Lymph node counts in endometrial cancer: expectations versus reality.

It has been proposed that an adequate lymph node (LN) dissection in cases of endometrial carcinoma (CA) should contain a determined number of pelvic (P) and/or para-aortic (PA) LNs. As a result, our surgeons have certain expectations regarding the number of PLNs and PALNs reported per case. Failure to meet these expectations has become a challenge in our practice. In an attempt to solve this problem, we wanted to ascertain whether a pathology factor such as disregarding small LNs not detected on gross examination was responsible for any discrepancy between expected and reported LN counts. To achieve this goal, we evaluated the impact of the microscopic examination of residual adipose tissue (AT) after the routine processing of LN dissections performed as part of the staging procedure for patients with endometrial CA (endometrioid, serous, and clear cell CA) on the LN counts and status for hysterectomies performed from 2006 to the present. In addition, we assessed whether other factors such as surgical procedure type, operating surgeon, histologic subtype of CA, depth of myometrial invasion, or body mass index had an impact on the number of LNs obtained. The number of PLNs and PALNs were recorded. All LN specimens were processed by dissecting LNs from the surrounding AT. The number of LNs submitted per cassette was recorded in the section code. In cases in which residual AT was submitted, hematoxylin and eosin-stained slides of the additional tissue were reviewed to determine the number and size of any additional LNs and their status. Two hundred fifty-eight patients had a median of 11 PLNs (range, 1 to 38) and 6 PALNs (range, 1 to 25). Fifty of 78 cases (64%) in which residual AT was submitted had additional LNs (median size, 4.0 mm): median 2 PLNs and 3 PALNs. There was no significant association between the number of LNs obtained and whether the residual AT was submitted (PLN, P=0.2; PALN, P=0.78). There were no cases in which metastatic endometrial CA was present exclusively in the additional LNs. Compared with open hysterectomy, laparoscopically and robotically obtained lymphadenectomy specimens had an average of 3 and 0.8 more PALNs, respectively (P=0.002). No similar association was found for PLNs or total LNs. Evidence for some difference in LN counts between surgeons was observed. No evidence of an association between body mass index, histologic subtype of endometrial CA or depth of myometrial invasion and LN count was identified. In our experience, the standard processing of lymphadenectomy specimens adequately reflects the actual numbers of LNs obtained in cases of endometrial CA. Submitting the residual AT does not increase the number of reported LNs or the detection of positive LNs. Additional studies are required to determine the actual numbers of PLNs and PALNs present and to determine whether a revision of the number of LNs required to consider a lymphadenectomy as adequate is necessary.