Abstract
Ovarian cancer is the leading cause of death among cancers of the female reproductive tract. Most women present with regionally advanced intraperitoneal disease with superficial spread of tumor along the peritoneal surfaces. Distant metastasis rarely occurs until late in the disease process; however, long-term survival of patients with advanced disease remains poor. Intraperitoneal chemotherapy allows higher concentration and prolonged half-life of chemotherapy within the peritoneal cavity compared with intravenous administration, while therapeutic intravenous concentrations are obtained. Three large randomized Phase III trials and a meta-analysis have firmly established the superiority of combination intraperitoneal and intravenous chemotherapy compared with intravenous chemotherapy alone in women with optimally debulked advanced epithelial ovarian carcinoma. In January of 2006, the National Cancer Institute issued a clinical alert recommending that these women receive a combination of intravenous and intraperitoneal chemotherapy; however, resistance to adopt this as a new standard of care results from concerns regarding toxicity and lack of experience with intraperitoneal catheters. Here we review the largest trials of combination intraperitoneal and intravenous chemotherapy, address the concerns regarding intraperitoneal chemotherapy, and provide further resources for implementing intraperitoneal chemotherapy into practice.
In the year 2007, approximately 22,430 women will be diagnosed and 15,280 women will die of ovarian cancer [1]. It is the second most common gynecologic malignancy, but accounts for more deaths than any other malignancy of the female reproductive tract. There are no accurate screening tests and only one out of five women presents with highly curable localized disease. The 5-year survival rates for women with regional and distant disease across all stages are 68 and 29%, respectively. Standard treatment includes debulking surgery followed by chemotherapy. The prognosis has been slowly improving over the past 25 years. The recent demonstration in prospective randomized controlled clinical trials of substantially improved survival using intraperitoneal chemotherapy provides a powerful addition to the treatment armamentarium.
The first prospective randomized Phase III trial showing improved efficacy of combination intraperitoneal and intravenous chemotherapy over intravenous chemotherapy alone was published in 1996 [2]. Two additional large Phase III randomized trials and a recent meta-analysis have now firmly established the superiority of intraperitoneal and intravenous chemotherapy over intravenous chemotherapy alone [3–5]. Resistance to adopt intraperitoneal chemotherapy as a new standard of care results from concerns regarding toxicity and lack of experience with intraperitoneal catheters.
Prior to 1993, the standard of care for patients with advanced ovarian cancer included debulking surgery followed by intravenous combination chemotherapy with cyclophosphamide and a platinum agent [6–8]. McGuire and colleagues subsequently reported an improved relapse-free survival in patients given paclitaxel in combination with cisplatin compared with cyclophosphamide and cisplatin in suboptimally debulked disease [9]. Further reports confirmed the superiority of paclitaxel over cyclophosphamide and the combination of intravenous paclitaxel and cisplatin became the standard of care [10,11]. Subsequently, Ozols and colleagues compared paclitaxel with either cisplatin or carboplatin and found similar efficacy and an improved toxicity analysis favoring carboplatin, with significantly less nausea, emesis and neurotoxicity, and an improved quality of life [12]. This was confirmed by later studies and the carboplatin and taxane combination has been adopted as the standard intravenous chemotherapy combination [13–15]. Although many intravenous chemotherapeutic advances have been made over the last 13 years, the long-term survival for those with advanced disease at diagnosis is still only 10–30% [16].
Rational for intraperitoneal chemotherapy
A total of 70–80% of women with newly diagnosed ovarian cancer will have advanced disease (cancer that has spread beyond the ovary). Most ovarian cancer patients present with regionally advanced intraperitoneal disease with superficial spread of tumor along the peritoneal surfaces [16]. Ovarian cancer rarely metastasizes to the liver or outside of the peritoneal cavity until very late in the natural history of the disease process. Owing to this specific natural history, and based on observations from the pharmokinetics of peritoneal dialysis, Dedrick and colleagues first theorized in 1978 that the intraperitoneal administration of chemotherapy may provide a ‘peritoneal advantage’, i.e., a greater tumor exposure of the cancer to antineoplastic drugs within the peritoneal cavity than results from intravenous administration, and may be advantageous in the treatment of ovarian carcinoma [17]. Further evaluation of this hypothesis resulted in clinical trials examining the pharmacokinetics of the intraperitoneal administration of chemo-therapy. It was demonstrated that administration of intraperitoneal cisplatin results in a 10–20-fold greater exposure of the drug to the tumor as opposed to intravenous infusion alone [18–23]. Furthermore, adequate concentrations of cisplatin were absorbed into the systemic circulation so that those components of the tumor supplied only by capillary flow were exposed to cytotoxic concentrations of cisplatin [18,19].
