Results and Discussion
Literature search
The literature search identified 4336 articles possibly fulfilling the inclusion criteria
([Fig. 1]). After the screening of titles and abstracts, 4254 articles were excluded because
of non-TSA-related topics. The remaining 82 articles were evaluated by reading the
full text. Of these, 62 articles were excluded because they did not focus on TSA and
the underlying risk factors. The remaining 20 articles were included and are discussed
in the review. Of these, eleven articles were excluded because they did not fulfill
the inclusion criteria that would allow for a direct comparison. The remaining nine
articles that reported on frequencies of TSA and potential risk factors according
to the inclusion criteria were selected and included for direct comparison of their
results [6]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]. The frequencies of TSM and basic patient characteristics of these nine studies
are presented in [Table 1].
Table 1
Frequencies of TSM and basic patient characteristics in selected studies.
Tab. 1 Häufigkeit der TSA in ausgewählten Studien und wesentliche Charakteristika der eingeschlossenen
Patienten.
|
author
|
Davenport et al.
|
Pietryga et al.
|
Davenport et al.
|
Kim et al.
|
Hayashi et al.
|
Bashir et al.
|
Motosugi et al.
|
Shah et al.
|
Well et al.
|
|
site A
|
site B
|
|
frequency of TSM
|
17/99 (17 %)
|
37/345 (11 %)
|
67/559 (12 %)
|
46/357 (13 %)
|
22/458 (5 %)
|
14/170[1] (8 %)
|
32/146 (22 %)
|
17/130 (13 %)
|
15/200 (8 %)
|
19/89 (21 %)
|
|
publication date
|
2013
|
2014
|
2014
|
2015
|
2015
|
2015
|
2016
|
2017
|
2017
|
|
journal
|
Radiology
|
Radiology
|
AJR
|
AJR
|
Radiology
|
Radiology
|
Radiology
|
Radiology
|
Clin Imaging
|
RöFo
|
|
country
|
USA
|
USA
|
USA
|
USA
|
Japan
|
USA
|
USA
|
Japan
|
USA
|
Germany
|
|
mean age
|
56 male
|
58 female
|
55 ± 14
|
56 ± 0.55
|
64 ± 7 TSM
|
62 ± 8 no TSM
|
61 ± 1 TSM
|
63 ± 2 no TSM
|
57 ± 12
|
52 ± 15
|
52 ± 15
|
57 ± 13
|
55 ± 15
|
|
mean BMI
|
29
|
28.3 ± 7
|
29 ± 0.3
|
27 ± 5
|
29 ± 6
|
24.3 ± 3.9
|
23.4 ± 3.9
|
29 ± 7
|
n/a
|
n/a
|
29 ± 7
|
26 ± 5
|
1 occurrence of TSM at initial MR imaging.
Frequency of TSM
In the first report on TSM in 2013 by Davenport et al., the frequency of severely
degraded arterial phase image quality after injection of gadoxetate disodium was described
as high as 17 %, compared to 2 % after injection of gadobenate disodium [6].
Initially, some radiologists doubted that the appearance of TSM is a reproducible
phenomenon [16]. Since 2013, our own and several other groups investigated the frequency of TSM:
except for one [17], all studies confirmed consistently that TSM occurs significantly more often after
injection of gadoxetate disodium than after injection of gadobenate disodium, or other
gadolinium-based contrast agents [6]
[7]
[8]
[10]
[11]
[13]
[14].
Interestingly, the results from different investigators show a wide range of frequencies
of TSM after injection of gadoxetate disodium. The highest frequency (22 %) was observed
by Motosugi et al. in the USA [13]. Hayashi et al. from Japan observed the lowest frequency of TSM (5 %) [11]. As shown in [Table 1], all other studies investigating TSM revealed frequencies over 10 %.
In summary, TSM occurs with a frequency of 5 – 22 % in patients undergoing gadoxetate
disodium-enhanced liver MRI.
Influence of contrast dose
One could assume that the dose of gadoxetate disodium influences the occurrence of
TSM. This assumption might explain the differences in observed frequencies of TSM.
