Gynecologic Oncology 135 (2014) 435–440
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Prognostic value of lymph node ratio in patients with advanced epithelial ovarian cancer Beyhan Ataseven ⁎, Christoph Grimm, Philipp Harter, Sonia Prader, Alexander Traut, Florian Heitz, Andreas du Bois Department of Gynecology and Gynecologic Oncology, Kliniken Essen-Mitte, Evangelische Huyssens-Stiftung, Henricistrasse 92, 45136 Essen, Germany
H I G H L I G H T S • Lymph node status is a prognostic factor in ovarian cancer. • Lymph node ratio reflects lymph node spread and surgical extent. • Lymph node ratio predicts overall survival more concisely.
a r t i c l e
i n f o
Article history: Received 21 August 2014 Accepted 1 October 2014 Available online 12 October 2014 Keywords: Lymph node ratio Ovarian cancer Prognosis Systematic lymphadenectomy
a b s t r a c t Objective. Lymph node status is an established prognostic factor in epithelial ovarian cancer (EOC). Lymph node ratio (number of positive LN/number of resected LN) reflects both qualitative and quantitative lymph node spread as well as surgical effort and extent of disease. We evaluated whether LNR is a more precise prognostic factor than conventional lymph node status in patients with EOC. Methods. The present retrospective study includes 809 patients with EOC, who underwent primary cytoreductive surgery between 2000–2013. Clinico-pathological parameters and survival data were extracted from a prospectively maintained tumor registry database. The optimal cut-off point for LNR was calculated by using Martingale residuals. Survival analyses were calculated using Kaplan–Meier method and Cox regression models. Results. Lymphadenectomy was performed in 693 (85.7%) out of 809 patients. Median number of removed LN was 64 (IQR 25–75%: 39–84). LNR of 0.25 was identified as the optimal prognostic cut-off value. The estimated 5-year-OS rates were 69.3% for patients with node-negative EOC compared to 33.1% for patients with any lymph node metastasis (p b 0.001). The estimated 5-year-OS rates were 42.5% for patients with LNR ≤ 0.25, and 18.0% for patients with LNR N 0.25 (p b 0.001). Additionally in multivariate analysis LNR N 0.25 was approved to be an independent prognostic factor for overall survival (adjusted HR 1.44, 95% CI 1.04–2.00; p = 0.028). Conclusion. LNR more precisely predicts overall survival than conventional lymph node status in EOC patients undergoing primary debulking surgery. © 2014 Elsevier Inc. All rights reserved.
Introduction Epithelial ovarian cancer (EOC) is the most lethal gynecologic malignancy [1]. EOC has a very unique tumor biology and pattern of metastasis with predominantly peritoneal and lymphovascular spread and rarely hematological spread. The role of complete tumor resection of mainly intraperitoneal disease has well been described, however, the prognostic value of lymph node metastasis, its impact on residual
⁎ Corresponding author. Fax: +49 20117434200. E-mail address: [email protected] (B. Ataseven).
http://dx.doi.org/10.1016/j.ygyno.2014.10.003 0090-8258/© 2014 Elsevier Inc. All rights reserved.
tumor status and thus, the therapeutic value of comprehensive pelvic and paraaortic lymphadenectomy have not been fully clarified in EOC. Age, tumor stage, and residual disease are the major prognostic factors in EOC [2,3]. The prognostic value of lymph node metastases is less well defined. However, in clinically apparent stage I disease, the presence of LN metastasis causes an upstaging possibly leading to a subsequent change in treatment. Studies addressing the therapeutic value of comprehensive pelvic and paraaortic lymphadenectomy in patients with EOC are limited. Retrospective analyses indicate a benefit particularly in optimally debulked ovarian cancer cases and early stage EOC [4,5]. In advanced EOC this benefit has not been consistently shown. Nonetheless,
436
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
complete gross resection of all visible disease after surgery is the surgical goal in treating EOC [6]. The therapeutic role of comprehensive lymphadenectomy in EOC patients is currently being addressed in a prospective randomized controlled trial (lymphadenectomy in ovarian neoplasms (LION) NCT00712218). Lymph node ratio (LNR) is the number of positive lymph nodes divided by the total number of removed lymph nodes. LNR has been proven to be an important prognostic factor in a variety of malignancies including breast, gastric, esophageal, and non-small cell lung cancer, and melanoma [7–11]. With respect to EOC, one retrospective analysis has reported that high LNR was independently associated with decreased overall survival in patients with advanced EOC [12]. Patients from the Surveillance, Epidemiology, and End Results Program (SEER) were included in this study. A major limitation of this retrospective analysis was the extent of lymphadenectomy. Only 58.0% and 21.1% of patients had ≤10 and 11–20 lymph nodes removed, respectively. Just a small amount of 20.8% of patients had a comprehensive lymphadenectomy with N20 lymph nodes removed. The aim of our present study was to evaluate the additional prognostic value of LNR compared to conventional lymph node status in patients with EOC that have undergone systematic pelvic and paraaortic lymphadenectomy during primary upfront debulking surgery. Methods Patients with first diagnosis of invasive epithelial ovarian, fallopian tube or serous primary peritoneal cancer who attended our tertiary gynecologic oncology center (Horst–Schmidt–Kliniken (HSK), Wiesbaden and Kliniken Essen-Mitte (KEM, Essen)) were prospectively documented in a clinical tumor registry from 2000 until present. All patients' follow-up data are updated annually. In the present study, we included only patients with EOC (n = 906; HSK n = 630; KEM n = 276) who underwent primary upfront cyctoreductive surgery in our institution. Patients receiving neo-adjuvant chemotherapy and subsequently attending our center for interval debulking (n = 97) were excluded. Furthermore, patients with borderline tumors and non-epithelial ovarian cancer were also not included in this analysis. All patients gave informed consent for the surgical procedure and documentation of clinical data in our clinical tumor registry. Surgery was performed by a dedicated surgical team. In all patients with macroscopic residual disease routine intra-operative 2nd opinion was obtained from a senior gynecological oncologist before wound closure. The indication for systematic pelvic and para-aortic lymphadenectomy changed over the study period. Between 2000–2006 systematic lymphadenectomy was performed in patients with age ≤75-years and surgical debulking without residual tumor intraabdominally or tumor residuals up to 10 mm. After 2006 systematic lymphadenectomy was performed only in patients with complete macroscopic tumor resection, whereas in patients with 1–10 mm residual tumor only bulky nodes were resected. For this retrospective analysis we defined systematic lymphadenectomy as a complete procedure when at least 30 lymph nodes were reported in the pathologic specimen. Platinum-based chemotherapy was given stage adapted after surgery to all patients. To evaluate the optimal cut-off point for LNR (number of positive LN/total resected LN) Martingale residuals [13] was used. Node positive patients were categorized either in low LNR group (≤0.25) or high LNR group (N0.25). For comparison of the total cohort we divided patients into four LN groups: NX (lymphadenectomy not performed), N0 (lymphadenectomy performed, negative for metastasis), LNR ≤ 0.25 (lymphadenectomy performed, positive for metastasis; LNR ≤ 0.25) and LNR N 0.25 (lymphadenectomy performed, positive for metastasis, LNR N 0.25). To further elucidate the prognostic value of LNR, we additionally performed analysis in a more restrictive patient cohort (complete gross resection — complete LN dissection (CGR-LND group)). Therefore we evaluated the impact of LNR only in patients with tumor stage pT3c,
