In Vivo Measurements of Tumor Metabolism and Growth after Administration of Enzastaurin Using Small Animal FDG Positron Emission Tomography

Background. The use of 2-[18F]fluoro-2-deoxy-D-glucose ([18F]FDG) may help to establish the antitumor activity of enzastaurin, a novel protein kinase C-beta II (PKC-βII) inhibitor, in mouse xenografts. Methods. The hematologic cell line RAJI and the solid tumor cell line U87MG were each implanted in NOD/SCID mice. Standard tumor growth measurements and [18F]FDG PET imaging were performed weekly for up to three weeks after tumor implantation and growth. Results. Concomitant with caliper measurements, [18F]FDG PET imaging was performed to monitor glucose metabolism. Heterogeneity of glucose uptake in various areas of the tumors was observed after vehicle or enzastaurin treatment. This heterogeneity may limit the use of [18F]FDG PET imaging to measure enzastaurin-associated changes in xenograft tumors. Conclusion. [18F]FDG PET imaging technique does not correlate with standard caliper assessments in xenografts to assess the antitumor activity of enzastaurin. Future studies are needed to determine the use of [18F]FDG PET imaging in preclinical models.


Introduction
Imaging techniques play an important role in the diagnosis, staging, and follow-up of cancer patients. While standard imaging techniques, such as computed tomography (CT), are based on differences in the anatomical structure of the tissues, positron emission tomography (PET) uses radiolabeled molecular probes to assess differences in biological or biochemical properties of tissues [1]. The most commonly used tracer for PET is the glucose analogue 2-[ 18 F]fluoro-2-deoxy-D-glucose ([ 18 F]FDG), which provides an estimate of tissue glucose utilization. Today, [ 18 F]FDG PET is widely used in clinical diagnoses of cancer, including lung cancer [2], non-Hodgkin lymphoma [3], and glioblastoma [4]. Because of this increased utility in oncology [5][6][7], [ 18 F]FDG PET is being evaluated as a tool to assess antitumor effects of standard or novel anticancer drugs in both human subjects and in animal models of cancer [8]. PET imaging may provide evidence of biological responses of novel anticancer compounds, which, in turn, can facilitate the transition of compounds from preclinical to clinical investigation. One such novel compound is enzastaurin, which was developed as a Protein kinase C-beta (PKC-β) inhibitor.
The family of Protein Kinase C (PKC) has been implicated in processes that control tumor growth, survival, and progression [9]. In particular, PKC-β activation has been recognized as an important contributor to malignant growth in diffuse large cell B cell lymphoma [10] and in glioblastoma [11]. Recently, enzastaurin has shown antitumor activity in xenograft models of the colorectal cancer cell line HCT116 and in the glioblastoma cell line U87MG [12]. While enzastaurin was designed as a selective PKC-β inhibitor, recent studies suggest that its antitumor activity is modulated by activation of GSK-3β and the PI3K/AKT pathway [12]. In this study we evaluated the extent to which [ 18 F]FDG PET imaging can accurately characterize the antitumor activity of enzastaurin in two different mouse xenograft tumor models.

Xenograft Tumor Studies.
Prior to subcutaneous (s.c.) injection, cells were resuspended in a 1 : 1 ratio of tumor cells : matrigel (BD Biosciences, Bedford, Mass, USA) and Burkitt's lymphoma cell line, RAJI was resuspended in a 1 : 2 ratio of tumor cells : matrigel. Each mouse was injected s.c. in the right flank with 5 × 10 6 cells. Mice were monitored daily for palpable tumors.

Enzastaurin Administration.
Enzastaurin treatment was initiated when the tumors reached a volume of at least 150 mm 3 . Mice with similar tumor sizes were matched in the control and enzastaurin treated groups. Enzastaurin was suspended in 10% acacia (Fisher Scientific, Fair Lawn, NJ, USA) in water and dosed by gavage twice daily at 75 mg/kg based upon weekly body measurements for each treated group. Control groups were treated only with vehicle.

PET and CT Imaging.
Each animal was anesthetized with acepromazine (1-2 mg/kg i.m.) and torbugesic (2 mg/kg i.m.) and placed on a custom bed for imaging. Animals were administered 0.5-1 mCi of [ 18 F]FDG via a tail vein injection. A static 15-minute PET study was performed using the IndyPET II scanner [14,15] at 45 minutes posttracer injection. Following the PET study, the animal bed was moved to and mounted on an EVS RS9 microCT scanner, and a volumetric image that encompassed the tumor volume was imaged at approximately 90 micron spatial resolution.

