Monitoring of anti-cancer treatment with ¹⁸F-FDG and ¹⁸F-FLT PET: a comprehensive review of pre-clinical studies

Abstract

Functional imaging of solid tumors with positron emission tomography (PET) imaging is an evolving field with continuous development of new PET tracers and discovery of new applications for already implemented PET tracers. During treatment of cancer patients, a general challenge is to measure treatment effect early in a treatment course and by that to stratify patients into responders and non-responders. With 2-deoxy-2-[¹⁸F]fluoro-D-glucose (¹⁸F-FDG) and 3’-deoxy-3’-[¹⁸F]fluorothymidine(¹⁸F-FLT) two of the cancer hallmarks, altered energy metabolism and increased cell proliferation, can be visualized and quantified non-invasively by PET. With ¹⁸F-FDG and ¹⁸F-FLT PET changes in energy metabolism and cell proliferation can thereby be determined after initiation of cancer treatment in both clinical and pre-clinical studies in order to predict, at an early time-point, treatment response. It is hypothesized that decreases in glycolysis and cell proliferation may occur in tumors that are sensitive to the applied cancer therapeutics and that tumors that are resistant to treatment will show unchanged glucose metabolism and cell proliferation. Whether ¹⁸F-FDG and/or ¹⁸F-FLT PET can be used for prediction of treatment response has been analyzed in many studies both following treatment with conventional chemotherapeutic agents but also following treatment with different targeted therapies, e.g. monoclonal antibodies and small molecules inhibitors. The results from these studies have been most variable; in some studies early changes in ¹⁸F-FDG and ¹⁸F-FLT uptake predicted later tumor regression whereas in other studies no change in tracer uptake was observed despite the treatment being effective. The present review gives an overview of pre-clinical studies that have used ¹⁸F-FDG and/or ¹⁸F-FLT PET for response monitoring of cancer therapeutics.

Introduction

During treatment of cancer patients a general challenge is to measure treatment effect early in a treatment course and by that to stratify patients into responders and non-responders. An advantage for non-responding patients is that shift to other therapies may be done early in the treatment course thereby avoiding unnecessary side-effects of inefficient treatment. However, this requires reliable biomarkers that can accurately predict the treatment outcome. The amount of patients responding to chemotherapy is in many cases 30% or less, this being both to conventional cytotoxic therapy and new targeted therapies [1-4]. Response monitoring has therefore become increasingly important as it can allow for individualized tailored therapy.

Functional imaging of solid tumors with positron emission tomography (PET) imaging is an evolving field with the continuous development of new PET tracers and new applications for existing PET tracers [5]. Two of the cancer hallmarks, altered energy metabolism and increased cell proliferation, can be visualized non-invasively with the widely used PET tracers 2-deoxy-2-[¹⁸F]fluoro-D-glucose (¹⁸F-FDG) and 3’-deoxy-3’-[¹⁸F]fluorothymidine (¹⁸F-FLT) [6-8].

It is hypothesized that decreases in glycolysis and cell proliferation may occur in tumors that are sensitive to the applied anti-cancer treatment and that tumors resistant to treatment will show unchanged glucose metabolism and cell proliferation. Furthermore, changes in physiological processes, e.g. metabolic or proliferative pathways are expected to precede morphological changes and changes in growth rate of the tumor. ¹⁸F-FDG and ¹⁸F-FLT have therefore, both in pre-clinical and clinical studies, been evaluated as imaging biomarkers that can predict and assess responses to various types of anti-cancer therapies including conventional chemotherapeutic drugs and newer targeted anti-cancer therapies in various tumor types. Accordingly, ¹⁸F-FDG PET has the potential to facilitate and accelerate drug development by shortening phase II and III using ¹⁸F-FDG PET as a surrogate for clinical outcome [9]. Pre-clinical studies in mice with human tumor xenografts may help to predict the expected ¹⁸F-FDG and ¹⁸F-FLT outcome for specific therapies in later clinical studies.

In this review we therefore present a comprehensive overview of pre-clinical studies that have used either ¹⁸F-FDG and/or ¹⁸F-FLT PET for response monitoring of cancer therapeutics.

PET imaging

PET is an imaging technique that allows for non-invasive functional imaging in living subjects and dependent on which PET tracer is used different molecular and cellular processes can be visualized without acquiring invasive biopsies. The most widely used PET tracer is the glucose analogue ¹⁸F-FDG and in oncology ¹⁸F-FDG PET is applied for tumor diagnosis, staging and monitoring of cancer as well as for monitoring of residual disease after completion of a treatment course [9]. ¹⁸F-FLT is a widely studied tracer for assessment of cell proliferation [8,10]. ¹⁸F-FDG PET is most often positive at a baseline scan of human solid tumors, whereas a wide range of tumor avidities is observed for ¹⁸F-FLT [11,12].

The change in tumor size after treatment is often used as a surrogate marker of survival in clinical studies, as for many cancer types, tumor shrinkage has been correlated with overall survival [1,13]. In clinical studies treatment monitoring can be performed using the Response Evaluation Criteria In Solid Tumors (RECIST) [14]. The RECIST criteria are based on anatomical tumor burden measured by computed tomography (CT). In 2009, an updated version of the RECIST guideline, RECIST 1.1, was published. In RECIST 1.1, ¹⁸F-FDG PET measurements have been incorporated; however, only as an adjunct to determination of progression by identification of new lesions. The new guideline includes comments on the possibility for future use of PET for treatment evaluation in clinical trials when the technique becomes more standardized and widespread available [14]. Furthermore, in 2009, a guideline proposing the use of ¹⁸F-FDG PET for tumor response assessment termed PET Response Criteria In Solid Tumors (PERCIST) was published. The PERCIST guideline proposes the use of ¹⁸F-FDG PET for tumor response assessment [4]. In this guideline it is argued that treatment can be effective despite minimal changes in tumor size which is a concern especially during treatment with cytostatic therapies.

¹⁸F-FDG-PET

In the 1970s the first whole-body ¹⁸F-FDG PET was acquired and ¹⁸F-FDG PET has subsequently become widely available and is frequently used in cancer diagnostics, staging and monitoring of recurrent and residual disease after completion of a treatment course [9]. ¹⁸F-FDG is a glucose analogue where the 2-carbon hydroxyl group has been substituted with an ¹⁸F isotope. Like glucose, ¹⁸F-FDG is taken up in cells by the glucose transporters (GLUT) and thereafter phosphorylated by hexokinases (HK) [15]. Further metabolism of ¹⁸F-FDG is prevented due to lack of the 2-carbon hydroxyl group and the phosphorylated ¹⁸F-FDG is trapped in the cells (Figure 1) [16]. The requirements of glucose are higher due to increased glycolysis in cancer cells compared to normal cells, a phenomenon known as the Warburg effect, and high expression of glucose transporters as well as hexokinases are characteristics of many cancers [9,17].

Figure 1.

¹⁸F-FLT and ¹⁸F-FDG uptake mechanism. ¹⁸F-FLT enters the cell through nucleoside transporters by the salvage pathway paralleled with thymidine. ¹⁸F-FDG enters the cells paralleled with glucose. ¹⁸F-FLT - 3’ -deoxy-3’-[¹⁸F] fluorothymidine; ¹⁸F-FLTMP - ¹⁸F-FLT-monophosphate; TK1 - thymidine kinase 1; dNT - 5’(3’)-deoxyribonucleotidase; TMP - thymidine monophosphate; TS - thymidylate synthase; dUMP - deoxyuridine monophosphate; TP - thymidine phosphorylase; ¹⁸F-FDG - 2-deoxy-2-[¹⁸F] fluoro-D-glucose; HK - hexokinase; glucose-6-P - glucose-6-phosphate; G6Pase - glucose-6-phosphatase.

