Editor,—I read the article on the comparison of serum, salivary, and rapid whole blood diagnostic tests forHelicobacter pylori and their validation against endoscopy based tests, by Reilly et al (Gut1997;40:454–8) with considerable interest because, in addition to other comments, they would not recommend the routine use of salivary antibodies as a method of choosing appropriate endoscopy inH pylori positive patients being considered for this test.

Our results on patients with peptic ulcer, published in 1996,1 suggest that saliva urease testing seems to be the preferred method for screening purposes and for following the effectiveness of eradication therapy for H pyloriinfection. Indeed, we followed two groups of patients with peptic ulcer treated for H pylori infection to see, mainly, whether monitoring treatment by the non-endoscopic means of a urease test and serology was informative. All patients were treated with amoxycillin and lansoprazole and followed mainly using the saliva urease test (CLO, Delta West) and serology (IgG anti-H pylori by using an ELISA) after one and three months’ treatment in the first group (33 patients), and after three months’ treatment in the second group (60 patients). In the first group, H pylori was detected by the urease test in saliva and gastric mucosa and by a rise in serum IgG antibody levels against H pylori in 100%, 91% and 88% of cases, respectively, before treatment. H pylori was detected histologically in 45% of the cases, including the three patients whose gastric mucosa was H pylori negative. After one month of treatment the saliva test was negative and there was a >50% drop in serum IgG antibodies, predicting eradication of H pylori,2 ,3 in 55% and 52% of patients respectively. After three months these values were at 67% and 67% respectively. These findings were associated with clinical or endoscopic, or both, remission of ulcer disease and a negative gastric mucosa urease test. A similar profile was also obtained before and after three months’ treatment in the second group of patients. The saliva test was easy to perform and a positive result was available within three hours. Therefore, because of the simplicity and cost effectiveness of the saliva test, this method is preferable to endoscopy or serology, or both, particularly where the availability of endoscopy and serological testing kits is limited or in certain cases (for example, patients with coronary heart disease) where the endoscopic procedure may put the patient at risk. This method is currently applied by us as a routine diagnostic test in these patients, particularly in dyspeptic patients with symptoms suggestive of peptic ulcer disease, with one exception: when the symptomatology indicates malignancy, endoscopy is the diagnostic procedure of choice.

An important question is whether the urease detected in the saliva was really produced by H pylori or by other organisms usually present in the human mouth. It has been proposed4 that if a urease test is positive within the first three hours, it is definitely due to the presence of H pylori because the other urease producing organisms (such as Proteus spp) have almost 100 times less urease activity than H pylori. Therefore, as all the infected saliva samples as well as the gastric mucosal samples were positive within the first three hours, we believe that H pylori was the organism responsible for producing the positive results.

In analysing more recent results with an intention to treat H pylori positive patients with omeprazole, amoxycillin, clarithromycin, and vitamin E or N-acetylcisteine, we observed an increased (90%) ulcer healing rate, with a similar picture on testing saliva. Finally, by comparing the salivary urease and anti-H pylori tests, we noticed that the saliva urease method was more accurate. Moreover, we know that the presence of antibodies to H pylori can indicate both current and past infection.5 ,6