLA media. Mean time for detection ofmycobacteria on TLA was9.04+1.66 days comparedto21.78+6.19 days on L-J media. The rate of contamination was higher (6%) inL-J media than in TLA media (4%). Conclusion: The TL7H11 media can be used as an alternative to the Lowenstein-Jensen medium for early isolation of mycobacteria in resource constrained settings.

Tuberculosis seenario of Bangladesh is not satisfactory at all as we stand in the 6th position amongst 22 MTB burdened countries3.According to the WHO surveillance report 2009 the estimated annual incidence of all TB cases are 225 per 100,000 people.The prevalence of all cases is 425 and mortality is 51 per 100,000 peoplea.The most apprehensive information is that about 50% adult population of Bangladesh are infected by Mycobacterium tuber- culosis.Direct microscopy and T owenstein-Jensen (L-J) medium are still considered the gold standard for the diagnosis of TB in low-income countriess.However, these methods have either low sensitivity, especially in pauci- bacillary pulmofiary and extra-pulmonary TB, or take a longer time (3-4 weeks) ,f,-or mycob actertal culture.Thinlayer agar and MODS (Microscopic observation broth drog susceptibility) are two promising modalities that Introduction: Despite the discovery of the tubercle bacillus more than a hundred years &Eo, and,all the advances in our knowledge of the disease made since then, tuberculosis still remains one of the major health problems facing mankind, particularly in developing countries.Presently, about one third of the world's population is infected with Mycobacterium tuberculosis.It is estimated that currently there are about 10 million new cases of tuberculosis every year with 3 million deaths occurring world-widel.Currently, more people die of tuberculosis than from arry other infectious disease.Death from tuberculosis comprises 25% of all avoidable deaths in developing countries.Ne arly 95% of all tuberculosis cases and 98% of deaths due to tuberculo-Address for Correspondence: Dr. Habiba Binte Alam Medical Officer Sir Salimullah Medical College, Dhaka, Bangladesh.E-mail : koly_hasan@yahoo.comMobile no-0 17 16582563 detect the presence of mycob actefia by microscopic observation of colonies.Both the methods can detect the colonies microscopi calty within l-2 weeks and can also give a presumptive identification of mycobacteria based on their characteristic colony morphology 6'7.The objective of the present study was to evaluate the TLA medium in terms of recovery rate, time for detection and contamination rate, comparing it with the Lowenstein-Jensen media.

Methods:
Ethical clearance was taken from Institutional review board of BSMMU.A cross sectional study conducted in National Tuberculosis Reference Lfuoratory NTRL) of National Institute of Disease of Chest and Hospital (NIDCH), Dhaka, from July 2010 to June 2011.Clinically suspected pulmonary tuberculosis patients referred to the NTRL by physicians were enrolled for sputum examination.Patient positive for AFB by sputum smear Z-N stain were considered as pulm onary fuberculosis and were included in this study.A total of 100 patients of different age and different categories of treatment were enrolled in this study.Z-N staining was done on all the sputum samples and was processed for digestion and decontamination by modified Petroff method as described by WHO for TLA and L-J culture.A predesigned data collection sheet was used as research instrument.
TLA medium : The TLA technique uses Middlebrook T}Jll agar for the detection of microcolonies of various mycob acteria.About 6m1 of the autoclaved 7H11 medium was poured into one out of 3 divided compartment of 96x15 mm plastic petri dish, L0% OADC (Oleic acid, Albumin, Dextrose, Catalase) and antibiotics, including Polymyxin B, Ticarcillin, Amphotericin and Trimetho- prim, were added at concentrations of 200,000 unttlL, 100mg lL,l.0mg/Lrespectively.A second set of 7HI1 medium containing 500 pg/ml of Paranitrob enzoLc acid (PNB) was prepared to differentiate between Mycobacterium tuberculosis complex and non-tuberculous mycobacteria (NTM); 0.1m1 of the concentrated specimen was 40 inoculated onto both TLA and PNB +TLA.A11 the plates were sealed with adhesive tape, leaving afi area of about 2 cm for ventilation, and incubated at 37'C in a 5-8% CO2 incubator for 6 weeks.The plates were examined for appearance of microcolonies (cording) once weekly by light microscopy.On appearance of visible colonies, colony morphology, rate of growth and pigment produc- tions were studied.After that the growths were confirmed by acid fast staining with Z-N method.
L-J medium: Three drops of pellet of each sputum sample after digestion and decontamrnation were inoculated into two slants of two set of Lowenstein-Jensen media using pasture pipettes.The inoculated media were incubated at 37"C in a slanted position with their screw caps loose for at least one day to ensure even distribution of inoculation.
Then those were kept upright with caps tightened to minimize evaporation and drying of media.The bottles were examined after 72 hours to exclude contamination andif there was no contamrnation, the bottles were exam- ined weekly up to 8 weekst.If no growth observed after 8 weeks of incubation then the culfure was reported as "no growth".In case of contaminated samples, result was interpreted from the other boffles.On appearance of visible colonies, colony morpholo gy, rate of growth and pigment productions were studied.Mycobacterium tuberculosis produce dry duff to cream colored rough surfaced colonies.The growths were confirmed by Z-N staining.
After collection, data was checked for inadequ&cy, irrelevancy, and inconsistency.A11 datawas analyzed with appropriate statistical stools and SPSS 15 program and presented as text, table and figure.

