Abstract

Background: The role of systematic aortic and pelvic lymphadenectomy in patients with optimally debulked advanced ovarian cancer is unclear and has not been addressed by randomized studies. We conducted a randomized clinical trial to determine whether systematic aortic and pelvic lymphadenectomy improves progression-free and overall survival compared with resection of bulky nodes only. Methods: From January 1991 through May 2003, 427 eligible patients with International Federation of Gynecology and Obstetrics (FIGO) stage IIIB-C and IV epithelial ovarian carcinoma were randomly assigned to undergo systematic pelvic and para-aortic lymphadenectomy (n = 216) or resection of bulky nodes only (n = 211). Progression-free survival and overall survival were analyzed using a log-rank statistic and a Cox multivariable regression analysis. All statistical tests were two-sided. Results: After a median follow-up of 68.4 months, 292 events (i.e., recurrences or deaths) were observed, and 202 patients had died. Sites of first recurrences were similar in both arms. The adjusted risk for first event was statistically significantly lower in the systematic lymphadenectomy arm (hazard ratio [HR] = .75, 95% confidence interval [CI] = 0.59 to 0.94; P = .01) than in the no-lymphadenectomy arm, corresponding to 5-year progression-free survival rates of 31.2 and 21.6% in the systematic lymphadenectomy and control arms, respectively (difference = 9.6%, 95% CI = 1.5% to 21.6%), and to median progression-free survival of 29.4 and 22.4 months, respectively (difference = 7 months, 95% CI = 1.0 to 14.4 months). The risk of death was similar in both arms (HR = 0.97, 95% CI = 0.74 to 1.29; P = .85), corresponding to 5-year overall survival rates of 48.5 and 47%, respectively (difference = 1.5%, 95% CI = −8.4% to 10.6%), and to median overall survival of 58.7 and 56.3 months, respectively (difference = 2.4 months, 95% CI = −11.8 to 21.0 months). Median operating time was longer, and the percentage of patients requiring blood transfusions was higher in the systematic lymphadenectomy arm than in the no-lymphadenectomy arm (300 versus 210 minutes, P <.001, and 72% versus 59%; P = .006, respectively). Conclusion: Systematic lymphadenectomy improves progression-free but not overall survival in women with optimally debulked advanced ovarian carcinoma.

Epithelial ovarian cancer is a clinically significant health problem in Western countries, ranking fifth highest in cancer incidence and fourth highest in site-specific causes of cancer deaths in women. In the United States, approximately 25 000 new ovarian cancer cases and 16 000 ovarian cancer deaths were expected in 2004. Although ovarian cancer is potentially curable by surgery and chemotherapy, most cases are still diagnosed at advanced stages, and the 5-year overall survival (OS) of the patients diagnosed lingers at approximately 30% ( 1 ) .

Primary cytoreductive surgery (i.e., the removal of as much of the tumor as possible at the time of initial surgery, with resection of bulky nodes only) has been an integral part of the treatment of advanced ovarian cancer since the observation that postoperative residual tumor is a clinically significant prognostic factor ( 2 , 3 ) . A recent meta-analysis ( 4 ) has confirmed that maximal surgical cytoreduction is one of the most powerful determinants of cohort survival in International Federation of Gynecology and Obstetrics (FIGO) stage III-IV ovarian cancer.

Whether systematic removal of retroperitoneal lymph nodes should be part of maximal cytoreductive surgery is still unclear. Retroperitoneal lymph node involvement occurs in approximately 50% to 80% of women with advanced ovarian cancer ( 5 ) . In 1988, in recognition of the prognostic importance of retroperitoneal spread, the FIGO staging classification was amended to include a substage for node involvement, ( 6 ) . Subsequent work ( 79 ) has illuminated the relevant surgical anatomy, which has allowed for identification of the role and technical aspects of lymph node dissection and clarification of the nomenclature.

Although systematic lymphadenectomy has been shown to be feasible and safe in the treatment of advanced ovarian cancer, its therapeutic role remains controversial. The core issue of the controversy is whether removal of the lymph nodes should be performed only to stage the disease or whether removal improves survival. In the former case, lymph node regions at risk would only be sampled, whereas in the latter case there would be a systematic effort to remove all accessible lymph node-bearing tissue. Retrospective studies ( 10 , 11 ) have suggested a clinically significant survival advantage following systematic lymphadenectomy in patients undergoing cytoreductive surgery for advanced disease; however, no prospective randomized clinical trial has been published. Therefore, in this study we conducted a multicenter, randomized clinical trial at 13 different sites in five countries to determine whether the addition of systematic aortic and pelvic lymphadenectomy to optimal cytoreduction surgery within the peritoneal cavity improves progression-free survival (PFS) and overall survival in patients with advanced ovarian cancer.

