Selective modified radical neck dissection for papillary thyroid cancer-is level I, II and V dissection always necessary?

摘要

BACKGROUND: There is ongoing controversy as to the indications for and extent of lateral cervical lymphadenectomy for patients with papillary thyroid cancer (PTC). While most now agree that prophylactic lymph node dissections (LND) play no role, at the University of California, San Francisco (UCSF) we limit LND selectively on a level by level basis, and resect only the levels thought to harbor disease or to be at increased risk of metastases. This initial 'selective LND' usually includes levels III and IV (due to the well-documented increased likelihood of metastases to these levels) and levels I, II, and V are included when there is clinical or radiological evidence of disease or increased risk of it. METHODS: A retrospective review of the clinical charts and hospital records of 106 consecutive patients who had metastatic PTC and who underwent at least one lateral cervical LND at UCSF between January 1995 and December 2003 was carried out. Data were collected to assess which patients had levels I, II,and/or V included in their initial ipsilateral and/or contralateral LND and to determine the recurrence rates at these levels if they had previously been excised compared with if they had not. Chi-squared and Fisher exact tests were utilized for statistical comparison, where appropriate. RESULTS: A total of 140 initial lateral LND were performed: 104 ipsilateral and 36 contralateral. In these initial LND, 3.9%, 72.5%, and 18.6% of patients had levels I, II, and V resected on the ipsilateral side, and 2.9%, 60.0%, and 37.1% of patients had levels I, II, and V resected on the contralateral side. Recurrence at levels I and V was uncommon in all patient populations. Recurrence at level II was 19% ipsilaterally and 10% contralaterally when the level was previously resected and 21% ipsilaterally and 14% contralaterally when the level was not previously resected. There was no statistically significant difference in recurrence at level II when the level had previously been resected compared with when it had not. CONCLUSIONS: If utilized in the appropriate patient population, a selective approach to lateral cervical LND for PTC can be a successful alternative to the routine modified radical LND. Levels I and V do not require resection unless there is clinical or radiological evidence of disease. Guidelines for which patients may be considered for this less aggressive approach to level II nodal metastases are suggested.