Abstract 2053: Dynamic 18F-FDG PET parameters variation in patient-derived breast cancer xenograft and correlation with outcome following treatment with cytotoxic agents

Abstract

Abstract Using pharmacokinetic modeling for dynamic 18FDG PET (D-PET) the temporal and spatial uptake pattern of 18FDG may be separated into components reflecting perfusion and metabolism. Patient-derived breast cancer xenografts (bcPDX) reflect morphology, gene and protein expression profiles of those in patients. The outcome for conventional treatment with cytotoxic agents may be dependent on the perfusion and tumor aggressiveness among others. The purpose of this study was to utilize D-PET for assessment of perfusion and glucose metabolism of 6 bcPDX and correlate these parameters with treatment outcome. Each of the 6 bcPDX, 3 basal-like (HBCx12a, HBCx28, HBCx39) and 3 luminal-like (HBCx3, HBCx34, MAS98.06), were inoculated bilateral in mammary fat pad of 6 athymic nude mice (ANM). D-PET (1hour scan) was performed with injections of 10 MBq 18FDG. A well-established three-compartment FDG distribution model was fitted to the uptake curves voxel-by-voxel. The rate constants k1 and k3, and vascular fraction vp were calculated. Each of 6 bcPDX was inoculated in 20 ANM. When tumors measured 500mm3, mice received either vehicle or paclitaxel 15mg/kg twice a week, doxorubicin 8mg/kg once every three week or carboplatin 10mg/kg once a week for 6 weeks. Tumor double time was calculated and correlated to k1, k3 and vp using a Spearman rank test. The median k1, k3, vp and tumor double time (TDT) (see table) were significantly different for the majority all bcPDX(p<0.05). TDT was not significantly correlated to k1, k3 and vp except for paclitaxel treatment, where TDT was increased for models with low k1, k3 and vp can be explained by higher treatment resistance of well vascularized and perfused tumor with high glucose metabolism. We showed that different bcPDX have different perfusion and glucose metabolism as measured by D-PET with low intermodal variations except HBCx28. D-PET appeared to be a robust technique and potentially can be used for monitoring of these parameters. Pharmacokinetic Modeling and treatment results and correlation between it.Median tumor double time (TDT)Median rate constants and vascular fractionsTreatment agentVehiclePaclitaxelDoxorubicinCarboplatinMedian perfusion rate constant, k1 (1/min)Median metabolic rate constant, k3 (1/min)Vascular fraction, Vp (%)Model CodeHBCx3Luminal-like, ER+2047731350.11980.02040.9540HBCx28Basal-like, tripple negative614182750.24840.05882.1815HBCx34Luminal Like, ER+11831536220.05080.00840.5200HBCx39Basal-like, tripple negative840216220.06510.00940.7125HBCx12ABasal-like, tripple negative1451631170.08810.03750.3.3470MAS98.06,Limunal-like, ER+1814931390.07650.10671.3190Spearman correlation coeficientk1 and TDT0.029-0.657-0.086-0.029p-value0.9570.156^0.8720.957k3 and TDT0.143-0.771-0.143-0.029p-value0.7870.072*0.7870.957Vp and TDT0.787-0.8860.2000.086p-value0.9570.019**0.7040.872 Citation Format: Alexandr Kristian, Mona-Elizabeth Revheim, Hong Qu, Gunhild Mælandsmo, Olav Engebraaten, Eirik Malinen. Dynamic 18F-FDG PET parameters variation in patient-derived breast cancer xenograft and correlation with outcome following treatment with cytotoxic agents. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2053. doi:10.1158/1538-7445.AM2014-2053