Risk assessment of lymph node metastasis before surgery in endometrial cancer: Do we need a clinical trial for low-risk patients?

bs_bs_banner doi:10.1111/jog.12281 J. Obstet. Gynaecol. Res. Vol. 40, No. 2: 322 326, February 2014 Risk assessment of lymph node metastasis before surgery in endometrial cancer: Do we need a clinical trial for low-risk patients? Sokbom Kang 1, Yukiharu Todo 2 and Hidemichi Watari 3 1 Center for Uterine Cancer, National Cancer Center, Goyang, Korea, 2 Division of Gynecology, Hokkaido Cancer Center, and 3 Department of Obstetrics and Gynecology, Hokkaido University Graduate School of Medicine, Sapporo, Japan Abstract Due to advances of radiological imaging and tumor biomarkers, the extent of information provided by preoperative assessment is rapidly growing. The Korean Gynecologic Oncology Group (KGOG) recently proposed new preoperative criteria to identify patients at low risk for lymph node metastasis in endometrial cancer. In the multicenter study, serum carbohydrate antigen 125 levels and three magnetic resonance imaging parameters were found to be independent risk factors for nodal metastasis, and classified 53% of patients as part of a low-risk group. The false-negative predictive value (NPV) was 1.7%, and was 1.4% in the validation set. Furthermore, the KGOG low-risk criteria were validated in 319 Japanese patients with endometrial cancer. The criteria identified 181 of 319 patients as a low-risk group (51%), and three false-negative cases were found (1.9%). These results indicate that we are able to identify low-risk patients with a negligible NPV before surgery. In addition, the low false NPV implies that there is great difficulty in performing a randomized trial to determine the efficacy of routine lymphadenectomy in patients at low risk of lymph node metastasis. Based on these data, the challenges and possible solutions for developing a consensus on the optimized management of low-risk endometrial cancer will be discussed in this review. Key words: endometrial cancer, lymphadenectomy, metastasis, risk factor, staging. Introduction Although endometrial cancer is the most common gynecologic malignancy observed in Western countries, it is also well known that most patients have low-risk disease. Therefore, identifying patients who may benefit from adjuvant treatment has been a major challenge in the disease. Since Sir John Stallworthy and his colleagues reported that a substantial number of patients with early stage endometrial cancer have lymph node metastasis, 1,2 gynecologic oncologists have studied the prognostic importance of lymph node metastasis even in apparently early stage disease. Moreover, since International Federation of Gynecology and Obstetrics (FIGO) decided to adopt surgical staging including systemic lymph node dissection, routine lymphadenectomy has been the standard treatment in the field. However, because the survival benefit of routine lymphadenectomy in patients with endometrial cancer has never been proven, the clinical value of routine lymphadenectomy remains controversial. Recently, two randomized clinical trials and one systemic meta-analysis indicated that routine lymphadenectomy provided no survival benefit in endometrial cancer. 3 5 In contrast to those publications, many experts have claimed a possible benefit of routine Received: July 18 2013. Accepted: August 21 2013. Reprint request to: Dr Sokbom Kang, Center for Uterine Cancer, National Cancer Center, Goyang 410-769, Korea. Email: sokbom@gmail.com Conflict of interest: The authors declare that there are no conflicts of interest. 322 2014 The Authors

Risk assessment of lymph node metastasis lymphadenectomy in patients with unfavorable risk factors. However, there has been no debate regarding the low-risk group. This raised a question whether we can safely omit lymphadenectomy in patients without risk factors. What Is a Negligible Rate of False Positivity? Few clinical tests are free from false negativity. Therefore, to adopt a preoperative risk assessment algorithm, gynecologists must ask themselves the following question: what level of false negativity do I ask my patients to accept? In other solid tumors, a falsenegative predictive value (NPV) of 5% is commonly suggested as an acceptable rate. The National Surgical Adjuvant Breast and Bowel Project trial B-32 was an extremely large randomized phase 3 trial that set out to determine the efficacy of sentinel node biopsy (SNB) in breast cancer. 