¹⁸F-FLT-PET

¹⁸F-FLT, a thymidine analogue labeled with an ¹⁸F isotope, was introduced in 1998 by Shields et al. [18]. ¹⁸F-FLT PET is used to study cell proliferation in vivo [18,19]. ¹⁸F-FLT is incorporated into cells by the pyrimidine salvage pathway paralleled with thymidine. After phosphorylation by thymidine kinase 1 (TK1) ¹⁸F-FLT is trapped intracellular; however, the phosphorylated ¹⁸F-FLT is not incorporated into DNA (Figure 1) [20]. TK1 is mainly expressed during the S-phase of cell cycle [21,22]. ¹⁸F-FLT uptake has shown to be positively correlated with cell growth and TK1 activity [21,23] and several studies have shown a positive correlation between ¹⁸F-FLT uptake and tumor cell proliferation measured by Ki67 protein expression [10,24-33]. The tracer uptake into cells is mediated by equilibrative nucleoside transporters (ENT) 1 and 2 and concentrative nucleoside transporters (CNT) 1 and 3 [34-36]. ¹⁸F-FLT uptake gives consequently a measure of the uptake and incorporation of thymidine into DNA and therefore the tracer uptake does not give a direct measure of cell proliferation but is a surrogate marker of the proliferative status of cells. The ratio of the salvage pathway versus the de novo synthesis of thymidine to fulfill the cancer cells demand for thymidine will determine baseline ¹⁸F-FLT uptake in a tumor. In cancer cells mainly relying on de novo synthesis of thymidine ¹⁸F-FLT uptake determined by PET will therefore not necessarily reflect the proliferative activity.

Response monitoring of targeted therapy

Many targeted therapies induce clinical responses; however, only in a subset of patients does the targeted therapy lead to tumor stasis or regression, increase in overall or progression free survival. The patients do not necessarily respond to the therapy even though the tumor expresses the target. Signaling pathways and cross-talks with other pathways can disturb identification of the ‘correct’ target and thereby how to predict the treatment outcome in an individual patient [37]. There is therefore clinical interest in understanding, which parameters are predictive for a positive treatment outcome and consequently if changes in ¹⁸F-FLT and/or ¹⁸F-FDG uptake measured by PET after initiation of a cancer treatment will be predictive for patient outcome.

Tyrosine kinase inhibitors

Various pre-clinical studies have analyzed ¹⁸F-FDG and/or ¹⁸F-FLT PET uptake following inhibition of different classes of tyrosine kinases (Tables 1, 2). Both treatment with small molecule inhibitors and monoclonal antibodies have been studied. Compounds inhibiting members of the human epidermal growth factor receptor (HER/ErbB) have gained most interest where especially studies with drugs targeting the human epidermal growth factor receptor 1 (EGFR) have been conducted.

Table 1.

¹⁸F-FDG PET of tyrosine kinase inhibitor therapy

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation)	Results	Same paper comparison with 18F-FLT
EGFR	erlotinib	[38]	sc	PC9, HCC827 and H1975 human lung adenocarcinoma	daily	2 and 4	→	+
	erlotonib	[41]	sc	CAL33 and CAL166 human head and neck carcinoma	one dose	1 and 3	↓↓	-
	gefitinib	[37]	sc	H3255, HCC4006, A549 and H1975 NSCLC and A431 human epithelial carcinoma	two doses	2	↓↓	-
HER2	trastuzumab	[46]	sc	syngenic MMTV/HER2 mammary, BT474 human breast cancer	twice weekly/3 weeks	1 weekly	→	+
	trastuzumab	[47]	sc	MDA-MB-361 and MDA-MB-231 human breast cancer	once weekly/3 weeks	2, 9 and 16	↓ (day 16)	-
Pan-HER	afatinib	[49]	sc	N87 human gastric cancer	daily/21 days	7, 14 and 21	→	-
	canertinib (CI-1033)	[48]	sc	A431 human squamous cell carcinoma	daily	3 and 7	↓↓	+
	PKI-166	[30]	sc	A431 human squamous cell carcinoma	daily	7, 14 and 21	↓↓	+
c-KIT	imatinib	[50]	sc	FDC-P1 murine hemopoietic cell line	twice daily	4 h and day 1 and 2	↓↓	-
	imatinib	[51]	sc	GIST882 gastrointestinal stromal cell line	twice daily	1 and 8	↓↓ (day 1)	-
c-MET	rilotumumab	[54]	sc	U87MG human glioblastoma	twice weekly/1 week	1, 2, 3, 4, and 7	↓↓	+
	CE-355621	[57]	sc	U87MG human glioblastoma	one dose	1, 3, 7 and 9	↓	-
	crizotinib	[55]	sc	U87MG human glioblastoma and GTL-16 human gastric cancer	daily	2, 5, 7 and 13	↓↓ (day 13 GTL-16)	+
							→ (U87MG)
	BAY 853474	[56]	sc	Hs746T human gastric cancer	twice daily	2 and 4	↓	+

sc: subcutaneous; NSCLC: non-small cell lung cancer; →: no change in ¹⁸F-FDG uptake; ↓↓: decrease in ¹⁸F-FDG uptake compared with baseline; ↓: decreases in ¹⁸F-FDG uptake compared with a control group; ↑↑: increases in ¹⁸F-FDG uptake compared with baseline; ↑: increases in ¹⁸F-FDG uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FDG results from the tumor sensitive cell lines were included in the result column.

Table 2.

¹⁸F-FLT PET of tyrosine kinase inhibitor therapy

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation	Results	Same paper comparison with 18F-FDG
EGFR	erlotinib	[38]	sc	PC9, HCC827 and H1975 human lung adenocarcinoma	daily	2 and 4	↓↓	+
	erlotinib	[39]	sc	A431 human squamous cell carcinoma	daily/4 days	3	↓↓	-
	erlotinib	[40]	sc	HCC827, H1975 and H1650 human NSCLC	daily	3	↓↓	-
	cetuximab	[39]	sc	SCC-1 human squamous cell carcinoma	every 3 days	6	↓↓	-
	cetuximab	[44]	sc	H1975 human NSCLC	one dose	3	↓↓	-
	cetuximab	[45]	sc	DiFi and HCT-116 human colorectal carcinoma	3 doses/week	7	→	-
	CL-387, 785	[40]	sc	H1975 human NSCLC	daily	3	↓↓	-
	WZ4002	[40]	sc	H1975 human NSCLC	daily	3	↓↓	-
HER2	trastuzumab	[46]	sc	syngenic MMTV/HER2 mammary cancer, BT474 human breast cancer	twice weekly/3 weeks	1 weekly	↓↓	+
Pan-HER	CI-1033	[48]	sc	A431 human squamous cell carcinoma	daily	3 and 6	↓↓	+
	PKI-166	[30]	sc	A431 human squamous cell carcinoma	daily	6 h and day 1, 2, 7, 14 and 21	→ (6 h, day 1)	+
							↓↓ (day 2, 7, 14 and 21)
c-MET	rilotumumab	[54]	sc	U87MG human glioblastoma	twice weekly/1 week	1, 2, 3, 4 and 7	↓↓ (from day 4)	+
	crizotinib	[55]	sc	U87MG human glioblastoma and GTL-16 human gastric cancer	daily	2, 4/5 and 7/8	↓↓ (day 4 and 7 GTL-16 and day 8 U87MG)	+
	BAY 853474	[56]	sc	Hs746T human gastric cancer	twice daily	2 and 4	↓	+

sc: subcutaneous; NSCLC: non-small cell lung cancer; →: no change in ¹⁸F-FLT uptake; ↓↓: decrease in ¹⁸F-FLT uptake compared with baseline; ↓: decreases in ¹⁸F-FLT uptake compared with a control group; ↑↑: increases in ¹⁸F-FLT uptake compared with baseline; ↑: increases in ¹⁸F-FLT uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FLT results from the tumor sensitive cell lines were included in the result column.