optimal debulking surgery (defined as complete resection) and systematic lymphadenectomy (number of resected LN N 30). This cohort is henceforward referred to as the “CGR-LND group”. To eliminate bias caused by a different extent of lymphadenectomy in the low and high LNR groups, we evaluated the number of total lymph nodes removed. Moreover, we compared both groups with respect to length of operation, blood loss, and quantification of surgical effort using the Surgical Complexity Score (SCS) previously described [14]. Statistical analysis was performed using SPSS™ version 20.0 (Chicago, Illinois, USA) and GraphPad Prism® 4 (La Jolla, CA, USA). Cut-off value for LNR was calculated by Martingale residuals. We performed multivariate Cox regression analyses comprising tumor stage, distant metastases, tumor grade, histologic type, patient's age, performance status, and surgical outcome/residual tumor to determine whether the LNR had any significant impact on overall survival (OS). Univariate survival analyses were provided as Kaplan–Meier curves and differences calculated by log-rank tests. Association between LNR and OS was calculated by using a Cox proportional hazard regression model. Results 809 patients met inclusion criteria for this analysis. Patient and tumor characteristics are shown in Table 1. Two hundred seventynine (34.5%) patients were 65 years or older. The majority of patients (n = 686; 84.8%) were in a good performance status (Eastern Cooperative Oncology group; ECOG 0), diagnosed with pT3-stage (n = 581;
Table 1 Patient and tumor characteristics. Variable
Total cohort
Sensitivity-analysis cohort #
Patients (n)
%
Patients (n)
%
Total Age⁎
809
100
210
100
≤64 yr N64 yr Performance status ECOG 0 ECOG N0 FIGO-stage FIGO I FIGO II FIGO III FIGO IV (=M1) Grade Low-grade High-grade Histology Serous Others pN-stage NX N0 N1 Lymph node resected 0 ≤30 N30 Lymph node ratio (LNR)‡⁎ NX N0 ≤0.25 N0.25 Residual disease after surgery 0 mm 1–10 mm N10 mm
530 279
65.5 34.5
155 55
73.8 26.2
686 123
84.8 15.2
200 10
95.2 4.8
133 50 404 222
16.4 6.2 49.9 27.4
0 0 132 78
62.9 37.1
57 752
7.0 93.0
4 206
1.9 98.1
589 220
72.8 27.2
189 21
90.0 10.0
116 295 398
14.3 36.5 49.2
0 45 165
0 21.4 78.6
116 152 541
14.3 18.8 66.9
0 0 210
116 295 244 154
14.3 36.5 30.2 19.0
0 45 123 42
573 170 66
70.8 21.0 8.2
210 0 0
0 0 100 0 21.4 58.6 20.0 100 0 0
⁎ Cut by martingale residue. ‡ LNR is defined as the number of positive nodes divided by the number of resected nodes. # Sensitivity-analysis cohort (stage pT3c, residual disease 0 mm, systematic lymphadenectomy).
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
71.8%), had a high-grade carcinoma (n = 752; 93.0%) and a serous histology (n = 589; 72.8%). Any kind of lymphadenectomy was performed in 693 (85.7%) patients (median number of removed LN 65 (IQR 25– 75%: 39–84)). In 152 (18.8%) patients 1–30 nodes were dissected with a median number of 7 dissected nodes (IQR 25–75%: 3–15). In 541 (66.9%) patients more than 30 nodes were dissected (systematic pelvic and para-aortic lymphadenectomy) with a median number of 73 dissected nodes (IQR 25–75%: 59–89). Lymph node metastases were evident in 389 of 693 (56.1%) patients. Complete gross cytoreduction (defined as no macroscopic tumor) was achieved in 573 (70.8%), cytoreduction with gross residual disease of 1–10 mm in 170 (21.0%) and gross residual disease N 10 mm in 66 (8.2%) patients. According to the LN groups 116 (14.3%) patients were categorized as NX, 295 (36.5%) as N0, 244 (30.2%) as LNR ≤ 0.25 and 154 (19.0%) as LNR N 0.25. Patient and tumor characteristics for the more restrictive sensitivityanalysis group differed slightly from the total cohort in terms of included patients (n = 210), age (N 64 years: 26.2%), performance status (ECOG 0: 95.2%), tumor grade (high grade: 98.1%), histology (serous: 90.0%) and LN group (N0: 21.4%; LNR ≤ 0.25: 58.6%; LNR N 0.25: 20.0%). After a median follow up of 45 months (1–173) 332 (41.0%) patients had died. Estimated 5-year OS rate for the entire cohort was 46.5%. Based on conventional nodal status, patients with no evidence for LN metastasis (N0) showed more favorable 5-year OS (69.3%) compared to patients without lymphadenectomy (32.0%) and patients with positive LN (33.1%) (Fig. 1; p = b0.001). After stratification of lymph node positive patients according to LNR, 5-year OS rates were 42.5% for patients with LNR ≤ 0.25 and 18.0% for patients with LNR N 0.25 (Fig. 2; p b 0.001). Further prognostic factors in univariate analysis were patient's age, performance status, tumor stage, distant metastases, FIGO stage, tumor grade, histologic type, nodal status, number of resected lymph nodes, and residual tumor (Table 2). In multivariate Cox regression analysis, LNR was an independent prognostic factor for OS after adjusting for other clinically relevant factors in univariate analysis (Table 2). Patients with LNR N 0.25 had a significantly worse prognosis HR 1.44 (95% CI 1.04–2.00; p = 0.028). Moreover, we performed sensitivity analysis for LNR and evaluated its prognostic impact in the CGR-LND group. Only patients with systematic lymphadenectomy, tumor stage pT3c, and without any residual tumor were included. Again LNR N 0.25 remained an independent prognostic factor for OS with a hazard ratio of 2.10 (95% CI 1.06–4.16; p = 0.034; Table 3). After stratification for LNR, 5-year OS rates in this group were 33.0% for patients with LNR ≤ 0.25 and 22.0% for patients with LNR N 0.25 (p = 0.178; Fig. 3), respectively. To rule out a potential
437
bias regarding surgical efforts, we compared the number of resected lymphnodes in subgroups with high or low LNR. We could not find a significant difference between the LNR ≤ 0.25 and LNR N 0.25 cohort regarding mean number of resected lymph nodes (n = 81 versus 83; p = 0.743), mean blood loss (ml = 1065 versus 1247; p = 0.102), and mean length of operation (minutes = 446 versus 486; p = 0.133), respectively. In terms of extent and complexity of surgery no significant difference was detected between both groups as 98.4% of patients in the LNR ≤ 0.25 and 97.6% in the LNR N 0.25 group were categorized in high Surgical Complexity Score group (Supplementary data Table 2). Discussion In the present study, we show that the lymph-node-ratio (LNR) could serve as an additional prognostic parameter to better predict outcome in node positive patients with epithelial ovarian cancer (EOC). Overall, nodal positive patients had an impaired prognosis compared to node negative EOC patients. Furthermore, nodal positive patients with a higher LNR had an even worse prognosis. A LNR cut-off value of 0.25 was identified to maximally discriminate prognosis. The estimated 5-year-OS rates were 69.3% for patients with node-negative cancer compared to 33.1% for patients with any lymph node metastasis. Stratified by LNR the estimated 5-year-OS rates were 42.5% for patients with LNR ≤ 0.25 compared to 18.0% for patients with LNR N 0.25. Lymph node status is a well-established prognostic parameter in patients with EOC [4,15] and this is also reflected in the new FIGO classification, where lymph node status and size of lymph node metastases define FIGO tumor stage [16]. LNR has the potential to add another dimension besides positivity, location and size of LN metastasis to the prognostic value of nodal spread in EOC. The potential value of LNR may be based on its composition of both tumor biology and extent of surgical effort. The latter may also be one of the theoretical weaknesses of this parameter. In the present study, LNR provided additional information compared to lymph node status alone. Patients with any lymph node metastasis presented with a median overall survival of 44 months. If stratified by LNR patients with LNR ≤ 0.25 compared to patients with LNR N 0.25 presented with a median overall survival of 52 and 30 months, respectively. However, all patients were operated rather homogeneously in a high volume center with high rates of complete resection and high numbers of resected nodes. Further studies in other institutions and multicenter trials may elucidate the generalizability of our observations. Nevertheless, we tried to challenge our first
Fig. 1. Overall survival by lymph node status. Kaplan–Meier.