6,7 After analyzing 5611 SNB procedures, the authors concluded that the false NPV and falsenegative rate for SNB in breast cancer were 4% and 10%, respectively. 6 Its performance resulted in equivalent overall survival, disease-free survival, and regional control between the SNB arm and the full dissection arm. 7 In addition, in vulvar cancer, the recent Gynecologic Oncology Group (GOG) study reported that the SNB procedure had a false NPV of 4%. 8 Based on its the diagnostic performance, the authors concluded that SNB is a reasonable alternative to inguinal femoral lymphadenectomy in selected women with vulvar cancer. These reports indicate that a false NPV of 4% is commonly adopted as an acceptable false negativity for lymph node metastasis. In endometrial cancer, the GOG reported a landmark clinicopathological study, GOG-33. Based on the data, the authors suggested a risk grouping according to the risk of lymph node metastasis: low risk (<5%), moderate risk (5 10%) and high risk (>10%). Other experts in the field also indicated that a negligible risk for lymph node metastasis was 0 4% and that low-risk patients accounted for 75% of the entire population with the disease. 9,10 However, a comparison of the false negativity of lymph node assessment between disease sites may not be justified if the morbidity of lymphadenectomy and the consequence of neglected node metastasis differ among them. Can a false NPV of 4% safely ensure an equivalent outcome between two strategies (routine lymphadenectomy and no lymphadenectomy)? Based on the results of two recent studies, the answer is probably yes 3,4 for the following reasons. First, in the two randomized trials reporting equivalent overall survival and disease-free survival between routine lymphadenectomy and no lymphadenectomy, the observed lymph node metastasis rates in the routine lymphadenectomy groups far exceeded those of the no lymphadenectomy groups. The difference was 10% (13% vs 3%) for the Italian multicenter randomized trial 3 and 7% (8% vs 1%) for the ASTEC trial. 4 If we assume that the lymph node metastasis rates were similar between the two groups in the two trials, we find that 7 10% of false negativity did not result in a survival difference. Second, in those two trials, both low-risk and high-risk patients were included. It is therefore evident that it would be nearly impossible to observe any survival difference between the two strategy groups if only lowrisk patients are selected. Although some researchers have criticized the randomized trials, noting that they were underpowered to detect a possible survival benefit associated with routine lymphadenectomy, a recent decision analysis indicated that at least a 10% rate of neglected pelvic node metastases would be required to produce a 5% difference in the predicted overall survival between a staging and selective radiotherapy arm and a hysterectomy and radiotherapy in a high-risk group arm. 11 According to the analysis, 4% of neglected node metastases would result in a predicted difference of 1% between the two strategies. Obviously, to perform a randomized trial discriminating such a small difference in outcome may be nearly impossible without an extremely large number of subjects and vast resources. Preoperative Criteria for Identifying a Low-risk Group Even in the GOG-33 study, the authors suggested that a low-risk group existed in which routine lymphadenectomy was not recommended. 12 However, after FIGO included lymphadenectomy in the surgical staging of endometrial cancer in 1988, the procedure was endorsed as a standard surgical treatment by many practice guidelines. 13,14 However, because of the low rates of lymph node metastasis, many researchers have attempted to identify a low-risk group to avoid morbidity associated with unnecessary lymphadenectomy. The most well-known assessment algorithm was an intraoperative decision criteria proposed by researchers from the Mayo Clinic. In a review of 815 patients, Mariani et al. indicated that a low-risk group could be 2014 The Authors 323

