Preventive Strategies in Epithelial Ovarian Cancer

Despite advances in surgery and chemotherapy, epithelial ovarian cancer (EOC) remains the most lethal gynecologic malignancy 1. Due to lack of a specific prodromal symptomatology as well as effective screening strategies, the vast majority of EOC patients present with advanced stage disease will ultimately die from their disease 1. Furthermore, while up to 80% of patients will respond to conventional primary platinum/taxane chemotherapy, greater than 60% will experience disease recurrence, and current reports indicate discouraging response rates of 20% in women with resistance to platinum-based chemotherapy 2. Preventive strategies, including improvements in early detection of disease as well as in preventing disease recurrence, are, therefore, crucial to improving prognosis. Ovarian cancer prevention can be defined by two main strategies: 1) early detection of cancer in at-risk patients and 2) prevention of recurrent disease in patients with an established diagnosis of cancer. Through the use of screening tools, such as serum biomarkers and medical imaging, early disease detection offers the promise of identifying cancer while still localized and potentially curable 3. Secondary preventive approaches aim to maintain patients without evidence of active disease and thereby extend their disease-free survival. Surveillance methods including serial biomarker measurements as well as therapeutic vaccinations have been examined for their impact on survival outcomes. Finally, risk stratification is critical to the success of any cancer prevention strategy; capitalizing on risk-reducing behaviors and intensive screening is most likely to improve individuals at greatest risk for disease-related morbidity and mortality. Current research in the early detection of ovarian cancer largely focuses on biomarker discovery, using transcriptome analysis, proteomics, epigenomics, metabolomics and glycomics of differentially expressed molecules between women with disease and healthy controls. Biomarkers already approved by the FDA (i.e., CA125 and HE4) or those under investigation, including osteopontin, MUC1 and mesothelin, offer hope for women at risk for disease development, especially those with predisposing genetic mutations 4. As part of this effort, we have generated site-specific biotinylated recombinant antibodies secreted by yeast (Biobodies 5) to costand time-effectively generate antibodies for developing screening tools for large populations of women 6. In addition, biomarkers, especially tumor associated antigens, may also serve as targets for vaccination 7.

Immune-driven therapies are currently under investigation for the prevention of ovarian cancer recurrence [8][9][10][11][12] .Therapeutic vaccinations, targeting molecules specific to an individuals' disease through the use of whole tumor lysates and tumor-pulsed dendritic cells, are currently under investigation for women with recurrent disease; such immunotherapeutic strategies are an additional research interest in our group (NCT00683241, NCT01132014, NCT00603460) 13,14 .Preventive approaches targeting individuals at risk for ovarian cancer as well as those with advanced stage disease may significantly impact disease incidence and prognostic outcomes.In this chapter, we will discuss these current approaches in detail.

Challenges in ovarian cancer prevention
In 1968, the World Health Organization (WHO) established guidelines for disease screening, including that the screened condition should be an important health problem with available treatment 15 .Ovarian cancer arguably satisfies these principles as it ranks as one of the top ten most common cancers amongst women in the US with more than 21,000 diagnosed annually 1 .Further, ovarian cancer is the fifth most common cause of cancer mortality and remains the most lethal gynecologic cancer 1 .Platinum/taxane chemotherapy is available for women with this disease, and approximately 70-80% will respond to this regimen 1 .However, more than 75% of women are diagnosed when disease has already spread from the ovary, and advanced stage disease at presentation carries an overall poor prognosis 1 .Improved ovarian cancer screening methods are therefore needed to detect disease in its earliest stages when treatment is more effective, translating into improved overall five year survival rates ranging from 60% to 90% 3,16,17 .The prevention strategy applied in cervical cancer demonstrates that successful disease screening significantly diminishes disease-related morbidity and mortality.The understanding of the natural history of cervical cancer led to the introduction of screening cervical cytology via Papnicolaou smears and guidelines for the early detection of preneoplastic cervical lesions.Since the introduction of these strategies, the incidence of cervical cancer has declined by more than 75% 18 .Furthermore, vaccination against the oncogenic Human papillomavirus (HPV) will also aid in eliminating this disease.However, preventive strategies in ovarian cancer, unlike those in cervical cancer, have been met with several challenges.First, compared to cervical cancer which ranks as the second most common gynecologic cancer worldwide, ovarian cancer has a low prevalence with 40 cases per year per 100,000 women over the age of 50 years; this mandates that an effective screening test for ovarian cancer has both a high sensitivity and specificity in order to significantly impact disease incidence 19 .Second, current screening methods for early detection of ovarian cancer, including routine physical examination, CA125 serum assessment, and transvaginal ultrasound, have high false-positive rates and low positive predictive values (Table 1) 20 .In fact, for a positive predictive value (PPV) of 10%, an ovarian cancer screening test would require a sensitivity of at least 75% and a specificity of greater than 99% 22 .Further, current methods of screening have not resulted in a significant impact on disease morbidity or mortality 21 .While it is known that persistent HPV infection is responsible for virtually all cervical cancer and its immediate precursors worldwide 23 , the exact etiology for ovarian cancer remains largely debated.Precursors for ovarian cancer should be "morphologically recognizable lesions that are reproducible thereby permitting early clinical intervention" 24 .It has been generally accepted that ovarian cancer originates from the ovarian surface epithelium (OSE) or from postovulatory inclusion cysts, and one hypothesis is that incessant ovulation is the main pathogenic mechanism 25,26 .Yet, recent evaluation of pathologic specimens has also suggested that a greater proportion of "ovarian" cancers may actually originate in the fimbriated end of the fallopian tube with metastasis to the ovary 26,27 .Further, a dualistic classification system has been proposed in which ovarian cancers are divided into two groups: type I tumors which consist of low-grade neoplasms and type II tumor which are aggressive and progress rapidly 28 .Precursor lesions, including borderline malignant tumors and endometriosis, have been identified for type I tumors and may serve to improve early detection of these ovarian tumors especially given their indolent nature; a slower transition time between early and later stage of disease may afford opportunities to detect disease when it is still localized to the ovary.However, type II tumors do not have well-characterized precursor lesions, which is perhaps due to their high level of genetic instability 24 .Because the transition time between stage I and stage III is unknown, it is uncertain whether these tumors rapidly progress from an early stage to an advanced stage or whether these tumors develop as a result of a diffuse peritoneal process 19 .At this point in time and despite a large body of work, no consensus has been reached regarding ovarian cancer precursors, which contributes to the challenge of creating an effective preventive strategy for ovarian cancer.Finally, screening can carry some significant disadvantages, including an increased cost to society for over-utilized medical resources as well as psychological stress/anxiety especially in cases of false positive screening resulting in unnecessary operative intervention for benign pathology.However, thanks to stratifying approaches based on reproducible risk factors enabling maximized efficiency and balanced cost-effectiveness, this last hurdle may be easier to overcome.

