with UPSC, and 79% of patients with EC. LND was performed in 82% ofstage I-II and in 68% of stage III–IV cases. The median total LNC was 13(range, 1–75), and there was no significant difference in the total LNCamong the different histologies. In patients with stage III and IV diseasewho had LND, the rate of positive LN was higher in UPSC (29%) andMMMT (19%) as compared to EC (15%) (P b 0.01). In the group withpositive LN, a moderately positive association between the total andpositive LNC was present (Pearson coefficient 0.34, P b 0.001). Thecohort was divided in quartiles based on the total LNC and a Kaplan–Meier survival analysis was performed. A continuum of improved OSwas noted in correlation with increased LNC. OS was 27 months for thegroup with 0 nodes, 112 months for the group with 1–8 nodes,117 months for the group with 9–16 nodes, and 196 months for thegroupwith N17 nodes. Doubling the total LNCwas associatedwith a 28%risk of death reduction (HR 0.724, 95% CI 0.66–0.794, P b 0.001) for thefirst year and 14% risk reduction (HR 0.858, CI 0.761–0.967, P=0.012)for the second year. This effect was independent of stage, histology, typeof adjuvant treatment, age, and race.Conclusions: In this cohort, the performance of LND was associatedwith improved OS. The extent of the LND was inversely correlatedwith the risk of death for the first 2 years. This effect appears to beuniform across pathology types.

doi:10.1016/j.ygyno.2014.03.339

320 - Poster Session AAnalyzing the learning curve of robotic-assisted sentinel lymphnode dissection for endometrial cancerP.H. Desai, P. Hughes, K. Graebe, N. Tchabo, D.H. Tobias, P.B. Heller,B.M. Slomovitz. Women's Cancer Center, Carol G Simon Cancer Center,Morristown, NJ, USA.

Objectives: Wide-ranging controversy surrounds performing lymphaticstaging for endometrial cancer (EC). Sentinel lymph node dissection(SLND) has emerged as a feasible option not only to avoid comorbiditiesassociated with regional lymphadenectomy but to aid in determiningwhich patient population would benefit from adjuvant therapy.However, performing SLND with minimally invasive technique such asrobotics involves a learning curve that has exceptionally high challenges.We sought to evaluate the learning curve of SLND with robotic-assistedlaparoscopic hysterectomy for EC.Methods: A retrospective database of the patients who underwentcervical blue dye injection followed by intraoperative SLN mappingperformed by single surgeon with an experience of N500 robotic

procedures was reviewed. SLN was initially examined by routinehematoxylin and eosin stain if negative, with ultrastaging by immuno-histochemistry (IHC). CUSUM curve of failure for the learning curve wasconstructed for detection rate (DR) and negative predictive value (NPV)by using following formula: Sn= Σ (Xi− Xo), where Xi = 0 for asuccess and 1 for an observed failure. Xo is the predicted risk of failure.Results: A total of 120 patients with EC underwent Robotic SLNDwithout any conversion to open procedure between April 2011 and June2013. Only 1 of 120 patients underwent SLND for fertility preservation.The DR for SLND was 86% (103/120). Bilateral SLN were detected in 52%(62/120). Positive nodes were identified in 8% (10/120) of the patients.SLN and regional lymphadenectomywere performed in 17.5 % (21/120)cases and theNPVwas 100%. CUSUMcurve for DR (Fig. 1B) andNPV (Fig.1C) trended below the baseline from the beginning, corresponding tozone 3 of adequate experience in Fig. 1A.Conclusions: This is the largest cohort for robotic SLND for EC. Ourresults demonstrated that the initial expected learning curve can besurpassed, and SLND by robotic-assisted procedure for DR and NPVcan be mastered with an adequate prior experience in roboticsurgery. However, further large studies are required to confirm theDR and NPV for SLND with robotic-assisted procedures for EC.

doi:10.1016/j.ygyno.2014.03.340

321 - Poster Session ADoes ultrastaging improve detection of micrometastasisfor early-stage endometrial cancer?P.H. Desai, P. Hughes, S. Edmee, N. Tchabo, D.H. Tobias, P.B. Heller,B.M. Slomovitz. Women's Cancer Center, Carol G Simon Cancer Center,Morristown, NJ, USA.

Objectives: The prognostic value of lymph node dissection for allpatients with endometrial cancer (EC) remains controversial.Sentinel lymph node dissection (SLND) avoids a full lymphadenec-tomy with its associated increased morbidity, but it offers additionalprognostic information that may help to better determine adjuvanttherapy. The value of ultrastaging the SLN to reveal micrometastasisremains unclear. The objective of this study was to evaluate a largeseries of patients who underwent SLND for EC and to determine theimpact of ultrastaging on lymph node positivity.Methods: We performed a retrospective analysis of patients whounderwent SLND for EC using the robotic platform. All patientsunderwent intraoperative sentinel lymph node mapping by injecting

Abstracts / Gynecologic Oncology 133 (2014) 2–207130