In addition to the PLCO and UKCTOCS screening trials, a number of additional clinical trials have examined the efficacy of imaging and CA-125. Imaging of the ovary has been proposed as a strategy to detect changes in size and architecture that might precede the development of symptoms and detection by pelvic examination. Studies involving healthy women have established the upper limit of ovarian size as greater than 5 cm and a set of morphologic characteristics of ovarian masses in order to identify malignant neoplasms [131, 132]. Therefore, considering the above model, these trials could be expected to preferentially diagnose cancer at invasive late stages, thus limiting their efficacy as screening tools [133]. In fact, several prospective studies of transvaginal ultrasonography that involved more than 136,000 women had mixed results, with some but not others showing a nonsignificant trend toward the diagnosis of ovarian cancer at an early stage [12, 32, 134–139]. The PPV of ultrasonography in these studies ranged from 1.0 to 27 % [137]. The interpretation of these results is complicated by differences with respect to inclusion criteria, with some studies enrolling only women at high risk (on the basis of a family history) [12, 135], some enrolling only women at average risk [32, 134, 138, 139], and others enrolling both women at high risk and those at average risk [136, 137].

An alternative strategy is to evaluate serum tumor markers for the early detection of ovarian cancer. This strategy is potentially attractive in that measurement of markers is broadly available, can be repeated at appropriate intervals, and is minimally invasive. In addition, biomarker testing does not depend on operator interpretation and is less costly than ultrasonography. As of yet, the most robust serum biomarker for detection of ovarian cancer is CA-125. In clinical trials, CA-125 as a single screening biomarker demonstrated limited utility when examined in a retrospective analysis of serum samples from 5,550 women enrolled in a population-based registry in Sweden [140]. Later it was suggested that measurement of CA-125 values in an individual patient over time could improve the estimation of a patient’s risk of ovarian cancer [141]. To evaluate CA-125 dynamics, the ROCA was developed based on the slope of serial CA-125 measurements drawn at regular intervals [34]. This algorithm was based on the observation that irrespective of the initial level, CA-125 measurements are stable in non-cases for periods of more than 5 years, indicating that each woman has her own baseline level of CA-125 [34]. In contrast, exponentially increasing serial values readily identify cases. When the ROCA score exceeds a 1 % risk of having ovarian cancer, patients undergo TVU to determine whether additional intervention is warranted. In a retrospective examination of 33,621 serum samples from 9,233 women, ROCA provided an SN of 86 % at a fixed SP of 98 % for the preclinical detection of ovarian cancer, as compared with an SN of 62 % for a single CA-125 value [34]. This algorithm was confirmed to have a reasonably high PPV (19 %) in a subsequent prospective pilot study involving more than 13,000 postmenopausal women [142]. This strategy is currently being prospectively studied among more than 2,600 high-risk women by the Gynecologic Oncology Group (protocol #199) as well as in a parallel trial being conducted by the National Cancer Institute-Cancer Genetics Network [143]. Another algorithm used in screening of high-risk women is the Parametric Empirical Bayes (PEB) algorithm [45, 144]. In the PEB model, each woman’s complete screening history, not only her most recent index value, acts as her own control, and the slightest deviation from that person-specific norm is used to detect elevated values at a fixed SP. The PEB model will assign the great majority of subjects lower thresholds, allowing their cancers to be detected earlier. Additionally, PEB allows for assessment of multiple markers.

While many of the reports represented in Table 3.4 are initially encouraging, several hurdles remain before any of the identified models can be implemented clinically on a widespread basis. Foremost is the need to evaluate the most promising panels in prospective randomized clinical trials. Additional preclinical validation will be required to fully characterize the efficacy of selected panels before this significant next step is warranted. A key component of this validation process is the evaluation of panels in samples obtained from prediagnostic ovarian cancer patients. Progress toward this type of validation is greatly hindered by the rarity of this sample type; however, several significant findings have been reported by a group under the direction of Nicole Urban. In a pair of reports, this group first describes elevated levels of CA-125, HE4, and mesothelin in the sera of symptomatic ovarian cancer patients and then in the sera of patients 0–3 years prior to diagnosis, noting an optimal lead time of 1 year [150, 151]. In a separate study, the investigators utilize a combinatorial approach including CA-125 and HE4 measurements in addition to the Symptom Index (SI) to prospectively classify benign from malignant pelvic masses with an SN of 84 % at an SP of 98.5 % (Table 3.4) [148].

