Annals of Oncology Advance Access published online on November 15, 2007
Annals of Oncology, doi:10.1093/annonc/mdm518
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2007 European Society for Medical Oncology
Why i.p. therapy cannot yet be considered as a standard of care for the first-line treatment of ovarian cancer: a systematic review
1 Centre for Clinical Pharmacology, Division of Medicine University College
2 Cancer Group
3 Meta-analysis Group, MRC Clinical Trials Unit, London
4 Cancer Research UK, UCL Cancer Trials Centre, London
5 Department of Respiratory Diseases, Health Protection Agency Centre for Infections, London, UK
* Correspondence to: Dr A. M. C. Swart, Cancer Group, MRC Clinical Trials Unit, 222 Euston Road, London, NW1 2DA, UK. Tel: +44 207 670 4700; Fax: +44 207 670 4818; E-mail: ams{at}ctu.mrc.ac.uk
 |
Abstract |
|---|
 Top Abstract introductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
A National Cancer Institute (NCI) clinical announcement recommended i.p. therapy for women with optimally debulked ovarian cancer. Its basis was a summary of eight randomised controlled trials and two systematic reviews, which appear to indicate benefit of i.p. therapy. However, the systematic reviews that inform the recommendations have been inappropriately presented and interpreted. The systematic reviews inappropriately pooled results from ‘confounded’ trials in which different drugs and different doses of drugs were given in the control and i.p. treatment arms. Therefore, it is not possible to assess which component of treatment is responsible for improving outcome. In addition, none of the trials use a control arm of the internationally accepted standard of care. Using just the unconfounded trials, indirect comparisons show that the magnitude of benefit observed when i.p. regimens are compared with older i.v. regimens [hazard ratio (HR) for overall survival (OS) 0.75; 95% confidence interval (CI) 0.60–0.92, P = 0.006] is smaller than the magnitude of benefit achieved with modern day standard of i.v. treatment compared with the same i.v. regimen used as control in the unconfounded i.p. trials (HR for OS 0.68; 95% CI 0.58–0.80, P < 0.001). A further difficulty is that the reviews cannot recommend an i.p. regimen for standard use. Drug-related toxicity and catheter complications that occur with i.p. therapy are considerable. The NCI recommendations have major implications for the treatment of women with ovarian cancer and for the next generation of clinical trials. We do not believe that the body of evidence currently available supports the recommendation that i.p. therapy should form part of routine care. The choice of treatment of women with newly diagnosed, optimally debulked, ovarian cancer, where therapy has the best chance of influencing OS, is too important to be left with this uncertainty. A clinical trial that investigates a practical and acceptable regimen which gives some or all chemotherapy by the i.p. route and compares this with standard i.v. chemotherapy should be a priority for those who wish to promote its use.
intraperitoneal therapy, ovarian cancer, systematic review
|
introduction |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
In January 2006, USA National Cancer Institute (NCI) issued a clinical announcement on the use of i.p. chemotherapy in the treatment of women with optimally debulked ovarian cancer [1]. The announcement was timed to coincide with the publication of a small but ‘positive’ randomised clinical trial assessing i.p. therapy [2]. Posted with the announcement was a review of the randomised trials of i.p. chemotherapy. Since 2006, there have been four further systematic reviews published which include an assessment of i.p. chemotherapy for the initial management of primary epithelial cancer [3–6]. We have appraised the methods used in these reviews and highlighted the methodological flaws which affect their validity and conclusions.
