In summary, the controversies surrounding the advisable limits of surgical radicality, and relative importance of maximal surgical effort and biological behaviour of the tumour in the final outcome of patients are supposed to persist unless more prospective randomized data can be accumulated. Nevertheless it could be stated that the standard of care of patient with stage III ovarian cancer should include an attempt at the optimal cytoreduction of disease based on prospective data reported by van der Burg regarding interval debulking.
1.3. Preoperative evaluation of cytoreducible disease
Despite advances in intraoperative and postoperative care, with development of specific surgical procedures to maximize the optimal cytoreduction rate and indirectly pursue a survival advantage, a substantial proportion of patients do not present enough clinical condition to undergo a radical surgery. In this subset of patients, extensive surgery may result in serious morbidity, which occasionally precludes or significantly delays the initiation of postoperative chemotherapy. In order to establish a preoperative criteria predictive of optimall cytoreduction, Nelson et al, carried out a retrospective study in which they analysed the medical records and preoperative abdominopelvic CT scans of 51 patients who underwent primary CRS [46]. Adopting the criteria outlined in table 10 (see appendix), they verified that the CT findings accurately predicted surgical outcome with a sensitivity of 92.3% and specificity of 79.3%. The positive predicitve value was 67% and the negative predictive value was 96%.
1.4. First line chemotherapy
Following the surgical staging systemic chemotherapy with paclitaxel (Taxol) plus cisplatin or carboplatin is commonly used [10,47-52]. Platinum based chemotherapy regimens have been shown to produce higher response rates and, in some studies, have produced a statistically significant survival advantage compared to drug regimens without platinum. A meta-analysis addressing this comparison in 1,400 patients revealed a strong trend in favour of platinum-containing combinations with respect to response, but not survival [53]. Single-agent cisplatin seems to be equally effective and less toxic than platinum-containing combinations. This meta-analysis, however, suggests that the combination confers a 15% survival advantage out to the eighth year over the use of single-agent platinum.
The Gynecologic Oncology Group (GOG) has carried out a randomized, phase III clinical trial comparing paclitaxel and cisplatin (TP) with cyclophosphamide and cisplatin (CP) in sub-optimally debulked (>1 cm residual mass) stage III/IV patients who had no prior chemotherapy [10]. There was a statistically significant higher clinical response rate in the TP arm as compared to CP arm (73% vs 60%). Median survival was also significantly better in the TP arm (24 months vs 38 months; P=0.001). Differences in surgically documented complete response were not statistically significant (20% for CP and 26% for TP). Further, in a European-Canadian trial carried out in patients with both optimally and sub-optimally debulked tumours the relapse-free and overall survival advantages of TP over CP were confirmed [54] and were seen in both groups of patients (i.e., those with large-bulk and low-bulk disease).
Another randomized trial that compared paclitaxel (135 mg/m2) combined with cisplatin (75 mg/m2) with paclitaxel monotherapy (200 mg/m2) given over 24 hours or cisplatin monotherapy (100 mg/m2) found equivalent survival for all 3 regimens. Paclitaxel monotherapy was inferior in response rate and progression-free interval while cisplatin monotherapy was associated with significant neurotoxicity [55].
In the European-Canadian study, paclitaxel was administered over a more convenient 3 hours at a dose of 175 mg/m2. This dose and schedule were previously found to be equivalent to a dose of 135 mg/m2 over 24 hours in terms of response and disease-free survival in patients with advanced disease [56]. Because the 3-hour regimen of paclitaxel is associated with substantial neurotoxicity when given with cisplatin [54], carboplatin has frequently been substituted for cisplatin in this regimen. Clinical trials assessing the efficacy of this substitution are in progress. Initial reports indicate no loss of efficacy [57], and in a meta-analysis, carboplatin was found to be as effective as cisplatin alone and in combination. Thus, many investigators consider the 3-hour regimen of paclitaxel plus carboplatin (AUC 5-7) to be an acceptable alternative to the GOG regimen of paclitaxel and cisplatin as the preferred initial chemotherapy for patients with stage III/IV ovarian cancer.
