Journal of Nuclear Medicine, published on February 8, 2024 as doi:10.2967/jnumed.123.

266601

Somatostatin Receptor Imaging with [18F]FET-bAG-TOCA

PET/CT and [68Ga]Ga-DOTA-Peptide PET/CT in Patients
with Neuroendocrine Tumors: A Prospective, Phase 2
Comparative Study
Suraiya Dubash*1, Tara D. Barwick*1,2, Kasia Kozlowski1, Andrea G. Rockall1, Sairah Khan2, Sameer Khan2,
Siraj Yusuf 3, Angela Lamarca4,5, Juan W. Valle4,5, Richard A. Hubner4,5, Mair
ead G. McNamara4,5, Andrea Frilling1,
Tricia Tan6, Florian Wernig6, Jeannie Todd6, Karim Meeran6, Bhavesh Pratap1, Saleem Azeem7, Michael Huiban7,
Nicholas Keat7, Jingky P. Lozano-Kuehne1,8, Eric O. Aboagye1, and Rohini Sharma1
1
Department of Surgery and Cancer, Imperial College London, London, United Kingdom; 2Department of Imaging, Imperial College
Healthcare NHS Trust, London, United Kingdom; 3Radiology and Nuclear Medicine Department, Royal Marsden NHS Foundation
Trust, London, United Kingdom; 4Division of Cancer Sciences, University of Manchester, Manchester, United Kingdom; 5Department
of Medical Oncology, The Christie NHS Foundation Trust, Manchester, United Kingdom; 6Department of Endocrinology, Imperial
College Healthcare NHS Trust, London, United Kingdom; 7Invicro-London, Imperial College London, London, United Kingdom; and
8
Population Health Sciences Institute, Faculty of Medical Sciences, University of Newcastle, Newcastle, United Kingdom

[68Ga]Ga-DOTA-peptide in visualizing NET and may be considered in

18 routine clinical practice given the longer half-life and availability of the
There is a clinical need for F-labeled somatostatin analogs for the
imaging of neuroendocrine tumors (NET), given the limitations of using cyclotron-produced fluorine radioisotope.
[68Ga]Ga-DOTA-peptides, particularly with regard to widespread Key Words: [18F]FET-bAG-TOCA; [68Ga]Ga-DOTA-peptide; PET;
accessibility. We have shown that [18F]fluoroethyl-triazole-[Tyr3]- neuroendocrine tumors; somatostatin receptor
octreotate ([18F]FET-bAG-TOCA) has favorable dosimetry and biodis-
tribution. As a step toward clinical implementation, we conducted a J Nucl Med 2024; 00:1–7
prospective, noninferiority study of [18F]FET-bAG-TOCA PET/CT DOI: 10.2967/jnumed.123.266601
compared with [68Ga]Ga-DOTA- peptide PET/CT in patients with
NET. Methods: Forty-five patients with histologically confirmed NET,
grades 1 and 2, underwent PET/CT imaging with both [18F]FET-bAG-
TOCA and [68Ga]Ga-peptide performed within a 6-mo window
(median, 77 d; range, 6–180 d). Whole-body PET/CT was conducted
50 min after injection of 165 MBq of [18F]FET-bAG-TOCA. Tracer
N euroendocrine neoplasms (NEN) are a heterogeneous group
of malignancies arising from cells of the diffuse neuroendocrine
uptake was evaluated by comparing SUVmax and tumor-to- system. Accurate diagnosis of primary lesion and staging the extent
background ratios at both lesion and regional levels by 2 unblinded, of disease dictates both management and prognosis, whereby
experienced readers. A randomized, blinded reading of both scans patients with limited disease can undergo radical locoregional ther-
was also then undertaken by 3 experienced readers, and consensus apy, including surgery or ablation with curative intent, whereas sys-
was assessed at a regional level. The ability of both tracers to visualize
temic therapy is reserved for those with metastatic disease given
liver metastases was also assessed. Results: A total of 285 lesions
were detected on both imaging modalities. An additional 13 tumor
with palliative intent (1). Accurate imaging is crucial. As 20%–
deposits were seen in 8 patients on [18F]FET-bAG-TOCA PET/CT, and 50% of patients with NEN will have metastatic disease at presenta-
[68Ga]Ga-DOTA-peptide PET/CT detected an additional 7 lesions in tion (2), there is a need for an imaging methodology that is both
5 patients. Excellent correlation in SUVmax was observed between both sensitive and widely accessible.
tracers (r 5 0.91; P , 0.001). No difference was observed between A unique characteristic of NEN is the expression of somato-
median SUVmax across regions, except in the liver, where the median statin receptors (SSTRs) on the tumor surface (3). The presence of
tumor-to-background ratio of [18F]FET-bAG-TOCA was significantly SSTRs has long been exploited for imaging NEN initially with
lower than that of [68Ga]Ga-DOTA-peptide (2.5 6 1.9 vs. 3.5 6 2.3; planar or SPECT imaging using [111In]In-diethylenetriaminepenta-
P , 0.001). Conclusion: [18F]FET-bAG-TOCA was not inferior to acetic acid-octreotide and, more recently, with PET/CT using
radiolabeled somatostatin analogs (SSAs). PET imaging has
greater sensitivity, enhanced resolution, and better accuracy in
Received Sep. 22, 2023; revision accepted Dec. 19, 2023.
For correspondence or reprints, contact Rohini Sharma (r.sharma@imperial.
detecting NEN compared with SPECT imaging (4,5). The most
ac.uk). commonly used PET tracers used for the visualization of NEN are
*Contributed equally to this work. SSAs labeled with [68Ga]Ga-DOTA-peptides, including [68Ga]Ga-
Published online Feb. 8, 2024.
Immediate Open Access: Creative Commons Attribution 4.0 International
DOTA-0-Tyr3-octreotate ([68Ga]Ga-DOTATATE) and [68Ga]Ga-
License (CC BY) allows users to share and adapt with attribution, excluding DOTA-0-Phe1-Tyr3-octreotide ([68Ga]Ga-DOTATOC). Defining
materials credited to previous publications. License: https://creativecommons. the presence of SSTRs on the tumor surface is also important for
org/licenses/by/4.0/. Details: http://jnm.snmjournals.org/site/misc/permission.
xhtml. therapeutic decision making, whereby patients with SSTR-positive
COPYRIGHT ß 2024 by the Society of Nuclear Medicine and Molecular Imaging. NEN on [68Ga]Ga-DOTA PET may be candidates for [177Lu]Lu-

