Ovarian cancer is notorious for its late presentation. It has been called ‘the silent killer’ because the majority of women with ovarian cancer present very late, that is, they already have disseminated disease by the time they are first seen by a doctor. Diagnosis is often delayed because of the wide variety of symptoms experienced by these patients. Ovarian cancer can present as a result of symptoms or may be detected by screening.
Symptomatic Presentation
The range of symptoms associated with ovarian cancer is wide and nonspecific. Symptoms may be related to the presence of a pelvic mass, such as urinary frequency, pain, and constipation; or to other intra-abdominal disease including disease on or invading the bowel, such as rectal bleeding or altered bowel habit; or to the presence of ascites, leading to abdominal distension; or to general symptoms related to cancer, such as nausea, vomiting, anorexia, and cachexia. Although it is important for all physicians to have a high index of suspicion, given that the variety of nonspecific symptoms is so great, it is not surprising that diagnosis is often delayed and patients are often referred to general surgical teams and other specialists before referral to a gynecologist.
Pelvic examination is useful in identifying the presence of pelvic masses, but it does not provide information about the nature of such a mass, so imaging is required for further evaluation. Transvaginal ultrasound remains the most useful modality because it is cheap, well-tolerated, and provides useful information about the contents and nature of a pelvic mass or the presence of free fluid in the pelvis. Transvaginal ultrasound is often combined with transabdominal ultrasound to identify intra-abdominal metastases. Other imaging modalities include computerized tomography (CT) scan and magnetic resonance imaging (MRI), but overall these confer little advantage over ultrasound.
Since its identification in 1983 (Bast et al., 1983), serum levels of CA 125 have been used widely in the diagnosis and monitoring of ovarian cancer. CA 125 is a protein produced by mesodermally derived tissues and for this reason is relatively nonspecific; CA 125 levels can also be raised in nonmalignant ascites and cardiac failure, as well as pregnancy and benign conditions including endometriosis and fibroids. Furthermore, some ovarian cancers, particularly early-stage and mucinous tumors, do not produce CA 125.
Nevertheless, CA 125 remains the most useful tumor marker for ovarian cancer, and the test is commonly performed alongside carcinoembryonic antigen (CEA), levels of which help to differentiate ovarian tumors from colorectal malignancies.
There is increasing evidence that patients treated by specialist gynecological oncologists have a better prognosis than patients who undergo surgery by general gynecologists or general surgeons ( Junor et al., 1999). For this reason, there is much interest in devising mechanisms for improving preoperative diagnosis in order that patients who have ovarian cancer are referred to specialist gynecological oncologists prior to surgery. Using a combination of the patient’s menopausal status, the results of ultrasound, and the CA 125 level, an algorithm has been devised (the ‘risk of malignancy’ index) that can predict the presence of invasive ovarian cancer ( Jacobs et al., 1996). Such algorithms have been adopted by many centers in the United Kingdom, with a reasonable degree of success (Kirwan et al., 2002).
Screening
The concept of screening for ovarian cancer has attracted much interest, given that the disease tends to present clinically at such a late stage, when it has a very high fatality. Unfortunately, our knowledge of the natural history of the disease is limited, and there is no evidence to support the presence of a latent or premalignant phase of the disease. Both of these factors limit the ability to develop an effective screening program.
Efforts have nevertheless been made to develop a suitable screening test. Those that have attracted the most attention are transvaginal ultrasound and serum CA 125. In isolation, these tests tend to have low specificity, which would lead to large numbers of women without cancer being investigated with surgery. This has led to the development of a multimodal approach in which patients are initially screened with a CA 125 test and only subjected to ultrasound when the CA 125 measurement is abnormal. Pilot studies of this approach have been encouraging, and a large multicenter trial is currently under way in the United Kingdom to assess whether screening can be effective in reducing mortality from this disease.
Despite the lack of evidence to support the use of these tests in a mass screening program, they have tended to be offered to women who are thought to be at high risk of developing ovarian cancer, because of a strong family history. In this context, both CA 125 measurement and transvaginal ultrasound are often offered, on an annual basis.
Primary Management
Surgery
The primary treatment for epithelial ovarian cancer has traditionally been surgical. It is based on the principle of cytoreductive or debulking surgery. At the time of surgery, disease is often disseminated throughout the peritoneal cavity. This makes it impossible to perform a complete removal of the tumor with a clear margin of normal tissue, often considered to be the gold standard for surgical oncology. Several studies have demonstrated that patients’ overall prognosis is improved if most of the disease is removed at surgery compared with patients in whom it is impossible to remove disease. This has led to the concept of optimal surgery, often defined as residual disease less than 1–2 cm in greatest diameter.
More recently, the concept of cytoreductive surgery has been questioned: Is the altered outcome following cytoreductive surgery merely a reflection of the biology of the disease? That is, do patients who have had optimal cytoreductive surgery do better than patients who are suboptimally debulked, not because their disease has been removed but simply because their disease can be removed?
This question has not yet been fully answered, but there are interesting data emerging regarding the concept of ultraradical surgery. If one accepts the argument that outcome following surgery is principally related to the biology of the disease, then more aggressive surgery and removal of more disease should have no effect on overall outcome (Bristow and Berek, 2006). However, data from Eisenkop et al. (2003) have demonstrated that an ultraradical approach, that is, removing the maximal amount of disease possible, thereby achieving optimal debulking rates of up to 96%, is associated with a 5-year survival rate as high as 49%. It is important to note that these data are not from randomized studies, but then it is unlikely that a randomized study could or will ever be carried out to address this question.
Surgical intervention has also been used as an ‘interval’ procedure, that is, part way through a course of chemotherapy. Evidence from randomized studies suggests that for patients who are suboptimally debulked at initial surgery and respond to chemotherapy, there is a prognostic benefit from undergoing interval surgery followed by further chemotherapy (van der Burg et al., 1995).
Laparoscopy has gained in popularity in the management of ovarian cancer, principally as a diagnostic tool, but also for prophylactic surgery in women at high risk of developing ovarian cancer as a result of a genetic predisposition.
Chemotherapy
While surgery is the primary treatment modality for ovarian cancer, chemotherapy is used in two settings in the initial management of this disease: first as adjuvant treatment following surgery, and second as neoadjuvant or primary treatment in those patients for whom surgery is inadvisable because of significant comorbidity or extensively disseminated disease.
The most effective form of chemotherapy is based on platinum drugs. It was first used in the 1970s in the combination of cisplatin, cyclophosphamide, and Adriamycin (CAP). After a number of trials questioned whether this three-drug combination was necessary, the ICON-2 trial confirmed that single-agent carboplatin was indeed as effective as the combination, and caused significantly less toxicity (ICON, 1998). In the 1990s, the addition of paclitaxel to the regimen was found to improve efficacy still further in the GOG-111 and OV-010 trials (McGuire et al., 1996; Piccart et al., 2000). These trials both used the combination of cisplatin with paclitaxel, but a number of later trials comparing this regimen with carboplatin/paclitaxel have found the two to be equivalent in efficacy, with the carboplatin regimen producing significantly less toxicity. It is important to note that two other studies looking at the addition of paclitaxel to platinum-based regimens, ICON-3 and GOG-132, did not show a survival advantage from the addition of paclitaxel (ICON, 2002; Muggia, 2000). One reason proposed to explain this is the extent of crossover to taxane-based regimens in the control groups of these studies. Another possible explanation is that paclitaxel may reduce the response to the platinum agent if given at the same time. Indeed, in the GOG-132 trial, patients treated with the combination regimen experienced a myeloprotective effect compared to those treated with single-agent cisplatin. It is suggested that this may be because paclitaxel was attenuating the effects of the platinum.
Proponents of this theory suggest that the optimal regimen may be sequential treatment with a platinum agent and a taxane rather than simultaneous treatment with the two drugs (Moss and Kaye, 2002). As well as comparing the efficacy of alternative platinum drugs, the use of alternative taxanes has been studied. The SCOTROC trial compared paclitaxel/carboplatin with docetaxel/carboplatin and found that the docetaxel regimen had a similar efficacy to paclitaxel/carboplatin with less severe neurotoxicity. However there was more neutropenia in this group (Vasey et al., 2004). In addition, there are ongoing and recently closed trials of new combinations of agents such as the five-arm ICON-5 trial, which looked at gemcitabine, liposomal doxorubicin, and topotecan in various combinations and sequences, compared to standard carboplatin and paclitaxel.
The trials mentioned above have all looked at patients with advanced ovarian cancer. The ICON-1 and ACTION trials looked at the effect of adjuvant chemotherapy in high-risk early-stage ovarian cancer and found a survival advantage for adjuvant treatment in these patients also (ICON, 2003; Trimbos et al., 2003).
In an effort to try to improve the outcome in patients with ovarian cancer, intraperitoneal chemotherapy and maintenance chemotherapy have also been investigated. Intraperitoneal chemotherapy has been extensively studied in the past. Three large randomized trials have found a survival benefit in women treated with intraperitoneal chemotherapy compared with those treated with intravenous chemotherapy. However, it is associated with high levels of gastrointestinal and systemic toxicity. A GOG Phase II study is due to begin in 2007 to investigate whether less toxic regimens can be given with the same efficacy (Ozols, 2006). The use of maintenance chemotherapy has also been investigated, with extended cycles of induction chemotherapy and prolonged administration of a single chemotherapeutic agent, but no improvement in survival has been found (Ozols, 2006).
The current standard regimen following debulking surgery is therefore paclitaxel and carboplatin, given intravenously every 3 weeks for a total of six cycles. Combination chemotherapy is generally offered to patients with FIGO Stage Ic disease or worse.
Chemotherapy would usually be discussed with those patients with Stage Ia or Ib disease with poor prognostic features, in the form of single-agent platinum treatment. The CA 125 level is measured at the start of each cycle of chemotherapy. If this falls to within normal levels within the first three cycles, there is an improved prognosis. If, at the end of six cycles, the CA 125 has fallen but not completely normalized, consideration may be given to treating with a further two cycles of chemotherapy, although there is no published evidence showing that this is of benefit.
Neoadjuvant chemotherapy is currently used in patients who are not fit to undergo surgery, or in patients with extensive disease, to try to downstage the disease so that optimal debulking surgery can be performed. A recent study of neoadjuvant chemotherapy found a higher tumor resection rate and a survival advantage in women who were treated with chemotherapy in the first instance (Kuhn et al., 2001). The use of neoadjuvant chemotherapy is therefore currently being investigated in a randomized UK study (CHORUS) in which patients who are healthy enough and suitable for either chemotherapy or surgery are randomized to primary treatment with one of these modalities. Those receiving primary surgery subsequently receive adjuvant chemotherapy, while those receiving primary chemotherapy (usually with carboplatin and paclitaxel) undergo interval debulking surgery midway through their chemotherapy course.
Management Of Relapse
Over 75% of women with ovarian cancer have advanced disease when they are first diagnosed. While the disease is very chemosensitive and the majority of patients achieve complete remission following surgery and chemotherapy, cancer will recur in most patients (Ozols, 2006). For this reason, following completion of chemotherapy, patients are routinely followed up on a regular basis with physical examination and recording of their CA 125 level. In patients with recurrent ovarian cancer, the aim of treatment is palliation of symptoms rather than cure. There is debate about whether measuring CA 125 levels is of any value in the asymptomatic patient. Currently the OVO5 study is investigating this by double-blinded estimation of CA 125 levels. At present, if a patient is symptomatic and/ or shows a rapidly rising CA 125, relapse is confirmed by CT scan. The initiation of second-line treatment would usually be on the development of symptoms. However, the response to treatment is affected by tumor bulk and the performance status of the patient, and it may be that by waiting for symptoms to develop, tumor bulk has increased and performance decreased such that response is adversely affected. Equally, starting treatment too soon may cause significant toxicity in a patient who had previously been asymptomatic.
The choice of second-line agent depends largely on the woman’s response to her original chemotherapy, the length of time since it was completed, and the toxicities experienced, as well as her current performance status and preference. Treatment for relapsed disease may continue intermittently for months or even years, and the effect that the treatment will have on quality of life is therefore even more important and must be carefully considered.
Patients who relapse more than 6 months after their initial platinum regimen are said to be platinum-sensitive and should be rechallenged with carboplatin with or without paclitaxel. The ICON-4 trial showed an improved 2-year survival after further treatment with the combination regimen compared with carboplatin alone (ICON/AGO, 2003), but toxicity is greater with the combination, and in the palliative setting this needs to be borne in mind. In patients who are platinum-resistant (relapse less than 6 months after their original platinum regimen) or platinum-refractory (had no response to the original platinum regimen), there is no standard second-line treatment. A number of treatments are known to have activity (Moss and Kaye, 2002). Liposomal doxorubicin and topotecan have been recommended by NICE for second-line treatment in this group of patients. Tamoxifen may also be considered (Williams and Simera, 2001). Paclitaxel should be considered if it was not used in the initial regimen prior to relapse. Other agents with activity include gemcitabine and oral etoposide. The response to second-line agents is poor – of the order of 15–20%. Therefore, patients should also be considered for entry into Phase I and II trials of investigational agents.
Although the treatment of relapsed ovarian cancer is essentially with chemotherapy, both surgery and radiotherapy do have a role to play. Surgery is used for the excision of bulky disease or for the removal of lesions causing specific symptoms. In addition, in the patient with a long disease-free interval, further debulking surgery may be an option. Ovarian cancers are generally not radiosensitive, and therefore radiotherapy is only indicated for the treatment of specific symptoms such as vaginal bleeding or pain.
Palliation
Given that 75% of women with ovarian cancer will die of their disease, it is important to recognize that most patients will require palliative input at some point in their care. In addition to pain, one of the most common problems encountered is bowel dysfunction, with symptoms such as nausea and vomiting, but in many cases it may develop into bowel obstruction.
Intermittent, partial obstruction can often be managed conservatively, and treatment centers on control of nausea and vomiting and promotion of bowel motility using antiemetics such as metoclopramide. Conservative management of complete obstruction utilizes antispasmodics and antisecretory drugs such as hyoscine butylbromide and octreotide, respectively. Surgery is generally reserved for those patients who have a discrete level of obstruction in whom a defunctioning or bypass operation can be performed. It is important to note that such defunctioning surgery does not increase longevity but merely acts as a palliative procedure and is associated with a high perioperative morbidity and mortality. Nevertheless, some patients will benefit from this, allowing them to be discharged from hospital and managed at home.
Prevention
Combined oral contraception has been shown to be protective against the development of ovarian cancer, and it has therefore attracted attention as a possible preventative treatment. Unfortunately, because of the increased risk of breast cancer in users of combined oral contraception, it cannot be recommended even for the population at greatest risk, namely those patients with mutations in the BRCA genes.
Similarly, prophylactic surgery is often considered for these high-risk patients. Salpingo-oophorectomy can be performed laparoscopically, as a day case, with minimal morbidity. The subsequent risk of developing ovarian cancer is hugely reduced, and, if performed before the menopause, removal of the ovaries and fallopian tubes (salpingo-oophorectomy) has been shown to reduce the risk of developing breast cancer. There is still great controversy regarding the timing of surgery, particularly because surgery will render the woman menopausal and the use of HRT in this group is controversial (Griffiths et al., 2005).
Summary And Conclusions
Epithelial ovarian cancer remains a significant health problem with incidence and mortality rates that have barely changed over the last 40 years, partly due to the late presentation of the disease. Nevertheless, some progress has been made, not least in the area of chemotherapy. The increase in number of active drugs has allowed patients to live longer with their disease, although ultimately the majority of patients will still succumb to the cancer.
Challenges for the future lie in developing our ability to predict chemo sensitivity to the different agents available and therefore to be able to individualize therapies for patients. Longer-term strategies are to detect the disease earlier and ultimately, of course, to prevent the development of the cancer in the first place.
Bibliography:
- Auersperg N, Wong A, Choi K, Kang S, and Leung P (2001) Ovarian surface epithelium: Biology, endocrinology, and pathology. Endocrine Reviews 22: 255–288.
- Bast R, Klug T, and St John E (1983) A radioimmunoassay using a monoclonal antibody to monitor the course of ovarian cancer. New England Journal of Medicine 309: 883–887.