P ATIENTS AND M ETHODS

Patient Eligibility

Patients with histologically proven and optimally debulked (i.e., residual tumor of ≤1 cm) epithelial ovarian carcinoma with FIGO stages IIIB and IIIC were eligible for participation in the study. Stage IV patients were eligible if the only evidence of stage IV disease was malignant cells in pleural effusion. Additional eligibility criteria included age of less than 75 years, Karnofsky performance status of ≥80, and no previous chemotherapy or radiation therapy. The study protocol was revised and accepted by local ethics committees, and appropriate written informed consent was obtained from all patients.

Randomization Procedures

Random assignment of patients, who had an equal probability of assignment to either treatment arm, was carried out by a block arrangement that balances the treatment assignment within each site. Randomization was performed centrally by telephone at seven sites (Monza, Milan [three sites], Aviano, Treviglio, and Padua) and with sealed envelopes at six sites (Rome, Graz, Erlangen, Brescia, Sydney, and Gateshead). Randomization codes were generated at the Mario Negri Institute, Milan. Patients were randomly assigned intraoperatively (i.e., the patient's envelope, which was identified by a registration number, was opened or a telephone call was made) at the end of primary cytoreductive surgery, only after the surgeon was sure the tumor was optimally debulked. Data from all eligible patients were analyzed for survival on an intention-to-treat basis.

Patient Characteristics

Pretreatment clinical and tumor characteristics and operative details were collected soon after surgery. Chemotherapy and initial follow-up data were collected 6 months after surgery, and further follow-up data were collected annually thereafter. Data were sent either to the Mario Negri Institute, Milan, or to the Campus Biomedico, Rome.

Surgical Procedures

Control arm.

Primary cytoreductive surgery was aimed at removing the primary tumor and any metastatic implants, including total abdominal hysterectomy, bilateral salpingo-oophorectomy, radical omentectomy, appendectomy, and the attempted removal of all macroscopic intra-abdominal tumor (including all lymph nodes of ≥1 cm in diameter).

Systematic lymphadenectomy arm.

Primary cytoreductive surgery as detailed for the control arm was followed by systematic pelvic and aortic lymphadenectomy. Pelvic lymphadenectomy dissection began at the origin of the external iliac vessels and continued caudally along the medial border of the psoas muscle, with the lower limit of the external iliac lymphadenectomy being represented by the deep inferior epigastric vessels. The lateral boundaries of the lymphadenectomy were delineated superficially by the fascia covering the psoas muscle and deeply by the fascia covering the internal obturator and levator ani muscles. The medial margin of the lymphadenectomy was represented by an imaginary plane parallel to the umbilical artery, the umbilico-pubic fascia, the bladder, and the rectum. The clearing of the obturator fossa began with mobilization of the superficial obturator nodes, which were removed en bloc (i.e., all together) with the lymphatic fatty tissue that had been previously separated from the internal iliac vessels to the origin of the internal pudendal vessels. Lymphadenectomy continued with the dissection of the deep obturator nodes and gluteal nodes. Pelvic lymphadenectomy was considered appropriate when at least 25 nodes were removed.

Aortic lymphadenectomy dissection began at the aortic bifurcation by removing the superficial intercavoaortic, precaval, and preaortic nodal groups. Lymph nodes located lateral to the vena cava (i.e., paracaval nodal group) were separated from the vena cava, and the renal capsule and psoas muscle were removed en bloc. Lymph nodes behind the vena cava (i.e., retrocaval nodal group) and the lumbar vessels (i.e., deep intercavo-aortic nodal group) were separated and removed from the prevertebral fascia after displacing the vena cava and the aorta laterally and medially. Removal of the cranial nodes, both behind and under the left renal vein, was performed after entering the plane between the Toldt's and Gerota's fasciae; mobilizing the descending colon from the renal capsule, the psoas muscle, and the ovarian pedicle; and displacing the ureter laterally. Aortic lymphadenectomy was considered appropriate when at least 15 nodes were removed.

Adjuvant Chemotherapy

After primary cytoreductive surgery, all patients, regardless of the trial arm, were recommended to have adjuvant treatment with platinum-based chemotherapy. Patients found to have progressive disease were provided with further treatment at the discretion of the treating physician–individual investigator.

Statistical Analysis

The primary endpoint of this study was overall survival, defined as the time from randomization to death from any cause. Secondary endpoints included progression-free survival, defined as the time from randomization to the earliest occurrence of progression or death from any cause, and surgical morbidity. In pretrial analysis, we calculated that, with a type I error limited to .05 (two-tailed test), a total of 440 patients would need to be accrued to afford 80% power to detect a 30% relative reduction in the mortality hazard rate of patients in the systematic lymphadenectomy arm (i.e., an increase in the 5-year survival rate from 45% to 57%, which corresponds to a hazard ratio [HR] of 0.7).