[18F]FET-bAG-TOCA PET/CT TO DETECT NET  Dubash et al. 1

DOTA0-Tyr3-octreotate ([177Lu]Lu-DOTATATE), a targeted No clinically significant differences in DOTA tracers have been
radiotherapeutic that significantly improves progression-free sur- reported (11,12), and these patients were all included for the pri-
vival in patients with metastatic disease (6). mary analysis. Imaging with [18F]FET-bAG-TOCA was con-
Although [68Ga]Ga-DOTA analogs have good resolution, the ducted after [68Ga]Ga-DOTA-peptide PET in most cases.
availability and scalability of production is limited due to the [18F]FET-bAG-TOCA was synthesized by Invicro-London (8);
necessity of an on-site generator pertaining to the short half-life of the mean dose injected was 157.7 MBq, with a mean uptake
[68Ga]Ga. Furthermore, the [68Ga]Ga-radiometal may accumulate period of 37.4 min (range, 30–51 min). Images were acquired on a
within the uncinate process of the pancreas, leading to a false- Siemens Biograph 6 TruePoint PET/CT scanner (with TrueV;
positive diagnosis (7). Clinically, a [18F]F-radioligand would extended field of view) at 50 min after injection (8). An attenua-
overcome the limited capacity of [68Ga]Ga-DOTA production tion CT scan was obtained from vertex to midthigh, immediately
while exploiting existing worldwide cyclotron manufacturing. We followed by a PET emission study at 4 min per bed position (CT
developed a novel [18F]F-octreotate radioligand, [18F]fluoroethyl- settings: tube potential, 130 kV; exposure, 15 effective mAs; pitch,
triazole-[Tyr3]-octreotate ([18F]FET-bAG-TOCA) (8), to obviate 1.5; slice thickness, 5 mm; rotation time, 0.6 s). Images were recon-
these limitations of [68Ga]Ga-DOTA ligands. Previously, we structed using the ordered-subsets expectation maximization algo-
showed that tumor uptake of [18F]FET-bAG-TOCA was superior to rithm (4 iterations and 8 subsets) with corrections for dead time,
that of [68Ga]Ga-DOTATATE in vivo with good spatial resolution scatter, attenuation, and radioactive decay. All images were viewed
(9). Clinically, [18F]FET-bAG-TOCA has favorable dosimetry and on a dedicated PET workstation (Hermes Medical Solutions).
biodistribution (8). We therefore performed a prospective study, the Image Interpretation. Images were reviewed by 2 observers: a
primary objective of which was to assess uptake of [18F]FET-bAG- radiation oncologist with greater than 15 y of experience in imaging
TOCA both at lesion and regional levels. Evaluation of interreader and tumor outlining and an experienced radiologist (with dual
agreement between [18F]FET-bAG-TOCA and [68Ga]Ga-DOTA- accreditation in radiology and nuclear medicine) with greater than
peptide PET was assessed as a secondary endpoint. 20 y of experience. To ensure a methodical and consistent approach,
comparison of [18F]FET-bAG-TOCA with [68Ga]Ga-DOTA-peptide
MATERIALS AND METHODS PET/CT on a patient-by-patient and lesion-by-lesion analysis was
Study Design and Participants performed. Due to the large number of metastases, lesions were ana-
A prospective, multicenter, open-label, single-arm comparative lyzed within the context of anatomic regions. Seven regions were
imaging study consisting of an initial safety run phase (part A) fol- defined as being the most common sites for both primary tumors and
lowed by a noninferiority phase (part B) was conducted. The metastases: head and neck, lung, liver, pancreas, abdomen/pelvis,
safety and biodistribution study (part A) has been reported (8). bone and lymph nodes. Any organ with greater than 5 lesions
Patients from part A (n 5 9) were included in the noninferiority were truncated at 5 target lesions as in previous studies (13,14).
analysis. Key eligibility criteria include histologically confirmed SUV measurements (SUVmax, SUVmean, and tumor-to-background
diagnosis of locally advanced or metastatic grade 1 or 2 neuroen- ratio [TBR]) were obtained for lesion-by-lesion analysis by manually
docrine tumors (NET), measurable disease with at least 1 lesion outlining whole tumor volumes on side-by-side analysis of both studies
with longest diameter $ 10 mm on conventional imaging, and to ensure, in cases with innumerable lesions, that the same lesions
positive SSTR imaging within 6 mo of study enrollment with were selected for comparative quantitative analysis. For comparative
[68Ga]Ga-DOTA-peptide PET. Patients were not required to stop SUV analysis, only those lesions that were visible on both [18F]FET-
SSAs before either PET scan. Patients were recruited from 2 U.K. bAG-TOCA and [68Ga]Ga-DOTA-peptide PET/CT were included
European Neuroendocrine Tumor Society Centers of Excellence, reference (normal background) tissue were outlined using a spheric ref-
Imperial College Health Care NHS Trust and Christie NHS Foun- erence volume of interest (3 cm3 for background liver; 2 cm3 for
dation Trust, Manchester. All diagnostic tissue samples underwent spleen, bone, and mediastinal blood pool); 1 cm3 for spheric volumes
central pathology review to assess eligibility. The study was in the pancreas and sum of 3 slices manually drawn around each adre-
approved by the Leeds East, Yorkshire and Humber National nal gland). TBR was calculated using tumor lesion SUVmax/back-
Research Committee (13/YH/0281). The administration of radio- ground tissue SUVmean using background liver for liver metastases,
activity was approved by the Administration of Radioactive Sub- background bone marrow for bone metastases, and background medi-
stances Advisory Committee (United Kingdom) (RPC 630/2892/ astinal blood pool for soft-tissue, nodal, and pulmonary metastases.
30595). The Medicines and Health Care Products Regulatory As the presence of liver metastases is an independent prognostic
Agency (United Kingdom) gave permission to administer the factor (15), subgroup analysis of SUV and TBR measurements of
investigational medicinal product (European Clinical Trials no. the liver lesions based on tumor size was performed.
2013-003152-20). All patients provided written informed consent. Independent Reader Evaluations. PET/CT scans were reviewed
The study was conducted in accordance with the Declaration of by 3 independent imaging experts to obtain an objective interrea-
Helsinki and registered with EudraCT (2013-003152-20). der lesion detection rate between [18F]FET-bAG-TOCA PET/CT
and [68Ga]Ga-DOTA-peptide PET/CT scans. To avoid recall bias,
Procedures [18F]FET-bAG-TOCA PET/CT and [68Ga]Ga-DOTA-peptide
PET Imaging Protocol. At Imperial College Health Care NHS PET/CT scans for each subject were reviewed at least 4 wk apart
Trust, clinical PET/CT imaging was performed using [68Ga]Ga- in random order. Readers were blinded to clinical details, type of
DOTATATE, as previously described (10) (mean dose injected, scan and results of other imaging modalities. Readers documented
134.1 MBq; mean uptake period, 37.7 min [range, 27–82 min]). At the presence or absence of lesions in each of the 7 previously
Christie NHS Foundation Trust, imaging was performed using defined areas. Comparison was made between individual readers
[68Ga]Ga-DOTANOC PET (mean dose injected: 136.2 MBq and across both imaging modalities for interobserver agreement. After
mean uptake period of 65.2 min (range, 30–82 min) and a single locking findings, readers then performed a final side-by-side visual
case [68Ga]Ga-DOTATOC (dose, 143 MBq; uptake time, 75 min). analysis of the 2 sets of scans to document any discordant lesions

