++
Ovarian cancer is the second most common cancer of the female genital tract, with approximately 21,290 cases expected in the United States in 2015 (1). Epithelial tumors comprise 90% of ovarian cancers, and the most common histologic subtype is high-grade serous carcinoma, followed by endometrioid, clear cell, and mucinous tumors. Ovarian cancer remains the number one cause of death due to gynecologic cancers in the United States, accounting for 14,180 deaths this year. Among women, ovarian cancer is the fifth most common cancer-related cause of death in the United States (1). The lifetime risk of a woman in the United States developing ovarian cancer is approximately 1 in 70 (1.37%). Ovarian cancer is also more common among white women compared to African American or Asian American women in the United States, although the differences are narrowing. In most parts of Europe and North America, the incidence of ovarian cancer was constant during the decades prior to the 1990s. However, among white women, ovarian cancer incidence rates are reported to have declined from 2001 to 2010 by 2.2% per year (2).
++
This cancer is predominantly a cancer of the perimenopausal and the postmenopausal period, with 80% to 90% of cases occurring after the age of 40. The incidence is higher in older women, and the median age at diagnosis is 63 years.
++
Ovarian cancer accounts for 5.5% of the deaths from cancer that occur in women between 60 and 79 years of age (1). Prognosis among women with ovarian cancer is dependent on the stage of disease at the time of diagnosis. Five-year survival rates among women with localized, regional, and distant disease at the time of diagnosis are 92%, 72%, and 27%, respectively (1). Relative survival rates for ovarian cancer have improved substantially over the last decade by an average of 2% per year, and modern 5-year survival estimates are between 45% and 50% (3). Survival among white women with ovarian cancer in the United States is reportedly better than survival among black women, and the improvement noted was not observed in black women (1,4).
++
The etiology and the tissue of origin of ovarian cancer are not fully understood. Over the past decade, there has been an increased appreciation that epithelial ovarian cancers (EOCs) represent a heterogeneous group of malignancies. Some of this heterogeneity is related to distinct pathophysiology associated with the development of different histologic subtypes. For example, the majority of high-grade serous ovarian cancers are now believed to arise from fallopian tube fimbria rather than ovarian surface epithelium. This fallopian tube hypothesis originated from observations in women undergoing risk-reducing (prophylactic) salpingo-oophorectomy due to hereditary breast-ovarian cancer syndromes. Approximately 5% to 10% of these women are diagnosed with an occult ovarian cancer (5,6,7). The majority of these early cancers are either located in the fimbrial portion of the fallopian tube or, on close histologic examination, have a coexisting carcinoma in situ component in the fallopian tube fimbria. Subsequent investigations revealed that careful sectioning of fallopian tubes from high-risk women frequently revealed areas of marked cytologic atypia and disorganized growth within the fimbria. These areas have been called serous tubal intraepithelial carcinoma (STIC) or tubal dysplasia and are characterized by cytologic atypia, positive p53 immunostaining (which correlates with mutations in the TP53 gene), abnormal proliferation (as evident by Ki67 staining), and DNA damage (Fig. 31-1) (8,9,10).
++
++
Based on their distinct molecular features and clinicopathologic characteristics, other carcinogenesis models have been proposed for endometrioid, clear cell, mucinous, and low-grade serous ovarian cancers. Endometrioid and clear cell tumors have a strong epidemiologic link with endometriosis, and there is accumulating evidence that they may arise from endometriotic cysts or areas of atypical endometriosis. Low-grade serous carcinomas are thought to arise from borderline neoplasms. However, this remains a controversial area, and much research is focused on better understanding the pathophysiology of different subtypes of EOC.
++
Numerous studies have attempted to demonstrate possible links between environmental, dietary, reproductive, endocrine, viral, and hereditary factors and the risk of developing ovarian cancer. These factors are summarized in Table 31-1.
++
++
The strongest risk factor for ovarian cancer is a genetic predisposition. Women with BRCA1 and BRCA2 mutations have a 39% to 46% and 10% to 27% lifetime risk of developing ovarian cancer, respectively, which is 18 to 36 times higher than that of the background risk. A smaller proportion of cases of familial ovarian cancer are associated with mutations in mismatch repair genes (MLH1, MSH2, MHS6, PMS2) related to Lynch syndrome, with a lifetime risk of ovarian cancer ranging from 0% to 24%. Patients with germline mutations in MLH1 and MSH2 seem to be at the highest risk for ovarian cancer compared to patients with MSH6 and PMS2 mutations. Most recently, germline mutations in BRIP1, RAD51D, and RAD51C have been associated with an increased lifetime risk of ovarian cancer, ranging from 10% to 15%.
++
Other factors associated with an increased risk of ovarian cancer include age, early menarche, late menopause, and obesity (11,12,13).
++
Protective factors that have been shown to reduce the risk of ovarian cancer include the use of oral contraceptives, multiparity, breast feeding, hysterectomy, and tubal ligation (14). Other factors, including exercise, perineal talc exposure, infertility treatment, and use of postmenopausal hormone replacement therapy, have not been definitively shown to alter a woman’s risk of developing ovarian cancer.