S. Kang et al. defined as patients with: (i) endometrioid subtype; (ii) myometrial invasion of 50% or less; (iii) histological grade 1 2; and (iv) no intraoperative evidence of macroscopic disease. 15 The authors claimed that these patients can be treated optimally with hysterectomy only. In 2004, they revised the criteria and tumor size (diameter of 2 cm) was newly introduced. 16 After a follow-up of 1393 patients, they identified 385 (27.6%) patients that composed a low-risk group. 17 When they omitted lymphadenectomy in 305 of the 385 patients (79%), the 5-year cause-specific survival rate was 99.0%. The intraoperative algorithm was validated in two independent cohorts. In the ancillary analysis of the GOG-LAP2 trial, a false NPV was 0.8% (3/389; 95% confidence interval [CI], 0.16 2.2) in a low-risk group identified using the Mayo criteria. 18 Another report observed a false NPV of 1.8% (2/110; 95% CI, 0.2 6.4). However, these two studies were not based on the data from intraoperative fresh frozen pathological examination, but on the final pathological review. Thus, it is still unclear whether the intraoperative criteria will show good performance in other clinical settings. Indeed, some experts have claimed that the feasibility of making such an accurate pathological decision intraoperatively has only been proven in the Mayo Clinic. 19 Another attempt to tailor lymphadenectomy according to preoperative assessment was made by Japanese researchers. Todo et al. reported that patients with no risk factors (elevated serum carbohydrate antigen [CA]-125 level [ 70 U/mL, if aged <50 years; 28, if aged 50 years], volume index <25, serous histological type, and histological grade 3) have a low risk of pelvic lymph node metastasis (4/110 cases, 3.6%). 20 Based on these data, the authors claimed that para-aortic lymph node metastasis can be safely omitted in the low-risk group. The authors modified these algorithms into a scoring system with four risk factors (grade 3 or serous histological subtype, volume index 36, magnetic resonance imaging [MRI]-based myometrial invasion 1/2, and a high serum CA-125 level [ 70 U/ml, if aged <50 years; 28, if aged 50 years]). 21 Using the scoring system, the authors found lymph node metastasis in only 3.3% of the low-risk patients. KCOG Low-risk Criteria Although these preoperative or intra-operative risk assessment algorithms showed that the prediction of a low-risk group with a node metastasis rate of less than 4% was feasible, many gynecologists did not adopt those algorithms in clinical practice. Therefore, the Korean Gynecologic Oncology Group (KGOG) initiated a multicenter observational study to develop a well-validated guideline for identifying a low-risk group before surgery. 22 The study had two end-points. First, the criteria should identify a low-risk group with a node metastasis rate of less than 4%. Second, the criteria should identify as many low-risk patients as possible. In addition, the KGOG preferred preoperative criteria to intraoperative criteria because of several reasons: First, preoperative criteria are more helpful in patient counseling and shared decision-making. Second, preoperative criteria are more useful in future clinical trials in terms of the random allocation of enrolled subjects. Third, preoperative criteria are more convenient for surgeons because time is not spent waiting for pathological results. Using the MRI data and serum CA-125 levels of 360 patients from four institutions, a multivariate logistic regression model was developed. 22 The analysis revealed four risk factors: (i) myometrial invasion of 50% or more on MRI; (ii) lymph node enlargement on MRI; (iii) extrauterine tumor spread, including cervix, on MRI; and (iv) serum CA-125 greater than 35 U/mL. The low-risk group was characterized as patients without any of these risk factors. These criteria identified 53% of patients as a low-risk group, and the false NPV was 1.7%. When the low-risk criteria were further validated in 180 patients from two independent institutions, they identified 43% of patients as a low-risk group, and the false NPV was 1.4%. However, although the KGOG criteria were validated both internally and externally, there is a need to test whether the good performance of the criteria is reproducible in diverse clinical settings such as those with different surgical policies, imaging instruments, assay methods or pathologist s experience. Therefore, the KGOG designed a collaborative study to validate the efficacy of the KGOG low-risk model in a Japanese cohort. 23 The KGOG low-risk criteria identified 51% of patients (181/319) from two Japanese hospitals. The false NPV was 1.4% (95% CI, 0.5 4.3%). Randomized Trial in a Low-risk Group: Is It Necessary? The accumulated data showing the feasibility of pre- or intraoperative prediction may prompt us to initiate a randomized trial to assess the equality of the survival outcome between routine lymphadenectomy and no lymphadenectomy strategies. However, such a trial 324 2014 The Authors