2 THE JOURNAL OF NUCLEAR MEDICINE  Vol. 00  No. 00  XXX 2024

detected on 1 scan and not the other, to arrive at consensus TABLE 1
between the 3 readers. Baseline Characteristics of Patient Cohort (n 5 45)
Clinical cross-sectional imaging (contrast-enhanced CT or
MRI) performed within 3 mo of [18F]FET-bAG-TOCA was Variable Value*
reviewed by a single experienced observer with more than 20 y of
experience. Age (y)
Median 57
Statistical Analysis Range 29–81
A total of 56 patients were required based on a hypothesized
Sex
90% sensitivity, a noninferiority margin of 10%, power of 80%,
and a level of significance of 5%. Descriptive statistics such as the Male 23 (51)
median and interquartile range were calculated for numeric out- Female 22 (49)
comes. Wilcoxon test was used for comparison of results. Pearson Stage
linear correlation test was used to evaluate correlation between Locally advanced 4 (9)
SUVmax values. Groups were compared using the x2 test. The Metastatic 41 (91)
Cohen k and the Fleiss k were used to determine the level of
Site of primary tumor
agreement among 2 and more than 2 readers of [18F]FET-bAG-
TOCA and [68Ga]Ga-DOTA-peptide PET/CT, respectively. A Pancreas 15 (33)
P value lower than 0.05 was taken to be significant. All statistical Small bowel 20 (44)
analyses were performed using SPSS version 27.0 (IBM Inc.) and Lung 3 (7)
Stata 16 (StataCorp LLC). Other 7 (16)
Grade
RESULTS
1 15 (33)
Baseline Characteristics
2 21 (47)
A total of 56 patients were enrolled to the study. Eleven patients
Unknown 9 (20)
were excluded from the primary analysis: 2 patients underwent
octreotide scan, 6 patients did not have a [68Ga]Ga-DOTA-peptide Site of metastatic disease
PET/CT within 6 mo of [18F]FET-bAG-TOCA PET/CT and in a Liver 27 (60)
further 3 patients, central pathology review after [68Ga]Ga-DOTA- Bone 12 (27)
peptide and [18F]FET-bAG-TOCA imaging reported high-grade Nodes 10 (22)
neuroendocrine carcinoma. A total of 45 patients were included in
Lung 3 (7)
the final analysis. Four patients had a [68Ga]Ga-DOTANOC
PET/CT and 1 patient had [68Ga]Ga-DOTATOC PET/CT. All Other 17 (38)
remaining patients underwent [68Ga]Ga-DOTATATE PET/CT. Median Ki-67 (%) 3 (7)†
The median age of the enrolled population was 57 y (range, 29– Chromogranin A (ng/mL) 72 (92)†
81 y) and most had a diagnosis of small-bowel (44%) NET. All Previous treatment
patients had either locally advanced (9%) or metastatic disease Surgery 24 (53)
(91%), the commonest site of metastases being the liver (58%).
Somatostatin analogs 20 (44)
Demographics and clinical characteristics of the study population
are presented in Table 1. The median interval between [18F]FET- Chemotherapy 9 (20)
bAG-TOCA and [68 Ga]Ga-DOTA PET/CT was 77 d (range, PRRT 7 (16)
6–180 d). RFA 6 (13)
Other 3 (7)
Comparison of [18F]FET-bAG-TOCA and
[68Ga]Ga-DOTA-Peptide PET
Lesion Analysis. On per-lesion analysis, 285 lesions were seen *Data are reported as numbers of patients, with percentages of
both on [18F]FET-bAG-TOCA PET/CT and [68Ga]Ga-DOTA- patients in parentheses.
†
peptide PET/CT. Most lesions were within the liver (38.6% for Value in parentheses is interquartile range.
both imaging modalities) followed by nodal involvement (16.8%) PRRT 5 peptide receptor radiotherapy; RFA 5 radiofrequency
and bone metastases (17.1%). After unblinding of readers, side- ablation.
by-side visual analysis illustrated 20 discordant lesions in 11
patients; in 6 patients additional lesions were detected on For comparative SUV analysis, 285 lesions were included. Excel-
[18F]FET-bAG-TOCA in comparison to [68Ga]Ga-DOTA-peptide, lent correlation in lesion SUVmax between imaging modalities was
conversely additional lesions were detected on [68Ga]Ga-DOTA- observed (r 5 0.91; P , 0.001) (Fig. 2). We then assessed the
peptide in 3 patients compared with [18F]FET-bAG-TOCA and in impact of the use of SSAs on tracer uptake. Twenty-three patients
2 patients there was a mixture with some lesions detected by 1 (51%) were receiving monthly injections with SSAs. No difference
tracer and not the other and vice versa (Table 2). [18F]FET-bAG- was observed in median SUVmax (6 SD) of [18F]FET-bAG-TOCA
TOCA detected an additional 13 lesions (6 liver metastases, 4 of those receiving SSAs (19.2 6 21.1) compared with those who
bone metastases, 2 nodes, and 1 small-bowel lesion) in 8 patients were not (15.8 6 15.9) (P 5 0.06).
and [68Ga]DOTA-peptide PET/CT detected an additional 7 lesions Regional Analysis. No significant difference was noted in the
(4 liver, 1 bone, 1 pancreas, and 1 node) in 5 patients (Fig. 1B). median SUVmax across all tumor regions between the 2 imaging

[18F]FET-bAG-TOCA PET/CT TO DETECT NET  Dubash et al. 3

 TABLE 2
Discordant Lesions Between [18F]FET-bAG-TOCA (FETO) and [68Ga]Ga-DOTA-peptide (DOTA) PET/CT

Patient Time between scans (mo) Congruent site(s) Lesion Discordant lesion site Lesion size (mm) Scan

1 4.6 Liver 1 Liver 7 FETO

2 0.2 Liver, nodal 2 Node 5 FETO
3 Node 5 FETO
3 4.8 Liver, peritoneal 4 Liver 7 FETO
4 0.6 Bone 5 Bone 3 FETO
5 4.3 Liver, nodal 6 Liver 10 DOTA
6 2.5 Gastric, liver 7 Liver 8 DOTA
8 Liver 8 DOTA
7 0.2 Liver, nodal 9 Liver 12 DOTA
10 Pancreas 13 DOTA
8 1.6 Bone, nodal 11 Bone 10 FETO
12 Bone 4 FETO
13 Bone 4 FETO
14 Small bowel 5 FETO
15 Node 15 DOTA
9 1.1 Liver 16 Liver 9 FETO
17 Liver 9 FETO
10 3.7 Liver, nodal 18 Liver 5 FETO
11 4.4 Liver, bone 19 Bone 3 DOTA
20 Liver 8 FETO

modalities (Table 3). The highest median [18F]FET-bAG-TOCA

SUVmax was observed in pancreatic lesions (median SUVmax,
24.5 6 24.9) and the lowest was observed in bone (median
SUVmax, 9.7 6 8.8).
Both tracers demonstrated comparable distribution in back-
ground organs (spleen, pancreas, adrenals, bone) except for
increased background hepatic activity on [18F]FET-bAG-TOCA
PET/CT (Supplemental Fig. 1) (supplemental materials are avail-
able at http://jnm.snmjournals.org). Low physiologic uptake of
[18F]FET-bAG-TOCA was observed, as previously described in
the pituitary, salivary glands, spleen and thyroid gland (8). There

