The development of cancer screening has led to the discovert of smaller tum
ours and less frequent dissemination to lymph nodes and organs that require
s special techniques for detection. Numerous papers on micrometastases refl
ect a considerable amount of work devoted to detection methods, technical p
roblems and the prognostic value of these lesions. Apart from cytological t
echniques, the pathologist can rely on two methods for the detection of mic
rometastases: serial slicing of paraffin-embedded blocks and immunohistoche
mistry. When these methods are combined, the detection rate is similar to t
hat of biological methods and can attain levels as high as 60% for the sent
inel node with the added vantage of being able to visualise cells. Despite
an impressive body of studies, major disparities are found in detection rat
es and the prognostic value of micrometastases is not firmly established. I
n order to facilitate comparisons and analyses, it is essential to adopt a
common terminology with precise definitions. The UICC advocates the use of
the term micrometastasis which denotes a metastasis smaller than or equal t
o 2 mm in size. The potential aggressiveness of micrometastases is dependen
t on other poorly explored parameters such as the number of cells detected
in the bone marrow or lymph node and the location of micrometastases. The n
ew pTNM classification takes into account this latter parameter and disting
uishes two categories of micrometastases: "isolated tumor cells" located in
the lumen of vessels or sinuses and "micrometastasis" which has already in
vaded an organ. This classification warrants further analysis to determine
the prognostic value of these categories. The next challenging problem cons
ists in determining the key biological properties that account for distant
dissemination.