|
background |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
Worldwide there are >200 000 cases of ovarian cancer per year, just under half of which are reported from developed countries [7]. In North America, there are 25 000 cases per year with 16 000 deaths making ovarian cancer the seventh most common cancer and the leading cause of death from gynaecological cancer [8]. Nearly 70% of women present with disease which has spread beyond the ovary. Primary treatment is surgery with the aim of removing as much of the tumour as possible, with better survival rates reported for women left with small amounts of disease after surgery (optimal debulking) [9]. While there have been considerable improvements in outcome for women with ovarian cancer following the introduction of platinum-based chemotherapy in the 1980s, further advances in the last decade have been modest and the mortality rate for the large majority of women who present with advanced stages of this disease remains high. The major trials of platinum-based chemotherapy published over the past 15 years report that the median progression-free survival (PFS) for patients with advanced disease ranges between 14 and 18 months while the median overall survival (OS) lies between 26 and 40 months [10–13]. Two large trials (GOG111 [11] and OV10 [13]) demonstrated superiority of paclitaxel (Taxol; Bristol Myers Squibb) and cisplatin (CP) over cyclophosphamide and CP on response rate, PFS and OS. Two subsequent large trials (du Bois et al. [14] and Ozols et al. [15]) then showed no significant differences in efficacy between the combination of paclitaxel and CP compared with paclitaxel and carboplatin and also noted an advantage for the carboplatin combination in terms of administration, toxicity and quality of life. Data from all these trials set the standard for an internationally accepted reference arm of carboplatin and paclitaxel [16]. GOG182/ICON5 [17], one of the most recently completed trials, confirms that after 25 years, i.v. platinum remains the mainstay of treatment.
Ovarian cancer spreads primarily via the i.p. route and therefore chemotherapy given via the i.p. route could offer advantages in terms of local dose intensity with higher concentrations of chemotherapy being delivered directly to the tumour. Supporting this theory are the observations that peritoneal/plasma concentration ratios give an 
20-fold advantage for platinum drugs and a 1000-fold advantage for paclitaxel [18–21].
Intraperitoneal therapy has now been investigated in eight randomised clinical trials [2, 22–28] published over a 20-year period from 1986 to 2006 but the results of these trials individually were unconvincing and did not influence clinical practice. The systematic reviews [3–6] include pooled data from the eight trials and conclude that a combination of i.v. and i.p. administration of chemotherapy improves OS in women with optimally debulked epithelial ovarian cancer compared with i.v. administration alone. Systematic reviews, which may or may not contain a meta-analysis, summarise results of sufficiently similar studies which include similar patients and assess the same treatment in a related area. They can be a powerful tool to provide a basis for evidence-based medicine. They require a well-defined question, and robust inclusion and exclusion criteria are essential in order to avoid a wide range of biases. Critically, they depend on the availability of large-scale randomised trials which have examined the interventions of interest. Large-scale randomised trials are the gold standard when comparing new with existing treatments, but a well-conducted systematic review can ‘allow for objective appraisal and resolution of uncertainty’ [29]. On the basis of first of the reviews of i.p. therapy [1], USA NCI advised that women with optimally debulked ovarian cancer should be treated with a regimen containing CP (100 mg/m2) given i.p. and a taxane given i.v. only, or by an i.v. plus i.p. route. The review and recommendations are on the basis of an implicit pooling of the results with all the i.p. trials (whatever their design) considered to be equally relevant to the conclusion. A number of issues make undertaking a meta-analysis of all these trials into a single pooled estimate inappropriate. First, the majority of the trials used different drugs and/or different doses of drugs in the reference and experimental arms. Thus, the comparison of i.p. and i.v. therapy is not a straightforward comparison of route of administration of a chemotherapy regimen alone, it is confounded by two other major differences in treatment in the i.p. and i.v. arms (choice of drugs and the dose of platinum used). These differences (i.e. heterogeneity) give rise to confounding, a type of bias, which is usually eliminated in randomised trials by the process of randomisation. Secondly, and equally importantly, none of the trials used an appropriate control or reference arm, one of which would be comparable with a current standard of care. The conclusions cannot therefore be applied to current practice. In this paper, we present what we consider to be a more appropriate analysis and consequently more appropriate conclusions for the present day.
|
patients and methods |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
Recently published trials were found by carrying out searches in Medline using established search strategies [30] up to July 2007. The primary outcome measure was OS with the secondary outcome measures of PFS and summaries of toxicity data. Where possible, analyses were of all reported randomised patients.
