Standardized uptake value (SUV) by positron emission tomography/computed tomography (PET/CT) as a prognostic variable in metastatic breast cancer (MBC).

Abstract

567 Background: The accurate prediction of outcome from MBC is challenging. PET/CT could be useful as it combines anatomical with functional imaging. This could enable greater individualization of treatment. However, there is substantial SUV variation by site. In this retrospective, single-institution study, we examine baseline SUV on PET/CT as a predictor of outcome from MBC. Methods: Patients (Pts) with ≥1 metastatic lesion on PET/CT performed ≤60 days of diagnosis of MBC from 01/01/2001 to 12/31/2008 were identified through institutional databases. Pts who received chemotherapy < 30 days prior to PET/CT were excluded. Electronic medical record reports were reviewed and maximum SUV (SUV-MAX) by site for lesions in bone, liver, lung, and lymph node (LN) was recorded. In a secondary analysis, PET/CT scans were reviewed and SUV-MAX recalculated. Relationships between SUV-MAX tertiles and OS were assessed using Cox regression by site. Results: We identified 285 pts, median age 57 yrs (range 27-90) who had PET/CT at median of 2.3 yrs (range 0-41) from primary BC (67% ER+ and 21% HER2+). The median time between PET/CT and MBC diagnosis was -9 days (range -58 to 59). At median follow-up of 53 mths, 163 pts have died and median OS is 41 mths (95%CI 34-48). The SUV-MAX by site was; bone (N=159) median 7.0 (range 2.1 - 29.6); liver (N=55) median 8.2 (range 2.9 - 51.2); lung (N=89) median 4.7 (range 1.1 - 24.0); LN (N=180) median 6.9 (range 1.2 - 34.0). On univariate analysis, higher SUV in bone was associated with shorter survival (p<0.001); table. This was maintained in multivariate analyses after adjusting for known prognostic variables (p=0.02). A similar trend for shorter survival for higher SUV was noted in liver (p=0.07). However, no relationship between SUV and OS was noted in lung (p =0.34) and LN (p=0.6). Conclusions: This large retrospective study of pts with chemotherapy-naïve MBC suggests that SUV-MAX in bone strongly correlates with prognosis. SUV in bone: Univariate analysis for OS. Variable N (%) N P value HR Max SUV Tertiles 159 <.001 1 54 (34%) – (ref) 2 53 (33%) – 1.42 3 52 (33%) – 2.76 ER neg 38 (24%) 159 <.001 2.55 HER2 neg 122 (80%) 152 0.56 0.86 Grade 3 103 (75%) 138 0.03 1.95