Staging the axillae in breast cancer patients with 18F-FDG PET/CT: How small are the metastases that we can detect with new generation clinical PET/CT?

Objectives PSF reconstruction improves the spatial resolution throughout the entire FOV of PET and could detect smaller metastatic deposits than conventional algorithms such as OSEM. We assessed the impact of PSF reconstruction on quantitative values and diagnostic accuracy for the axillary staging of breast cancer patients, as compared to an OSEM algorithm and compared these results to the size of nodal metastases.

Methods This was a single-referral-centre prospective study in which 50 patients underwent a PET examination prior to axillary lymph node dissection. Data were reconstructed with an OSEM algorithm and PSF reconstruction. PET data were analyzed blindly and validated by a single pathologist who measured the largest nodal metastasis per axillae. This size was used to dichotomise PET diagnostic performance.

Results Thirty-four patients had nodal involvement at pathology. Overall, median size for the largest nodal metastasis per axillae was 7 mm. PSF- and OSEM-PET results were concordant in 46 patients and discordant in 4 patients in whom PSF-PET was positive (3 true positive and 1 false positive) and OSEM-PET was negative. PSF-PET detected more involved nodes than OSEM-PET (P=0.003). The mean ratio between SUVmax for PSF and OSEM reconstructions was 1.66 (95% CI: 1.01-2.32). In patients with largest nodal metastasis > 7mm, sensitivity and specificity were 96% and 69% for PSF-PET and 92% and 75% for OSEM-PET, respectively. In patients with largest nodal metastasis < 7mm, sensitivity and specificity were 60% and 69% for PSF-PET and 40% and 75% for OSEM-PET, respectively.

Conclusions Thanks to an improved activity recovery, PSF-PET performs better than OSEM-PET for detection of axillary lymph node metastasis < 7 mm in breast cancer patients.