EGFR

Decrease in ¹⁸F-FLT uptake has been observed as early as day 2-3 after initiation of treatment with the small molecule EFGR inhibitor erlotinib (Table 2) [38-40]. Ullrich et al. compared ¹⁸F-FLT uptake with ¹⁸F-FDG uptake and ¹⁸F-FDG uptake was observed to be unchanged following treatment with erlotinib [38]. The suggestion was that ¹⁸F-FDG more indirectly reflected tumor cell proliferation and therefore a therapy induced reduction in ¹⁸F-FDG uptake was likely to be a later event; however, analysis of time points beyond day 4 was not covered in the study. In contrast to this, other studies found decreases in ¹⁸F-FDG uptake 1 or 2 days after treatment initiation with the small molecule EFGR inhibitors erlotinib and gefitinib, respectively [37,41]. Targeting of the mitogen-activated protein kinase (MAPK) signaling pathway by mitogen-activated protein kinase kinase (MEK)/Raf inhibitors, downstream of EGFR, also showed decrease in ¹⁸F-FDG uptake [42,43] (Table 7). Decrease in ¹⁸F-FDG uptake after EGFR inhibition has been associated with translocation of GLUTs from the plasma membrane to the cytosol in some studies [37,43] whereas another study observed unchanged GLUT-1 expression despite decreased ¹⁸F-FDG uptake [41].

Table 7.

¹⁸F-FDG PET of other targeted therapies

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation)	Results	Same paper comparison with 18F-FLT
MAPK pathway
MEK1/2	PD0325901	[42]	sc	SKMEL-28 human melanoma, BT-474 human breast cancer	5 days weekly/3 weeks	week 1, 2, and 3	↓	+
MEK/Raf	RO5126766	[43]	sc	HCT116, COLO205, COLO320DM human colon cacinoma	daily	1, 2 and 3	↓↓	-
BRAF	PLX4720	[87]	sc	Lim2405 and HT29 human colorectal carcinoma	daily	3	→	+
Metabolism
NAMPRT	daporinad (APO866)	[89]	sc	A2780 human ovarian carcinoma	twice daily	1, 2 and 7	↓↓ (day 7)	+
AMPK	metformin	[88]	sc	HT29 human colorectal carcinoma	one dose	1	↑↑	+
Amino acid metabolism	Top216	[90]	sc	A2780 human ovarian carcinoma	day 0 and 2	6 hours and day 1 and 6	↓↓	+
Aurora kinases
Aurora B kinase	barasertib (AZD1152)	[93]	sc	HCT116 and SW620 human colorectal carcinoma	2 consecutive days/week/3 weeks	7, 14, 21, 26, 36, 43	→	+
Aurora B kinase	TAK-901	[94]	sc	HCT116 human colorectal carcinoma	twice daily for 2 days/week/2 weeks	4, 8, 11 and 15	→	+
HSP90	luminespib (AUY922)	[59]	sc	SUDHL-1 human lymphoma	daily	5	→	+
	tanespimycin (17AAG)	[96]	sc	BT474 human breast carcinoma	3 doses/one day	1, 8, 15, 22	→	-
Topoisomerase I	irinotecan	[97]	sc	HCT116 human colorectal carcinoma	weekly/3 weeks	1, 5, 8, and 15	↑ (day 8 and 15)	+
HDAC	belinostat	[100]	sc	A2780 human ovarian carcinoma	day 0-4 and 6-10	3, 6, and 10	↓ (day 10)	+
EMMPRIN	Anti-EMMPRIN	[113]	orthotopic	MIA PaCa-2 human pancreas carcinoma	day 0, 2, 7 and 10	7 and 14	↓↓ (day 14)	-
Proteasome	bortezomib	[118]	sc	CWR22 human prostate carcinoma	day 0, 2, 7, 10 and 14	1, 4, 8, 15 and 18	↓↓ (day 8)	-

sc: subcutaneous; →: no change in ¹⁸F-FDG uptake; ↓↓: decrease in ¹⁸F-FDG uptake compared with baseline; ↓: decreases in ¹⁸F-FDG uptake compared with a control group; ↑↑: increases in ¹⁸F-FDG uptake compared with baseline; ↑: increases in ¹⁸F-FDG uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FDG results from the tumor sensitive cell lines were included in the result column.

¹⁸F-FLT uptake has been analyzed following treatment with the EGFR targeting monoclonal antibody cetuximab. ¹⁸F-FLT uptake was found to decrease day 3 and day 6 after start of treatment with cetuximab in a human squamous cell carcinoma and human non-small cell lung cancer (NSCLC) tumor model, respectively [39,44].Another study observed no change in ¹⁸F-FLT uptake despite treatment with cetuximab in a cetuximab-sensitive human colorectal carcinoma tumor model for a period of 7 days [45]. None of the studies compared ¹⁸F-FLT uptake with ¹⁸F-FDG.

HER2

Two studies analyzed uptake of ¹⁸F-FDG after inhibition of the human epidermal growth factor receptor 2 (HER2) pathway (Table 1). Treatment with trastuzumab, a monoclonal antibody targeting HER2, did not change the uptake of ¹⁸F-FDG during a three week treatment course in one study [46]. Contrary to this, another study found differences between a treatment and a control group after trastuzumab treatment, but not until 16 days after treatment initiation [47]. In a study by Shah et al. ¹⁸F-FDG uptake was compared with ¹⁸F-FLT uptake where ¹⁸F-FLT uptake was decreased following one week of treatment with trastuzumab [46].

Pan-HER

Drugs targeting several members of the HER/ErbB family simultaneously have also been tested for their ability to change uptake of ¹⁸F-FDG and ¹⁸F-FLT early after treatment initiation (Tables 1, 2). Both ¹⁸F-FLT and ¹⁸F-FDG uptake were decreased 3 and 6 days after start of daily treatments with canertinib (CI-1033), a pan-HER inhibitor targeting all four members of the HER family [48]. Treatment with PKI-166 targeting both EGFR and HER2 resulted in decreases in ¹⁸F-FDG uptake from day 7 and decreases in ¹⁸F-FLT uptake from day 2 [30]. No change in ¹⁸F-FDG uptake was observed after treatment with afatinib a selective inhibitor of EGFR, HER2 and HER4 and no comparison was made with ¹⁸F-FLT [49].

c-KIT

Treatment with the c-KIT inhibitor imatinib resulted in dramatic and early decreases in ¹⁸F-FDG uptake in tumor models displaying mutations in c-KIT, which is often observed in gastrointestinal stromal tumors (GIST) (Table 1) [50,51]. Both number and activity of glucose transporters on the cell surface were decreased after imatinib treatment which were comparable with, and probably the cause of, the decrease in ¹⁸F-FDG uptake [50]. Changes in ¹⁸F-FDG uptake early after initiation of treatment with imatinib in GIST patients is a well-known predictor of patient outcome [52,53].

c-MET

Inhibition of c-MET activation by the monoclonal antibody rilotumumab, as inhibits the binding of hepatocyte growth factor (HGF) to the c-MET receptor, induced decreases in both ¹⁸F-FDG and ¹⁸F-FLT uptake in the U87MG human glioblastoma model. ¹⁸F-FDG and ¹⁸F-FLT uptake were decreased day 2 and 4 after treatment start, respectively [54]. Furthermore, a dose-response relationship was evaluated with increasing doses of rilotumumab in U87MG tumor-bearing mice with ¹⁸F-FDG PET [54]. Treatment effect was evaluated at baseline and day 7 after treatment with 10, 30, 100, 300 or 500 µg rilotumumab. Doses of 300 and 500 µg were similarly effective at reducing tumor growth and this was further reflected in a comparable inhibition of ¹⁸F-FDG uptake [54]. In contrast to the early decrease in both ¹⁸F-FDG and ¹⁸F-FLT uptake after treatment with rilotumumab, inhibition of c-MET by the small molecule inhibitor crizotinib induced no change in ¹⁸F-FDG uptake and ¹⁸F-FLT uptake was not decreased until day 8 after treatment initiation with crizotinibin the U87MG tumor model [55]. Inhibition of c-MET with crizotinib in a GTL-16 human gastric cancer model caused ¹⁸F-FLT decrease day 4 after treatment start whereas ¹⁸F-FDG uptake was unchanged until day 13 [55]. Inhibition of c-MET with another small-molecule inhibitor, BAY 853474, induced reductions in both ¹⁸F-FDG and ¹⁸F-FLT uptake already from day 2 [56]. The monoclonal antibody against c-MET, CE-355621, inhibits ¹⁸F-FDG accumulation in the U87MG human glioblastoma model [57]. Inhibition of ¹⁸F-FDG accumulation following injection of CE-355621 was not compared with ¹⁸F-FLT uptake.

