It is well established that change in FDG tumor uptake after chemotherapy can predict pathological response in primary LABC tumors. Assessment of tumor response to NACT by traditional methods like clinical examination (CE), mammography (MAM) and ultrasonography (USG) are unreliable.

Although the presence of residual tumors in axillary lymph nodes after chemotherapy is an independent risk factor for loco-regional recurrence in LABC, response monitoring data in the axillary status are scarce ; no cut-off values for accurate prediction of lymph node responders have been reported.

Material and methods: 6 patients (age range 29-58 years) with non-inflammatory, large (>3 cm), or locally advanced cancer received four doses of primary chemotherapy. PET imaging was performed in the same conditions, without significant change of blood sugar levels or total body weights, before treatment and 21 days after the second dose or treatment. We used high dose FDG (8MBq per Kg), 8 minutes per bed position, and delayed acquisition (1 hour and 3 hours post injection) in order to increase potential sensitivity of FDG.

Primary tumors and involved axillary lymph nodes were initially identified by CE, MAM and USG and confirmed by evaluation of surgical resection specimens and axillary lymph node dissection (ALND). PET results were evaluated visually and semi-quantitatively by measuring standardized uptake value (SUV), and were compared to the tumor response to chemotherapy as judged by histological evaluation. Patients were classified as responders or non responders taking in account the results of ALND.

Results: All axillary lymph nodes detected by either CE and/or USG were identified by PET-FDG/CT. Disappearance of some of the lymph node images after treatment was observed in 3 patients, but this result was not correlated with ALND findings.

All diagnosed breast malignant lesions were characterized by high uptake of FDG in all lesions before treatment. Histological evaluation was 3 responders and 3 non responders.

There was a clear difference between SUV before chemotherapy (15.3 +/- 7.4 (mean +/- SD) at 1 hour and 18.42 +/- 9.2 at 3 hours) and after second course of chemotherapy (3.32 +/- 3.8 at 1 hour and 3.86 +/- 3.8 at 3 hours) in responders.

There was a lesser difference between SUV before (29.9 +/- 4.2 at 1 hour and 34.5 +/- 4.6 at 3 hours) and after second course of chemotherapy (12.9 +/- 1.4 at 1 hour and 14.6 +/- 1.6 at 3 hours) in non-responders.

Conclusion: These preliminary results are consistent with the expected correlation between the decrease in metabolism and the histological response for tumor, but not for axillary lymph node. Usefulness of delayed (3 hours) acquisition in terms of predictive value remains unclear. Methodological improvements appear clearly necessary for lymph node study, and advisable for tumor study.