Intraperitoneal pharmacokinetics of paclitaxel were later evaluated by Markman and colleagues [3], revealing a 1000-fold greater concentration in the intraperitoneal space compared with that measured in the plasma [24,25]. Intraperitoneal paclitaxel had a prolonged half-life in the peritoneal space but, despite these properties, therapeutic intravenous concentrations of paclitaxel were also achieved.
Randomized trials
There have been seven reported randomized trials of intravenous vs intraperitoneal therapy of ovarian carcinoma [2–4,26–29], and one trial utilizing intraperitoneal administration as consolidation therapy compared with surveillance [30]. We will review the three largest randomized trials of intravenous versus intraperitoneal therapy [2–4] (
Phase III trials of intraperitoneal chemotherapy.
ip.: Intraperitoneal; iv.: Intravenous; NR: Not recorded; OS: Overall survival; PFS: Progression-free survival.
The first large randomized Phase III trial of intraperitoneal versus intravenous cisplatin chemotherapy was conducted by the Southwest Oncology Group (SWOG) and the Gynecologic Oncology Group (GOG) and reported by Alberts and colleagues in 1996 [2]. Between 1986 and 1992, 546 stage III epithelial ovarian cancer patients who had undergone optimal debulking surgery to less than 2 cm residual tumor were randomized. All patients received intravenous cyclophosphamide. Both arms contained cisplatin 100 mg/m2 administered intravenous or intraperitoneal at 3-week intervals for six planned cycles.
The median survival was superior in the intraperitoneal arm (41 versus 49 months, p = 0.02). In both groups, only 58% of patients were able to complete all six cycles of chemotherapy. A total of 297 patients underwent a second-look surgery, revealing pathologic complete responses in 36% in the intravenous arm versus 47% in the intraperitoneal arm.
Overall toxicity was more common in the intravenous group, which was attributed to higher systemic concentrations of cisplatin. Significantly more patients in the intravenous group had grade 3 or greater granulocytopenia (p = 0.002) and leucopenia (p = 0.04). Moderate-to-severe tinnitus and hearing loss, and grade 2 or 3 neuromuscular toxic effects (25 versus 15%, p = 0.02) were also more frequent. Increased grade 2 abdominal pain was noted in the intraperitoneal group (p = 0.001) but usually persisted for less than 24 h after treatment. One patient had grade 4 abdominal pain. Transient dyspnea was infrequent but occurred in significantly more patients in the intraperitoneal group (3 versus 0.4%, p = 0.002). This was felt to result from compression of the base of the lung by the fluid-filled intraperitoneal cavity.
However, this study was controversial owing to several issues. Firstly, accrual of patients having low-volume (0.5 cm) residual disease was expanded during the study following recommendations by the GOG in order to increase statistical power in the subgroup. Secondly, the documented pathologic complete response rate with second-look surgery in the intravenous group was lower than that reported in similar studies. Last, many experts suggested that because patients were not treated with paclitaxel, the potential benefit of the intraperitoneal administration of cisplatin therapy would be eliminated with the addition of the intravenous taxane, which had not yet been approved for the treatment of ovarian cancer at the time that this study was designed.
