Fax: (905) 575-6343
Systematic review of adjuvant care for women with Stage I ovarian carcinoma†
Article first published online: 27 SEP 2004
Copyright © 2004 American Cancer Society
Volume 101, Issue 9, pages 1926–1935, 1 November 2004
How to Cite
Elit, L., Chambers, A., Fyles, A., Covens, A., Carey, M. and Kee Fung, M. F. (2004), Systematic review of adjuvant care for women with Stage I ovarian carcinoma. Cancer, 101: 1926–1935. doi: 10.1002/cncr.20595
The current study was performed on behalf of the Cancer Care Ontario Practice Guidelines Initiative's Gynecology Cancer Disease Site Group.
- Issue published online: 18 OCT 2004
- Article first published online: 27 SEP 2004
- Manuscript Revised: 14 JUL 2004
- Manuscript Accepted: 14 JUL 2004
- Manuscript Received: 1 APR 2004
- Cancer Care Ontario Practice Guidelines Initiative's Gynecology Cancer Disease Site Group
- Ontario Ministry of Health and Long-Term Care
- ovarian carcinoma;
- adjuvant chemotherapy;
- adjuvant radiotherapy;
- neoplasm staging
Several adjuvant care interventions to treat women with Stage I ovarian carcinoma have been studied. The aim of the current systematic review was to determine the optimal strategy for adjuvant care for women with Stage I ovarian carcinoma.
A systematic search was conducted to find randomized controlled trials published between 1965 and April 2004 that examined adjuvant therapy (e.g., chemotherapy and radiotherapy) for women with Stage I ovarian carcinoma.
Thirteen randomized controlled trials were identified that compared adjuvant therapies for women with Stage I ovarian carcinoma. Eight of these trials reported results only for patients with Stage I disease. The majority of patients in the five randomized trials that compared adjuvant chemotherapy with no chemotherapy did not receive lymphadenectomy as part of their surgical staging. The pooled results for Stage I patients indicated a survival benefit (relative risk [RR], 0.74; 95% confidence interval [CI], 0.58–0.94; P = 0.01), and a benefit in terms of a reduced risk of developing disease recurrence (RR, 0.70; 95% CI, 0.58–0.86; P = 0.0004) favoring adjuvant chemotherapy. Platinum-based adjuvant chemotherapy was reported to improve overall 5-year survival (absolute survival difference 8%; 95% CI, 2–12%; hazard ratio, 0.67; 95% CI, 0.50–0.90; P = 0.008).
Adjuvant platinum-based chemotherapy for women with Stage I ovarian carcinoma improved survival and reduced the risk of recurrent disease. The optimally staged group accounted for approximately 10% of women with Stage I disease. The role of adjuvant chemotherapy in optimally staged patients (especially those with good prognostic factors) has not been assessed adequately. Cancer 2004. © 2004 American Cancer Society.
Ovarian carcinoma is the fourth leading cause of cancer mortality among women. Approximately 27% of patients present with cancer confined to the ovary (Stage I), and their 5-year survival rate is 85%.1 There is controversy concerning the benefit of adjuvant therapy for women with Stage I disease. Some physicians would choose not to offer adjuvant therapy because for 80% of women, surgery alone is adequate treatment.2, 3 They would debate that adjuvant therapy for all Stage I patients could lead to unnecessary adverse events with no survival benefit for the majority of the patients. In contrast, other physicians would offer adjuvant therapy, as optimal surgical staging is only conducted for 10% of women with presumptive Stage I disease according to a population study by Munoz et al.4 Young et al.5 showed that in inadequately staged patients, a reoperation upstages approximately 25% of women, and 70% of these women are upstaged to Stage III disease. Adjuvant chemotherapy is the standard of care for patients with higher stage disease. A compromise position is to allow prognostic factors to guide the decision for adjuvant treatment for patients with Stage I disease. If a patient receives poor initial staging surgery but has a tumor with favorable prognostic factors, reoperation to complete the surgical staging may be used to avoid chemotherapy. If a woman has a tumor with poor prognostic factors,6–12 a second surgery may be avoided and adjuvant chemotherapy administered. (The poor prognostic factors that could influence this decision include: poor differentiation,6, 7 clear cell histology, large volume ascites, ascites that are positive for malignancy,8 vegetations, dense adhesions, substage rupture before surgery,8 bilaterality;8 aneuploidy;6 CA125 > 65 units per mL; mitosis; necrosis; anisocaryosis; Bax-negative,9 p53-positive, Ki-67-positive, or HER-2-positive10 status; high microvessel density;11 the presence of vascular endothelial growth factor or epithelial growth factor receptors; and Cox-2-positive status.12) Some prognostic factors have significant shortcomings, due to their subjectivity, lack of reproducibility, and low prognostic power.13 Given the variation in decision-making and the recent publication of two large randomized trials investigating adjuvant therapy for women with Stage I ovarian carcinoma,14, 15 the Gynecologic Cancer Disease Site Group of Cancer Care Ontario believed that a systematic review of the literature was warranted to inform evidence-based practice. Important issues that were considered when reviewing the studies included the degree of surgical staging received by women, treatment, age of the study (differences over time in standard of care), and stages of disease eligible for inclusion in the studies.
