Additional reverse transcription-polymerase chain reaction of peripheral slices is not superior to analysis of the central slice in sentinel lymph nodes from melanoma patients.

Abstract

The status of the sentinel lymph node (SLN) is an important prognostic factor in patients with cutaneous melanoma. Reverse transcription-polymerase chain reaction (RT-PCR) has been used as a sensitive means of detecting tumour cells in SLNs. To determine whether RT-PCR analysis of the SLN using both the central and the peripheral slices is more sensitive than molecular analysis of the central slice only. Eighty-three SLNs from 59 patients with primary cutaneous melanoma were identified by SLN mapping. All SLNs were bisected along their longitudinal axis to produce two equal halves. One half was used for histology and immunohistochemistry, and the other was analysed by RT-PCR for tyrosinase and MelanA. Parallel to the longitudinal axis, one central slice (approximately 2 mm in thickness) was cut manually. This central slice was used for our standard RT-PCR protocol. In the current study, up to eight additional peripheral slices (each approximately 2 mm in thickness) were cut parallel to the existing cut surface. These peripheral slices were analysed by additional RT-PCR. Standard RT-PCR of the central slice yielded positive results in 34 of 59 patients (57%). Additional RT-PCR of peripheral slices demonstrated positive findings in six additional patients (10%) who were initially negative by standard RT-PCR of the central slice. In detail, seven of those 34 patients positive by standard RT-PCR of the central slice had positive histological findings. In each of these seven patients, RT-PCR was positive both in the central slice as well as in the peripheral slices. The remaining 27 patients with positive RT-PCR results of the central slice showed negative histological findings. Only nine (33%) of these 27 patients had a positive RT-PCR also in the peripheral slices. Finally, all 25 patients with negative RT-PCR results in the central slice showed negative histological findings. Six of these patients (24%) revealed positive RT-PCR results on the analysis of peripheral slices. However, three of these patients expressed only MelanA but not tyrosinase. Thirty lymph nodes from 24 nonmelanoma patients served as negative controls for RT-PCR. In three of these 24 patients (13%) expression of MelanA but not tyrosinase was detected by RT-PCR. Molecular analysis of peripheral slices yielded six additional patients (10%) positive by RT-PCR who were initially negative by standard RT-PCR of the central slice. However, three of these six patients were found to express only MelanA but not tyrosinase. As MelanA expression was also found in 13% of control lymph nodes, positive MelanA expression alone in SLNs of melanoma patients requires cautious interpretation in order to avoid false-positive findings. Thus, additional molecular processing of peripheral slices did not significantly increase the number of patients with RT-PCR-positive SLNs.