Long-term follow-up of the Stockholm screening study on ovarian cancer

Mortality impact, risks, and benefits of general population screening for ovarian cancer: the UKCTOCS randomised controlled trial

Background

Ovarian and tubal cancers are lethal gynaecological cancers, with over 50% of the patients diagnosed at advanced stage.

Trial design

Randomised controlled trial involving 27 primary care trusts adjacent to 13 trial centres based at NHS Trusts in England, Wales and Northern Ireland.

Methods

Postmenopausal average-risk women, aged 50-74, with intact ovaries and no previous ovarian or current non-ovarian cancer.

Interventions

One of two annual screening strategies: (1) multimodal screening (MMS) using a longitudinal CA125 algorithm with repeat CA125 testing and transvaginal scan (TVS) as second line test (2) ultrasound screening (USS) using TVS alone with repeat scan to confirm any abnormality. The control (C) group had no screening. Follow-up was through linkage to national registries, postal follow-up questionnaires and direct communication with trial centres and participants.

Objective

To assess comprehensively risks and benefits of ovarian cancer screening in the general population.

Outcome

Primary outcome was death due to ovarian or tubal cancer as assigned by an independent outcomes review committee. Secondary outcomes included incidence and stage at diagnosis of ovarian and tubal cancer, compliance, performance characteristics, harms and cost-effectiveness of the two screening strategies and a bioresource for future research.

Randomisation

The trial management system confirmed eligibility and randomly allocated participants using computer-generated random numbers to MMS, USS and C groups in a 1:1:2 ratio.

Blinding

Investigators and participants were unblinded and outcomes review committee was masked to randomisation group.

Analyses

Primary analyses were by intention to screen, comparing separately MMS and USS with C using the Versatile test.

Results

1,243,282 women were invited and 205,090 attended for recruitment between April 2001 and September 2005.

Randomised

202,638 women: 50,640 MMS, 50,639 USS and 101,359 C group.

Numbers analysed for primary outcome

202,562 (>99.9%): 50,625 (>99.9%) MMS, 50,623 (>99.9%) USS, and 101,314 (>99.9%) C group.

Outcome

Women in MMS and USS groups underwent 345,570 and 327,775 annual screens between randomisation and 31 December 2011. At median follow-up of 16.3 (IQR 15.1-17.3) years, 2055 women developed ovarian or tubal cancer: 522 (1.0% of 50,625) MMS, 517 (1.0% of 50,623) USS, and 1016 (1.0% of 101314) in C group. Compared to the C group, in the MMS group, the incidence of Stage I/II disease was 39.2% (95% CI 16.1 to 66.9) higher and stage III/IV 10.2% (95% CI -21.3 to 2.4) lower. There was no difference in stage in the USS group. 1206 women died of the disease: 296 (0.6%) MMS, 291 (0.6%) USS, and 619 (0.6%) C group. There was no significant reduction in ovarian and tubal cancer deaths in either MMS (p = 0.580) or USS (p = 0.360) groups compared to the C group. Overall compliance with annual screening episode was 80.8% (345,570/420,047) in the MMS and 78.0% (327,775/420,047) in the USS group. For ovarian and tubal cancers diagnosed within one year of the last test in a screening episode, in the MMS group, the sensitivity, specificity and positive predictive values were 83.8% (95% CI 78.7 to 88.1), 99.8% (95% CI 99.8 to 99.9), and 28.8% (95% CI 25.5 to 32.2) and in the USS group, 72.2% (95% CI 65.9 to 78.0), 99.5% (95% CI 99.5 to 99.5), and 9.1% (95% CI 7.8 to 10.5) respectively. The final within-trial cost-effectiveness analysis was not undertaken as there was no mortality reduction. A bioresource (UKCTOCS Longitudinal Women's Cohort) of longitudinal outcome data and over 0.5 million serum samples including serial annual samples in women in the MMS group was established and to date has been used in many new studies, mainly focused on early detection of cancer.

Harms

Both screening tests (venepuncture and TVS) were associated with minor complications with low (8.6/100,000 screens MMS; 18.6/100,000 screens USS) complication rates. Screening itself did not cause anxiety unless more intense repeat testing was required following abnormal screens. In the MMS group, for each screen-detected ovarian or tubal cancer, an additional 2.3 (489 false positives; 212 cancers) women in the MMS group had unnecessary false-positive (benign adnexal pathology or normal adnexa) surgery. Overall, 14 (489/345,572 annual screens) underwent unnecessary surgery per 10,000 screens. In the USS group, for each screen-detected ovarian or tubal cancer, an additional 10 (1630 false positives; 164 cancers) underwent unnecessary false-positive surgery. Overall, 50 (1630/327,775 annual screens) women underwent unnecessary surgery per 10,000 screens.

