The utility of sentinel lymph node mapping in the management of endometrial atypical hyperplasia.

OBJECTIVES: To determine the risk of endometrial cancer (EC) and lymph node involvement in patients with a preoperative diagnosis of "AH-only" versus "AH - cannot rule out carcinoma" and to study the value of SLN mapping.

METHODS: We reviewed all patients with a preoperative diagnosis of atypical hyperplasia, who underwent primary surgery with SLN mapping followed by pelvic lymphadenectomy. Sensitivity and negative predictive value (NPV) of SLN and rates of endometrial cancer were calculated.

RESULTS: Overall, 64/120 (53.3%) patients were found to have EC on final pathology: 58 stage IA, 3 IB, and 3 IIIC1. In patients with preoperative diagnosis of "AH", 44.3% (31/70) had EC on final pathology compared to 66% (33/50) in patients with "AH - cannot rule out cancer" (p=0.02). Overall, 3.3% of the patients (4/120) had lymph node involvement. In patients with EC with a pre-operative diagnosis of "AH", none had lymph node metastasis (0/31), compared to 12.1% (4/33) in patients with "AH - cannot rule out cancer" (p=0.06). Elevated preoperative CA125 levels (>25U/mL) were statistically associated with the risk of lymph node metastasis on final pathology (p=0.024). Unilateral and bilateral SLN detection occurred in 93.7% and 78.1% respectively. In patients with EC and bilateral SLN mapping, sensitivity and NPV were respectively 66.6% and 97.9%. There was one false negative (ITCs in non-SLN).

CONCLUSION: Our data indicate that the risk of lymph node involvement in patients with a preoperative diagnosis of "AH-only" is null. Lymph node assessment could be omitted in those patients. Conversely this risk is significant in patients with "AH - cannot rule out cancer". SLN mapping could be a valuable staging procedure in these patients.