Achmad et al. BMC Medical Imaging (2017) 17:66
DOI 10.1186/s12880-017-0237-1
RESEARCH ARTICLE
Open Access
The diagnostic performance of 18F-FAMT
PET and 18F-FDG PET for malignancy
detection: a meta-analysis
Arifudin Achmad1,2* , Anu Bhattarai1, Ryan Yudistiro1,3, Yusri Dwi Heryanto1, Tetsuya Higuchi1
and Yoshito Tsushima1
Abstract
Background: This meta-analysis aims to compare the diagnostic performance of L-3-18F-α-methyl tyrosine (18F-FAMT)
positron emission tomography (PET) and 2-deoxy-2-[18F]fluoro-D-glucose (18F-FDG) PET for malignancy detection.
Methods: The workflow of this study follows Cochrane Collaboration Guidelines of a systematic review of diagnostic
test accuracy studies. An electronic search was performed for clinical diagnostic studies directly comparing 18F-FAMT
and 18F-FDG PET for malignant tumors. Study quality, the risks of bias and sources of variation among studies were
assessed using the QUADAS (Quality Assessment of Diagnostic Accuracy Studies) assessment tool. A separate metaanalysis was performed for diagnostic performance based on visual assessment and diagnostic cut-off values.
Whenever possible, a bivariate random-effect model was used for analysis and pooling of diagnostic measures
across studies.
Results: Electronic search revealed 56 peer-reviewed basic science investigations and clinical studies. Six eligible studies
(272 patients) of various type of cancer were meta-analyzed. The 18F-FAMT diagnostic accuracy for malignancy was
higher than 18F-FDG based on both visual assessment (diagnostic odd ratio (DOR): 8.90, 95% confidence interval (CI) [2.4,
32.5]) vs 4.63, 95% CI [1.8, 12.2], area under curve (AUC): 77.4% vs 72.8%) and diagnostic cut-off (DOR: 13.83, 95% CI [6.3,
30.6] vs 7.85, 95% CI [3.7, 16.8], AUC: 85.6% vs 80.2%), respectively. While the average sensitivity and specificity of
18
F-FAMT and 18F-FDG based on visual assessment were similar, 18F-FAMT was significantly more specific than 18F-FDG
(p < 0.05) based on diagnostic cut-off values.
Conclusions: 18F-FAMT is more specific for malignancy than 18F-FDG, while their sensitivity is comparable. 18F-FAMT PET
is equal to 18F-FDG PET in diagnostic performance for malignancy detection in several cancer types.
Keywords: 18F-FAMT, 18F-FDG, Malignancy, Meta-analysis, Diagnostic accuracy
Background
Since its introduction as a positron emission tomography (PET) tracer back in the early 1970’s, [18F]-fluorodeoxyglucose (18F-FDG) has been widely utilized and
now comprises more than 96% of PET studies worldwide
[1]. Even though 18F-FDG is mainly a radiotracer for
* Correspondence: aachmad@gunma-u.ac.jp; m09702036@gunma-u.ac.jp
1
Department of Diagnostic Radiology and Nuclear Medicine, Gunma
University Graduate School of Medicine, 3-39-22 Showa-machi, Maebashi,
Gunma 371-8511, Japan
2
Department of Nuclear Medicine and Molecular Imaging, Faculty of
Medicine, Padjadjaran University, Jl. Professor Eyckman No.38, Bandung, West
Java 40161, Indonesia
Full list of author information is available at the end of the article
oncology, it is not a tumor-specific PET tracer, since it is
essentially based on the presence of elevated glucose uptake [2]. Many malignant lesions, in fact, are poorly imaged with 18F-FDG; some due to their slow growth or
low metabolic nature, and others due to their location
within highly metabolic organs such as the brain and
liver [3]. Various alternative PET tracers have been synthesized and evaluated over the last decade to overcome
the limitations of 18F-FDG, including tracers based on
amino acid metabolism such as L-3-18F-α-methyl tyrosine (18F-FAMT) [1, 4].
18
F-FAMT has been validated in several clinical studies
to be useful for the prediction of cancer prognosis and
© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Achmad et al. BMC Medical Imaging (2017) 17:66
to rule out benign lesions from malignant neoplasms
[5–13]. The tumor accumulation of 18F-FAMT is exclusively facilitated by the L-type amino acid transporter 1
(LAT1), which is highly upregulated in malignant cells
[14]. Unlike other amino acid PET tracers that are not
specific to a single amino acid transporter, 18F-FAMT
has a α-methyl moiety that allows it to be transported
only by LAT1, making it highly specific for malignancies
[15]. Although a handful of clinical studies have investigated its potential in malignant tumor detection, the
overall diagnostic performance of 18F-FAMT remains
unknown. The present meta-analysis aimed to determine
the diagnostic performance of 18F-FAMT PET for detection and evaluation of malignant lesions in a direct sideby-side comparison to 18F-FDG PET.
Methods
Search strategy and study selection
The design of this study followed the current recommendations for systematic review of diagnostic test accuracy studies from the Cochrane Collaboration [16, 17].