FIGURE 1. (A) Congruent imaging: [68Ga]Ga-DOTATATE imaging and

[18F]FET-bAG-TOCA imaging (maximum-intensity projection [MIP], axial
PET, fused and CT images) in metastatic small-bowel NEN with wide-
spread liver and bone metastases. (B) Incongruent imaging: [68Ga]Ga-
DOTATATE imaging and [18F]FET-bAG-TOCA imaging (MIP, axial PET,
fused and CT images) performed 4 wk apart in metastatic ileal NEN with
liver metastases (green arrows), which are more visible on [18F]FET-bAG- FIGURE 2. Bland–Altman plot of difference in SUVmax between
TOCA. Additional lesion is detected on [18F]FET-bAG-TOCA (blue arrow). [18F]FET-bAG-TOCA and [68Ga]Ga-DOTA-peptide.

4 THE JOURNAL OF NUCLEAR MEDICINE  Vol. 00  No. 00  XXX 2024

 TABLE 3
Median Tumor Uptake (SUVmax) and Tumor-to-Background Ratio (TBR) of [18F]FET-bAG-TOCA and
[68Ga]Ga-DOTA-peptide per Anatomic Region

[18F]FET-bAG-TOCA [68Ga]Ga-DOTA-peptide Median TBR with:

18
No. of Median Median [ F]FET- [68Ga]Ga-
Region lesions SUVmax Range SUVmax Range P bAG-TOCA DOTA-peptide P

Head and neck 3 12.4 10.4–27.7 6.9 6.4–23.4 0.5 29.8 12.2 0.3
Liver 110 19.59 7.2–132.4 20.6 6.7–95.1 0.5 2.5 3.5 ,0.001
Bone 49 9.7 2.2–37.0 7.2 1.9–38.8 0.5 12.5 10.1 0.5
Lung 11 10.4 5.3–42.1 9.4 2.2–38.0 0.9 14.6 15.4 0.4
Pancreas 28 24.5 4.2–85.8 21.9 6.4–88.4 0.8 35.2 36.6 0.6
Abdomen/pelvis 36 18.8 2.7–152.3 21.2 4.1–110.3 0.6 23.5 23.6 0.1
Lymph nodes 48 18.0 3.4–102.4 17.0 3.1–122.9 0.7 21.1 29.7 0.5