A second large, randomized Phase III intergroup study comparing the efficacy of intravenous versus intraperitoneal chemotherapy was conducted by the SWOG, GOG and Eastern Cooperative Oncology Group (ECOG) and reported by Markman and colleagues in 2001 [3]. This study was conducted between 1992 and 1995. A total of 462 evaluable stage III epithelial ovarian cancer patients who had undergone optimal debulking surgery to less than 1 cm residual tumor were included. All patients randomized to intravenous treatment alone received intravenous paclitaxel 135mg/m2 over 24 h, followed by intravenous cisplatin 75mg/m2 on day 2. The intraperitoneal arm also incorporated an initial two cycles of moderately high-dose intravenous carboplatin (AUC = 9) to evaluate the potential utility of ‘chemical debulking’ prior to the administration of six cycles of paclitaxel 135 mg/m2 over 24 h followed by intraperitoneal cisplatin 100 mg/m2 on day 2. The cycles were administered every 3 weeks for six cycles. The median progression-free survival was statistically superior in the intraperitoneal cisplatin arm (22 versus 28 months, p = 0.01, one tail). There was a borderline improvement in median overall survival in the intraperitoneal arm (52 versus 63 months, p = 0.05, one tail). Notably, this was the first study to show a median survival of greater than 5 years. Furthermore, 18% of patients only received two or fewer total courses of intraperitoneal therapy, making the significant survival difference even more impressive. The associated myelosuppression, however, prevented any recommendation for routine clinical use, though this was attributed to the initial two cycles of high-dose intravenous carboplatin as opposed to the intraperitoneal therapy. Gastrointestinal and metabolic toxicity was also more common in the intraperitoneal arm.
One month before the aforementioned study was reported in 2001, a third large, randomized Phase III study completed accrual. Many were pessimistic as to the potential of intraperitoneal chemotherapy. William McGuire, co-Chairman of the GOG Ovarian Cancer Committee in 2001, compared the Markman trial to a ‘sacrifice bunt’, and suggested that any future trial would have to be “solid enough” to convince that there was a role for regional intraperitoneal administration of chemotherapy:
“Thus, this sacrifice bunt may have moved the runner from first to second and into scoring position, even though this batter was thrown out at first. If the recently completed trial, however, does not show a significant advantage in therapeutic index for the regional therapy approach, most will logically conclude that the batter who is up simply hit into a 6-5-3 double play… So even after this sacrifice bunt, the next hit will have to be solid enough to bring the man on second home.” [31].
The ‘next hit’, reported by Armstrong and colleagues in the New England Journal of Medicine in 2006 [4], did appear to “bring the man on second home”. Based on the impressive pharmacokinetic data from early phase trials with paclitaxel, this third large, randomized, Phase III trial administered intraperitoneal paclitaxel in addition to intraperitoneal cisplatin. The study was conducted between 1998 and 2001. A total of 415 eligible patients with stage III epithelial ovarian or primary peritoneal carcinoma who had undergone optimal debulking surgery to 1 cm or less residual tumor were randomized. All patients received intravenous paclitaxel, 135 mg/m2, over 24 h on day 1. The control arm received intravenous cisplatin 75 mg/m2 on day 2. The experimental arm received intraperitoneal cisplatin 100 mg/m2 on day 2 and intraperitoneal paclitaxel 60 mg/m2 on day 8. Cycles were administered every 3 weeks for six cycles.
Although only a borderline improvement in the median duration of progression-free survival on the intraperitoneal arm (18.3 vs 23.8 months, p = 0.05) was described, the median duration of overall survival was statistically significantly improved in the intraperitoneal arm (49.7 vs 65.6 months, p = 0.03). The potential benefit may have been diminished by the fact that more than half of patients were unable to complete all six cycles of intraperitoneal therapy. Women who received intraperitoneal treatment had a 25% reduction in the risk of death. Of those who underwent a second-look laparotomy; the complete response was superior in the intraperitoneal arm (41 vs 57%).
The experimental intraperitoneal cisplatin and paclitaxel arm experienced more toxicity, having increased grade 3/4 pain, fatigue and hematologic, gastrointestinal, metabolic and neurologic side effects (p ≤ 0.001). Only 42% of the patients in the intraperitoneal therapy group completed all six cycles of intraperitoneal therapy, primarily due to catheter-related complications, which occurred in 39 of 118 patients (33%). Infection was observed in 21 women, catheter blockage in nine, catheter leak in three, access problems in five and drainage per vagina in one.