Risk assessment of lymph node metastasis may be challenged for the following reasons. As mentioned above, the two randomized trials reported equivalent survival between the two strategies despite a 7 10% rate of neglected lymph node metastasis. 3,4 Moreover, those two trials included both low-risk and high-risk patients. Thus, it would be extremely difficult to observe any survival difference between the two strategies in low-risk patients. If a survival difference cannot be an ideal end-point, then, can improved quality of life be an alternative end-point? Unfortunately, it does not appear so. A recent meta-analysis of Cochrane Database System indicated that lymphadenectomy significantly increased systemic morbidity (relative risk, 3.7) and lymphedema formation (relative risk, 8.4). 5 Another population-based study from the Netherlands reported that women receiving lymphadenectomy reported no clinically relevant better healthrelated quality of life. 24 Therefore, neither improved survival nor improved quality of life would be expected as a merit of routine lymphadenectomy. However, although a conventional randomized trial does not appear feasible, we can plan an alternative study such as the GROINSS-VI study in vulvar cancer. 25 The GROINSS-VI study was designed as a multicenter, international, observational study investigating the safety and clinical utility of SNB in early stage vulvar cancer patients. The authors omitted inguinofemoral lymphadenectomy when the sentinel node was found to be negative. They defined the groin recurrence rate after inguinofemoral lymphadenectomy and no evidence of nodal metastasis as 2% in early stage vulvar cancer. They also defined the maximum acceptable increase in the groin recurrence rate as 6% in light of an anticipated significant decrease in treatment-related morbidity. Indeed, they observed a 2.3% groin recurrence rate and a 97% 3-year survival rate. However, there was also concern that even a 2% groin recurrence rate was high compared to the GOG historical controls. 26 Nonetheless, the truly important indicator is what patients really perceive as the maximum acceptable risk. Frequently, patients do not want to accept an additional increase of risk and are willing to accept substantial toxicity related to treatment. 27 Therefore, to design such a study scheme, a sufficient consensus on the extent of acceptable risk should be drawn from both patients and gynecologists. Conclusion In endometrial cancer, less than a 4% lymph node metastasis risk may be regarded as acceptable. Recent evidence has repeatedly suggested that to identify such a subset before or during surgery is feasible. In particular, the KGOG low-risk criteria successfully identified such a low-risk subset in both the model-developing cohort and several independent cohorts regardless of geographic differences and surgical policy. Because there is no evidence to suggest an improved survival outcome or improved quality of life in the low-risk group as a result of lymph node dissection, it should not be performed routinely in these low-risk patients. It is therefore crucially important for gynecologic oncologists to give sufficient information to their patients and involve them in shared decision-making. In the lowrisk group, lymph node dissection is recommended to be performed only in the context of clinical trials. References 1. Stallworthy JA. Surgery of endometrial cancer in the Bonney tradition. Ann R Coll Surg Engl 1971; 48: 293 305. 2. Lewis BV, Stallworthy JA, Cowdell R. Adenocarcinoma of the body of the uterus. J Obstet Gynaecol Br Commonw 1970; 77: 343 348. 3. Benedetti Panici P, Basile S, Maneschi F et al. Systematic pelvic lymphadenectomy vs. no lymphadenectomy in earlystage endometrial carcinoma: Randomized clinical trial. J Natl Cancer Inst 2008; 100: 1707 1716. 4. Kitchener H, Swart AM, Qian Q et al. Efficacy of systematic pelvic lymphadenectomy in endometrial cancer (MRC ASTEC trial): A randomised study. Lancet 2009; 373: 125 136. 