was a statistical difference observed in median TBR for liver visualization of NET is well established (4,5). However, the use of
lesions with [18F]FET-bAG-TOCA compared with [68Ga]Ga- [68Ga]Ga necessitates the presence of an onsite (limited life span)
DOTA-peptide PET/CT (2.52 6 1.88 vs. 3.50 6 . 2.35; P , generator, limiting the scalability and availability of [68Ga]Ga-
0.001). No other differences in regional TBR were observed DOTA-peptide radioligands, such that many patients are not able
(Table 3). to access [68Ga]Ga-DOTA-peptide for diagnosis, treatment plan-
ning or assessment of disease progression. To alleviate these
Interreader Agreement issues, we developed a GMP compliant [18F]F-octreotate radio-
Interreader agreement across tumor sites was considered. It was pharmaceutical, [18F]FET-bAG-TOCA. We have previously
possible to estimate with 95% confidence a k-agreement of 86% reported [18F]FET-bAG-TOCA to be safe, with good dosimetry
with an SE of 10% assuming 90% positive ratings among raters and biodistribution, that highlights tumor lesions with high con-
for a total of 45 subjects. Agreement was significantly higher in trast (8). In this prospective study, we have shown that [18F]FET-
the liver with [68Ga]Ga-DOTA-peptide (k 5 0.3) than with bAG-TOCA is excellent in detecting lesions and is not inferior to
[18F]FET-bAG-TOCA (k 5 0.05) (P , 0.001). In particular, [68Ga]Ga-DOTA-peptide- PET/CT for the detection of NET. We
when considering the liver, discrepancies in reads were noted in 4 have also shown the ability of [18F]FET-bAG-TOCA in detecting
patients on [18F]FET-bAG-TOCA imaging, 3 of whom did not small liver lesions, an important consideration given the prognos-
have liver metastases but were thought to be present by 1 of the 3 tic impact of liver metastases (15).
readers. In contrast, only 1 patient was felt to have liver metastases We observed no significant difference in tumoral SUVmax both
on [68Ga]Ga-DOTA-peptide PET, where none were present by 1 on lesion and regional bases between scan types confirming the
of the 3 readers. No significant differences in agreement were noninferiority of [18F]FET-bAG-TOCA for imaging NET.
observed across other sites (Table 4). Observed SUVmax values of [18F]FET-bAG-TOCA are consistent
with the high affinity of [18F]FET-bAG-TOCA for SSTR2 binding
Liver Metastases
(IC50,1.6 6 0.2 nM) (16). The use of SSAs had no impact on
As the presence of liver metastases is an independent prognostic
[18F]FET-bAG-TOCA SUVmax, an important consideration, given
factor, we performed subgroup analysis of uptake in the liver lesions
the widespread use of these agents. Moreover, there was excellent
based on lesion size (,1 cm, 1.0–2 cm, .2.1 cm). Of the 110 liver
correlation between the 2 tracers as confirmed by interobserver
metastases, 28 lesions were smaller than 1 cm, 52 were 1–2 cm, and
agreement across most regions.
30 were larger than 2.1 cm. When considering SUVmax, no signifi-
The liver is the commonest site of metastases and is an indepen-
cant difference in uptake was observed with [18F]FET-bAG-TOCA
dent prognostic factor in patients with NET (15). Background liver
in lesions smaller than 1 cm (15.1 6 7.9) and those 1–2 cm
uptake was significantly lower with [18F]FET-bAG-TOCA com-
(22.7 6 19.9) compared with [68Ga]Ga-DOTATATE (,1 cm,
pared with [68Ga]Ga-DOTA-peptide. This difference in uptake
12.2 6 6.9 [P 5 0.2]; 1–2 cm, 22.4 6 14.5 [P 5 0.4]). A signifi-
can be attributed to differences in elimination. [18F]FET-bAG-
cantly lower median TBR was observed for lesions 1–2 cm with
TOCA is eliminated by both the biliary and renal system, whereas
[18F]FET-bAG-TOCA (3.3 6 2.1) compared with [68Ga]Ga-DOTA-
[68Ga]Ga-DOTA-peptide is eliminated predominantly through the
peptide (4.5 6 2.4) (P 5 0.050. No difference was observed in
kidneys. Hepatic clearance and slow clearance through the com-
median TBR for lesions smaller than 1 cm ([18F]FET-bAG-TOCA,
mon bile duct may contribute to the higher background uptake
1.9 6 0.8; [68Ga]Ga-DOTA-peptide, 2.3 6 1.3) (P 5 0.4). Overall,
observed on [18F]FET-bAG-TOCA imaging. As a result of this
the [18F]FET-bAG-TOCA median TBR was significantly lower in
difference in background uptake, a significant difference in TBR
the liver than the [68Ga]Ga-DOTA-peptide median TBR (2.5 6 1.9 in the liver between the 2 tracers was observed. The higher liver
vs. 3.5 6 2.3; P , 0.001). background activity may have contributed to the difference
observed on interreader agreement within the liver, whereby
DISCUSSION
observers were less confident in 3 cases about the absence of
The superiority of [68Ga]Ga-DOTA-peptide PET/CT over metastases in “normal liver”, a concept that needs exploring in
111
[ In]In-octreotide SPECT/CT and contrast CT imaging for the future work. However, of the 20 discordant lesions, 10 were in the

[18F]FET-bAG-TOCA PET/CT TO DETECT NET  Dubash et al. 5

 liver, 6 were only detected on [18F]FET-bAG-TOCA and 4 with
[68Ga]Ga-DOTA-peptide imaging. The management in these cases

P for [18F]FET-

DOTA-peptide
vs. [68Ga]Ga-
bAG-TOCA
did not change as the patients already had multiple liver metastases.

,0.001
0.04
Since 18F has a shorter positron range and higher positron yield

NC

NC
0.2
0.3

0.7
than 68Ga, one might postulate that [18F]FET-bAG-TOCA imag-
ing could detect smaller lesions compared with [68Ga]Ga-DOTA-
peptide imaging. On the 20 discordant lesions, 13 were detected
only on [18F]FET-bAG-TOCA, and all were less than or equal to
10 mm in size, whereas 7 were detected only on[68Ga]Ga-DOTA-

,0.001

0.008
0.05
0.04
Interrater Agreement for [18F]FET-bAG-TOCA and [68Ga]Ga-DOTA-peptide per Anatomic Region Between All 4 Raters

peptide imaging, of which all were greater than or equal to 8 mm

NC
0.6
0.5
P in size except for 1 (Table 2). The latest digital PET detector tech-
nology may improve detection of small lesions.
The use of [18F]FET-bAG-TOCA PET/CT may be considered in
the clinical setting where difficulties accessing [68Ga]Ga-DOTA-
20.01 (20.02 to 20.006)

peptide have led to longer waiting times for patients, particularly

0.1 (20.03 to 0.3)
0.1 (20.02 to 0.3)

0.2 (20.02 to 0.3)

[68Ga]Ga-DOTA-peptide

where delivery of [68Ga]Ga-DOTA-peptide is limited to those cen-

ters within close proximity to the gallium generator. Delivery of low
NC
yields of [68Ga]Ga-DOTA-peptide is also a common problem and
k*