Conclusions

Population screening for ovarian and tubal cancer for average-risk women using these strategies should not be undertaken. Decreased incidence of Stage III/IV cancers during multimodal screening did not translate to mortality reduction. Researchers should be cautious about using early stage as a surrogate outcome in screening trials. Meanwhile the bioresource provides a unique opportunity to evaluate early cancer detection tests.

Funding

Long-term follow-up UKCTOCS (2015-2020) - National Institute for Health and Care Research (NIHR HTA grant 16/46/01), Cancer Research UK, and The Eve Appeal. UKCTOCS (2001-2014) - Medical Research Council (MRC) (G9901012/G0801228), Cancer Research UK (C1479/A2884), and the UK Department of Health, with additional support from The Eve Appeal. Researchers at UCL were supported by the NIHR UCL Hospitals Biomedical Research Centre and by MRC Clinical Trials Unit at UCL core funding (MR_UU_12023).

MUC16 (CA125): tumor biomarker to cancer therapy, a work in progress

Over three decades have passed since the first report on the expression of CA125 by ovarian tumors. Since that time our understanding of ovarian cancer biology has changed significantly to the point that these tumors are now classified based on molecular phenotype and not purely on histological attributes. However, CA125 continues to be, with the recent exception of HE4, the only clinically reliable diagnostic marker for ovarian cancer. Many large-scale clinical trials have been conducted or are underway to determine potential use of serum CA125 levels as a screening modality or to distinguish between benign and malignant pelvic masses. CA125 is a peptide epitope of a 3-5 million Da mucin, MUC16. Here we provide an in-depth review of the literature to highlight the importance of CA125 as a prognostic and diagnostic marker for ovarian cancer. We focus on the increasing body of literature describing the biological role of MUC16 in the progression and metastasis of ovarian tumors. Finally, we consider previous and on-going efforts to develop therapeutic approaches to eradicate ovarian tumors by targeting MUC16. Even though CA125 is a crucial marker for ovarian cancer, the exact structural definition of this antigen continues to be elusive. The importance of MUC16/CA125 in the diagnosis, progression and therapy of ovarian cancer warrants the need for in-depth research on the biochemistry and biology of this mucin. A renewed focus on MUC16 is likely to culminate in novel and more efficient strategies for the detection and treatment of ovarian cancer.

Ovarian cancer in endometriosis: epidemiology, natural history, and clinical diagnosis

We review whether endometriosis-associated ovarian cancer is a specific entity compared with ovarian cancer not associated with endometriosis, with respect to epidemiology, natural history, and clinical diagnosis; we present a review of the English-language literature for ovarian cancer in endometriosis with respect to these three features. A recent prospective study in Japan directly showed that, during a follow-up of up to 17 years of an ovarian endometrioma cohort (n = 6398), 46 incident ovarian cancers were identified, showing that the ovarian cancer risk was significantly elevated in patients with ovarian endometrioma (standardized incidence ratio [SIR], 8.95; 95% confidence interval [CI], 4.12 to 15.3). Advancing age (>40 years) and the size of the endometriomas (>9 cm) were independent predictors of the development of ovarian cancer among the women with ovarian endometrioma. Although some endometriosis lesions may predispose to clear cell carcinoma (CCC) and endometrioid adenocarcinoma (EAC) of the ovary, both of these cancers differ from the other histological types with respect to their clinical characteristics and carcinogenesis. In patients with endometriosis-associated ovarian cancer, benign-appearing ovarian masses are typically present several years before the diagnosis of the cancer. A slightly elevated carbohydrate antigen [CA] 125 level is also typically present many years before the diagnosis in these patients. However, serous-type ovarian cancer may exhibit a rapid progression possibly through de-novo carcinogenesis. Ovarian endometrioma could be viewed as a neoplastic process, particularly in perimenopausal women. Understanding the mechanisms of the development of endometriosis and elucidating its pathogenesis and pathophysiology are intrinsic to the prevention of endometriosis-associated ovarian cancer and the search for effective therapies.

Screening, epidemiology, molecular biology, and treatment strategies for endometriosis-associated ovarian cancer

Objectives

This article reviews recent data on the biology, pathogenesis and pathophysiology of the different entity of epithelial ovarian cancer (EOC).