++
Stage I ovarian cancer is associated with excellent survival; however, more than two-thirds of patients are diagnosed with stage III or IV disease. These observations have provided a compelling rationale in support of screening for early-stage disease. The most commonly used screening strategies include a combination of serum cancer antigen 125 (CA125) levels and pelvic sonography. While several large trials have demonstrated that screening can detect cancer in asymptomatic women, there are concerns regarding the poor positive predictive value for these strategies and lack of proven survival benefit. The findings of four key trials are summarized in Table 31-2 (15,16,17,18,19). Currently, the US Preventive Services Task Force recommends against screening for ovarian cancer in asymptomatic women of average risk. However, the United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) is using the Risk of Ovarian Cancer (ROC) time series algorithm to interpret CA125, which has shown an encouraging sensitivity and specificity, and the mortality data are anticipated in 2015.
++
+++
The Molecular Landscape of Epithelial Ovarian Cancer
++
Although the different subtypes of EOC possess unique molecular aberrations (Table 31-3) and transcriptional signatures, their morphologic features resemble the specialized epithelia of the reproductive tract that derive from the Müllerian ducts.
++
++
As noted, accumulating evidence points to the distal fallopian tube epithelium as the tissue of origin for most high-grade serous carcinomas. The most common molecular alterations in serous carcinomas are mutations in TP53, which are nearly ubiquitous. The Cancer Genome Atlas (TCGA) project has also significantly advanced our understanding of other molecular and genetic alterations in high-grade serous carcinoma. In addition to the expected TP53 mutations in 96% of tumors, low prevalence recurrent somatic mutations in NF1, BRCA1, BRCA2, RB1, and CDK12 were also observed. Serous carcinomas are also characterized by a high degree of chromosomal instability (gene copy number amplifications and deletions), and both total and regional instability are associated with tumor grade and altered patient outcomes (20). Somatic copy number analysis performed as part of TCGA also confirmed 8 and 22 chromosomal regions of recurrent gain and loss, respectively. Five of the gains and 18 of the losses occurred in more than half of the tumors. Although such aberrant areas of DNA frequently carry multiple genes, it is presently thought that only a limited number of genes are “key drivers” of the process. These key drivers are thought to be the most critical markers and potential treatment targets. Candidate drivers at areas of copy number gain and loss are frequently proposed. For example, it has been suggested that 45% of high serous cancers harbor altered phosphatidylinositol 3-kinase (PI3K)/RAS signaling mediated by multiple copy number alterations, including PTEN deletion and PIK3CA, KRAS, AKT1, and AKT2 amplification. Low-grade serous carcinomas have been found to have alterations in the mitogen-activated protein kinase (MAPK) pathway. Approximately 20% to 40% of tumors have a KRAS mutation, while a smaller proportion of tumors demonstrate mutations in BRAF, accounting for 5% of low-grade serous carcinomas (21).
++
Clear cell and endometrioid cancers are epidemiologically and molecularly linked to endometriosis. Frequent somatic mutations of PIK3CA and ARID1A (AT-rich interactive domain-containing protein 1A) have been documented in tumors associated with endometriosis (22,23). Common genetic abnormalities identified in endometrioid ovarian carcinomas include somatic mutations of CTNNB1 and PTEN (24).
+++
Cytokines and Growth Factors
++
Several cytokines and growth factors have been studied in ovarian carcinogenesis. For instance, levels of interleukin 10 (IL-10) and interleukin 6 (IL-6) are particularly elevated in ovarian cancer ascites. Endogenously produced IL-6 can protect tumor cells from natural killer cell–mediated killing, and IL-6 expression by immunohistochemistry was associated with poor prognosis (25). Furthermore, IL-6 has been identified as an etiologic paracrine factor in paraneoplastic thrombocytosis and associated poor prognosis in ovarian cancer (26). When compared to high-grade serous tumors, ovarian clear cell carcinomas are associated with higher circulating levels of IL-6 (27).
++
A detailed review of growth factor pathways targeted in ovarian cancer is beyond the scope of this chapter. Instead, we briefly highlight the vascular endothelial growth factor (VEGF) pathway, which has proven to be the most clinically useful target to date. The VEGF signaling cascade is mediate through a partially redundant set of ligands and receptors, which have emerged as promising targets for antiangiogenic cancer therapy. The VEGF ligand family consists of seven ligands: VEGF A-E, placenta growth factor 1 (PlGF1), and PlGF2. The receptor tyrosine kinases involved in this signaling cascade include VEGF receptor type 1 (VEGFR1), VEGFR2, and VEGFR3. Vascular endothelial growth factor ligands are overexpressed in EOC cells, while the receptors are present mainly on the tumor endothelial cells (28). Vascular endothelial growth factor is a key mediator of angiogenesis, which is stimulated by hypoxia. Bevacizumab, a monoclonal anti-VEGF-A antibody, is the prototypical member of this class, and as a single agent has been the most promising biological compound for the treatment of recurrent ovarian cancer.
++
Prognostic factors are tumor-related characteristics that determine the biologic behavior and risk of death from the disease; their predictive value may change during the course of treatment and thereafter.
++
Factors associated with poor prognosis in advanced ovarian cancer (stage III or IV) fall into two subgroups (as determined by multivariate analysis in clinical trials):
++
Variables prior to systemic treatment predictive of survival: age, stage at diagnosis, performance status, residual tumor volume, and tumor histology
Variables at the time of relapse predictive of time to progression: less than 6 months from last chemotherapy (platinum-resistant disease), poorer performance status, mucinous histology, larger number of sites of disease, best previous response to chemotherapy versus progression, serum CA 125 levels
++
Stage is a dominant prognostic factor in ovarian cancer. The main prognostic factors in early-stage ovarian cancer (stages I-IIA) are International Federation of Gynecology and Obstetrics (FIGO) stage, histologic grade, histologic type, and patient’s age. Early-stage ovarian cancer is discovered early in fewer than 30% of patients; in such cases, the 5-year survival is good, ranging from 51% to 98% (1).