Studies evaluating 18F-FAMT PET or PET/CT as a diagnostic tool for evaluation of malignancy were electronically searched in Pubmed/MEDLINE, Web of Science,
ScienceDirect, and Google Scholar databases from the
inception of 18F-FAMT to December 2016 without language restriction. The search algorithm was based on a
combination of the following terms: 18F-FAMT or
18
F-FMT or “alpha-methyltyrosine.” To find more potential studies, we also screened references of the
retrieved studies. Articles without raw clinical data such
as reviews, conference abstracts, editorial, comments,
preclinical, animal and non-radiopharmaceutical studies,
or clinical studies with fewer than ten patients were
excluded. The following information was extracted: first
author’s name, year of publication, study design, study
population, types/subtypes of malignancies, injected
dose, imaging parameters, cut-off values of quantitative
parameters, study and follow-up period, final diagnosis,
and the reference standard.
The clinical studies obtained were subject to inclusion
criteria for further analysis: (a) both 18F-FAMT and
18
F-FDG were used to differentiate malignant tumors
from benign lesions, (b) histopathological analysis and/
or close clinical and imaging follow-up were used as reference standards, (c) when data or subsets of data were
presented in more than one article, the article with the
most detailed/recent data was chosen, and (d) only articles in which at least 10 of the 14 questions in the
QUADAS (Quality Assessment Tool for Diagnostic
Accuracy Studies) questionnaire were answered ‘yes’
were included [18]. Studies were screened for eligibility,
the risk of bias, and source of variations by three authors
Page 2 of 9
(AA, AB, RY) independently. Disagreements regarding
the eligibility of a study were resolved by consensus.
Meta-analysis
Meta-analysis of the diagnostic performance of
F-FAMT and 18F-FDG in recognizing malignancies
was performed following the current recommendations
[17] and was conducted separately for two diagnostic
methods: 1) by visual assessment, and 2) by diagnostic
cut-off values applied in each study. From each study included, the number of true positives, false positives, true
negatives, and false negatives were extracted to construct
a 2 × 2 contingency table. If studies lacked clear data to
produce such tables, the first authors were contacted
when possible. This main data were described on forest
plots of specificity and sensitivity.
Heterogeneity and between-study variability were evaluated, and subgroup study (meta-regression analysis)
was used to investigate the source, if any. A Higgins’ inconsistency I2 up to 30% was considered little evidence
of heterogeneity. To determine whether different thresholds were used to define positive and negative test
results (either explicitly or implicitly), the Spearman ρ
between the logit of sensitivity and logit of 1 − specificity
was calculated to assess the presence of a threshold effect. A strong positive correlation (Spearman ρ > 0.6)
would suggest the presence of a threshold effect. Whenever possible, a bivariate random-effect model metaanalysis method was used to obtain summary estimates
of sensitivity and specificity across studies instead of univariate approaches.
The hierarchical summary of the receiver operating
characteristic (HSROC) curve was plotted following the
method of Rücker and Schumacher [19]. The area under
the curve (AUC), which is the average true-positive rate
over the entire range of false-positive rate, serves as a
global measure of test performance, while the diagnostic
odd ratio (DOR) is calculated to describe the diagnostic
value [20]. Note that the DOR is a single overall indicator of diagnostic performance and is, unlike sensitivity
and specificity, independent of any threshold value.
Meta-analysis was performed using the ‘mada’ (MetaAnalysis of Diagnostic Accuracy) package in R statistical
software version 3.2.2 [21, 22].
18
Results
Literature search
The systematic search was performed to collect diagnostic test studies using 18F-FAMT and 18F-FDG PET for
malignancy detection. The search yielded 65 studies involving 18F-FAMT as PET radiotracer in basic science
investigations and clinical studies. There were three
radiochemistry studies, nine in vitro and animal studies,
four review articles, and 49 clinical studies. Thirty
Achmad et al. BMC Medical Imaging (2017) 17:66
studies among these 49 clinical studies were original articles in which both PET radiotracers were employed.
Figure 1 summarizes the systematic study selection.
Study eligibility, quality, and risk of bias
Nine eligible studies according to the inclusion criteria
(Table 1) were further evaluated with QUADAS tool. All
were prospective studies of good quality (QUADAS
Scores >10) involving at least 19 patients (patient number range: 19–74) and 21 lesions (lesion number range:
21–75). Overall, the nine eligible studies had a low risk
of bias, except in blinding from the index test results
(Additional file 1: Table S1). Blinding from the index test
results was sometimes unavoidable in the clinical workflow, since histopathological diagnosis is established after
the primary surgery or biopsy, while PET imaging is an
early step in workups to establish the clinical diagnosis.
In one study, the histopathology (biopsy) diagnosis was
known before the PET study was performed [7].
However, this study was later excluded from the metaanalysis (Table 1). The other important potential source
of bias was the use of other imaging studies (CT, MRI or
bone scans) and close clinical monitoring as verification
methods in one study [5]. However, in this study, only
two patients (from 19 patients, total 57 lesions) had their
lesions diagnosed without any histological examination:
Fig. 1 The study selection
Page 3 of 9
one had malignant melanoma in the foot (single lesion),
and the other had diffuse malignant melanoma (lesions
in the brain and spinal cord).