Indeed, the study sites addressing TSM used different doses of gadoxetate disodium:
weight-adapted doses of 0.025 or 0.05 mmol/kg [11]
[13] fixed doses of 10 ml (package size) or 20 ml [6]
[8]
[10] ([Table 2]).
Table 2
Frequency of TSM and contrast injection protocols
Tab. 2 Häufigkeit der TSA und Kontrastmittel-Injektionsprotokolle.
|
author
|
Davenport et al.
|
Pietryga et al.
|
Davenport et al.
|
Kim et al.
|
Hayashi et al.
|
Bashir et al.
|
Motosugi et al.
|
Shah et al.
|
Well et al.
|
|
Site A
|
Site B
|
|
frequency of TSM
|
17/99 (17 %)
|
37/345 (11 %)
|
67/559 (12 %)
|
46/357 (13 %)
|
22/458 (5 %)
|
14/170 (8 %)
|
32/146 (22 %)
|
17/130 (13 %)
|
15/200 (8 %)
|
19/89 (21 %)
|
|
concentration of contrast agent (mmol/kg)
|
fixed
|
fixed
|
fixed
|
fixed
|
0.025
|
fixed
|
0.05
|
0.025
|
fixed
|
0.025
|
|
mean volume of contrast agent (ml)
|
10
|
10
|
10/20
|
10
|
n/a
|
10
|
n/a
|
n/a
|
10
|
8.1
|
|
injection rate (ml/s)
|
1 or 2
|
2
|
1 or 2
|
1
|
1
|
2
|
2
|
1
|
2
|
2
|
|
saline flush (ml)
|
equivalent to contrast agent
|
20
|
10/20
|
25
|
40
|
20
|
50
|
20
|
20 – 30
|
20
|
Many institutions use the recommended dose of 0.025 mmol/kg body weight for gadoxetate
disodium. This is only a quarter of the dose of conventional extracellular contrast
agents (0.1 mmol/kg) [18]
[19]. The weight-adjusted contrast volume of gadoxetate disodium (concentration: 0.25mol)
results in half the volume of other extracellular contrast agents (concentration:
0.5mol) [20]. The smaller volume shortens the duration of arterial enhancement and reduces signal
intensity [21]. Therefore, some institutions use higher doses of 0.05 mmol/kg [18]
[22]
[23]. Other institutions use a fixed dose of 10 ml equaling the package size, resulting
in varying concentrations of gadoxetate disodium, depending on the patient’s body
weight [6]
[8]
[9]
[10]
[12]
[14].
Davenport et al. demonstrated that gadoxetate disodium-associated TSM is more frequent
(15 %) after application of a fixed dose of 20 ml compared to a fixed dose of 10 ml
(10 %) [24]. Other studies, however, were not able to confirm a dose-dependent relationship
of gadoxetate disodium and TSM [8]
[10]
[12]. Also, no correlation between the dose of gadoxetate disodium and the frequency
of TSM can be derived when indirectly comparing results from different studies and
different sites ([Table 2]). For example, Hayashi et al. found the lowest frequency of TSM (5 %), while our
group found one of the highest frequencies (21 %), despite the fact that both groups
used the same weight-adjusted dose of 0.025 mmol/kg [11].
Taken together, a relation between the dose of gadoxetate disodium and the frequency
of TSM has been proposed, but only few studies support this hypothesis.
Influence of contrast injection rate
The injection rate of gadoxetate disodium is another factor potentially influencing
the occurrence of TSM. Indeed, the study sites addressing TSM used different injection
rates, which may explain the differences in observed frequencies of TSM.
Some institutions administer gadoxetate disodium with a high injection rate of 2 ml/s
to achieve a concentrated bolus for improved arterial phase contrast [6]
[8]
[9]
[12]
[13]. Other institutions administer it with a slower injection rate of 1 ml/s to spread
the contrast bolus, thereby reducing the risk of a poorly timed arterial phase [6]
[10]
[11]
[13]
[25]. Dilution of gadoxetate disodium with saline is an alternative approach to spread
the contrast bolus [9]. Another difference with regard to the injection of contrast agent is the amount
of the saline chaser ([Table 2]).
