Michael V Seiden
Associate Professor of Medicine
Division of Medical Oncology
Massachusetts General Hospital
Epithelial ovarian cancer is the most lethal gynaecological malignancy in countries that have cervical cancer screening programmes. The median age at diagnosis is 63 years, and disease is usually advanced by the time of diagnosis. The overall incidence has not changed significantly in the last two decades, nor has the overall survival. The five-year survival rate, however, has improved from 34% in 1974 to 50% in 1994 for all stages of the disease. This ranges from 28% for women with stage III or IV disease to 95% for those with stage I disease with favourable prognostic characteristics.(1,2)
Epithelial ovarian cancer (the most common ovarian cancer) arises from the surface epithelium of the ovary and begins as a cystic mass. It is unusual for symptoms to occur at this stage, making early diagnosis a challenge. Tumour cells then exfoliate from the ovary or when the cystic mass ruptures, spreading malignant cells throughout the peritoneum. Tumour cells adhere throughout this cavity, and tumour deposits can be found on any intraperitoneal surface, including the hemidiaphragm, which is often associated with pleural effusions. Tumour cells can also spread to the retroperitoneum via the lymphatic vessels that drain the ovaries.(2) Initial symptoms of ovarian cancer tend to be subtle and migratory, and typically ignored or ascribed to much less serious health problems by patients and, unfortunately, by their physicians. The majority of women present with abdominal complaints, including dyspepsia, early satiety, postprandial cramping and change in bowel pattern, which is often thought to be secondary to irritable bowel syndrome, diverticulitis, stress, depression or menopause. Symptoms usually persist for months before a definitive diagnosis can be established.
The cornerstone in the management of ovarian cancer is the initial surgical intervention. Aggressive surgical debulking is essential and is the most important prognostic factor in long-term outcomes.(3) Its goal is optimal reduction of tumour, defined as no residual tumour nodule greater than 1cm at the end of surgery. Because ovarian cancer is a disease that spreads along the surface of intraperitoneal organs, the surface of the small and large intestines is often involved. Bowel resection is common and should be taken into consideration when managing future gastrointestinal symptoms. Several studies have demonstrated that the best surgical outcome can be expected when patients have their surgery performed by a gynaecological oncologist; yet the majority of women worldwide still have initial surgery performed by gynaecologists or general surgeons, who are less likely to achieve best surgical outcomes.(4,5) In these cases a second surgical procedure by a gynaecological oncologist should be considered, often after a few cycles of chemotherapy, especially if the first operation left significant intra-abdominal disease.(6)
Surgical resection of tumour is rarely complete, and all women with ovarian cancer should be offered chemotherapy after the surgical procedure or before surgery in the event of their not having access to the appropriate surgical expertise or having extensive and unresectable disease. The medical management of ovarian cancer has evolved over the last 30 years, with the current standard of care being paclitaxel and carboplatin. The use of a taxane with platinum was initially supported by Gynecologic Oncology Group (GOG) 111, a pivotal phase III trial completed in the early 1990s demonstrating that the substitution of cyclophosphamide with paclitaxel resulted in improved response rate and statistically significant progression- and disease-free survival in women with advanced ovarian cancer.(7,8)
Over the next decade, cisplatin was often substituted by carboplatin due to the more favourable toxicity profile of the latter and the general feeling that the two compounds had similar efficacy. This practice was validated by GOG 158 and two European studies.(9–11) GOG 158 was a noninferiority study comparing the toxicity profile and efficacy of the carboplatin–paclitaxel combination with that of cisplatin–paclitaxel. Toxicity, especially metabolic and gastrointestinal, was far less significant in women assigned to the carboplatin arm. Median overall survival was 48.7 months, and 57.4 months for the cisplatin and carboplatin arms, respectively.(11) This demonstrated that the combination of paclitaxel and carboplatin is not inferior to paclitaxel and cisplatin.