Increased toxicity was partially attributed to the increased dose of cisplatin in the intraperitoneal group and possibly to the intraperitoneal paclitaxel, which has been reported to elicit abdominal pain. However, the toxicity associated with intraperitoneal treatment was of short duration, manageable and self-limited. The quality of life measurements were significantly worse in the intraperitoneal group prior to cycle 4 and 3–6 weeks following completion of treatment, however, had returned to baseline at 1 year following treatment, and at that time were not statistically significantly different than women who received intravenous chemotherapy alone.
Meta-analysis
Kyrgiou and colleagues published a recent meta-analysis of all randomized chemotherapeutic trials over the last 40 years in ovarian cancer [5]. A total of 198 trials with 120 different regimens were included. A platinum and taxane-based combination utilizing intraperitoneal chemotherapy resulted in a relative risk reduction of 55% compared with intravenous nonplatinum or taxane monotherapy. Monte Carlo simulations revealed a 92% probability that combinations of platinum and taxane with intraperitoneal administration are the most effective regimens. Given an expected median survival of 2.5 years without treatment, more than 3 years of median survival are gained with the most effective treatment.
Administration of intraperitoneal chemotherapy
In January of 2006, the National Cancer Institute issued a clinical alert recommending that women with advanced ovarian cancer who undergo optimal surgical debulking receive a combination of intravenous and intraperitoneal chemotherapy and specific recommendations were given on the use and administration of intraperitoneal chemotherapy [101,102]. The following is a summary of these recommendations.
The potential risks and benefits of intraperitoneal chemotherapy should be discussed with the patient before the primary surgery. The port for intraperitoneal infusion may be placed at the time of primary surgery, unless there are contraindications such as serious comorbidity limiting ability to tolerate intraperitoneal chemotherapy, uncertain pathologic diagnosis, gross bacterial contamination of the peritoneal cavity or serious intraoperative complications precluding the placement of the port. Bowel resection and reanasta-mosis is not an absolute contraindication to intraperitoneal port placement at the time of surgery, but the option of placing the port during a second procedure at a later time is left to the discretion of the surgeon. The performance of a supracervical hysterectomy is not recommended as leakage of the peritoneal instillate from the vagina has occurred. Vaginal defects should be closed with delayed absorbable sutures. The abdominal wound may also leak instillate and therefore should be carefully closed with semipermanent or permanent sutures.
There are two commonly used types of intraperitoneal catheters. Many experts recommend the use of a semi-permanent subcutaneous venous access port connected to a single-lumen venous catheter. Some data suggest that peritoneal catheters with fenestrations and dacron cuffs may be associated with a greater incidence of bowel adhesions and erosion into the bowel, however, this remains controversial [32,33].
The catheter should be located on the inferior thorax at the midclavicular line, overlying the ribs, low enough to avoid irritation from a brassiere. A transverse incision slightly larger than the port should be made overlying the ribs. The port should be sutured with permanent 2–0 suture at four corners to the fascia, to prevent rotation or migration. The catheter itself should be tunneled subcutaneously, above the fascia, to a point 6 cm lateral to the umbilicus. At this point, the catheter may be pulled into the peritoneal cavity through a small incision. The catheter should be cut to an adequate length (approximately 10 cm) to ensure that it remains in the abdominal cavity, but not so long as to promote adherence to the bowel or kinking. The port should be flushed with 10 ml of heparin (100 units per ml).
The optimal volume of infusate utilized to administer the intraperitoneal chemotherapy has not been established. Armstrong and colleagues reconstituted the paclitaxel and cisplatin in 2 l of normal saline warmed to body temperature followed by infusion through the peritoneal catheter as rapidly as possible [4]. After completion of the infusion, patients were encouraged to change position at 15-minute intervals for 2 h in order to facilitate intra-abdominal distribution. Others recommend reconstitution of the chemotherapy in 1 l of instillate followed by an additional liter of normal saline. If the patient becomes uncomfortable, the second liter may not be entirely infused. There is no need to drain the infused fluid from the abdominal cavity. Patients with malignant ascites should undergo drainage of their ascites prior to intraperitoneal installation of the infusate. These patients may require additional intravenous fluid [4].