20.01 (NC)
0.3 (NC)

can lead to last minute cancellation of scanning slots with an ever-

increasing burden on nuclear medicine departments. Recent work
Agreement between raters 1, 2, 3, and 4 for:

has explored the utility of [18F]F-AIF-1,4,7-triazacyclononane-1,4,7-

tri-acetate-octreotide ([18F]F-AIF-NOTA-octreotide compared with
[68Ga]Ga-DOTATATE/NOC in patients with NET (17). In this
study the noninferiority of 18F-labeled AIF-NOTA-octreotide was
Agreement (%)

illustrated; the authors reported high physiologic uptake in the pan-

creas, necessitating the need for additional cross-sectional imaging
97.8
98.9
98.5
91.5

100.0
95.2

96.3

to delineate any pancreatic lesion, a feature not observed with

[18F]FET-bAG-TOCA (8). Moreover, SUVmax was lower than
[68Ga]Ga-DOTATATE and TBR within the bone was particularly
TABLE 4

lower, which may have implications in assessing disease response

to therapy within the bone. [64Cu]Cu-DOTATATE has also recently
been studied, with comparable results to [68Ga]Ga-DOTATOC,
<0.001

,0.001

,0.001
,0.001
0.6
0.7

0.5
P

albeit with a higher radiation burden, which may not be acceptable

to users, particularly as patients typically undergo multiple PET/CT
studies during their disease journey (18).
However, there are some key limitations. As [18F]FET-bAG-
20.03 (20.05 to 20.006)

TOCA imaging was performed after [68Ga]Ga-DOTA-peptide imag-

20.01 (20.02 to 20.01)

ing in most patients, potential sequence effects cannot be excluded,

0.2 (20.01 to 0.3)

0.0005 (20.07 to 0.2)

0.3 (20.01 to 0.3)

[ F]FET-bAG-TOCA

but most were performed within 6 mo and no change in treatment

0.2 (0.1 to 0.3)

occurred between both scans. Moreover, due to the variation in time

k*

interval between the 2 scans, changes in tumor composition or size

0.3 (NC)

and the possible change in SSTR density cannot be excluded (14).

Although most patients underwent [68Ga]Ga-DOTATATE imaging,
a number were imaged with other [68Ga]Ga-DOTA-peptide radioli-
18

gands, the impact of which remains unclear. Finally, PET findings

were not validated by a reference imaging standard such that sensi-
tivity or specificity cannot be established.
Agreement (%)

*Values in parentheses are 95% CIs.

CONCLUSION
97.4
97.8
94.4
89.3
88.5
98.5
97.4

In this prospective head-to-head comparison of [18F]FET-bAG-

TOCA PET/CT and [68Ga]Ga-DOTA-peptide PET/CT we have
shown excellent tumoural uptake and noninferiority at both lesion
and regional levels. [18F]FET-bAG-TOCA could potentially be
NC 5 noncalculable.

used clinically as an alternate to [68Ga]Ga-DOTA-peptide. Further

Abdomen/pelvis
Head and neck

developments could lead to its use as a theranostic agent in locally

Lymph nodes

advanced and metastatic NET.