Survival curves were estimated by the Kaplan–Meier method ( 12 ) and compared using the log-rank test ( 13 ) . We used Kaplan–Meier estimates of overall and progression-free survival in the no-lymphadenectomy arm at specific time points and the hazard ratio to calculate absolute benefits at those time points according to the following formula: absolute benefit = exp {HR × log[control survival]} − control survival. Although this approach implicitly assumes proportional hazards, it is preferable to comparing differences between Kaplan–Meier survival curves of two groups at individual time points. In fact, the entire survival curve is a good summary of the survival experience of two groups, whereas individual points on the survival curves are often unreliable ( 14 ) . Differences in median progression-free and overall survival times between the two groups were calculated in a similar way, except that we used the following formula: difference in medians = (control group median/HR) − control group median. This approach assumes approximately exponentially distributed survival curves. Because absolute benefits at different time points or medians for PFS or OS were estimated using the same hazard ratio and its 95% confidence intervals (CIs), the P values of such estimates are the same as for the hazard ratios. Additional analyses were performed with the Cox proportional-hazards model adjusting for multiple baseline characteristics. Proportional hazards assumptions were checked by plotting log{−log [S(t)]} against log t for each group and found to be satisfied ( 14 ) .

Per trial protocol, progression-free survival and overall survival were also analyzed in only those patients who underwent appropriate surgical procedures (see Fig. 1 ). In this per-protocol analysis, survival analysis was performed on 195 control patients and 189 systematic lymphadenectomy patients. Comparison of proportions between groups was performed using a two-sided chi-square (χ 2 ) test or, if the number of patients in a given category was less than five, a two-sided Fisher's exact test. Continuous data, including number of resected nodes, operating time, blood loss, and hospital stay, were expressed as medians with interquartile ranges and were compared using a two-sided Kruskal–Wallis (nonparametric) test.

Fig. 1.

CONSORT trial flow diagram for patients with stage III-IV ovarian cancer who were accrued into the trial. *Protocol violations: no-lymphadenectomy (control) arm = more than 25 pelvic or 15 lumbo-aortic nodes resected; systematic lymphadenectomy arm = less than 25 pelvic and 15 lumbo-aortic nodes resected.

Fig. 1.

CONSORT trial flow diagram for patients with stage III-IV ovarian cancer who were accrued into the trial. *Protocol violations: no-lymphadenectomy (control) arm = more than 25 pelvic or 15 lumbo-aortic nodes resected; systematic lymphadenectomy arm = less than 25 pelvic and 15 lumbo-aortic nodes resected.

R ESULTS

Patient Accrual

Between January 1991 and May 2003, 452 patients were enrolled at 13 centers in five countries (Italy, 379 patients; Austria, 41 patients; Germany, 22 patients; Australia, 8 patients; and United Kingdom, 2 patients). After pathologic and clinical review, 25 patients were deemed ineligible. Figure 1 shows the trial flow diagram and details reasons for patient ineligibility.

Patient Characteristics

The clinical and tumor characteristics of eligible patients are listed in Table 1 . Characteristics appear to be well balanced across treatment arms. Approximately 75% of patients (i.e., 328) had FIGO stage IIIC disease, and 96% of patients had residual disease of ≤1 cm. Data on residual tumor were missing for four patients, and the remaining 16 patients had residual tumors of >1 cm but <2 cm.

Table 1.

Clinical and tumor characteristics by treatment arm

  No lymphadenectomy (n = 211)
 
  Lymphadenectomy (n = 216)
 
 
Characteristics 
Median age (25th–75th percentiles) 56.0 (47–62)  53.0 (45–61)  
FIGO * stage      
    IIIb 37 17.5 41 19.0 
    IIIc 162 76.8 166 76.9 
    IV 12 5.7 4.2 
    Missing data 
Residual tumor     
    None 79 37.4 80 37.0 
    ≤1 cm 118 55.9 130 60.2 
    1–2 cm 12 5.7 1.9 
    Missing data 0.9 0.9 
Tumor grade     
    1 (well differentiated) 11 5.2 19 8.8 
    2 (moderately well differentiated) 37 17.5 50 23.1 
    3 (poorly differentiated) 160 75.8 142 65.7 
    Missing data 1.4 2.3 
Cell type     
    Serous 132 62.6 155 71.8 
    Endometrioid 28 13.3 21 9.6 
    Mucinous 2.8 1.9 
    Clear-cell 12 5.7 1.9 
    Undifferentiated 23 10.9 18 8.3 
    Other 3.8 12 5.6 
    Missing data 0.9 0.9 
  No lymphadenectomy (n = 211)
 
  Lymphadenectomy (n = 216)
 
 
Characteristics 
Median age (25th–75th percentiles) 56.0 (47–62)  53.0 (45–61)  
FIGO * stage      
    IIIb 37 17.5 41 19.0 
    IIIc 162 76.8 166 76.9 
    IV 12 5.7 4.2 
    Missing data 
Residual tumor     
    None 79 37.4 80 37.0 
    ≤1 cm 118 55.9 130 60.2 
    1–2 cm 12 5.7 1.9 
    Missing data 0.9 0.9 
Tumor grade     
    1 (well differentiated) 11 5.2 19 8.8 
    2 (moderately well differentiated) 37 17.5 50 23.1 
    3 (poorly differentiated) 160 75.8 142 65.7 
    Missing data 1.4 2.3 
Cell type     
    Serous 132 62.6 155 71.8 
    Endometrioid 28 13.3 21 9.6 
    Mucinous 2.8 1.9 
    Clear-cell 12 5.7 1.9 
    Undifferentiated 23 10.9 18 8.3 
    Other 3.8 12 5.6 
    Missing data 0.9 0.9 
*

International Federation of Gynecology and Obstetrics.

