Ferritinaemia in cancer

Abstract

SMALL amounts of ferritin are found in serum in normal and pathological states^1,2, usually at levels directly related to the amount of tissue storage iron^3,4. This relationship, however, does not hold in many cancers where grossly elevated levels can occur without a corresponding increase in iron stores^5–9. These latter findings suggest a possible use of ferritinaemia in cancer diagnosis but the value of this test has not been clearly established Such use is complicated by the multiple potential sources of the serum ferritin which could affect its type as well as its amount. Much of the ferritin in cancer sera may reflect increased iron stores arising from chronic anaemia or from transfusions in treated patients. Some may also come from nonspecific tissue damage. The high levels of ferritin synthesis in some cancer cells¹⁰ and the presence in serum of isoferritins characteristic of some cancer and foetal tissues^11,12, however suggest that much of the ferritin may come directly from cancer cells. If so, selective quantitation of these isoferritins may provide specificity in possible serodiagnosis. The multiple isoferritins found in most human tissues seem to be hybrid molecules composed of different proportions of two subunit types¹³. One type predominates in liver and spleen isoferritins, and the other in the more acidic isoferritins in heart, tumours and HeLa cells¹⁴. Different isoferritin populations can be distinguished immunologically, apparently on the basis of their subunit composition^14,15. Ferritin is usually assayed by immunological methods, with crystalline liver or spleen ferritins as immunising antigens and reference standards. These ferritins consist almost entirely of the liver-type subunit¹⁵. Consequently, most assays based on these ferritins have only a low reactivity for the more acidic ‘carcinofoetal’ isoferritins which may contain less than 20% of the liver-type subunit^14,15. We have, therefore, tested a variety of cancer sera and compared the apparent levels of ferritin given by a standard ‘liver-type’ assay with levels given by an assay whose specificity is directed against the more acidic isoferritins¹⁵. We report here preliminary evidence that the latter type of assay may significantly improve the usefulness of ferritinaemia as a tumour marker.