DALLAS--Estradiol scans, currently being used in research for the diagnosis of early breast cancer (see illustration on page 1), have in some cases identified breast cancer 2 years prior to diagnosis by conventional means, such as mam-mography, physical examination, and ultrasound, David F. Preston, MD, said in an interview with Oncology News International.
DALLAS--Estradiol scans, currently being used in research forthe diagnosis of early breast cancer (see illustration on page1), have in some cases identified breast cancer 2 years priorto diagnosis by conventional means, such as mam-mography, physicalexamination, and ultrasound, David F. Preston, MD, said in aninterview with Oncology News International.
"We believe we are able to identify the regional estrogen-receptorexcess of early breast cancer prior to histological changes asseen by the light microscope," said Dr. Preston, professorof radiology and nuclear medicine, and head of the Division ofNuclear Medicine, University of Kansas Medical Center, KansasCity.
Even tumors later labeled as estrogen-receptor negative have beenseen using radiolabeled estradiol SPECT (single photon emissioncomputed tomography), he added.
In a poster presentation at the American College of Radiology's27th National Conference on Breast Cancer, Dr. Preston describeda study of 44 patients with breast cancer or a high risk of breastcancer who were referred from the oncology department or highrisk breast clinic. The study, which is ongoing, has funds toenroll up to 60 patients.
After patients give informed consent, their thyroid is blockedwith stable iodine for 3 days. Then, 40 µCi/kg of I-123 E-2is given intravenously. This compound is a high-spe-cific-activityiodine halogen (16-alpha-iodine I-123) that is attached to iodo-estradiol17-beta. Two hours later, standard SPECT is performed.
"There is rapid blood clearance of the estradiol followinginjection," Dr. Preston said. Each dose includes approximately2.5 mCi of 123I attached to approximately 100 ng of estradiol.The greatest radiation exposure is approximately 4 rads (0.04Gy) by the descending colon, "approximately the dose receivedduring a barium enema," he said.
Regions of breast tissue on the scan that are found to have numericalvalues (counts per pixel) two times that of background lung tissuevalues are considered to be cancer. "Thus, the techniquecould offer a numerical way to diagnose early-stage breast cancer,"Dr. Preston said.
The actual presence, location, and type of breast disease in thesepatients was later established by autopsy, biopsy, aspiration,estrogen-receptor assay, or clinical follow-up.
Using the numerical criterion for determining cancer, in one patient,estradiol SPECT identified a cancer 2 years before clinical detection.Two cancers were identified with SPECT months prior to clinicaldetection.
The study found the sensitivity of estradiol SPECT to be 0.88;specificity, 0.86; positive predictive value, 0.78; and negativepredictive value, 0.92. "It proved more specific and sensitivethan mammography," he said.
In this study, estradiol SPECT imaging was able to detect bothestrogen-receptor-positive tumors and tumors called estrogen negative."There is evidence that breast cancers begin as an estrogen-receptor-positiveentity, and, as they age, the older parts become estrogen-receptornegative. Because of this, estrogen-negative cancers may havesome sections that do contain an estrogen-receptor excess,"Dr. Preston said to explain the paradox.
Almost half of the tumors identified to date in the study patientshave been labeled estrogen-receptor negative by assay, "yetwe have been able to identify them by imagable estradiol SPECTscanning because we see the receptor-positive regions," hesaid.
Although the technique is still experimental, Dr. Preston believesit is an extremely promising method for the early detection ofbreast cancer.
"If we see areas of concern, we can then direct attentionto those areas using conventional techniques," he said. "Withearly detection, women have more treatment options."
Dr. Preston's colleagues in the study are Jay Spicer, MS, RichBaranczuk, PhD, Margie Schiefelbein, RT, Kirkman Baxter, MD, MarkRedick, MD, and Norman Martin, MD, all from the University ofKansas Medical Center.