Surgical Procedures

Sixteen patients allocated to the no-lymphadenectomy arm had more than 25 pelvic or 15 lumbo-aortic nodes removed; conversely, 27 patients allocated to systematic lymphadenectomy had fewer than 25 pelvic or 15 lumbo-aortic nodes resected ( Fig. 1 ). These 43 patients were excluded from the per-protocol survival analyses for inappropriate surgical treatment.

Table 2 shows the median number and interquartile range of resected nodes by treatment arm. In the systematic lymphadenectomy arm, the median number of removed pelvic and aortic lymph nodes was 28.5 (interquartile range = 22 to 41) and 23 (interquartile range = 16 to 32), respectively. Overall, the median number of removed nodes was 51.5 (interquartile range = 41 to 70) in the lymphadenectomy arm and 4 (interquartile range = 0 to 11) in the control arm. As a consequence of the high number of resected nodes in the systematic lymphadenectomy arm, more of these patients had positive nodes at histologic examination than patients in the no-lymphadenectomy arm (70% vs. 42%, P <.001).

Table 2.

Median number (25th–75th percentiles) of resected nodes by treatment arm

Nodal site No lymphadenectomy (n = 211) Lymphadenectomy (n = 216) P 
Pelvic 1 (0–6) 28.5 (22–41) <.001 
Lumbo-aortic 1 (0–5) 23 (16–32) <.001 
Pelvic and lumbo-aortic 4 (0–11) 51.5 (41–70) <.001 
Missing data  
Nodal site No lymphadenectomy (n = 211) Lymphadenectomy (n = 216) P 
Pelvic 1 (0–6) 28.5 (22–41) <.001 
Lumbo-aortic 1 (0–5) 23 (16–32) <.001 
Pelvic and lumbo-aortic 4 (0–11) 51.5 (41–70) <.001 
Missing data  

Systematic lymphadenectomy had a statistically significant impact on surgical parameters such as median operative time, median blood loss, and the proportion of patients undergoing blood transfusions ( Table 3 ). However, the difference in the number of hospital days between women in the two trial arms was not statistically significant. Although the number of intra-operative complications was similar in the two arms (i.e., 15 in the control arm and 17 in lymphadenectomy arm), systematic lymphadenectomy had greater perioperative and late morbidity (60 patients vs. 39 patients in the control arm, P = .014). Most of the difference in morbidity between the two trial arms was due to formation of lymphocysts and lymphedema that occurred in 14 patients in the systematic lymphadenectomy group versus no patients in the control arm. The frequency of leakage of colorectal anostomosis (two patients vs. five patients), intestinal fistula (four patients vs. four patients), and adhesive small bowel obstruction (three patients vs. one patient) was similar in the trial arms. No surgery-related deaths occurred.

Table 3.

Operative details and postoperative hospital stay

Surgical outcome No lymphadenectomy (n = 211) Lymphadenectomy (n = 216) P 
Median operating time (min) (25th–75th percentiles) 210 (170–280) 300 (250–360) <.001 
    Missing data  
Median blood loss (mL) (25th–75th percentiles) 650 (400–1200) 1000 (600–1500) <.001 
    Missing data 14 10  
Patients transfused (%) 59.2 71.7 .006 
Median hospital stay (days) (25th–75th percentiles) 9 (7–12) 9 (7–13) .21 
    Missing data 10  
Surgical outcome No lymphadenectomy (n = 211) Lymphadenectomy (n = 216) P 
Median operating time (min) (25th–75th percentiles) 210 (170–280) 300 (250–360) <.001 
    Missing data  
Median blood loss (mL) (25th–75th percentiles) 650 (400–1200) 1000 (600–1500) <.001 
    Missing data 14 10  
Patients transfused (%) 59.2 71.7 .006 
Median hospital stay (days) (25th–75th percentiles) 9 (7–12) 9 (7–13) .21 
    Missing data 10  

Adjuvant Chemotherapies

The vast majority of patients received adjuvant chemotherapy (94% in the control arm and 96% in the systematic lymphadenectomy arm) after cytoreductive surgery. No differences in chemotherapy schedules were found between the two trial arms; 93% of patients underwent platinum-based mono- (17%) or multi- (76%) agent chemotherapy regimens, and only 7% of patients received non–platinum-based chemotherapy regimens. Taxol was used in combination with platinum-based agents in 39% of patients.