438
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
Fig. 2. Overall survival by lymph node ratio (LNR). Kaplan–Meier.
Table 2 Univariate and multivariate analyses of prognostic factors for overall survival (OS) (total cohort; n = 809). Variable
Patients (n)
%
Univariate OS event n (%)
Total Age⁎ ≤64 yr N64 yr Performance status ECOG 0 ECOG N0 Secondary malignancy Yes No Tumor size T1 T2 T3 M-stage M0 M1 (FIGO IV) Grade Low-grade High-grade Histology Others Serous pN-stage NX N0 N1 Lymph node resected 0 ≤30 N30 Lymph node ratio (LNR)‡⁎ N0 ≤0.25 N0.25 NX Residual disease after surgery 0 mm 1–10 mm N10 mm
809
Multivariate p-Value
HR
95% CI
p-Value
332 (41.0)
530 279
65.5 34.5
191 (36) 141 (50.5)
b0.001
1 1.34
1.06–1.70
0.016
686 123
84.8 15.2
244 (35.6) 88 (71.5)
b0.001
1 1.38
1.04–1.80
0.023
124 685
15.3 84.7
46 (37.1) 286 (41.8)
0.160
153 75 581
18.9 9.3 71.8
14 (9.2) 15 (20) 303 (52.2)
b0.001
1 2.47 7.36
1.16–5.26 4.04–13.4
b0.001 0.019 b0.001
587 222
72.6 27.4
214 (36.5) 118 (53.2)
b0.001
1 1.30
1.03–1.66
0.030
57 752
7.0 93.0
6 (10.5) 326 (43.4)
b0.001
1 2.19
0.97–4.95
220 589
27.2 72.8
64 (29.1) 268 (45.5)
b0.001
1 0.85
0.64–1.13
0.255
116 295 398
14.3 36.5 49.2
67 (57.8) 79 (26.8) 186 (46.7)
b0.001
116 152 541
14.3 18.8 66.9
67 (57.8) 78 (51.3) 187 (34.6)
b0.001
295 244 154 116
36.5 30.2 19.0 14.3
79 (26.8) 98 (40.2) 88 (57.1) 67 (57.8)
1 1.30 1.44 1.25
0.95–1.77 1.04–2.00 0.87–1.81
0.157 0.096 0.028 0.230
573 170 66
70.8 21.0 8.2
168 (29.3) 114 (67.1) 50 (75.8)
1 1.84 3.07
1.40–2.43 2.07–4.54
b0.001 b0.001 b0.001
⁎ Cut by martingale residue. ‡ LNR is defined as the number of positive nodes divided by the number of resected nodes.
b0.001
b0.001
0060
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
439
Table 3 Univariate and multivariate analysis of prognostic factors for overall survival (OS) (Complete gross resection — complete LN dissection; CGR-LND group; n = 210). Variable
Patients (n)
%
Univariate OS event n (%)
Total events PFS/OS Age⁎ ≤64 yr N64 yr Performance status ECOG 0 ECOG N0 M-stage M0 M1 (FIGO IV) Grade Low-grade High-grade Histology Others Serous pN-stage N0 N1 Lymph node ratio (LNR)‡⁎ N0 ≤0.25 N0.25 Surgical complexity scoring (median split) ≤13 N13
210
Multivariate p-Value
HR
95% CI
p-Value
88 (41.9)
155 55
73.8 26.2
58 (37.4) 30 (54.5)
0.058
1 1.46
0.90–2.37
0.122
200 10
95.2 4.8
80 (40) 8 (80)
0.021
1 1.87
0.83–4.20
0.129
132 78
62.9 37.1
55 (41.7) 33 (42.3)
0.019
1 1.17
0.73–1.83
0.525
4 206
1.9 98.1
1 (25) 87 (42.2)
0.461
1 1.68
0.23–12.4
0.612
21 189
10 90
8 (38.1) 80 (42.3)
0.734
1 1.27
0.60–2.71
0.655
45 165
21.4 78.6
45 (21.4) 68 (41.2)
0.071
45 123 42
21.4 58.6 20
20 (44.4) 48 (39) 20 (47.6)
0.050
1 1.45 2.10
0.82–2.56 1.06–4.16
0.105 0.202 0.034
142 68
67.6 32.4
63(44.4) 25(36.8)
0.135
1 1.48
0.90–2.44
0.125
⁎ Cut by martingale residue. ‡ LNR is defined as the number of positive nodes divided by the number of resected nodes.
results and performed a sensitivity analysis in a more restricted CGRLND group. This CGR-LND group consisted of EOC patients with tumor stage pT3c, optimal debulking surgery, and no residual tumor. Even in this restrictive analysis LNR remained an independent prognostic parameter for OS in uni- and multivariate analyses. Mahdi et al. [12] had also evaluated the prognostic impact of LNR in 6310 patients with advanced ovarian cancer (FIGO IIIC and IV) with and without macroscopic peritoneal disease obtained from SEER database. Patients were stratified according to the LNR in three groups (LNR b 10%, LNR 10–50% and LNR N 50%). LNR was an independent prognostic factor for survival, particularly in patients with no peritoneal
macroscopic disease. However, the SEER analysis differed from our study: The authors defined comprehensive lymphadenectomy as dissection of N10 lymph nodes. Fifty-eight percent of patients had a lymphadenectomy with less than 10 LNs and only 21% of the patients had N20 LN dissected. This may rather describe lymph node sampling in a considerable portion of patients and therefore LNR ratios may differ in both cohorts. The role of systematic lymphadenectomy has been mentioned by Pereira et al. [17] who concluded that the probability of obtaining a positive LN could only be predicted if at least 22 LNs (15 pelvic and 7 aortic) were dissected. In our study we defined systematic lymphadenectomy if N30 LNs were reported. This is based on the
Fig. 3. Sensitivity analysis with estimated overall survival for lymph node ratio (LNR). Other covariates are fixed at their average values.
