Tumour markers are substances which can be measured, commonly in blood or urine, to assist in the detection and monitoring of cancers. This paper summarises features of some commonly-used tumour markers.

The ideal tumour marker should be (1) specific for the cancer for which it is testing; (2) not present in any other conditions; and (3) the concentration should change with the amount of malignant tissue present. An ideal tumour maker could be used for screening, diagnosis and monitoring of disease progression. Unfortunately there are no ideal tumour markers and most are restricted to monitoring cancers once they have been detected and diagnosed using other methods.

There are no tumour makers which are currently recommended for screening of the general population. The most likely candidate is PSA as a screen for prostate cancer, however there is no agreement about whether screening of the general population reduces premature mortality. Most tumour markers have too many false positives from benign conditions to make screening feasible (see table over). Many only clearly identify malignancy once the cancer is sufficiently advanced to make this detection of limited use. There are some circumstances where screening is useful in defined populations. The use of AFP to screen for hepatocellular carcinoma in patients with chronic liver disease or calcitonin in patients at high risk of thyroid cancer are examples of this.

Tumour markers are occasionally useful as a pointer towards a diagnosis but never replace the need for formal tissue diagnosis. Urine catecholamines and their metabolites can strongly suggest a phaeochromocytoma and very high levels of some serum markers make some cancers exceedingly likely (eg very high CA 125 and ovarian cancer). It is very important that tumour markers should never be used on their own to make a diagnosis of malignancy.

This is the main clinical use for tumour markers. Appropriate markers should be measured after the diagnosis of a cancer and prior to commencement of treatment. Regular measurement of a marker shown to be elevated in that patient may then assist in demonstrating the effectiveness of a treatment intervention (medical, surgical or with radiotherapy) by showing a reduction in the levels of the marker after therapy. The timing of such measurements should be guided by the half-life of the marker. In the longer term markers can be used to try and identify disease recurrence or progression prior to the appearance of clinical signs or symptoms which may allow earlier treatment of such a condition.