Kim et al. demonstrated that injection of 1:1 diluted gadoxetate disodium significantly
reduces overall and severe imaging artifacts in the arterial phase of contrast-enhanced
liver MRI [26]. Polanec et al. demonstrated that a slow injection rate of 1 ml/s in combination
with a 1:1 dilution of gadoxetate disodium can significantly reduce the frequency
of TSM without a relevant decrease in image quality compared to undiluted injection
protocols [27].
Both Hayashi et al. and Motosugi et al. applied the same amount of weight-adapted
gadoxetate disodium at a slow rate of 1 ml/s but observed different frequencies of
TSM (5 % vs. 13 %) [11]
[13]. Studies with faster injection rates of 2 ml/s also showed varying, but higher frequencies
of TSM, ranging from 10 % to 22 % with either weight-adjusted dosing or fixed doses
of 10 ml gadoxetate disodium ([Table 2]) [6]
[8]
[13]
.
Altogether, dilution of the contrast agent and slow injection rates seem to reduce
the frequency of TSM, but further studies are needed to confirm this observation.
Influence of arterial phase imaging sequence length
The length of the imaging sequence is another factor that potentially influences the
frequency of observed TSM in the arterial phase of gadoxetate disodium-enhanced liver
MRI.
This hypothesis is supported by a recent prospective multicenter placebo-controlled
study in healthy volunteers. The study revealed that the maximal breath-hold duration
is reduced after gadoxetate disodium administration, which in turn was associated
with motion artifacts in the arterial imaging phase [7]. Thus, an imaging sequence that requires a longer breath-hold may more likely be
affected by gadoxetate disodium-associated TSM.
Most of the published studies on TSM use three-dimensional, fat-suppressed, T1-weighted
spoiled gradient echo sequences for dynamic liver MRI [6]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[24] ([Table 3]). These sequences allow image acquisition during a single breath-hold with adequate
signal-to-noise ratio and spatial resolution [28]. The acquisition times for dynamic liver MRI sequences at the different study sites
ranged from 12 to 22 seconds [8]
[12]
[15], hence another potential explanation for the difference in observed frequencies
of TSM.
Table 3
Frequency of TSM and MR sequence acquisition protocols
Tab. 3 Häufigkeit der TSA und MR-Sequenzprotokolle.
|
author
|
Davenport et al.
|
Pietryga et al.
|
Davenport et al.
|
Kim et al.
|
Hayashi et al.
|
Bashir et al.
|
Motosugi et al.
|
Shah et al.
|
Well et al.
|
|
Site A
|
Site B
|
|
frequency of TSM
|
17/99 (17 %)
|
37/345 (11 %)
|
67/559 (12 %)
|
46/357 (13 %)
|
22/458 (5 %)
|
14/170 (8 %)
|
32/146 (22 %)
|
17/130 (13 %)
|
15/200 (8 %)
|
19/89 (21 %)
|
|
imaging sequence
|
3 D spoiled GRE
|
VIBE[1]
|
3 D spoiled GRE
|
GRAPPA[2]
|
GRAPPA2
|
3 D spoiled GRE
|
LAVA[3]
|
LAVA3
|
LAVA3/THRIVE[4]/mDixon[5]
|
eTHRIVE4/mDixon5
|
|
TR
|
3.6
|
3.7 – 4.4
|
3.6/3.7 – 4.4
|
3.8
4.3
|
3.98
|
3.7 – 4.4
|
3.7 – 4.1
|
3.4 – 4.7
|
4.2 – 4.5
3.0 – 4.2 3.4
|
3.9
3.5
|
|
TE
|
1.3
|
1.3 – 2.1
|
1.3
1.3 – 2.1
|
1.7
1.9
|
1.5
|
1.3 – 2.1
|
1.7 – 1.9
|
1.4 – 2.1
|
1.7 – 2.1 1.4 – 2.1 1.7
|
1.8 1.2/2.3
|
|
FA
|
12
|
9 – 12
|
12
9 – 12
|
12
9
|
12
|
9 – 12
|
12 – 15
|
12 – 15
|
12
10 – 70
15 – 32
|
10
|
|
FOV(mm)
|
entire liver
|
256 × 156 – 192
|
entire liver
|
370 × 300
|
entire liver
|
256 × 156 – 192
|
360 × 380
|
340 × 380
|
entire liver
|
330 × 330/400 × 400
|
|
section thickness (mm)
|
4
|
4
|
4
|
3
|
–
|
4
|
5/3.4
|
5/4
|
3 – 7
|
4.4/1.7
|
|
arterial phase acquisition time (s)