Progression-Free and Overall Survival

After a median follow-up of 68.4 months (interquartile range = 35.2 to 90.7 months), tumors had recurred in 292 patients (68.4%) and 202 patients (47.3%) had died, 11 without evidence of recurrent disease (2.6%). Site of disease recurrence by treatment arm is shown in Table 4 ; Figs. 2 and 3 depict progression-free and overall survival, respectively, for all eligible patients. Comparison of the Kaplan–Meier curves for occurrence of first event ( Fig. 2 ) gave a hazard ratio of 0.76 (95% CI = 0.60 to 0.96, P = .02), which translated into increases in 3-year progression-free survival, from 31.6% to 41.6% (difference = 10.0%, 95% CI = 1.6% to 18.3%) and in 5-year progression-free survival, from 21.6% to 31.2% (difference = 9.6%, 95% CI = 1.5% to 21.6%). Median progression-free survival was 22.4 months (interquartile range = 11.8 to 52.0 months) for patients in the control arm and 27.4 months (interquartile range = 14.3 to 83.9 months) for patients in the systematic lymphadenectomy arm. When we calculated the improvement in median progression-free survival using the hazard ratio and median progression-free survival in the control arm, the difference was 7.0 months (95% CI = 1.0 to 14.4 months), i.e., it improved from 22.4 months to 29.4 months.

Table 4.

Site of disease recurrence by treatment arm

  No lymphadenectomy (n = 211)
 
  Lymphadenectomy (n = 216)
 
 
Recurrence site No. No. 
No recurrence 65 30.8 81 37.5 
Recurrence 146 69.2 135 62.5 
    Pelvic 28 13.3 26 12.0 
    Intraperitoneal 58 27.5 56 25.9 
    Retroperitoneal 2.4 2.3 
    Distant site 16 7.6 10 4.6 
    Multiple sites 30 14.2 26 12.0 
    Missing data 4.3 12 5.5 
  No lymphadenectomy (n = 211)
 
  Lymphadenectomy (n = 216)
 
 
Recurrence site No. No. 
No recurrence 65 30.8 81 37.5 
Recurrence 146 69.2 135 62.5 
    Pelvic 28 13.3 26 12.0 
    Intraperitoneal 58 27.5 56 25.9 
    Retroperitoneal 2.4 2.3 
    Distant site 16 7.6 10 4.6 
    Multiple sites 30 14.2 26 12.0 
    Missing data 4.3 12 5.5 
Fig. 2.

Progression-free survival (PFS) for patients with optimally debulked advanced ovarian carcinoma undergoing systematic aortic and pelvic lymphadenectomy (Lymphad.) versus resection of bulky nodes only (No lymphad.). Median PFS times were 27.4 months (interquartile range = 14.3 to 83.9 months) in the systematic lymphadenectomy arm and 22.4 months (interquartile range = 11.8 to 52.0 months) in the no-lymphadenectomy arm.

Fig. 2.

Progression-free survival (PFS) for patients with optimally debulked advanced ovarian carcinoma undergoing systematic aortic and pelvic lymphadenectomy (Lymphad.) versus resection of bulky nodes only (No lymphad.). Median PFS times were 27.4 months (interquartile range = 14.3 to 83.9 months) in the systematic lymphadenectomy arm and 22.4 months (interquartile range = 11.8 to 52.0 months) in the no-lymphadenectomy arm.

Fig. 3.

Overall survival (OS) for patients with optimally debulked advanced ovarian carcinoma undergoing systematic aortic and pelvic lymphadenectomy (Lymphad.) versus resection of bulky nodes only (No lymphad.). Median OS times were 62.1 months (interquartile range = 30.9 months to still not reached) in the systematic lymphadenectomy arm and 56.3 months (interquartile range = 31.3 to 123.6 months) in the no-lymphadenectomy arm.

Fig. 3.

Overall survival (OS) for patients with optimally debulked advanced ovarian carcinoma undergoing systematic aortic and pelvic lymphadenectomy (Lymphad.) versus resection of bulky nodes only (No lymphad.). Median OS times were 62.1 months (interquartile range = 30.9 months to still not reached) in the systematic lymphadenectomy arm and 56.3 months (interquartile range = 31.3 to 123.6 months) in the no-lymphadenectomy arm.

Comparison of the Kaplan–Meier curves for mortality from any cause ( Fig. 3 ) gave a hazard ratio of 0.96 (95% CI = 0.73 to 1.26, P = .77), which translated into non–statistically significant absolute increases in 3-year overall survival, from 67.5 to 68.6% (difference = 1.1%, 95% CI = −6.6 to 7.6%), and in 5-year overall survival, from 48.0 to 49.5% (difference = 1.5%, 95% CI = −8.4 to 10.6%). Median overall survival was 56.3 months (interquartile range = 31.3 to 123.6 months) for patients in the no-lymphadenectomy arm and 62.1 months (interquartile range = 30.9 months to still not reached) for patients in the systematic lymphadenectomy arm. Similar to the calculations for progression-free survival, we calculated the increase in overall survival in the systematic lymphadenectomy arm using the hazard ratio and median overall survival in the control arm; the difference was 2.4 months (95% CI = −11.8 to 21.0 months), corresponding to an increase from 56.3 months to 58.7 months.