MATERIALS AND METHODS
The MEDLINE, CANCERLIT, Cochrane Library, Physician Data Query, Canadian Medical Association, and National Guidelines Clearinghouse databases (covering the period from 1965 to April 2004) were searched using variations on the following search terms: neoplasms, ovarian, combined with early stage, Stage I, chemotherapy, surgery, radiotherapy, practice guidelines, randomized controlled trials, and controlled clinical trials. Proceedings of the 1997–2003 meetings of the American Society of Clinical Oncology, as well as reference lists of papers and review articles, were scanned for additional citations.
Articles were selected for inclusion in our systematic review if they were fully published reports or abstracts of randomized controlled trials (RCTs) comparing two or more adjuvant treatments (e.g., chemotherapy and radiotherapy) for women with Stage I ovarian carcinoma. The RCTs had to report outcomes specifically for women with Stage I disease, or > 60% of the women included in the trial had to have Stage I disease. Articles published in languages other than English were excluded from the systematic review.
Synthesizing the Evidence
To estimate the overall effect of the interventions on survival and disease recurrence for women with Stage I ovarian carcinoma, mortality data were abstracted from the published reports of individual RCTs and pooled using the Review Manager software (RevMan 4.1) provided by the Cochrane Collaboration (Metaview Update Software). Results are expressed as relative risks (RR) with 95% confidence intervals (CI). The random-effects model was used for comparative testing of the pooled results across studies, in preference to the fixed-effects model. Random effects models are more conservative than the fixed-effects models due to calculations that produce wider CIs, as it is more difficult to detect a statistically significant difference with wider CIs.16
Literature Search Results
Thirteen published RCTs were identified that compared treatments for women with Stage I ovarian carcinoma. Two publications each reported two separate RCTs.2, 3 Another two RCTs had three treatment arms.17, 18 Two RCTs compared radiotherapy with no adjuvant treatment.18, 19 Seven RCTs compared an adjuvant chemotherapy regimen with no adjuvant chemotherapy.2, 3, 6, 14, 17, 18, 20 Three RCTs compared adjuvant chemotherapy with radiotherapy.17, 18, 21 Four RCTs were identified that compared an adjuvant chemotherapy regimen with intraperitoneal (IP) radioactive chromic phosphate (32P).2, 3, 22, 23 The other two RCTs compared two different forms of adjuvant chemotherapy.24, 25 Table 1 summarizes these RCTs.
|Study||FIGO stages/grades considered||No. of patients||Treatment A||Treatment B||Median follow-up (mos)|
|Adjuvant RT vs. no RT|
|Dembo et al., 197919||Stage I||41||Pelvic RT||No treatment||NR|
|Hreshchyshyn et al., 198018||Stage I||52||Pelvic RT||No treatment||36|
|Adjuvant CT vs. no CT|
|Trimbos et al., 200314||Stage IA–B, Grade 2–3; Stage IC–IIA; Stage I–IIA clear cell carcinoma||448||Platinum-based chemotherapy||No treatment||59|
|ICONI Collaborators, 200315||Stage I–III||477||Platinum-based chemotherapy||No treatment||51|
|Trope et al., 20006||Stage I, Grade 2–3 disease or Grade I aneuploid or clear cell carcinoma||162||Carboplatin||No treatment||46|
|Bolis et al., 19953||Stage IA–B, Grade 2–3||83||Cisplatin||No treatment||76|
|Young et al., 19902||Stage IA–B, Grade 1–2||81||Melphalan||No treatment||>72|
|Gronroos et al., 198417||Stage I||75||CT||No treatment||36 (overall)|
|Hreshchyshyn et al., 198018||Stage IA–B||63||Melphalan||No treatment||36|
|Adjuvant CT vs. RT|
|Chiara et al., 199421||Stage I–II||69||Cisplatin||WAR||60|
|Gronroos et al., 198417||Stage I||65||CT||Pelvic RT||36 (overall)|
|Hreshchyshyn et al., 198018||Stage IA–B||57||Melphalan||Pelvic RT||36|
|Adjuvant CT vs. IP 32P|
|Young et al., 200322||Stage IA–IIC||251||Cyclophosphamide + cisplatin||32P||120|
|Bolis et al., 19953||Stages IA2, IB2, IC||152||Cisplatin||32P||76|