440
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
currently ongoing LION trial, in which systematic lymphadenetomy is defined as comprising at least 20 pelvic lymph nodes and 10 paraaortic lymph nodes. Future studies on LNR should include a precise definition of the surgical scenario — in particular the systematic approach and extent of lymphadenectomy. This is crucial, as diverging surgical effort can impair the informative value when the prognostic value of LNR is investigated. Our selection criteria of the sensitivity-analysis-cohort and the definition of the LNR cut-off value may be used in further studies to obtain a more standardized model that can be used for comparison and validation of our findings. The interest in LNR as prognosticator in EOC is supported by other analyses which showed that LNR was an important prognostic factor in gastric cancer, breast cancer, esophageal cancer, melanoma, thyroid papillary carcinoma, gallbladder carcinoma and non-small cell lung cancer [7–11,18–21]. The present study has some limitations which underscore that this is a hypothesis building analysis and far from any definitive and conclusive statement. As already mentioned, this is a retrospective single-team study bearing potential risk for selection bias and a bias with respect to our surgical school which may not be comparable to ovarian cancer surgery worldwide [3,22]. Nevertheless, this potential bias could also be regarded as strengths of this series: all surgeries had been performed by high-volume surgeons at a high-volume center with an expertise in gynecologic oncology. Due to the structured quality management program on surgical training and the 2nd opinion procedure, as described before [3], we were able to compare the operative procedures and their impact in a very standardized way and may rule out a bias regarding the extent of lymphadenectomy performed from case to case and by a heterogeneous cohort of surgeons. Nonetheless, the statistical calculation of cut-off values by Martingale residuals is a well-established statistical approach to identify the optimal cut-off value. At any case, a validation of our result with data from another high-volume institution is necessary to support the power of our study and define more precisely the impact of LNR on the prognosis of EOC patients. If confirmed, LNR could be an easily assessable parameter in operated EOC, which allows a favorable and further prognostic discrimination of node-positive patients. Conflict of interest The authors declare no conflict of interest.
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ygyno.2014.10.003. References [1] Jemal A, Simard EP, Dorell C, Noone AM, Markowitz LE, Kohler B, et al. Annual Report to the Nation on the Status of Cancer, 1975–2009, featuring the burden and trends in human papillomavirus (HPV)-associated cancers and HPV vaccination coverage levels. J Natl Cancer Inst 2013;105:175–201.
[2] du Bois A, Reuss A, Pujade-Lauraine E, Harter P, Ray-Coquard I, Pfisterer J. Role of surgical outcome as prognostic factor in advanced epithelial ovarian cancer: a combined exploratory analysis of 3 prospectively randomized phase 3 multicenter trials: by the Arbeitsgemeinschaft Gynaekologische Onkologie Studiengruppe Ovarialkarzinom (AGO-OVAR) and the Groupe d'Investigateurs Nationaux Pour les Etudes des Cancers de l'Ovaire (GINECO). Cancer 2009;115:1234–44. [3] Harter P, Muallem ZM, Buhrmann C, Lorenz D, Kaub C, Hils R, et al. Impact of a structured quality management program on surgical outcome in primary advanced ovarian cancer. Gynecol Oncol 2011;121:615–9. [4] du Bois A, Reuss A, Harter P, Pujade-Lauraine E, Ray-Coquard I, Pfisterer J. Potential role of lymphadenectomy in advanced ovarian cancer: a combined exploratory analysis of three prospectively randomized phase III multicenter trials. J Clin Oncol 2010; 28:1733–9. [5] Trimbos B, Timmers P, Pecorelli S, Coens C, Ven K, van der Burg M, et al. Surgical staging and treatment of early ovarian cancer: long-term analysis from a randomized trial. J Natl Cancer Inst 2010;102:982–7. [6] Thigpen T, duBois A, McAlpine J, DiSaia P, Fujiwara K, Hoskins W, et al. First-line therapy in ovarian cancer trials. Int J Gynecol Cancer 2011;21:756–62. [7] Deng J, Sun D, Pan Y, Zhang L, Zhang R, Wang D, et al. Ratio between negative and positive lymph nodes is suitable for evaluation the prognosis of gastric cancer patients with positive node metastasis. PLoS ONE 2012;7:e43925. [8] Dings PJ, Elferink MA, Strobbe LJ, de Wilt JH. The prognostic value of lymph node ratio in node-positive breast cancer: a Dutch nationwide population-based study. Ann Surg Oncol 2013;20:2607–14. [9] Hou X, Wei JC, Xu Y, Luo RZ, Fu JH, Zhang LJ, et al. The positive lymph node ratio predicts long-term survival in patients with operable thoracic esophageal squamous cell carcinoma in China. Ann Surg Oncol 2013;20:1653–9. [10] Xing Y, Badgwell BD, Ross MI, Gershenwald JE, Lee JE, Mansfield PF, et al. Lymph node ratio predicts disease-specific survival in melanoma patients. Cancer 2009; 115:2505–13. [11] Taylor MD, Lapar DJ, Thomas CJ, Persinger M, Stelow EB, Kozower BD, et al. Lymph node ratio predicts recurrence and survival after R0 resection for non-small cell lung cancer. Ann Thorac Surg 2013;96:1163–70. [12] Mahdi H, Thrall M, Kumar S, Hanna R, Seward S, Lockhart D, et al. The prognostic impact of the ratio of positive lymph nodes on survival of epithelial ovarian cancer patients. J Surg Oncol 2011;103:724–9. [13] Therneau TM, Grambsch PM, Fleming TR. Martingale-based residuals for survival models. Biometrika 1990;77:147–60. [14] Aletti GD, Santillan A, Eisenhauer EL, Hu J, Aletti G, Podratz KC, et al. A new frontier for quality of care in gynecologic oncology surgery: multi-institutional assessment of short-term outcomes for ovarian cancer using a risk-adjusted model. Gynecol Oncol 2007;107:99–106. [15] Panici PB, Maggioni A, Hacker N, Landoni F, Ackermann S, Campagnutta E, et al. Systematic aortic and pelvic lymphadenectomy versus resection of bulky nodes only in optimally debulked advanced ovarian cancer: a randomized clinical trial. J Natl Cancer Inst 2005;97:560–6. [16] Prat J. Staging classification for cancer of the ovary, fallopian tube, and peritoneum. Int J Gynaecol Obstet 2014;124:1–5. [17] Pereira A, Irishina N, Perez-Medina T, Magrina JF, Magtibay PM, Kovaleva A, et al. Defining the optimal lymphadenectomy cut-off value in epithelial ovarian cancer staging surgery utilizing a mathematical model of validation. Eur J Surg Oncol 2013;39:290–6. [18] Madbouly KM, Abbas KS, Hussein AM. Metastatic lymph node ratio in stage III rectal carcinoma is a valuable prognostic factor even with less than 12 lymph nodes retrieved: a prospective study. American journal of surgery 2014;207:824–31. [19] Choi BG, Kim CY, Cho SH, Kim HJ, Koh YS, Kim JC, et al. Impact of lymph node ratio as a valuable prognostic factor in gallbladder carcinoma, focusing on stage IIIB gallbladder carcinoma. J Korean Surg Soc 2013;84:168–77. [20] Jeon MJ, Yoon JH, Han JM, Yim JH, Hong SJ, Song DE, et al. The prognostic value of the metastatic lymph node ratio and maximal metastatic tumor size in pathological N1a papillary thyroid carcinoma. Eur J Endocrinol 2013;168:219–25. [21] Ryu IS, Song CI, Choi SH, Roh JL, Nam SY, Kim SY. Lymph node ratio of the central compartment is a significant predictor for locoregional recurrence after prophylactic central neck dissection in patients with thyroid papillary carcinoma. Ann Surg Oncol 2014;21:277–83. [22] du Bois A, Rochon J, Pfisterer J, Hoskins WJ. Variations in institutional infrastructure, physician specialization and experience, and outcome in ovarian cancer: a systematic review. Gynecol Oncol 2009;112:422–36.
