Optimal Cytoreduction
Definitions of optimal surgical cytoreduction following completion of the initial surgery have varied in the literature from no macroscopic residual disease to 3 cm gross residual disease.[3-8] In the United States, the GOG has settled on 1 cm or less residual disease as a criterion for optimum cytoreductive surgery. Using this definition, “ultraradical surgery” is now being recommended by several centers in the United States. Ultraradical surgery incorporates splenectomies, partial pancreatectomies, diaphragm resections, and multiple bowel resections into the surgical treatment for patients with advanced-stage (FIGO stage IIIC–IV) ovarian cancer.[14,15]
Reports by advocates for this approach reveal that they have accomplished at least 70% optimum surgical cytoreduction in one series and up to 92% optimum surgical cytoreduction in another series.[14,15] However, when one evaluates these series based on no macroscopic residual disease at the completion of the initial surgery, both series report only a 22% incidence of patients left with no macroscopic residual disease. It would appear that the ultraradical surgery has been most effective in surgically cytoreducing patients who might have > 1 cm residual disease in the past into the ≤ 1 cm disease category, but not into the no macroscopic residual disease category.[14-16]
CT Scan vs Surgical Assessment
Recently, a prospective study of 78 women with advanced ovarian cancer who were optimally cytoreduced to ≤ 1 cm macroscopic residual disease compared postoperative, prechemotherapy CT findings with the surgeon’s assessment of residual disease left at completion of the initial surgical cytoreduction.[17] In 20 (26%) of the 78 cases, residual “definitely malignant disease” > 1 cm was identified by CT scan. In 10 additional cases (13%) findings of “probably malignant residual disease” > 1 cm was identified by CT scan. “Indeterminate findings” were identified in 7 cases (9%). In 41 cases (52%), no disease > 1 cm was identified.
Statistical analysis found that no macroscopic disease left at the completion of the initial surgery was the only factor associated with postoperative, prechemotherapy CT scans showing < 1-cm lesions. Residual disease on postoperative CT scans was significantly more likely to be found in patients with 0.6 to 1 cm residual disease at completion of the initial “optimal” cytoreductive surgery. The authors could not state whether these findings reflect underestimation of residual disease by surgeons, overestimation of residual disease by radiologists, or rapid regrowth of cancer. Regardless of the cause, it supports the importance of cytoreduction to no gross residual disease at the initial surgery.
GOG Studies
Two recent GOG studies have been reported that cause us to reflect on what is being accomplished with surgical cytoreduction in the management of advanced ovarian cancer.[18,19] The first series included 1,895 stage III disease patients, the largest series of stage III patients reported to date.[18] It involved patients participating in one arm of six different prospective, randomized GOG trials. All of the patients in these treatment arms had received cisplatin and paclitaxel chemotherapy in standard intravenous regimens. The GOG studies are significant for being rigidly controlled regarding type and dosage of chemotherapy administered and for treating a rather homogeneous group of patients.
Among patients with stage III disease, 23% were optimally cytoreduced to no macroscopic residual disease—a similar percentage to that achieved by those proposing ultraradical surgical techniques to optimally cytoreduce patients with advanced ovarian cancer.[14,15] A major impact on both progression-free and overall survival (OS) was seen in these optimally cytoreduced patients (Figure 1). PFS was 33.0 months for those with no macroscopic residual disease and OS was 71.9 months, a survival status significantly better (P < .001) than the PFS and OS for those with 0.1 to 1 cm residual disease (16.8 and 42.4 months, respectively) and for those with > 1 cm residual disease (PFS = 14.1 months; OS = 35.0 months).
The GOG subsequently reported on 360 stage IV patients participating in one of four prospective, randomized GOG trials, all of whom received the same regimen of intravenous cisplatin and paclitaxel.[19] Only 8% of the patients with stage IV disease were cytoreduced to no macroscopic residual tumor. PFS for those with no macroscopic residual disease was 20.1 months and OS for this group was 64.1 months (Figure 2). Surprisingly, there was no difference in survival for patients with ≤ 1 cm macroscopic residual disease (PFS = 13.0 months; OS = 28.7 months) and those with 1.1 to 5 cm macroscopic residual disease (PFS = 13.0 months; OS = 31 months). The only difference in survival when macroscopic residual disease was encountered was for the group of patients who had more than 5 cm macroscopic residual disease; their PFS was 8.9 months and their OS, 22.5 months.
Once again, the major benefit of cytoreduction surgery accrued to the few patients (8%) with no macroscopic residual disease. Patients with metastases to the supraclavicular, axillary, and mediastinal lymph nodes and those with subcutaneous metastases had a better survival than those with malignant pleural effusions, multiple hepatic metastases, or metastases to multiple sites. Approximately 48% of the patients in this series had stage IV disease by virtue of a malignant pleural effusion. The operative results of this report were consistent with the recent report in which ultraradical surgical cytoreduction for stage IV patients resulted in only 6% who were surgically cytoreduced to no macroscopic residual disease.[16]
Best Cytoreduction Results With Initial Surgery
Ideally, surgical cytoreduction to no macroscopic residual disease should approach 100% in patients undergoing initial surgery for advanced ovarian cancer. Only two groups of investigators have reported a greater than 80% initial surgical cytoreduction rate to no macroscopic residual disease when treating patients with advanced ovarian cancer (Table 1).[5,9,10]
Table 1 | |||
Complete Surgical Cytoreduction in Advanced Ovarian Cancer | |||
Parameter | Eisenkop | Eisenkop | Scholz |
Disease stage | IIIC, IV | IIIC | IIIC, IV |
Number of patients | 160 | 213 | 55 |
Mean operative | 254 min | 180 min | 372 min |
Mean estimated | 1,190 mL | 980 mL | — |
Mean number of | 4 (0–22) | 4 (0–24) | 9 (2–22) |
Median hospital stay | 12 d (2–61 d) | 12 d (3–64 d) | 35 d (13–73 d) |
Deaths ≤ 30 d | 3 | 4 | 0 |
No visible | 85.3% | 87.7% | 83% |
Median survival | 54 mo | 75.8 mo | 47 mo |
Eisenkop and colleagues’ most recent report on their success in applying ultraradical cytoreduction to remove all gross disease was approximately 85% no macroscopic residual disease in stage IIIC patients, with ≤ 1 cm residual disease achieved in 12%, and > 1 cm residual disease in the very few remaining patients.[9] These investigators believed that patients cytoreduced to no macroscopic residual disease did well regardless of which agents were combined with platinum as adjuvant therapy. Their median overall survival for stage IIIC disease was 75.8 months.[9] Survival was statistically influenced only by the extent of peritoneal metastatic implants that had to be removed. Mean operative time was 180 minutes (range = 90–380 minutes), estimated blood loss was 980 mL (30–7,000 mL), mean number of transfusions was 4 units (0–24 units), and median hospital stay 12 days (2–61 days). The authors reported four deaths within 30 days of surgery.
Scholz et al recently reported their success in treating 55 patients with stage IIIC and IV disease, in whom 83% were surgical cytoreduced to no residual tumor (Table 1).[10] Mean operative time was 372 minutes (160–590 minutes). Patients participating in this series received an average of 9 units of blood (2–22 units), 8 units of fresh frozen plasma intraoperatively (0–24 units), and 2 units of blood postoperatively (0–9 units). Five patients required reoperation (two for leakage of anastamosis, one for bleeding, one for bowel leakage, and one for a stomach lesion). Mean hospital stay was 35 days (13–73 days). No deaths occurred within the first 30 days postoperatively. The authors reported a 28-month PFS, but their patients’ overall survival was only 48 months.