Available modalities in prevention of late stage ovarian cancers and of disease recurrence
The goal of preventive strategies is to reduce ovarian cancer-related morbidity and mortality.Disease screening aims to detect ovarian cancer while it is still confined to the ovary and the five-year survival rates are 80-90% 1 , thus to prevent incurable, late stage disease.Disease surveillance following conventional adjuvant chemotherapy allows for early detection of recurrent ovarian cancer and therefore permits prevention of clinically apparent recurrence.Current screening modalities include symptom recognition, bimanual exam, serial CA125 levels and pelvic ultrasound, while disease surveillance typically relies on physical exam, CA125 levels and imaging.

Do symptoms correspond with the onset of disease or with recurrence?
Ovarian cancer is referred to as the "silent killer" due to non-specific symptoms which often go unrecognized until the disease has significantly spread.Although there is limited data to support symptomatology as a sole screening modality for ovarian cancer, recognition of ovarian cancer symptoms by both patients and caregivers may help to identify individuals with early stage disease 29 and in 2007 the Gynecologic Cancer Foundation, American Cancer Society and Society of Gynecologic Oncologists issued a consensus statement supporting the recognition of symptoms as a modality in the evaluation of ovarian cancer 30 .
Patient symptoms have been correlated with the onset of disease [31][32][33] .Symptoms commonly attributed to ovarian cancer include abdominal bloating, increased abdominal size and urinary symptoms 32 .In a case-control study of women at risk for developing ovarian cancer symptoms, specifically pelvic/abdominal pain, urinary urgency/frequency, increased abdominal size/bloating and difficulty eating/early satiety, were significantly associated with ovarian cancer when they occurred more than 12 days per month for less than one year duration 31 .Further, a symptom index was more sensitive in women with advanced stage disease (79.5% vs. 56.7%early stage disease) and more specific in women greater than 50 years of age (90% vs. 86.7%for women less than 50 years old).The authors also applied this symptom index to a sample of 1709 women at average risk and reported a positive screening rate of 2.6%.
In a large population-based study 34 , Rossing and colleagues reported a positive symptom index in 62.3% of women with early stage disease compared to 70.7% with late-stage disease and 5.1% of controls.While symptoms were more likely to occur in women with ovarian cancer, there only was a short interval (less than 5 months) from symptom onset to diagnosis.This suggests that the symptom index may not provide a critical help to diagnose early stage ovarian cancer.In addition, the PPV of the symptom index was approximately 1%; thus the use of a positive symptom index alone would only result in the diagnosis of ovarian cancer in 1 out of 100 women in the general population presenting with the same symptoms.
Further complicating this screening technique is the fact that symptom presentation and duration may be influenced by the histological subtype of ovarian cancer 35 .In a recent population-base study, women with serous histology (the major histologic subtype ) were less likely to report symptoms, were more often diagnosed at advanced stage (compared to mucinous tumors) and had a shorter duration of symptoms compared to women with early stage disease.This study also further highlights the difficulty in diagnosing ovarian cancer at an early stage due to rapid progression of disease.Finally, monitoring symptoms in women with established ovarian cancer has also been considered for early intervention for disease recurrence.However, in a recent systematic review 36 , approximately 67% of a patients identified with recurrent disease had no concurrent clinical symptoms.Other surveillance modalities, including clinical examination, biomarker determination and imaging, should therefore be used in conjunction with symptoms in order to diagnose recurrent ovarian cancer.