Paclitaxel may be replaced by docetaxel without compromising response rate and survival, as demonstrated by investigators in Scotland through the large SCOTROC (Scottish Randomised Trial in Ovarian Cancer) trial. This randomised trial demonstrated that docetaxel was less neurotoxic, but more myelosuppressive, than paclitaxel. It is a logical approach in women with pre-existing neuropathy or in whom neuropathy would be intolerable.(12)
Although 75% of women with advanced stage disease achieve clinical remission after initial therapy, approximately 75% of them will have a recurrence of the disease.(13) As ovarian cancer is a chemotherapy-sensitive tumour, subsequent lines of chemotherapy are often successful in achieving clinical response. Choosing therapy in this setting is dependent upon platinum sensitivity, with potential platinum sensitivity being determined by the interval from completion of platinum-containing regimens to recurrence. Women who have a recurrence of the disease less than six months from completing platinum-based therapy are defined as having platinum-resistant tumours, while those with longer remissions have tumours that are potentially platinum-sensitive. Unsurprisingly, the likelihood or response to retreatment with platinum increases with progressively longer platinum-free intervals. Women who have tumour recurrence greater than 12 months from the last platinum-based therapy are defined as having platinum-sensitive tumours. Tumour recurrence (on average 6–12 months after platinum-based therapy) demonstrates intermediate sensitivity to platinum. Clinical judgement and patient preference should be used when choosing therapy.(14)
Treatment of women with platinum-sensitive tumours
While there is general consensus that patients with potentially platinum-sensitive tumours should be considered for retreatment with platinum, the use of single-agent platinum (typically carboplatin) versus platinum in combination with a second active drug is controversial.
In a multicentre randomised trial, the European ICON (International Collaborative Ovarian Neoplasm) and AGO (Arbeitsgemeinschaft Gynaekologische Onkologie) collaborators compared women with platinum-sensitive ovarian cancer who were assigned conventional platinum-based chemotherapy or paclitaxel plus platinum. This study showed improved outcome in the paclitaxel arm, with a two-year absolute survival difference of 7% (57% vs 50%), a difference of five months in median survival (29 vs 24 months) and a difference of three months in median progression-free survival (12 vs nine months).(15) Pfisterer and colleagues also demonstrated that combination platinum-based chemotherapy (in this study with gemcitabine) improves response rate.(16) The overall response rate for those treated with combination therapy was 47.2%, versus 30.9% in the single-agent arm.(16) This latter study was not powered to detect a survival difference. Both studies demonstrate improved outcome associated with treatment using a platinum-containing doublet compared with single-agent platinum in women with recurrent and platinum-sensitive ovarian cancer. As data become more mature, this may be incorporated as the standard of care for this subset of patients.
The likelihood of achieving a significant and/or durable response to chemotherapy in women who are thought to have recurrent and platinum-resistant tumours is low. Although there are many active drugs in ovarian cancer, their response rates are similar, with most responses being partial and response rates falling between 10% and 20%. Agents with randomised data supporting their use in this setting include liposomal doxorubicin, topotecan and paclitaxel.(17,18) A large body of phase II literature supports the use of gemcitabine and etoposide, with data also supporting docetaxel, vinorelbine, hexamethylmelamine, irinotecan and capecitabine.(19–26) When choosing a drug, clinical judgement, patient preference and the toxicity profile of each drug need to be taken into consideration. Treatments may be given daily (oral), weekly or monthly. Performance status, tolerance of gastrointestinal toxicity and desire to avoid alopecia all tend to factor into best choices for palliative chemotherapy. Some women may need closer monitoring due to disease status; weekly treatment may be a better option for these patients. Bone marrow reserve may be an issue, particularly in women who have had multiple cycles of carboplatin. The ability to absorb oral medications is often an issue, due to partial or intermittent bowel obstruction. Table 1 reviews the schedule, route of delivery and some of the major toxicities associated with each of the commonly used agents. Importantly, many women with recurrent disease will receive many of these therapies, typically in sequential fashion, in an attempt to palliate or delay the onset of disease-related symptoms.