|Vergote et al., 199223||Stage IA–III||340||Cisplatin||32P||62|
|Young et al., 19902||Stage I–II, Grade 2–3||141||Melphalan||32P||>72|
|Two different forms of adjuvant CT|
|Bell et al., 200324||Stage IA–B, Grade 3; Stage IC clear cell carcinoma; Stage II (completely resected)||321||Three cycles of paclitaxel + carboplatin||Six cycles of paclitaxel + carboplatin||54|
|Hatae et al., 199825||Stage IA||96||Intravenous CT||Oral CT||19|
To understand whether the trials should be combined and compared, it was important to determine the adequacy of surgical staging. The European Organization for Research and Treatment of Cancer (EORTC) scale of completeness of surgery was used to describe the study population (optimal—complete staging including lymph node sampling; modified—everything that falls between minimal and optimal; minimal—careful inspection and palpation of all peritoneal surfaces and biopsy specimens of suspected lesions and washing and omentectomy; and inadequate—careful inspection and palpation of all peritoneal surfaces and biopsy specimens of suspected lesions). Eleven RCTs indicated the surgical procedure.2, 3, 6, 14, 15, 17–19, 21–23 One RCT met the minimum requirement,23 four RCTs met the modified requirement,2, 3, 6, 21 and two met the optimal requirement14, 22 (Table 2).
|Study||Treatment||TAH + BSO||Laparotomy (vertical incision)||Tumor capsule examined||Peritoneal washings||Suspicious lesions biopsied||Pelvic and paraaortic lymph node sampling||EORTC staging classification|
|Trimbos et al., 200314||CT vs. no CT||✓a||—||✓||✓a||✓||✓a||Optimal|
|ICONI Collaborators, 200315||CT vs. no CT||✓||—||—||—||—||—||Inadequate|
|Young et al., 200322||CT vs. 32P||✓||✓||✓||✓||✓||✓||Optimal|
|Trope et al., 20006||CT vs. no CT||✓||✓||✓||✓||✓||—||Modified|
|Bolis et al., 19953||CT vs. no CT||✓||✓||✓||✓||✓||✓||Modified|
|Chiara et al., 199421||CT vs. WAR||✓a||—||—||—||✓a||—||Modifieda|
|Vergote et al., 199223||CT vs. 32P||✓||✓||—||—||✓||—||Minimal|
|Young et al., 19902||CT vs. no CT||✓||✓||✓||✓||✓||✓||Modified|
|Gronroos et al., 198417||CT vs. no CT||✓||—||—||—||—||—||Inadequate|
|Hreshchyshyn et al., 198018||CT vs. no CT||✓||—||—||—||—||—||Inadequate|
|Dembo et al., 197931||RT vs. WAR vs. RT + CT||✓||—||—||—||—||—||Inadequate|
Survival and Disease Recurrence
Adjuvant radiotherapy versus no adjuvant radiotherapy
Two RCTs compared adjuvant radiotherapy with no adjuvant radiotherapy in women with Stage I ovarian carcinoma18, 19 (Table 3). Both studies were published before the development of surgical staging guidelines. Therefore, patients may have been understaged, as the upper abdomen and retroperitoneum were not assessed formally for disease. According to the EORTC surgical staging guidelines, women were staged inadequately in both studies. The RCT by Dembo et al.19 included women with Stage I and II disease, and compared no adjuvant radiotherapy (observation) with pelvic radiotherapy, 45 gray in 20 fractions. Twenty-seven women with Stage I disease were included in each treatment arm. Although there was no benefit in terms of disease recurrence among the women with Stage I disease, this was a small study that was not sufficiently powered to detect a clinically meaningful difference between treatment groups. The Hreshchyshyn et al. RCT18 included only women with Stage I disease and compared three treatment arms: adjuvant chemotherapy, adjuvant radiotherapy, and no adjuvant therapy. Seventeen percent of the patients receiving no adjuvant therapy had disease recurrence compared with 30% of the patients receiving adjuvant radiotherapy (P < 0.05). Although 168 patients were recruited for that study, a major limitation was that 82 patients were excluded (49%) because they were judged to have a tumor of low malignant potential, refused the prescribed treatment, or were removed by their physician. When these patients were removed from the analysis, the treatment arms no longer matched with respect to prognostic factors. Therefore, the results were not internally valid.