Can bimanual examination diagnose early stage ovarian cancer and/or recurrent disease?
Routine pelvic examination is a key component of annual gynecologic health assessment.Palpation of the uterus and ovaries by bimanual examination may allow for the earlier detection of ovarian cancer; exam findings may initiate further evaluation with ultrasound and ultimately surgery, potentially detecting cancers before they become clinically evident.Further, a pelvic exam has little adverse consequences 37 .However, pelvic examination is generally recommended only for the evaluation of symptomatic patients and only in conjunction with ultrasonography 38 .Routine pelvic exam is considered as being neither a sensitive nor a specific means for detecting ovarian cancer in asymptomatic women 39,40 and may thus result in unnecessary surgical intervention for benign ovarian lesions.Further supporting this view, bimanual examination, which was originally included in the screening protocol of asymptomatic, postmenopausal women in the Prostate, Lung, Colorectal and Ovarian Screening Trial of the National Cancer Institute (NCI), was eliminated as a screening modality from the trial as it became evident that it failed to detect the first onset of ovarian cancer 41 .
In contrast, pelvic examination is recommended for disease surveillance of ovarian cancer per the NCCN guidelines, as 26-50% of recurrences occur within the pelvis 42 .Physical examination is an inexpensive, safe and practical tool that can trigger further evaluation with other modalities, but it must be kept in mind that the detection rates of recurrent ovarian cancer vary widely 43,44 and physical examination may fail to detect common sites of recurrence, including the upper abdomen, the retroperitoneum and the thorax 45 .

How effective is ultrasound in detecting early stage ovarian cancer?
Pelvic ultrasound has been utilized for predicting the likelihood of malignancy, especially in women with a known pelvic mass.Transvaginal ultrasound (TVUS) can detect changes in ovarian size and morphology and is superior to physical examination in evaluating ovarian size, especially in women who are postmenopausal, obese or who have an enlarged uterus 46 .
Primary screening studies with TVUS in both asymptomatic and symptomatic at-risk women have been successful in identifying early stage ovarian cancers [47][48][49][50] .
Ovarian volume is inversely related to age; thus an enlarged ovary in post-menopausal women can be a sign of an evolving ovarian cancer.Mean ovarian volume is significantly greater in premenopausal women compared to postmenopausal women 51 ; the upper limit of normal ovarian volume is 20 cm 3 and 10cm 3 in premenopausal and postmenopausal women, respectively.Other ovarian characteristics, including complex or solid morphology, cyst papillations, septae and increased blood flow, have also been suggested as findings suspicious for malignancy 52,53 .
To decrease the number of false-positive results, morphology scoring indices have been introduced for ovarian cancer screening.Investigators at the University of Kentucky have developed a morphology scoring index based on ovarian volume, wall structure and septal structure as a means to improve the PPV of TVUS for ovarian cancer screening 54 .In a multiinstitutional sample of patients undergoing surgical intervention for ovarian tumors, this morphology index implemented during preoperative ultrasound evaluation, yielded a sensitivity of 89%, a specificity of 73%, and a positive predictive value of 46% 55 .The International Ovarian Tumor Analysis (IOTA) study has also provided a reproducible standardized methodology for the ultrasound evaluation of adnexal masses and has further identified features with increased risk of malignancy: the presence of an irregular solid tumor, the presence of papillary or solid components, the presence of ascites, an irregular multilocular solid tumor and the presence of pronounced blood flow 53 .Prospective validation of these simple ultrasound rules in a sample of women with adnexal masses yielded a sensitivity of 95%, a specificity of 91%, positive likelihood ratio of 10.37 and negative likelihood ratio of 0.06

www.intechopen.com
Ovarian Cancer -Clinical and Therapeutic Perspectives 20 56 .This study has also demonstrated that although pattern recognition of ultrasound findings by an experienced examiner can not only reproducibly discriminate between benign and malignant adnexal masses, it is superior to serum CA125 57 .
The United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) is a randomized control trial evaluating the effect of screening on mortality 58 .Patients are randomized to screening with CA125 and transvaginal ultrasound or with transvaginal ultrasound alone.At the prevalence screen, the results were promising for transvaginal ultrasound, which yielded a sensitivity, specificity and PPV of 75.0%, 98.2%, and 2.8%, respectively, for primary invasive epithelial ovarian and tubal cancers 58 .The impact of ultrasound screening on mortality is still pending at this time.
Ultrasound has also been examined for its role in the detection of ovarian cancer recurrence 45 .While sensitivity ranges 45-85% and specificity ranges 60-100% 45 , ultrasound has user variability and limited visibility 59 .For this reason, CT scans are often employed in surveillance protocols 42 and are typically only performed when indicated by clinical findings.In summary, although it is not the imaging modality of choice for ovarian cancer surveillance, TVUS is a useful tool to prevent the discovery of late stage ovarian cancer in women who are at increased risk for developing ovarian cancer.