Routine supportive care and premedications for intraperitoneal paclitaxel or intraperitoneal cisplatin should be at least equivalent to those used for intravenous administration. Armstrong and colleagues prescribed dexamethsone 20 mg orally 12 and 6 hours prior to paclitaxel or 20 mg intravenously 30 min before the paclitaxel infusion [4]. Both diphenhydramine 50 mg and cime-tidine 300 mg (or an alternative) were administered intravenously 30 min before the paclitaxel infusion. Routine intravenous hydration with at least 1 l of intravenous saline and antiemetics should be given before and after cisplatin. In addition, delayed nausea is common with intraperitoneal administration of cisplatin and appropriate anti-emetics should be prescribed.
If a patient develops intolerable toxicity from intraperitoneal chemotherapy, such as abdominal pain, ileus or infection, and so on, the intraperitoneal therapy should be abandoned and intravenous chemotherapy alone may be substituted. intraperitoneal catheters may be removed at the completion of intraperitoneal therapy.
Conclusion
Three large randomized Phase III trials and a meta-analysis have firmly established the superiority of combination intraperitoneal and intravenous chemotherapy compared with intravenous chemotherapy alone in women with optimally debulked advanced epithelial ovarian carcinoma [2–5]. Women should be counseled on the benefits and risks of intraperitoneal therapy prior to their initial surgery.
The optimal regimen of combination intravenous and intraperitoneal chemotherapy remains to be established; however, all three trials utilized intraperitoneal cisplatin, 100 mg/m2. The most recent utilized intravenous paclitaxel 135 mg/m2 intravenous on day 1 over 24 hours, intraperitoneal cisplatin 100 mg/m2 on day 2, and intraperitoneal paclitaxel at 60 mg/m2 on day 8 [4]. This resulted in a 16-month improvement in overall survival compared with the use of intravenous chemotherapy alone.
Combination intraperitoneal and intravenous chemotherapy is associated with significantly increased short-term toxicity compared with intravenous chemotherapy alone. However, the toxicity is limited and manageable, although many patients are unable to tolerate the full six cycles of intraperitoneal chemotherapy. Armstrong and colleagues reported that only 42% of patients were able to undergo the full six cycles of intraperitoneal therapy, although even this foreshortened course of intraperitoneal chemotherapy translated into a marked survival benefit.
Despite more than two decades of research of the use of intraperitoneal chemotherapy, there are multiple areas that require further research. Improved attempts to decrease toxicity without sacrificing efficacy must be investigated. For example, it is unknown whether decreasing intraperitoneal cisplatin from 100 mg/m2 to 75 mg/m2 will sacrifice benefit, or if the substitution of carboplatin, with its improved therapeutic index, will be possible. Previous pharmacokinetic trials of intraperitoneal carboplatin have been promising and randomized trials should be performed comparing intraperitoneal cisplatin and carboplatin [34–36]. There is currently an ongoing Phase II trial by Massachusetts General Hospital Cancer Center using intraperitoneal paclitaxel and carboplatin [103]. SWOG S009 completed accrual in February of 2006 [105]. This Phase II trial in suboptimally debulked stage III or intravenous patients administered 3 cycles of neoadjuvant intravenous paclitaxel and carboplatin followed by secondary debulking surgery. Post-operative chemotherapy consisted of intravenous paclitaxel over 3 hours followed by intraperitoneal carboplatin on day 1 and intraperitoneal paclitaxel on day 8.
Executive summary
Intravenous chemotherapy for advanced epithelial ovarian cancer has prolonged median survival, but long-term survival remains only 10–30%.
Most patients with epithelial ovarian cancer are found to have regional disease, with spread of tumor along peritoneal surfaces at the time of diagnosis.
Intraperitoneal chemotherapy allows higher concentrations of chemotherapy within the peritoneal cavity compared with intravenous administration, and allows a prolonged half-life in the peritoneal space, while therapeutic intravenous concentrations are obtained.
The first large randomized trial of intraperitoneal and intravenous chemotherapy versus intravenous chemotherapy alone by Alberts and colleagues in 1996, showed a significant survival benefit of intraperitoneal cisplatin compared with intravenous cisplatin, with both arms containing intravenous cyclophosphamide.