|Study||Treatment||No. of patients||5 yr overall survival||HR/P value||5 yr disease-free survival||HR/P value|
|Adjuvant RT vs. no RT|
|Dembo et al., 197919||Pelvic RT vs. no treatment||41||87%||No difference between groups||NR|
|Hreshchyshyn et al., 198018||Pelvic RT vs. no treatment||23 vs. 29||Recurrence rate: 30% vs. 17%||P < 0.05|
|Adjuvant CT vs. no CT|
|Trimbos et al., 200314||Platinum-based CT vs. no treatment||224 vs. 224||85% vs. 78%||HR, 0.69 (95% CI, 0.44–1.08); P = 0.10||RFS: 68% vs. 76%||HR, 0.63 (95% CI, 0.43–0.92); P = 0.02|
|ICONI Collaborators, 200315||Platinum-based CT vs. no treatment||241 vs. 236||79% vs. 70%||HR, 0.66 (95% CI, 0.45–0.97); P = 0.03||RFS: 73% vs. 62%||HR, 0.65 (95% CI, 0.46–0.91); P = 0.01|
|Trope et al., 20006||Carboplatin vs. no treatment||81 vs. 81||DSS: 86% vs. 85%||HR, 0.94 (95% CI, 0.37–2.36)||70% vs. 71%||HR, 0.98 (95% CI, 0.52–1.83)|
|Bolis et al., 19953||Cisplatin vs. no treatment||41 vs. 44||88% vs. 82%||P = NS||NR|
|Young et al., 19902||Melphalan vs. no treatment||38 vs. 43||98% vs. 94%||P = 0.43||98% vs. 91%||P = 0.41|
|Gronroos et al., 198417||CT vs. no treatment||38 vs. 37||2 yr survival: 74.1% vs. 95.8%||P = 0.02||NR|
|Hreshchyshyn et al., 198018||Melphalan vs. no treatment||34 vs. 29||Recurrence rate: 6% vs. 17%||P < 0.05|
|Adjuvant CT vs. RT|
|Chiara et al., 199421||Cisplatin vs. WAR||36 vs. 34||71% vs. 53%||P = 0.16||RFS: 74% vs. 50%||P = 0.07|
|Gronroos et al., 198417||CT vs. pelvic RT||38 vs. 27||74% vs. 87%||P = NS||NR|
|Hreshchyshyn et al., 198018||Melphalan vs. pelvic RT||34 vs. 23||Recurrence rate: 6% vs. 30%||P < 0.05|
|Adjuvant CT vs. IP 32P|
|Young et al., 200322||Cyclophosphamide + cisplatin vs. 32P||251||81% vs. 79%||P = NS||10 yr RFS: 70% vs. 76%||P = NS|
|Bolis et al., 19953||Cisplatin vs. 32P||82 vs. 79||81% vs. 79%||P = NS||NR|
|Vergote et al., 199223||Cisplatin vs. 32P||171 vs. 169||NR||79% vs. 82%||P = NS|
|Young et al., 19902||Melphalan vs. 32P||68 vs. 73||81% vs. 78%||P = 0.48||80% vs. 80%||P = NS|
|Two different forms of adjuvant CT|
|Bell et al., 200324||3 cycles of paclitaxel + carboplatin vs. 6 cycles of paclitaxel + carboplatin||321||Probability of recurrence within 5 yrs: 27% vs. 19%||P = NS|
|Hatae et al., 199825||i.v. CT vs. oral CT||96||No difference between groups||No difference between groups|
Adjuvant chemotherapy versus no adjuvant chemotherapy
Seven RCTs compared adjuvant chemotherapy with no adjuvant chemotherapy for women with Stage I ovarian carcinoma (Table 3). According to the EORTC surgical staging guidelines, one RCT was classified as optimal,14 three were classified as modified,2, 3, 6 and three were classified as inadequate.15, 17, 18
A metaanalysis of the RCTs that compared adjuvant chemotherapy with no chemotherapy was completed because none of the RCTs had sufficient power to detect a significant difference in survival on their own between patients who received adjuvant chemotherapy and those who did not. A metaanalysis allows us to increase statistical power. Two studies were excluded from the metaanalysis. The RCT by Hreshchyshyn et al.18 was omitted because the removal of 49% of the randomized patients from the analysis left the treatment arms no longer balanced with respect to prognostic factors. The RCT by Gronroos et al.17 was also excluded because Stage I-specific, 5-year survival information was not available. When the remaining 5 RCTs comparing adjuvant chemotherapy with no adjuvant chemotherapy were pooled for mortality,2, 3, 6, 14, 15 the analysis detected a significant difference in mortality in favor of adjuvant chemotherapy (RR, 0.74; 95% CI, 0.58–0.94; P = 0.01; Fig. 1). The Adjuvant Chemotherapy in Ovarian Neoplasm (ACTION) trial14 was the only study that included patients who had received optimal surgical staging, and this trial did not detect a significant 5-year overall survival difference between the adjuvant chemotherapy group and the no adjuvant therapy group (hazard ratio [HR], 0.69; 95% CI, 0.44–1.08; P = 0.10). If the studies with optimal or modified surgery are pooled (i.e., excluding the International Collaborative Ovarian Neoplasm 1 [ICON1],15 trial), there is no significant difference in 5-year survival (RR, 0.81; 95% CI, 0.58–1.21).