Six studies were included in the final meta-analysis
due to the availability of individual patient data to
construct 2 × 2 contingency tables (Table 1 and
Additional file 1: Table S2). All studies employed maximum standardized uptake value (SUVmax) for quantitative interpretation of the PET images. Four explicitly
described SUVmax cut-off value for discrimination
between malignant and benign lesions. The SUVmax
cut-offs of 18F-FAMT studies ranged from 1 to 1.45
while in 18F-FDG studies, they ranged from 0.81 to 4.72.
Six studies with a total sample size of 272 patients (278
lesions) with malignancy from musculoskeletal [12, 23],
fatty tumors [11], maxillofacial tumors [9], lung cancer
[24], and several different tumors [5] were included.
Descriptive statistics
Figure 2 described the paired sensitivity and specificity
of 18F-FAMT and 18F-FDG of each study in forest plots.
The sensitivity of both radiotracers was homogeneous
either based on the visual assessment or diagnostic cutoff values. Their specificity was heterogeneous based on
visual assessment. The Spearman correlation (ρ) between
sensitivity and the logit of 1-specificity suggest that
Study
(year) [Ref]
N
Mean/
Median
Age (range)
Inoue
(1999) [6]
20 41 ± 21
(1–71)
Watanabe 74 44 (12–83)
(2000) [12]a
Inoue
(2001) [5]a
19 58 (20–84)
F-FAMT and
18
F-FDG
Sex
(M/F)
Tumour & other pathology examined
No. of lesions
18
F-FAMT 18F-FDG
dose
dose
8/12
Brain tumour
23
185 MBq
22 mlg, 53 bgn
Follow-up Gold
period
Standard &
verification
Study Study
design Period
QUADAS
Scoreb
200 MBq Within
1 wk.
> 4 mo.
H (16), Img
& Cln (4)
Pro
ND
12 (6,11)
185–350
MBq
185–
350
MBq
ND
> 1 y.
H
Pro
2/‘98–6/‘99
13 (11)
Lung cancer (10), mlg myeloma (2),
57
Chondrosarcoma (1), Prostate (1),
mlg lymphoma (1), mlg of unknown
origin (1), Schwannoma (1), Sarcoidosis (2)
200–370
MBq
200–
370
MBq
Within
1 wk.
> 8 mo.
H (31), Img
& Cln (26)
Pro
ND
10 (1,6,7,11)
- Gliomatosis cerebri (8): Anaplastic (1),
grade II astrocytoma (4) grade III
astrocytoma (3)
- Non-neoplastic diseases (6)
ND
185 MBq
ND
1–5 wk.
(detailed)
ND
H (8) for
Pro
gliomatosis
Img & Cln for
non-neoplasms
ND
12 (6,11)
29/28 Fatty tissue tumour
32 lipoma,
25 liposarcoma
185–350
MBq
185–
350
MBq
ND
ND
H (57)
Pro
9/‘97–12/‘03 12 (11,4Un)
9 AC, 6 SQC, 2
NSCLC, 16 LNM,
24 sarcoidosis
4–5
MBq/kg
5–6
MBq/kg
ND for lung > 2 y.
cancer.
Sarcoidosis:
1 wk. after
diagnosis
established.
H (41)
Pro
Sarcoidosis:
9/‘98–8/‘05
12 (10,12)
5–6
MBq/kg
5–6
MBq/kg
2 wk.
H (43)
Pro
5/‘99–7/‘06
13 (11)
1–14 d.
ND
(mean: 4 d.)
H (43)
Pro
5/‘07–3/‘08
12 (11,12)
H, Img
& Cln (36)
Pro
ND
13 (11)
37/38 Musculoskeletal tumours: 24 bone,
48 soft tissue
13/6
Sato (2003) 14 Gliomatosis:
4/4
[13]
5/1
45 (15–60)
Non-neoplastic:
41.2 (23–76)
57 58.1 (27–87)
Suzuki
(2005) [11]a
18
Kaira
(2007) [7]
41 61 (45–82)
43 (20–73)
13/4
9/15
Miyakubo
(2007) [9] a
43 ND (31–90)
ND (32–81)
16/20 Maxillofacial tumours:
5/2
mlg. (36), bgn. (7) and LNM (14)
34 SQC, 1
rhabdomyosarcoma,
1 mucoepidermoid
carcinoma, 14 LNM,
7 bgn
33/10 Thoracic tumours: mlg. (37), bgn. (6)
19 AC, 9 SQC, 1 LCC,
ND
2 atypical SQC,
3 Bronchoalveolar
carci-noma, 1 carcinoid,
6 bgn
5
MBq/kg
22/14 Musculoskeletal tumour
13 mlg, 23 bgn
320 MBq Maximum
2 wk.
43 67 (41–79)
Kaira
(2009) [24]a
Tian (2011)
[23]a
36 ND (11–84)
Lung cancer (17)
Sarcoidosis (24)
PET studies
interval
260 MBq
> 6 mo.