1.5. Second line therapies
When previous effective drug combinations fail, there is virtually no chance of inducing a significant response with second-line treatment. A partial response and control of malignant effusions can be achieved occasionally and are usually associated with a short survival [2].
Several groups of investigators have studied innovative forms of second-line or salvage therapy, such as new drugs or high-dose chemotherapy with autologous bone marrow support or intraperitoneal chemotherapy. Available data (sees Appendix, Table 1) shows a somewhat higher response rate to carboplatin+ifosfamide and intraperitoneal chemotherapy in platinum-sensitive patients compared to platinum-resistant ones. According to different studies [47,58-72], the response rate to salvage systemic chemotherapies in the latter subgroup never exceeds 26.8%, and median survival ranges from 8.8 months to 15 months.
Following high-dose chemotherapy, a fairly high response rate has been observed that, however, did not reflect in a higher median survival, at least in the one study performed on an adequate series of patients [61]. Conversely, in platinum-resistant patients treated by intraperitoneal chemotherapy, a low response rate was observed, whereas the median survival was about two-three times longer than that observed following any other treatment.
1.6. Secondary CRS
Besides the second line antiblastic therapies presented above, another treatment option for chemoresistant or recurrent disease is the secondary CRS. Unfortunately, the benefits of this alternative have not been clearly established, yet. The only prospective randomised clinical trial evaluating the impact of secondary CRS in advanced ovarian cancer is still ongoing [73]. To separate patients based on differences in biologic tumour behaviour, secondary operations for ovarian cancer can be classified into four clinical settings [74].
a) Progressive disease: patients who have evidence of clinical disease progression while receiving first line therapy;
b) Recurrent disease: those patients who enjoy a prolonged clinical disease free interval (>6 months) after completing primary therapy, and then develop recurrent disease;
c) Second-look operation: patients who are clinically and radiologically free of disease after primary surgery and first-line chemotherapy, who are found to have macroscopic disease at second-look operation (SLO);
d) Interval debulking: patients with bulky, unresectable tumour discovered at initial surgery after neoadjuvant chemotherapy.
There is limited data on secondary CRS in patients with progressive disease. Many oncologists present reservations about subjecting these patients with a uniformly poor prognosis to the additional morbidity associated with a major surgical procedure. Morris et al. at the MDACC performed secondary CRS prior to completion of first line chemotherapy on 31 patients with tumour progression and 2 patients with stable disease [75]. Optimal debulking (<2cm) was accomplished in 55%. Median survival for <1cm residual disease (RD) patients was 12 months, compared with 7.8 months for patients with larger RD (p<0.03). This apparent survival advantage disappeared after 2 years of follow-up. Operative morbidity occurred in 24% of patients, with 22 of 31 patients requiring a small or large bowel resection. The authors concluded that in patients with stable disease unresponsive to first line therapy, secondary CRS is associated with unacceptable surgical morbidity considering the limited survival benefit. A second report from Michel et al. described similar outcomes for 77 patients undergoing additional surgery for progressive ovarian cancer prior to completing first-line chemotherapy [76]. Median survival for the 32 optimally resected (<2cm) patients was no different from that of patients left with larger residual tumour (12 moths each). Based on these studies, there is little evidence to support a role for secondary CRS in the setting of progressive disease resistant to primary chemotherapy.
Second look operation was originally defined as a systematic surgical reexploration of patients with ovarian cancer who are clinically free of disease after completing a planned treatment program of primary CRS and chemotherapy. Despite the clinical and radiologic absence of disease, approximately 40% of patients reaching SLO will have macroscopic disease detected at the time of surgery [77]. There are no consistent data showing that therapeutic decisions based upon results of this procedure alter (or does not) outcomes for the patient. In a large non-randomised trial, there was no survival advantage in patients who received a SLO as compared to those who did not [77] and the only randomised trial albeit statistically underpowered, was negative [78].
