Effect of screening on ovarian cancer mortality in the prostate, lung, colorectal, and ovarian (PLCO) cancer randomized screening trial

Reporter: Gita Suneja, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 5, 2011

Share article


Presenter: Saundra Buys, MD
Presenter Affiliation: University of Utah

Background

  • Ovarian cancer is a deadly disease with high case fatality rate.
  • Approximately 22,000 women are diagnosed yearly in the U.S, and 14,000 die from the disease.
  • Women presenting with lower stage disease have better survival outcomes, but unfortunately many women present with symptoms at late stages.
  • Theoretically, screening in asymptomatic women may improve survival, however no screening strategy providing clear benefit to this end has been identified.
  • A pilot randomized trial conducted in the United Kingdom in the 1990s suggested a survival benefit for CA-125 screening (Jacobs, Lancet 1999).
  • A large, multi-center randomized trial of CA-125 and transvaginal ultrasound (TVU) conducted in Japan from 1985-1999 demonstrated that although there was a rise in the detection of early-stage ovarian cancer in asymptomatic postmenopausal women, this difference was not statistically significant (Kobayashi, Int J Gynecol Cancer 2008).
  • Currently, the U.S. Preventative Services Task Force (USPSTF) recommends against routine screening for women.
  • The PLCO randomized screening trial began in 1993 to determine whether screening tests reduced deaths from prostate, lung, colorectal and ovarian cancer.
  • The primary objective of this study was to evaluate the effect of CA-125 and TVU screening on ovarian cancer-specific mortality.

Methods

  • Study participants were 55 to 74 years of age at the time of enrollment.
  • Patients were enrolled from 10 U.S. centers between November 1993 to July 2001 and followed through February 2010.
  • Screening interventions included:
    • CA-125 annually for 6 yrs
    • TVU annually for 4 years
  • Further evaluation after screening tests was left to physician discretion.
  • The study was designed to have 88% power to detect a 35% reduction in mortality.
  • Contamination, or screening of patients at a non-study center, was planned to be < 10%.

Results

  • 78,216 patients were randomized
    • 39,105 had screening intervention
    • 39,111 had usual care
  • The two groups were well-matched in terms of age, ethnicity, education, prior hysterectomy, oral contraceptive use, hormone therapy exposure, nulliparity, personal history of breast cancer, and family history of breast cancer or ovarian cancer.
  • Median time to follow-up was 12.6 yrs, and over 95% of patients were followed in each arm.
  • Compliance with screening was as anticipated, but CA-125 screening decreased slightly over the study period from 85% at baseline to 73% at year 6. TVU screening decreased from 84% at baseline to 78% at year 4.
  • Contamination was much lower than expected, 2.3-4.6% as compared to an estimated 10%.
  • Incidence of new cancers was higher in the screening intervention arm as compared to the usual care arm:
    • 212 new cases were detected in intervention arm
    • 176 new cases were detected in the usual care arm
    • Rate ratio for new detection was 1.21 (95% CI, 0.99-1.48) in the intervention arm compared to the usual care arm
  • Of the detected cancers, 9% were detected at baseline (first screening), 25% during years 1-5 of the study, 17% in the interval between cancer screenings, and 37% after the screening phase of the study was completed.
  • Most of the tumors were primary invasive cancer in both the intervention and control groups. High-grade serous histology was the most common.
  • Demographic, tumor, and treatment characteristics of the screened and usual care groups were not significantly different from each other. Most tumors were stage III-IV, and most patients received surgery and chemotherapy.
  • There were 118 deaths in the intervention arm and 100 deaths in the usual care arm. The rate ratio was 1.18 (95% CI, 0-91-1.54).
  • There was no difference in survival between the two groups analyzed from the date of randomization.
  • Harms related to screening were incurred: 3,285 women had false positive results, and of those 1,080 underwent surgery. 163 of these women had major complications such as infection, direct surgical complication (i.e. bowel injury), and cardiopulmonary disease.

Authors' conclusions

  • In this study, screening with CA-125 annually for 6 years and TVU annually for 4 years did not reduce disease-specific mortality.
  • In addition, evidence of harm from false-positive screenings was observed.

Implications

  • National media attention has focused on the development of a successful screening strategy for ovarian cancer, however no such screening program has been identified to date.
  • This large, prospective randomized trial has many strengths, including:
    • Power sufficient to provide meaningful results on mortality
    • Clearly defined primary study endpoints
    • Extensive follow-up to ascertain information about all study endpoints
    • Minimal contaminations in the control arm
    • Transparent data analysis
  • There are also several limitations of this study:
    • Study design: since patients were screened for the same number of years regardless of when they entered the study, the early enrollment patients were screened for 4-6 years, then followed up without any additional screening during the study period. This may have contributed to “dilution” of the screening effect, as some of these women developed new ovarian cancers during a period in follow-up when they were no longer actively being screened.  This flaw in the study design is reflected in the fact that 40% of cancers detected in this study were diagnosed after screening was completed.
    • Evaluation and follow-up of positive screening tests were at the discretion of the treating physician. The more common practice is to have well-defined guidelines for post-screening work-up.
    • Due to the multi-center nature of this trial, a centrally organized trial management system could have improved quality assurance measures. In this study, blood work was performed in various labs and many different practitioners performed TVUs. Data suggests that TVU is highly operator dependant, and the study investigators could have incorporated more ongoing quality assurance measures, such as training and audits.
  • The natural history of ovarian cancer is not well understood, and therefore the benefit likely to be derived from screening is unclear.
  • At present, ovarian cancer screening with CA-125 or TVU cannot be recommended in asymptomatic women. Use of routine screening may lead to harm due to false positives results, and does not appear to offer clear benefit in terms of outcomes associated with ovarian cancer.
  • Clearly, the need for more sensitive and specific biomarkers is critical.
  • The UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) is ongoing to address the efficacy of ovarian cancer screening using a different screening strategy than PLCO. This data will be available in 2013 and may provide a new perspective on screening of asymptomatic women.