Positron emission mammography outperforms standard imaging in mapping newly diagnosed breast cancer

In 77 women with 80 surgically confirmed malignant breast lesions, positron emission mammography, or PEM, achieved almost perfect agreement with post-operative pathology in determining whether disease was confined to a single tumor, spread as multiple foci within one quadrant, or scattered across different quadrants of the breast. Conventional sono-mammography, by comparison, reached only fair agreement, frequently underestimating the true extent of disease and thereby risking undertreatment at surgery.

The authors report a single-site study conducted between October 2022 and January 2023, in which 115 patients with pathology-proven breast cancer were assessed and 77 met the inclusion criteria. The authors report that 27 patients with locally advanced disease requiring neoadjuvant chemotherapy were excluded, as were 11 patients who already had distant metastatic deposits at diagnosis. The 77 included participants, with a reported mean age of 51.4 years, underwent PEM in addition to standard sono-mammography, core biopsy, and up-front surgery for all 80 lesions. The authors report that imaging findings for lesion size and multiplicity were correlated and compared with the surgical pathology specimen results.

PEM is a functional imaging technique specialized for breast tissue. It relies on the radioactive glucose analog fluorodeoxyglucose, labeled with fluorine-18, which is taken up by cancer cells through glucose transporter-1, a protein overexpressed in breast malignancies. Once inside the cell, the tracer is phosphorylated and trapped, accumulating in proportion to metabolic activity. As the tracer decays, it emits positrons that collide with electrons, producing pairs of 511-keV gamma rays that dedicated detectors record while the breast is gently compressed between paddles in positions mirroring standard mammography views. Because PEM and conventional imaging examine the breast in the same cranio-caudal and mediolateral oblique orientations, correlating precise tumor location and size between the two becomes substantially easier.

According to the reported protocol, patients fasted for four to six hours, serum glucose was checked to ensure it was below 150 mg/dL, and an average dose of approximately 0.154 mCi/kg of FDG was administered. The reported protocol states that after a 45-to-90-minute rest, approximately 12 images of each breast were obtained per view, with an emission scan time of 10 minutes per orientation and a total scan time of 60 minutes. The authors report that two readers with seven and twelve years of breast imaging experience interpreted the images in consensus, classifying each abnormality as a focus, a mass, or non-mass uptake, and calculating the maximum PEM uptake value and the lesion-to-background ratio.

The headline result concerned lesion multiplicity. Against post-operative pathology, PEM achieved a kappa statistic of 0.869, indicating almost perfect agreement, while sono-mammography managed only 0.401, a fair level. The gap in raw performance was striking. Of 55 breasts with a single malignant lesion, PEM agreed with pathology in 53 cases, or 96.4 percent. For multifocal disease, confirmed by pathology in six breasts, PEM correctly identified four, whereas sono-mammography caught only two. For multicentric disease, present in 18 breasts, PEM agreed with pathology in 17 of 18 cases, a 94.4 percent sensitivity, while sono-mammography identified just 7, often mislabeling extensive disease as single or multifocal.

Standard diagnostic metrics reinforced the contrast. For multifocal disease, PEM achieved 66.7 percent sensitivity, 100 percent specificity, a 100 percent positive predictive value, and 97.5 percent accuracy, compared with 33.3 percent sensitivity and 86.2 percent accuracy for sono-mammography. For multicentric disease, PEM reached 94.4 percent sensitivity, 95.2 percent specificity, and 95.0 percent accuracy, against 38.9 percent sensitivity and 80.0 percent accuracy for conventional imaging. The authors report that this improved accuracy directly changed surgical management for 18 patients, 22.5 percent of the group, ensuring appropriate selection between breast-conserving surgery and modified radical mastectomy, for example by switching to mastectomy when multicentric disease was uncovered.

Tumor size measurement told a parallel story. Size is a primary prognostic factor in breast cancer, a key component of T staging, and a determinant of whether a patient is offered mastectomy, neoadjuvant chemotherapy, or breast conservation. Using intraclass correlation analysis, PEM showed substantial agreement with pathology for lesion size, with a coefficient of 0.724, while sono-mammography showed only fair agreement at 0.275. After excluding two patients, one with an occult primary and one whose multicentric invasive lobular carcinoma PEM failed to detect, the smallest lesion seen by PEM was 4 millimeters, closely matching the smallest invasive lesion of 5 millimeters found on pathology. In roughly four cases, PEM overestimated the largest lesion, a discrepancy the authors attribute to the restricted angular sampling of PEM’s planar detectors.

The study also examined how tracer uptake relates to tumor biology. Invasive ductal carcinomas exhibited higher uptake values and lesion-to-background ratios than invasive lobular carcinomas, and lobular tumors showed reduced metabolic sensitivity. The difference in lesion-to-background ratios between the subtypes was statistically significant, while the difference in maximum uptake values did not reach significance. One multicentric invasive lobular carcinoma, grade II, was entirely invisible on PEM despite being detected by sono-mammography, reflecting the diffuse growth pattern and low metabolic activity characteristic of lobular cancers. The authors caution that PEM findings must be interpreted carefully whenever lobular carcinoma, especially multicentric disease, is suspected.

The investigators are explicit about their study’s limitations. PEM images were interpreted by consensus rather than independently, so interobserver variability could not be assessed. The cohort included only malignant lesions, meaning no cutoff value for distinguishing benign from malignant uptake could be established, and lymph node metastasis was excluded from the evaluation, a gap that could affect overall staging decisions. The authors also stress that PEM is a functional adjunct, not a standalone diagnostic tool or a definitive means of tumor size estimation, and should complement rather than replace conventional imaging in pre-operative assessment.

Within those constraints, the findings position PEM as a targeted problem-solving tool rather than a routine screening test. The authors suggest it is particularly suited to patients considering breast-conserving surgery, those with inconclusive conventional imaging, and patients with contraindications to magnetic resonance imaging or contrast-enhanced mammography, since PEM offers a functional assessment of disease extent without those examinations’ requirements. Given that mammography’s sensitivity drops in dense breast tissue and that accurate distinction between multifocal and multicentric disease shapes the surgical approach, an imaging modality that reliably reveals additional tumor foci before the operating room carries real clinical weight. For the selected patients in this study, adding PEM to the pre-operative workup changed what the surgeon saw coming, and in nearly a quarter of cases, changed what the surgeon did.

Subject of Research: Diagnostic accuracy of pre-operative positron emission mammography for tumor size and multiplicity assessment in newly diagnosed breast cancer

Article Title: Accuracy of positron emission mammography for tumor size assessment and detection of other breast lesions in newly diagnosed breast cancer

Article References: Kamal, E., Mekhaimar, M., Hamid, R. A., Abdel-Salam, G., Sakr, S., Moustafa, A., Gomaa, M., Elsayed, H. F., Mikhael, H. S. W., & Kamal, R. (2026). Accuracy of positron emission mammography for tumor size assessment and detection of other breast lesions in newly diagnosed breast cancer. Egyptian Journal of Radiology and Nuclear Medicine, 57(1), Article 220. https://doi.org/10.1186/s43055-026-01871-4

Image Credits: AI Generated

DOI: 10.1186/s43055-026-01871-4

Keywords: positron emission mammography, breast cancer, tumor size, multifocal disease, multicentric disease, sono-mammography, FDG imaging, pre-operative staging, breast-conserving surgery, invasive lobular carcinoma, diagnostic accuracy, surgical management