Study design

The present article reviews the English language literature for screening, epidemiology, clinical diagnosis, natural history, preclinical and clinical trials, and promising molecular targets on EOC, particularly for clear cell EOC (cEOC) based on the gene expression profiling studies.

Results

Prospective ovarian cancer screening trials in Japan showed that (1) serous-type EOC (sEOC) may exhibit a rapid progression possibly through de novo carcinogenesis, (2) the EOC risk was elevated significantly among patients with ovarian endometrioma (SIR = 8.95), (3) advancing age (>40 years) and the diameter of endometriomas (>9 cm) were independent predictors of development of EOC, (4) the benign-appearing ovarian masses are present several years before the EOC diagnosis in patients with endometriosis-associated EOC, and (5) the slightly elevated CA125 level is also typically present many years (>3 years) before the diagnosis in these patients. Upregulation of HNF-1beta and PLK-Emi1 genes were specifically detected in cEOC. In addition, the therapy currently used in renal cell carcinoma (RCC) should be considered as an attractive therapeutic option for cEOC.

Conclusions

Ovarian endometrioma could be viewed as a neoplastic process, particularly in perimenopausal women. Understanding the mechanisms of endometriosis development and elucidating its pathogenesis and pathophysiology are intrinsic to prevention and the search for effective therapies of endometriosis-associated EOC.

Annual surveillance by CA125 and transvaginal ultrasound for ovarian cancer in both high-risk and population risk women is ineffective

OBJECTIVE

To assess the efficacy of annual CA125 and transvaginal ultrasound (TVU) scan as surveillance for ovarian cancer.

DESIGN

Retrospective audit.

SETTING

NHS Trust.

POPULATION

Three hundred and forty-one asymptomatic women enrolled for ovarian cancer screening: 179 were in a high-risk group (>10% lifetime risk of developing ovarian cancer), 77 in a moderate risk group (4-10% lifetime risk of developing ovarian cancer) and 71 in a near population risk group (<4% lifetime risk).

METHODS

Retrospective audit of case records, laboratory CA125 results, radiology reports, histology records and local cancer registry data.

MAIN OUTCOME MEASURES

Ovarian cancers occurring in study population. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of TVU, and CA125 as a screening tool for ovarian cancer.

RESULTS

Four ovarian cancers and one endometrial cancer occurred. One ovarian cancer was detected at surveillance, three occurred in women who presented symptomatically between screenings. Thirty women underwent exploratory surgery because of abnormal findings at surveillance. Two women had cancer (PPV = 6.7%); one had ovarian cancer and the other endometrial cancer. Twenty-eight women (93.3%) had no malignancy. Sensitivity, specificity, PPV and NPV for TVU in the whole cohort were 33.3, 85.8, 0.6 and 99.8%, respectively. For high-risk individuals, the figures for TVU were 33.3, 84.5, 1.1 and 99.6, respectively. Combining both modalities for the whole cohort, the sensitivity, specificity, PPV and NPV were 66.7, 82.9, 1.5 and 99.8% and 50.0, 82.8, 1.3 and 99.7%, respectively, for the high-risk group alone.

CONCLUSIONS

Ovarian screening by annual TVU and CA125 is inefficient at detecting early-stage ovarian cancers.

Ovarian cancer

Imaging has become an essential part of the clinical management of patients with ovarian cancer, contributing to tumor detection, characterization, staging, treatment planning, and follow-up. Imaging findings incorporated into the clinical impression assist in creating a treatment plan specific for an individual patient. Advances in cross-sectional imaging and nuclear medicine (PET) have yielded new insights into the evaluation of tumor prognostic factors. A multimodality approach can satisfy the complex imaging needs of a patient with ovarian cancer; however, the success of such an approach always depends on available resources and on the skills of the physicians involved.

Testing for ovarian cancer

Ovarian cancer is responsible for more deaths per annum than cervical and endometrial cancer combined. Patients are often diagnosed at a late stage because of the non-specific symptoms of this disease. It can be difficult to differentiate between benign and malignant ovarian pathology, and a malignancy risk index has been developed to guide clinicians. The accuracy of CA125 and ultrasound scans as screening tests is being assessed in randomised controlled trials and proteomic technology shows promise for the early detection of cancers. At present, without accurate screening and early diagnostic techniques, high-risk patients often chose to have prophylactic surgery.

BRCA1 genetic mutation and its link to ovarian cancer: implications for advanced practice nurses

PURPOSE

The purpose of this paper is to review (a) the linkage between the BRCA1 gene and ovarian cancer and (b) BRCA1 testing and its related issues. This review is aimed for nurse practitioners (NPs), who may be in positions to identify those at risk for BRCA1-associated ovarian cancer and to assist patients with related issues.

