And isolated tumor cells

Gabor Cserni

"... there are truths but there is not truth"

A Camus: Le mythe de Sisyphe


This chapter will summarize several aspects of nodal micrometastases in breast cancer. It will briefly deal with the term itself, the methods used for the detection of these low-volume metastases, and some interpretation issues. Finally, it will discuss the prognostic and predictive implications of micrometastases: whether they have a prognostic impact on survival, and whether they are predictive of nonsentinel node involvement when found within a sentinel lymph node. Obviously, predictive, in this sense is different from the general use of this term, but is nevertheless justified.


Metastases (from Greek: changing state/change of state) are tumor deposits away from the primary neoplasm. Although instances of metastasis are known in benign conditions, this is generally the most important hallmark of malignant behavior. The prefix micro- (from Greek: small) suggests that micrometastases are indeed small metastases, and in this context the term reflects that they cannot be identified clinically or by naked-eye observation.

If one delves in the fields of medical history, it may happen that who was the first to make a statement in some area is challenged by others. To my knowledge, micrometastases were originally named so by Huvos et al,1 because they had been supposed to be detectable only by microscopy instead of being picked up at gross examination. A noninclusive size limit of 2 mm was suggested by these authors, who found no survival disadvantage for breast cancer patients with only micrometastases as compared to those with no metastasis at all, after a minimum follow-up of 8 years. Although this study suffers from low patient numbers (only 18 patients with micrometastasis) and lack of detail on the pathologic assessment of the lymph nodes (probably meaning that a single hematoxylin and eosin (H&E)-stained slide was assessed for each), it can be credited for a definition of micrometastases. It also suggested that not only the number of metastatic nodes but also the tumor burden in the lymph nodes may be an important aspect of nodal involvement.

Occult metastases have been well known by pathologists for more than 50 years2,3 and were often reported to be of no real prognostic value. However, the fact of being occult, i.e. not disclosed by first microscopic inspection but identified later by a more thorough or more sensitive work-up (this definition of "occult metastasis" will be used), does not reflect a size or tumor burden adequately. Such metastases may be smaller or larger than 2 mm, depending on what size the lymph node investigated is and what the "standard" or original pathologic examination consisted of.4 Because the size of a metastasis influences its chances of being detected by limited sampling, micrometastases are linked to occult metastases since many of them would remain occult if a single H&E section were to be assessed per lymph node; both terms reflect low-volume metastatic involvement.

The definitions of micrometastasis are not uniform. Most investigators use the 2 mm inclusive upper cut-off limit, as suggested by the tumor node metastasis (TNM) classification of malignant tumors,5,6 whereas others use a 1 mm cut-off size,7,8 a larger 0.2 cm2 or a <1 mm2 cut-off area.9,10 The European Working Group for Breast Screening Pathology recently assessed the pathology practice relating to sentinel lymph nodes (SLNs) by means of a questionnaire and found that the term micrometastasis was used by 93% of the responders; the definitions given for this category included 17 somewhat different enti-ties.11 For uniformity of use, adherence to the TNM definition of micrometastasis, last updated in 2002, which includes a lower cutoff value of 0.2 mm, is strongly recommended. However, it must be noted that this definition is not unanimously reflected in current and earlier publications, and one should always check definitions before interpreting retrospective data.

According to the current definition, the former group of micrometastasis was split into two diagnostic categories: the micrometas-tases per se and a disputed category labeled with the misnomer of "isolated tumor cells" (ITCs).5,6,12 The latter is also called submi-crometastasis by some,13 but the name of nanometastasis has also been suggested.14

Although the Union Internationale Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC) are supposed to use the same TNM system for the determination of the anatomic extent of malignant disease, the wording in the two main relevant publications of these bodies differ in a minimal extent, and this may be the source of some differences in interpretation and hence classification. The sixth edition of the UICC TNM Classification of MalignantTumours5 defines ITC as "single tumor cells or small clusters not more than 0.2 mm in greatest dimension, that are usually detected by immunohistochemistry or molecular methods; they do not typically show evidence of metastatic activity or penetration of vascular or lymphatic sinus walls." The AJCC Cancer Staging Manual6 and its abridged variant used a somewhat different wording, and defined ITC as "single cells or small clusters of cells not greater than 0.2 mm in largest dimension, usually with no histological evidence of malignant activity (such as proliferation or stromal reaction)." The relation to sinus walls is not mentioned in the latter description.6 The use of the pN0(i+) symbol to denote ITC was also different in the AJCC publications6 from the one suggested in the UICC publications.5,12 However, in a revision it was made clear that the "(i+)" denoted the presence of ITC and not immunohistochemistry (IHC) as the method of its detection (Figure 6.1).15,16 Although it may seem that size is a major criterion to distinguish between ITC and micrometastases, and therefore a node-negative and a node-positive stage (from staging and management aspects), both major publications5,6 cite a UICC paper where the tumor cell location within the lymph node, and the so-called metastatic activity, were also listed as distinguishing features.12

The present TNM categories were also endorsed by the United Kingdom National Health Service Breast Screening Programme17 and their respective labels are given below:

• pN1mi: micrometastasis;

• pN0(i+): isolated tumor cells, identified by microscopy, i.e. morphological methods, including H&E staining and/or IHC;

• pN0(mol+): evidence of ITCs by molecular methods, most commonly reverse transcription polymerase chain reaction (RT-PCR) only;

• whenever ITCs are looked for by morphologic or molecular studies, but these yield negative results, the pN0(i—) and pN0(mol—) symbols are to be used;

• pN0 is to be used when no metastases are found, but no special methods are used for the search of "occult" metastases.