For meta-analyses of time-to-event outcomes, the most appropriate statistic is the hazard ratio (HR). If provided in a trial report, the HR and associated variances were used directly in the meta-analysis. Alternatively, using the methods described by Parmar et al. [31], they were estimated indirectly from other summary statistics [confidence intervals (CIs), P values, total number of events] or from data extracted from published Kaplan–Meier curves [32]. In the latter method, the numbers at risk are adjusted (reduced), where appropriate, to allow for immature follow-up. Where feasible, a number of methods were used to indirectly estimate the trial HR to check its reliability. The estimated log HR and variance were then combined across all trials using the fixed-effect model to give a pooled HR [31]. This represents the overall risk of an event on i.v. chemotherapy versus i.p. chemotherapy.
OS was estimated primarily using the fixed-effect model. Absolute differences at relevant time points were calculated by multiplying the log of the baseline event rate for patients not allocated to i.p. chemotherapy by the HR. This value was then subtracted from the baseline event rate to give a percentage improvement or detriment in the outcome. CIs for absolute differences were similarly calculated from the baseline event rate and the HR at the 99% CI boundary values. Toxicity was examined qualitatively and quantitatively summarised by estimates of proportions of patients who received all planned cycles of chemotherapy.
|
results |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
Eight randomised controlled trials (RCTs) [2, 22–28] (including 1826 patients) testing i.p. therapy were identified as eligible for inclusion in the systematic review. A summary of the eligible trials, publication dates, outcome measures and the wide range of regimens used both in reference (i.v. chemotherapy) and in experimental (i.p. chemotherapy) arms is shown in Table 1. Of the eight RCTs, limited information in the publication did not allow inclusion of any data from two small trials [23, 28]. Data on OS could be extracted from six RCTs [2, 22, 24–27] (1716 patients) representing 94% of all patients entered into the relevant RCTs. Data on PFS could be extracted from four RCTs [2, 24, 26, 27] (1052 patients) representing 58% of all patients entered into the relevant RCTs.
|
The regimens used in the reference arms of the trials reflect changing practice over the 20-year period that the trials were conducted. The third International Gynecologic Cancer Intergroup Ovarian Cancer Consensus Conference in 2004 [16] recommended that the control arm in international clinical trials of systemic treatment of ovarian cancer should be six cycles of combination i.v. chemotherapy with carboplatin [area under the curve (AUC) 5 to AUC 7.5] and paclitaxel (175 mg/m2 given as a 3-h infusion) administered every 3 weeks. While all the i.p. regimens did include platinum-based chemotherapy, only two trials used regimens that would now be considered close to the internationally accepted standard of care of carboplatin and paclitaxel [2, 26].
The experimental arms of the randomized trials vary widely with respect to the chemotherapeutic agents used and the dose of platinum and other drugs. All contained platinum-based chemotherapy; the i.p. platinum compound in seven of the trials was CP, only one trial used i.p. carboplatin. Two of the trials used i.p. platinum in combination with i.v. paclitaxel.
The rational behind considering the trials as one homogenous group is that for all or the majority of improvement associated with the experimental arm is due to therapy being given by the i.p. therapy route and not to any other component of therapy (such as an increased dose of platinum in the i.p. arm). With the aim of reducing clinical heterogeneity, we separated the trials into four groups according to similarities between the reference and experimental arms with respect to the chemotherapy drugs used, the dose of platinum and the dose of paclitaxel (Table 1).
We also classified groups as unconfounded or confounded depending on whether more than one major component of treatment (choice of drugs and dose of platinum used) was changed between the reference and experimental arms. In the unconfounded trials, the only difference between the experimental and reference arms is the route of administration. In the confounded trials, it is not possible to assess which component of treatment is responsible for any difference in outcome.
Figures 1 and 2 show HRs for OS (Figure 1) and PFS (Figure 2) with 95% (inner ticks) and 99% (outer ticks) CIs for individual i.p. trials. Trials are classified (as in Table 1) according to whether or not they are confounded and according to similarities in the reference and experimental arms with respect to the chemotherapy drugs used, the dose of platinum and the dose of paclitaxel. Information on OS is more complete and less biased in these open-label trials than information on PFS. Summary statistics for each subgroup are presented only if more than one trial is included in the group of trials.
|
|
In Figures 1 and 2, group A (i) shows the trials with the only unconfounded comparisons, where the only change between the experimental and reference arms of the trial is the route of administration of platinum, i.p. or i.v. The combined HR for the two trials with data on OS is 0.75 (95% CI 0.60–0.92, P = 0.006). For PFS, data were only available for one trial. In both trials, the reference arm was i.v. cyclophosphamide plus CP.