ND
Page 4 of 9
Abbreviations: MBq/kg (MegaBecquerel/kg), d. Days, wk. Weeks, mo. Months, y. Year, H Histopathology, Img Imaging, Cln Clinical follow-up, Pro Prospective, Retro Retrospective, mlg malignant, bgn benign,
LNM lymph node metastases, AC adenocarcinoma, SQC squamous cell carcinoma, NSCLC non-small cell lung cancer, LCC large cell carcinoma, ND not determined
a
Studies included in the meta-analysis
b
QUADAS Score was presented as ‘total score (item number which answered ‘No’ or ‘Unclear’ (Un))’. QUADAS tool items were described in Tablze S1
Achmad et al. BMC Medical Imaging (2017) 17:66
Table 1 Characteristics of Diagnostic Comparison Studies of
Achmad et al. BMC Medical Imaging (2017) 17:66
Page 5 of 9
Fig. 2 Sensitivity and specificity of 18F-FAMT and 18F-FDG for malignancy detection
accuracy of both radiotracers based on visual assessment
may be influenced by threshold effects (≥ 0.6). However,
their accuracy was less affected by threshold effect when
the diagnostic cut-off value was implemented.
4.63, while those based on diagnostic cut-off were 13.83
and 7.85, respectively. The heterogeneity between
studies as well as inter-study was observed only mildly
on 18F-FAMT studies based on visual assessment
(Higgins’ I2: 11.76%, τ2: 1.46) while it was not observed
in other studies.
Meta-analysis
Due to the small number of studies included, both univariate and bivariate approach meta-analysis was performed. The bivariate approach is the method currently
recommended; however, it cannot handle small sample
sizes [17]. Meta-regression or subgroup analysis (to explore the source of heterogeneity) was also irrelevant
due to the limited number of studies.
Table 2 described the summary estimates from the
random effects univariate analysis. DOR of 18F-FAMT
and 18F-FDG based on visual assessment were 8.90 and
Table 2 Summary estimates from univariate meta-analysis
Summary estimates
(95% CI)
Based on visual
assessment
Based on
diagnostic cut-off
18
F-FAMT
18
18
Between-study
heterogeneity
2
I : 11.76%
2
I : 0%
2
I : 0%
I2: 0%
Inter-study
heterogeneity
τ2: 1.46
τ2: 0.30
τ2: 0.00
τ2: 0.00
DOR
8.90
(2.4–32.5)
4.63
(1.8–12.2)
13.83
(6.3–30.6)
7.85
(3.7–16.8)
F-FDG
F-FAMT
18
F-FDG
Achmad et al. BMC Medical Imaging (2017) 17:66
Page 6 of 9
The summary estimate measures of the random effects
bivariate model are described in Table 3. There was no
significant difference in average sensitivity and specificity
between 18F-FAMT and 18F-FDG based on visual assessment (p = 0.181 and 0.207, respectively). However,
18
F-FAMT was significantly more specific than 18F-FDG
(p < 0.01) based on diagnostic cut-off values. DOR of
18
F-FAMT and 18F-FDG based on visual assessment
were 8.33 and 3.88 while based on diagnostic cut-off
were 16.70 and 8.17, respectively.
The HSROC curves of diagnostic performance comparison are shown in Fig. 3. The AUC of diagnostic performance of 18F-FAMT and 18F-FDG based on visual
assessment was 77.4% and 72.8%, while those based on
diagnostic cut-off were 85.6% and 80.2%, respectively.
The estimated SROC curves from the bivariate model
(Rutter-Gatsonis method) were also plotted as a reference (Fig. 3, dashed lines). The summary operating
points of 18F-FAMT were on the left side of those of
18
F-FDG in both HSROC curves comparison, which indicated that 18F-FAMT provided more specificity. Meanwhile, their similar heights of the summary operating
points on the Y-axis showed that their sensitivities were
comparable.
Discussion
This meta-analysis summarized the diagnostic performance of 18F-FAMT PET for detection of various malignancies in six studies with total 278 patients. Overall,
the included studies have a low risk of bias with good
methodological quality based on QUADAS tool. Our results demonstrated that 18F-FAMT is comparable with
18
F-FDG for its diagnostic performance in detecting
Table 3 Summary estimates from bivariate meta-analysis
Summary
estimates
(95% CI)
Based on visual
assessment
Average
Sensitivity
80.7%
88.8%
74.1%
78.3%
(72.4–87.0%) (80.2–93.9%) (63.0–82.7%) (67.8–86.1%)
p values
Average
Specificity
p values
18
F-FAMT
Based on
diagnostic cut-off
18
F-FDG
18
F-FAMT
18
F-FDG
0.181
0.542
60.7%
29.2%
(25.3–87.6%) (9.2–62.5%)
84.4%
68.1%
(75.7–90.4%) (58.1–76.6%)
0.207
0.009
Positive
Likelihood
2.46
(1.11–6.23)
1.34
(1.00–2.25)
4.90
(2.96–7.92)
2.48
(1.81–3.44)
Negative
Likelihood
0.36
(0.20–0.70)
0.44
(0.20–0.98)
0.31
(0.20–0.45)
0.33
(0.20–0.49)
DOR
8.33
3.88
16.70
8.19
(1.60–26.10) (1.02–10.40) (7.25–33.40) (3.86–15.40)
AUCa
77.4%
72.8%
85.6%
80.2%
λ (mean
accuracy)
3.81
3.08
3.44
2.99
a
approximated following Rucker-Schumacher’s method [19]
malignancies by either visual assessments or diagnostic
cut-off values. Moreover, 18F-FAMT capability is coherent in several types of tumors, where all individual
diagnostic test studies directly compared the two radiotracers on the same patients in a prospective study design. Additionally, the potential for selection bias can be
safely ignored due to the sufficient number of lesions
evaluated in each study included (n > 20). Another
strength of this meta-analysis is that even though the
study number is limited, heterogeneity was not substantial. The source of observed mild heterogeneity was
likely due to threshold effects, which was found in studies based on visual assessment. However, other potential
sources of heterogeneity should not be neglected since
subgroup analysis was not applicable [25]. Publication
bias is an important consideration in any meta-analysis.