++
Cancer antigen 125 is a high molecular weight glycoprotein that is elevated in 80% of EOCs (29). There is no definitive evidence that pretherapeutic CA125 levels correlate with survival in EOC (30). The most aggressive tumors are not necessarily those with the highest CA125 levels. However, there has been evidence that the kinetics of an individual’s CA125 level during treatment may be related to best response to treatment as well as survival (31).
++
It is logical to assume that the extent of postoperative residual tumor volume is affected by both the biology and the history of the disease, as well as the radicality, emphasis, and effort involved in the tumor reductive surgery. What remains controversial is the relative contribution of these factors to the prognostic significance of residual disease. On one hand, the tumors that are more aggressive and disseminated are more difficult to resect and therefore associated with larger residual disease. Therefore, how advanced the tumor was before debulking may be more important than how much disease was left behind. Other features—such as the type of chemotherapy, the intrinsic chemosensitivity of the tumor, and the presence of other biological variables—may be as important as or even more important than the extent of the surgery. Proponents of the importance of maximal surgical effort point to a wealth of retrospective data and the evolution of “optimal cytoreductive surgery” from less than 2 cm to R0 (no visible residual disease) as evidence for the importance of the surgical result (32). The only prospective randomized trial of the neoadjuvant approach versus up-front tumor reductive surgery in patients with advanced-stage EOC was carried out by the European Organization for Research and Treatment of Cancer (EORTC)–Gynaecological Cancer Group. In this trial, optimal resection (as defined by residual tumor of 1 cm or less) was noted in 41.6% of patients who underwent primary debulking compared to 80.6% of patients who underwent neoadjuvant chemotherapy interval debulking (33).
++
Serous carcinoma is the most common histologic subtype of EOC, and this subtype can further be divided into high-grade and low-grade serous carcinomas (34). Recently, MD Anderson developed a two-tier grading system in serous carcinomas based on nuclear atypia and mitotic rate to distinguish high-grade from low-grade tumors (Fig. 31-2). This has been adopted by the gynecologic oncology community to further define these two diseases.
++
+++
High-Grade Serous Carcinoma
++
High-grade serous carcinomas account for 70% to 80% of all ovarian cancers and are the most common type of EOC. High-grade serous carcinomas may present with varying architectural patterns, but the defining characteristic of these tumors is high mitotic activity (>12 mitoses per 10 high-power fields [HPFs]) and the presence of multinucleated cells (34).
+++
Low-Grade Serous Carcinoma
++
Low-grade serous carcinoma accounts for 6% to 10% of serous ovarian cancers and 5% to 8% of all ovarian cancers. These tumors are now thought to arise from borderline tumors have distinct molecular aberrations and clinical behavior when compared to their high-grade counterpart. Low-grade serous tumors have low mitotic activity (<12 mitoses per 10 HPFs) and are distinguished from borderline tumors based on destructive invasion of more than 5 mm (35).
++
In older references, the proportion of mucinous ovarian cancers is as high as 30%. However, there is now increased recognition that many such tumors, on careful evaluation, are thought to represent metastases from the gastrointestinal tract. Contemporary estimates of the prevalence of primary ovarian mucinous carcinoma is approximately 3%. Primary ovarian origin is favored by unilaterality, large size greater than 12 cm, smooth external surface, and association with other ovarian pathology. Conversely, metastases tend to be bilateral, be less than 10 cm in size, exhibit surface involvement, and have colloid and signet ring morphology. True mucinous ovarian tumors are low-grade malignancies that have a low propensity for metastatic spread and are usually diagnosed as unilateral stage IA tumors even though they may reach enormous size (Fig. 31-3) (36). Newer studies have also dispelled the notion that pseudomyxoma peritonei is commonly secondary to an ovarian mucinous carcinoma. It is now recognized that pseudomyxoma peritonei is almost always associated with an appendiceal mucinous lesion (37).
++
+++
Endometrioid Adenocarcinoma
++
Overall, 10% of epithelial ovarian tumors are of endometrioid histology and resemble endometrial adenocarcinoma, and both types occur simultaneously as synchronous primary tumors in as many as 25% of cases in patients younger than 45 years of age (38). Identification of multifocal disease is important because patients with disease metastatic from the uterus to the ovaries have a 5-year survival rate of 30% to 40%. Those with synchronous multifocal disease have a 5-year survival rate of 70% to 80%. Concurrent endometriosis is present in 10% of cases. The malignant potential of endometriosis is very low, although a transition from benign to malignant epithelium may be seen.
++
Clear cell carcinomas comprise 5% to 10% of ovarian cancers and, like endometrioid tumors, may be associated with endometriosis or endometrial cancer. Some studies indicated that clear cell carcinoma may be resistant to standard carboplatin-paclitaxel–based chemotherapy regimens. However, other investigators have shown that when the pathology was carefully reviewed by a gynecologic pathologist, only suboptimal cytoreduction and spread of disease were associated with a significantly increased risk of platinum resistance (39).