The two trials in group A (ii) gave the same drugs in both reference and experimental arms but gave higher doses of platinum in the i.p. arms compared with the i.v. arms. A higher platinum dose i.p. alone seems to yield only a very modest possible improvement in OS, the combined result of these trials for OS was HR 0.87 (95% CI 0.69–1.09, P = 0.227). PFS data were only available for one trial.
In group A (iii), doses of both platinum and paclitaxel are different in the two arms with a higher dose of both platinum and paclitaxel given in the i.p. therapy arm. There is only one trial in this group, but it is the most recently reported publication and was the motivation for the NCI clinical announcement. HR for OS was 0.75 (95% CI 0.58–0.97, P = 0.028) and HR for PFS 0.80 (95% CI 0.64–1.00, P = 0.50). There are other methodological difficulties with this trial including the exclusion of more randomised patients and more patients lost to follow-up in the i.p. arm. There is also a larger improvement in PFS compared with OS which is counterintuitive and a suggestion that the estimates of median OS may be unreliable at the extremes of the Kaplan–Meier plot because there are relatively few patients at risk at this time. In addition, there are questions about feasibility and acceptability of a treatment in which 42% of patients received less than two cycles of i.p. therapy.
In the two trials in group A (iv), different drugs as well as different doses of platinum were included in the reference and experimental arms. Outcome measures were only available for one of the two trials in this group.
The overall results shown in Figures 1 and 2, across all trials using i.p. therapy (whether unconfounded or confounded) give an overall summary result of HR 0.79 (95% CI 0.69–0.90, P = 0.003) and HR 0.81 (95% CI 0.70–0.93, P < 0.001) for OS and PFS, respectively. It is not possible to interpret the meaning of the summary statistic with respect to the improvement in outcome that could be obtained with the introduction of i.p. therapy against current international standards because of the wide range of reference and experimental arms across the trials. In none of the trials did the i.p. therapy arm receive less platinum, less paclitaxel or a simpler regimen. In fact, in all but the three unconfounded trials in group A (i), patients receiving i.p. therapy received more treatment than the reference arm. The only trials which can be considered to give a reliable summary of the relative value of i.p. therapy over i.v. therapy are the three trials (of which we have data from the two largest for OS) in the unconfounded comparison. The estimate of the effect of i.p. therapy on OS (HR 0.75; 95% CI 0.60–0.92) obtained for the unconfounded trials is similar to the overall (uninterpretable) estimate, but there are other reasons why i.p. therapy can still not be considered as an appropriate standard therapy.
choice of control arm and relevance of the results of i.p. trials to current practice
The most important observation in our review is that none of the trials which consider i.p. therapy compare it against the international standard reference arm of carboplatin and paclitaxel or other widely used regimens (e.g. carboplatin alone). It is therefore not possible to say how the variety of i.p. regimens tested would compare with these regimens. Since no such direct comparison is possible, the only way to consider this issue is to make an indirect comparison with appropriate trials. To allow indirect comparison of the results of the i.p. trials with current standard treatments, we summarise the results of the two trials which compare i.v. therapy with cyclophosphamide and CP with i.v. paclitaxel and CP (Table 2 and Figures 1 and 2—group B). The reference arms of these pivotal trials (GOG111 [11] and OV-10 [13]) which investigated the efficacy of platinum plus paclitaxel were CP 75 mg/m2 and cyclophosphamide 750 mg/m2. The experimental arm of GOG111 was CP 75 mg/m2 and paclitaxel 135 mg/m2 >24 h and that of OV10 was CP 75 mg/m2 and paclitaxel 175 mg/m2 >3 h. Both trials gave six cycles every 3 weeks. Although such indirect comparisons should be treated with caution, the magnitude of improvement in OS seen with i.v. CP and paclitaxel compared with i.v. CP in these two trials (HR 0.68; 95% CI 0.58–0.80) is actually slightly larger than the magnitude of improvement seen with the unconfounded i.p. trials where the reference arms were i.v. CP and i.v. CP plus doxorubicin (which has been shown to be approximately equivalent to i.v. CP [23]) (HR 0.75; 95% CI 0.60–0.92).
|
For PFS, data are more limited for the unconfounded comparison of i.p. therapy and so the overall estimate (HR 0.81; 95% CI 0.70–0.93) is used. The magnitude of improvement in PFS with i.v. platinum and paclitaxel compared with CP is also larger than the improvement seen with i.p. therapy (HR 0.71; 95% CI 0.62–0.81, P = 0.001).