In conclusion, the changes in ¹⁸F-FDG and ¹⁸F-FLT uptake after c-MET inhibition were most variable (Tables 1, 2).

mTOR inhibitors

Activation of the phosphatidyl-inositol-3-kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR) cascade signaling pathway regulates anti-proliferative and apoptotic functions and are involved in regulation of cell metabolism. Inhibition of mTOR and also other targets in this pathway will, at least theoretically, affect the cellular metabolism, expression of hexokinases and GLUT transporters and thereby the ¹⁸F-FDG uptake in cancer cells [9,41,58]. Much interest has therefore been on analyzing ¹⁸F-FDG uptake following inhibition of the PI3K/AKT/mTOR pathway and many pre-clinical studies have been conducted.

Everolimus, rapamycin and temsirolimus, small molecule inhibitors of mTOR, induced early reductions in both ¹⁸F-FLT and ¹⁸F-FDG from day 1 or 2 after initiation of treatment (Tables 3, 4) [59-67].

Table 3.

¹⁸F-FDG PET of PI3K/AKT/mTOR inhibitor therapy

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation)	Results	Same paper comparison with 18F-FLT
mTOR	everolimus	[59]	sc	SUDHL-1 and Karpas299 human lymphoma	daily/14 days	5	→	+
	everolimus	[67]	sc	NCI-N87 human gastric cancer	daily	1, 2, 8 and 15	↓↓	-
	everolimus	[66]	Lymph node metastasis (B16/BL6)sc (H596, HCT116, KB13)	B16/BL6 murine melanoma, H596 human lung carcinoma, HCT116 human colorectal carcinoma and KB13 human cervical cancer	daily	2/3 and 6/7	↓↓ (B16/BL6 and H596)	+
							→ (HCT116 and KB13)
	everolimus	[68]	sc	H727 human carcinoid cancer	daily	1, 3 and 10	→	+
	rapamycin	[65]		U87MG and LN-299 human malignant glioblastoma	one dose	2	↓↓	+
	temsirolimus	[63]	sc	Daudi human B-lymphoblast	NA	2, 4, 7, 9 and 14	↓↓	+
	temsirolimus	[62]	sc	Daudi human B-lymphoblast	NA	2, 4, 7, 9 and 14	↓↓	-
	temsirolimus	[64]	sc	Granta-519 human mantle cell lymphoma	NA	1, 2, 4, 7, 9, 11 and 14	↓↓ (day 1 and 2)	+
	temsirolimus	[60]	sc	ACHN human renal cell adenocarcinoma	twice daily	1	↓↓	-
	AZD8055	[58]	sc	U87MG human glioblastoma	daily	1 h and day 4	↓	+
AKT	AZD5363	[69]	sc	U87MG human glioblastoma, BT474C human breast cancer and Calu-6 human lung cancer	daily	4 h, day 4 (U87MG) day 3 (all models)	↓ (4 h) U87MG	-
							↓↓ (day 4) U87MG
							↓ (day 3) U87MG and BT474C

sc: subcutaneous; NA: not available; →: no change in ¹⁸F-FDG uptake; ↓↓: decrease in ¹⁸F-FDG uptake compared with baseline; ↓: decreases in ¹⁸F-FDG uptake compared with a control group; ↑↑: increases in ¹⁸F-FDG uptake compared with baseline; ↑: increases in ¹⁸F-FDG uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FDG results from the tumor sensitive cell lines were included in the result column.

Table 4.

¹⁸F-FLT PET of PI3K/AKT/mTOR inhibitor therapy

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation	Results	Same paper comparison with 18F-FDG
mTOR	everolimus	[61]	sc	SKOV3 human ovarian adenocarcinoma	daily	2 and 7	↓↓	-
	everolimus	[59]	sc	SUDHL-1 and Karpas299 human lymphoma	daily/14 days	2 (Karpas299) and 5 (SUDHL)	↓↓	+
	everolimus	[66]	sc	H596 human lung carcinoma and HCT116 human colorectal carcinoma	daily	2/3 and 7/10	↓↓ (H596)	+
							→ (HCT116)
	everolimus	[68]	sc	H727 human carcinoid cancer	daily	1, 3 and 10	↓ (day 10)	+
	rapamycin	[65]	sc	U87MG and LN-299 human glioblastoma	one dose	2	↓↓	+
	temsirolimus	[63]	sc	Daudi human B-lymphoblast	NA	2, 4, 7, 9 and 14	↓↓	+
	temsirolimus	[64]	sc	Granta-519 human mantle cell lymphoma	NA	1, 2, 4, 7, 9, 11 and 14	↓↓	+
	AZD8055	[58]	sc	U87MG human malignant glioma	daily	4	↓	+
PI3K/mTOR	BEZ235	[70]	sc	N87, MKN28 and MKN45 human gastric cancer	daily	2	↓	-
PI3-kinase	pictilisib (GDC-0941)	[71]	sc and orthotopic	HCT116 human colorectal carcinoma and U87 human glioma	twice daily/8 days	18 and 186 hours	↓↓ (18 h)	-

sc: subcutaneous; NA: not available; →: no change in ¹⁸F-FLT uptake; ↓↓: decrease in ¹⁸F-FLT uptake compared with baseline; ↓: decreases in ¹⁸F-FLT uptake compared with a control group; ↑↑: increases in ¹⁸F-FLT uptake compared with baseline; ↑: increases in ¹⁸F-FLT uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FLT results from the tumor sensitive cell lines were included in the result column.

Several studies have compared ¹⁸F-FDG and ¹⁸F-FLT uptake [59,63-66]. Different mouse models of human cancer have been used, but despite the heterogeneous tumor models the early decreases in both ¹⁸F-FDG and ¹⁸F-FLT uptake were comparable. Most of the studies showed decreases in both ¹⁸F-FLT and ¹⁸F-FDG uptake after treatment initiation; however, in two studies ¹⁸F-FDG was unchanged despite effective treatment of everolimus sensitive tumors [59,68]. In one study ¹⁸F-FDG and ¹⁸F-FLT uptake on an individual tumor level day 1 and day 3 after treatment initiation predicted tumor growth despite no difference between the control and treatment group was observed [68]. Honer et al. analyzed the effect on ¹⁸F-FLT and ¹⁸F-FDG uptake in tumors arising from different cell lines characterized in advance as either sensitive or insensitive to everolimus treatment from in vitro assays [66]. When grown as tumor xenografts in nude mice both the growth of sensitive and insensitive tumors was inhibited with everolimus treatment. The growth inhibition of the insensitive tumors was suggested to be due to anti-angiogenic/vascular effects of everolimus, which was not evident in vitro. Interestingly, in the insensitive tumor models, in which everolimus had an effect on tumor growth, no change in either ¹⁸F-FDG or ¹⁸F-FLT uptake was observed and that led the authors to conclude that ¹⁸F-FLT and ¹⁸F-FDG PET may result in false-negative prediction of the possible anti-angiogenic/vascular effect of everolimus [66]. Inhibition of the mTOR kinase with AZD8055 resulted in decreases in both ¹⁸F-FLT and ¹⁸F-FDG uptake day 4 after treatment initiation. As early as one hour after injection with AZD8055 the ¹⁸F-FDG uptake was reduced [58].

Inhibition of the PI3K/AKT/mTOR pathway by the AKT inhibitor AZD5363 resulted in decreases in ¹⁸F-FDG uptake in two AZD5363-sensitive but not a AZD5363-resistant mouse tumor model 3 days after treatment initiation. Additionally, already 4 hours after one dose the ¹⁸F-FDG uptake was decreased in a sensitive tumor model [69].

The dual PI3K/mTOR inhibitor BEZ235 significantly reduced ¹⁸F-FLT uptake in a treatment-sensitive N87 human gastric cancer xenograft model whereas no change in ¹⁸F-FLT uptake was observed in treatment-resistant tumor models [70]. Furthermore, Cawthorne et al. observed decreases in ¹⁸F-FLT uptake already 16 hours after initiation of therapy with the PI3K inhibitor pictilisib (GDC-0941) in two pictilisib-sensitive tumor models but not a pictilisib-resistant tumor model [71].