The current data from literature (see Appendix, Table 2) show somewhat more favourable results in terms of feasibility and survival impact when secondary CRS is performed in a SLO setting rather than in progressive disease. Cytoreductive attempts in SLO setting have often proven to be successful. Approximately 40% of patients with macroscopic disease are able to undergo complete resection to be left with only microscopic RD, 30% are able to be partially debulked and left with minimal RD (<2cm), and 30% are left with bulky RD [74] (see Appendix, Table 2) [79-91].
Some difficulties surround the interpretation of data outlined in Table 2. First, most of studies have been retrospective and have included heterogeneous study populations. Second, selection of patients for secondary operations has usually been dependent on the individual physicians’ bias. Third, several studies have been conducted over long periods, and the chemotherapy regimens used have varied considerably. Fourth, not all of these investigations were specifically designed to evaluate effects of CRS at SLO.
Nevertheless, the majority demonstrate some survival advantage for patients who can be debulked to microscopic RD. Cytoreduction that leaves patients with small macroscopic disease may provide some survival benefit. As noted, the role of SLO in the standard management of patients with advanced ovarian cancer is still under investigation because the overall survival benefit of the procedure has not been demonstrated.
On balance, when SLO is performed and macroscopic residual tumour is detected, the weight of evidence suggests that the removal of all macroscopic disease should be accomplished if technically feasible. Because the presence of bulky residual tumour after SLO portends a uniformly poor prognosis, attempts at resection are probably not indicated unless the patient can be left with an optimal RD.
1.7. LRT for peritoneal surface malignancies
Loco regional therapy in the current study will mean the combination of cytoreduction and intraperitoneal hyperthermic perfusion (IPHP). This new treatment strategy was conceived for advanced gastric cancer and peritoneal carcinomatosis [92-95], and a slight increase in morbidity in patients treated by this aggressive loco regional approach has been reported [96,97,98]. Recently, it has also been considered as second-line and salvage therapy in Phase I/II clinical studies in the management of advanced ovarian cancer with some promising results [99-102,129].
For the CRS the peritonectomy technique is performed. Firstly described by Sugarbaker [103], the procedure encompasses 6 different visceral and/or parietal peritoneum resections and is a useful resource for the achievement of minimal residual disease. After the cytoreduction the patient is submitted to the second phase of procedure, which is IPHP. Before describing the rationale of IPHP, some aspects regarding the intraperitoneal drug delivery under normothermic condition in ovarian epithelial tumour will be considered.
1.7.1. Intraperitoneal chemotherapy under normothermia
Ovarian cancer dissemination most frequently occurs intraperitoneally. The disease remains in the cavity for most of its natural history [104,105], and this biological behaviour provides the opportunity for increasing drug concentration selectively in the tumour area by direct intraperitoneal instillation [106,107] in order to overcome intrinsic or acquired drug resistance and simultaneously reduce systemic side-effects.
A large intergroup trial randomized 654 stage III patients with optimal residual disease (defined as largest nodule 2 cm or less after cytoreduction) to intraperitoneal cisplatin plus intravenous cyclophosphamide or intravenous cisplatin plus intravenous cyclophosphamide [108]. Intraperitoneal therapy was associated with a significantly improved median survival (49 versus 41 months) and fewer toxic side effects. The results of this trial have however, not substantially altered clinical practice. First, the onset of the study was in 1986, before the advent of paclitaxel. Second, somewhat counter intuitively, survival was not dependent on the extent of tumour residual mass. In a subsequent Gynecologic Oncology Group trial, 523 patients were randomized to intravenous cisplatin/paclitaxel of high-dose carboplatin followed by intraperitoneal cisplatin plus intravenous paclitaxel. The preliminary results demonstrated a significant increase in recurrence-free interval (28 versus 22 months), without the same favourable impact on overall survival.