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Prognostic value of lymph node ratio in patients with advanced epithelial ovarian cancer Beyhan Ataseven ⁎, Christoph Grimm, Philipp Harter, Sonia Prader, Alexander Traut, Florian Heitz, Andreas du Bois Department of Gynecology and Gynecologic Oncology, Kliniken Essen-Mitte, Evangelische Huyssens-Stiftung, Henricistrasse 92, 45136 Essen, Germany
H I G H L I G H T S • Lymph node status is a prognostic factor in ovarian cancer. • Lymph node ratio reflects lymph node spread and surgical extent. • Lymph node ratio predicts overall survival more concisely.
a r t i c l e
i n f o
Article history: Received 21 August 2014 Accepted 1 October 2014 Available online 12 October 2014 Keywords: Lymph node ratio Ovarian cancer Prognosis Systematic lymphadenectomy
a b s t r a c t Objective. Lymph node status is an established prognostic factor in epithelial ovarian cancer (EOC). Lymph node ratio (number of positive LN/number of resected LN) reflects both qualitative and quantitative lymph node spread as well as surgical effort and extent of disease. We evaluated whether LNR is a more precise prognostic factor than conventional lymph node status in patients with EOC. Methods. The present retrospective study includes 809 patients with EOC, who underwent primary cytoreductive surgery between 2000–2013. Clinico-pathological parameters and survival data were extracted from a prospectively maintained tumor registry database. The optimal cut-off point for LNR was calculated by using Martingale residuals. Survival analyses were calculated using Kaplan–Meier method and Cox regression models. Results. Lymphadenectomy was performed in 693 (85.7%) out of 809 patients. Median number of removed LN was 64 (IQR 25–75%: 39–84). LNR of 0.25 was identified as the optimal prognostic cut-off value. The estimated 5-year-OS rates were 69.3% for patients with node-negative EOC compared to 33.1% for patients with any lymph node metastasis (p b 0.001). The estimated 5-year-OS rates were 42.5% for patients with LNR ≤ 0.25, and 18.0% for patients with LNR N 0.25 (p b 0.001). Additionally in multivariate analysis LNR N 0.25 was approved to be an independent prognostic factor for overall survival (adjusted HR 1.44, 95% CI 1.04–2.00; p = 0.028). Conclusion. LNR more precisely predicts overall survival than conventional lymph node status in EOC patients undergoing primary debulking surgery. © 2014 Elsevier Inc. All rights reserved.
Introduction Epithelial ovarian cancer (EOC) is the most lethal gynecologic malignancy [1]. EOC has a very unique tumor biology and pattern of metastasis with predominantly peritoneal and lymphovascular spread and rarely hematological spread. The role of complete tumor resection of mainly intraperitoneal disease has well been described, however, the prognostic value of lymph node metastasis, its impact on residual
⁎ Corresponding author. Fax: +49 20117434200. E-mail address: [email protected] (B. Ataseven).
http://dx.doi.org/10.1016/j.ygyno.2014.10.003 0090-8258/© 2014 Elsevier Inc. All rights reserved.
tumor status and thus, the therapeutic value of comprehensive pelvic and paraaortic lymphadenectomy have not been fully clarified in EOC. Age, tumor stage, and residual disease are the major prognostic factors in EOC [2,3]. The prognostic value of lymph node metastases is less well defined. However, in clinically apparent stage I disease, the presence of LN metastasis causes an upstaging possibly leading to a subsequent change in treatment. Studies addressing the therapeutic value of comprehensive pelvic and paraaortic lymphadenectomy in patients with EOC are limited. Retrospective analyses indicate a benefit particularly in optimally debulked ovarian cancer cases and early stage EOC [4,5]. In advanced EOC this benefit has not been consistently shown. Nonetheless,
436
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
complete gross resection of all visible disease after surgery is the surgical goal in treating EOC [6]. The therapeutic role of comprehensive lymphadenectomy in EOC patients is currently being addressed in a prospective randomized controlled trial (lymphadenectomy in ovarian neoplasms (LION) NCT00712218). Lymph node ratio (LNR) is the number of positive lymph nodes divided by the total number of removed lymph nodes. LNR has been proven to be an important prognostic factor in a variety of malignancies including breast, gastric, esophageal, and non-small cell lung cancer, and melanoma [7–11]. With respect to EOC, one retrospective analysis has reported that high LNR was independently associated with decreased overall survival in patients with advanced EOC [12]. Patients from the Surveillance, Epidemiology, and End Results Program (SEER) were included in this study. A major limitation of this retrospective analysis was the extent of lymphadenectomy. Only 58.0% and 21.1% of patients had ≤10 and 11–20 lymph nodes removed, respectively. Just a small amount of 20.8% of patients had a comprehensive lymphadenectomy with N20 lymph nodes removed. The aim of our present study was to evaluate the additional prognostic value of LNR compared to conventional lymph node status in patients with EOC that have undergone systematic pelvic and paraaortic lymphadenectomy during primary upfront debulking surgery. Methods Patients with first diagnosis of invasive epithelial ovarian, fallopian tube or serous primary peritoneal cancer who attended our tertiary gynecologic oncology center (Horst–Schmidt–Kliniken (HSK), Wiesbaden and Kliniken Essen-Mitte (KEM, Essen)) were prospectively documented in a clinical tumor registry from 2000 until present. All patients' follow-up data are updated annually. In the present study, we included only patients with EOC (n = 906; HSK n = 630; KEM n = 276) who underwent primary upfront cyctoreductive surgery in our institution. Patients receiving neo-adjuvant chemotherapy and subsequently attending our center for interval debulking (n = 97) were excluded. Furthermore, patients with borderline tumors and non-epithelial ovarian cancer were also not included in this analysis. All patients gave informed consent for the surgical procedure and documentation of clinical data in our clinical tumor registry. Surgery was performed by a dedicated surgical team. In all patients with macroscopic residual disease routine intra-operative 2nd opinion was obtained from a senior gynecological oncologist before wound closure. The indication for systematic pelvic and para-aortic lymphadenectomy changed over the study period. Between 2000–2006 systematic lymphadenectomy was performed in patients with age ≤75-years and surgical debulking without residual tumor intraabdominally or tumor residuals up to 10 mm. After 2006 systematic lymphadenectomy was performed only in patients with complete macroscopic tumor resection, whereas in patients with 1–10 mm residual tumor only bulky nodes were resected. For this retrospective analysis we defined systematic lymphadenectomy as a complete procedure when at least 30 lymph nodes were reported in the pathologic specimen. Platinum-based chemotherapy was given stage adapted after surgery to all patients. To evaluate the optimal cut-off point for LNR (number of positive LN/total resected LN) Martingale residuals [13] was used. Node positive patients were categorized either in low LNR group (≤0.25) or high LNR group (N0.25). For comparison of the total cohort we divided patients into four LN groups: NX (lymphadenectomy not performed), N0 (lymphadenectomy performed, negative for metastasis), LNR ≤ 0.25 (lymphadenectomy performed, positive for metastasis; LNR ≤ 0.25) and LNR N 0.25 (lymphadenectomy performed, positive for metastasis, LNR N 0.25). To further elucidate the prognostic value of LNR, we additionally performed analysis in a more restrictive patient cohort (complete gross resection — complete LN dissection (CGR-LND group)). Therefore we evaluated the impact of LNR only in patients with tumor stage pT3c,