The results of five RCTs that compared adjuvant chemotherapy with no adjuvant chemotherapy were pooled for disease recurrence. The RCTs by Young et al.2 and Trope et al.6 did not detect a significant difference in disease recurrence between women who received adjuvant chemotherapy and those who did not. In contrast, the remaining 3 RCTs detected a significant difference in disease recurrence in favor of adjuvant chemotherapy.3, 14, 15 The ICON115 and ACTION trials14 both reported that disease recurrence-free survival is significantly improved for women who receive adjuvant chemotherapy (P < 0.02; Level 1 evidence). In addition, the nonoptimally surgically staged women who received adjuvant chemotherapy had longer disease recurrence-free survival than did the nonoptimally staged women who did not receive adjuvant chemotherapy (P = 0.008). The Bolis et al. RCT3 detected that adjuvant chemotherapy significantly decrease the disease recurrence rate by 65% compared with the no adjuvant chemotherapy group (HR, 0.35; 95% CI, 0.14–0.89; P = 0.02). When the results for disease recurrence from the 5 trials are pooled, there is a significant difference favoring adjuvant chemotherapy (RR, 0.70; 95% CI, 0.58–0.86; P = 0.0004; Fig. 2). Disease recurrence in the optimally staged ACTION trial favored chemotherapy compared with no adjuvant therapy (RR, 0.70; 95% CI, 0.50–0.97). However, disease recurrence in the optimally and modified groups was not significant (RR, 0.73; 95% CI, 0.52–1.02).
Chemotherapy versus pelvic radiotherapy or whole abdominopelvic radiotherapy
Three RCTs were identified that compared adjuvant chemotherapy with radiotherapy17, 18, 21 (Table 3). Two of the RCTs were classified as having inadequate surgical staging17, 18 and one RCT was classified as having modified surgical staging according to the EORTC surgical staging guidelines.21 The RCTs were not pooled because 2 RCTs were published > 20 years ago, and the chemotherapy regimens described are not the standard of care used today. None of the RCTs detected a significant survival difference between treatment groups. However, the RCT by Hreshchyshyn et al. reported a significant decrease in the disease recurrence rate among the women in the adjuvant chemotherapy group compared with women in the adjuvant radiotherapy group (6% vs. 30%, respectively, P < 0.05).
Of the three RCTs identified, the highest-quality study is that conducted by Chiara et al.21 However, this RCT contains flaws. For example, the analysis was conducted by treatment received rather than by intent, and only 60 women were included in the study, 47 of whom had Stage I disease (68%). Chiara et al. concluded that disease recurrence-free survival (74% vs. 50%) and overall survival (71% vs. 53%) were improved in the chemotherapy arm compared with the radiotherapy arm, but did not reach statistical significance.
Adjuvant chemotherapy versus radioactive chromic phosphate
Four RCTs compared adjuvant chemotherapy with IP 32P2, 3, 22, 23 (Table 3). According to the EORTC surgical staging guidelines, one RCT had optimal surgical staging,22 one had minimal surgical staging,23 and the other two were classified as having modified surgical staging.2, 3 None of the studies detected a significant difference in survival between groups. The RCT by Bolis et al.3 was the only one to detect a reduction (61%) in the number of disease recurrences in women treated with chemotherapy compared with women treated with 32P (P = 0.007). The most recently published RCT, by Young et al.,22 reported that the probability of disease recurrence within 10 years was 35% for women treated with 32P and 28% for women treated with adjuvant chemotherapy (not significant). This RCT included women with Stage I or II disease. Approximately 70% of the women included in the trial had Stage I disease. This was one of two trials included in the systematic review in which eligible women received optimal surgical staging. When disease recurrence was adjusted for stage and grade of disease, Young et al. reported that there was a 29% decrease in the disease recurrence rate among women receiving adjuvant chemotherapy compared with women receiving 32P. Unfortunately, this trial did not report results separately for the women with Stage I disease.
