Carcinoma of the uterus is the most common gynecologic cancer diagnosed in the United States, with more than 40,000 women affected annually.1 Despite this high prevalence, management is an issue of significant debate and controversy. Balancing complete staging information for both prognostic and potential therapeutic benefits against potential perioperative morbidity and mortality has fostered numerous studies to estimate the relationship between clinical and pathologic characteristics in endometrial cancer.
It has been almost 25 years since the Gynecologic Oncology Group first published their seminal article on surgical pathologic spread patterns in 621 patients with endometrial cancer.2 Key findings of this study include the importance of grade and depth of invasion, which enabled the authors to stratify patients into three risk categories for nodal metastases.
In 2000, Mariani et al4 from the Mayo Clinic published their study that demonstrated a subset of endometrial cancer with favorable characteristics, which included three low-risk features involving tumor size 2 cm or smaller, grade 1 or 2 tumors, and depth of invasion of 50% or less. In their study, they demonstrated a 5% risk for nodal metastasis and 97% cancer-specific survival in this low-risk group.4 A subsequent recent prospective study from the Mayo Clinic correlated tumor size, grade, and depth of invasion with lymph node metastasis in 422 consecutive patients undergoing routine formal surgical staging.5 The “Mayo criteria” has led to several studies examining the potential applicability of the low-risk criteria at other medical centers. Most recently, a retrospective multicenter review demonstrated the potential applicability of frozen and final pathology to predict low-risk endometrial cancer with a 98.2% negative predictive value.6 Although the rate of unexpected diagnoses of endometrial cancer is not clear, a recent study utilizing the Surveillance, Epidemiology, and End Results program demonstrated that approximately 62% (24,436 of 39,396) of women underwent a hysterectomy without an associated lymphadenectomy.9
Currently, there are no clear management strategies for those unstaged cases in patients identified with endometrial cancer in the postoperative setting; the American Congress of Obstetricians and Gynecologists guidelines defer to individual practice patterns.7 Recent evidence suggests that staging of cases in patients with endometrial cancer may have a prognostic role, whereas a therapeutic role at this time is less clear.10–12 Risk–benefit profiles of surgical intervention often guide staging, but clear information to patients and clinicians is not readily available.13 In light of these concerns, we conducted a review of LAP2, a large prospective trial in endometrial cancer staging conducted in the United States, using a modified Mayo criteria to assess the risk of nodal disease in women with endometrial cancer.13
MATERIALS AND METHODS
This is a post hoc analysis from the previously identified 2,516 women enrolled in the Institutional Review Board-approved Gynecologic Oncology Group study (Gynecologic Oncology Group 2222 or LAP2).13 All patients signed a locally approved informed consent and authorization permitting release of personal health information. The current study was approved by the University of Louisville Medical Center institutional review board. Inclusion criteria for the patients in this analysis were uterine cancer of endometrioid histology and complete clinicopathologic data involving the modified risk criteria. According to the modified Mayo criteria, we identified each patient as at low risk or at high risk for extra uterine disease requiring systematic lymphadenectomy. Patients identified as low risk for nodal metastasis were modified from the Mayo criteria endometrial cancer characteristics with three specific criteria on final pathology reports: less than 50% invasion; tumor size smaller than 2 cm; and well or moderately differentiated (grade 1 or grade 2) endometrioid histology.
Myometrial invasion and primary tumor data were abstracted by chart review, and histologic classification was determined by central review by the Gynecologic Oncology Group Pathology Committee. Unlike the previously reported Mayo criteria, frozen-section analysis was not available to us for review.5
Differences between patient characteristics for the low-risk and high-risk groups were examined using the Pearson χ2 test or Fisher exact test for categorical variables and the Wilcoxon–Mann–Whitney test for continuous variables, with statistical significance defined as P<.05. A logistic model was used to estimate a patient's risk for nodal metastasis based on risk-group membership. The model adjusted for relevant prognostic factors and baseline characteristics that differed significantly between the risk groups (tumor grade, depth of invasion, age); tumor extent and surgical stage were considered but were not used when shown to contribute little to the variation in the model. Multicollinearity among the model covariates was examined through the method of generalized variance inflation factors described by Fox and Monette14 and were found to be negligible. The Hosmer–Lemeshow–Cressie goodness-of-fit test failed to find a significant lack of fit (P=.47) in the model. By the “rare-disease assumption,” ie, if the proportion of nodal metastases in the low-risk group is less than 0.10 (0.008), the odds ratio approximates the relative risk of nodal metastasis. All statistical analyses were performed using the R programming language and environment.15
Only patients enrolled in LAP2 were included in this study. Among the 2,516 eligible patients from the original LAP2 study,13 we included only those who had complete information criteria of grade, myometrial invasion, and primary tumor diameter, and those whose tumors were classified by prospective pathology review as endometrioid adenocarcinomas; those restrictions left a patient population of 971 (39%), as shown in the CONSORT diagram in Figure 1. A total of 2,023 patients with endometrioid uterine cancer were reviewed. All patients had documented tumor grade. There were 51.3% (1,038 of 2,023) of patients missing data regarding tumor size; 51.4% (1,041 of 2,023) of patients were missing data regarding depth of invasion; and 52% (1,052 of 2,023) of patients were missing data for one or both of size and invasion, leading to the 971 patients included in this study. The majority of patients were white, and the median age was 60 years.