+++
Transitional Cell Carcinoma
++
Transitional cell carcinomas were previously thought to represent malignant Brenner tumors; however, recent studies have shown that they are molecularly similar to high-grade serous carcinomas. These tumors are now classified as a subtype of high-grade serous carcinomas (40).
+++
Undifferentiated Carcinoma
++
Undifferentiated carcinomas are thought to represent the most poorly differentiated high-grade serous carcinomas rather than a separate entity (41).
++
Previously, a three-tier grading system was used in the diagnosis of EOC; however, there was no consensus on the definition of this system. More recently, two-tier grading systems (high grade vs low grade) have been shown to have better prognostic and interobserver variability (34,42). Consistent with their distinct pathophysiology and tumor biology, low-grade serous carcinomas are more resistant to chemotherapy and are associated with significantly higher rates of platinum-resistant disease (43).
++
Accurate staging is critical to the success of surgery and adjuvant therapy. The staging of ovarian cancer is based on the gross and pathologic findings of the initial surgical evaluation. The FIGO classification uses the sites and extent of the disease, including capsule rupture and ascites, to categorize ovarian cancer into four stages. This is summarized in Table 31-4.
++
++
Symptoms of ovarian cancer are nonspecific and include early satiety, bloating, constipation, and weight loss. It is not uncommon for patients to have been referred for a gastrointestinal evaluation before the correct diagnosis is reached. Objective signs of ovarian carcinoma are also nonspecific and include a pelvic mass, ascites, carcinomatosis, possible pleural effusion(s), and occasionally supraclavicular lymphadenopathy.
++
In general, the initial management and staging of EOC is surgical. In early-stage ovarian cancer, comprehensive staging allows for proper triage to adjuvant therapy. When comprehensive staging is performed, a substantial number of patients initially believed to have disease confined to the pelvis will be staged upward (44).
++
In advanced EOC, surgery and chemotherapy are both utilized in initial management. However, there remains debate regarding the sequence of these interventions in the treatment of advanced ovarian cancer. Patients have historically been candidates for neoadjuvant chemotherapy if they have multiple medical comorbidities, poor performance status, or extensive disease on imaging that is not felt to be amenable to up-front surgery. Despite this, there is no current consensus regarding which patients should have up-front cytoreduction or neoadjuvant chemotherapy. The current approach at MD Anderson is for all patients with suspected advanced ovarian cancer (based on computed tomographic [CT] imaging) to undergo a preoperative laparoscopic assessment (Fig. 31-4).
++
++
This laparoscopic evaluation provides the following: surgicopathologic diagnosis, assessment of metastatic disease burden and likelihood of complete resection (modified Fagotti score) (45), and research tissue acquisition. Our laparoscopic triage is accomplished by scoring made by two independent and blinded surgeons. Those patients scored less than 8 undergo primary cytoreduction (with up to a 2-week interval). Those with scores of 8 or more undergo neoadjuvant chemotherapy (consisting of three cycles of carboplatin and paclitaxel), followed by consideration of interval cytoreductive surgery and three additional cycles of chemotherapy (± maintenance therapy) (46). An outline of our treatment algorithm is shown in Fig. 31-5.
++
++
In addition to a preoperative evaluation with lab testing, imaging, and office examination, the National Comprehensive Cancer Network (NCCN) now recommends genetic testing for all newly diagnosed high-grade serous ovarian cancers.
+++
Primary Cytoreductive Surgery
++
A staging laparotomy involves the following steps:
++
Midline vertical incision
Evacuation and cytologic analysis of ascites
Inspection and palpation of all peritoneal (intraperitoneal and retroperitoneal) surfaces, including the subdiaphragmatic areas
Total abdominal hysterectomy and bilateral salpingo-oophorectomy
Omentectomy with debulking to no gross residual disease
If disease limited to ovaries, bilateral pelvic and para-aortic lymph node sampling; multiple biopsies, including the paracolonic gutters, cul-de-sac, lateral pelvic walls, vesicouterine reflection, subdiaphragmatic sites, and intra-abdominal areas
Appendectomy if a mucinous tumor is found
++
The goal of an initial cytoreductive surgery is to remove all visible tumor because the amount of residual disease left behind is inversely correlated with patient survival. Currently, an optimal tumor-reductive surgery is defined as no residual nodules greater than 1 cm in size. More recent studies have tried to refine the definition of optimal cytoreductive surgery by dividing patients into those with no visible residual disease and others as these patients have improved overall survival (OS), even compared to patients with less than 1 cm of residual disease (32).
++
The only prospective randomized trial of the neoadjuvant approach versus up-front tumor-reductive surgery in patients with advanced-stage EOC carried out by the EORTC demonstrated the noninferiority of the neoadjuvant approach with respect to progression-free survival (PFS) and OS (33). However, this study has been criticized by the lower rate (41.6%) of optimal cytoreductive surgery in the primary debulking group (which was defined as largest single tumor nodule <1cm in this study) compared to reports from institutions in the United States and elsewhere. Hence, this remains a controversial area in the management of ovarian cancer.
