LETTER

Top Letter References

Chong et al. recently evaluated the use of primers Hp1 and Hp2 for detection of Helicobacter pylori (1). They found many unexpectedly positive PCR results by using these two primers for detection of H. pylori from clinical specimens; these results were attributed to false amplification from the human genome. In their study, the 109-bp products were detected in all the gastric biopsy specimens obtained from the CLO tests (Delta West, Bentley, Western Australia, Australia) of 40 patients, irrespective of their CLO test result. This was very different from our experience.

We examined by PCR specimens from patients with bleeding ulcers which gave false-negative test results. Forty patients were determined to be positive for H. pylori by histology and/or the [¹³C]urea breath test but negative by the rapid urease test (CLO test). The gastric biopsy specimens with these false-negative urease test results were taken out and processed for DNA extraction by using a standard proteinase K and phenol-chloroform extraction protocol. The PCR amplification protocol was slightly different from that described by Chong et al. Nested PCR using two sets of primers was performed as described previously (2, 3). Thirty cycles of amplification with Hp1 and Hp3 (5'-AGG ATG AAG GTT TAA GGA TT-3', positions 407 to 426) at an annealing temperature of 55°C were followed by another 30 cycles with Hp1 and Hp2 at an annealing temperature of 62°C. Five gastric biopsy specimens which gave genuine positive and negative CLO test results were used as controls. Among the 40 test specimens, only 3 showed the 109-bp amplicons after the nested PCR. In the control group, only the three positive controls showed the expected PCR product; none of the negative-control specimens showed the product.

The nested PCR that was originally described by Ho et al. had a much higher sensitivity than the single-step amplification used by Chong et al. (2, 3). However, the 109-bp PCR product was detected only in three gastric biopsy specimens which were supposed to be positive. False-positive amplification was not encountered with the use of this set of primers in our study. If human genomic DNA caused false-positive results or nonspecific amplification, more positive results should be expected. Our results were more consistent with previous experiences of other investigators (3-5). The results of Chong et al. were very unexpected and probably cannot be explained by nonspecific amplification from human DNA alone.