optimal debulking surgery (defined as complete resection) and systematic lymphadenectomy (number of resected LN N 30). This cohort is henceforward referred to as the “CGR-LND group”. To eliminate bias caused by a different extent of lymphadenectomy in the low and high LNR groups, we evaluated the number of total lymph nodes removed. Moreover, we compared both groups with respect to length of operation, blood loss, and quantification of surgical effort using the Surgical Complexity Score (SCS) previously described [14]. Statistical analysis was performed using SPSS™ version 20.0 (Chicago, Illinois, USA) and GraphPad Prism® 4 (La Jolla, CA, USA). Cut-off value for LNR was calculated by Martingale residuals. We performed multivariate Cox regression analyses comprising tumor stage, distant metastases, tumor grade, histologic type, patient's age, performance status, and surgical outcome/residual tumor to determine whether the LNR had any significant impact on overall survival (OS). Univariate survival analyses were provided as Kaplan–Meier curves and differences calculated by log-rank tests. Association between LNR and OS was calculated by using a Cox proportional hazard regression model. Results 809 patients met inclusion criteria for this analysis. Patient and tumor characteristics are shown in Table 1. Two hundred seventynine (34.5%) patients were 65 years or older. The majority of patients (n = 686; 84.8%) were in a good performance status (Eastern Cooperative Oncology group; ECOG 0), diagnosed with pT3-stage (n = 581;
Table 1 Patient and tumor characteristics. Variable
Total cohort
Sensitivity-analysis cohort #
Patients (n)
%
Patients (n)
%
Total Age⁎
809
100
210
100
≤64 yr N64 yr Performance status ECOG 0 ECOG N0 FIGO-stage FIGO I FIGO II FIGO III FIGO IV (=M1) Grade Low-grade High-grade Histology Serous Others pN-stage NX N0 N1 Lymph node resected 0 ≤30 N30 Lymph node ratio (LNR)‡⁎ NX N0 ≤0.25 N0.25 Residual disease after surgery 0 mm 1–10 mm N10 mm
530 279
65.5 34.5
155 55
73.8 26.2
686 123
84.8 15.2
200 10
95.2 4.8
133 50 404 222
16.4 6.2 49.9 27.4
0 0 132 78
62.9 37.1
57 752
7.0 93.0
4 206
1.9 98.1
589 220
72.8 27.2
189 21
90.0 10.0
116 295 398
14.3 36.5 49.2
0 45 165
0 21.4 78.6
116 152 541
14.3 18.8 66.9
0 0 210
116 295 244 154
14.3 36.5 30.2 19.0
0 45 123 42
573 170 66
70.8 21.0 8.2
210 0 0
0 0 100 0 21.4 58.6 20.0 100 0 0
⁎ Cut by martingale residue. ‡ LNR is defined as the number of positive nodes divided by the number of resected nodes. # Sensitivity-analysis cohort (stage pT3c, residual disease 0 mm, systematic lymphadenectomy).
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
71.8%), had a high-grade carcinoma (n = 752; 93.0%) and a serous histology (n = 589; 72.8%). Any kind of lymphadenectomy was performed in 693 (85.7%) patients (median number of removed LN 65 (IQR 25– 75%: 39–84)). In 152 (18.8%) patients 1–30 nodes were dissected with a median number of 7 dissected nodes (IQR 25–75%: 3–15). In 541 (66.9%) patients more than 30 nodes were dissected (systematic pelvic and para-aortic lymphadenectomy) with a median number of 73 dissected nodes (IQR 25–75%: 59–89). Lymph node metastases were evident in 389 of 693 (56.1%) patients. Complete gross cytoreduction (defined as no macroscopic tumor) was achieved in 573 (70.8%), cytoreduction with gross residual disease of 1–10 mm in 170 (21.0%) and gross residual disease N 10 mm in 66 (8.2%) patients. According to the LN groups 116 (14.3%) patients were categorized as NX, 295 (36.5%) as N0, 244 (30.2%) as LNR ≤ 0.25 and 154 (19.0%) as LNR N 0.25. Patient and tumor characteristics for the more restrictive sensitivityanalysis group differed slightly from the total cohort in terms of included patients (n = 210), age (N 64 years: 26.2%), performance status (ECOG 0: 95.2%), tumor grade (high grade: 98.1%), histology (serous: 90.0%) and LN group (N0: 21.4%; LNR ≤ 0.25: 58.6%; LNR N 0.25: 20.0%). After a median follow up of 45 months (1–173) 332 (41.0%) patients had died. Estimated 5-year OS rate for the entire cohort was 46.5%. Based on conventional nodal status, patients with no evidence for LN metastasis (N0) showed more favorable 5-year OS (69.3%) compared to patients without lymphadenectomy (32.0%) and patients with positive LN (33.1%) (Fig. 1; p = b0.001). After stratification of lymph node positive patients according to LNR, 5-year OS rates were 42.5% for patients with LNR ≤ 0.25 and 18.0% for patients with LNR N 0.25 (Fig. 2; p b 0.001). Further prognostic factors in univariate analysis were patient's age, performance status, tumor stage, distant metastases, FIGO stage, tumor grade, histologic type, nodal status, number of resected lymph nodes, and residual tumor (Table 2). In multivariate Cox regression analysis, LNR was an independent prognostic factor for OS after adjusting for other clinically relevant factors in univariate analysis (Table 2). Patients with LNR N 0.25 had a significantly worse prognosis HR 1.44 (95% CI 1.04–2.00; p = 0.028). Moreover, we performed sensitivity analysis for LNR and evaluated its prognostic impact in the CGR-LND group. Only patients with systematic lymphadenectomy, tumor stage pT3c, and without any residual tumor were included. Again LNR N 0.25 remained an independent prognostic factor for OS with a hazard ratio of 2.10 (95% CI 1.06–4.16; p = 0.034; Table 3). After stratification for LNR, 5-year OS rates in this group were 33.0% for patients with LNR ≤ 0.25 and 22.0% for patients with LNR N 0.25 (p = 0.178; Fig. 3), respectively. To rule out a potential
437
bias regarding surgical efforts, we compared the number of resected lymphnodes in subgroups with high or low LNR. We could not find a significant difference between the LNR ≤ 0.25 and LNR N 0.25 cohort regarding mean number of resected lymph nodes (n = 81 versus 83; p = 0.743), mean blood loss (ml = 1065 versus 1247; p = 0.102), and mean length of operation (minutes = 446 versus 486; p = 0.133), respectively. In terms of extent and complexity of surgery no significant difference was detected between both groups as 98.4% of patients in the LNR ≤ 0.25 and 97.6% in the LNR N 0.25 group were categorized in high Surgical Complexity Score group (Supplementary data Table 2). Discussion In the present study, we show that the lymph-node-ratio (LNR) could serve as an additional prognostic parameter to better predict outcome in node positive patients with epithelial ovarian cancer (EOC). Overall, nodal positive patients had an impaired prognosis compared to node negative EOC patients. Furthermore, nodal positive patients with a higher LNR had an even worse prognosis. A LNR cut-off value of 0.25 was identified to maximally discriminate prognosis. The estimated 5-year-OS rates were 69.3% for patients with node-negative cancer compared to 33.1% for patients with any lymph node metastasis. Stratified by LNR the estimated 5-year-OS rates were 42.5% for patients with LNR ≤ 0.25 compared to 18.0% for patients with LNR N 0.25. Lymph node status is a well-established prognostic parameter in patients with EOC [4,15] and this is also reflected in the new FIGO classification, where lymph node status and size of lymph node metastases define FIGO tumor stage [16]. LNR has the potential to add another dimension besides positivity, location and size of LN metastasis to the prognostic value of nodal spread in EOC. The potential value of LNR may be based on its composition of both tumor biology and extent of surgical effort. The latter may also be one of the theoretical weaknesses of this parameter. In the present study, LNR provided additional information compared to lymph node status alone. Patients with any lymph node metastasis presented with a median overall survival of 44 months. If stratified by LNR patients with LNR ≤ 0.25 compared to patients with LNR N 0.25 presented with a median overall survival of 52 and 30 months, respectively. However, all patients were operated rather homogeneously in a high volume center with high rates of complete resection and high numbers of resected nodes. Further studies in other institutions and multicenter trials may elucidate the generalizability of our observations. Nevertheless, we tried to challenge our first
Fig. 1. Overall survival by lymph node status. Kaplan–Meier.