+++
Secondary Cytoreduction for Recurrent Disease
++
The benefits of secondary cytoreduction for recurrent disease are also unclear, although accumulating data suggest that, in certain circumstances, secondary cytoreduction may lead to a survival benefit. Eisenkop et al published a large series of patients with recurrent ovarian cancer undergoing cytoreduction for recurrent disease (47). There were 106 patients who had a complete clinical response to initial therapy. These patients had a disease-free interval of at least 6 months prior to recurrence. Sixty percent underwent reexploration and debulking, and 82% were rendered disease free at that time. The authors evaluated factors that might be predictive of surgical outcome (optimal debulking) as well as those that might be indicative of survival. Predictors of complete cytoreduction included the following:
++
Size of recurrent tumor less than 10 cm
The use of chemotherapy before cytoreduction
Good Karnofsky performance status
++
A study from MD Anderson Cancer Center (MDACC) looked at a similar group of patients who also had a disease-free interval of at least 6 months, with similar results. These investigators noted that there was significantly improved survival in women who underwent optimal cytoreduction of tumor to less than 2 cm (19.5 vs 8.3 months). Others have published similar findings, all noting that the duration of prior clinical response is important in terms of survival and chances of optimal cytoreduction.
++
Previous studies have defined a subset of patients with early-stage disease who are at increased risk of relapse. These high-risk features include:
++
Stage IC disease
Clear cell histology
High-grade tumor
++
Patients with these features should be considered for adjuvant therapy. Two large meta-analyses have shown that adjuvant chemotherapy improved both recurrence-free survival (RFS) and OS in patients with early-stage disease; however, women who had no gross residual disease following surgery did not benefit from adjuvant treatment (48). Subgroup analysis showed that women with high-risk features who received chemotherapy did have improved OS compared to women who received no adjuvant therapy (48). Most recently, the ACTION trial randomized patients with early-stage disease to platinum-based chemotherapy or observation. Overall, adjuvant chemotherapy improved RFS (70% vs 62%); however, there was no improvement in patients who had complete surgical staging (RFS 78% vs 72%) (49). This trial supports the need for complete surgical staging in early-stage ovarian cancer to appropriately triage patients for adjuvant chemotherapy as well as prognostic counseling for the patient.
++
There is no consensus on the number of cycles of adjuvant chemotherapy to give in early disease. The main clinical trial targeting early stage ovarian cancer in the platinum and taxane era was conducted by the Gynecologic Oncology Group (GOG). The goal was to determine whether six cycles of carboplatin and paclitaxel would significantly lower the rate of cancer recurrence, compared to three cycles of the same agents following surgical staging operations on patients with stage IA grade 3, stage IB grade 3, stage IC, and completely resected stage II ovarian epithelial cancer. A secondary objective was to compare the toxicities of the two treatments. Following surgical staging, 321 patients were randomized to either three or six cycles of paclitaxel 175 mg/m2 infused over 3 hours followed by carboplatin 7.5 AUC (area under the curve) infused over 30 minutes. Cycles were repeated every 21 days. A total of 70% of these patients had stage I disease. In the standard three-cycle arm, the estimated probability of cancer recurring within 5 years was 27%, compared to 19% in the six-cycle arm. The recurrence rate for six cycles was 24% lower (hazard ratio [HR] 0.761; 95% CI, 0.51–1.13; P = .18). It was concluded that the addition of three cycles of carboplatin and paclitaxel over the standard three cycles did not significantly alter the rate of cancer recurrence in patients with early-stage ovarian epithelial carcinoma. In addition, six cycles caused significantly more toxicity than three cycles (50). However, potential weaknesses of this study include the fact that it was likely underpowered to result in statistically significant differences, and that toxicity outcomes lacked external validity given that the high dose of carboplatin (AUC = 7.5) used in this trial is rarely used in clinical practice. At MD Anderson, we currently administer six cycles of neoadjuvant chemotherapy if utilized in the adjuvant setting for early-stage disease.
+++
Advanced-Stage Disease
++
Carboplatin and paclitaxel remain the gold standard drugs in primary adjuvant treatment of EOC. The effect of incorporating additional cytotoxic agents (gemcitabine, liposomal doxorubicin, topotecan) in combination with carboplatin and paclitaxel was evaluated in GOG182. This randomized four-arm trial showed equivalent PFS and OS in all the arms and a better toxicity profile in the control carboplatin and paclitaxel alone arm (51). Addition of bevacizumab to carboplatin and paclitaxel in the frontline setting was tested in GOG218 and ICON7 (52,53). Both studies showed that the addition of bevacizumab resulted in a modest improvement in PFS, but without an improvement in OS. Furthermore, the 4-month improvement in PFS observed in GOG218 was only associated with the arm receiving prolonged administration. There was no difference in the PFS of the control and carboplatin paclitaxel plus bevacizumab arms. Incidences of serious bowel toxicity (perforation or fistulas) and hypertension were about 2% and 23%, respectively (52). Given the significant costs of bevacizumab and the lack of OS improvement, cost-effective analysis has shown that the use of bevacizumab as part of frontline therapy in ovarian cancer is not cost-effective (54).
+++
Intraperitoneal Chemotherapy
++
The frequent metastasis of EOC within the peritoneal cavity and the ability to achieve a much higher concentration of platinum and taxane drugs following intraperitoneal administration are the principal rationale for the use of intraperitoneal chemotherapy for the treatment of patients with advanced-stage EOC who have undergone optimal resection (largest residual tumor nodule less than 1 cm). Three randomized trials have revealed improved PFS and OS in patients receiving intraperitoneal chemotherapy (55,56,57). While all three trials have been criticized, the publication of GOG172 results, which revealed a 17-month OS benefit in the intraperitoneal arm, led the National Cancer Institute (NCI) to issue a Clinical Announcement recommending that women with stage III ovarian cancer who undergo optimal surgical cytoreduction be considered for intraperitoneal chemotherapy.