The HR for OS with i.p. therapy of 0.75 would translate into an improvement in median OS of 12 months (for optimally debulked patients), but with an overall HR of 0.68 for the i.v. CP plus paclitaxel trials, a larger improvement of 16 months would be expected for similar patients.
toxicity
Summary measures of toxicity are difficult to obtain and even when reported are difficult to interpret. This is because few trials reported toxicity in detail, different scales of measurement were used and information was frequently combined by site or grade. Interpretation of comparisons across trials is difficult as toxicity is highly dependent on chemotherapy regimen and dose as well as method of administration. Consideration was therefore given to a number of approaches. The method used in the Cochrane review [3] calculated a pooled odds ratio for types of reported toxicity classified in broad common toxicity criteria categories. Pooled estimates of toxicity favoured i.v. therapy in the adverse outcomes classified as fever, fatigue, gastrointestinal, infection, metabolic and pain. No pooled effect favoured i.p. therapy. There was considerable heterogeneity as a result of the different regimens that were used. Data published with the NCI alert included a summary of proportions of patients reported with different types and grades of toxicity. Much of the reported toxicity for i.v. therapy is haematological which is easily managed and has a limited impact on quality of life. Interestingly in the one trial where a high proportion of patients in the i.v. therapy reported abdominal symptoms, closer examination of the trial report revealed that 80% of patients in the i.v. group had an i.p. catheter placed for peritoneal fluid sampling.
Compliance with planned chemotherapy can be regarded as a composite measure of toxicity and efficacy as trial treatment is stopped if patients are failing therapy which can be due to either a toxicity or disease progression or a combination of the two. This outcome measure was reported consistently for the majority of trials and is therefore the approach we used. The proportion of patients receiving all cycles of planned chemotherapy is summarised in Table 3. Overall, compliance and deliverability of i.p. regimens are generally worse than i.v. reference regimens in the i.p. trials and worse than i.v. CP and paclitaxel (85%) in the taxane trials.
|
For the i.p. trials, data were available for two of the three unconfounded trials [group A (i)]. In one trial, the same proportion of patients in both arms completed all cycles (58%) and in the other a higher proportion of patients allocated to i.v. received all cycles (96%) than those allocated to i.p. (64%). In the confounded trials which gave the same chemotherapy drugs but different dose of platinum [group A (ii)], higher proportions of patients in both trials allocated to i.v. chemotherapy received all six cycles.
In the confounded trial where different doses of platinum and paclitaxel were given [group A (iii)], only 42% of patients completed all cycles of the i.p. regimen compared with 83% of patients assigned to i.v. This was the only trial where patient-assessed outcomes were measured. Quality of life measured by Functional Assessment of Cancer Therapy-Ovarian module (adjusted for baseline imbalances) was worse in patients receiving i.p. therapy before cycle 4 and 3–6 weeks after completion of treatment although at 12 months after the sixth cycle, no significant difference was detected. Toxicity was only reported in one of the trials in group A (iv), the trials where very different regimens were given. Gastrointestinal toxicity was very high in both groups but 78% of patients in the i.v. group assigned to i.v. therapy had peritoneal catheters in situ which were used for peritoneal cytology sampling and two of these patients developed peritoneal-vaginal fistulas.
relative toxicity of i.p. therapy and i.v. carboplatin and paclitaxel
No trials compare i.p. therapy with the current standard of i.v. carboplatin and paclitaxel. Indirect comparisons from Table 2, however, show that overall, giving some drugs via the i.p. route (platinum alone or platinum plus paclitaxel), therapy is more toxic and less deliverable than therapy with i.v. platinum and paclitaxel. Catheters cause abdominal pain, nausea and vomiting and there is an increased risk of infection and fever [26]. Intraperitoneal paclitaxel has been associated with increased haematological, metabolic and neurological toxicity [2].