Recently, a collaborative study was performed to assess the prediagnostic performance of candidate biomarkers in the PLCO trial [152, 153]. Forty-one biomarkers selected by five leading ovarian cancer biomarker laboratories from Harvard University, Fred Hutchinson Medical Center, MD Anderson Cancer Center, University of Pittsburgh, and Yale University were utilized in the analysis of 118 cases and 473 matched healthy controls [153, 154]. The selection of these markers proceeded from a Phase 2 validation study using a common reference set of 160 ovarian cancer cases (approximately 50 % stage I/II) and 480 general population controls based solely on individual performance (SN at 95 % SP, AUC). The study demonstrated that none of these biomarkers or multimarker panels added any appreciable improvement in SN over that of CA-125 alone. In addition, levels of CA-125 began to increase, on average, less than 1 year prior to diagnosis with ovarian cancer (Fig. 3.1). Subsequent analyses revealed that 10 of the selected biomarkers, HE4, mesothelin, CA 15-3, CA 72-4, KLK6, spondin-2, TTR, Cyfra 21-1, IGFBP-2, and YKL-40, were strongly (r > 0.35) correlated with CA-125. This lack of feature independence could be credited as a limiting factor in the ability of these additional biomarkers to complement the diagnostic performance of CA-125. Similarly, Anderson et al. [151] measured several biomarkers over time in prediagnostic samples obtained from ovarian cancer patients and matched controls and observed that CA-125 (along with HE4 and mesothelin, but not of B7-H4, DcR3, and spondin-2) demonstrated appreciable increases in patients approximately 3 years before diagnosis. However, this CA-125 velocity reached detectable elevations only within the final year before diagnosis implying that a more sensitive assay could possibly detect early individual changes.

In nearly all reports, efforts aimed at novel biomarker development have relied upon the use of biological fluid samples collected at the time of first diagnosis. It would appear from the disappointing results of the PLCO study and others that this approach is not optimized for the development of prediagnostic tests offering a potential window of early successful intervention. Ongoing efforts of this type are unlikely to be productive, and the incorporation of new strategies focused on the characterization and detection of preclinical disease will be necessary for true advancement. Recent work suggests that a precursor lesion for serous carcinomas, termed serous tubal intraepithelial carcinoma (STIC), does indeed exist [155]. STICs originate from the fimbriated end of the fallopian tube, indicating a tubal, rather than ovarian epithelial, origin for serous carcinomas. The identification of the potential STIC precursor of serous carcinomas represents an important finding that may lead to advances in detecting these tumors at an earlier stage. Perhaps, analysis of blood samples from patients who had undergone prophylactic oophorectomy where preneoplastic lesions were discovered could prove useful although substantial effort would be needed to collect a statistically meaningful set of such samples. In fact, a recent publication based on mathematical modeling of ovarian cancer development using pathological findings from prophylactic bilateral salpingo-oophorectomies (PBSOs) suggested that ovarian cancers spend more than 4.3 years as in situ or stage I–II disease and that serous ovarian cancers have already progressed to a late stage by approximately 1 year prior to their discovery [133]. Therefore, if, as suggested above, changes in CA-125 levels are detectable within 3–12 months before clinical diagnosis, then current strategies are doomed to identify lower-volume high-grade disease, at best. Indeed, the authors conclude that in order to achieve 50 % SN in detecting tumors before they advance to stage III, an annual screen would need to detect tumors of less than 1.3 cm in diameter, whereas 80 % SN would require detecting tumors of less than 0.4 cm in diameter. No imaging modality could presently detect such small lesions, and therefore improved biomarker testing including the use of additional biomarkers and assays capable of detecting CA-125 with greater SN represents the most promising tools for early detection.