However, DOR heterogeneity observed in our results precludes the necessity for a funnel plot asymmetry test [26].
In the current recommendation for meta-analysis of
diagnostic test accuracy from The Cochrane Collaboration, bivariate approach meta-analysis is preferred over
the traditional univariate meta-analysis [17]. However,
guidance for determining methodological approaches for
meta-analysis with small numbers of studies is currently
lacking. In this case, Doebler et al. and Takwoingi et al.
encouraged the use of univariate approaches excluding
pooling sensitivities and specificities [21, 27]. Eventually,
both univariate and bivariate methods were conducted
in the current study, and the diagnostic performance of
18
F-FAMT against 18F-FDG was consistent under both
approaches. The more conservative approach for
HSROC estimation (Rücker-Schumacher’s method) also
showed a similar tendency to the traditional HSROC
parametrization (Rutter-Gatsonis’s method) [19].
Despite the limited number of studies included, results
of our meta-analysis reflect the natural characteristics of
both radiotracers that assess malignant lesions via different metabolic processes. The key feature of 18F-FDG is
its superior capability to depict increased metabolic activity reflected by cell glucose consumption. The price of
this high sensitivity is the detection accuracy that is
prone to being obscured by normal physiological uptake,
inflammation, and active benign tumors [2]. In a recent
large-size meta-analysis, 18F-FDG PET failed to maintain
its diagnostic accuracy for lung cancer in populations
with endemic infectious lung disease [28]. 18F-FDG PET
was also only moderately accurate for differentiating benign from malignant pleural effusions [29].
In another meta-analysis, whole-body 18F-FDG PET/
CT remained superior to conventional imaging in the
detection of distant malignancies, regardless of the primary tumor site and type [30]. However, the diagnostic
accuracy of a PET radiotracer for lesions in the thorax
and abdomen, where most primary lesions are located, is
Achmad et al. BMC Medical Imaging (2017) 17:66
Page 7 of 9
Fig. 3 Summary ROC plots obtained from the bivariate model of the diagnostic performance of 18F-FAMT and 18F-FDG based on (a) visual assessment
and (b) diagnostic cut-off value. Oval regions are the 95% confidence regions around the summary operating points. The SROC curves from
parametrization according to Rutter and Gatsonis are also presented
essential. It is well known that the role of 18F-FDG PET
in oncology is often mitigated by many pitfalls, including
background physiological uptake of major organs [31].
On the other hand, 18F-FAMT specific uptake depicted
the actual malignant process. 18F-FAMT uptake reflects
excessive transport of amino acids via LAT1, which is
absent in normal cells and pathology other than malignancy [15]. However, the trade-off of 18F-FAMT’s high
specificity is the relatively small absolute uptake in
tumor cells, as a consequence of the nature of the LAT1
transporter. The influx of one amino acid substrate into
tumor cells via LAT1 is mandatoryly coupled to the efflux of another amino acid substrate, resulting in
18
F-FAMT’s relatively fast clearance from the tumor
[14]. Nonetheless, the advantage of 18F-FAMT is the
minimal background uptake in all organs except kidney
and urinary tracts, allowing one to obtain high contrast
images clearly depicting various types of malignancy including brain tumors [6, 13].
Meta-analyses evaluating the diagnostic performance
of 18F-FDG PET in malignancy detection were mostly
limited to a particular cancer type, or in comparison
with conventional imaging (CT or MRI) or hybrid
imaging (PET/CT or PET/MRI). Currently, only a few
tumor-specific PET radiotracers are continuously
investigated in a clinical setting for various type of
cancers [32]. 18F-FET is probably the closest to
18
F-FAMT in terms of chemical compound, radiochemistry, and clinical applicability. While 18F-FET has higher
diagnostic accuracy than 18F-FDG, its effectiveness is
limited for brain tumors [33]. L-[methyl-11C]-methionine
(11C-MET), the most popular amino acid-based PET radiotracer to date, also has excellent diagnostic accuracy
for glioma compared to 18F-FDG [34]. However, both
18
F-FET and 11C-MET are also substrates for LAT2
transporters, which is also expressed in normal cells [14,
35]. The low kidney uptake PET tracer anti-1-amino-318
F-fluorocyclobutane-1-carboxylic acid (18F-FACBC)
has recently been meta-analyzed for its accuracy in prostate cancer recurrence detection. However, the specificity of 18F-FACBC is lower than 11C-choline PET and
even T2-weighted MRI [36]. Therefore, 18F-FAMT probably the most versatile oncologic PET radiotracer
currently available.