++
However, the better outcomes associated with intraperitoneal chemotherapy are associated with higher toxicity, illustrated by the fact that fewer than half of women randomized to the intraperitoneal arm were able to complete the six prescribed cycles (55). Furthermore, the regimen used in GOG172 included an inpatient administration of day 1 and 2 (for 24-hour paclitaxel and intraperitoneal cisplatin). While most centers have substituted a better-tolerated and outpatient regimen from the original GOG172, legitimate concerns regarding the equal efficacy of such regimens remain. In hopes of arriving at a better-tolerated and more efficacious intraperitoneal regimen, GOG252 was conducted to compare the GOG172-derived outpatient regimen to intraperitoneal carboplatin and weekly dose-dense paclitaxel and standard intravenous regimen (with bevacizumab included in all three arms). The results of this trial are anticipated in 2015.
++
The Japanese GOG protocol 3016 (JGOG3016) compared the administration of carboplatin (AUC = 6) with paclitaxel (180 mg/m2) every 3 weeks or weekly paclitaxel (80 mg/m2 on days 1, 8, and 15 of a 21-day cycle). This study showed a significantly better median PFS and OS in the weekly dose regimen (PFS: 18.2 vs 17.5 months; OS: 100.5 vs 62.2 months) (58,59). However, a peculiarity of this trial was that even the 62.2-month median OS survival in the control arm was much longer than in any of the platinum era GOG or European trials, especially given that the JGOG trial included patients with suboptimal cytoreductive surgery.
++
Additional questions about the external validity of the JGOG3016 trial were raised when the results of the MITO7 became available. This trial compared weekly carboplatin and paclitaxel (AUC = 2 and 60 mg/m2, respectively) to conventional therapy (carboplatin AUC = 6; paclitaxel 175 mg/m2) every 3 weeks. There was no significant difference in the median PFS between the two arms (median PFS 17.3 months in the every 3 weeks group vs 18.3 months in the weekly arm) (60). In addition, preliminary results from GOG262 showed no difference in PFS in patients with stage I to IV ovarian cancer. This study included both optimally and suboptimally debulked patients, but both arms also allowed bevacizumab administration. Interestingly, patients not treated with bevacizumab had a 4-month improvement in PFS if they received the dose-dense regimen, while there was no difference in PFS in patients who received bevacizumab in addition to the assigned treatment regimen (61).
+++
Alternative Chemotherapeutic Agents
++
Docetaxel (Taxotere) is a semisynthetic compound structurally related to paclitaxel. The toxicity of docetaxel is similar to that of paclitaxel. However, docetaxel is associated with less neuropathy and more myelosuppression compared to paclitaxel. In addition, prolonged treatment with docetaxel increases skin and nail toxicity and can produce fluid retention and significant edema. Another important distinguishing feature is that docetaxel does not share the cremophor EL diluent used in the preparation of paclitaxel, which is the etiologic component responsible for many cases of paclitaxel hypersensitivity. A trial conducted by the Scottish Gynaecological Cancer Trials Group included 1,077 patients with FIGO stage IC to IV EOC who were randomized to receive carboplatin in combination with either paclitaxel or docetaxel. The median PFS for both arms was approximately 15 months (62). The authors concluded that the docetaxel combination appears to be a viable alternative to the paclitaxel combination as first-line chemotherapy in EOC because of an improved therapeutic index while maintaining similar efficacy.
++
The high risk of recurrent disease after treatment of advanced-stage EOC has prompted an intensive search for therapeutic strategies that can be given after standard-of-care therapy to improve patient outcomes. More than 12 phase III trials have been undertaken in this setting, including extension of frontline agents, administration of short-duration non–cross-resistant chemotherapy, high-dose chemotherapy, whole-abdominal or intraperitoneal radiotherapy, immunotherapy, vaccine therapy, biologic therapy, and single-agent paclitaxel; however, none has shown a survival advantage against various controls (usually no treatment) (63). The S9701/GOG178 phase III trial that administered paclitaxel intravenously for 12 months (vs 3 months) after an initial response to first-line chemotherapy showed improved PFS. However, the design of the trial (with PFS as the primary end point) led to early closure and lack of data on the effect of this consolidation strategy on OS. This improvement in PFS was associated with significant neurotoxicity, and overall this strategy has not been widely adopted in routine clinical practice (64). The chemotherapy regimens utilized at MD Anderson are outlined in Fig. 31-6.
++
+++
Follow-up and Treatment of Recurrent Disease
++
Despite multimodality therapy, 75% to 80% of women with advanced-stage epithelial cancer will have a recurrence. Regular follow-up with tumor marker can detect disease recurrence (65,66). In women with previously treated ovarian cancer that is in clinical remission, the NCCN has recommended assessment of serum CA125 concentration at every follow-up visit if this concentration was raised at initial diagnosis. After documentation of CA125 increase in such women, the median time to a clinical relapse of ovarian cancer is 2 to 6 months. There is, however, controversy over the benefit of early treatment versus treatment later. The results of a large study showed no survival benefit from early treatment on the basis of a raised serum CA125 concentration alone and therefore questioned the value of routine measurement of CA125 in the follow-up of patients with EOC (66). Some authors also suggested that smaller tumors are more often responsive to treatment, but this does not eliminate the lead-time bias. It is also argued that larger tumors have an inferior primary response and grow rapidly. Recurrent ovarian cancer is a mortal disease, but—in the absence of data showing that treatment improves quality of life—this does not justify haste in treating patients. At MD Anderson, we follow patients with CA125 levels in surveillance but in general do not treat patients based solely on a rising value (see Fig. 31-5).