DATA SOURCES

Data sources include reviews and original research from scholarly journals and Internet sites.

CONCLUSIONS

Ovarian cancer is a deadly disease. Identification of those at risk because of BRCA1 mutation is possible through genetic testing. Testing for BRCA1 gene mutations has many implications whether results are positive or negative. Those with positive results will be faced with decisions regarding the best management strategies. Negative results do not completely eliminate ovarian cancer risk. Current management options for carriers of the BRCA1 mutation include taking no action, increasing surveillance for ovarian cancer, and chemoprevention with oral contraceptives or prophylactic oophorectomy for those who have completed childbearing. It is essential that NPs have knowledge underlying the issues and concerns of patients and their families at risk for BRCA1-associated ovarian cancer.

IMPLICATIONS FOR PRACTICE

NPs are in a unique position to help identify BRCA1 mutation carriers and to assist them and their families with the complex issues involving genetic testing and management options. Understanding these issues will allow NPs to give appropriate care that may include making appropriate referrals to certified genetic counselors and having balanced discussions on treatment options. Such measurements may improve early diagnosis of ovarian cancer and increase survival from this disease.

Outcome of surveillance and prophylactic salpingo-oophorectomy in asymptomatic women at high risk for ovarian cancer

OBJECTIVE

Women at high risk of ovarian cancer are currently offered two options: either surveillance or prophylactic bilateral salpingo-oophorectomy. The efficacy and outcome of surveillance remain unclear.

METHODS

We performed a retrospective study. Between 1994 and 2000, we screened 383 high-risk women, of which 152 were BRCA1/2 mutation carriers. Surveillance consisted of annual gynecological examination, transvaginal ultrasound, and serum CA125 measurement. Exploratory or prophylactic surgery was performed in selected cases.

RESULTS

There were no screen-detected primary ovarian cancers. Abnormal results at surveillance were observed in 74 (19.3%) of women; in 47 (63.5%), the abnormalities disappeared spontaneously. Exploratory surgery was performed in 20 (27.0%) women in whom one malignancy was found (metastatic breast cancer in the ovary). A rising CA125 value prompted further (non-surgical) evaluation in three women with a history of breast cancer: recurrent breast cancer was diagnosed in two women; in the third, a chondrosarcoma was found. 133 women opted for prophylactic bilateral salpingo-oophorectomy, whereby two unexpected malignancies were found (fallopian tube cancer and metastatic breast cancer). One interval primary ovarian cancer occurred, presenting as papillary serous carcinoma of the peritoneum 14 months after prophylactic bilateral salpingo-oophorectomy. Complications of prophylactic surgery were encountered in 15 (11.5%) women.

CONCLUSIONS

Ovarian cancer surveillance has limited sensitivity, and a high number of false positive findings. This can lead to unnecessary surgical interventions, possibly resulting in surgery-related complications. It is important to inform high-risk women of these limitations. For now, prophylactic bilateral salpingo-oophorectomy remains the optimal risk-reducing strategy for women at high risk.

Screening postmenopausal women for ovarian cancer: a systematic review

OBJECTIVES

To assess ovarian cancer screening in asymptomatic, general-risk postmenopausal women. Outcomes of interest were the screening tests assessed (predictive values, sensitivity, and specificity), the stage of screen-detected disease at diagnosis, psychological effects of screening, and survival.

METHODS

MEDLINE, CANCERLIT, and the Cochrane Library databases were searched to June 2003 using the terms "ovarian," "cancer," "neoplasms," "screening," "clinical trial," "meta-analysis," and "systematic review." Studies were included if they were clinical trials, meta-analyses, or systematic reviews that evaluated tests used to detect ovarian cancer in asymptomatic women in the general population. Studies investigating women at increased risk for ovarian cancer (e.g., family history) and those with symptoms suggestive of ovarian cancer were excluded.

TABULATION, INTEGRATION, AND RESULTS

Seventeen prospective cohort studies and 3 pilot randomized controlled trials were included in this review. Screening tests for cancer antigen 125 (CA125) and ultrasound had low positive predictive values, resulting in healthy women being recalled and a false-positive rate of 0.01% to 5.8%. Of every 10,000 women participating in an annual screening program with CA125 for 3 years, 800 will have an ultrasound scan because of an elevated CA125, 30 will undergo surgery because of an abnormal ultrasound, and 6 will have ovarian cancer detected at surgery (3 will be diagnosed at early-stage disease and have a chance of a cure).