438
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
Fig. 2. Overall survival by lymph node ratio (LNR). Kaplan–Meier.
Table 2 Univariate and multivariate analyses of prognostic factors for overall survival (OS) (total cohort; n = 809). Variable
Patients (n)
%
Univariate OS event n (%)
Total Age⁎ ≤64 yr N64 yr Performance status ECOG 0 ECOG N0 Secondary malignancy Yes No Tumor size T1 T2 T3 M-stage M0 M1 (FIGO IV) Grade Low-grade High-grade Histology Others Serous pN-stage NX N0 N1 Lymph node resected 0 ≤30 N30 Lymph node ratio (LNR)‡⁎ N0 ≤0.25 N0.25 NX Residual disease after surgery 0 mm 1–10 mm N10 mm
809
Multivariate p-Value
HR
95% CI
p-Value
332 (41.0)
530 279
65.5 34.5
191 (36) 141 (50.5)
b0.001
1 1.34
1.06–1.70
0.016
686 123
84.8 15.2
244 (35.6) 88 (71.5)
b0.001
1 1.38
1.04–1.80
0.023
124 685
15.3 84.7
46 (37.1) 286 (41.8)
0.160
153 75 581
18.9 9.3 71.8
14 (9.2) 15 (20) 303 (52.2)
b0.001
1 2.47 7.36
1.16–5.26 4.04–13.4
b0.001 0.019 b0.001
587 222
72.6 27.4
214 (36.5) 118 (53.2)
b0.001
1 1.30
1.03–1.66
0.030
57 752
7.0 93.0
6 (10.5) 326 (43.4)
b0.001
1 2.19
0.97–4.95
220 589
27.2 72.8
64 (29.1) 268 (45.5)
b0.001
1 0.85
0.64–1.13
0.255
116 295 398
14.3 36.5 49.2
67 (57.8) 79 (26.8) 186 (46.7)
b0.001
116 152 541
14.3 18.8 66.9
67 (57.8) 78 (51.3) 187 (34.6)
b0.001
295 244 154 116
36.5 30.2 19.0 14.3
79 (26.8) 98 (40.2) 88 (57.1) 67 (57.8)
1 1.30 1.44 1.25
0.95–1.77 1.04–2.00 0.87–1.81
0.157 0.096 0.028 0.230
573 170 66
70.8 21.0 8.2
168 (29.3) 114 (67.1) 50 (75.8)
1 1.84 3.07
1.40–2.43 2.07–4.54
b0.001 b0.001 b0.001
⁎ Cut by martingale residue. ‡ LNR is defined as the number of positive nodes divided by the number of resected nodes.
b0.001
b0.001
0060
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
439
Table 3 Univariate and multivariate analysis of prognostic factors for overall survival (OS) (Complete gross resection — complete LN dissection; CGR-LND group; n = 210). Variable
Patients (n)
%
Univariate OS event n (%)
Total events PFS/OS Age⁎ ≤64 yr N64 yr Performance status ECOG 0 ECOG N0 M-stage M0 M1 (FIGO IV) Grade Low-grade High-grade Histology Others Serous pN-stage N0 N1 Lymph node ratio (LNR)‡⁎ N0 ≤0.25 N0.25 Surgical complexity scoring (median split) ≤13 N13
210
Multivariate p-Value
HR
95% CI
p-Value
88 (41.9)
155 55
73.8 26.2
58 (37.4) 30 (54.5)
0.058
1 1.46
0.90–2.37
0.122
200 10
95.2 4.8
80 (40) 8 (80)
0.021
1 1.87
0.83–4.20
0.129
132 78
62.9 37.1
55 (41.7) 33 (42.3)
0.019
1 1.17
0.73–1.83
0.525
4 206
1.9 98.1
1 (25) 87 (42.2)
0.461
1 1.68
0.23–12.4
0.612
21 189
10 90
8 (38.1) 80 (42.3)
0.734
1 1.27
0.60–2.71
0.655
45 165
21.4 78.6
45 (21.4) 68 (41.2)
0.071
45 123 42
21.4 58.6 20
20 (44.4) 48 (39) 20 (47.6)
0.050
1 1.45 2.10
0.82–2.56 1.06–4.16
0.105 0.202 0.034
142 68
67.6 32.4
63(44.4) 25(36.8)
0.135
1 1.48
0.90–2.44
0.125
⁎ Cut by martingale residue. ‡ LNR is defined as the number of positive nodes divided by the number of resected nodes.
results and performed a sensitivity analysis in a more restricted CGRLND group. This CGR-LND group consisted of EOC patients with tumor stage pT3c, optimal debulking surgery, and no residual tumor. Even in this restrictive analysis LNR remained an independent prognostic parameter for OS in uni- and multivariate analyses. Mahdi et al. [12] had also evaluated the prognostic impact of LNR in 6310 patients with advanced ovarian cancer (FIGO IIIC and IV) with and without macroscopic peritoneal disease obtained from SEER database. Patients were stratified according to the LNR in three groups (LNR b 10%, LNR 10–50% and LNR N 50%). LNR was an independent prognostic factor for survival, particularly in patients with no peritoneal
macroscopic disease. However, the SEER analysis differed from our study: The authors defined comprehensive lymphadenectomy as dissection of N10 lymph nodes. Fifty-eight percent of patients had a lymphadenectomy with less than 10 LNs and only 21% of the patients had N20 LN dissected. This may rather describe lymph node sampling in a considerable portion of patients and therefore LNR ratios may differ in both cohorts. The role of systematic lymphadenectomy has been mentioned by Pereira et al. [17] who concluded that the probability of obtaining a positive LN could only be predicted if at least 22 LNs (15 pelvic and 7 aortic) were dissected. In our study we defined systematic lymphadenectomy if N30 LNs were reported. This is based on the
Fig. 3. Sensitivity analysis with estimated overall survival for lymph node ratio (LNR). Other covariates are fixed at their average values.