Cox proportional hazards analysis was performed to adjust the treatment comparison for baseline characteristics ( Table 5 ). When histologic grade and residual disease were taken into account, the hazard ratios for occurrence of first event and mortality from any cause were almost unchanged (i.e., 0.75 [95% CI = 0.59 to 0.94] and 0.97 [95% CI = 0.74 to 1.29], respectively).

Table 5.

Multivariable Cox proportional hazards analysis for progression-free survival (PFS) and overall survival (OS) to adjust the risk associated with therapy for various prognostic factors *

Prognostic factor  PFS
 
  OS
 
 
 HR (95% CI) P HR (95% CI) P 
Treatment arm     
    No lymphadenectomy 1.0 (Referent) .01 1.0 (Referent) .85 
    Lymphadenectomy 0.75 (0.59–0.94)  0.97 (0.74–1.29)  
Tumor grade     
    1 or 2 1.0 (Referent) .76 1.0 (Referent) .25 
    3 1.04 (0.81–1.34)  1.20 (0.88–1.64)  
Residual tumor     
    No 1.0 (Referent) <.001 1.0 (Referent) .002 
    Yes 1.65 (1.29–2.12)  1.59 (1.18–2.15)  
Prognostic factor  PFS
 
  OS
 
 
 HR (95% CI) P HR (95% CI) P 
Treatment arm     
    No lymphadenectomy 1.0 (Referent) .01 1.0 (Referent) .85 
    Lymphadenectomy 0.75 (0.59–0.94)  0.97 (0.74–1.29)  
Tumor grade     
    1 or 2 1.0 (Referent) .76 1.0 (Referent) .25 
    3 1.04 (0.81–1.34)  1.20 (0.88–1.64)  
Residual tumor     
    No 1.0 (Referent) <.001 1.0 (Referent) .002 
    Yes 1.65 (1.29–2.12)  1.59 (1.18–2.15)  
*

HR = hazard ratio; CI = confidence interval.

Per the trial protocol, we also performed an analysis to evaluate the effect of systemic lymphadenectomy on progression-free survival and overall survival among the 384 patients who received appropriate surgery (n = 195 in the control arm and n = 189 in the systematic lymphadenectomy arm; see Fig. 1 ). In a multivariable Cox model, the hazard ratios for occurrence of first event and mortality from any cause were 0.69 (95% CI = 0.54 to 0.89) and 0.93 (95% CI = 0.69 to 1.25), respectively.

The prognostic value of positive aortic and/or pelvic lymph nodes in the systematic lymphadenectomy arm was also evaluated. The hazard ratio for death from any cause among patients with positive aortic and/or pelvic lymph nodes in the systematic lymphadenectomy arm (as compared with patients with negative nodes) was 1.61 (95% CI = 1.01 to 2.56). This increased mortality rate is in keeping with the observation that the metastatic involvement of aortic and pelvic nodes is a predictor of survival.

D ISCUSSION

The therapeutic value of systematic lymphadenectomy in women with advanced ovarian cancer remains controversial. Retrospective studies ( 10 , 11 ) have suggested a clinically significant survival advantage following systematic lymphadenectomy in patients undergoing cytoreductive surgery for advanced disease; however, no prospective studies have been published. Therefore, we conducted what is, to our knowledge, the first multicenter randomized clinical trial to determine whether the addition of systematic aortic and pelvic lymphadenectomy to optimal cytoreductive surgery would improve progression-free and overall survival in women with advanced ovarian cancer. We found that, although systematic lymphadenectomy statistically significantly improved progression-free survival, overall survival was similar in both the systematic lymphadenectomy and control arms. In addition, this study showed that a larger number of patients in the systematic lymphadenectomy arm than in the control arm had metastasis to lymph nodes and confirmed that metastatic lymph node involvement is a statistically significant prognostic factor for survival.

Surgical management of patients with ovarian cancer is challenging, and systematic lymphadenectomy is a major surgical procedure. This study provides evidence that the procedure itself is feasible in the framework of a multicenter, randomized clinical trial. Women who received systematic lymphadenectomy had a higher incidence of postoperative complications, although almost all of these were mild, consisting of lymphocysts or lymphedema. The median operating time was about 90 minutes longer and the blood loss about 350 mL higher in the lymphadenectomy arm than in the control arm, and 12% more patients underwent a blood transfusion in the experimental group. There was no difference in the duration of postoperative hospital stay between trial arms, and no perioperative deaths occurred.