|
conclusions |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
The reviews of i.p. therapy and the recommendations that they have generated have major implications for the practical management for a large proportion of women with ovarian cancer who are having to make treatment decisions on the basis of incomplete data. They also have important implications for the design of the next generation of clinical trials which will set the standard treatments for the future. Before addressing the challenge of incorporating the newer angiogenesis inhibitors, with potential problems of wound healing and bowel perforations, into regimens which include i.p. therapy, it is very important that we know that i.p. regimens are genuinely better than standard i.v. therapy. The reviews that have been carried out have indicated that i.p. therapy should be considered as part of the standard approach to treating optimally debulked International Federation of Gynecology and Obstetrics (FIGO) stage III ovarian cancer and the NCI recommendation is that ‘after primary surgery, women with optimally debulked FIGO stage III ovarian cancer should be counselled about the clinical benefit associated with combined IV and IP administration of chemotherapy’. The recommendations are included on the NCI's Web site with a designation of 1iiA for the level of evidence [33]. This level of evidence is designated for randomised (level 1), unblinded (ii) trials with mortality as an end point (A). The designation does not take into account deficiencies in both the control and the experimental arms of the i.p. trials. A closer and more critical and appropriate consideration of the data, however, shows that the apparently ‘clear’ body of evidence in favour of i.p. therapy is not as strong as indicated. Compared with current standard of care, the benefits of i.p. therapy may have been overestimated. We believe that it is impossible at this stage to give women accurate information on the risks and benefits of i.p. therapy compared with the modern standard of care. The clinical heterogeneity of the regimens in the trials has led the NCI to recommend as standard practice i.p. CP (100 mg/m2) and a taxane (whether given by an i.v. only or i.v. and i.p.). It is not clear whether giving a taxane i.v. only or a taxane i.v. and i.p. is equivalent in terms of toxicity and efficacy, or how the choice should be made for individual patients. The regimen in the i.p. arm of GOG172 [2] included i.p. CP and i.v. and i.p. paclitaxel. The only other trial which included i.p. CP and i.v. paclitaxel also included two cycles of high-dose i.v. carboplatin in the experimental arm. Thus, it is not clear what the standard i.p. regimen should be. Any i.p. CP/paclitaxel regimen that deviates from that used in GOG172 cannot be said to have been assessed in any trial. The NCI recommendation is a direct consequence of the lack of clarity in the design of the individual trials and the inappropriate presentation and interpretation of systematic reviews. No new treatment would be approved by regulatory authorities or should be adopted as standard clinical practice on the basis of this evidence. Allied to this, i.p. regimens are considerably more toxic than the current standard of care and indirect comparisons strongly indicate that they may actually be less effective than a current standard of care of i.v. platinum and paclitaxel. Others have attempted to use indirect approaches to assess the benefit of the i.p. regimen used in GOG172 and concluded that if the i.v. therapy reference arm in GOG172 had been treated with i.v. carboplatin and paclitaxel instead of i.v. CP and paclitaxel, the difference in OS may have been considerably reduced (8.2 months, instead of 15.9 months) [34].
Consequently, we do not believe that i.p. therapy should form part of the routine care of women with ovarian cancer. Until problems with drug-related toxicity and catheter complications are resolved, even proposals for further randomised trials of i.p. therapy against modern standard of care could be considered as speculative. The choice of treatment of women with newly diagnosed, optimally debulked, ovarian cancer, where therapy has the best chance influencing OS, is too important to be left with these uncertainties.
|
funding |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
MRC Clinical Trials Unit to A.M.S., S.B., P.M. and M.P.; University College London to J. L.
Received for publication October 17, 2006. Revision received August 14, 2007. Accepted for publication October 9, 2007.
|
References |
|---|
|
TopAbstractintroductionbackgroundpatients and methodsresultsconclusionsfundingReferences |
|---|
1. NCI Clinical Announcement. Intraperitoneal chemotherapy for ovarian cancer [Web Page]. http://ctep.cancer.gov/highlights/ovarian.html (30 October 2007, data last accessed).