FDG Utilization
Estimates. FDG uptake estimates are generated by placing ROIs on PET images acquired over the time period from 45-60 minutes posttracer administration. Indices of tumor FDG uptake were generated by calculating a ratio of the tumor to muscle uptake and by calculating the standardized uptake value (SUV): In (1) C t (t) is the concentration of [F-18]

Results
We first investigated whether two tumor cell lines provided acceptable tumor growth in NOD/SCID mice to allow reproducible imaging with [ 18 F]FDG PET ( Figure 1). We used U87MG cells as a representative cell line for solid tumors, and RAJI as a model for hematologic cancers. Both the U87MG (Figure 1(a)) and the RAJI (data not shown) tumor types grew consistently in NOD/SCID mice prior to drug treatment and [ 18 F]FDG PET imaging reliably detected glucose uptake in xenografts as determined by Standardized Uptake Value (SUV) (see Materials and Methods). In these feasibility studies, we also found that SUV correlated with U87MG tumor size as measured by CT (Figure 1(b)). Based on these preliminary observations of initial tumor growth, we elected to use both tumor cell lines to evaluate enzastaurin-induced metabolic changes as detected by Enzastaurin-induced metabolic changes were evaluated using [ 18 F]FDG uptake in two independent experiments for each tumor cell line, U87MG and RAJI, respectively (Table 1). Consistent with previous studies in nude mice [12], enzastaurin induced tumor growth delay in NOD/SCID mice implanted with U87MG and RAJI (Figures 2(a) and 2(b)). U87MG cells grew slower than RAJI cells, and enzastaurin had a tumor growth delay mainly in the U87MG tumor cell model (Figures 2(a) and 2(b)). A significant tumor growth delay was seen in U87MG after treatment with enzastaurin over the period of 3 weeks (Figure 2(a)). In the RAJI xenograft model, there was a trend in the tumor growth delay, but not a statistically significant difference between  (Figure 2(c)). Compared to RAJI, the SUV changes in U87MG xenografts occurred at smaller signal intensities, and thus contributed to an overlap of SUV measurements between vehicle-and enzastaurin-treated mice (compare Figures 2(c) and 2(d)). In RAJI, SUV uptake appeared to be increased at weeks 2 and 3 in enzastaurintreated compared to vehicle-treated mice (P < .10 at weeks 2 and 3; Figure 2(d)). It is possible that metabolic heterogeneity of the tumor glucose metabolism in different areas of the tumor may have caused this difference in SUV uptake. In addition to SUV, we also used tumor/muscle ratio to determine the metabolic effect of enzastaurin in tumors. The tumor/muscle ratio uses muscle tissue with its low metabolic rate to normalize the tumor tissue [ 18 F]FDG uptake. Based on this analysis there was no clear evidence of a metabolic change induced by enzastaurin compared to vehicle treatment (Figures 2(e) and 2(f)). However, for U87MG there was a trend in FDG uptake in enzastaurintreated mice (P < .10 at week 3), which was not observed in RAJI xenografts (Figures 2(e) and 2(f), resp.). Next, we determined tumor volume by using standard volumetric measurements based on CT. After enzastaurin treatment, we observed a trend in tumor size reduction for RAJI xenografts but not for U87MG xenografts at week 2 (P < .10) (Figures 2(g) and 2(h)). While the collective analysis of all animals from 2 independent experiments did not reveal a clear distinction between enzastaurin-and vehicle-treated animals, there were some animals that did show a metabolic change after enzastaurin treatment. In these cases we saw a decline only after two weeks of treatment that was essentially undetectable after the third week of treatment ( Figures  3(a)-3(f) and Figures 4(a)-4(f)). The changes observed in U87MG were different from changes observed in RAJI. In U87MG xenografts, [ 18 F]FDG uptake declined after the 2nd week of treatment. The RAJI tumors were fast growing and had large areas of necrotic tissue and some areas with newly increased [ 18 F]FDG uptake ( Figure 5). The growth pattern of these tumors is partly responsible for the intertumoral heterogeneity seen in this study and may contribute to the lack of detecting enzastaurin-induced changes in the tumor.