Patients with nodal metastasis did not demonstrate any significant clinical differences from those patients without nodal metastasis regarding age, race or ethnicity, or performance status (Table 1). As expected, patients with nodal metastases were more likely to have higher-grade tumors, increased depth of invasion, and larger tumor size (Table 1). Individual low-risk uterine pathologic characteristics were associated with an up to 4.8% risk for lymph node metastasis (Table 1). According to the modified Mayo criteria, we identified each patient as being at low risk or at high risk for dissemination (metastatic disease in the lymph nodes) requiring systematic lymphadenectomy. The patient characteristics for the low-risk and high-risk groups are shown in Table 2, with patients in the high-risk group being significantly older (high-risk median age of 61.8 years compared with low-risk median age of 59.6 years, P=.008) than patients in the low-risk group. Because this was a post hoc analysis, patients with missing data were excluded; however, we found no statistically significant difference between the patients missing data by presence of nodal metastasis (Table 3).
Approximately 40% (or 389 of 971) of patients in this study were found to be at low risk based on the modified Mayo criteria, with a rate of nodal metastasis of only 0.8% (3 of 389; exact 95% confidence interval [CI] 0.16–2.2). We analyzed the risk of a patient having nodal metastasis by determining the odds of metastasis for patients in the low-risk and in the high-risk group (Table 4) using logistic regression. The regression controlled for the following potential confounders: tumor grade, depth of myometrial invasion, and patient age, all of which were significantly different between the groups. The odds of nodal metastasis were greater for patients in the high-risk group (odds ratio 6.3, 95% CI 1.67–23, P=.007). All three patients in the low-risk group who had nodal metastasis were white and had moderately differentiated endometrial histology. We also identified the correlative effects of positive pelvic lymph node involvement with periaortic lymph node involvement (Table 5). Patients with high-risk modified Mayo criteria but negative pelvic lymph nodes had a statistically decreased risk for periaortic lymph node metastasis (12 of 520 or 2.3% compared with 19 of 50 or 38%; P<.001).
Complications and adverse events by risk-group membership (Table 6) suggest a statistically higher perioperative complication rate in the high-risk modified Mayo criteria group, but the P values have not been adjusted for multiple testing and should be interpreted with caution. It is important to note, however, the perioperative complication risk (defined as readmission, reoperation, and death) in the low-risk modified Mayo criteria group was higher than that in the identified nodal metastasis risk (6.4% or 25 of 389 compared with 0.8% or 3 of 389; P<.001).
Low-risk criteria in endometrioid endometrial cancer hysterectomy specimens have been associated with a decreased risk for nodal metastasis. Approximately 40% (or 389 of 971) of patients in this study were found to be at low risk based on the modified Mayo criteria, with a rate of nodal metastasis of only 0.8% (3 of 389; exact 95% CI 0.16–2.2).
Our study demonstrates that low-risk criteria correspond to a low rate of nodal disease and are associated with a decreased risk of lymph node metastasis in patients with endometrioid endometrial cancer. Our study supports the previously published associations between risk criteria involving tumor size, grade, and myometrial invasion with nodal metastasis risk.3,5,16 However, those previous studies were single institution reviews with limited ability to translate to general clinical practice.3,5,16 Heterogeneity of tumor specimens included nonendometrioid tumors and, therefore, added to potential risk of extra uterine disease.3,5,16
Recent studies have demonstrated the limitation of preoperative testing to identify patients at risk for nodal metastasis in early stage endometrial cancer, which often leaves staging to individual surgeon preference.17 Management of clearly identified endometrial cancer is even now heterogeneous, with a wide range of management strategies.18 Clinical and pathologic factors that predict reduced risk lymph node involvement could help guide more conservative therapy, including simple hysterectomies, whereas those that predict increased risk for lymph node involvement may justify aggressive surgical management to help with identification of nodal metastasis. Patients with a postoperative hysterectomy diagnosis of endometrial cancer could be triaged into categories of patients who may benefit from surgical staging with a clearer discussion of risks and benefits before surgical intervention.