However, the intraperitoneal chemotherapy carries some problems such as limited drug absorption into the tumour tissue in normothermic conditions [109] and incomplete drug distribution due to the abdominal postoperative adhesion [110].
1.7.2. The rationale of IPHP
In an attempt to overcome the drawbacks outlined in the last paragraph, intraperitoneal chemotherapy was combined with hyperthermia. The performance of regional drug delivery under supra-normal temperature is known as IPHP and it has become an area of growing interest supported by experimental observations. Cisplatin has been shown to penetrate much deeper in the tumour tissue under hyperthermic conditions [111, 112]. Moreover, at 40-42?C, neoplastic cells become more chemosensitive due to an increase in the intracellular concentration of drugs and in their activation process, especially for alkylating agents, and to alterations in the DNA repair process [113,114]. In addition, it has been shown that these events have a greater intensity in cisplatin-resistant rather than cisplatin-sensitive ovarian cancer cell lines. Formation of platinum-DNA adducts after cisplatin exposure is enhanced and/or adduct removal is increased in heated cells, resulting in a relatively higher DNA damage [115].
IPHP is the natural evolution of intraperitoneal chemotherapy that has been increasingly used over the last 2 decades, and still represents an intriguing area of clinical research. The peritoneal barrier, consisting of sub-mesothelial tissue and the capillary basement membrane, limits the reabsorption of hydrophilic and high molecular weight drugs such as MMC and CDDP, permitting a longer drug exposure within the peritoneal cavity [116]. Finally, IPHP favours the drug diffusion into the peritoneal cavity and the elimination of microscopic cancer residues by circuit filters [117].
1.7.3. Technical variations of LRT
Since LRT is a relatively recent technique for the treatment of peritoneal surface malignancies, several technical aspects have been a matter of debate. Consensus has not been reached about the modality of IPHP (Open vs Closed), optimal target temperature, optimal timing for bowel anastomosis (before or after the IPHP) and optimal drug combination.
1.7.3.1. Timing for bowel anastomosis
Another technical variation is the optimal timing for bowel anastomosis. They can be performed during the CRS just before the IPHP or after the completion of IPHP. Proponents of first alternative argue that delaying the anastomosis permits a better distribution of heat and drugs inside the peritoneal cavity, during the IPHP. In addition, they state that the risk of postoperative bowel complications can be diminished due to avoidance of potential adverse effects of heat and chemotherapy on the suture healing. On the other hand, others have proposed the second alternative supported by experimental and clinical evidence. In fact, the influence of chemotherapy on the suture healing depends on the type of drug. In animal studies, anastomotic healing can be impaired by intraperitoneal MMC but not by 5-fluorouracil, at normal temperature [118,119] or paclitaxel [120]. Local hyperthermia in itself has no adverse effect on rat anastomotic healing [121]. Moreover, there seems to be no increased morbidity, due to post-operative bowel fistula and/or anastomotic leak when anastomosis are constructed before the IPHP [96,102,122].
1.7.3.2. Modality of IPHP: Open vs Closed
One of major issue for debate has been the modality of IPHP execution. Investigators have not achieved a consensus about opened or closed abdomen techniques. Proponents of Coliseum technique [97] claim better drug and heat distribution by continuous manipulation of the abdominal organs. Deficiencies were noted in the distribution of methylene blue dye with the closed technique, which, in its turn, was blamed for higher rate of complications [123].