Two different forms of chemotherapy
Two RCTs (abstracts) compared two different forms of chemotherapy in women with early-stage ovarian carcinoma24, 25 (Table 3). Because both of these RCTs only have been presented in abstracts at this time, it is difficult to measure the adequacy of surgical staging required for patients to be eligible for these trials. The Hatae et al. RCT25 compared the role of intravenous (i.v.) cisplatin (75 mg/m2) with either i.v. (500 mg/m2) or oral cyclophosphamide (500 mg/m2) in patients with Stage IA disease. There was no significant difference detected in the response rate or overall survival rate. The other RCT by Bell et al. compared 3 cycles of paclitaxel and carboplatin with 6 cycles of the same agents in 321 women with Stage I or II disease (70% of women had Stage I disease). The risk of disease recurrence was 33% lower among the women receiving 6 cycles compared with those receiving 3 cycles (HR, 0.762; 95% CI, 0.499–1.16; P = not significant). This difference is not statistically significant. However, there is no information provided in the abstract regarding the power of the study to detect a statistically significant difference between groups. Bell et al. did report that toxicity was substantially higher among the women receiving six cycles compared with those receiving three cycles.
The results presented below are a cumulative report of adverse events from all patients in the trial and not just patients with Stage I disease. None of the RCTs reported the scale used to measure adverse effects, a factor that interferes with the comparison of adverse event rates across studies.
Adjuvant radiotherapy versus no radiotherapy
Hreshchyshyn et al.18 (18) provided no information regarding complications. Dembo et al.19 reported 1 death, caused by surgery-related bowel complications, in the whole-abdominal radiotherapy (WAR) group (n = 75). The most frequently reported acute adverse events of radiotherapy were myelosuppression (66%), bowel cramps or diarrhea (64%), nausea or emesis (54%), and neutropenia (45%).
Adjuvant chemotherapy versus no adjuvant chemotherapy
Of the seven RCTs that compared adjuvant chemotherapy with no adjuvant chemotherapy, the only studies that provided details of complications were Gronroos et al.,17 who reported no deaths, and Young et al.,2 who reported mild to moderate hematologic and gastrointestinal toxicities in the melphalan group (n = 43). Young et al. reported one case of aplastic anemia. Some degree of myelosuppression occurred in 79% of patients, with 7 patients having severe thrombocytopenia. This study raised concern about the long-term sequelae of alkylating agents.
Adjuvant chemotherapy versus radiotherapy
Two of the RCTs that compared adjuvant chemotherapy with radiotherapy reported complication rates.17, 21 Chiara et al.21 reported that 71% of the women receiving adjuvant chemotherapy experienced Grade 3 or 4 nausea and emesis, compared with 14% of the women receiving WAR (P < 0.01). There were no Grade 3 or 4 toxicities reported among the women receiving adjuvant chemotherapy. However, 28% of women receiving WAR reported Grade 3 or 4 diarrhea, and 14% of women receiving WAR reported Grade 3 or 4 enteritis. Gronroos et al.17 reported that there were no deaths and that nausea, emesis, gastrointestinal complications, and hematologic toxicities were of anticipated severity, but provided no further details.
Adjuvant chemotherapy versus radioactive chromic phosphate
All four RCTs that compared adjuvant chemotherapy with 32P reported complication rates. Young et al.22 reported that among the women receiving adjuvant chemotherapy, the Grade 3 or 4 adverse effects included leukopenia (70%), granulocytopenia (65%), thrombocytopenia (9%), gastrointestinal effects (12%), renal toxicity (2%), oxotoxicity (1%), pulmonary emobolism (1%), allergic reaction (1%), and hematuria (1%).
In the trial by Bolis et al.,3 < 1% of the women receiving adjuvant chemotherapy experienced Grade 3 or 4 myelosuppression. Of the patients receiving 32P, one patient experienced intestinal occlusion. Bolis et al. reported that compliance in the 32P arm was low, because the isotope could not be implanted in 15 of 75 patients. In an earlier study by Young et al.,2 20% of the patients in the adjuvant chemotherapy arm reported severe myelosuppression. No severe adverse effects were reported in the 32P arm. In the study by Vergote et al.,23 there were no reports of Grade 3 or 4 adverse effects among the women receiving adjuvant chemotherapy. One woman receiving 32P had a lung embolism. No other severe adverse effects were reported.