CONCLUSION

There is insufficient evidence to support the introduction of screening for ovarian cancer in the asymptomatic general-risk postmenopausal population. Screening is associated with increased rates of surgery and patient anxiety.

Cost-effectiveness studies in ovarian cancer

Ovarian cancer is the fifth leading cause of cancer-related deaths. The costs associated with this cancer impact both on the affected individual and on the health system. Screening is currently unproven as a strategy for improving outcomes for women with ovarian cancer. Randomized controlled trials, however, are underway, estimating any impact of screening with ultrasound and CA125 on ovarian cancer mortality. Paclitaxel and carboplatin combination, the standard first-line chemotherapy regimen for ovarian cancer, has not been compared with cisplatin and cyclophosphamide regarding the cost-effectiveness and cost-utility, but for paclitaxel and cisplatin, numerous studies have addressed these issues. The estimated incremental costs resulting from these studies fall well within the generally accepted range for new therapies. Although acquisition costs of new chemotherapy drugs exceed those of older drugs, the impact of costly drugs on total costs may be cost saving due to less costs related to supportive and palliative care. The most important costs for the patient, the pain and suffering associated with ovarian cancer and its treatment, are hard to quantify. Nevertheless, patients' quality of life must be considered when making a clinical decision to treat this disease. A review of available cost-effectiveness studies is presented and discussed.

BRCA1: mechanisms of inactivation and implications for management of patients

The BRCA1 gene was cloned in 1994 as one of the genes that conferred genetic predisposition to early-onset breast and ovarian cancer. Since then, a genetic test for identification of high-risk individuals has been developed. Despite being implicated in many important cellular pathways, including DNA repair and regulation of transcription, the exact mechanism by which inactivation of BRCA1 might lead to malignant transformation of cells remains unknown. We examine the mechanisms that underlie inactivation of BRCA1 and assess how they affect management of patients, in terms of both primary and secondary cancer prevention strategies. Furthermore, we look at the potential usefulness of BRCA1 as a prognostic tool and as a predictive marker of response to different classes of drugs. Finally, throughout this review, we draw links between the functional consequences of BRCA1 inactivation, in terms of key cellular signalling pathways, and how they might explain specific clinical observations in individuals who carry mutations in the gene.

The significance of serum CA 125 elevation in malignant and nonmalignant diseases

OBJECTIVES

The aims of the study were to investigate whether an elevated CA 125 level signals malignancies other than ovarian cancer and to find the cause of death for 247 women with elevated values among the 5550 women screened in 1986-1988 in the Stockholm population.

METHODS

The Swedish Regional Cancer Registry delivered malignancy diagnoses among the 5550 women screened. The Cause of Death Registry gave the cause of death among the women with elevated CA 125 values.

RESULTS

Patients with ovarian cancer were excluded. In 44 women with elevated CA 125 values other malignancies were reported to the Cancer Registry. They represent 18% of the entire group with elevated values. Among the 5297 women with normal CA 125 values 13% developed various malignancies. The difference between incidence of malignant disease in women with elevated values and women with normal values is significant, P = 0.02. Especially during the test-related period, from 1 year before to 1 year after the test, malignancies were detected in 6.9% of the population with elevated values and in only 1.6% with normal values (P = < 0.001). Breast cancer and lung cancer were overrepresented among women with elevated CA 125 values (P = 0.015 and < 0.001, respectively). Of the total 5500 women screened, 358 women died with different diagnoses. Elevated CA 125 values had been noted earlier in 25 women, and of these 20 died of malignant diseases, predominantly ovarian, breast, and lung cancer.

CONCLUSIONS

Asymptomatic postmenopausal women with elevated CA 125 levels in ovarian carcinoma screening trials should, if ovarian cancer is excluded, be investigated for possible breast or lung cancer. The findings also suggest that elevated CA 125 level is a risk factor for death from malignant disease.

Porous polymer implant for repair of meniscal lesions: a preliminary study in dogs

Artificial meniscal lesions extending into the avascular part of the meniscus, which do not heal by any other means, were repaired by suturing either a porous polymer implant or a synovial flap into the defect. The implant guided the ingrowth of vascular repair tissue into the defect. This fibrous tissue later on transformed into fibrocartilage. Reconstruction with a synovial flap was not successful. It appeared that healing can be achieved by implantation of a porous polymer implant in a large number of cases. Future research will be aiming at improvement of the results of meniscal repair and application of this type of polymer for repair of cartilage defects.