|
18 – 22
|
23
|
18 – 22
|
16
|
20
|
23
|
22/20
|
16
|
n/a
|
12 – 15
|
|
triggering method
|
manual fluoroscopic/automated bolus tracking
|
fixed delay
|
manual fluoroscopic/automatedbolus tracking
|
manual fluoroscopic trigger
|
manual fluoroscopic trigger
|
fixed delay/fluoroscopic trigger/automated bolus tracking
|
n/a
|
n/a
|
fixed delay/fluoroscopic trigger
|
fixed delay
|
1 VIBE = volume interpolated breath-hold examination.
2 GRAPPA = Generalized autocalibrating partially parallel acquisition.
3 LAVA = spectrally selective intermittent fat inversion.
4 eTHRIVE = Enhanced T1 High Resolution Isotropic Volume Excitation.
5 mDixon = multi-echo Dixon.
It seems conceivable that shorter scan times would decrease the frequency of TSM.
Indeed, Luetkens et al. used short acquisition times (i. e., breath-hold times) between
14 and 15 seconds and observed a reduced frequency (6 %) and severity of TSM compared
to other studies [17]. However, our group used an equally short imaging protocol (12 – 15 seconds) but
observed one of the highest frequencies of TSM (21 %) [15]. At the same time, Hayashi et al. used a rather long imaging protocol (20 s) and
observed the overall lowest frequency of TSM (5 %) ([Table 3]) [11].
Taken together, a relation between the duration of the arterial phase imaging sequence
and frequency of TSM has been proposed, but future prospective studies are needed
to confirm this hypothesis.
Potential individual risk factors
As outlined above, the techniques of contrast application and the length of imaging
protocols cannot fully explain the frequency of the occurrence of TSM after the injection
of gadoxetate disodium. A different explanation for the occurrence of TSM may therefore
be related to specific individual risk factors. Identifying such potential risk factors
is essential to develop strategies for avoiding TSM. Several groups investigated the
correlation of TSM with potential predisposing risk factors, such as sex, pulmonary
disease, body mass index (BMI), previous episodes of TSM, allergies, and other factors
([Table 4]) [6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[24]
.
Table 4
Frequency of TSM and potential risk factors
Tab. 4 Häufigkeit der TSA und potentielle ursächliche Risikofaktoren.
|
author
|
Davenport et al.
|
Pietryga et al.
|
Davenport et al.
|
Kim et al.
|
Hayashi et al.
|
Bashir et al.
|
Motosugi et al.
|
Shah et al.
|
Well et al.
|
|
Site A
|
Site B
|
|
frequency of TSM
|
17/99 (17 %)
|
37/345 (11 %)
|
67/559 (12 %)
|
46/357 (13 %)
|
22/458 (5 %)
|
14/170 (8 %)
|
32/146 (22 %)
|
17/130 (13 %)
|
15/200 (8 %)
|
19/89 (21 %)
|
|
age
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
old age[1]
,
[2]
|
n.s.
|
|
sex
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
male2
|
male2
|
female1, 2
|
n.s.
|
|
BMI
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
high BMI2
|
n.s.
|
high BMI2
|
high BMI2
|
n.s.
|
n.s.
|
|
ascites
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
|
pleural effusion
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
|
lung disease
|
n.s.
|
n.s.
|
COPD[3]
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
–
|
n.s.
|
|
previous exposure
|
–
|
–
|
–
|
n.s.
|
gadolinium exposure[4]
|
n.s.
|
–
|
–
|
–
|
n.s.
|
|
previous TSM
|
–
|
–
|
–
|
previous TSM2,
[5]
|
n.s.
|
previous TSM3
|
–
|
–
|
–
|
–
|
|
allergies
|
n.s.
