BACKGROUND:
Sentinel lymph node status is considered the most important prognostic indicator for disease-specific survival as well as the best predictor of recurrence in patients with primary cutaneous melanoma. However, up to 20% of patients with an initial negative sentinel lymph node biopsy (SLNB) have recurrence, and distant metastasis is the most common site. This study aims to determine the pattern of recurrences including distant sites for patients with a negative SLNB.
METHODS:
We performed a retrospective review of our melanoma database collected in an academic tertiary-care medical center following IRB approval for the period between 1993 and 2009 which included 2,098 patients who underwent SLNB (16.9% positive vs. 83.1% negative). Patients with primary cutaneous melanoma and an initial negative SLNB were included in this study (n=1,744). The median follow-up of the patients was 6.9 years. Statistical analysis was performed using Stata v14.1 software.
RESULTS:
The 1,744 patients with an initial negative SLNB were divided into four groups: 1,446 (82.9%) of these patients had no recurrence, 130 (7.4%) had local (within 5 cm of the original lesion) or in-transit (between the primary site and regional LN basin) recurrence, 54 (3.1%) had recurrence that involved regional LN basins, and 114 (6.5%) had distant recurrences with or without local or regional recurrences. The false negative SLNB rate of our study was 6.7% (any nodal recurrences in the SLNB sites).
Of the 114 patients with a negative SLNB who developed distant metastasis, 100 patients (87.7%) presented with distant metastasis without evidence of local, in-transit, or regional LN recurrence. Nine patients (7.9%) presented with either local or in-transit recurrence in addition to distant metastasis. Five patients (4.4%) presented with recurrence within the regional LN basin along with distant metastasis.
Of the patients that had distant metastasis without local, in-transit, or regional LN recurrence, either with single or multiple location recurrences, the following sites were the most common: 52 patients (61.9%) had lung metastases, 30 patients (35.7%) had brain metastases, and 14 patients (16.6%) had liver metastases. Only 8 of these patients (9.5%) had any evidence of distant LN involvement (outside of regional LN basins).
CONCLUSION:
Only a minority of patients with a negative SLNB (17.1%) developed recurrences. The majority of recurrences in this cohort of patients occurred without evidence of local, in-transit, or regional LN involvement. The most common sites for distant metastases were the lung and brain, and less than 10% of these patients had any evidence of distant LN involvement. These findings suggest that hematogenous rather than lymphatic dissemination may play a role in the cancer progression and development of metastatic disease in some melanoma patients. Future molecular studies may shed light to the mechanisms of lymphatic versus hematogenous spread of cancer.