Figure 6.1 Isolated tumor cells (ITC; left side images) and micrometastases (right side images) can be detected by both hematoxylin and eosin staining (HE; top images) and cytokeratin immunohistochemistry (IHC; bottom images). Note that ITC are generally visualized by IHC but larger clusters close to 0.2 mm can also be first seen on HE slides. Likewise small micrometastases may escape HE detection and may require IHC for identification. Bars, 0.2 mm; all images at 400x magnification, except top right at 100x.

Figure 6.1 Isolated tumor cells (ITC; left side images) and micrometastases (right side images) can be detected by both hematoxylin and eosin staining (HE; top images) and cytokeratin immunohistochemistry (IHC; bottom images). Note that ITC are generally visualized by IHC but larger clusters close to 0.2 mm can also be first seen on HE slides. Likewise small micrometastases may escape HE detection and may require IHC for identification. Bars, 0.2 mm; all images at 400x magnification, except top right at 100x.

The TNM classification also suggests that whenever there are multiple distinct foci of metastatic involvement, only the largest should be considered for classification.

With all this in mind, it must be mentioned that the reproducibility of these staging categories is less than optimal. A group of European pathologists with expertise in breast pathology examined and interpreted 50 cases represented by digital images, all approaching the differential diagnosis of micrometastases and ITCs. The kappa value for the consistency of categorizing low-volume nodal load into micrometastasis, ITCs or none of these was 0.39 (reflecting fair reproducibility) when each participant used his/her interpretation of the TNM definitions. This figure changed to 0.49 (still reflecting only moderate repro-ducibility) in a second circulation of the same set of images performed after a discussion aimed at making the interpretation of the definitions more uniform.18 The kappa value was <0.57 even on a single institutional level.19 One source of interpretative trouble stems from the classification of nodal metastases of invasive lobular carcinomas, which often infiltrate the lymph nodes by a noncohesive single cell pattern (Figure 6.2). A group of patholo-gists could achieve a very good consistency in diagnosing such lesions following the partly

Figure 6.2 Discohesive pattern of a micrometastasis from lobular carcinoma. Note that some pathologists would consider this as multiple "isolated tumor cells" (requiring a comment about overall nodal tumor volume) because none of the cells or touching cell clusters are >0.2 mm. Bar, approximately 0.2 mm; 400x magnification, cytokeratin. Courtesy of Professor Simonetta Bianchi, Florence, Italy.

Figure 6.2 Discohesive pattern of a micrometastasis from lobular carcinoma. Note that some pathologists would consider this as multiple "isolated tumor cells" (requiring a comment about overall nodal tumor volume) because none of the cells or touching cell clusters are >0.2 mm. Bar, approximately 0.2 mm; 400x magnification, cytokeratin. Courtesy of Professor Simonetta Bianchi, Florence, Italy.

image-guided interpretative guidelines agreed on before testing, but several pathologists within that study, including the present author, raised concerns about diagnosing a relatively high total metastatic nodal volume as ITCs on the basis of the infiltrative pattern (Roderick Turner, personal communication, June 2007). Another source of confusion comes from the localization of low volume (<0.2 mm in greatest dimension) nodal involvement; some would not consider this a matter of diagnostic distinction between micrometa-stases and ITCs, and would make the distinction simply on the basis of size criteria, whereas others would consider extrasinus-oidal and intraparenchymal lesions of this size as micrometastases.18,20 Such interpretative issues may seriously affect prognostic, prediction and consequently therapy-related conclusions of most studies lacking a uniform pathology review of SLN slides. These problems in interpretation also highlight the need for a more precise definition of ITCs and micrometastasis categories, preferably with visual aid and examples on how to classify challenging cases.


The presence of occult metastases can be confirmed by a more detailed sampling of the lymph nodes and/or the use of a more sensitive method to detect them. As a consequence, examining several slices or step sections from a lymph node and/or introducing immunostains in their investigation leads to an increase in occult nodal involvement and an upstaging from node-negative to node-positive status when compared to a methodology which does not use these tools.