On the other hand, the closed technique permits an increase in the intrabdominal pressure that might lead to increased convection driven drug penetration of macromolecular agents such as TNF ? inside the tumour [124-126]. Moreover, a series of 94 patients with abdominal sarcomatosis, pseudomyxoma peritonei, peritoneal mesothelioma from four Italian oncological institutes (Milan, Rome, Padua, Turin) treated by CRS and intraperitoneal hyperthermic perfusion had recently been analysed. One of the endpoint of the investigation was rates of morbidity/mortality related to the procedure. Seventy-three patients were submitted to the closed procedure and it was verified that there was no statistically significant link between the modality of IPHP (closed versus open) with morbidity grades III/IV (p=0.6) [127]. Since up to date no prospective controlled clinical trial has been conducted addressing specifically the superiority of one technique over the other, the issue remains unclear. The accumulated data suggests that there is no striking difference between both in terms of operative morbidity.
1.7.3.3. Optimal target temperature
Consensus is lacking also about the optimal target temperature. Intra-abdominal temperatures ranging from 41 to 44 oC have been described; due to the dose-effect curve of hyperthermia a temperature of at least 42.5 oC seems to be optimal.
1.7.3.4. Optimal drug combination
The final technical issue worth discussing is the optimal IPHP drug regimen. Various drug combinations for ovarian cancer have been tested by experimental and phase I/II clinical studies: cisplatin alone [101,128], carboplatin alone [129], mitoxantrone alone [130], cisplatin+doxorubicin [102]. The criteria for choosing the ideal combination should be based on the pharmacokinetic profile of drugs, tumour chemo sensibility and toxicity. Ideally the drug must be water-soluble and of high molecular weight in order to guarantee a low peritoneal clearance. This, combined with a high systemic clearance, will result in pharmacological advantage expressed by a higher exposure of tumour to the agent (high AUCpe/AUCpl ratio). For intraperitoneal therapy to be effective against intraperitoneal tumours, the drug must also diffuse inward from the periphery of the tumour mass. The penetration ability of drug in the tumour is a function of passive diffusion, removal from by the capillary blood flow and temperature modulation. Penetration by passive diffusion is related to the AUCpe/AUCpl ratio, although this may not hold true for all drugs. Finally, the influence of temperature in the cytotoxicity should also be of concern, so that the higher the cell killing capacity of the drug due to the hyperthermia the better.
In 1976, the activity of cisplatin against epithelial ovarian cancer was first described [131]. Since then, cisplatin has become the most widely used agent in the systemic treatment of ovarian cancer with the response rate of 50% [132]. Combination chemotherapy with cisplatin and other cytotoxic drugs, most commonly doxorubicin and cyclophosphamide with or without hexamethymelanine, became standard systemic treatment, before the advent of paclitaxel, with response rate of 70-80%, [133-136]. Cisplatin combinations were found to be more effective than alkylating agents as a single agent of combinations, when measured by clinical response rates and progression free intervals [135-139].
When cisplatin was employed in a loco regional setting, in the treatment of epithelial ovarian cancer, a comparable distinctive antiblastic effect was shown. Cisplatin has a high AUCpe/AUCpl ratio, as compared to other cytostatic drugs, a deep tumour penetration ability (table 3) and partial response rate of up to 65% in normothermic condition [144] (see Appendix, table 4).
Another eligible agent for IPHP is carboplatin. Despite a better therapeutic index than cisplatin, with substantially less renal toxicity, nausea and neurotoxicity, carboplatin has not a favourable pharmacokinetic profile as cisplatin (see Appendix, table 3) [111,129,149]. In fact, the AUCpe/AUCpl ratio, tumour penetration capacity and response rate are markedly lower [148,150,151].
Doxorubicin has one of the highest AUCpe/AUCpl ratio of about 80 (see Appendix, table 3)[152,153]. Irrespective of limited tumour diffusion ability, not more than several cell layers, a response rate of 30% was reported when doxorubicin was administered intraperitoneally, under normothermic condition [154]. The dose limiting toxicity, chemical induced peritonitis, makes doxorubicin feasible for loco regional employment only at very low dose.
Other agents are sketched in table 3 and 4 (see Appendix) [109,130,155-164]. Oxaliplatin; paclitaxel and gemcitabine are also promising for IPHP in the treatment of ovarian cancer However, since they are still under experimental and/or phase I clinical investigations, they should be further investigated before been evaluated in a prospective phase III trial.