Cejka et al. used ¹⁸F-FDG PET to determine the optimal treatment dose of the mTOR inhibitor everolimus. Doses from 0.05 to 15 mg/kg were administered to mice bearing N87 gastric cancer xenografts and ¹⁸F-FDG PET revealed that doses above 5 mg/kg did not reduce ¹⁸F-FDG uptake further [67]. This was reflected in the anti-tumor activity of everolimus that reached a plateau with doses from 5 mg/kg and above.

Haagensen et al. investigated whether ¹⁸F-FLT could be used to analyze the enhanced activity of a combination of a PI3K and a MEK inhibitor.They observed that treatment with the PI3K inhibitor pictilisib or the MEK inhibitor PD0325901 alone did not induce changes in ¹⁸F-FLT uptake day 2 post injection whereas the combination treatment decreased the ¹⁸F-FLT uptake [72].

Anti-angiogenic/vascular therapy

Treatment with anti-angiogenic compounds does only have an effect in a limited amount of a patients and identification of the subgroup of patients who is benefitting from anti-angiogenic therapy is difficult [73,74]. Accordingly, several pre-clinical studies have investigated if either ¹⁸F-FDG or ¹⁸F-FLT PET could be of value in assessing whether or not a patient is responding to anti-angiogenic therapy. One of the treatment effects of anti-angiogenic therapy is normalization of the tumor vasculature [75]. Theoretically, it is therefore difficult to predict the outcome of both ¹⁸F-FDG and ¹⁸F-FLT uptake early after initiation of an anti-angiogenic therapy because normalization of tumor vasculature could cause a transient increase in cell proliferation and glucose consumption resulting from an enhancement of oxygen and nutrient supply to the cancer cells [75].

Following inhibition of tumor angiogenesis by the vascular endothelial growth factor A (VEGF-A) targeting antibody bevacizumab, ¹⁸F-FDG uptake was decreased (Table 5) [76-78]. By dynamic ¹⁸F-FDG PET analyses, treatment with bevacizumab was shown to reduce both the tumor perfusion and metabolism 24 hours post-treatment [77]. Targeting of VEGF-A by the PEGylated Anticalin Angiocal PRS-050-PEG40 did also reduce uptake of ¹⁸F-FDG [76]. No influence on ¹⁸F-FLT uptake was observed following inhibition of VEGF-A with bevacizumab [78].

Table 5.

¹⁸F-FDG PET of angiogenic/vascular inhibitor therapy

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation)	Results	Same paper comparison with 18F-FLT
VEGF-A	bevacizumab	[78]	orthotopic	U87 and U251 human glioblastoma	0, 3, 7 and 10	6	↓	+
	bevacizumab	[77]	orthotopic	MAS98.12 human breast carcinoma	one dose	1 and 3	↓↓ (day 1)	-
	bevacizumab	[76]	sc	A673 human rhabdomyosarcoma	one dose	2	↓↓	-
	PRS-050-PEG40	[76]	sc	A673 human rhabdomyosarcoma	one dose	2	↓↓	-
RAF/VEGFR2	AAL881	[66]	orthotopic	BN472 rat mammary tumors	daily	2 and 7	→	-
Tubulin	ombrabulin (AVE8062)	[79]	ip	HeyA8 human ovarian cancer	one dose	2 and 24 h	↓↓	-
VEGFR/PDGFR	sunitinib	[80]	orthotopic	U87MG human glioblastoma	5 dose/week for 2 weeks	3, 7, 10, 14 and 16	↓ (day 16)	+
Mulitkinase: RAF, VEGFR, PDGF, c-KIT, RET	sorafenib	[82]	sc	A673 human sarcoma	daily	1 and 6	↓↓ (day 6)	+
VEGFR1-3	axitinib	[84]	sc	U87MG human glioblastoma and MDA-MB-231 human breast cancer	daily/10 days	1, 3, 7 and 10	↓ (day 10)	+
VEGFR-2	ZD4190	[85]	sc	MDA-MB-435 human breast cancer	daily/3 days	1, 3 and 7	→	+

sc: subcutaneous; ip: intra peritoneal; →: no change in ¹⁸F-FDG uptake; ↓↓: decrease in ¹⁸F-FDG uptake compared with baseline; ↓: decreases in ¹⁸F-FDG uptake compared with a control group; ↑↑: increases in ¹⁸F-FDG uptake compared with baseline; ↑: increases in ¹⁸F-FDG uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FDG results from the tumor sensitive cell lines were included in the result column.

Honer et al. found no change in ¹⁸F-FDG uptake after treatment with AAL881, a dual RAF/VEGFR2 inhibitor [66]. As described in the previous section, anti-angiogenic/vascular effects of everolimus did not result in ¹⁸F-FDG changes [66]. Targeting of the vasculature with the tubulin-binding agent ombrabulin (AVE8062) did very early after treatment initiation (2 and 24 hours) induce decrease in ¹⁸F-FDG uptake [79].

Treatment with the compounds sunitinib [80,81], sorafenib [82,83], axitinib [84] and ZD4190 [85], all targeting one or several of the VEGF receptors, induced decreases in ¹⁸F-FLT uptake (Table 6). Sunitinib, sorafenib and axitinib did also decrease ¹⁸F-FDG uptake; however, the reductions in ¹⁸F-FDG uptake was a later event compared with the ¹⁸F-FLT uptake (Table 5). Treatment with the VEGFR-2 targeting compound ZD4190 did not change the tumor ¹⁸F-FDG uptake [85].

Table 6.

¹⁸F-FLT PET of angiogenic/vascular inhibitor therapy

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation	Results	Same paper comparison with 18F-FDG
VEGF-A	bevacizumab	[78]	orthotopic	U87MG and U251MG human glioblastoma	0, 3, 7 and 10	6	→	+
VEGFR/PDGFR	sunitinib	[80]	orthotopic	U87MG human glioblastoma	5/2 schedule	3, 7, 10, 14 and 16	↓ (from day 7)	+
	sunitinib	[81]	sc	U87MG human glioblastoma	daily/7 days	1, 3, 7 and 13	↓↓ (day 3 and 7)	-
Mulitkinase: RAF, VEGFR, PDGF, c-KIT, RET	sorafenib	[82]	sc	A673 human sarcoma	daily	1 and 5	↓↓	+
	sorafenib	[83]	im	FSall mouse fibrosarcoma	day 0 and 1	2 and 3	↓↓	-
VEGFR1-3	axitinib	[84]	sc	U87MG human glioblastoma and MDA-MB-231 human breast cancer	daily/10 days	1, 3, 7 and 10	↓ (day 3) U87-MG	+
							↓↓ (day 7) MDA-MB-231
VEGFR-2	ZD4190	[85]	sc	MDA-MB-435 human breast cancer	daily/3 days	1, 3 and 7	↓↓ (day 1 and 3)	+

sc: subcutaneous; im: intra muscular; →: no change in ¹⁸F-FLT uptake; ↓↓: decrease in ¹⁸F-FLT uptake compared with baseline; ↓: decreases in ¹⁸F-FLT uptake compared with a control group; ↑↑: increases in ¹⁸F-FLT uptake compared with baseline; ↑: increases in ¹⁸F-FLT uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FLT results from the tumor sensitive cell lines were included in the result column.

MAPK cascade

A few studies have analyzed how inhibitors of the MAPK signaling pathway influence the ¹⁸F-FDG and ¹⁸F-FLT uptake. Inhibition of MEK induced decreases in both ¹⁸F-FDG and ¹⁸F-FLT uptake (Tables 7, 8) [42,43,86]. Inhibition of BRAF with the small molecule inhibitor PLX4720 induced decreases in ¹⁸F-FLT uptake day 4 after treatment initiation in a mouse model of human colorectal cancer whereas no change in ¹⁸F-FDG uptake was observed at day 3 after therapy initiation [87].