It has been known for a long time that IHC aimed at highlighting the presence of epithelial markers (mainly cytokeratins) in lymph nodes may increase the detection rate of small metastases.21 However, it must be remembered that not everything that is cytokeratin-positive represents metastatic nodal deposits; cytokeratin-positive dendritic cells are normal constituents of the lymph nodes,22,23 and benign epithelial inclusions24 or dislodged papillary fragments25 are also positive. Rarely, macrophages and plasma cells may also stain nonspecifically, and different contaminants and carry-over may hinder interpretation of cytokeratin immunostaining.26 In cases of uncertainty, the TNM general rules suggest that the lower category (less advanced disease stage) should be opted for. Therefore, a node-negative status should be the conclusion if one is in doubt about the nature of the cyto-keratin positivity.27

SLNs can be defined as lymph nodes with direct lymph drainage from the primary tumor site.26 Sentinel lymph node biopsy (SLNB) seems an ideal surgical method for selecting the most likely sites of regional nodal metastases, and therefore selecting the few or only lymph nodes which should be subjected to more scrutiny in order to identify otherwise occult metastases. According to the first relevant report on this issue, the addition (per half

SLN) of two extra H&E stained levels and one stained with cytokeratin IHC resulted in a 13% nodal upstaging as compared to a previous series of comparable breast cancer patients staged by axillary dissection with a standard pathological assessment of the lymph nodes.28 The upstaging was mainly due to the higher detection rate of micrometastases (38% in the SLNB group vs 10% in the axillary clearance group). Many other studies have confirmed these observations, and the upstaging rates vary between 9 and 47%.29

The use of cytokeratin IHC is one of the most controversial issues in the examination of axillary lymph nodes. SLNs are often subjected to this method when negative by H&E. Although several guidelines do not recommend the routine use of IHC for the evaluation of SLNs,30 a questionnaire based survey by the European Working Group for Breast Screening Pathology suggested that 71% of the 240 pathology laboratories assessed used IHC regularly for SLNs negative by H&E.11 This rate may be even higher in the United States (Roderick Turner, personal communication, at Sentinel Node 2004, Santa Monica, December 2004).

The value of cytokeratin staining in highlighting occult nodal involvement depends on several factors. When step sectioning is done, the smaller the steps, the lower is the extra yield in nodal positivity detected by adding IHC to the standard histological stains.31-33 The size of the metastatic involvement is also important, since most of the cases identified by IHC belong to the micrometastasis or ITC categories. The pattern of nodal involvement of lobular carcinomas is often prone to escape traditional detection by light microscopy: therefore, the addition of cytokeratin IHC increases the rate of nodal positivity more than in cases of ductal carcinomas.34-36 IHC of SLNs also suggests a higher upstaging rate if the primary tumor is of the lobular type, and with this histological type immunostaining may highlight not only ITCs and micrometas-tases but also some smaller macrometastases.37 Although cytokeratin IHC may increase the detection rate of nodal involvement, it disproportionally increases the detection rate of ITCs, many of which may escape detection even if visible. The results of automated analysis of SLNs immunostained with cytokeratin antibodies suggest that very small volumes may remain unnoticed by conventional microscopic examination.38,39

Once it is established that the cytokeratin-positive cells are from the tumor, histopathol-ogy cannot differentiate between cells which got to the lymph node by an active metastatic process and those which were dislodged by prior manipulation of the tumor. Indeed, diagnostic and localization procedures (needle biopsies, wire or radioisotopic localizations) and massage, sometimes used in order to promote the migration of the tracers during SLNB, have been reported as, or supposed to be, possible ways of tumoral seeding.25,40,41 If noticed, such seeding would most likely be labeled as ITC, giving further support for including these in the pN0 category.

The combination of multilevel assessment and IHC yields the highest rate of upstaging when compared to a single H&E level assessment.29 Most macrometastases are discovered on the first or first few levels of sectioning SLN tissue blocks. Micrometastases may need a more intensive search, whereas ITCs are rather randomly distributed in the lymph nodes.31-33,42 Targeting the search for nodal involvement and concentrating it to the area around the junction of the tumor-draining afferent lymphatic channel, visualized by either the blue coloration or the highest intranodal radioactivity count, may help in finding most nodal involvement with less effort,43,44 but this needs further prospective investigation. It seems obvious that no histology protocol can aim at detecting all ITCs in a lymph node, and histology protocols should be devised in a way to be able to exclude a given size metastasis with a reasonable accuracy.45 Whether micrometastases should be the category to exclude is a matter of debate, and this will be discussed further at the end of this chapter.

Molecular methods to highlight nodal involvement of very low volume include flow cytometry46 and RT-PCR assays.29 These are very sensitive but lack sufficient specificity: a challenge which mandates the use of several markers. Most reports on RT-PCR in lymph nodes report false-negative cases, i.e. lymph nodes with histologically proven metastatic involvement but negative molecular testing. Whereas such nodes are obviously positive, those which test positive with the RT-PCR assay but are negative by histology may either represent false-positive testing or histologi-cally occult nodal involvement.47 This may also justify labeling them as pN0(mol+). Realtime quantitative RT-PCR assays are more promising, as manipulations of the cut-off levels may make the tests more specific.48,49 However, it must be remembered that molecular staging may also be biased by sampling errors43,44,50 and nontumoral epithelial tissue presence,24 and this strengthens the need for histological verification of molecular results.

Histopathology will therefore remain the standard diagnostic approach to SLNs and nodal staging. The differences in histology protocols may however cause differences in treatment options,51 and perhaps even outcome. Such differences may be overcome by devising regional, national or international guidelines on the minimum requirements in SLN histopathology. Without such standardization, outcome data will not be consistently comparable.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment