PET measure sheds light on response to chemo for breast cancer
The change in total lesion glycolysis measured prior to and after two cycles of chemotherapy provided a better prediction of response to chemotherapy than other PET measures, including change in standardized uptake value (SUV), according to a study published online Jan. 17 in The Journal of Nuclear Medicine.
Most studies assessing tumor response to neoadjuvant chemotherapy via PET measures have focused on the decrease in SUV measurements after two cycles of chemotherapy. However, recent studies have suggested that other 18F-FDG PET-derived measures may provide a statistically significant higher predictive value. Thus, Mathieu Hatt, MD, from INSERM in Brest, France, and colleagues devised a study to evaluate the predictive value of several PET-derived measures at baseline and after two cycles of chemotherapy.
The study cohort included 51 patients diagnosed with breast cancer who underwent PET/CT at baseline and after the second round of neoadjuvant chemotherapy at Saint Louis Hospital in Paris between July 2007 and May 2009.
Histopathologic response was determined on surgical specimens at the completion of chemotherapy. Researchers also determined pathologic complete response (pCR) and assessed the power of PET measures to predict pCR.
Among the 51 patients, 13 had triple-negative tumors, 12 had human epidermal growth factors receptor 2 (HER2)-positive tumors and 26 had estrogen receptor (ER)-positive/HER2-negative tumors. The pathologic response at surgery established 27 women as pathologic responders and 24 as nonresponders. pCR was reported in three patients with triple-negative tumors and four with HER2-positive tumors.
Hatt and colleagues compared maximum and mean SUV, metabolically active tumor volume (MATV) and total lesion glycolysis (TLG) at baseline and their variation after two cycles of chemotherapy. None of the parameters correlated with response at baseline.
After two cycles of chemotherapy, the reduction of each measure was significantly associated with response. The change in TLG best predicted response and provided 96 percent sensitivity, 92 percent specificity and 94 percent accuracy. MATV also outperformed SUV. Among the SUV measures, maximum SUV offered the highest predictive value, with 77 percent accuracy.
When the researchers evaluated subgroups, they found significantly higher accuracy for TLG change than SUV change for ER-positive/HER2-negative tumors but not for triple-negative and HER2-positive tumors.
“Future validation studies on larger groups of patients will be focused on ER-positive/HER2-negative patients since it is for this subgroup that the added value of change in MATV and change in TLG over change in SUV seems to be the most significant,” wrote Hatt et al.
The clinical value and efficiency of FDG-PET for early prediction of response to neoadjuvant chemotherapy may increase if the current results can be confirmed in a larger cohort of ER-positive/HER2-negative women, the researchers concluded.