What role do biomarkers play in the screening of ovarian cancer and the detection of disease recurrence?
Biomarkers are substances which help to indicate the presence of a disease.Soluble biomarkers differentially expressed between individuals with disease and normal controls are convenient tools of disease detection.The perceived advantages of biomarkers compared to other disease screening modalities such as physical exam or ultrasound, include availability, reproducibility, objectivity (operator-independent) and costeffectiveness.Biomarkers for early detection aim to identify ovarian cancer in individuals who are symptomatic (Phase II specimens) or who are asymptomatic before a clinical diagnosis is made (Phase III specimens).However, the identification of such biomarkers is challenging.Discovery methods often use patient samples with clinically diagnosed and advanced stage disease, thus making it necessary to extrapolate findings of advanced disease to early-stage disease, and biomarkers discovered in diagnostic samples may not be validated in prediagnostic samples 60 .CA125 (or MUC16) glycoprotein is the most studied tumor marker, alone and/or in combination with other biomarkers, for ovarian cancer screening.Approximately 80% of ovarian cancer tumors are CA125 positive 61 .Elevated serum CA125 levels (>35 units/mL) can be detected in asymptomatic women with ovarian cancer using a monoclonal antibody (OC 125) 62 and carry a specificity of 98.5% for postmenopausal women 63 .An elevation in CA125 levels, especially twice the upper limits of normal, can often occur 2 to 5 months prior the clinical detection of an ovarian cancer recurrence 45 , with sensitivity and specificity for recurrence detection ranging from 62-94% and 91-100% 45,[64][65][66] .Recent work has further shown that CA125 levels may begin to rise as early as 3 years prior to clinical diagnosis, but will likely only reach detectable levels in the final year before diagnosis 67 .While CA125 is the most predictive marker of ovarian cancer 67 and remains the single-best marker 68 , studies have generally indicated that CA125 serum testing performs poorly in the detection of early stage disease 69 .CA125 levels are only elevated in approximately 50% of stage I ovarian cancers 62 .Further, false positive CA125 levels can occur in women with benign conditions, including menstruation, appendicitis, benign ovarian cysts, endometriosis and pelvic inflammatory disease, as well as with other malignancies, including breast, lung, endometrial and pancreatic cancers 61 .Thus, multimodal strategies, particularly the combination of CA125 with pelvic ultrasound, have been examined in order to improve sensitivity and PPV of ovarian cancer screening.The combination of CA125 and ultrasound has been examined in several studies 70 .In a prospective pilot study, 144 women with an elevated risk of ovarian cancer, as defined by a Receiver Operating Characteristic (ROC) curve based on age and CA125, underwent TVUS 71 .Sixteen women were recommended for surgery and 3 women were found to have primary invasive ovarian cancer, thus yielding a specificity of 99.8% and a PPV of 19%.This algorithm was subsequently incorporated into the United Kingdom Collaborative Trial of Ovarian Cancer Screening, which is a randomized controlled trial designed to assess the effect of screening on mortality 58 .Women are randomized to three arms: no treatment, CA125 with TVUS screening or TVUS alone screening.At the prevalence screen, CA125 combined with TVUS achieved sensitivity, specificity, and positive-predictive values of 89.5%, 99.8% and 35.1%, respectively 58 .The specificity was higher in this combined screening group compared to the TVUS alone group (89.4% vs. 75.0%),suggesting that this screening would result in lower rates of repeat testing and surgery.In an additional study, an elevated serum CA125 (≥35 units/mL) and preoperative ultrasound findings of solid or complex tumors yielded a PPV of 84.7%, a NPV of 92.4% and correctly identified 77.3% of patients with early stage disease 70 .Additional potential serum biomarkers have been identified [72][73][74] and extensively examined for the detection of ovarian cancer [75][76][77] .Human epididymis protein 4 (HE4) is a biomarker overexpressed by both serous and endometrioid ovarian cancers 78 and is expressed by 32% of ovarian cancers lacking CA125 expression 76 .HE4 has been FDA approved to monitor for disease recurrence (June 2008) and was recently incorporated into the clinical evaluation of ovarian cancer patients.Studies have also indicated that HE4 may also improve prediction of malignancy in ovarian masses when combined with CA125 measurements [75][76] .Furthermore, Anderson and colleagues have demonstrated an increase in CA125, HE4 and mesothelin in ovarian cancer patients compared to matched controls, with a differential expression noted as early as 3 years preceding diagnosis; these results suggest that a multimarker profile may improve detection of early stage disease 67 .Several panels of biomarkers have been published during the past ten years.One of them, a multiplex, bead-based, immunoassay system, examined serum concentrations of leptin, prolactin, osteopontin, insulin-like growth factor II, macrophage inhibitory factor and CA125.This blood test, called OvaSure™, was reported to achieve a sensitivity of 95.3% and specificity of 99.4%, providing a significant improvement over CA125 alone for ovarian cancer detection in a cohort of women newly diagnosed with ovarian cancer compared to healthy controls 79 .
OvaSure™ was proposed as a screening tool for women at risk for ovarian cancer, but, due to some concerns 80 , further investigation is warranted prior to the commercial use of this biomarker panel as a screening tool for the early detection of ovarian cancer.The use of CA125 for detection of relapsed disease is not supported by the recent results of a randomized control trial 81 .This multi-institutional European randomized control trial failed to demonstrate a survival advantage for women with recurrent disease who received early intervention based on rising CA125 levels compared to those who received treatment when symptoms developed 81 .The authors thus questioned the value of routine CA125 measurements for surveillance of women with ovarian cancer who attain a complete response after first line therapy.Yet, the conclusions of this study have been underplayed by several concerns, including failure to address the role of secondary cytoreduction, lack of stratification by residual disease following primary cytoreduction, lack of radiographic confirmation of recurrence and non-standardized second-line therapies.Thus, the prevention of clinically detectable relapses using serial CA125 measurements will likely continue at the discretion of the patient and her physician 82 .