However, there a few limitations in this study and also
in 18F-FAMT itself. First, all studies were from a single
institution, which was potentially affected by publication
bias despite the authors of each study belonging to various departments and evaluating different types of tumors. Even though studies by Watanabe et al. and Tian
et al. focused on musculoskeletal tumors, they were
separated by more than a decade, eliminating the possibility of overlapping patients [12, 23]. A study of various
tumors by Inoue et al., however, included two patients
with chondrosarcoma and schwannoma that might also
be involved in the Watanabe et al. study, since these studies were from the same period [5, 12]. Unfortunately, this
is difficult to confirm. Second, not all types of malignancies were evaluated; in particular, lymphoma, melanoma,
pancreas and thyroid cancer, which are tumor types for
Achmad et al. BMC Medical Imaging (2017) 17:66
which 18F-FDG PET is recommended to improve diagnostic accuracy [3]. Tumors in the pelvic area and abdomen
were also poorly represented in this study.
Another drawback of the current 18F-FAMT studies is
the absence of dynamic PET data. Currently 18F-FAMT
PET scan is performed at 40–60 min post injection.
However, phases as early as 5–15 min post injection
might show higher tumor detection accuracy for any
amino acid PET tracer considering the two-waydirectional characteristic of amino acid uptake by their
transporters [37]. A dynamic 18F-FAMT PET study in an
animal tumor model showed that tumor-to-muscle uptake ratio is highest at 20 min and remains high at
60 min [38]. However, clinical dynamic PET studies are
necessary to obtain optimal scan times.
Our current findings emphasize the need for prospective multicenter studies to overcome limitations of the single center report. This can only be achieved when the
18
F-FAMT synthesis method is optimized and becomes
widely used. The current 18F-FAMT radiofluorination
method yields a low radioactivity that is only enough for
PET scans for a mere three to four patients in each radiosynthesis [39]. Recently, a modified method of 18F-FAMT
synthesis allows production to achieve high radioactivity
for routine use [40]. However, a more practical approach
is warranted. The twenty years of anticipation might soon
be realized with the recent rapid development of fluorination methods. Of particular interest are the so-called
late-stage fluorination methods which allow optimized
synthesis of previously inaccessible PET radiotracers [41].
These novel radiofluorination approaches which make
possible large-scale synthesis allow reconsideration of
promising but underutilized radiotracers, like 18F-FAMT.
Hence, revisiting the diagnostic performance of
18
F-FAMT is a major step in the quest for an ideal general
oncology PET tracer. Once these impediments are resolved, which we foresee shortly, the future may bring increased clinical impact of 18F-FAMT in oncology.
Page 8 of 9
tyrosine; DOR: Diagnostic odds ratio; LAT(1–4): L-type amino acid transporter
(1–4); PET: Positron emission tomography; QUADAS: Quality Assessment of
Diagnostic Accuracy Studies; SPECT: Single-photon emission computed
tomography; SROC: Summary receiver operating characteristics;
SUVmax: Maximum standardized uptake value
Acknowledgements
We are very grateful to Dr. Ayako Taketomi-Takahashi for her assistance in
improving the English in this manuscript.
Funding
Not applicable.
Availability of data and materials
All the data is contained within the manuscript.
Authors’ contributions
AA designed the study, selected articles for inclusion, performed data extraction,
assessed study quality, undertook meta-analysis, and wrote the paper. AB and RY
selected articles for inclusion, assisted with data extraction, and assessed study
quality. YDH performed additional statistical analysis. TH and YT edited and
reviewed the final manuscript. All authors contributed during manuscript
preparation, and read and approved the final manuscript.
Ethics approval and consent to participate
All studies included in this meta-analysis had been approved by their institutional
review boards.
Consent for publication
Not applicable.
Competing interests
Authors are educational staffs (AA, TH, YT) and graduate students (AB, RY, YDH) at
the academic institution where the 18F-FAMT is developed. None of the authors
are among the first authors of the studies included in this meta-analysis.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Author details
Department of Diagnostic Radiology and Nuclear Medicine, Gunma
University Graduate School of Medicine, 3-39-22 Showa-machi, Maebashi,
Gunma 371-8511, Japan. 2Department of Nuclear Medicine and Molecular
Imaging, Faculty of Medicine, Padjadjaran University, Jl. Professor Eyckman
No.38, Bandung, West Java 40161, Indonesia. 3Department of Nuclear
Medicine, Mochtar Riady Comprehensive Cancer Center, Jl. Garnisun Dalam
No. 2–3, Semanggi, Jakarta 12930, Indonesia.
1
Received: 3 June 2017 Accepted: 13 December 2017
Conclusion
F-FAMT has diagnostic performance equal to or perhaps even better than 18F-FDG for malignancy detection
in several cancer types. Future development in
18
F-FAMT radiosynthesis might allow this tracer to be
evaluated in other tumor types.
18
Additional file
Additional file 1: Table S1. QUADAS tool assessment results (n = 9).
