The N-nitrosoproline test as a measure of cancer risk in geographical comparison studies: results from Italy and an overall comparison.

The N-nitrosoproline (NPRO) test has been used in studies in which populations at high risk of cancer have been compared with equivalent populations at lower risk, to examine whether the geographical variation in cancer risk correlates with propensity for endogenous nitrosation, as assessed by the NPRO test. The usual method employed has been to determine NPRO in 12- or 24-h urine samples, after ingestion of L-proline, in a representative sample of the general population. We present results from one such geographical study conducted in two regions of Italy (Florence and Cagliari) with an approximately three-fold variation in gastric cancer mortality. The nonsignificant difference in mean NPRO excretion between the two populations was insufficient to explain the difference in cancer risk. The fact that there are appreciable international differences in formation of NPRO suggests, firstly, that nitrosation may be of relevance to cancer risk in some countries but not in others and, secondly, that variations within one country may not be large enough for significant geographical differences to be evident. Multivariate analysis of individual, rather than grouped, results from our Italian study made it possible to quantify the relevance of different factors to NPRO formation: a major factor is exposure to nitrate. Important relationships may be missed by analysing only grouped data.