Biomarker discovery for ovarian cancer prevention
Various techniques are currently under investigation in order to identify new biomarkers which may improve the detection of early stage ovarian cancer as well as improve the detection of recurrent disease [83][84][85][86][87] .Proteomic analysis of serum and ascites samples by mass spectrometry is a strategy under investigation for the detection of differentially expressed proteins or protein fragments in women with ovarian cancer compared to healthy controls 83, 88, 89 .Biomarkers and respective panels identified with proteomics have the potential to influence ovarian cancer prevention; further development and validation, however, are necessary before they may introduced into clinical practice 89 .
Evolving technologies, including transcriptomics 84 , epigenomics 85,86 , metabolomics 90 and glycomics 87 , are also under investigation in ovarian cancer.Transcriptomics, or expression profiling, studies the impact of RNA molecules, including mRNA, rRNA, tRNA and noncoding RNAs, in diseases.Using techniques based on DNA microarrays, these molecules can be identified to help pinpoint genes which may be differentially expressed in ovarian cancer compared to normal tissue 84 .Gene expression profiling can be performed using both serum and formalin-fixed paraffin-embedded tissue biopsies and may help to identify genes associated with early-stage disease thereby improving screening 84 .Epigenomics focuses on the role of DNA methylation, histone modifications, RNA interference and nucleosome remodeling in the development and progression of ovarian cancer 86 .Epigenetic alterations can be used as candidate targets for early detection and for monitoring of ovarian cancer recurrence 85 .Aberrant DNA hypermethylation of CpG islands in the promoter of tumor suppressor genes and other cancer genes as well as microRNAs (miRNAs) are currently being identified in both body fluids and tissue biopsies and may help to demonstrate the importance of specific genes involved in ovarian tumorigenesis 85 .Metabolomics examines the role of small molecules ("metabolites") which are unique to a specific cellular process.Metabolic fingerprints of ovarian cancer can be measured in serum and/or other bodily fluids using mass spectrometry and has the potential to improve detection of early stage and recurrent disease 90 .Lysophosphatidic acid 91 and lipid associated sialic acid 92 are metabolites which are currently under investigation for ovarian cancer detection.Glycosylation is the most common post-translational modification of proteins.Aberrant glycosylation patterns of proteins, such as MUC1 93 , have been identified in ovarian cancer and may play a key role in promoting tumor cell invasion and metastasis as well as stimulating anti-tumor immune responses 94 .Therefore, glycoproteins are currently being examined for their potential as biomarker as well as for treatment 87 .
In addition to these efforts, we have generated a cost-and time-effective method for generating site-specific in vivo biotinylated recombinant antibodies secreted by yeast (Biobodies 5,95 ).Biobodies have been generated against HE4 95 and mesothelin 96 .We have also demonstrated that this technology can be used reliably in a highly-sensitive bead-based ELISA assay for screening large populations of women for ovarian cancer 5,67 and for serum biomarker discovery 6 .These novel approaches to biomarker discovery offers promise for improved ovarian cancer screening and for detection of recurrences.The impact of these biomarkers on clinical outcomes warrants further investigation in prospective clinical trials.

Current recommendations for ovarian cancer prevention
Given its low prevalence in the general population, universal screening for ovarian cancer is neither feasible nor cost-effective.Risk assessment is inherent to the success of any screening approach, and women at highest risk for disease are likely to benefit the most from preventive strategies.Several risk factors have been identified for epithelial ovarian cancer (Table 2) 25 , and current screening recommendations are often stratified by an individual's risk of developing disease.While the exact pathogenesis of this disease is still unclear, it is generally postulated that an increase in ovulation and/or an increase in estrogen exposure is associated with an increased lifetime risk of disease.Perhaps, the single most important risk factor for ovarian cancer is family history.Hereditary ovarian cancers account for approximately 10% of all EOC cases.Compared to controls, women with one first or second-degree relative with ovarian cancer have a three-fold increase in risk 104 .Hereditary ovarian cancers are commonly attributed to genetic mutations which are transmitted in families in an autosomal dominant fashion.Germline mutations in BRCA1 and BRCA2, tumor suppressors which participate in homologous recombination repair of double-stranded DNA breaks, account for approximately 95% of all hereditary EOC cases 105 and carry a 25 to 50% lifetime risk of ovarian cancer 106 .Further, BRCA1/2 mutations are highly prevalent amongst women of Ashkenazi Jewish descent; 35-40% of Ashkenazi women with ovarian cancers have a BRCA1 or BRCA2 mutation 107 .These mutations may also be suspected in individuals with a personal history of breast cancer before age 50, dual breast cancer or ovarian cancer 108 .Women with BRCA-associated ovarian cancer typically present with high grade serous cancers at an earlier age compared to non-hereditary controls; however, these individuals more often have higher response rates to platinum-based chemotherapy and improved overall survival 109 .
The remaining hereditary EOC cases are attributed to Lynch Syndrome II, also referred to as hereditary nonpolyposis colorectal cancer (HNPCC) syndrome; these individuals with mutations in DNA mismatch repair genes MLH1, MSH2, MSH6 and PMS2 are at increased risk for colon cancer as well as numerous other cancers, including endometrial and ovarian cancer 110 .Women with this autosomal dominant genetic background have a 3 to 14% lifetime risk of ovarian cancer 110 .