Results:
Out of 100 study population, 74 (74%) were new cases and 26 (26%) were re-treatment cases.A total of 99 (99%) mycob actertal isolates (M.tuberculosis, rr99; NTM-O) were recovered from the 100 AFB smear positive qpeci- mens.Rates of recovery of mycobacteria on TLA media and L-J media were ,90oA and 97% respectively.No growth was observed in 3 smear positive sputum in L-J media.Out of these 3 sputum samples, 2 specimens Thin-laver 7lil1) for rapid isolation of Mycobacterium tuberculosis in showed growth of mycobacterium in TLA media.Ten smear positive sputum samples yielded no growth in TLA media.Among them, 9 samples showed growth of myco- bacterium in L-J media but one sample did not grow in L-J media.culture positivity in both media was BB% (Tabel-I).Figure 1 shows a typical cord formation charac- teristic for M. tuberculosis on TLA. Figure 2 shows comparing time to growth detection among samples with different bacillary loads.The time for detection of positive cultures in TLA was g.04 days which was faster than L-J 21.78 days (p<0'05).Rate of contarnination observed in L-J media was 6% and in TLA media was 4%. (either in TLA or in L-J media) was 99%.Similar finding was observed in a study ofAmencain 1993 and in Colom- bia in 2004 tL' t2.In their studies isolation rate of MTB in TLA media was 83% and 74% and in L-J media was 93% and 9I% respectively.On the contrary another study in Spain by Idigoras et aI (1995), overall isolation rate of MTB in TLA and L-J media was g0% and 77% respectivelyl3.In house preparation of TLA medium rather than obtained from commercial source and incubating in 5% CO, might be the reasons for higher isolation rate of MTB in TLA media.In this study, it was observed that in average after 20 days of incubation, TLA media became dry and shrunken and sometimes cracked.Some relatively slow growing mycobacteria might be missed for these reasons in TLA media.
In present study, TLA showed statistically signific ant shorter growth detection time than L-J media ( TLA media observed under tight microscope (otbjective media was 11 days and in L-J media was 23 days had also been reported by Welch et al (1993) in Amertcatl.It was not found any study contrary to this study, regarding growth detection time.Shorter growth detection time of MTB in TLA media might be due to growth supplements such as Oleic acid, Albumin, Dextrose and catalase (OADC), varieties of inorganic salts, glycerol and CO, containing atmosphere.
Mean growth detection time for MTB varied depending on the bacillary load in sputum in both TLA and L-J media.
But the variation of time was statistically not significant in the present study.Robledo et al (2A0Q in LatinAmerica, showed similar finding 6.They found mean growth detec- tion time of MTB in grade 1+, 2+ and 3+ specimen in TLA media was 8 days, 7 .5days and 7 days respectively.In L-J media mean growth detection time was 28 days, 22 days and 2l days respectively.Relatively shorter mean growth detection time in higher bacillary load is natural and need not to be explained. In present sfudy rate of contamination in L-J media was 60/o and in TLA media was 4%.However, Martin with his associates (2009) revealed that, the contamination rate was 16% and 26% respectively in TLA mediar4.Contamination rute was much higher in these studies than this one.
The probable cause may be that in these studies they used Trimethoprim, Amphotericin and Piperacillin antibiotics (TAP) in TLA media for prevention of contamination.
Due to unavailability of TAP preparation in Bangladesh in present study Polymyxin B, Ticarcillin, Amphotericin, Trimethoprim (MAST selectatab) were used.Before introducing MAST selectatab apilot study was carried out to evaluate its inhibitory effects on the growth of Myco- bacterium tuberculosis and its inhibitory effects on contaminating organisms.. A11 the samples were processed freshly within 24 hours after collection, to minim tze c ontaminati on.

Conclusion:
TLA media was evaluated in this study for isolation of MTB.Results of this study revealed that TLA media is a reliable alternative to L-J media for the rapid isolation of M.tuberculosis .Rate of contamination was also low in TLA media and it required no additional equipments or materials.The simplicity of this one-plate format may favor its implementation and applicability in mycobacterral diagnostic laboratories, especially in resource limited countries like Bangladesh.

Table - I
Growth of Mycobacterium tuberculosis in LJ and TLA media, duration of growth detection time and rate of contamination (r: I 00).Early diagnosis and prompt treatment have become a priority due to the rapid increase in the spread of TB world wide.Various studies have been carried out in various contexts to evaluate the performance of TLAe'10.Higher isolation rate of Mycobacterium fuberculosis was observed in L-J media in comparison with TLA media (97% versus 90%) in this study.Combined isolation rate