Empirical data supporting a therapeutic benefit of systematic lymphadenectomy in women with advanced ovarian cancer are largely indirect. Most of the early observational studies ( 4 , 10 , 1518 ) compared the survival of patients who underwent cytoreductive surgery and lymphadenectomy, which ranged from systematic lymphadenectomy to node sampling, with the survival of patients undergoing cytoreductive surgery only. These studies showed considerable differences in overall survival between groups, with all of the studies favoring lymphadenectomy ( 4 , 10 , 1518 ) . These studies were limited, however, by small numbers of patients and a failure to reasonably account for selection bias. Selection bias creates a type of confounding that can occur because women undergoing successful systematic lymphadenectomy might have had a lower rate of comorbidity or a tumor with more favorable prognostic features than women undergoing node sampling or removal of bulky nodes only. The findings of this study are discordant with these early observational studies. Moreover, when we performed a per-protocol analysis that excluded patients in the systematic lymphadenectomy arm who did not fulfill the criteria for appropriate systematic lymphadenectomy (i.e., not enough nodes removed) and the patients in the control arm who did not fulfill the criteria for appropriate cytoreductive surgery (i.e., too many nodes removed), we were unable to detect any difference in overall survival between the two treatment arms (i.e., hazard ratios for mortality did not change). However, in this subanalysis, the risk of first event (i.e., disease recurrence or death) associated with systematic lymphadenectomy was further reduced (HR = 0.69, 95% CI = 0.54 to 0.89), seemingly reaffirming the impact of systematic lymphadenectomy on the natural history of the disease.

The findings of this study confirm that there is a high prevalence of both pelvic and para-aortic lymph node metastases in patients with advanced ovarian cancer ( 5 ) . The prevalence of lymph node involvement depends closely on the number of lymph nodes removed and examined. In this study, the proportion of patients with lymph node involvement was much higher in the systematic lymphadenectomy arm (70%) than in the control arm (42%). Given that, in the control arm, any clinically suspicious lymph nodes were removed, this difference in the proportion of patients with lymph node involvement may have been due to the low sensitivity and specificity of nodal size as an indicator of ovarian cancer metastases ( 19 , 20 ) . An association between lymph node involvement and clinical outcome has been established ( 10 ) , and patients with nodal metastases have a worse prognosis than patients with negative lymph nodes.

The unfavorable prognosis of patients with positive lymph nodes suggests either that nodal involvement is a marker of the tumor's biologic aggressiveness or that nodal metastases may be resistant to chemotherapy because of diminished blood supply, i.e., the “pharmacologic sanctuary” hypothesis ( 21 ) . The data in this study do not support the latter hypothesis because lymph node involvement in the systematic lymphadenectomy arm remained a statistically significant negative prognostic factor (HR = 1.61, 95% CI = 1.01 to 2.56) for overall survival.

This trial was designed to assess progression-free and overall survival as primary endpoints. We found that patients who underwent systemic lymphadenectomy had a 25% improvement in progression-free survival when compared with patients who had removal of bulky nodes only (HR = 0.75, 95% CI = 0.59 to 0.94). However, this improvement in progression-free survival did not translate into an improvement in overall survival. Two possible explanations for this discrepancy are that 1) follow-up may have been too short and 2) there were differences in second-line chemotherapy. Median overall survival for both trial arms in this study was good because all patients had optimal residual disease at the completion of their cytoreductive surgery, in keeping with other studies ( 22 ) . However, a longer follow-up will be needed before long-term survival values are definitive. In addition, although we did not track our patients for information on second-line chemotherapies, the median times to disease recurrence in the systematic lymphadenectomy arm (27.4 months) and control arm (22.4 months) suggest that most of the patients did not have platinum-resistant disease because the patients relapsed at least 6 months after first-line treatment. Indeed, only 11 more patients relapsed within 12 months from surgery in the control arm (51 patients) than in the systematic lymphadenectomy arm (40 patients). Hence, we do not expect differences in the use of second-line chemotherapy to be a confounding variable in this study because most patients would have received platinum-containing regimens. Therefore, it seems that second-line chemotherapies, with or without surgery at the time of relapse, may confer a survival benefit that is similar in the two trial arms and is independent of the length of the progression-free survival interval.

Maximal cytoreductive surgery remains a mainstay in the treatment of women with advanced ovarian cancer. The results of this study provide the first direct comparison of systematic lymphadenectomy with standard cytoreductive surgery (i.e., no lymphadenectomy). This study found that 1) the addition of systematic lymphadenectomy to cytoreductive surgery prolonged progression-free survival, which, in turn, may have an important impact on the quality of life of patients with advanced ovarian cancer; 2) systematic lymphadenectomy did not prolong overall survival, probably because effective platinum-based first- and second-line (with or without salvage surgery) chemotherapies might have diluted the impact of systematic lymphadenectomy on the risk of death; and 3) patients in the systematic lymphadenectomy arm had a higher number of nodal metastases than patients in the no-lymphadenectomy arm. The superior assessment of node status in patients undergoing systematic lymphadenectomy could help refine the prognosis of patients with advanced ovarian cancer and to tailor their follow-up. Moreover, surgical treatments for advanced ovarian cancer patients, including systematic lymphadenectomy, should be performed in selected institutions that specialize in gynecologic oncology care to minimize the risk of surgical morbidity.