2. Armstrong DK, Bundy B, Wenzel L, et al. Intraperitoneal cisplatin and paclitaxel in ovarian cancer. N Engl J Med (2006) 354:34–43.
3. Jaaback K, Johnson N. Intraperitoneal chemotherapy for the initial management of primary epithelial ovarian cancer. Cochrane Database Syst Rev (2006) 1. CD005340. doi: 10.1002/14651858.CD005340.pub2.
4. Kyrgiou M, Salanti G, Pavlidis N, et al. Survival benefits with diverse chemotherapy regimens for ovarian cancer: meta-analysis of multiple treatments. J Natl Cancer Inst (2006) 98(22):1655–1663.
5. Elit L, Oliver TK, Covens A, et al. Intraperitoneal chemotherapy in the first-line treatment of women with stage III epithelial ovarian cancer: a systematic review with metaanalyses. Cancer (2007) 109(4):692–702.[CrossRef][Medline]
6. Fung-Kee-Fung M, Provencher D, Rosen B, et al. Intraperitoneal chemotherapy for patients with advanced ovarian cancer: a review of the evidence and standards for the delivery of care. Gynecol Oncol (2007) 105(3):747–756.[CrossRef][Web of Science][Medline]
7. Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. IARC CancerBase No. 5, version 2, IARC Press, Lyon: 2004.
8. U.S. Cancer Statistics Working Group. United States Cancer Statistics: 1999–2002 Incidence and Mortality Web based Report Version (2005) Atlanta: Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute.
9. Bristow RE, Tomacruz RS, Armstrong DK, et al. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: a meta-analysis. J Clin Oncol (2002) 20(5):1248–1259.
10. The International Collaborative Ovarian Neoplasm (ICON) Group. Paclitaxel plus carboplatin versus standard chemotherapy with either single-agent carboplatin or cyclophosphamide, doxorubicin, and cisplatin in women with ovarian cancer: the ICON3 randomised trial. Lancet (2002) 360(9332):505–515.[CrossRef][Web of Science][Medline]
11. McGuire WP, Hoskins WJ, Brady MF, et al. Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med (1996) 334:1–6.
12. Muggia FM, Braly PS, Brady MF, et al. Phase III randomized study of cisplatin versus paclitaxel versus cisplatin and paclitaxel in patients with suboptimal stage III or IV ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol (2000) 18:106.
13. Piccart MJ, Bertelsen K, James K, et al. Randomized intergroup trial of cisplatin–paclitaxel versus cisplatin–cyclophosphamide in women with advanced epithelial ovarian cancer: three-year results. J Natl Cancer Inst (2000) 92:699–708.
14. du Bois A, Luck HJ, Meier W, et al. A randomized clinical trial of cisplatin/paclitaxel versus carboplatin/paclitaxel as first-line treatment of ovarian cancer. J Natl Cancer Inst (2003) 95(17):1320–1329.
15. Ozols RF, Bundy BN, Greer BE, et al. Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol (2003) 21(17):3194–3200.
16. du Bois A, Quinn M, Thigpen T, et al. 2004 consensus statements on the management of ovarian cancer: final document of the 3rd International Gynecologic Cancer Intergroup Ovarian Cancer Consensus Conference (GCIG OCCC 2004). Ann Oncol (2005) 16(Suppl 1):viii7–viii12.
17. Bookman MA, for the Gynecologic Cancer InterGroup(GCIG). GOG0182-ICON5: 5-arm phase III randomized trial of paclitaxel (P) and carboplatin (C) vs combinations with gemcitabine (G), PEG-lipososomal doxorubicin (D), or topotecan (T) in patients (pts) with advanced-stage epithelial ovarian (EOC) or primary peritoneal (PPC) carcinoma. J Clin Oncol (2006) 24(18S):5002.
18. Goel R, Cleary SM, Horton C, et al. Effect of sodium thiosulfate on the pharmacokinetics and toxicity of cisplatin. J Natl Cancer Inst (1989) 81:1552–1560.