|
–
|
n.s.
|
iodinated contrast4, 5
|
n.s.
|
–
|
n.s.
|
n.s.
|
–
|
n.s.
|
|
chronic liver disease
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
–
|
n.s.
|
|
hepatocellular carcinoma
|
n.s.
|
–
|
–
|
n.s.
|
–
|
–
|
–
|
–
|
–
|
n.s.
|
|
hypertonus
|
–
|
–
|
–
|
n.s.
|
–
|
–
|
n.s.
|
n.s.
|
–
|
n.s.
|
|
volume of contrast agent
|
n.s.
|
n.s.
|
20 ml4
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
|
injection rate
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s.
|
n.s. not significant.
1 associated with respiratory motion artifacts.
2 p ≤ 0.05.
3 p ≤ 0.0001.
4 p ≤ 0.01.
5 statistically significant only in univariate analysis.
With regard to gender, Motosugi et al. showed a predisposition towards male sex [29]. In contrast, Shah et al. proposed a correlation between TSM and female sex [14]. None of the other groups confirmed predisposition of either gender towards TSM
[6]
[10]
[11]
[12]
[14]
[15]. ([Table 4])
Concerning pulmonary disease, Davenport et al. revealed that patients with chronic
obstructive pulmonary disease have an increased risk of TSM after injection of gadoxetate
disodium [9]. Our group confirmed a trend toward significance in patients with pulmonary disease
for the occurrence of TSM [15]. Other studies, however, did not replicate this observation [10]
[13]. The results of McClellan et al. indicate that patients with low baseline breath-hold
duration may be the most affected by gadoxetate disodium-associated TSM [7]. None of the published studies prospectively assessed the effect of poor breath-hold
capability on frequency of TSM. Retrospective analyses of other risk factors, i. e.
pleural effusions, ascites, cardiac conditions that are potentially associated with
a reduction of breath-hold capability, showed no significant correlation with TSM
[10]
[11]
[12].
Motosugi et al. found a significantly higher risk of TSM in patients with a high BMI
[30]. Hayashi et al. confirmed that body weight contributes to the occurrence of TSM
[11]. One has to keep in mind that these results are biased by weight-adapted dosing
of gadoxetate disodium, resulting in higher contrast doses in patients with higher
BMI. None of the other groups confirmed predisposition of high BMI towards TSM [10].
Bashir et al. found a significantly higher risk of TSM in patients who had experienced
TSM before in previous liver MRI examination [12]. Kim et al. found that a known allergy towards iodinated contrast agents may be
a possible contributor for TSM. However, this relationship was identified only on
univariate analysis and could not be confirmed in a multivariate model [10].
Regarding age, Shah et al. found that the risk of arterial phase motion artifacts
increases with age [14], but none of the other groups confirmed a predisposition of higher age towards TSM
[15]
[24].
A large number of other factors, such as pleural effusions, ascites, liver cirrhosis,
model of end stage liver disease (MELD) score, and hepatocellular carcinoma have been
investigated as potential risk factors of TSM [8]
[9]
[12]. However, none of these potential risk factors showed a significant correlation
with the occurrence of TSM.
In summary, several potential predisposing individual risk factors for the appearance
of TSM have been described, but results are inconsistent. Conceivably, pulmonary disease
may represent a potential risk factor for TSM, but future prospective studies are
needed to confirm this hypothesis.
Effects of gadoxetic disodium injection on physiological parameters
As outlined above, no clear predisposing individual risk factor for the occurrence
of TSM has been identified. Understanding the mechanism of TSM is therefore another
essential strategy to avoid TSM. As a first step, several investigators assessed the
effects of gadoxetic disodium injection on physiological parameters. It should be
noted that breath-hold failure after injection of gadoxetate disodium is a self-terminating
effect with a duration of less than 60 – 90 seconds. After this interval, the acquisition
of the portal venous phase begins and is not affected by motion artifacts.
Several studies were able to show that injection of gadoxetate disodium leads to an
increased rate of breath-hold failures that are associated with TSM [13]
[30]. Moreover, McClellan et al. revealed that the maximal breath-hold duration is significantly
reduced in healthy volunteers after gadoxetate disodium administration. The reduction
in maximal achievable breath-hold duration was associated with TSM [7].