References

(1)
Jemal A, Tiwari RC, Murray T, Ghafoor A, Samuels A, Ward E, et al. Cancer statistics, 2004.
CA Cancer J Clin
 
2004
;
54
:
8
–29.
(2)
Griffith CT. Surgical resection of tumor bulk in the primary treatment of ovarian cancer.
J Natl Cancer Inst Monogr
 
1975
;
42
:
101
–4.
(3)
Hacker NF, Berek JS, Lagasse LD, et al. Primary cytoreductive surgery for epithelial ovarian cancer.
Obstet Gynecol
 
1983
;
61
:
413
–20.
(4)
Bristow RE, Tomacruz RS, Armstrong DK, Trimble EL, Monz FJ. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during platinum era: a meta-analysis.
J Clin Oncol
 
2002
;
20
:
1248
–59.
(5)
Burghardt E, Girardi F, Lahousen M, Tamussino K, Stettner H. Patterns of pelvic and paraaortic lymph node involvement in ovarian cancer.
Gynecol Oncol
 
1991
;
40
:
103
–6.
(6)
FIGO Cancer Committee.
Gynecol Oncol
 
1986
;
25
:
383
.
(7)
Panici PB, Scambia G, Baiocchi G, Matonti G, Capelli A, Mancuso S, et al. Anatomical study of para-aortic and pelvic lymph-nodes in gynecologic malignancies.
Obstet Gynecol
 
1992
;
79
:
498
–502.
(8)
Ferraris G, Lanza A, D'Addato F, et al. Techniques of pelvic and para-aortic lymphadenectomy in the surgical treatment of cervix carcinoma.
Eur J Gynecol Oncol
 
1988
;
9
:
83
–6.
(9)
Benedetti-Panici P, Scambia G, Baiocchi G, Greggi S, Mancuso S. Technique and feasibility of systematic paraaortic and pelvic lymphadenectomy in gynecologic malignancies. A prospective study.
Int J Gynecol Cancer
 
1991
;
1
:
133
–40.
(10)
Chen SS. Survival of ovarian carcinoma with or without lymph node metastasis.
Gynecol Oncol
 
1987
;
27
:
368
–72.
(11)
Scarabelli C, Gallo A, Zarrelli A, Visentin C, Campagnutta E. Systematic pelvic and para-aortic lymphadenectomy during cytoreductive surgery in advanced ovarian cancer: potential benefit on survival.
Gynecol Oncol
 
1995
;
56
:
328
–37.
(12)
Kaplan EL, Meier P. Nonparametric estimation from incomplete observations.
J Am Stat Assoc
 
1958
;
53
:
457
–81.
(13)
Peto R, Pike MC, Armitage P, Breslow NE, Cox DR, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples.
Br J Cancer
 
1977
;
35
:
1
–39.
(14)
Parmar MKB, Machin D. Survival analysis: a practical approach. Chichester (England): John Wiley,
1995
.
(15)
Spirtos NM, Gross GM, Freddo JL, Ballon SC. Cytoreductive surgery in advanced epithelial cancer of the ovary: the impact of aortic and pelvic lymphadenectomy.
Gynecol Oncol
 
1995
;
56
:
345
–52.
(16)
di Re F, Baiocchi G, Fontanelli R, et al. Systematic pelvic and paraaortic lymphadenectomy for advanced ovarian cancer: prognostic significance of node metastases.
Gynecol Oncol
 
1996
;
62
:
360
–5.
(17)
Kikkawa F, Ishikawa H, Tamakoshi K, Suganuma N, Mizuno K, Kawai M, Arii Y, et al. Prognostic evaluation of lymphadenectomy for epithelial ovarian cancer.
J Surg Oncol
 
1995
;
60
:
227
–31.
(18)
Kigawa J, Minagawa Y, Itamochi H, Kanamori Y, Ishihara H, Terakawa N. Retroperitoneal lymphadenectomy including the para-aortic nodes in patients with stage III ovarian cancer.
Am J Obstet Gynecol
 
1994
;
17
:
230
–3.
(19)
Tangjitgamol S, Manusirivithaja S, Sheanakul C, Leelahakorn S, Sripramote M, Thawaramara T, et al. Can we rely on the size of the lymph node in determining nodal metastasis in ovarian carcinoma?
Int J Gynecol Cancer
 
2003
;
13
:
297
–302.
(20)
Arango HA, Hoffman MS, Roberts WS, Decesare SL, Fiorica JV, Drake J. Accuracy of lymph node palpation to determine need for lymphadenectomy in gynecologic malignancies.
Gynecol Oncol
 
2000
;
95
:
553
–6.
(21)
Eisenkop SM, Spirtos NM. The clinical significance of occult macroscopically positive retroperitoneal nodes in patients with epithelial ovarian cancer.
Gynecol Oncol
 
2001
;
82
:
143
–9.
(22)
Markman M, Kennedy A, Webster K, Peterson G, Kulp B, Belinson J. Combination chemotherapy with carboplatin and docetaxel in the treatment of cancers of the ovary and fallopian tube and primary carcinoma of the peritoneum.
J Clin Oncol
 
2001
;
19
:
1901
–5.