Radiology: PEM complements MR, but gaps remain
With previous data indicating that PEM may be more specific and possibly more accurate than MR, researchers designed the current study to detail the performance characteristics of PEM compared with MR, including the effect on surgical breast management, in ipsilateral breasts with cancer, wrote lead author Wendie A. Berg, MD, from the American College of Radiology Imaging Network in Lutherville, Md.
The researchers at six sites recruited 388 female candidates for breast-conserving surgery (median age, 58 years) with newly diagnosed invasive and/or intraductal breast cancer to undergo PEM and MR exams. Imaging results were independently reviewed by different investigators blinded to the results of the other exam. After all surgeries, Berg and colleagues documented and characterized mastectomies as appropriate or inappropriate, obtained data about modifications in surgical management based on imaging results and gauged the accuracy of estimates of disease extent by PEM and MR.
“PEM tended to better depict cancer when it was present, depicting 357 of the 386 foci versus 344 foci depicted with MR imaging. Biopsy sites were more readily identified with MR imaging,” wrote Berg. Eighty-two of the 388 breasts were found to have additional tumor foci, with 60 percent of these identified via MR, 51 percent identified via PEM and 27 percent identified with conventional imaging. “The addition of conventional imaging review to either MRI or PEM significantly improved the sensitivity of MR imaging or PEM alone. Integrating PEM and MR findings significantly improved the detection of additional cancer: to 61 of 82 breasts versus 49 of 82 breasts with MR imaging alone,” continued the investigators.
MR imaging characterized 61 of the 116 malignant lesions unknown at study entry as suspicious. PEM characterized 47 and conventional imaging 24 as suspicious. However, PEM delivered greater specificity than MR in the characterization of benign lesions, characterizing 151 of the 189 as negative compared to 124 of the 189 via MR. PEM was less likely than MR to prompt unnecessary biopsies.
PEM yielded a superior positive predictive value versus MR when it prompted biopsies at 66 percent versus 53 percent, respectively. The addition of PEM to MR significantly improved the detection of DCIS over MR alone from 39 percent to 57 percent, and the addition of PEM to MR also improved detection of invasive cancer over MR alone from 64 percent to 73 percent.
MR outperformed PEM among the 56 participants who required mastectomy based on the extent of disease. MR correctly identified 71 percent of the cases, whereas PEM correctly identified 36 percent. However, MR findings prompted five inappropriate mastectomies, and PEM prompted one.
Berg and colleagues reported that MR was more accurate than PEM in surgical planning, with MR accurate in 75 percent of breasts and PEM accurate in 67 percent.
Given the improved detection of additional malignancies with conventional image review, Berg wrote, “[o]ur results suggest that a coordinated review of all breast imaging studies at the time of PEM or MR image interpretation is important for improved diagnosis.”
The authors surmised, “PEM proved to be complementary to MR imaging for defining preoperative disease extent in the ipsilateral breast of women with newly diagnosed breast cancer … even the combination of PEM and MR imaging did not fully depict disease extent, particularly in cases with an extensive intraductal component, multifocal disease, or multicentric disease--that is, the very patient populations anticipated to benefit most from accurate preoperative assessment of disease extent.”