5. May K, Bryant A, Dickinson HO et al. Lymphadenectomy for the management of endometrial cancer. Cochrane Database Syst Rev 2010; (1): CD007585. 6. Krag DN, Anderson SJ, Julian TB et al. Technical outcomes of sentinel-lymph-node resection and conventional axillarylymph-node dissection in patients with clinically nodenegative breast cancer: Results from the NSABP B-32 randomised phase III trial. Lancet Oncol 2007; 8: 881 888. 7. Krag DN, Anderson SJ, Julian TB et al. Sentinel-lymph-node resection compared with conventional axillary-lymph-node dissection in clinically node-negative patients with breast cancer: Overall survival findings from the NSABP B-32 randomised phase 3 trial. Lancet Oncol 2010; 11: 927 933. 8. Levenback CF, Ali S, Coleman RL et al. Lymphatic mapping and sentinel lymph node biopsy in women with squamous cell carcinoma of the vulva: A Gynecologic Oncology Group study. J Clin Oncol 2012; 30: 3786 3791. 9. Boronow RC. Surgical staging of endometrial cancer: Evolution, evaluation, and responsible challenge a personal perspective. Gynecol Oncol 1997; 66: 179 189. 10. Sakuragi N. Emerging concept of tailored lymphadenectomy in endometrial cancer. J Gynecol Oncol 2012; 23: 210 212. 11. Naumann RW. The role of lymphadenectomy in endometrial cancer: Was the ASTEC trial doomed by design and are we destined to repeat that mistake? Gynecol Oncol 2012; 126: 5 11. 12. Creasman WT, Morrow CP, Bundy BN et al. Surgical pathologic spread patterns of endometrial cancer. A Gynecologic Oncology Group Study. Cancer 1987; 60 (8 Suppl): 2035 2041. 2014 The Authors 325

S. Kang et al. 13. Greer BE, Koh WJ, Abu-Rustum N et al. Uterine neoplasms. Clinical practice guidelines in oncology. J Natl Compr Canc Netw 2009; 7: 498 531. 14. ACOG. ACOG practice bulletin, clinical management guidelines for obstetrician-gynecologists, number 65, August 2005: Management of endometrial cancer. Obstet Gynecol 2005; 106: 413 425. 15. Mariani A, Webb MJ, Keeney GL et al. Low-risk corpus cancer: Is lymphadenectomy or radiotherapy necessary? Am J Obstet Gynecol 2000; 182: 1506 1519. 16. Mariani A, Dowdy SC, Keeney GL et al. High-risk endometrial cancer subgroups: Candidates for target-based adjuvant therapy. Gynecol Oncol 2004; 95: 120 126. 17. Dowdy SC, Borah BJ, Bakkum-Gamez JN et al. Prospective assessment of survival, morbidity, and cost associated with lymphadenectomy in low-risk endometrial cancer. Gynecol Oncol 2012; 127: 5 10. 18. Milam MR, Java J, Walker JL et al. Nodal metastasis risk in endometrioid endometrial cancer. Obstet Gynecol 2012; 119 (2 Pt 1): 286 292. 19. Walker JL. Re: Clinical practice guidelines for the management of patients with endometrial cancer in France. Int J Gynecol Cancer 2012; 22: 179 181. 20. Todo Y, Sakuragi N, Nishida R et al. Combined use of magnetic resonance imaging, CA 125 assay, histologic type, and histologic grade in the prediction of lymph node metastasis in endometrial carcinoma. Am J Obstet Gynecol 2003; 188: 1265 1272. 21. Todo Y, Okamoto K, Hayashi M et al. A validation study of a scoring system to estimate the risk of lymph node metastasis for patients with endometrial cancer for tailoring the indication of lymphadenectomy. Gynecol Oncol 2007; 104: 623 628. 22. Kang S, Kang WD, Chung HH et al. Preoperative identification of a low-risk group for lymph node metastasis in endometrial cancer: A Korean gynecologic oncology group study. J Clin Oncol 2012; 30: 1329 1334. 23. Kang S, Todo Y, Odagiri T et al. A low-risk group for lymph node metastasis is accurately identified by Korean gynecologic oncology group criteria in two Japanese cohorts with endometrial cancer. Gynecol Oncol 2013; 129: 33 37. 24. van de Poll-Franse LV, Pijnenborg JM, Boll D et al. Health related quality of life and symptoms after pelvic lymphadenectomy or radiotherapy vs. no adjuvant regional treatment in early-stage endometrial carcinoma: A large populationbased study. Gynecol Oncol 2012; 127: 153 160. 25. Van der Zee AG, Oonk MH, De Hullu JA et al. Sentinel node dissection is safe in the treatment of early-stage vulvar cancer. J Clin Oncol 2008; 26: 884 889. 26. Levenback CF. How safe is sentinel lymph node biopsy in patients with vulvar cancer? J Clin Oncol 2008; 26: 828 829. 27. de Hullu JA, Ansink AC, Tymstra T et al. What doctors and patients think about false-negative sentinel lymph nodes in vulvar cancer. J Psychosom Obstet Gynaecol 2001; 22: 199 203. 326 2014 The Authors