Pancreas
Scan site

DISCLOSURE
Bone
Lung
Liver

This work was funded by Medical Research Council Developmen-

tal Clinical Studies grant MR/J007986/1. The authors acknowledge

6 THE JOURNAL OF NUCLEAR MEDICINE  Vol. 00  No. 00  XXX 2024

infrastructure support from the National Institute for Health Research REFERENCES
Imperial Biomedical Research Centre and the Imperial College 1. Pavel M, Oberg K, Falconi M, et al. Gastroenteropancreatic neuroendocrine neo-
Experimental Cancer Medicine Centre. Rohini Sharma received grant plasms: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up.
funding from Incyte, AAA, Boston Scientific, and Terumo. Mair
ead Ann Oncol. 2020;31:844–860.
McNamara received research grant support from Servier, Ipsen, 2. Yao JC, Hassan M, Phan A, et al. One hundred years after “carcinoid”: epidemiol-
ogy of and prognostic factors for neuroendocrine tumors in 35,825 cases in the
NuCana, and AstraZeneca, travel and accommodation support from
United States. J Clin Oncol. 2008;26:3063–3072.
AAA, Ipsen, and AstraZeneca, and speaker honoraria from AAA 3. Kl€oppel G, Couvelard A, Perren A, et al. ENETS Consensus Guidelines for the
and AstraZeneca and is on the advisory boards for Incyte and Astra- Standards of Care in Neuroendocrine Tumors: towards a standardized approach to
Zeneca. Andrea Frilling received grants from Novartis, AAA, Ipsen, the diagnosis of gastroenteropancreatic neuroendocrine tumors and their prognostic
Sirtex, and Clifton, all unrelated to the submitted work. Angela stratification. Neuroendocrinology. 2009;90:162–166.
4. Sadowski SM, Neychev V, Millo C, et al. Prospective study of 68Ga-DOTATATE
Lamarca received travel and educational support from Ipsen, Pfizer, positron emission tomography/computed tomography for detecting gastro-entero-
Bayer, AAA, SirtEx, Novartis, Mylan, Delcath Advanz Pharma, pancreatic neuroendocrine tumors and unknown primary sites. J Clin Oncol. 2016;
and Roche, speaker honoraria from Merck, Pfizer, Ipsen, Incyte, 34:588–596.
AAA/Novartis, QED, Servier, AstraZeneca, EISAI, Roche, Advanz 5. Krausz Y, Freedman N, Rubinstein R, et al. 68Ga-DOTA-NOC PET/CT imaging of
neuroendocrine tumors: comparison with 111In-DTPA-octreotide (OctreoScanV R ).
Pharma, and MSD, advisory and consultancy honoraria from EISAI,
Mol Imaging Biol. 2011;13:583–593.
Nutricia, Ipsen, QED, Roche, Servier, Boston Scientific, Albireo 6. Strosberg J, El-Haddad G, Wolin E, et al. Phase 3 trial of 177Lu-Dotatate for mid-
Pharma, AstraZeneca, Boehringer Ingelheim, GENFIT, TransThera gut neuroendocrine tumors. N Engl J Med. 2017;376:125–135.
Biosciences, Taiho, and MSD, and principal investigator–associated 7. Mapelli P, Tam HH, Sharma R, Aboagye EO, Al-Nahhas A. Frequency and signifi-
institutional funding from QED, Merck, Boehringer Ingelheim, Ser- cance of physiological versus pathological uptake of 68Ga-DOTATATE in the pan-
creas: validation with morphological imaging. Nucl Med Commun. 2014;35:613–619.
vier, AstraZeneca, GenFit, Albireo Pharma, Taiho, TransThera, and 8. Dubash SR, Keat N, Mapelli P, et al. Clinical translation of a click-labeled 18F-
