Kirby et al are correct in their statement that continued controversy surrounds the comprehensive surgical staging of all patients with clinical stage I endometrial adenocarcinoma. Such is the case because lymph node metastasis is found in only 10% of these patients. The proportion of patients found to have lymph node metastasis is even lower among those with grade 1 and 2 tumors with minimal or no invasion. A high proportion of patients with endometrial adenocarcinoma fall into this group. In the classic paper by Creasman et al,[1] of the 621 women with endometrial adenocarcinoma who underwent surgical staging, 468 had grade 1 or 2 tumors and the tumor invaded to less than one-third of the myometrial thickness or not at all in 367. The incidence of pelvic lymph node metastasis in patients with grade 1 tumors that invaded to less than onethird of the myometrial thickness was 3%. Among women with grade 2 tumors with less than one-third myometrial invasion, the proportion with pelvic lymph node metastasis was 5%. The incidence of periaortic lymph node metastasis was 1% and 4%, respectively. These numbers were obtained from analysis of the final pathology diagnosis. Kirby et al argue that there is no accurate way of knowing the tumor grade and depth of invasion until the permanent tissue sections from the hysterectomy specimen are studied. Therefore, they suggest that every patient with a biopsy or curettage diagnosis of grade 1 endometrial adenocarcinoma and a frozen section evaluation that confirms the histologic grade and shows minimal or no myometrial invasion should undergo a comprehensive lymphadenectomy, to benefit the few who, on final analysis, may be found to have a poorly differentiated or deeply invasive tumor. They use selected studies from the literature to support their recommendation. Evaluation Methods, Morbidity, and Survival
Is the predictive value of a preoperative and intraoperative tumor evaluation as poor as suggested by Kirby et al? Will the final pathology diagnosis cause a significant change in therapy, and more importantly, will not performing a lymphadenectomy in selected cases have an impact on survival? Is a comprehensive lymphadenectomy always a safe procedure with no associated morbidity, as Kirby et al would like us to believe? I will approach the last question first. Kirby et al tell us that the morbidity associated with a total abdominal hysterectomy and a pelvic and periaortic lymphadenectomy is no different from that associated with hysterectomy alone. If the lymphadenectomy is such a safe and simple procedure, we should be training our gynecology residents-who will diagnose the majority of the 40,000 new cases of endometrial cancer each year-to do them. They may do a better job than the general surgeons, who do not have adequate training or experience in performing pelvic or periaortic lymphadenectomies. Perhaps we gynecologic oncologists, as a group, are gifted technicians who are able to do these procedures fast and safely. However, data from other centers suggest otherwise. The gynecologic oncology group from Duke University showed that women who underwent pelvic lymph node sampling sustained increased blood loss and transfusions, longer operative time, and increased morbidity compared to those who only underwent a hysterectomy.[2] In addition, Lewandowski et al[3] showed that 11% of patients who received whole-pelvic irradiation after total abdominal hysterectomy, bilateral salpingo- oophorectomy, and pelvic and periaortic lymph node dissection developed severe intestinal complications compared to none of those who did not undergo a lymphadenectomy. It is likely that the reported differences in morbidity associated with the lymphadenectomy could be explained in part by the composition of the population studied. It has been shown that the success of completing a periaortic lymphadenectomy in patients with endometrial carcinoma steadily decreases as the patients' body mass index increases and with the presence of other severe comorbid conditions. This is true whether the procedure is done via laparotomy or laparoscopy.[4] Omitting the Lymphadenectomy
Several studies suggest that omitting the lymphadenectomy in selected cases of endometrial adenocarcinoma has no negative impact on survival. Larson et al[5] reported on 102 women with grade 1 endometrial adenocarcinoma confined to the endometrium or that did not invade deeper than the inner half of the myometrium. These patients did not undergo lymphadenectomy and did not receive adjuvant therapy. Their cumulative 5-year survival was 87%. Belinson et al[6] reported on 216 women with clinical stage I endometrial adenocarcinoma who did not undergo pelvic lymphadenectomy. Adjuvant radiation therapy and periaortic lymph node dissection was performed in those who had deep myometrial invasion, grossly positive pelvic lymph nodes, or cervical involvement. With a median follow-up of 61 months, only 21 (10%) recurrences were encountered. Only two (1%) of the recurrences occurred in the pelvis. How many patients in whom the frozen section evaluation of the hysterectomy specimen shows a grade 1 tumor with no invasion will be at significant risk of having lymph node metastasis? In a recent study from the M. D. Anderson Cancer Center, of 36 patients with stage IA, grade 1 adenocarcinoma on frozen section, 4 were found to have grade 1 tumors with positive peritoneal cytology.[7] Of the remaining 32 patients, 9 had a grade 1 tumor and three had a grade 2 tumor with less than 50% myometrial invasion. The authors estimated that for patients with a frozen section diagnosis of stage IA, grade 1 endometrial adenocarcinoma, the probability of nodal metastasis is only 1% to 2%. They also studied patients with a prehysterectomy biopsy diagnosis of grade 1 endometrial adenocarcinoma and an intraoperative gross assessment of no myometrial invasion.[8] Of these patients with presumed stage IA, grade 1 endometrial adenocarcinoma, only 1 (4%) was significantly upstaged. Others have shown that combining the histologic grade obtained on the presurgery endometrial biopsy and the frozen section evaluation of the hysterectomy specimen could further improve the correlation with the final pathology diagnosis.[9,10] Extent of Excision
Kirby et al are also correct in that the appropriate extent of the lymphadenectomy varies. They describe a complete retroperitoneal lymph node dissection that extends from the origin of the inferior mesenteric artery (or the superior mesenteric artery) down to the entry of the superficial circumflex iliac vein into the external iliac vein. If this procedure was indeed associated with no increased morbidity, then it may be justified to perform it in women with comorbid conditions who have a small chance of lymph node metastasis (2 cm or less, grade 1 tumors with no invasion or minimal invasion on pre- and intraoperative evaluation). I suggest that it will be impossible to show that the benefit (ie, survival) outweighs the risk (ie, morbidity) in these patients.

—Enrique Hernandez, MD, FACOG, FACS