These findings should be validated in a prospective multicenter trial that includes the predictive value of low-risk in the frozen section and final pathology in the setting of diagnosed endometrial cancer. Our hypothesis would be that criteria, such as the modified Mayo criteria, could be utilized and reproducible in both frozen and final pathology in a multicenter community setting.
This study is limited by its retrospective nature and missing data, but the findings are still suggestive of an association and need to evaluate the uterine specimen to help guide therapy. Another potential limitation of this study is that although histology and grade were centrally reviewed, tumor size and depth of invasion in general were not unless rereview was warranted (eg, a staging discrepancy), and the missing data may limit the conclusions. Future research will also be needed to elucidate the effect on the modified Mayo criteria in those patients who undergo dilation and curettage compared with those patients who undergo an endometrial biopsy to diagnose endometrial cancer. The key component of this study regarding tumor size should be explored and multicenter research emphasizing its importance will be needed so clinicians and pathologists will begin to incorporate size as a factor, along with grade and depth of invasion. Other important criteria to consider regarding lymph node dissection includes nodal count to ensure decreased risk of nodal metastases. Nodal count is often limited by retrospective single-institution reviews and pathology technique.19–21 Adequate nodal counts to identify a positive lymph node have also been reported from the Surveillance, Epidemiology, and End Results database from the National Cancer Institute to have a range of 21–25 lymph nodes.22 In our article, we did have a statistically significant difference in median nodal count between open or laparoscopic technique (22.5 compared with 21.0; P=.017), low-risk or high-risk modified Mayo criteria (20.0 compared with 22.0; P=.023), and identified nodal metastases (negative nodes of 21.0 compared with positive nodes 25; P=.018); however, the median count for these patients were within and or above previously published adequate nodal count criteria.
The receiver-operator characteristic curve for our logistic model (Fig. 2) shows a promising area under the curve of 0.822 (95% bootstrap CI 0.77–0.86), but the low prevalence of confirmed nodal metastasis among the study population (6.27%; 95% CI 0.05–0.08) suggests that the risk-group test sensitivity and specificity, which depend only on the test itself, are inadequate for characterizing the study population. A simpler model that discriminates for nodal metastasis only by high-risk or low-risk group membership has a positive predictive value of 0.11 (95% CI 0.083–0.134), but a negative predictive value of 0.992 (95% CI 0.978–0.998). Therefore, neither model is particularly useful for identifying patients at risk for nodal metastasis; however, the simpler model seems potentially useful for identifying patients without nodal metastasis. It is important to note that no single risk-group criterion (grade, depth of invasion, or size) was, by itself or paired with another criterion, capable of generating an adequately large positive predictive value or negative predictive value in the study population. For example, tumor grade of 1 or 2 was found in 60% of the patients with identified nodal metastases (Table 1). This alone would limit the utility of endometrial sampling before surgery to help predict lymph node metastases. Frozen-section data were not reported in this multicenter review. Previous reports have suggested limitations of frozen section in endometrial cancer; however, the use of binary (modified Mayo criteria low-risk compared with high-risk) criteria was not used, and a streamlined management may lead to more consistent intraoperative consultation.17
Minimally invasive surgical technique, including robotic surgery, may be associated with decreased perioperative complication rates, and the morbidity reported in this study may be higher than what is currently being seen in practice. However, given the median body mass index (calculated as weight (kg)/[height (m)]2) of 28.6 in this study and 89.5% asymptomatic performance status (Table 1) seen in this study, surgical improvements in technique may be offset by current patient comorbid conditions, which are often worse outside of a clinical trial.
In this multicenter post hoc analysis, modified Mayo criteria low-risk endometrioid uterine cancer characteristics were associated with a 0.8% rate of nodal metastasis. These data add further support to the utilization of tumor size as a criterion for lymph node dissection promoted by the Mayo Clinic. These criteria should be used to help guide treatment planning for reoperation in patients with incomplete surgical staging information.
1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin 2010;60:277–300.
2. Creasman WT, Morrow CP, Bundy BN, Homesley HD, Graham JE, Heller PB. Surgical pathologic spread patterns of endometrial cancer. A Gynecologic Oncology Group Study. Cancer 1987;60:2035–41.
3. Schink JC, Lurain JR, Wallemark CB, Chmiel JS. Tumor size in endometrial cancer: a prognostic factor for lymph node metastasis. Obstet Gynecol 1987;70:216–9.
4. Mariani A, Webb MJ, Keeney GL, Haddock MG, Calori G, Podratz KC. Low-risk corpus cancer: is lymphadenectomy or radiotherapy necessary? Am J Obstet Gynecol 2000;182:1506–19.