++
One of the problems with recurrent ovarian cancer is the lack of a truly effective salvage therapy. The other problem is the inability to identify a predictive marker for recurrence. The treatment of recurrent ovarian cancer is stratified by the amount of time from completion of primary platinum-based chemotherapy. Patients who have a recurrence with a PFS of 6 months or longer are defined as having platinum-sensitive disease. Patients who have a recurrence less than 6 months from primary chemotherapy are classified as having platinum-resistant disease. Patients whose disease progresses on primary chemotherapy have tumors classified as platinum refractory and have a poor prognosis (67). A list of treatment regimens utilized at MD Anderson in platinum-sensitive and platinum-resistant disease can be found in Fig. 31-6.
+++
Treatment of Platinum-Sensitive Disease
++
As mentioned, patients who have a treatment-free interval of 6 months or greater prior to recurrence are more likely to respond to repeat platinum treatment. As noted, depending on the pattern and extent of recurrence, some of these patients may also be good candidates for secondary cytoreductive surgery. While there are several options for second-line chemotherapy in platinum-sensitive disease, all regimens have a platinum backbone. Combination therapy has been shown to have an improved OS compared to single-agent therapy in this patient population and should be considered in patients with an acceptable performance status. As long as patients continue to respond to platinum and have more than a 6-month interval between each treatment, it seems reasonable to continue treatment with platinum, either as a single agent or in a combination regimen (see Fig. 31-5).
++
Two phase III trials have shown the benefit of carboplatin and paclitaxel in the setting of platinum-sensitive recurrence, and the results were published together. The ICON4 and AGO-OVAR-2.2 studies randomized patients to receive conventional platinum-based chemotherapy or platinum plus paclitaxel or platinum plus a nontaxane (68). Results from this study showed an OS benefit (HR 0.82, P = .02), which led to a 7% absolute survival difference (57% vs 50%) and a 5-month improvement in median survival (29 vs 24 months) in favor of platinum plus paclitaxel. There were higher rates of neuropathy and alopecia in the platinum-plus-paclitaxel treatment group but a lower rate of myelosuppression (68). One critique of this study is that a large proportion of patients (30%) were not previously treated with taxane therapy, which may have led to an improved response in this patient population. Other options that may be utilized in the setting of platinum-sensitive ovarian cancer include docetaxel as well as the use of weekly paclitaxel, both in combination with carboplatin (69,70).
++
A phase III study (CALYPSO) showed that a pegylated liposomal doxorubicin and carboplatin combination was better than paclitaxel with carboplatin in terms of PFS in relapsed platinum-sensitive cancer (11.3 vs 9.4 months); however, there was no difference in OS (30.7 vs 33 months). The combination of carboplatin plus liposomal doxorubicin led to fewer cases of severe neutropenia, fewer episodes of mild myalgias/arthalgias, and less neuropathy but did lead to more cases of severe thrombocytopenia, nausea/vomiting, hand-foot syndrome, and mucositis. Interestingly, the study combination also led to fewer episodes of carboplatin hypersensitivity reactions (71).
++
A phase III trial was performed by the Gynecologic Cancer InterGroup (GCIG) to evaluate carboplatin alone versus carboplatin plus gemcitabine in patients with platinum-sensitive disease. Treatment with the combination regimen showed an improved PFS (8.6 vs 5.8 months), but no improvement in OS. Not surprisingly, there more toxicities seen in the combination arm, with increased neutropenia and thrombocytopenia. A second trial, OCEANS, was a randomized phase III trial that evaluated carboplatin plus gemcitabine with or without the addition of bevacizumab. Patients received the assigned treatment for 6 to 10 cycles and were then continued on bevacizumab or placebo maintenance until progression. The addition of bevacizumab led to an improved PFS (12.4 vs 8.4 months) with an improved objective response rate (78.5% vs 57.4%; P < .001). As seen with other antiangiogenesis studies, the addition of bevacizumab led to higher rates of hypertension and proteinuria (72).
+++
Poly-ADP Ribose Inhibitors
++
The role for maintenance therapy with poly-ADP ribose (PARP) inhibitors has been evaluated in platinum-sensitive disease. In a randomized trial (Study 19), maintenance therapy with olaparib was given to patients who achieved a complete response to second-line therapy (73). Patients with germline BRCA mutations were found to have improved PFS (11 vs 4 months) with few adverse events (74). While there was no difference in OS, there was a 22.6% crossover rate in this study, which may prevent a true difference from being seen in this study population. While this did have significant results, olaparib is currently only approved by the Food and Drug Administration (FDA) for the treatment of advanced ovarian cancer in patients with germline BRCA mutation who have failed three or more prior lines of chemotherapy in the United States.