Table S2. Diagnostic test results (DOCX 26 kb)
Abbreviations
11
C-MET: L-[methyl-11C] methionine; 18F-FACBC: anti-1-amino-318
F-fluorocyclobutane-1-carboxylic acid; 18F-FAMT: L-3-18F-α-methyl tyrosine;
18
F-FDG: 2-deoxy-2-[18F]fluoro-D-glucose; 18F-FET: O-(2-[18F]fluoroethyl)-L-
References
1. Farwell MD, Pryma DA, Mankoff DA. PET/CT imaging in cancer: current
applications and future directions. Cancer. 2014;120:3433–45.
2. Gillies RJ, Robey I, Gatenby RA. Causes and consequences of increased
glucose metabolism of cancers. J Nucl Med. 2008;49(Suppl 2):24S–42S.
3. Fletcher JW, Djulbegovic B, Soares HP, Siegel BA, Lowe VJ, Lyman GH, et al.
Recommendations on the use of F-18-FDG PET in oncology. J Nucl Med.
2008;49:480–508.
4. Huang C, McConathy J. Radiolabeled amino acids for oncologic imaging. J
Nucl Med. 2013;54:1007–10.
5. Inoue T, Koyama K, Oriuchi N, Alyafei S, Yuan Z, Suzuki H, et al. Detection of
malignant tumors: whole-body PET with fluorine 18 alpha-methyl tyrosine
versus FDG - preliminary study. Radiology. 2001;220:54–62.
6. Inoue T, Shibasaki T, Oriuchi N, Aoyagi K, Tomiyoshi K, Amano S, et al. F-18
alpha-methyl tyrosine PET studies in patients with brain tumors. J Nucl Med.
1999;40:399–405.
7. Kaira K, Oriuchi N, Otani Y, Yanagitani N, Sunaga N, Hisada T, et al.
Diagnostic usefulness of fluorine-18-alpha-methyltyrosine positron emission
Achmad et al. BMC Medical Imaging (2017) 17:66
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
tomography in combination with F-18-fluorodeoxyglucose in sarcoidosis
patients. Chest. 2007;131:1019–27.
Kaira K, Oriuchi N, Shimizu K, Tominaga H, Yanagitani N, Sunaga N, et al. F18-FMT uptake seen within primary cancer on PET helps predict outcome
of non-small cell lung cancer. J Nucl Med. 2009;50:1770–6.
Miyakubo M, Oriuchi N, Tsushima Y, Higuchi T, Koyama K, Arai K, et al.
Diagnosis of maxillofacial tumor with L-3-[F-18]-fluoro-alpha-methyltyrosine
(FMT) PET: a comparative study with FDG-PET. Ann Nucl Med. 2007;21:129–35.
Sohda M, Sakai M, Honjyo H, Hara K, Ozawa D, Suzuki S, et al. Use of pretreatment F-18-FAMT PET to predict patient survival in Squamous cell
carcinoma of the esophagus treated by curative surgery. Anticancer Res.
2014;34(7):3623–8.
Suzuki R, Watanabe H, Yanagawa T, Sato J, Shinozaki T, Suzuki H, et al. PET
evaluation of fatty tumors in the extremity: possibility of using the
standardized uptake value (SUV) to differentiate benign tumors from
liposarcoma. Ann Nucl Med. 2005;19:661–70.
Watanabe H, Inoue T, Shinozaki T, Yanagawa T, Ahmed AR, Tomiyoshi K,
et al. PET imaging of musculoskeletal tumours with fluorine-18 alphamethyltyrosine: comparison with fluorine-18 fluorodeoxyglucose PET. Eur J
Nucl Med. 2000;27:1509–17.
Sato N, Inoue T, Tomiyoshi K, Aoki J, Oriuchi N, Takahashi A, et al.
Gliomatosis cerebri evaluated by F-18 alpha-methyl tyrosine positronemission tomography. Neuroradiology. 2003;45:700–7.
Wiriyasermkul P, Nagamori S, Tominaga H, Oriuchi N, Kaira K, Nakao H, et al.
Transport of 3-Fluoro-L-alpha-methyl-tyrosine by tumor-Upregulated L-type
amino acid transporter 1: a cause of the tumor uptake in PET. J Nucl Med.
2012;53:1253–61.
Wei L, Tominaga H, Ohgaki R, Wiriyasermkul P, Hagiwara K, Okuda S,
et al. Specific transport of 3-fluoro-l-α-methyl-tyrosine by LAT1 explains
its specificity to malignant tumors in imaging. Cancer Sci. 2016;107:
347–52.
Reitsma JB, Moons KGM, Bossuyt PMM, Linnet K. Systematic reviews of
studies quantifying the accuracy of diagnostic tests and markers. Clin Chem.
2012;58:1534–45.
Macaskill P, Gatsonis C, Deeks J, Harbord R, Takwoingi Y. Chapter 10:
Analysing and Presenting Results. In: Cochrane Handbook for Systematic
Reviews of Diagnostic Test Accuracy Version 1. Deeks J, Bossuyt P, Gatsonis
C, editors. The Cochrane Collaboration. 2010. http://methods.cochrane.org/
sdt/handbook-dta-reviews. Accessed 15 Dec 2015.
Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al.
QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy
studies. Ann Intern Med. 2011;155:529–36.