Discussion
In this study we used a specialized PET imaging approach which was developed to assess activity of anti-cancer agents in small animals [14]. Changes in [ 18 F]FDG uptake were generally found to correlate with reduction in tumor size as determined by caliper measurements [17,18]. Our study uses a metabolic kinase inhibitor to compare drugmediated tumor growth reduction with metabolic alterations in vivo. While previous imaging studies evaluated anti-tumor activity of cytotoxic agents, it is not clear how [ 18   . Tumor glucose metabolism changes were measured by SUV (Panels (c), (d)) and tumor/muscle ratio (Panels (e), (f)) in U87MG and RAJI xenografts. Only in RAJI xenografts enzastaurin treatment showed a trend in increased SUV (Panel (d); P < .10 at weeks 2 and 3). Using tumor/muscle ratio indicated a trend for FDG uptake in U87MG (P < .10 at week 3) (Panel (e)), but not in RAJI xenografts (Panel (f)). The tumor size assessment based on CT scan (Panels (g) and (h)) indicated a trend for detecting reduced tumor size only in RAJI xenografts after enzastaurin treatment at week 2 (P < .10). Mice treated with vehicle alone are shown in green; mice treated with enzastaurin are shown in red, representation of 2 independent experiments.  inhibitors. For instance, imatinib activity was evaluated in mice with limited success [19]. In addition to its cost, the use of PET imaging may be limited due to the spatial resolution of current PET scanners [20]. Thus, few studies have been published which evaluate the use of small animal imaging in drug discovery. For the first time, we assessed the anti-tumor activity of the serine/threonine kinase inhibitor enzastaurin by [ 18 F]FDG uptake in mice. Because we used NOD/SCID instead of conventional nude mice, we first confirmed that [ 18 F]FDG PET images could reproducibly be obtained in xenografts of glioblastoma and lymphoma. This feasibility assessment was important to establish the tumor size at which [ 18 F]FDG uptake is detectable in mice. Tumors had to be at least 150 mm 3 in volume to be visualized by the scanner, and the best assessment was observed in tumors that were more than 400 mm 3 (Figure 1). This is consistent with other studies which reported on the need of specific tumor sizes for scanner assessment [21]. The subsequent studies with enzastaurin treatment did not provide clear evidence of enzastaurin-induced metabolic changes in either of the two tumor types examined. However, there are several possibilities why enzastaurin-induced changes were not detected by [ 18 F]FDG PET imaging.
First, it is possible that enzastaurin may not have a homogenous impact on the metabolic rate in the tumor tissue. Although PKC isoenzymes have been implicated in cell proliferation [9], their specific role during glucose metabolism is still not understood. On one hand, glucose can induce PKC-β expression [22], and on the other hand, overexpression of PKC-β reduces glucose uptake in cells [23]. Hence, selective PKC-β inhibitors have been investigated as potential treatments for diabetes [24]. Whether such a PKC-β-dependent glucose regulation exists in tumor cells  is not known [25]. Considering the observation of this study, in which tumor growth delay and glucose metabolism are not correlated with enzastaurin activity in xenograft tumors, it is possible that enzastaurin is not able to modulate the complex glucose regulation in the tumor cells [26].
Recently, the antiangiogenic kinase inhibitor AZD2171 was also evaluated in a small animal study for its metabolic change in tumors.  study, when animals were sacrificed, and tumors had become heterogeneous.
Finally, the observed [ 18 F]FDG uptake signals are considerably lower relative to what is measured in humans. The low baseline level in the present xenograft study negatively impacted the ability to assess early changes in metabolism associated with treatment effect. Other factors influencing the low uptake signal are heterogeneity of glucose uptake in tumors, possible emergence of drug-resistant tumor cells, variability of pharmacokinetic exposure of the inhibitor, and the diffusion of tracer to the tumor site.
In summary, our data are consistent with the hypothesis that shrinking tumor size (assessed by caliper measurements) does not equate to reduction in overall tumor metabolism. In fact, "pockets" of drug-resistant tumor cells with increased glucose uptake and growth kinetics may become prevalent in certain areas of the tumor, while glucose reduction will be present in other areas. Additional longitudinal PET imaging studies will have to examine in detail the variables which affect the accurate measurement of tumor response to drug treatment. Our study implies that [ 18 F]FDG PET imaging will be useful only in a select type of tumors and perhaps detect antitumor activity only for a limited number of drugs.