Two different forms of adjuvant chemotherapy
Hatae et al.25 reported higher toxicities in the intravenous group, including myelosuppression, gastrointestinal sequelae, and hair loss. Given the side effect profile and the lack of difference in outcomes, Hatae et al. concluded that oral adjuvant therapy was superior. The RCT by Bell et al.24 reported substantially higher toxicity among the women receiving six cycles of chemotherapy compared with the women receiving three cycles of chemotherapy. More women receiving six cycles of chemotherapy had experienced anemia, granulocytopenia, and neurotoxicity than women who received three cycles.
Thirteen RCTs addressing the adjuvant management of patients with Stage I ovarian carcinomas were identified. There is evidence that adjuvant chemotherapy improves survival and decreases the disease recurrence rate. What remains to be defined is the population of Stage I patients who would most benefit from adjuvant therapy, the optimal chemotherapy regimen, and the dose and duration of treatment.
The completeness of surgical staging is a prognostic factor for ovarian carcinoma survival. Young et al.5 showed that in inadequately staged patients, a reoperation upstages approximately 25% of women, and 70% of these women are upstaged to Stage III. Thus, to determine the role of adjuvant therapy in patients with Stage I ovarian carcinoma, it is important that all the studies compare similar populations of patients. Optimal surgery for Stage I as defined by the EORTC is hysterectomy, bilateral salpingo-oophorectomy, peritoneal biopsies, omentectomy, and pelvic and paraaortic lymph node sampling.14 A subgroup analysis of the ACTION trial17 demonstrated that completeness of surgical staging, histologic type, and tumor grade were independent prognostic factors for progression-free and overall survival. The ACTION trial14 is the only study that provides information on optimally staged patients. A subgroup analysis suggests a lack of survival benefit for optimally staged patients who receive adjuvant therapy. This analysis was unplanned and the distribution of prognostic factors may be biased. The analysis was underpowered and, therefore, a specific recommendation for adjuvant chemotherapy in optimally staged patients is not possible. If the data for optimally and modified surgically staged patients are merged, again the merged data are underpowered to identify a survival or disease recurrence benefit for adjuvant treatment. The current sample size would only have a 20% power to detect the 3.3% difference seen in this group. In view of the overall benefit of chemotherapy and its modest toxicity, careful consideration should be given before omitting this therapy. Based on earlier work by Young et al.2 and the prognostic study by Vergote et al.,8 the standard of care for managing Stage IA and 1B and Grade 1 ovarian carcinoma has been optimal surgical staging with no adjuvant therapy. Thus, restaging surgery for women who fit this criterion, instead of adjuvant chemotherapy, is an option.
Adjuvant chemotherapy should include a platinum-based regimen. Currently, there is no consensus concerning the use of single versus combination treatment. The advocates of single-agent carboplatin would cite the ICON2 trial (n = 1526), which supported the use of single-agent platinum chemotherapy26 (carboplatin vs. CAP [cyclophosphamide, doxorubicin, and cisplatin]) in all stages of the disease. The 2074 patients in the ICON3 trial27 (paclitaxel plus carboplatin vs. carboplatin alone or CAP) represented all stages of ovarian carcinoma, and showed that carboplatin was as effective as the combination therapies. There is no adequately powered trial of carboplatin versus carboplatin and paclitaxel in patients with Stage I disease. Advocates of a platinum-paclitaxel combination are supported by two well conducted, North America–based randomized trials (Gynecologic Oncology Group [GOG] 11128 and EORTC/National Cancer Institute of Canada [NCIC] OV1029). These two trials showed that the platinum-paclitaxel combination provides a superior survival advantage compared with standard treatment in women with advanced-stage disease. By virtue of the Goldie hypothesis, this regimen should work even better in women with early-stage ovarian carcinoma.
In terms of the number of treatment cycles, Bell et al.'s adequately powered RCT24 of three versus six cycles of carboplatin-paclitaxel in Stage I disease showed no survival difference based on the number of treatment cycles. Young30 completed a randomized trial through the GOG that is published in abstract form. Of 457 women with Stage IA, Grade 3, Stage IB, Grade 3, Stage IC, or completely resected Stage II epithelial ovarian carcinoma, 70% had surgically staged Stage I disease. Again, three versus six cycles of carbotaxol was assessed. The risk of disease recurrence was lower for patients treated with 6 cycles of chemotherapy (RR, 0.672; 95% CI, 0.416–1.08). The toxicity rates were significantly greater in the arm with six cycles. The GOG is currently addressing the use of carbotaxol for 3 cycles with or without maintenance paclitaxel for 26 weeks in patients with completely resected Stage I and II ovarian carcinoma. Thus, clearer information on the number of treatment cycles will be forth coming shortly.
