Gastric or stomach cancer causes about 800,000 deaths worldwide every year, which makes it the second most common cause of cancer death. Its prevalence, however, varies enormously; it is rare in the West but common in many Far Eastern countries, accounting for over 20% of cancer cases in Korea. Japan has introduced a gastric cancer screening programme for people over 40 using endoscopy or a barium meal. However, many cancers are still diagnosed at a late stage there, partly because these tests are invasive and therefore unpopular. A screening test using biomarkers that is less invasive, easy and cheap to run would be likely to pick up more cases of early cancer.

Most, if not all cases of gastric cancer arise after Helicobacter pylori infection and the resulting chronic atrophic gastritis. This non-malignant condition causes changes in the stomach histology and protein expression, including increases in the expression of a group of small proteins known as trefoil factors (TFF1, TFF2 and TFF3) that are secreted by the gastrointestinal tract. A group of Japanese researchers led by Sachiyo Nomura of the University of Tokyo, Tokyo, Japan has now investigated whether serum levels of these proteins could be used as biomarkers for gastric cancer.

Nomura and his co-workers initially recruited 183 gastric cancer patients treated at the University of Tokyo hospital and 280 healthy controls. Serum levels of the three trefoil factor proteins and several other proteins associated or potentially associated with gastric cancer, including anti- H. pylori immunoglobulin G (IgG) and pepsinogens, were collected from all participants and assayed using the enzyme-linked immunosorbent assay (ELISA) method. The IgG levels were used to diagnose H. pylori infection. The average age of the cancer group was higher than that of the controls, and a higher percentage of cancer patients were female; 62.3% of cancer patients compared to 34.9% of controls tested positive for H. pylori.

The mean values of each protein in patients and controls were compared using a t-test, and receiver operator characteristic (ROC) curves calculated to measure the ability of each protein to distinguish between cancer patients and controls. Firstly, the level of trefoil factor protein 1 (TFF1) was found to correlate strongly with H. pylori infection status, with infected individuals with and without cancer expressing this protein at higher levels than non-infected ones.

Pepsinogen is sometimes used to test for gastric cancer in Japan, with a high serum pepsinogen I/II ratio correlating most strongly with the presence of disease. However, when the trefoil factor proteins were compared with these markers in their ability to distinguish between cancer patients and controls, both TFF1 and TFF3 levels were significantly more discriminatory than the pepsinogen I/II test. The odds ratios for TFF1 and TFF3 were 18.1 (95% CI 10.5-31.0) and 18.1 (95% CI 11.2-29.2) respectively, compared to 5.61 (95% CI 3.53-8.91) for the pepsinogen ratio. Of the single markers, TFF3 was the most discriminatory marker in both H. pylori infected and non-infected individuals, with a sensitivity of 80.9% and a specificity of 81.0%. These results were validated in a further, smaller set of cancer patients between 30 and 70 and age-matched controls.

Despite these promising results, a small number of patients with cancer were not detected using TFF3 levels, and a subset of these cancers could be detected using the pepsinogen test. Nomura and his co-workers concluded, therefore, that if these results are confirmed in prospective studies, TFF3 would be best used as a biomarker in non-invasive screening for gastric cancer alongside pepsinogen.

Reference

Aikou, S., Ohmoto, Y., Gunji, T. and 10 others (2011). Tests for serum levels of trefoil factor family proteins can improve gastric cancer screening. Gastroenterology 141(3): 837-845. doi:10.1053/j.gastro.2011.05.040