In summary, the best chemotherapy combination for IPHP for patients with ovarian cancer is still to be defined. However, experimental and phase I/II clinical studies suggests the combination of cisplatin and doxorubicin the currently most advisable regimen for LRT for epithelial ovarian cancer. A complete response rate of up to 59% has been reported with the employment of this combination in IPHP for advanced and recurrent epithelial ovarian cancer in a phase II clinical trial [102].
1.8. LRT for ovarian cancer
Investigators from Netherlands Cancer Institute have published a study [128] where 5 heavily pre-treated patients with extensive abdominal ovarian bulky tumour were submitted to aggressive cytoreduction followed by perfusion of the abdominal cavity with hyperthermic cisplatin 50-70 mg/m2 for 90 min. During perfusion the intra-abdominal temperature was maintained at 400 C. No major intra- or post-operative complications emerged. Median post-operative ileus (resuming of soft diet) was 11 days (9-13 days). The mean period of hospitalisation was 25 days (range 17-42). Toxicity due to intraperitoneal cisplatin was mainly metabolic and of grade 1-2, while no nephrotoxicity was observed. They concluded that aggressive cytoreduction combined with hyperthermic intra-operative intraperitoneal cisplatin was feasible in a small group of heavily pre-treated ovarian cancer patients with extensive tumour bulk with acceptable morbidity and toxicity. They stated that further studies should be carried out in larger groups of patients to further establish the feasibility of this intensified treatment strategy. The effectiveness of this combined treatment is likely to be dependent on the effectiveness of post-operative adjuvant chemotherapeutic regimens.
Hagar et al. in 2001 [165] conducted a prospective clinical trial evaluating the feasibility, efficacy and impact of IPHP chemotherapy on survival and quality of life of patients with advanced, peritoneal disseminated ovarian cancer. Thirty-six patients with ovarian cancer were accrued for the study, their selection being based on their progression following different systemic chemotherapies. The patients' average Karnofsky-performance status was 60% and 17/36 patients had ascites before IPHP chemotherapy. The intraperitoneal temperature was 42-430 C. Median overall survival time from first diagnosis of disease was 49 +/- 8 months and from the first IPHP chemotherapy treatment 19 +/- 4 months. The observed 1-year overall survival rate of all patients from the start of the first IPHP was 65 +/- 8% and the 5-year overall survival was 16 +/- 7%. Malignant ascites vanished within less than 3-5 procedures. Quality of life could be improved. The adverse effects were mild especially compared to systemic chemotherapy. In 3 out of 162 treatments, peritoneal disturbances with symptoms of subileus were observed. The authors concluded that IPHP is technically feasible, safe, and may improve the treatment outcome of patients with advanced ovarian cancer as salvage therapy, in second-line treatment or even as consolidation or maintenance therapy following induction chemotherapy.
At the National Cancer Institute of Milan the currently discussed treatment was evaluated in 27 patients with advanced recurrent ovarian carcinoma [102]. Median times to overall progression and local progression were 21.8 months and 16 months, respectively. Variables that affected the overall survival or time to progression were as follows: residual disease (p=.00025), patient age (p=.04), and lag time between diagnosis and CRS+IPHP (p=.04). Treatment-related morbidity, mortality and toxicity were 11%, 4% and 27%, respectively. Eight (89%) of 9 patients had ascites resolution.
It is hard to ascertain in which extent the apparent survival advantage reported by these uncontrolled clinical studies resulted from selection bias. Moreover, the investigations involved heterogeneous patient population, at different stages of disease evolution, usually heavily pre-treated, and therefore, with different tumour chemo resistances. Anyway, to clarify these arguments, all of these investigators agreed with respect the need for a prospective randomized clinical trial to confirm their encouraging findings.

1- 2 - 3 - 4 - 5 - 6 - 7