440
B. Ataseven et al. / Gynecologic Oncology 135 (2014) 435–440
currently ongoing LION trial, in which systematic lymphadenetomy is defined as comprising at least 20 pelvic lymph nodes and 10 paraaortic lymph nodes. Future studies on LNR should include a precise definition of the surgical scenario — in particular the systematic approach and extent of lymphadenectomy. This is crucial, as diverging surgical effort can impair the informative value when the prognostic value of LNR is investigated. Our selection criteria of the sensitivity-analysis-cohort and the definition of the LNR cut-off value may be used in further studies to obtain a more standardized model that can be used for comparison and validation of our findings. The interest in LNR as prognosticator in EOC is supported by other analyses which showed that LNR was an important prognostic factor in gastric cancer, breast cancer, esophageal cancer, melanoma, thyroid papillary carcinoma, gallbladder carcinoma and non-small cell lung cancer [7–11,18–21]. The present study has some limitations which underscore that this is a hypothesis building analysis and far from any definitive and conclusive statement. As already mentioned, this is a retrospective single-team study bearing potential risk for selection bias and a bias with respect to our surgical school which may not be comparable to ovarian cancer surgery worldwide [3,22]. Nevertheless, this potential bias could also be regarded as strengths of this series: all surgeries had been performed by high-volume surgeons at a high-volume center with an expertise in gynecologic oncology. Due to the structured quality management program on surgical training and the 2nd opinion procedure, as described before [3], we were able to compare the operative procedures and their impact in a very standardized way and may rule out a bias regarding the extent of lymphadenectomy performed from case to case and by a heterogeneous cohort of surgeons. Nonetheless, the statistical calculation of cut-off values by Martingale residuals is a well-established statistical approach to identify the optimal cut-off value. At any case, a validation of our result with data from another high-volume institution is necessary to support the power of our study and define more precisely the impact of LNR on the prognosis of EOC patients. If confirmed, LNR could be an easily assessable parameter in operated EOC, which allows a favorable and further prognostic discrimination of node-positive patients. Conflict of interest The authors declare no conflict of interest.
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ygyno.2014.10.003. References [1] Jemal A, Simard EP, Dorell C, Noone AM, Markowitz LE, Kohler B, et al. Annual Report to the Nation on the Status of Cancer, 1975–2009, featuring the burden and trends in human papillomavirus (HPV)-associated cancers and HPV vaccination coverage levels. J Natl Cancer Inst 2013;105:175–201.
[2] du Bois A, Reuss A, Pujade-Lauraine E, Harter P, Ray-Coquard I, Pfisterer J. Role of surgical outcome as prognostic factor in advanced epithelial ovarian cancer: a combined exploratory analysis of 3 prospectively randomized phase 3 multicenter trials: by the Arbeitsgemeinschaft Gynaekologische Onkologie Studiengruppe Ovarialkarzinom (AGO-OVAR) and the Groupe d'Investigateurs Nationaux Pour les Etudes des Cancers de l'Ovaire (GINECO). Cancer 2009;115:1234–44. [3] Harter P, Muallem ZM, Buhrmann C, Lorenz D, Kaub C, Hils R, et al. Impact of a structured quality management program on surgical outcome in primary advanced ovarian cancer. Gynecol Oncol 2011;121:615–9. [4] du Bois A, Reuss A, Harter P, Pujade-Lauraine E, Ray-Coquard I, Pfisterer J. Potential role of lymphadenectomy in advanced ovarian cancer: a combined exploratory analysis of three prospectively randomized phase III multicenter trials. J Clin Oncol 2010; 28:1733–9. [5] Trimbos B, Timmers P, Pecorelli S, Coens C, Ven K, van der Burg M, et al. Surgical staging and treatment of early ovarian cancer: long-term analysis from a randomized trial. J Natl Cancer Inst 2010;102:982–7. [6] Thigpen T, duBois A, McAlpine J, DiSaia P, Fujiwara K, Hoskins W, et al. First-line therapy in ovarian cancer trials. Int J Gynecol Cancer 2011;21:756–62. [7] Deng J, Sun D, Pan Y, Zhang L, Zhang R, Wang D, et al. Ratio between negative and positive lymph nodes is suitable for evaluation the prognosis of gastric cancer patients with positive node metastasis. PLoS ONE 2012;7:e43925. [8] Dings PJ, Elferink MA, Strobbe LJ, de Wilt JH. The prognostic value of lymph node ratio in node-positive breast cancer: a Dutch nationwide population-based study. Ann Surg Oncol 2013;20:2607–14. [9] Hou X, Wei JC, Xu Y, Luo RZ, Fu JH, Zhang LJ, et al. The positive lymph node ratio predicts long-term survival in patients with operable thoracic esophageal squamous cell carcinoma in China. Ann Surg Oncol 2013;20:1653–9. [10] Xing Y, Badgwell BD, Ross MI, Gershenwald JE, Lee JE, Mansfield PF, et al. Lymph node ratio predicts disease-specific survival in melanoma patients. Cancer 2009; 115:2505–13. [11] Taylor MD, Lapar DJ, Thomas CJ, Persinger M, Stelow EB, Kozower BD, et al. Lymph node ratio predicts recurrence and survival after R0 resection for non-small cell lung cancer. Ann Thorac Surg 2013;96:1163–70. [12] Mahdi H, Thrall M, Kumar S, Hanna R, Seward S, Lockhart D, et al. The prognostic impact of the ratio of positive lymph nodes on survival of epithelial ovarian cancer patients. J Surg Oncol 2011;103:724–9. [13] Therneau TM, Grambsch PM, Fleming TR. Martingale-based residuals for survival models. Biometrika 1990;77:147–60. [14] Aletti GD, Santillan A, Eisenhauer EL, Hu J, Aletti G, Podratz KC, et al. A new frontier for quality of care in gynecologic oncology surgery: multi-institutional assessment of short-term outcomes for ovarian cancer using a risk-adjusted model. Gynecol Oncol 2007;107:99–106. [15] Panici PB, Maggioni A, Hacker N, Landoni F, Ackermann S, Campagnutta E, et al. Systematic aortic and pelvic lymphadenectomy versus resection of bulky nodes only in optimally debulked advanced ovarian cancer: a randomized clinical trial. J Natl Cancer Inst 2005;97:560–6. [16] Prat J. Staging classification for cancer of the ovary, fallopian tube, and peritoneum. Int J Gynaecol Obstet 2014;124:1–5. [17] Pereira A, Irishina N, Perez-Medina T, Magrina JF, Magtibay PM, Kovaleva A, et al. Defining the optimal lymphadenectomy cut-off value in epithelial ovarian cancer staging surgery utilizing a mathematical model of validation. Eur J Surg Oncol 2013;39:290–6. [18] Madbouly KM, Abbas KS, Hussein AM. Metastatic lymph node ratio in stage III rectal carcinoma is a valuable prognostic factor even with less than 12 lymph nodes retrieved: a prospective study. American journal of surgery 2014;207:824–31. [19] Choi BG, Kim CY, Cho SH, Kim HJ, Koh YS, Kim JC, et al. Impact of lymph node ratio as a valuable prognostic factor in gallbladder carcinoma, focusing on stage IIIB gallbladder carcinoma. J Korean Surg Soc 2013;84:168–77. [20] Jeon MJ, Yoon JH, Han JM, Yim JH, Hong SJ, Song DE, et al. The prognostic value of the metastatic lymph node ratio and maximal metastatic tumor size in pathological N1a papillary thyroid carcinoma. Eur J Endocrinol 2013;168:219–25. [21] Ryu IS, Song CI, Choi SH, Roh JL, Nam SY, Kim SY. Lymph node ratio of the central compartment is a significant predictor for locoregional recurrence after prophylactic central neck dissection in patients with thyroid papillary carcinoma. Ann Surg Oncol 2014;21:277–83. [22] du Bois A, Rochon J, Pfisterer J, Hoskins WJ. Variations in institutional infrastructure, physician specialization and experience, and outcome in ovarian cancer: a systematic review. Gynecol Oncol 2009;112:422–36.