Platinum-containing regimens used to treat ovarian cancer are moderately emetogenic, and nausea and vomiting may occur. This is certainly less frequent with carboplatin than with cisplatin. 5-HT(3) antagonists, such as ondansetron, granisetron and palonosetron, and the neurokinin-1 antagonist aprepitant are often needed before treatment and for 1–3 days afterwards.(27–29) This can present challenges, as constipation is a common symptom of ovarian cancer, and patients should be advised to use stool softeners and/or laxatives. Prochlorperazine, dexamethasone and lorazepam may be adequate to treat patients with mild nausea.(30) In the USA, growth factors, including filgrastim, pegfilgrastim, epoetin and darbepoetin, are often used to minimise haematological toxicities associated with carboplatin and other chemotherapy regimens. These agents may allow for optimal dosing of chemotherapy, potentially improving outcomes and decreasing morbidity.(31–34)
As previously discussed, bowel obstruction is a common occurrence in women with ovarian cancer. Octreotide, a somatostatin analogue, is used with some success to reduce gastric and small intestine secretions, thereby minimising bowel distention, contractility, nausea and vomiting, particularly in women with inoperable bowel obstruction.(35) Octreotide can now be delivered monthly in long-acting depo form as an intramuscular injection.
Despite advances made over the last 30 years, five-year survival rates still remain low in women with advanced ovarian cancer. New classes of agents that target numerous pathways of possible importance in the biology of ovarian cancer are currently under investigation. Currently, the most intense investigation is focused on drugs that interfere with the epidermal growth factor and vascular endothelial growth factor pathways. It is hoped that these studies, or studies on other novel agents, will improve the outcome of women with epithelial ovarian cancer in the near future.
- CA Cancer J Clin 2004;54:8-29.
- Epithelial ovarian cancer. In: Hoskins WJ, et al, editors. Principles and practices of gynecologic oncology. New York: Lippincott, Williams & Wilkins; 2000. p. 981-2.
- Natl Cancer Inst Monogr 1975;42:101-4.
- Semin Oncol 2002;29 Suppl 1:3-8.
- Obstet Gynecol 1985;65:568-72.
- Gynecol Oncol 1994;55:s91-6.
- British J Cancer 2003;89. Suppl 3:s3-8.
- N Engl J Med 1996;334:1-6.
- J Clin Oncol 2000;18: 3084-92.
- J Natl Cancer Inst 2003;95:1320-9.
- J Clin Onc 2003;117:3194-200.
- J Clin Oncol 2003;21:136s-44s.
- CA Cancer J Clin 1995;45:88-101.
- Oncologist 2000;5:26-35.
- Lancet 2003;361: 2099-106.
- Proc Am Soc Clin Oncol 2004;22:450s (Abs 5005).
- J Clin Oncol 1997;15:2183-93.
- J Clin Oncol 2001;19:3312-22.
- Gynecol Oncol 2004;92:813-8.
- Gynecol Oncol 2003;90:593-6.
- Gynecol Oncol 2003;88:118-22.
- J Clin Oncol 1998;16:405-10.
- Br J Cancer 2003; 89 Suppl 3:S9-15.
- Gynecol Oncol 1999;72:148-53.
- Gynecol Oncol 2004;93:417-21.
- J Clin Oncol 2003;21:291-7.
- Support Care Cancer 2004;12:550-4.
- Ann Oncol 2003;14:1570-7.
- J Clin Oncol 2003;21:4112-9.
- N Engl J Med 1993;329:1790-6.
- Blood 2002;10:2303-20.
- Cancer 2004;100:859-68.
- Semin Oncol 2004;31 Suppl 8:27-34.
- Cancer Nurs 200;26 Suppl:32s-7s.
- Supportive Oncology 2004;2:357-64.
National Cancer Institute
Society of Gynecologic Oncologists