Sentinel node biopsy in papillary thyroid cancer--what is the potential?

PURPOSE

Sentinel node biopsy (SNB) may identify lymph node metastases in patients with papillary thyroid cancer (PTC), enabling selective application of central node dissection (CND). The aim of this study was to assess the feasibility of implementing SNB in patients undergoing thyroidectomy for a cytologically indeterminate/suspicious/malignant thyroid nodule and to determine the potential improvement in clinical outcomes and the costs associated with the SNB technique.

METHODS

The treatment strategies and clinical and pathological outcomes of two retrospective cohorts of patients who underwent preoperative thyroid FNA over a 5-year period in two different centres were studied. The potential for implementing the SNB technique and the benefits and costs associated with implementation were estimated.

RESULTS

In centre 1, in 819 adult patients who had thyroid fine-needle aspiration cytology, the final cytology was indeterminate, suspicious and diagnostic of malignancy in 113, 29 and 28 patients, respectively. One hundred eight patients were 'suitable' for SNB. Twenty-three of these patients had PTC, six of whom underwent CND. Of these six patients, node metastasis was absent in five--the cohort in whom prophylactic CND may have been avoided consequent to a negative 'sentinel node' biopsy. Morbidity attributable to CND may have been avoided in up to four patients over a 5-year period. Costs associated with implementation of SNB outweighed any potential savings. Analysis of 491 patients in centre 2 confirmed that the benefit of SNB in PTC was similarly limited; morbidity attributable to CND may have been avoided in up to seven patients over a 5-year period.

CONCLUSIONS

Even under ideal conditions (assuming 100 % node identification rate and 0 % false negative rate), the potential short- to medium-term benefit of sentinel node biopsy in patients with thyroid cancer in centres implementing a policy of selective or routine prophylactic CND is low.