Background Excision of the primary melanoma and biopsy of sentinel lymph nodes (SLNs) has become the therapy of choice at many institutions since extensive lymphadenectomies may be avoided if the biopsies are histologically normal. Preoperative lymphoscintigram (LS) is performed to map the lymphatic drainage and identify the SLNs, which are typically located in standard lymph node basins (SLNBs), such as the axillary, groin and cervical regions. However, so called interval (or in-transit) sentinel nodes (iSLN) may be identified at atypical sites in as much as 7 % of the cases. We describe, in detail, our recent experience in the management of iSLN in cutaneous melanoma.

Material and Methods All patients seen at the Yale Melanoma Unit from September 1997 to February 2003 with melanomas of at least 1.0 mm depth, or a Clark IV or greater and without palpable lymph nodes were evaluated with preoperative LS using Tc-99m sulfur colloid. Surgeon-adminstered intradermal injections around the primary melanoma using Tc-99m sulfur colloid and Lymphazurin Blue dye were performed intraoperatively. After ten minutes, radioactive uptake was traced using a gamma probe.

Results All SLNs were removed from each of 374/374 (100 %) patients based on the preoperative LS and intraoperative findings, and unequivocal iSLN were identified in 7/374 (1.9%) patient. The average age of the seven patients (three females) with iSLN was 46.3 yrs (range 25-65 yrs.). Mean thickness of the primary lesion was 1.7 mm (range 0.9-3.25 mm). Six patients had stage I disease and one patient had stage II disease with a lesion thickness measuring 3.3 mm. None of the patients had distant metastasis as determined by preoperative CT scan. A summary of the patients are compiled in the table:

Pt	Age/Sex	Melanoma Site	Depth	SLNB	pos	pos	iSLN location
#1	65/M	L Lower Abd	1.6 mm	L Inguinal	0/4	0/1	L Chest wall
#2	36/M	Mid Lumbar	1.0 mm	R Inguinal	0/2	0/2	R Flank
#3	55/F	R 5th Finger Subungual	3.3 mm	R Axilla	1/1	1/1	R Epitrochlear
#4	65/M	L Flank	0.9 mm Clark IV	L Axilla	0/3	0/2	L Trunk
#5	35/F	L Scapula	2.8 mm	L Cervical	0/1	0/1	L Subtrapezius
#6	25/F	R Flank	1.0 mm	R Axilla	0/5	0/1	R Chest wall
#7	43/M	L Chest	1.3 mm	L Axilla	N/A	1/1	L Pectoralis

The intraoperative findings in all cases were consistent with the preoperative LS. In three cases, the iSLN was located between the primary lesion and the SLNB, consistent with lymphatics first draining to the interval node and then to the "standard" sentinel node. However, in four cases, the interval nodes were not located in the anticipated pathway between the primary tumor and the SLNB. In patient # 3, both the epitrochlear and axillary sentinel nodes were positive, whereas in patient # 7 the breast interval node was positive. This patient had, however, undergone a completion L axillary lymphadenectomy as part of a previous operation. In the remaining 5 patients, both the interval and "standard" sentinel nodes were negative. All patients remain alive and disease-free upon follow-up for 2 to 36 months.

Conclusion Although iSLNs are rarely reported, they seem to be as likely to contain micrometastatic disease as those in the expected SLNB. Furthermore, in a majority of our patients, the iSLN was not found in the anticipated lymphatic pathway between the primary tumor and the SLNB. Surgical excision of both interval and "standard" SLN should be performed in these patients and adequate preoperative imaging and intraoperative recognition of these nodes are of paramount importance.

This work was supported by an NIH research grant CA-16359 from the National Cancer Institute to S.A.