Roche and is a member of the Knowledge Network and NETCon- octreotate radioligand for imaging neuroendocrine tumors. J Nucl Med. 2016;57:
nect Initiatives funded by Ipsen. Juan Valle received personal fees 1207–1213.
from Agios, AstraZeneca, Baxter, Genoscience Pharma, Hutchison 9. Iddon L, Leyton J, Indrevoll B, et al. Synthesis and in vitro evaluation of
[18F]fluoroethyl triazole labelled [Tyr3]octreotate analogues using click chemistry.
Medipharma, Imaging Equipment Ltd (AAA), Incyte, Ipsen, QED,
Bioorg Med Chem Lett. 2011;21:3122–3127.
Servier, Sirtex, and Zymeworks and grants, personal fees, and nonfi- 10. Sharma R, Wang WM, Yusuf S, et al. 68Ga-DOTATATE PET/CT parameters pre-
nancial support from NuCana, all outside the submitted work. Rich- dict response to peptide receptor radionuclide therapy in neuroendocrine tumours.
ard Hubner received consultancy fees from Beigene, Ipsen, and Radiother Oncol. 2019;141:108–115.
Novartis and travel and accommodation support from Roche. The 11. Koopmans KP, de Vries EG, Kema IP, et al. Staging of carcinoid tumours with 18F-
DOPA PET: a prospective, diagnostic accuracy study. Lancet Oncol. 2006;7:728–734.
views expressed are those of the authors and not necessarily those of 12. Wild D, Bomanji JB, Benkert P, et al. Comparison of 68Ga-DOTANOC and 68Ga-
the NIHR or the Department of Health and Social Care. No other DOTATATE PET/CT within patients with gastroenteropancreatic neuroendocrine
potential conflict of interest relevant to this article was reported. tumors. J Nucl Med. 2013;54:364–372.
13. Koopmans KP, Neels ON, Kema IP, et al. Molecular imaging in neuroendocrine
tumors: molecular uptake mechanisms and clinical results. Crit Rev Oncol Hema-
tol. 2009;71:199–213.
KEY POINTS 14. Poeppel TD, Binse I, Petersenn S, et al. 68Ga-DOTATOC versus 68Ga-DOTA-
TATE PET/CT in functional imaging of neuroendocrine tumors. J Nucl Med.
2011;52:1864–1870.
QUESTION: How does the novel fluorine-labeled PET tracer 15. Dasari A, Shen C, Halperin D, et al. Trends in the incidence, prevalence, and sur-
[18F]FET-bAG-TOCA compare with [68Ga]Ga-DOTA-peptide vival outcomes in patients with neuroendocrine tumors in the United States. JAMA
PET/CT for the detection of NET? Oncol. 2017;3:1335–1342.
16. Leyton J, Iddon L, Perumal M, et al. Targeting somatostatin receptors: preclinical
PERTINENT FINDINGS: In a prospective, noninferiority study evaluation of novel 18F-fluoroethyltriazole-Tyr3-octreotate analogs for PET. J Nucl
in 45 patients with histologically confirmed NET, we observed Med. 2011;52:1441–1448.
excellent correlations between both tracers with no difference 17. Pauwels E, Cleeren F, Tshibangu T, et al. 18F-AlF-NOTA-octreotide outperforms
68
across median SUVmax. Ga-DOTATATE/NOC PET in neuroendocrine tumor patients: results from a pro-
spective, multicenter study. J Nucl Med. 2023;64:632–638.
IMPLICATIONS FOR PATIENT CARE: [18F]FET-bAG-TOCA may 18. Johnbeck CB, Knigge U, Loft A, et al. Head-to-head comparison of 64Cu-DOTA-
be considered in routine clinical practice for imaging NET. TATE and 68Ga-DOTATOC PET/CT: a prospective study of 59 patients with neu-
roendocrine tumors. J Nucl Med. 2017;58:451–457.

[18F]FET-bAG-TOCA PET/CT TO DETECT NET  Dubash et al. 7