Table 8.

¹⁸F-FLT PET of other targeted therapies

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation)	Results	Same paper comparison with 18F-FDG
MAPK pathway
MEK1/2	PD0325901	[42]	sc	SKMEL-28 human melanoma, BT-474 human breast cancer	5 days weekly/3 weeks	week 1, 2, and 3	↓↓	+
MEK1/2	PD0325901	[86]	sc	SKMEL-28 human melanoma and HCT116 human colorectal carcinoma	daily	1 and 10	↓	-
BRAF	PLX4720	[87]	sc	Lim2405 and HT29 human colorectal carcinoma	daily	4	↓↓	+
Metabolism
NAMPRT	daporinad (APO866)	[89]	sc	A2780 human ovarian carcinoma	twice daily	1, 2 and 7	↓↓	+
AMPK	metformin	[88]	sc	HT29 human colorectal carcinoma	one dose	1	↓↓	+
Amino acid metabolism	arginine deiminase	[92]	sc	SKMEL-28 human melanoma	weekly/4 weeks	Once weekly prior to treatment	→	-
Amino acid metabolism	Top216	[119]	sc	A2780 human ovarian carcinoma	day 0 and 2	2 and 6 hours and day 1 and 6	↓↓	+
Amino acid metabolism	TP202377	[91]	sc	A2780 human ovarian carcinoma	one dose	6 hours, day 1 and 6	↓↓	-
Aurora kinases
Aurora B kinase	barasertib (AZD1152)	[93]	sc	HCT116 and SW620 human colorectal carcinoma	2 consecutive days/week/3 weeks	8, 15, 22, 29, 37	↓ (8, 15, 22)	+
Aurora B kinase	TAK-901	[94]	sc	HCT116 human colorectal carcinoma	twice daily for 2 days/week/2 weeks	4, 9, 11 and 15	↓↓ (day 4 and 11)	+
Aurora A/B kinase	CCT129202	[95]	sc	HCT116 human colorectal carcinoma	daily	2 or 7	↓ (day 7)	-
HSP90	luminespib (AUY922)	[59]	sc	SUDHL-1 human lymphoma	daily	5	↓↓	+
Topoisomerase I
Topoisomerase I	irinotecan	[97]	sc	HCT116 human colorectal carcinoma	weekly/3 weeks	1, 5, 8, and 15	↓↓ (1, 8, 15)	+
Topoisomerase I	irinotecan	[98]	sc	HCT116 human colorectal carcinoma	once weekly	8	↓↓	-
HDAC
HDAC	belinostat	[100]	sc	A2780 human ovarian carcinoma	day 0-4 and 6-10	3, 6, and 10	→	+
HDAC	belinostat	[98]	sc	HCT116 human colorectal carcinoma	day 1-5 and 8-12	8	↓	-
HDAC	dacinostat (LAQ824)	[28]	sc	HCT116 human colorectal carcinoma	daily	2, 4 and 10	↓ (4 and 10)	-
HDAC	vorinostat (SAHA)	[99]	sc	HepG2 human hepatoma	5 days a week/3 weeks	8	↓↓	-
HDAC	ISAHA	[99]	sc	HepG2 human hepatoma	5 days a week/3 weeks	8	↓↓	-
FGFR	PD173074	[120]	sc	H-69 human SCLC	daily	7, 14	↓ (7 and 14)	-

sc: subcutaneous; SCLC: small cell lung cancer; →: no change in ¹⁸F-FLT uptake; ↓↓: decrease in ¹⁸F-FLT uptake compared with baseline; ↓: decreases in ¹⁸F-FLT uptake compared with a control group; ↑↑: increases in ¹⁸F-FLT uptake compared with baseline; ↑: increases in ¹⁸F-FLT uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FLT results from the tumor sensitive cell lines were included in the result column.

Metabolism

The anti-diabetes drug metformin, which modulate the cellular metabolism through AMP-activated protein kinase (AMPK) activation, is currently in several clinical trials in combination with different chemotherapeutic. Treatment of a mouse model of human colorectal carcinoma with metformin had a divergent effect on the uptake of ¹⁸F-FDG and ¹⁸F-FLT. ¹⁸F-FDG uptake was found to be increased day 1 after initiation of therapy whereas the ¹⁸F-FLT uptake was decreased [88]. The increase in ¹⁸F-FDG uptake observed after treatment with metformin is probably because of externalization of GLUT transporters due to AMPK activation. Thus the positive effect of metformin on tumor growth would be overlooked if decreases in ¹⁸F-FDG uptake are used as a surrogate marker of effective treatment. Difficulties with interpretation of ¹⁸F-FDG uptake after initiation of cancer treatment could also arise in situations where glucose modulators are used either incidentally to treat other illness or as part of a combination treatment regime [88]. Treatment with another modulator of the cellular metabolism, the nicotinamide phosphoribosyltransferase (NAMPT) inhibitor daporinad (APO866) also decreased the ¹⁸F-FLT uptake. Following treatment with daporinad ¹⁸F-FDG uptake was also decreased, but later compared to ¹⁸F-FLT uptake [89].

Targeting amino acid metabolism with Top216/TP202377 decreased both ¹⁸F-FDG and ¹⁸F-FLT uptake (Figure 2) [90,91], whereas treatment with the arginine deiminase did not induce a response in the ¹⁸F-FLT uptake [92]. Treatment with both Top216 and TP202377 induced decrease in ¹⁸F-FLT uptake as early as 2 and 6 hours post injection where it was possible to separate responding from non-responding tumors (Figure 3) [90,91].

Figure 2.

Examples of ¹⁸F-FDG PET/CT images. Representative ¹⁸F-FDG PET/CT images of a treatment mouse (top panel) and a control mouse (lower panel) scanned with ¹⁸F-FDG at baseline and 6 hours, day 1 and day 5 after injections with one dose of Top216 or vehicle. Both mice carry the A2780 human ovarian carcinoma xenograft. The arrows point towards the tumors. The image is reproduced from [90].

Figure 3.

¹⁸F-FLT PET/CT images of treatment-sensitive and treatment-resistant tumors. Representative ¹⁸F-FLT PET/CT images of a TP202377-sensitive tumor xenograft (A2780, upper panel) and two TP202377-resistant tumor xenografts (A2780/Top216 and SW620, middle and lower panel). ¹⁸F-FLT uptake is measured by PET in the same mice at baseline and 6 hours, day 1 and day 6 after one injection with TP202377. The dotted circles delineate the tumors. The image is reproduced from [91].

Aurora kinases

Inhibition of mitosis through targeting of the aurora kinases, a family of serine kinases that play a role in the regulation of mitosis, has been investigated with both ¹⁸F-FDG and ¹⁸F-FLT PET. Treatment with the Aurora B kinase inhibitors barasertib (AZD1152) and TAK-901 induced reductions in ¹⁸F-FLT uptake (Table 8) [93,94]. For comparison, no change was observed in ¹⁸F-FDG uptake following treatment initiation with either barasertib or TAK-901 (Table 7). Likewise, did the dual Aurora A/B kinase inhibitor CCT129202 induce decreases in ¹⁸F-FLT uptake but no comparison was made with ¹⁸F-FDG [95].

HSP90

The heat shock protein 90 (HSP90) inhibitor luminespib (AUY922) inhibited uptake of ¹⁸F-FLT in a SUDHL-1 lymphoma model whereas no change was observed for ¹⁸F-FDG uptake [59]. Furthermore, did the HSP90 inhibitor tanespimycin (17AAG) not change ¹⁸F-FDG uptake in a BT474 breast cancer model after initiation of effective treatment [96].

Topoisomerase I

Inhibition of topoisomerase I by irinotecan decreased the ¹⁸F-FLT uptake already from day 1 after initiation of treatment (Table 8) [97,98]. This was in contrast to the ¹⁸F-FDG uptake which were shown to be increased from day 8 following treatment initiation with irinotecan [97]. Inhibition of DNA replication by targeting topoisomerase I with irinotecan did accordingly have the opposite effect on the ¹⁸F-FLT and ¹⁸F-FDG uptake.

Histone deacetylase

After treatment initiation with the histone deacetylase (HDAC) inhibitors belinostat, dacinostat (LAQ824) and vorinostat (SAHA)/ISAHA the uptake of ¹⁸F-FLT was shown to be decreased in all but one case (Table 8) [28,98,99]. In one study, the ¹⁸F-FLT uptake was compared with ¹⁸F-FDG after initiation of treatment with belinostat and it was observed that the ¹⁸F-FDG uptake was decreased whereas no change in ¹⁸F-FLT uptake was found after treatment with the compound for a period of 10 days [100].

Response monitoring of chemotherapeutics

Several studies have analyzed ¹⁸F-FDG and ¹⁸F-FLT uptake with PET in pre-clinical tumor models after treatment with compounds from different classes of chemotherapeutics. Many chemotherapeutics directly induces cell-cycle arrest and therefore uptake of the cell proliferation tracer ¹⁸F-FLT has been widely studied.

Microtubule targeting agents

The influence on ¹⁸F-FDG uptake following treatment with compounds disturbing the microtubules has shown to be quite variable. Following treatment with docetaxel, the ¹⁸F-FDG uptake was found to be unchanged despite effective treatment [101]. Treatment with nanoparticle albumin-bound paclitaxel induced increases in ¹⁸F-FDG uptake day 7 which were associated with an inflammatory reaction in the tumor tissue [102]. Oppositely, treatment with one dose of the microtubule stabilizer patupilone induced reductions in ¹⁸F-FDG uptake day 2 and 6 post injection (Table 9) [66].

Table 9.

¹⁸F-FDG PET of chemotherapeutics

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation)	Results	Same paper comparison with 18F-FLT
Microtubule	docetaxel	[101]	sc	22Rv1 human prostate carcinoma	weekly/2 weeks	2 weeks	→	+
	albumine-bound paclitaxel	[102]	orthotopic	MDA-MB-435 human melanoma	every other day/3 doses total	3, 7, 14 and 21	↑ (day 7)	-
	patupilone	[66]	orthotopic	BN472 rat mammary tumors	one dose	2 and 6	↓↓	-
Alkylating agents	cyclophosphamide	[64]	sc	Granta-519 human mantle cell lymphoma	one dose	1, 2, 4, 7, 9, 11 and 14	↓↓ (day 2 and 11)	+
	cyclophosphamide	[63]	sc	Daudi human B-lymphoblast	one dose	2, 4, 7, 9 and 24	↓↓	-
	cyclophosphamide	[62]	sc	Daudi human B-lymphoblast	NA	2, 4, 7, 9 and 14	↓↓ (day 2 and 4)	+
	temozolomide	[78]	orthotopic	U87 and U251 human glioblastoma	day 0, 3, 7 and 10	6	↓	+
Platinum analogues	cisplatin	[107]	sc	NCCIT human testicular embryonal carcinoma	one dose	2, 4 and 7	↓↓ (day 7)	-
	cisplatin	[108]	sc	PEO1 and PEO4 human ovarian adenocarcinoma	3 consecutive days	4	↓↓	+
	cisplatin	[27]	sc	RIF-1 murine fibrosarcoma	one dose	1 and 2	↓	+
Antimetabolites	5-fluorouracil	[24]	sc	RIF-1 murine fibrosarcoma	one dose	1 and 2	↓ (day 2)	+
	methotrexate	[112]	sc	MC4-L2 and MC7-L1 murine mammary ductal carcinoma	one dose	1, 7 and 14	→	-
	gemcitabine	[113]	orthotopic	MIA PaCa-2 human pancreas carcinoma	day 0 and 7	7 and 14	↓↓ (day 7 and 14)	-
Anthracyclines	doxorubicin	[117]	sc	SUDHL-4 human large B-cell lymphoma	one dose	2	↑↑	+
	doxorubicin	[112]	sc	MC4-L2 and MC7-L1 murine mammary ductal carcinoma	one dose	1, 7 and 14	↓↓	-
	liposomal doxorubicin	[115]	sc	UM-SCC-22B human head and neck squamous cell carcinoma	day 0 and 2	2 and 4	↓↓ (day 4)	+
	liposomal doxorubicin	[114]	sc	C26 murine colorectal carcinoma	one dose	1	→	+

sc: subcutaneous; NA: not available; →: no change in ¹⁸F-FDG uptake; ↓↓: decrease in ¹⁸F-FDG uptake compared with baseline; ↓: decreases in ¹⁸F-FDG uptake compared with a control group; ↑↑: increases in ¹⁸F-FDG uptake compared with baseline; ↑: increases in ¹⁸F-FDG uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FDG results from the tumor sensitive cell lines were included in the result column.

Treatment with chemotherapeutic compounds targeting the microtubules induced a more consistent response in the ¹⁸F-FLT uptake when compared with the ¹⁸F-FDG uptake. Treatment with docetaxel, JAC106 and patupilone all induced decreases in ¹⁸F-FLT uptake although there was variation in relation to at what time after first injection the decrease was observed (Table 10) [101,103,104]. Ebenhan et al. observed decreases in ¹⁸F-FLT uptake already from day 1 after one dose of patupilone, reductions were observed after 3 days of JAC106 treatment [103] and not until 2 weeks after treatment initiation with docetaxel was reductions in ¹⁸F-FLT uptake observed [101].

Table 10.

¹⁸F-FLT PET of chemotherapeutics

Target	Drug	Reference	Tumor placement	Cell line and origin	Treatments	Scan time (days after treatment initiation	Results	Same paper comparison with 18F-FDG
Microtubule	docetaxel	[101]	sc	22Rv1 human prostate carcinoma	weekly/2 weeks	2 weeks	↓↓	+
	JAC106	[103]	sc	SW620 human colorectal adenocarcinoma and KB-V1 human cervix carcinoma	day 0 and 7	3 and 7/8	↓↓ (day 3)	-
	patupilone	[104]	sc	RIF-1 mouse fibrosarcoma	one dose	1, 2, 3 and 6	↓	-
Alkylating agents	cyclophosphamide	[64]	sc	Granta-519 human mantle cell lymphoma	one dose	1, 2, 4, 7, 9, 11 and 14	↓↓ (day 2, 4, 7, 9, 11 and 14)	+
	cyclophosphamide	[62]	sc	Daudi human B-lymphoblast	NA	2, 4, 7, 9 and 14	↓↓ (day 7, 9 and 14)	+
	temozolomide	[78]	orthotopic	U87 and U251 human glioblastoma	day 0, 3, 7 and 10	6	↓	+
	temozolomide	[106]	sc and orthotopic	Gli36dEGFR-1 and Gli36dEGFR-2 human glioblastoma	daily/7 days	2	↓	-
Platinum analogues	cisplatin	[108]	sc	PEO1 and PEO4 human ovarian adenocarcinoma	3 consecutive days	4	↓↓	+
	cisplatin	[27]	sc	RIF-1 mouse fibrosarcoma	one dose	1 and 2	↓	+
Antimetabolites	5-fluorouracil	[24]	sc	RIF-1 mouse fibrosarcoma	one dose	1 and 2	↓	+
	5-fluorouracil	[109]	sc	RIF-1 mouse fibrosarcoma	one dose	1 hour	↑	-
	5-fluorouracil	[111]	sc	HT29 human colorectal carcinoma	one dose	1	↑↑	-
	5-fluorouracil	[110]	sc	HT29 human colorectal carcinoma	one dose	1	↑	-
Anthracyclines	doxorubicin	[117]	sc	SUDHL-4 human large B-cell lymphoma	one dose	2	↓↓	+
	doxorubicin	[116]	sc	SUDHL-4 human large B-cell lymphoma	one dose	1, 5 and 9	↓↓ (day 1 and 5)	-
	liposomal doxorubicin	[115]	sc	UM-SCC-22B human head and neck squamous cell carcinoma	day 0 and 2	2 and 4	↓↓ (day 4)	+
	liposomal doxorubicin	[114]	sc	C26 murine colorectal carcinoma	one dose	1	↓↓	+

sc: subcutaneous; NA: not available; →: no change in ¹⁸F-FLT uptake; ↓↓: decrease in ¹⁸F-FLT uptake compared with baseline; ↓: decreases in ¹⁸F-FLT uptake compared with a control group; ↑↑: increases in ¹⁸F-FLT uptake compared with baseline; ↑: increases in ¹⁸F-FLT uptake compared with a control group. If both treatment sensitive and treatment resistant tumor models were analyzed, only the ¹⁸F-FLT results from the tumor sensitive cell lines were included in the result column.