The second large randomized trial by Markman and colleagues in 2001 showed a significant survival benefit of intraperitoneal cisplatin compared with intravenous cisplatin, even though 18% of patients received two or fewer cycles of intraperitoneal chemotherapy. However, the intraperitoneal arm also was given two cycles of high-dose carboplatin, which likely contributed to increased myelotoxicity.
The third large randomized trial by Armstrong and colleagues in 2006 showed a significant survival benefit of intraperitoneal cisplatin and paclitaxel with intravenous paclitaxel compared with intravenous paclitaxel and cisplatin, even though only 42% of patients completed all six cycles of intraperitoneal chemotherapy. Overall, there was a 16-month survival benefit of intraperitoneal chemotherapy.
Monte Carlo simulations show a 92% probability that combinations of platinum and taxane with intraperitoneal administration are the most effective regimens. Given an expected median survival of 2.5 years without treatment, more than 3 years of median survival is gained with the combination of intraperitoneal and intravenous chemotherapy.
Patients who should be offered intraperitoneal chemotherapy are those with stage III advanced ovarian cancers who have undergone optimal surgical cytoreduction. The National Comprehensive Cancer Network also recommends the option for selected patients with stage II disease.
Resistance to adopting intraperitoneal chemotherapy as a new standard of care results from concerns regarding toxicity, and lack of experience with intraperitoneal catheters and intraperitoneal chemotherapy.
Toxicity of intraperitoneal chemotherapy is usually self-limited and well managed.
Specific recommendations for intraperitoneal chemotherapy are summarized above and are provided in detail from the National Cancer Institute clinical announcement on intraperitoneal chemotherapy in ovarian cancer [101] and published online at [102].
The Gynecologic Oncology Group has published online educational materials for both patients and professionals [104] including links to articles, nursing and supportive care guidelines, and intraperitoneal port placement video, and example order sheets.
It remains unclear whether the addition of intraperitoneal paclitaxel on day 8 is necessary; however, the marked pharmacologic advantage suggests that this agent is responsible for at least a portion of the improved efficacy observed in the Armstrong study. However, there are theoretical concerns that the prolonged half-life of intraperitoneal paclitaxel may delay wound healing [37]. Other challenges in the administration of intraperitoneal therapy must be investigated, including such components as the types, techniques and timing of placement of intraperitoneal catheters.
The large Phase III trials have only included stage III patients that have undergone optimal debulking surgery [2–4]. Potential improvements in outcome through the utilization of intraperitoneal therapy in stage II or intravenous patients or for patients who have undergone suboptimal tumor debulking are unknown. Combination intravenous and intraperitoneal chemotherapy is listed as an option. However, for patients with stage II disease under current National Comprehensive Cancer Network Guidelines [15].
The pharmacokinetics and potential peritoneal advantage of newer chemotherapeutic agents must continue to be evaluated. Phase I studies of sustained-release paclitaxel microspheres [38], intraperitoneal docetaxel [39] and intraperitoneal gemcitabine [40] have been reported.
Resistance to the adoption of intraperitoneal chemotherapy as a new standard in optimally debulked advanced ovarian epithelial ovarian carcinoma largely stems from unfamiliarity with the treatment and fear of toxicity. The Gynecologic Oncology Group has published online educational materials for both patients and professionals including links to articles, nursing and supportive care guidelines, an intraperitoneal port placement video, and examples of order sheets [103]. In conclusion, to quote Alberts and colleagues from a review in the Journal of Clinical Oncology in 2002, intraperitoneal therapy is ‘a therapy whose time has come’ [33].
Future perspective
Follow-up clinical trials utilizing intraperitoneal chemotherapy in epithelial ovarian carcinoma will further elucidate the best treatment regimens, dosing and administration techniques required to maximize efficacy and limit toxicity. Data will be available to make recommendations in a variety of settings and stages. Although the first intraperitoneal studies were published more than two decades ago, intraperitoneal therapy has yet to be generally adopted in the community. Given the well-established superior efficacy of this treatment, practioners and patients alike will make intravenous and intraperitoneal chemotherapy the gold standard in advanced epithelial ovarian cancer treatment.