Limitations of the metaanalysis include the use of overall survival and disease recurrence rate data as opposed to patient-specific information. The use of overall results may obscure important findings that might be present. The lack of surgical staging for many of the women is another limitation—a treatment may be assumed incorrectly to be ineffective when actually the patient's stage was wrong. Because many of the studies combined Stage I patients with other stages and did not report on Stage I-specific results, information on the Stage I patients cannot be retrieved. Many of the trials with radiotherapy and 32P are dated. Therefore, the population was not adequately surgically staged. There are major methodologic flaws in the conduct of the 32P trials, including the process of randomization, sample size, and compliance with the therapeutic maneuvers. Ignoring the methodologic flaws, the significant side effect profiles alone of the radiotherapy and 32P arms make these treatment modalities unlikely candidates for future development. The ACTION-ICON experience has taught us that meaningful work in early stage ovarian carcinoma will require large cooperative research groups persisting together over long periods of time.
Based on the systematic review of the evidence from randomized trials, the following recommendations are deemed reasonable. Women with suspected ovarian carcinoma should receive surgical staging. We suggest that optimal surgical staging (as outlined by the EORTC guidelines) would allow outcomes of future studies that could be compared in terms of exposure to a standard operation. Women with Stage I ovarian carcinoma who have not received optimal surgical staging may actually have higher stage disease and should be offered platinum-based chemotherapy to decrease their disease recurrence rate and increase their survival. Women with optimally staged Stage I ovarian carcinoma, a pathology review by a gynecologic pathologist, and good prognostic factors (Stage IA or Stage Ib, Grade 1) have an excellent prognosis without adjuvant chemotherapy. A subgroup analysis from the ACTION trial suggests that for women with optimally staged Stage I disease, there is no benefit from adjuvant chemotherapy. Unfortunately, this analysis was underpowered, and definitive recommendations cannot be made. Further study is needed to assess the role of adjuvant chemotherapy in optimally staged Stage I ovarian carcinoma. Insufficient evidence exists to make a recommendation addressing the role of adjuvant pelvic radiotherapy, WAR, or IP 32P.
- 3Adjuvant treatment for early epithelial ovarian cancer: results of two randomised clinical trials comparing cisplatin to no further treatment or chromic phosphate (32P). G.I.C.O.G.: Gruppo Interregionale Collaborativo in Ginecologia Oncologica. Ann Oncol. 1995; 6: 887–893., , , et al.
- 13Randomized trial on adjuvant carboplatin versus no treatment in Stage I high risk ovarian cancer by the Nordic Ovarian Cancer Study Group [abstract]. Proc Am Soc Clin Oncol. 1997; 16: 352a., , , et al.
- 20ICON1: a randomised trial of immediate platinum-based chemotherapy against chemotherapy delayed until indicated in women with ovarian cancer [abstract]. Br J Cancer. 2001; 85 Suppl 1: 28..
- 24A randomized phase III trial of three versus six cycles of carboplatin and paclitaxel as adjuvant treatment in early stage ovarian epithelial carcinoma: A Gynecologic Oncology Group Study [abstract]. Proc Soc Gynecolog Oncol. 2003. Available from URL: http://sgo.org/meetings/abstract_detail.cfm?ID=1001, , , et al.
- 25Randomized trial on adjuvant IV chemotherapy CDDP+CPA versus po chemotherapy CPA for Stage IA ovarian cancer by the Japanese Gynecologic Oncology and Chemotherapy Study Group [abstract]. Proc Am Soc Clin Oncol. 1998; 17: 366a., , , et al.
- 27Randomised trial of paclitaxel (PTX) and carboplatin (CBDCA) versus a control arm of carboplatin or CAP (cyclophosphamide, doxorubicin and cisplatin): the Third International Collaborative Ovarian Neoplasm Study (ICON3) [abstract]. Proc Am Soc Clin Oncol. 2000; 20: A1500..
- 29Long-term follow-up confirms a survival advantage of the paclitaxel-cisplatin regimen over the cyclophosphamide-cisplatin combination in advanced ovarian cancer. Int J Gynecolog Cancer. 2003; 13(Suppl 2): 144–148., , , et al.
- 30Randomized clinical trial of adjuvant treatment of women with early (FIGO 1 and 2A high-risk) ovarian cancer [abstract]. Am Soc Clin Oncol. 1999; 257a., , , et al.