5. Mariani A, Dowdy SC, Cliby WA, Gostout BS, Jones MB, Wilson TO, et al.. Prospective assessment of lymphatic dissemination in endometrial cancer: a paradigm shift in surgical staging. Gynecol Oncol 2008;109:11–8.
6. Convery PA, Cantrell LA, Di Santo N, Broadwater G, Modesitt SC, Secord AA, et al.. Retrospective review of an intraoperative algorithm to predict lymph node metastasis in low-grade endometrial adenocarcinoma. Gynecol Oncol 2011;123:65–70.
7. Clinical management guidelines for obstetrician-gynecologists, management of endometrial cancer. ACOG Practice Bulletin Number 65. American College of Obstetricians and Gynecologists. Obstet Gynecol 2005;106:413–25.
8. Trimble CL, Kauderer J, Zaino R, Silverberg S, Lim PC, Burke JJ II, et al.. Concurrent endometrial carcinoma in women with a biopsy diagnosis of atypical endometrial hyperplasia: a Gynecologic Oncology Group study. Cancer 2006;106:812–9.
9. Chan JK, Wu H, Cheung MK, Shin JY, Osann K, Kapp DS. The outcomes of 27,063 women with unstaged endometrioid uterine cancer. Gynecol Oncol 2007;106:282–8.
10. Chan JK, Cheung MK, Huh WK, Osann K, Husain A, Teng NN, et al.. Therapeutic role of lymph node resection in endometrioid corpus cancer: a study of 12,333 patients. Cancer 2006;107:1823–30.
11. ASTEC study group, Kitchener H, Swart AM, Qian Q, Amos C, Parmar MK. Efficacy of systematic pelvic lymphadenectomy in endometrial cancer (MRC ASTEC trial): a randomised study. Lancet 2009;373:125–36.
12. Todo Y, Kato H, Kaneuchi M, Watari H, Takeda M, Sakuragi N. Survival effect of para-aortic lymphadenectomy in endometrial cancer (SEPAL study): a retrospective cohort analysis. Lancet 2010;375:1165–72.
13. Walker JL, Piedmonte MR, Spirtos NM, Eisenkop SM, Schlaerth JB, Mannel RS, et al.. Laparoscopy compared with laparotomy for comprehensive surgical staging of uterine cancer: Gynecologic Oncology Group Study LAP2. J Clin Oncol 2009;27:5331–6.
14. Fox J, Monette G. General collinearity diagnostics. JASA 1992;87:178–83.
15. Team RDC. R: A language and environment for statistical computing. Vienna, Austria: Computing RFfS; 2009.
16. Lee KB, Ki KD, Lee JM, Lee JK, Kim JW, Cho CH, et al.. The risk of lymph node metastasis based on myometrial invasion and tumor grade in endometrioid uterine cancers: a multicenter, retrospective Korean study. Ann Surg Oncol 2009;16:2882–7.
17. Leitao MM Jr, Kehoe S, Barakat RR, Alektiar K, Gattoc LP, Rabitt C, et al.. Accuracy of preoperative endometrial sampling diagnosis of FIGO grade 1 endometrial adenocarcinoma. Gynecol Oncol 2008;111:244–8.
18. Soliman PT, Frumovitz M, Spannuth W, Greer MJ, Sharma S, Schmeler KM, et al.. Lymphadenectomy during endometrial cancer staging: practice patterns among gynecologic oncologists. Gynecol Oncol 2010;119:291–4.
19. Huang M, Chadha M, Musa F, Friedmann P, Kolev V, Holcomb K. Lymph nodes: is total number or station number a better predictor of lymph node metastasis in endometrial cancer? Gynecol Oncol 2010;119:295–8.
20. Lutman CV, Havrilesky LJ, Cragun JM, Secord AA, Calingaert B, Berchuck A, et al.. Pelvic lymph node count is an important prognostic variable for FIGO stage I and II endometrial carcinoma with high-risk histology. Gynecol Oncol 2006;102:92–7.
21. Milam MR, Abaid L, dos Reis R, Frumovitz M, Gehrig PA, Livasy C, et al.. Microscopic evaluation of lymph-node-bearing tissue in early-stage cervical cancer: a dual-institution review. Ann Surg Oncol 2010;17:1106–10.
22. Chan JK, Urban R, Cheung MK, Shin JY, Husain A, Teng NN, et al.. Lymphadenectomy in endometrioid uterine cancer staging: how many lymph nodes are enough? A study of 11,443 patients. Cancer 2007;109:2454–60.
Supplemental Digital Content
© 2012 by The American College of Obstetricians and Gynecologists.