+++
Treatment of Platinum-Resistant Disease
++
There are several treatment options for the treatment of platinum-resistant recurrent EOC. In general, single-agent treatment regimens are utilized in this patient population to minimize adverse effects, given incurable disease. While there are several options, no one therapy has been shown superior to others as first-line treatment for platinum-resistant disease. A Cochrane review of three of the most commonly used agents (paclitaxel, pegylated liposomal doxorubicin, and topotecan) showed similar efficacy. Thus, choice of first-line therapy is driven by side effect profiles of each of the therapies.
++
A pegylated liposomal formulation of doxorubicin was first tested in patients with platinum-refractory disease; the resulting response rate was approximately 26%. Compared to topotecan in a phase III setting, liposomal doxorubicin was associated with lower toxicities, including lower rates of grade 3/4 neutropenia and thrombocytopenia, but with equivalent efficacy, with similar response rate (20% vs 17%) and time to progression (22 vs 20 weeks) (75). Based on the data from Gordon et al, the drug was FDA approved for use in ovarian cancer.
++
In a phase III study comparing gemcitabine and liposomal doxorubicin, both drugs had similar overall response rates (6.1% vs 8.3%), PFS (3.6 vs 3.1 months), and OS (12.7 vs 13.5 months) (76). Gemcitabine has been frequently studied in combination with cisplatin. The response rates ranged from 40% to 70%. However, the small number of patients treated as well as their heterogeneity disallows any further conclusion. Furthermore, it is difficult to justify the use of combination chemotherapy in patients without evidence that responses correlated with improvements in quality of life.
++
Paclitaxel is part of the backbone of treatment for advanced ovarian cancer. Several studies have shown activity of paclitaxel in platinum-resistant patients; however, a proportion of these studies were performed prior to the incorporation of paclitaxel in primary adjuvant therapy. Most recently, weekly paclitaxel produced a 21% response rate in patients with platinum-resistant disease. Not surprisingly, the main toxicity associated with this regimen is peripheral neuropathy (77).
++
Compared to paclitaxel in patients with refractory ovarian cancer, topotecan was found to produce a response rate of 20%, compared to the 13% in patients who received paclitaxel. This resulted in its approval for use by the FDA. In patients with relapsed platinum-resistant ovarian cancer, the overall response rates on treatment with topotecan ranged from 5% to 18%. The proportion of these patients who achieved stable disease was 17%. In phase III studies, topotecan was shown to have an efficacy equal to both paclitaxel and liposomal doxorubicin as second-line therapy in patients with relapsed ovarian cancer (75).
++
Etoposide, a topoisomerase II inhibitor, has the advantage of being administered as an oral agent. In patients with platinum-refractory disease who were given 100-mg doses of etoposide orally for 14 days every 21 days, the response rate was about 26%. Lower doses of etoposide, at 50 mg/d, produced similar response rates, ranging from 10% to 27%. The combination of cyclophosphamide and bevacizumab has also been shown to be active in this patient population. A phase II trial showed partial response in 24% of patients, with 56% of patients predicted to be alive and progression free at 6 months (78).
++
The phase III AURELIA study evaluated bevacizumab in addition to single-agent chemotherapy in platinum-resistant ovarian cancer. Patients with disease that progressed less than 6 months after completing adjuvant therapy were eligible, but platinum-refractory patients were excluded. The patients received the treating physician’s choice of liposomal doxorubicin, weekly paclitaxel, or topotecan and were then randomized to chemotherapy alone or chemotherapy plus bevacizumab. Patients who received bevacizumab had an improved response rate (31% vs 13%) and a reduction in risk of disease progression (6.7 vs 3.4 months). There was no difference in OS; however, patients were allowed to cross over to receive bevacizumab following progression on chemotherapy alone, which may mask the survival advantage (79). Based on the results of this trial, bevacizumab is now FDA approved for use in combination with chemotherapy for the treatment of platinum-resistant recurrent EOC.
+++
Low-Grade Serous Ovarian Carcinoma
++
The primary management of low-grade serous ovarian carcinoma, like its high-grade counterpart, is surgical with comprehensive surgical staging. However, low-grade serous carcinoma is relatively chemoresistant when compared to high-grade serous carcinoma (43,80). Despite this, carboplatin and paclitaxel remain first-line adjuvant therapy in advanced disease as there is currently no better alternative. Recurrent disease can be treated with surgery, chemotherapy, hormone therapy, and targeted therapies, including MEK inhibitors. In a retrospective review of recurrent low-grade serous carcinoma, 78% of patients who had secondary cytoreduction had no gross residual disease at the conclusion of surgery, which translated into an improved PFS compared to patients who were left with gross residual disease (81). Similar to the adjuvant setting, low-grade serous carcinoma is also chemoresistant in the recurrent setting, with a response rate of less than 4%; however, 60% of patients did have stable disease (82). Numerous hormonal agents have been utilized in the treatment of recurrent low-grade serous carcinoma, with response rates that are only slightly improved compared to chemotherapy. Retrospective series have shown an overall response rate of 9%; however, similar to chemotherapy, 62% of patients achieved stable disease (82). Given the frequency of KRAS and BRAF mutations in this histologic subtype, a single-arm phase II trial of selumetinib, a MEK 1/2 inhibitor, was performed in recurrent low-grade serous carcinoma. The objective response rate was 15%, and 65% of patients had stable disease. Interestingly, there was no correlation between KRAS and BRAF mutational status and response to the study drug. Current trials are ongoing to evaluate therapy with MEK inhibitors in comparison to chemotherapy and hormonal therapy in the recurrent setting.