Rapid culture of Mycobacterium tuberculosis on blood agar in resource limited setting

MATERIAL AND METHODS

Blood was collected from jugular vein of sheep and 25 ml of it was mixed with 5 ml of sodium citrate (3.8%) solution. It was mixed well and stored at 2-8 °C, 100 ml of deionized water with 4.4 g of Columbia blood agar base (Central Drug House Ltd. New Delhi, India) and 1 ml of crystal violet solution (0.01%) was boiled to dissolve the medium completely. It was sterilized by autoclaving at 15 lbs pressure (121 °C) for 15 minutes and cooled to 45-50 °C, 7 ml of citrated sheep blood was mixed in it. To avoid growth of contaminants two solutions were added to the sample: 1) 90 microlitres of nystatin solution (prepared by dissolving 18 mg nystatin dissolved in 9 ml methanol) and 2) 50 microlitres of antibiotic solution (prepared by dissolving 88.8 mg Polymixin-B, 5 mg trimethoprim and 20 mg nalidixic acid). Contents were mixed well by vortexing for a minute and approximately 12-13 ml was quickly dispensed to each of 30 ml, flat bottom culture tubes (McCartney bottles).

These bottles were kept obliquely and allowed to cool till slants of blood agar were ready. Instead of using petri plates, we used McCartney bottles for blood agar to avoid desiccation. These bottles were incubated at 37 °C for 24 hours for sterility check and those showing any growth were discarded. Remaining blood agar slants were then stored at 2-8 °C. LJ medium was prepared as per WHO guidelines [1] using locally available ingredients.

Smear positive sputum samples of cases of pulmonary tuberculosis were obtained from laboratory of K. N. Chest Hospital Jodhpur on the day of collection from patient. These were processed with the N-acetyl L-cysteine sodium hydroxide (NALC-NaOH) method for homogenisation and decontamination. In brief equal volumes of sodium hydroxide (4%) and sodium citrate (2.9%) solutions were mixed. To this mixture N-acteyl-L-cysteine was added to make a final concentration of 0.5%; 5 ml of this mixture was added to 5 ml sputum; contents were mixed well and kept at room temperature for 15 minutes. Then these were centrifuged for 15 minutes at 3000 rpm. Supernatant was discarded in 5% Lysol and sediment was washed with deionized water. The sediment pellet was resuspended in 7H9 Middlebrook broth or deionized water by vortexing to make a homogeneous inoculum; 50 microlitres of this inoculum was inoculated on LJ slants as well as blood agar slants in triplicate. These McCartney bottles were sealed with parafilm.

The inoculated slants were incubated at 37 °C for eight weeks. These were observed daily for appearance of macroscopic growth. The growths observed on LJ medium as well as blood agar slants were observed after keeping the slants in diffuse sunlight for about one hour and no pigmentation was ever observed. Time to grow macroscopic colonies on LJ slants and blood agar slants as monitored by naked eye examination of McCartney bottles was recorded. Paired t test was used to compare number of days taken in appearance of visible growth on blood agar and LJ medium. Smears prepared from colonies grown were stained by the Ziehl-Neelsen method and were examined microscopically.

RESULTS

After treating with the NALC-NaOH method, 70 sputum samples were inoculated on three blood agar (7% sheep blood) slants as well as three slants of control LJ medium. Growth patterns were not different within three slants of any medium. However Mtb colonies were light grayish, glistening and easily recognized against red blood agar as compared to cream-coloured rough colonies on the green LJ medium. Colonies did not appear on blood agar from four sputum specimens which were smear grade +1 and showed colonies on LJ slants. Two sputum samples did not grow colonies of MTB on LJ medium, and showed growth on blood agar only. LJ medium recovered 68/70 isolates (97.1%) as compared to 66/70 by blood agar (94.2%). The difference was not significant. Rate of contamination was 1.6% on blood agar and 7.8% on LJ medium, the difference may be because of antibiotics added to blood agar; 64 sputum samples grew visible colonies of Mtb on LJ medium as well as blood agar. Time to grow macroscopic colonies on LJ medium and blood agar was compared in these samples.

Mean time to detect macroscopic colonies of Mtb on blood agar was 13.6 ±5.2 days as compared to 20.4± 5.1 days on LJ medium (p=0.0001) ( Table 1 ).

The median time to culture positivity of Mtb was 13 days (range: 5-25) on blood-agar slants and 19 days (range: 12-31) on LJ slants (p=0.0001). More colonies were observed on blood agar than on LJ medium.

DISCUSSION

The present study has highlighted that time to detect Mtb from smear positive sputum specimens was found to be significantly shorter with blood agar slants (13.6 ±5.2 days) as compared to LJ medium (20.4± 5.1 days). Many evaluation studies have been carried out comparing performance of automated and semi-automated complex culture systems in terms of detection time for Mtb. Some of them have compared detection time for Mtb from smear positive sputum samples on a particular culture system and on LJ medium [13-21]. Primary isolation of Mtb from smear positive sputum samples by the VersaTREK system took 19.8 ±11.2 days [13], by the Liquid Bio FM (BIO-RAD) medium it took 10.4 (3-33) days [14], by the Biphasic system (Middlebrook 7H11 agar slant+Middlebrook 9H broth) it took 21 ±4.4 days [15], by MGIT 960 it took 9 (7-11) days [21], 11.9 days [16] and 12.6 days [18], by the MB-Check culture system (liquid phase) it took 14.8 ±8.0 days [19], by BACTEC 9000 MB it took 9 (4-12) days [20], and by BACTEC 460 TB it took 7 (3-27) days [21], 7 (4-12) days [20] and 13.8 days [18] in various studies. Looking at these observations made by using sophisticated equipments, expensive kits and skilled manpower, use of blood agar slants appears to be fairly simple and has shown a fairly comparable performance as far as time to detect Mtb is concerned.

Mean time to detect Mtb from smear positive pulmonary sample on LJ slants has been found to be in the range of 19-24 days by most workers [14, 16, 19-21]. Use of blood agar slants in place of LJ medium is likely to reduce this time to approximately two thirds in developing countries. Blood agar slants have shown sensitivity of 94.2% in the present study, which was not significantly different from that of LJ medium and is acceptable in resource limited settings.

The present study has a limitation as it did not assess performance of blood agar slants with smear negative sputa or extrapulmonary specimens. Furthermore performance of blood agar was not compared with BACTEC or other sophisticated systems, which detect Mtb from a significantly higher proportion of clinical samples as compared to LJ medium. Since sensitivity of blood agar slants was comparable to LJ slants in the present study, these may only be a good substitute of LJ medium and not of more sensitive systems like BACTEC, which are not readily available in developing countries. The present work is therefore a small step indicating utility of blood agar in the place of LJ medium for rapid detection of Mtb from sputum in resource limited settings.

Conclusions

Blood agar slants may be a good substitute of LJ medium for rapid detection of Mtb from sputum in resource limited settings, where sophisticated systems like BACTEC are not available. It may save about one third of the time.

Figure 1