Evaluation of a Rapid Air Thermal Cycler for Detection of Mycobacterium tuberculosis

Chapin and Lauderdale (1) describe their evaluation of a rapid air thermal cycler (ATC) (Idaho Technology, Idaho Falls, ID) for the detection of Mycobacterium tuberculosis by PCR. We also have compared the use of the ATC to that of a more conventional heat block thermocycler (HBTC) (GeneAmp PCR System 9600; Perkin Elmer, Cheshire, England) for the direct detection of M. tuberculosis in clinical samples. Our findings support the general conclusions of Chapin and Lauderdale but differ significantly in regard to inhibitors.

Samples were prepared by sonication using glass beads, and we used a 123-bp sequence of IS6110 as the target for DNA amplification, as previously described (4). However, we adapted the PCR cycling parameters for use with both the ATC and HBTC (Table 1). Since PCR of clinical samples has been hindered by inhibitors present in 3 to 20% of specimens (2, 5), each PCR mixture included an internal amplification control of 169 bp (Novocastra Laboratories Ltd., Newcastle, England) which underwent coamplification. To date, we have tested 26 clinical samples (including 24 sputum and 2 bronchoalveolar lavage samples) from 18 patients by PCR using both cyclers and compared the results with those of microscopy and culture (Table 2).

TABLE 1.

PCR cycling parameters

Cycler	Initial step	Cycling parameters	No. of cycles	Cycle time
ATC	94°C, 1 min	94°C, 0 s; 50°C, 0 s; 72°C, 2 s	45	23 min
HBTC	94°C, 1 min	68°C, 30 s; 72°C, 30 s; 94°C, 15 s	15	1 h 20 min
		60°C, 30 s; 72°C, 30 s; 94°C, 15 s	15
		60°C, 30 s; 72°C, 5 min	1

TABLE 2.

Comparison of microscopy, culture, and PCR for detection of M. tuberculosis in 26 clinical samples

Sample status (n)	ATC		HBTC
	No. of samples PCR+	No. of samples with inhibitors	No. of samples PCR+	No. of samples with inhibitors
Smear+, culture+ (8)	8	0	7	1a
Smear−, culture+ (6)	2	0	1	2a
Smear+, culture− (2)b	0	0	0	0
Smear−, culture− (10)	0	1	0	7

^aIncludes one specimen containing inhibitors not removed by chloroform treatment; PCR therefore nonevaluable. 

^bSpecimens grew M. malmoense. 

On initial testing, inhibitors (as evidenced by nonamplification of the internal control) were detected in 1 sample (3.9%) with the ATC and 10 samples (38.5%) with the HBTC. Following further purification by simple chloroform extraction (3), 2 samples still had evidence of inhibitors with the HBTC. Both of these samples were therefore nonevaluable by using the HBTC, and both were M. tuberculosis culture positive.

By using culture as the “gold standard” for M. tuberculosis, of 14 culture-positive samples studied, 10 (71.4%) were PCR positive with the ATC and 8 (57.1%) were PCR positive with the HBTC, following chloroform purification where necessary. All the culture-positive but PCR-negative samples were smear negative, suggesting a low bacillary load, although 2 of the 10 samples found to be PCR positive with the ATC were also smear negative, as was one of the 8 successfully amplified samples with the HBTC.

Of 12 culture-negative samples examined (2 of which contained M. malmoense), all were negative with both cyclers.

We therefore agree with Chapin and Lauderdale that the ATC is an excellent alternative to the HBTC in decreasing both overall cost and total assay time (3 versus 4.5 h). However, in our limited series, and in contrast to what Chapin and Lauderdale found, despite the smaller sample input for the ATC (1 versus 5 μl), it not only provided an assay as sensitive as the HBTC but also had fewer problems with inhibitors.