Validation of lymphatic mapping and sentinel node ...

OR IGIN A L PA P E R

Validation of lymphatic mapping and sentinel node biopsy in patients with symptomatic breast cancer

Validation of lymphatic mapping and sentinel node biopsy in patients with symptomatic breast cancer ABSTRACT

Background Lymphatic mapping and sentinel node biopsy in breast cancer aims to allow lymph node negative women to avoid axillary clearance by providing a minimally invasive means of staging the axilla. However, before its implementation into routine clinical practice, initiating departments must verify their expertise in each of the surgical, radiological and pathological components necessary for its successful performance. Here, we present our validation experience. Methods Thirty patients with breast cancer of any stage (but without clinical axillary lymphadenopathy) undergoing definitive resection of their primary tumour underwent lymphatic mapping (using blue dye alone or in combination with radioisotope) and sentinel node biopsy concurrent with standard level II axillary clearance over a ten month period. Results All patients had sentinel nodes identified intraoperatively. The sentinel node in 29 patients correctly predicted the status of axillary involvement. One patient had non-sentinel nodal disease without metastases being identified in their sentinel node. Such a single false negative out of thirty patients is considered acceptable by current guidelines. Conclusion Validation of expertise in sentinel node identification and analysis is feasible over a relatively short period of time in a regional symptomatic breast unit. We now feel confident in offering this procedure to selected patients with breast cancer in our catchment area in place of routine axillary clearance. INTRODUCTION

Lymphatic mapping and biopsy of the sentinel node (the first node to receive lymphatic flow from the tumour site) for breast cancer aims to advance the surgical management of the axilla by allowing women who do not have nodal metastases to benefit from the accurate accrual of this prognostic information without having to undergo the morbidity associated with formal axillary clearance. As the demographic profile of this neoplasm continues to alter with the advent of population screening programmes, an expanding proportion of women are being diagnosed with breast neoplasia at an early, pre-metastatic phase. This, therefore, increases both the need and demand for such a minimally invasive means of surgical staging and, perhaps, treating regional disease. Furthermore, it may well be that, in the future, evidence of lymphatic dissemination to the sentinel node will prioritise systemic therapies over further loco-regional surgery,

RA Cahill, L Diamond,1 R Landers,2 D Walsh,1 RGK Watson Depts of Surgery, Radiology & Nuclear Medicine 1 and Pathology,2 Breast Care Unit, Waterford Regional Hospital, Waterford

with formal lymph node basin dissection being held in reserve only for those with overtly apparent lymphatic disease.1 However, imprecise performance of any of the processes required for accurate sentinel node identification, excision and pathological scrutiny risks understaging the patient’s disease. This could deprive the patient of useful adjuvant therapy. It is essential therefore that departments initiating lymphatic mapping for their patients with breast cancer must ensure their expertise in the procedure by performing thirty cases of lymphatic mapping and sentinel node biopsy concurrent with axillary clearance with satisfactory concordance of results. Here, we describe the validation series of patients with this disease in our department (which deals primarily with symptomatic breast cancer) prior to the implementation into our clinical practice of selective sentinel node biopsy in place of axillary clearance.

IRISH JOURNAL OF MEDICAL SCIENCE • VOLUME 175 • NUMBER 2

59

OR IGIN A L PA P E R

METHODS

Any patient undergoing definitive resection of the primary cancer (either wide local excision or mastectomy) and level II axillary clearance for breast cancer of any stage commencing from the 13th of October 2002 was considered for concurrent lymphatic mapping and sentinel lymph node biopsy as part of the validation series. The only exclusion criteria were clinically palpable axillary nodes or a contraindication to axillary clearance. Every patient was fully informed of the rationale, benefits (that the selected nodes would be scrutinised more closely than usual by the pathologist) and risks of the procedure (including blue staining of the urine for 24 hours and skin tattooing as well as an anaphylaxis rate of 1/1000 associated with blue dye injection) both as an outpatient prior to admission and again the evening before surgery. Lymphatic mapping was performed using either blue dye alone (n=4) or in combination with radioisotope (n=26). Although use of radioisotope, in addition to blue dye injection, increases the radiation exposure to the patient, the risks to both the patient and the health-care professionals caring for her is considered minimal provided there is adherence to published guidelines.2 The blue dye (patent blue V for the first ten patients followed by lymphozurin thereafter) was aseptically injected (1-2 mls using a 22g needle) in four peritumoral locations in every patient after induction of anaesthesia but prior to the incision being made with a five minute massage of the injection site being performed immediately after instillation of the dye. Sentinel node identification by the blue dye was noted when either the node was found to be blue or else had a definitely blue staining afferent lymphatic channel. The radioisotope tracer used (99m-Tc-labelled colloidal rhenium sulphide- Nanocis, Cis Biointernational) was injected intradermally in four peritumoral quadrants either at 10 am on the morning of surgery (n=9, total radioactivity of injectate 20 mBq) or at 16.00 hrs the evening prior (n=21, radioactivity = 80 mBq). When possible, lymphoscintigraphy was performed either two hours after injection when the surgery was being performed on the same day or at 10 am the following morning (approximately 18 hours after injection) when injected the evening before. If the scan was found to be positive, the location of the node(s) as determined by collimetry in both the anterior and lateral oblique positions was marked on the skin