Recommendations for ovarian cancer prevention in women at average risk
In the absence of significant risk factors, a typical woman carries a 1 in 72 lifetime risk of ovarian cancer 111 and is considered at average risk of developing ovarian cancer.

Prevention by risk reducing behaviors
Epidemiologic studies of women with ovarian cancer risk have identified several protective factors, including oral contraceptive pill use (OCP), parity, lactation, and tubal ligation (Table 2).These protective factors should be considered for women with any risk of ovarian cancer as an additional preventive strategy.Patients should be counseled regarding the impact of these factors on their risk of ovarian cancer.Specifically, (1) the use of OCPs for 5 or more years results in a 50% reduction in the incidence of ovarian cancer 102 , and this benefit may last for up to 30 years following use 112 .This benefit has also been reported in women with BRCA1 or BRCA2 mutations 113 and for most histological subtypes 114 .It is estimated that OCPs have prevented 200,000 ovarian cancers and 100,000 deaths 25 .(2) Parity is a protective factor for ovarian cancer 25 .The risk for ovarian cancer decreases with each live birth, but there is no additional benefit once a women achieves grand multiparity 115 .Parous women with BRCA1 mutations can also experience a reduced risk of ovarian cancer with each additional full-term pregnancy 116 .(3) Lactation also results in a decreased incidence of ovarian cancer 117 .However, there is no additional benefit for individual episodes of lactation beyond 12 months.The relative risk of ovarian cancer decreases by 2% for each month of breastfeeding 118 .(4) Tubal ligation may also substantially reduce the risk of ovarian cancer 119 .Given a greater than 60% risk reduction, women with BRCA1 mutations should be counseled regarding this option especially when they have completed childbearing 120 .

Prevention by routine screening
Given a low incidence and prevalence of ovarian cancer in the general population, large study cohorts are necessary to evaluate the utility of an ovarian cancer screening test 121 .The www.intechopen.comresults of initial clinical trials, while failing to evaluate the impact of screening on cancerrelated mortality, emphasize limitations on the specificity and PPV of available screening strategies for women at average risk.
A pilot randomized control trial evaluated a multimodal screening approach with serial CA125 and pelvic ultrasound in a sample of almost 22,000 postmenopausal women 121 .Combined CA125 and ultrasound (US) screening was not only feasible but also preliminarily resulted in a survival advantage (median survival 72.9 months in the screened group vs. 41.8 months in the control group, p = 0.0112).Data from this trial have paved the way for larger randomized-control trials 21,58,122 which aim to examine the impact of screening on mortality.
The Shizuoka Cohort Study of Ovarian Cancer Screening (SCSOCS) trial was a prospective, randomized trial examining ovarian screening, via CA125 and US, in asymptomatic postmenopausal Japanese women between 1985 and 1999 122 .Of more than 41,000 women who underwent screening, only 27 had detected ovarian cancer; at the prevalent screen, screening produced a detection rate of 0.31 per 1000.Ovarian cancer screening also identified a higher proportion of stage I cancers (63% vs. 38%, p=0.23) when compared to the control group.The Prostate, Lung, Colorectal and Ovarian Cancer (PLCO) screening trial is a randomized controlled cancer screening trial evaluating screening tests for the 4 PLCO cancers 41 .More than 78,000 healthy women between 55 and 74 years of age from across the United States were randomized to a screening or usual care arm at 10 screening sites between 1993 and 2001.The primary objective of this trial was to determine whether routine screening via transvaginal ultrasound (TVUS) and/or CA125 can reduce ovarian cancer-specific mortality.Twenty-nine neoplasms were identified in almost 29,000 women who received any screening test, producing a PPV for TVUS of 1.0%, 3.7% for CA125, and 23.5% for combined TVUS and CA125.Overall, these screening tests were associated with a high number of false-positive tests, especially for women who were younger, heavier, and had a history of prior hysterectomy 123 .Further, TVUS and CA125 failed to produce a significant impact on ovarian cancer mortality, and 15% of women with false-positive screening experienced serious resulting complications 21 .The results of this trial suggest that routine screening with CA125 and TVS should not be performed in asymptomatic women at low-risk for ovarian cancer.The United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) is a large randomized control trial evaluating TVUS and/or CA125 versus no screening in sample of more than 200,000 postmenopausal women between 2001 and 2005 58 .The primary objective of this study is to determine whether screening affected ovarian cancer-related mortality.In the prevalence screen, 42 primary ovarian and 45 fallopian tube cancers were identified with 48.3% of these cancers reported as stage I or II disease.The sensitivity, specificity and PPV for primary invasive epithelial ovarian and tubal cancers were 89.5%, 99.8% and 35.1% for combined TVUS and CA125 versus 75.0%, 98.2%, and 2.8% for TVUS alone, respectively.Thus, combination screening methods yielded the lowest number of false-positive screens, translating into lower rates of repeat testing and surgery.While this initial screen indicates that these screening strategies are feasible, the impact of these tests on mortality is still pending at this time.In summary, the latest studies suggest that risk reducing behaviors can provide significant prevention of ovarian cancer, while routine screening for ovarian cancer in women at average risk does not improve the prevention of late-stage disease and is currently not recommended by any professional society 25 .

Recommendations for ovarian cancer prevention in women at increased risk
Women with a strong family history of either ovarian or breast cancer alone are considered to be at higher-than-average risk, while women with confirmed mutations in BRCA1 or BRCA2 and those with Lynch syndrome are at the highest risk of developing ovarian cancer.Genetic risk assessment should be performed in individuals to provide "individualized and quantified assessment of risk as well as options for tailored screening and prevention strategies" 108 .The Society of Gynecologic Oncologists recommends genetic screening in women with a 20-25% risk of having an inherited predisposition to breast and ovarian cancer: (1) women with a personal history of both breast and ovarian cancer (including those with primary peritoneal or fallopian tube cancers; (2) women with ovarian cancer and a close relative with breast cancer at ≤ 50 years or ovarian cancer at any age; (3) women with ovarian cancer at any age who are of Ashkenazi Jewish ancestry; (4) women with breast cancer at ≤50 years and a close relative with ovarian or male breast cancer at any age; (5) women of Ashkenazi Jewish ancestry and breast cancer at ≤ 40 years; or (6) women with a first or second degree relative with a known BRCA1 or BRCA2 mutation 108 .Further, risk assessment is recommended if women have a 20-25% of having an inherited predisposition to endometrial, colorectal and related cancers, including: those patients meeting the revised Amsterdam criteria 124 and those with personal or family history concerning for Lynch Syndrome 108 .

Prevention by risk reducing behaviors and surgery
In addition to the risk reducing behaviors described earlier for women at average risk,, prophylactic surgery should be strongly considered in women at high risk for ovarian cancer.Risk-reducing salpingo-oophorectomy (RRSO) is associated with an 80% risk reduction in BRCA1/2-associated ovarian, fallopian tube or primary peritoneal cancer 125,126 .Women with BRCA germline mutations have a significant survival advantage following risk-reducing surgery compared to disease surveillance 125,126 .This approach has also been reported as a cost-effective strategy 127,128 .Women with BRCA1/2 germline mutations should be counseled on risk-reducing strategies, and RRSO should be recommended upon completion of childbearing or by age 40 25 .Risk-reducing hysterectomy and bilateral salpingo-oophorectomy is also a feasible preventive approach in women with Lynch Syndrome, with risk reduction approaching 100% 129 .Recent cost-effective analyses demonstrated that risk-reducing surgery is the most cost-effective gynecologic cancer prevention strategy in this patient population 128,130 .

Prevention by routine screening
Current opinion suggests that screening may be appropriate for women in these increased risk categories.However, while intensive screening is recommended for women with BRCA1 and 2 mutations, studies have indicated that screening with CA125 and TVUS are ineffective 131,132 because the majority of cancers are still detected at advanced stages.In a retrospective study of 241 women with confirmed BRCA1 or BRCA2 mutations, surveillance with annual pelvic exam, transvaginal ultrasound and serum CA125 level failed to effectively identify women with early stage disease 131 .
Currently, women with HNPCC/Lynch Syndrome are offered active disease surveillance including annual TVUS, endometrial biopsy and CA125 108 .Auranen and colleagues performed a systematic review of the literature to determine the role of screening in women with HNPCC or with a family history of HNPCC 133 .Of five studies meeting inclusion criteria, only three examined the utility of CA125 surveillance for ovarian cancer in this patient population.In total, five ovarian cancer cases, none of which were reported as early stage disease, were detected by CA125 surveillance.Based on the current available published evidence, the authors concluded that there is no benefit for ovarian cancer screening in this patient population.In summary, while studies have failed to demonstrate a benefit for screening in high risk patients, risk-reducing surgery is the most cost-effective gynecologic cancer prevention strategy and screening with serial CA125 levels and TVUS is generally recommended until risk-reducing surgery can be performed.

Recommendations for ovarian cancer prevention in women with pelvic masses
Several investigators have introduced risk models which would allow for the preoperative risk assessment of women with pelvic masses [134][135][136][137] .The Risk of Malignancy Index (RMI) is a diagnostic model combining CA125 levels, imaging and menopausal status; at a cutoff level of 200, the RMI produced a sensitivity of 85% and a specificity of 97% and was an effective model for discriminating between cancer and benign lesions 137 .The Risk of Ovarian Malignancy Algorithm (ROMA) is another model which predicts the likelihood of ovarian cancer in women with pelvic masses by the combination of HE4 and CA125 serum levels with menopausal status 136 .This algorithm has shown promising diagnostic performance for the detection of ovarian cancer in postmenopausal women, with a sensitivity of 82.5% 136 , and has also been shown to perform better than the RMI model for risk prediction of ovarian cancer 134,135 .This model may therefore be an effective strategy for triaging patients with pelvic masses.

Current recommendations for preventing disease recurrence 5.4.1 Role of disease surveillance
Active disease surveillance aims to detect recurrent ovarian cancer in asymptomatic women in order to provide opportunities for early intervention and ultimately improved outcomes.However, current surveillance recommendations are often based on expert opinions and practice patterns.The National Comprehensive Cancer Network (NCCN) recommends routine visits every 2 to 4 months for 2 years, then every 3 to 6 months for 3 years, followed by annual visits after 5 years 42 .A physical examination, serum CA125 and laboratory and imaging (as clinically indicated) are to be performed at each visit.
In response to the MRC OV05/EORTC 55955 trial 81 , the Society of Gynecologic Oncologists issued a statement on the use of CA125 for monitoring ovarian cancer in June 2009: "Although there may not presently be a major survival advantage to the use of CA125 monitoring for earlier diagnosis of recurrence, patients and their physicians should still have the opportunity to choose this approach as integral to a philosophy of active management" and that "patients and their physicians should be encouraged to actively discuss the pros and cons of CA125 monitoring and the implications for subsequent treatment and quality of life" 82 .
A systematic review of the literature demonstrated that routine surveillance was able to detect 67% of asymptomatic recurrences with a lead time of 3 months but that published studies failed to demonstrate a survival advantage of early detection of ovarian cancer by routine surveillance 36 .The authors suggest that routine surveillance should be reconsidered in current practice.

Immunoprevention of disease recurrence
While 70-80% of patients with advanced EOC will initially respond to conventional platinum/taxane therapy, more than 60% will experience a recurrence of disease and 70-90% will ultimately die of their disease 2 .Immune-driven vaccines are currently under investigation for the prevention of ovarian cancer recurrence [8][9][10][11][12] .
Host anti-tumor immune responses have the potential to significantly influence prognosis in ovarian cancer patients.The presence of tumor-infiltrating lymphocytes (TILs) has been correlated with significantly improved progression-free and overall survival rates in women with advanced stage ovarian cancer compared to women without TILs 138,139 .Thus, given that ovarian cancer is intrinsically immunogenic, it may be possible to enhance host antitumor immune responses by using vaccines which strengthen TIL responses and thereby improve patient outcomes by preventing recurrent disease.Therapeutic vaccinations derived from autologous whole tumor cell lysates may help to enhance host antigen-specific anti-tumoral immune responses 14 .The main advantages of these vaccines are "the opportunity to induce immunity to a personalized and broad range of antigens" and the incorporation of yet unidentified tumor antigens 140 .A recent metaanalysis of 173 immunotherapy trials, including ovarian and other primary cancers, demonstrated a higher objective clinical response in individuals receiving whole tumor antigen-based vaccines compared to those receiving synthetic antigens (8.1% vs. 3.6%, respectively; p <0.001) 141 .The Penn Ovarian Cancer Research Center is currently conducting a phase I/II randomized study to determine the feasibility, safety and immunogenicity of a vaccine derived from autologous oxidized tumor cell lysate (OC-L) in combination with Ampligen, a Toll-like receptor 3 agonist (NCT01312389).
Vaccination with antigen-specific dendritic cells (DCs) can enhance anti-tumor immunity via specific tumor-antigen presentation and activation of effector T cells 142 .There are several approaches to DC-based vaccines, including exposure of DCs to whole tumor cell lysates, defined ovarian tumor peptides, and ovarian tumor cells, to induce a cytotoxic T lymphocyte (CTL) response 143 .In a phase I trial, three of six patients with progressive or recurrent ovarian cancer experienced stabilization of disease following administration of autologous tumor antigen-pulsed DCs with reported progression-free intervals of 8-45 months 144 .Given these promising data, DC-based vaccines are currently the focus of several new trials (NCT00703105, NCT00683241, and NCT01132014) which will hopefully demonstrate an impact on long-term prognosis.The Penn Ovarian Cancer Research Center is currently examining the feasibility and immunogenicity of a DC vaccine loaded with autologous tumor lysate administered intranodally, alone or in combination with intravenous Bevacizumab (NCT01132014).A phase I/II trial is also underway at our institution in which patients with recurrent EOC or primary peritoneal cancer will undergo adoptive transfer of ex vivo CD3/CD28-costimulated autologous peripheral blood T cells along with tumor lysate-pulsed DC vaccination (DCVax®-L) (NCT00603460) in order to determine the feasibility and safety of this combination and progression-free survival at 6 months.

Conclusion
Ovarian cancer is the most lethal gynecologic cancer in the United States.Given the low prevalence of this disease in the general population, risk assessment is crucial to the success of available preventive strategies.However, current primary preventive strategies, even in women at high risk, have not proven reliable in the detection of early stage disease nor have they significantly impacted disease related mortality.Thus, risk-reducing behaviors and surgery should be considered in women at high risk for ovarian cancer.
In the near future, novel technologies may help to better characterize critical pathways in ovarian carcinogenesis and therefore result in biomarkers and/or multimarker panels more effective than CA125 alone, in both detecting early stage disease as well as recurrences.
Validations of proposed strategies are under investigation in ongoing studies (Table 3).

Table 2 .
25us, factors, such as nulliparity, menarche at an early age, menopause at a late age, fertility drug use and hormone replacement therapy use, are believed to put individuals at risk for disease 25, 97-101 .Age, Caucasian race, ethnicity (especially Ashkenazi Jewish heritage), living in an industrialized country, and a history of endometriosis are other factors predisposing to ovarian cancer25.In addition, several factors, particularly multiparity, oral contraceptive use, breastfeeding and tubal ligation, have been linked with a decreased incidence of ovarian cancer and are therefore believed to be protective against developing ovarian cancer 102, 103 .Protective and Risk factors for Ovarian Cancer25.