DALLAS--A new study suggests that RODEO breast MRI can accurately determine the extent of DCIS (ductal carcinoma in situ) and identify characteristic patterns that can differentiate DCIS from more malignant breast cancers.
DALLAS--A new study suggests that RODEO breast MRI can accuratelydetermine the extent of DCIS (ductal carcinoma in situ) and identifycharacteristic patterns that can differentiate DCIS from moremalignant breast cancers.
DCIS is biologically a different lesion from infiltrating cancer,and, if accurately diagnosed, conservative treatment can producegood results, Steven E. Harms, MD, of the University of ArkansasCancer Research Center, Little Rock, said at the American Collegeof Radiology's National Conference on Breast Cancer.
However, both conventional MRI and mammography have proved inadequateat predicting the extent of DCIS and distinguishing DCIS frominfiltrating cancer, leading to a tendency to overtreat the disease.
To provide better detection of DCIS and other breast lesions,Dr. Harms, who was then at Baylor University Medical Center, Dallas,developed a high-contrast, high-resolution three-dimensional MRItechnique called RODEO (rotating delivery of excitationoff-resonance). [For a report of RODEO in lobular breastcarcinomas, see Oncology News International, July, 1996,"RODEOBreast MRI Promising in Lobular Cancer Diagnosis".]
The study presented at the symposium involved 22 breast cancersdiagnosed at pathology as pure DCIS (five cases), DCIS with microinvasion(six cases), or infiltrating ductal carcinomas with extensiveintraductal component (EIC) (11 cases).
The study looked at the new technique's ability to determine bothdisease morphology (infiltrating carcinoma or DCIS) and extent.
In terms of lesion morphology, mam-mography did not demonstratea consistent pattern, Dr. Harms said. On mam-mography, 56% ofinvasive ductal carcinomas with EIC appeared as a poorly circumscribedmass with malignant microcalcifications, while 33% appeared onlyas a poorly circumscribed mass.
He noted that in the cases of pure DCIS and DCIS with microinvasion,mammography showed malignant appearing calcifications in 78%,indetermi-nant calcifications in 11%, and asymmetric density in11%.
DCIS frequently produces calcification that can be identifiedby mammography, Dr. Harms said, but the extent is often difficultto determine accurately by mam-mography (see figures). "Themalignant microcalcifications seen on mam-mography are not specificfor DCIS and can be present in more malignant forms of breastcancer," he noted.
Previously it was thought that MRI could not demonstrate microcalcifica-tions,he said. However, of the 14 lesions with microcalcification seenon mam-mography, 12 demonstrated microcalcifi-cation on RODEOMRI.
There was a definite trend in morph-ologic distinction with RODEOMRI, Dr. Harms said. The technique showed clumped, linear, orspiculated enhancement in all lesions containing DCIS components,and, unlike mammography, MRI in most cases could differentiateinvasive carcinoma from DCIS because of spiculated enhancement."This finding can help direct biopsies and plan appropriatetreatment," he said.
In the pure DCIS cases, "we saw only clumped enhancementwith no evidence of spiculation," Dr. Harms said, addingthat 82% of infiltrating cases with EIC and 67% of DCIS lesionswith microin-vasion showed spiculation.
In terms of tumor extent, RODEO was accurate in 95% of cases vs74% with mammography. Although mammogra-phy showed abnormalitiesin all cases, the determination of tumor extent correlated poorlywith pathology results in pure DCIS and DCIS with microinvasion.
RODEO MRI studies were also abnormal in all cases, "and wehad excellent agreement with the pathology report in all categories,"Dr. Harms said. In one case, he added, MRI disagreed with pathologybecause MRI had overestimated the extent of disease.
To sum up, Dr. Harms said, RODEO breast MRI can detect pure DCISand DCIS with microcalcifications; it can often distinguish pureDCIS from invasive cancer; and it can accurately determine tumorextent.