Rücker G, Schumacher M. Summary ROC curve based on a weighted
Youden index for selecting an optimal cutpoint in meta-analysis of
diagnostic accuracy. Stat Med. 2010;29:3069–78.
Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PMM. The diagnostic odds
ratio: a single indicator of test performance. J Clin Epidemiol. 2003;56:1129–35.
Doebler P. Mada: meta-analysis of diagnostic accuracy. R package version 0.
5.7. 2015.
R-Core-Team. R: a language and environment for statistical computing.
Vienna: R Foundation for Statistical Computing; 2015.
Tian M, Zhang H, Endo K. Comparison of cell proliferation, protein, and
glucose metabolism in musculoskeletal tumors in a PET study. J Biomed
Biotechnol. 2011; 10.1155/2011/807929.
Kaira K, Oriuchi N, Shimizu K, Ishikita T, Higuchi T, Imai H, et al. Evaluation of
thoracic tumors with F-18-FMT and F-18-FDG PET-CT: a clinicopathological
study. Int J Cancer. 2009;124:1152–60.
Lijmer JG, Bossuyt PMM, Heisterkamp SH. Exploring sources of heterogeneity
in systematic reviews of diagnostic tests. Stat Med. 2002;21:1525–37.
Bürkner P-C, Doebler P. Testing for publication bias in diagnostic metaanalysis: a simulation study. Stat Med. 2014;33:3061–77.
Takwoingi Y, Guo B, Riley RD, Deeks JJ. Performance of methods for metaanalysis of diagnostic test accuracy with few studies or sparse data. Stat
Methods Med Res. 2015; 10.1177/0962280215592269.
Deppen SA, Blume JD, Kensinger CD, Morgan AM, Aldrich MC, Massion PP,
et al. Accuracy of FDG-PET to diagnose lung cancer in areas with infectious
lung disease a meta-analysis. JAMA. 2014;312:1227–36.
Porcel JM, Hernandez P, Martinez-Alonso M, Bielsa S, Salud A. Accuracy of
Fluorodeoxyglucose-PET imaging for differentiating benign from malignant
pleural effusions a meta-analysis. Chest. 2015;147:502–12.
Page 9 of 9
30. Xu G, Zhao L, He Z. Performance of whole-body PET/CT for the detection of
distant malignancies in various cancers: a systematic review and metaanalysis. J Nucl Med. 2012;53:1847–54.
31. Corrigan AJG, Schleyer PJ, Cook GJ. Pitfalls and artifacts in the use of PET/CT
in oncology imaging. Semin Nucl Med. 2015;45:481–99.
32. Rice SL, Roney CA, Daumar P, Lewis JS. The next generation of positron
emission tomography radiopharmaceuticals in oncology. Semin Nucl Med.
2011;41:265–82.
33. Dunet V, Rossier C, Buck A, Stupp R, Prior JO. Performance of F-18-Fluoroethyl-tyrosine (F-18-FET) PET for the differential diagnosis of primary brain
tumor: a systematic review and Metaanalysis. J Nucl Med. 2012;53:207–14.
34. Zhao C, Zhang Y, Wang J. A meta-analysis on the diagnostic performance
of 18F-FDG and 11C-methionine PET for differentiating brain tumors. AJNR
Am J Neuroradiol. 2014;35:1058–65.
35. Habermeier A, Graf J, Sandhofer BF, Boissel JP, Roesch F, Closs EI. System L
amino acid transporter LAT1 accumulates O-(2-fluoroethyl)-L-tyrosine (FET).
Amino Acids. 2015;47:335–44.
36. Ren J, Yuan L, Wen G, Yang J. The value of anti-1-amino-3-18Ffluorocyclobutane-1-carboxylic acid PET/CT in the diagnosis of recurrent
prostate carcinoma: a meta-analysis. Acta Radiol. 2016;57:487–93.
37. Kameyama M, Umeda-Kameyama Y. Strategy based on kinetics of O-(2-[18F]
fluoroethyl)-L-tyrosine ([18F] FET). Eur J Nucl Med Mol Imaging. 2016;43:
2267–8.
38. Yamaguchi A, Hanaoka H, Fujisawa Y, Zhao S, Suzue K, Morita A, et al.
Differentiation of malignant tumours from granuloma by using dynamic
[18F]-fluoro-L-α-methyltyrosine positron emission tomography. EJNMMI Res.
2015;5:29.
39. Tomiyoshi K, Amed K, Muhammad S, Higuchi T, Inoue T, Endo K, et al.
Synthesis of isomers of 18F-labelled amino acid radiopharmaceutical:
position 2- and 3-L-18F-α-methyltyrosine using a separation and purification
system. Nucl Med Commun. 1997;18:169–75.
40. Meleán JC, Humpert S, Ermert J, Coenen HH. Stereoselective radiosynthesis
of L- and D-3-[F-18]fluoro-alpha-methyltyrosine. J Fluor Chem. 2015;178:
202–7.
41. Preshlock S, Tredwell M, Gouverneur V. 18F-labeling of Arenes and
Heteroarenes for applications in positron emission tomography. Chem Rev.
2016;116:719–66.
Submit your next manuscript to BioMed Central
and we will help you at every step:
• We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit