Sentinel lymph node biopsy in breast cancer and melanoma

Summary and conclusions In the introduction, a short overview of the development of the sentinel lymph node biopsy concept is presented. In addition to melanoma and breast cancer, the usefulness of sentinel lymph node biopsy as a surgical assessment method for squamous cell carcinoma of penis and vulva, Merkel cell carcinoma and thyroid cancer is discussed in Chapter II. In the first part of this thesis the application of sentinel lymph node biopsy in patients with breast cancer is described; in the second part, the application of the technique in patients with cutaneous melanoma is discussed. BREAST CANCER Chapter III describes a validation study undertaken by the University Medical Centre Groningen (UMCG) in co-operation with the Netherlands Cancer Institute in Amsterdam. Preoperative dynamic and static lymphoscintigraphy was used as well as intraoperative use of both a vital dye and a gamma detection probe. The tracers were administered into the primary lesion. Sentinel lymph nodes as well as axillary nodes were examined by immunohistochemistry staining. Sentinel lymph node biopsy was performed in 136 patients with invasive breast cancer. A sentinel lymph node was visualised by lymphoscintigraphy in 118 patients (87%). During the operation a sentinel lymph node was localised in 126 patients (93%). The sentinel lymph node contained metastatic disease in 56 patients (44%). Three sentinel lymph nodes were false-negative (sensitivity 95%). The sentinel lymph node biopsy technique was validated by performing axillary lymph node dissection regardless of the tumour status of the sentinel lymph node(s). In January 1999, the decision was made to perform lymph node dissection only if the sentinel lymph node is tumour-positive or if it cannot be found. Chapter IV provides data on the follow-up of patients with primary operable breast carcinoma assessed with sentinel lymph node biopsy without axillary lymph node dissection if the sentinel lymph nodes were tumour-negative. A total of 185 patients were enrolled in this prospective study conducted by the UMCG. The technique as described earlier was adjusted for tracer volume and injection site. The radioactive tracer was injected in 1 mL of saline (versus 0.2 mL previously) and both tracers were injected peritumourally (previously intratumourally), in accordance with current literature. Patients with tumour-positive sentinel lymph nodes, or those in which no sentinel lymph nodes were identified received completion axillary lymph node dissection. All patients were monitored according to regional follow-up protocols. The sentinel lymph nodes were identified in 179 out of the 185 patients (97%). In 73 patients (41%) the sentinel lymph nodes were tumour-positive. The median follow-up was 35 months (range 17-59). In one sentinel lymph node-negative patient, axillary recurrence occurred 26 months after the sentinel lymph node biopsy (false-negative rate: 1%). Chapter V focuses on the role of lymphoscintigraphy immediately versus three hours after tracer injection in sentinel lymph node biopsy for breast cancer management. Immediate imaging enables a more accurate definition of sentinel lymph nodes, whereas imaging three hours after tracer injection may lead to overestimation of the number of sentinel lymph nodes in breast cancer patients. In 165 sentinel lymph node biopsy procedures preoperative immediate and three hours (3 h) post-injection lymphoscintigraphy were performed. The day after this procedure both a vital dye and a gamma ray detection probe were used intraoperatively for identification of axillary and non-axillary sentinel lymph nodes. The tracers were administered intraparenchymally in four locations around the primary tumour or biopsy scar site. Patients with a tumour-positive axillary sentinel lymph node underwent a completion axillary lymph node dissection. Immediate and 3 h post-injection lymphoscintigrams were evaluated with regard to the location of the draining lymph node basin and moment of appearance, location and number of sentinel lymph nodes. Lymph node visualisation occurred in 63 immediate procedures (38%) versus 163 cases in the 3 h post-injection procedures (99%). In 17 procedures (10%) in which immediate lymphoscintigraphy had visualised sentinel lymph nodes, additional lymph nodes had been seen on 3 h post-injection lymphoscintigraphy. In eight of these procedures (5%) all lymph nodes were detected in the same draining lymph node basin. In nine of these procedures (5%) additional lymph nodes were detected in a second draining lymph node basin on 3 h post-injection imaging. Non-axillary sentinel lymph nodes were identified by preoperative lymphoscintigraphy in 28 procedures (17%) and improved assessment in three patients (5%). The only positive impact of immediate lymphoscintigraphy was the possible omission of removal of one to two second-echelon nodes per patient in 5% of patients. We consider this yield too low to continue immediate lymphoscintigraphy in breast cancer patients. MELANOMA Chapter VI describes the reliability and clinical impact of sentinel lymph node biopsy in cutaneous melanoma of the head, neck, trunk and extremities. Two hundred patients with cutaneous melanoma  1.0 mm Breslow thickness and clinically negative lymph nodes participated in this single institutional prospective study. Detection techniques comprised preoperative dynamic and static lymphoscintigraphy, intraoperative blue dye and gamma-probe. Sentinel lymph node tumour-positive patients underwent completion lymphadenectomy. Sentinel lymph nodes could be identified in 197 patients (97%). Three procedures failed in the head and neck region. In total, 150 patients had tumour-negative sentinel lymph nodes. During a median follow-up of 47 months, nodal recurrence in a negative mapped basin was documented in six patients (false-negative rate 11%). Recurrence-free and overall survival curves were constructed by Kaplan-Meier. The estimated three-year recurrence-free survival rate in node-negative and node-positive patients was 83 and 66% respectively (P<0.05). The estimated overall survival at three years was 92 and 73% respectively (P