60

RA Cahill et al

with the patient positioned as for her operation (i.e. supine with the upper limb on the operative side abducted to 90°). Following this, or if the scan was found to negative, the patient was transferred to the operating theatre. In every case, the sentinel node(s) biopsy was performed prior to resection of the primary tumour, with the completion level II axillary clearance being left until the end of the procedure as this mimics the way it was planned to perform the procedure after its validation (an order which would allow pathological analysis of the identified nodes intraoperatively with a minimum of operative time prolongation). To do this, a 3 cm incision was made below the hairline in the axilla and the subcutaneous tissue divided. After the underlying fascia was opened, the gamma camera (Neoprobe, Neoprobe Corporation) was used to guide further dissection with frequent, careful inspection for blue lymphatic channels. The search for additional sentinel nodes was continued after the first one was found until radioactivity counts in the axilla were less than 10% of the hottest node identified and there was no further evidence of nodes discoloured blue. While palpable nodes found intraoperatively are considered an indication to “convert” to a formal axillary clearance when the patient is planned for sentinel node biopsy alone, such patients were included in our validation series. Throughout the procedure, all measured radioactivity counts were recorded on a database. Following completion of the operation, a member of the nuclear medicine staff checked contamination levels with a Berthold monitor. Specimens were then transferred to the pathology department for histological scrutiny. While the main specimens were H&E stained in the usual fashion, the sentinel node(s) were bivalved in a longitudinal direction and examined after H&E staining at six different intervals 10 µm apart.

RESULTS

Thirty patients underwent lymphatic mapping and sentinel node biopsy along with their definitive breast cancer surgery and level II axillary clearance between October 2002 and July 2003. The mean age was 61.3 years (range 39.5-73.2 years) while the mean tumour size of the final pathological specimen was 2.5 cm (range 0.8-7 cm) with 15 being grade III cancers. Thirteen patients underwent wide local excision of the primary while 17 had mastectomies due to large tumour:breast size IRISH JOURNAL OF MEDICAL SCIENCE • VOLUME 175 • NUMBER 2

ratio (n=10), central location of the tumour (n=5) or a contraindication to adjuvant postoperative radiotherapy (n=4). No patient offered the procedure declined it, however, during the same time period, twenty-three other patients were not considered because of clinical evident lymphadenopathy (n=13), unfitness for axillary clearance (n=7, three had severe cardiorespiratory conditions while four were judged to be generally too infirm) or lacking in the capacity to consent (n=2, one patient had Alzheimer’s disease while the other had cognitive impairment). In no case was the sentinel node not identified either by visual dye localisation or audible gamma detection (0% failure rate), although intraoperatively five cases had only “hot” nodes that were not blue while six were found because they were only blue (four of these cases did not have radioisotope injected preoperatively). The mean number of nodes harvested per axillary clearance was 13 (range 528). The majority of patients (n=25) had more than one sentinel node found intraoperatively, the mean number yielded being 2.4 (range 1-4). In total, twelve patients had lymphatic metastases, eleven of whom had such disease in their sentinel node(s). There was therefore one patient who had non-sentinel nodal disease that was not predicted by determination of the status of the sentinel node (i.e. a falsely negative sentinel node). While most patients with involved sentinel nodes had additional non-sentinel nodal disease, four patients only had evidence of metastases in their sentinel node. Of the ten patients with T1 tumours (