Online Calculator Estimates Long-Term Chance of Surviving Prostate Cancer
PSA Rising/ source: DETROIT, MI/ April 22, 2004. A study at the Josephine Ford Cancer Center has resulted in the most reliable long-term computerized prostate cancer survival model available to date. An interactive version of this survival model is available online at prostatecalculator.org.
Patients and doctors who visit the site can obtain a personalized 10-year survival estimate based on age, race, a few clinical measures, and the kind of treatment being pursued. Once data have been entered, a simple mouse-click provides the prognosis.
Doctors at Detroit, MI cancer center worked with a Denver, CO medical computer team called Artificial Neural Networks in Prostate Cancer (ANNs in CaP). Together they studied 1,611 patients with clinically localized prostate cancer and compared them with 4,538 without prostate cancer, who were matched for age, race, and co-morbidity (additional diseases). Risk scoring based on certain variables produced survival probability estimates for both patients and controls.
Because the calculator, and the companion look-up tables published in the April issue of the Journal of Urology, provide a comparison with men with similar characteristics but who do not have prostate cancer, users receive a realistic estimate of the impact of prostate cancer on long-term survival.
Prostate cancer is the most common solid-organ male cancer diagnosed in the United States, with an estimated 189,000 new cases each year.
This new prostate cancer survival model is the most comprehensive to date because it provides an estimate of the likelihood of survival taking into account a patient’s age, race, comorbities, and treatment type.
Dr. Ashutosh Tewari, the medical scientist behind the calculator, says: "While the study was not a randomized controlled trial comparing surgery, radiation therapy, and watchful waiting, the method we used has been shown to eliminate much of the bias introduced with a non-randomized study design. And the inclusion of a large, matched control group is a great strength of our study."
The study showed that a man’s level of co-morbidities (other health cobditions like diabetes, heart disease and such) can have as much or more of an impact on his chances of long-term survival than his prostate cancer alone.
Currently, African-American men have the highest incidence of prostate cancer in the world (137 per 100,000 per year), and are 2.5 times as likely to die as whites. While the reason for this is not known, some research suggests that black men are often diagnosed at later disease stages.
"Our research indicates that African-American men also tend to undergo less aggressive treatment than whites," Dr. Tewari says, "and additional studies by our group suggest that if they received the same treatments, their prostate cancer survival rates would be much closer to those of whites."
While adding to overall knowledge of the long-term survival of men with prostate cancer, the study specifically sought answers to new questions: Why do black men receive less aggressive treatment than whites? Are black men choosing less aggressive treatment themselves or are their doctors suggesting the treatment? And, finally, is it worthwhile for men with additional diseases to treat their prostate cancer?
No personal information is collected and patient privacy is respected. The site does not promote particular doctors or specific treatments.
About ANNs in CaP
The prostatecalculator.org web site was produced by the Artificial Neural Networks in Prostate Cancer (ANNs in CaP) Project. Based in Denver, Colo., this research collaborative is headed by leading prostate-cancer expert Dr. E. David Crawford and funded by the Institute for Clinical Research at the Veterans Affairs Medical Center in Washington, D.C. Under the guidance of Dr. Crawford, national and international authorities in the fields of prostate cancer and artificial intelligence are working together to investigate the use of ANNs in the diagnosis, prognosis, and treatment of prostate cancer. The project web site is located at www.annsincap.org.
What are the limitations of artificial neural networks?
An Artifical Neural Network can create a model based only on the data with which it is developed. Therefore, patients with clinical variables that are outside the range of variables used to develop the model will not be able to use this application. For example, the upper limit of PSA level for the lymph node spread model is 84.2 ng/mL. This model would not be valid for a patient with a pre-treatment PSA that is higher than 84.2 ng/mL. Further, this model is designed to predict the risk of lymph node spread in men with clinically localized prostate cancer only (not for advanced disease or for men who have not been diagnosed with prostate cancer).
ANNs in CaP is sponsored by the Institute for Clinical Research (ICR), a non-profit research and development organization affiliated with the Veterans Administration Medical Center in Washington, DC. ICR derives its funding from disease focused charities (AHA, NKF, etc.), government entities (DOD, NIH, etc.), and donations from private individuals, pharmaceutical companies and other organizations.
Additional financial support was provided by the National Cancer Institute, Specialized Program of Research Excellence (SPORE) grant NCI-CA-58236.
This service is a nonprofit endeavor, and does not receive any advertising revenue.
This page reported by J. Strax, last updated April 22, 2004
Prostate Cancer Treatment Guidelines for Patients