19. Miyagi Y, Fujiwara K, Kigawa J, et al. Intraperitoneal carboplatin infusion may be a pharmacologically more reasonable route than intravenous administration as a systemic chemotherapy. A comparative pharmacokinetic analysis of platinum using a new mathematical model after intraperitoneal vs. intravenous infusion of carboplatin. A Sankai Gynecology Study Group (SGSG) study. Gynecol Oncol (2005) 99(3):591–596.[CrossRef][Web of Science][Medline]
20. Markman M. Intraperitoneal drug delivery of antineoplastics. Drugs (2001) 61:1057–1065.[CrossRef][Web of Science][Medline]
21. Vermorken JB. The role of intraperitoneal chemotherapy in epithelial ovarian cancer. Int J Gynecol Oncol (2000) 10(Suppl 1):26–32.[CrossRef]
22. Alberts DS, Liu PY, Hannigan EV, et al. Intraperitoneal cisplatin plus intravenous cyclophosphamide versus intravenous cisplatin plus intravenous cyclophosphamide for stage III ovarian cancer. N Engl J Med (1996) 335:1950–1955.
23. Polyzos A, Tsavaris N, Kosmas C, et al. A comparative study of intraperitoneal carboplatin versus intravenous carboplatin with intravenous cyclophosphamide in both arms as initial chemotherapy for stage III ovarian cancer. Oncology (1999) 56(4):291–296.[CrossRef][Web of Science][Medline]
24. Gadducci A, Carnino F, Chiara S, et al. Intraperitoneal versus intravenous cisplatin in combination with intravenous cyclophosphamide and epidoxorubicin in optimally cytoreduced advanced epithelial ovarian cancer: a randomized trial of the Gruppo Oncologico Nord-Ovest. Gynecol Oncol (2000) 76(2):157–162.[CrossRef][Web of Science][Medline]
25. Yen MS, Juang CM, Lai CR, et al. Intraperitoneal cisplatin-based chemotherapy vs. intravenous cisplatin-based chemotherapy for stage III optimally cytoreduced epithelial ovarian cancer. Int J Gynaecol Obstet (2001) 72(1):55–60.[CrossRef][Medline]
26. Markman M, Bundy BN, Alberts DS, et al. Phase III trial of standard-dose intravenous cisplatin plus paclitaxel versus moderately high-dose carboplatin followed by intravenous paclitaxel and intraperitoneal cisplatin in small-volume stage III ovarian carcinoma: an intergroup study of the Gynecologic Oncology Group, Southwestern Oncology Group, and Eastern Cooperative Oncology Group. J Clin Oncol (2001) 19(4):1001–1007.
27. Kirmani S, Braly PS, McClay EF, et al. A comparison of intravenous versus intraperitoneal chemotherapy for the initial treatment of ovarian cancer. Gynecol Oncol (1994) 54(3):338–344.[CrossRef][Web of Science][Medline]
28. Zylberberg B, Ravina JH, Salat-Baroux J, et al. Polychemotherapy of ovarian cancer via combined intravenous and intraperitoneal routes. Technic and preliminary results. J Gynecol, Obstet Biol Reprod (1986) 15(5):671–676.[Medline]
29. Egger M, Smith GD. Meta-analysis: potentials and promise. Br Med J (1997) 315:1371–1374.
30. Lefebvre C, Clarke MJ. Identifying randomised trials. In: Systematic Reviews in Healthcare—Egger M, Davy-Smith G, Altman DG, eds. (2001) 2nd edition. London: BMJ Publishing Group. 69–87.
31. Parmar MKB, Torri V, Stewart L. Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med (1998) 17:2815–2834.[CrossRef][Web of Science][Medline]
32. Yusuf S, Peto R, Lewis JA, et al. Beta blockade during and after myocardial infarction: an overview of the randomized trials. Prog Cardiovasc Dis (1985) 27:335–371.[Web of Science][Medline]
33. National Cancer Institute (PDQ). Levels of Evidence for Adult Cancer Treatment Studies (2007) http://www.cancer.gov/cancertopics/pdq/levels-evidence-adult-treatment (30 October 2007, data last accessed).
34. Ozols RF, Bookman MA, Young RC. Intraperitoneal chemotherapy for ovarian cancer. N Engl J Med (2006) 354(15):1641–1642.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||