Treatment with paclitaxel in combination with carboplatin decreased uptake of both ¹⁸F-FDG and ¹⁸F-FLT in a mouse model of human ovarian cancer [105].

DNA damaging agents

Analysis of both ¹⁸F-FDG and ¹⁸F-FLT uptake by PET has been investigated following initiation of treatment with different DNA damaging compounds. Both alkylating agents and platinum analogues induce DNA damage resulting in cell cycle arrest and apoptosis and have therefore attracted interest with regard to response monitoring with both ¹⁸F-FDG and ¹⁸F-FLT PET.

Alkylating agents

Several studies have analyzed how different alkylating agents affect the uptake of ¹⁸F-FDG and ¹⁸F-FLT in pre-clinical models of human cancer. Treatment with cyclophosphamide induced decreases in ¹⁸F-FDG uptake already day 2 after treatment initiation in different tumor models (Table 9) [62-64]. ¹⁸F-FLT uptake was also shown to decrease early following treatment with cyclophosphamide, however in one study it was not observed until the 7^th day after treatment initiation (Table 10) [62,64].

Treatment with temozolomide caused reductions in both ¹⁸F-FDG and ¹⁸F-FLT uptake day 6 after treatment initiation in an orthotopic model of human glioblastoma [78]. In another study, 2 days after temozolomide therapy initiation, ¹⁸F-FLT uptake was decreased in both a subcutaneous and an intra-cranially implanted glioblastoma model [106]. Furthermore, a positive correlation was observed between changes in ¹⁸F-FLT accumulation day 2 and changes in tumor size later on.

Platinum analogues

Treatment with the platinum analogue cisplatin reduced both ¹⁸F-FDG and ¹⁸F-FLT uptake early after initiation of treatment in several independent studies (Tables 9, 10) [27,107,108].

Antimetabolites

The use of ¹⁸F-FLT PET for measurement of treatment effect with the pyrimidine analogue 5-fluorouracil (5-FU) has attracted a special attention because 5-FU directly affects the thymidylate synthase (TS), a key enzyme involved in the DNA synthesis. Inhibition of TS by 5-FU results in rapid decrease in the cellular thymidine phosphate pool and the cancer cells may respond to this shortage by increasing TK1 and nucleoside transporter activity. How ¹⁸F-FLT uptake will be influenced by 5-FU treatment is difficult to predict, because the cancer treatment with 5-FU can result in a temporary increase in ¹⁸F-FLT retention and different outcomes on the ¹⁸F-FLT uptake has consequently been observed following treatment with 5-FU.

Studies on ¹⁸F-FLT uptake in RIF-1 tumor bearing mice after injection with 5-FU have showed that one hour after injection of 5-FU, TS inhibition was imaged by an increase in the ¹⁸F-FLT uptake probably being due to redistribution of nucleoside transporters to the cell membrane [109]. Day 1 and 2 after 5-FU treatment of the same RIF-1 tumor model the ¹⁸F-FLT uptake was decreased compared with a vehicle treated control [24]. Response monitoring of 5-FU treatment with ¹⁸F-FLT PET of a HT29 human colorectal adenocarcinoma model has been analyzed in two studies (Table 10). Both studies observed increase in ¹⁸F-FLT uptake 24 hours after injection of 5-FU [110,111]. Hong et al analyzed the kinetic parameters of the ¹⁸F-FLT uptake by a two hours dynamic PET imaging 24 hours after injection with 5-FU and showed that the parameters related to ¹⁸F-FLT retention was significantly higher in the treatment compared to the control group whereas the parameters related to dephosphorylation of ¹⁸F-FLT monophosphate was conversely significantly lower in the treatment compared to the control group [110].

One study compared the ¹⁸F-FLT uptake following 5-FU treatment with ¹⁸F-FDG uptake and it was found that ¹⁸F-FDG uptake was decreased 2 days after initiation of treatment (Table 9) compared to FLT uptake which was decreased on both day 1 and 2 following treatment initiation [24].

The antifolate compound methotrexate did not change the ¹⁸F-FDG uptake until 14 days after treatment initiation [112] whereas the nucleoside analogue gemcitabine was shown to decrease uptake of ¹⁸F-FDG day 7 and day 14 after initiation of treatment [113].

Anthracyclines

Treatment with the anthracyclines doxorubicin or liposomal doxorubicin (pegylated liposome-encapsulated doxorubicin) both induced decreases in ¹⁸F-FLT within 1 to 4 days after initiation of treatment (Table 10) [114-117]. Interestingly, changes in ¹⁸F-FDG uptake varied considerably after initiation of doxorubicin or liposomal doxorubicin treatment (Table 9). Two days after injection with one dose of doxorubicin one study observed that ¹⁸F-FDG uptake in the treatment group was significantly increased compared with baseline uptake [117]. In contrast to this up-regulation another study observed significant decrease in ¹⁸F-FDG uptake already day 1 after initiation of doxorubicin treatment [112]. The two studies used different tumor models which could be an explanation on the divergent outcome. Following injection with liposomal doxorubicin one study failed to detect an ¹⁸F-FDG response already day 1 after liposomal doxorubicin injection, which is in line with another study in which FDG uptake was observed to be unchanged until day 4 after treatment initiation [114,115].

Conclusions

A non-invasive method to measure early treatment effect of cancer therapeutics is requested in many settings both during development of new therapies but also during treatment with already approved therapies. A comprehensive amount of pre-clinical studies have investigated the use of ¹⁸F-FDG and ¹⁸F-FLT PET for treatment monitoring. ¹⁸F-FDG and ¹⁸F-FLT PET have in both pre-clinical and clinical studies been evaluated as imaging biomarkers that can predict and assess responses to various types of anti-cancer therapies this being different targeted therapies but also conventional chemotherapeutics. The results from the pre-clinical studies are variable, in some studies early changes in ¹⁸F-FDG and ¹⁸F-FLT uptake predict later tumor regression and in other studies no changes in tracer uptake are observed despite the treatment being effective. Overall both ¹⁸F-FDG and/or ¹⁸F-FLT uptake were decreased following treatment initiation with different inhibitors targeting the HER family; however, some studies observed no change in tracer uptake despite effective treatment. Differences in ¹⁸F-FDG and ¹⁸F-FLT uptake following treatment with the same anti-cancer compound are probably due to variations in the experimental protocols or the use of different tumor models. Both ¹⁸F-FDG and ¹⁸F-FLT were reduced after treatment with inhibitors of the PI3K/AKT/mTOR pathway while after treatment with the chemotherapeutic 5-FU different responses in ¹⁸F-FLT uptake was observed.

The mechanism behind changes in tracer uptake after treatment initiation seems to be very complex and dependent on both tumor type, mode of action of the anti-cancer drug and at what time after treatment initiation the tracer uptake is being studied.

Selection of patients into responders and non-responders based on non-invasive PET scans holds a large potential and our prediction is that ¹⁸F-FDG PET and/or ¹⁸F-FLT PET will increasingly be included in future (adaptive) study designs when new cancer treatments are being developed and tested. The use of PET imaging for biological characteristics during the early pre-clinical animal experiments can improve knowledge of drug candidates and maybe help selecting which imaging biomarkers could be included in subsequent clinical studies [8]. Furthermore we foresee that ¹⁸F-FDG and ¹⁸F-FLT PET will be applied to predict treatment effect in more cancer types and for more treatment regimens than today thereby help implementing the practice of precision medicine. However, the current data clearly underlines that in each specific case, pre-clinical testing of ¹⁸F-FDG and ¹⁸F-FLT should be performed to validate the value of the PET tracers.