br Currently SOX is under evaluation as an alter
Currently, SOX-10 is under evaluation as an alter-native for S100 protein because it is more specific, but its sensitivity is not well-established [23,24].
Panmelanocytic cocktails may facilitate the interpre-tation of challenging cases, and improvement in detec-tion rate has been reported .
4. Technical general comments
The interpretation is very much dependent on the technical production of thin (3 mm) flat, complete and well-stained sections without artifact. In our experience, this is best achieved following using an extended fixation (24 h), Harris’ haematoxylin and S100 protein using an automated staining platform.
5. Guidance on histopathological interpretation
Nodal naevus cells present the most important problem in SLN interpretation (Fig. 3). Naevus cells can be seen in 10e20% of SLNs for melanoma [6,25,26]. When present, they are almost always located in the fibrous tissue of the PD98059 or trabeculae within the lymph node but can also be seen in a paralymphatic site outside the capsule. Perilymphatic naevus cells may be seen
Fig. 3. Nodal naevus cells located in the capsule and tracking down the nodal trabeculae in the capsule.
bulging into the lymphatic space but are still surrounded by endothelium.
Occasional collections of naevus cells in the capsule can be quite large and appear to bulge in continuity into the subcapsular space. Naevus cells from Spitz naevi, deep penetrating naevi or blue naevi may be seen rarely in sentinel lymph nodes and may not be so closely related to the fibrous tissue of the capsule or trabeculae. In these cases, a review of the primary cutaneous lesion is essential as it can help incorrect classification of the primary and putative metastasis. Indeed, a review of the primary pathology by the pathologists interpreting SLNs is desirable, although not always achievable.
The features of melanoma metastasis in SLNs can be of predictive value for the involvement of other lymph nodes. Tumour burden and the localisation is associated with distant metastasis-free and overall survival. One of these features is the pattern of distribution of the me-tastases within the sentinel node (Fig. 4). We recom-mend that melanoma metastases in a SLN can be classified as follows: subcapsular, parenchymal, com-bined (subcapsular and parenchymal), extensive confluent and extensive multifocal.
The 8th Edition of the American Joint Cancer Committee underscores the good prognosis of small metastases confined to the subcapsular zone  but also quotes ‘poor reproducibility of microanatomic location in one study’ . In that study, however, an accurate and detailed definition of the various patterns was not given in advance to the participants. The sub-capsular metastases should have a smooth parenchymal aspect rather than an irregular border or budding  to distinguish them from a subcapsular metastasis with parenchymal extension. According to such strict defini-tion, a subcapsular site has not been associated with further non-sentinel nodes involved on completion lymphadenectomy . The extensive multifocal microanatomical location is characterised by multiple metastatic melanoma foci spreading across more than 70% of lymph node area in the most representative slide. A number of deposits covering less than 70% qualifies as parenchymal. These distinctions are clearly arbitrary for the convenience of descriptive pattern classification.
The question whether paratrabecular metastases are equivalent to subcapsular metastases has been raised, and we recommend to classify paratrabecular metastases as parenchymal but add further specification. Whether paratrabecular site has a specific significance will be clarified in future studies.
In addition, we now recommend documenting extracapsular extension as a separate feature, not as a part of an extensive pattern, as previously described .
Although not currently used for staging, tumour burden is clearly important and is likely to be included in the future prognostic models in patients with the regional metastatic disease. The maximum diameter of the largest aggregate in the SLN (1.0 mm versus
Fig. 4. Topography of metastatic deposits within the sentinel lymph node: subcapsular, parenchymal, combined, extensive confluent and extensive multifocal. Immunostain with S100 pro-tein is shown.
>1.0 mm) has already been included as an additional inclusion criterion for participation in adjuvant clinical trials [31,32]. A range of procedures can be used to es-timate quantitatively tumour burden in the SLN, such as the maximal number of metastatic foci within the node, the maximum size of the largest metastasis, the maximum depth from the capsule of the tumour deposit (tumour penetrative depth) and the percentage cross-sectional area of the SLN involved by the tumour [3,14,16,30,33e39]. Some of these involve calculations at several levels [17,40].