WASHINGTON – A simple urine test could detect gastric cancer even at an early stage, the test’s developers say.
The test, which looks for the presence of two metalloprotease enzymes labeled ADAM 12 and MMP-9/NGAL had 77.1% sensitivity and 82.9% specificity for gastric cancer when tested in 35 patients with the malignancy and an equal number of healthy controls, reported Dr. Takaya Shimura from the department of surgery at Boston Children’s Hospital and Harvard Medical School in Boston.
Dr. Takaya Shimura
“This study represents the first demonstration of the presence of ADAM 12 and MMO-9/NGAL complex in the urine of gastric cancer patients,” he said at the annual Digestive Disease Week.
Dr. Stephen J. Meltzer of Johns Hopkins University, Baltimore, commented in an interview that the findings are convincing but preliminary.
A randomized clinical trial enrolling a larger number of patients and controls would be required before he would consider screening patients for the enzymes, said Dr. Meltzer, who was not involved in the study and comoderated the meeting session where the results were presented.
ADAM 12 (a disintegrin and metalloprotease 12) and MMP-9 (matrix metalloprotease 9) are both members of a family of enzymes involved in cellular adhesion, invasion, growth, and angiogenesis, Dr. Shimura explained. MMP-9, when complexed with NGAL (neutrophil gelatinase associated lipocalin) is protected from autodegradation.
The investigators, from the lab of Dr. Marsha A. Moses at Boston Children’s Hospital, and their collaborators in Japan had previously reported that MMPs in urine were independent predictors of both organ-confined and metastatic cancer.
Urinary assays are noninvasive, using easily accessed tissues that can be handled simply and inexpensively, making them ideal for cancer detection, Dr, Shimura said.
Current tests for gastric cancer, such as carcinoembryonic antigen (CEA) and cancer antigens (CA) 19-9 and 72-4, have poor sensitivity for detecting advanced disease, and are even worse at spotting early disease, he noted.
To see whether they could improve on the current lot of tests, the investigators enrolled 106 patients in a case-control study, settling eventually, after age and sex matching, on a cohort of 70 patients: 35 with primarily early-stage gastric cancer, and 35 healthy controls.
After screening the urine of participants for about 50 different antigenic proteins, they found that the patients with gastric cancer had significantly higher levels in their urine of both ADAM 12 (P < .001) and the MMP-9/NGAL complex (P = .020).
In a multivariate analysis, they showed that both enzymes were strong, independent predictors of gastric cancer, with an odds ratio for urinary MMO-9/NGAL of 6.71 (P = .002), and an OR of 15.4 for ADAM 12 (P = .002). In contrast, Helicobacter pylori infection was associated with a nonsignificant OR of 2.54.
In a receiver operating characteristic (ROC) analysis, they also found that MMP-9/NGAL was associated with an area-under-the curve (AUC) of 0.657 (P = .024), ADAM 12 was associated with an AUC of 0.757 (P < .001), and that the two combined had an AUC of 0.825 (P < .001).
As noted before, the sensitivity of the combined enzymes was 77%, and the specificity was 83%.
Finally, using immunohistochemical analysis, the investigators were able to show that gastric cancer tissues had high levels of coexpression of MMP-9 and NGAL (P <.001) and high expression levels of ADAM 12 (P < .001), compared with adjacent normal tissues.