With regard to subjective dyspnea, several studies observed that subjective self-reported
dyspnea occurs significantly more frequently with gadoxetate disodium than with gadobenate
dimeglumine [6]
[13]
[31]. Interestingly, self-reported dyspnea shows no correlation with the occurrence of
TSM [13]. In contrast to the above-mentioned studies, Hayashi et al. did not detect an increase
in self-reported dyspnea after injection of gadoxetate disodium [11].
None of the studies revealed a significant decrease in blood oxygen saturation during
the arterial phase. Also, no significant correlation was identified between changes
in blood oxygen saturation and the occurrence of TSM [7]
[11]
[13]
[30], neither did any of the studies reveal a significant change in heart rate after
injection of gadoxetate disodium [7]
[11]
[13].
Taken together, impairment of breath-hold ability after injection of gadoxetate disodium
is a reproducible result that has been confirmed by several studies. In addition to
that, the impaired breath-hold ability is associated with TSM. However, the underlying
reason for breath-hold failure and consecutive TSM remains unclear.
Strategies to avoid TSM-associated arterial phase image degradation
As outlined above, breath-hold failure after injection of gadoxetate disodium has
been identified as the cause for TSM. As long as the underlying mechanism for the
breath-hold failure is not identified, TSM cannot be avoided in the arterial phase
of liver MRI. Therefore, technical strategies to avoid or mitigate the effects of
TSM on the arterial phase image quality need to be identified. Several investigators
explored the possibilities of modified imaging protocols to reduce the impact of TSM
on image degradation [8]
[32]
[33].
Pietryga et al. assessed fast, multi-arterial phase imaging in a single breath-hold
(23 seconds) that provides three arterial phase image sets with reduced or absent
motion artifacts [8]. With this approach, Pietryga et al. were able to recover up to 80 % of arterial
phases that would otherwise have been degraded in a standard imaging setting. However,
the technique results in a reduced signal-to-noise ratio and/or spatial resolution.
Recently, Rahimi et al. proposed a time-resolved interleaved variable density (IVD)
MR sequence for arterial phase imaging of the liver [34]. The IVD technique provides five arterial phase image sets with both high spatial
and temporal resolution and warrants further evaluation in this setting.
The effects of motion can also be reduced by using inherently motion-resistant MR
imaging techniques, such as radial or spiral k-space filling trajectories [35]
[36]. Another promising approach for improved arterial phase images is the utilization
of free-breathing techniques, which allow contrast-enhanced dynamic liver examinations
without the need for breath-holds [32]. Kaltenbach et al. demonstrated that free breathing sequences do not impair image
quality and can mitigate the effect of TSM [33].
An additional aspect that has to be considered is the proper preparation and training
of patients by the responsible technician. Gutzeit et al. were able to show that use
of an extended breathing command previous to and during the injection of gadoxetate
disodium significantly reduced the amount of breathing artifacts compared to examinations
with standard breathing commands [37]
[38]. A combination of both, motion-resistant imaging sequences and modified breathing
commands can further reduce the number of compromised arterial phases [39].
In summary, short acquisition times, multi-arterial phase imaging, alternative k-space
filling trajectories, free-breathing techniques, and breath-hold training warrant
further prospective studies to assess their potential in reducing TSM-associated arterial
phase image degradation.
Directions of future research
Based on the results presented above, we can assume that gadoxetate disodium injection
impairs the breath-hold ability, which in turn is associated with the occurrence of
TSM [7]
[13]. The underlying mechanism for the impaired breath-hold ability has not been identified.
Several associated risk factors for TSM have been described, such as high dose of
gadoxetate disodium, previous episodes of TSM, high BMI, and pulmonary disease [6]
[9]
[11]
[13]
[14]. However, results are inconsistent, and no unanimously validated risk factor could
be identified. Therefore, future studies are needed to improve the understanding of
this phenomenon as well as the diagnostic and clinical impact of TSM.
First, it is important to elucidate the underlying reason for the breath-hold failure
after injection of gadoxetate disodium. More research is needed regarding the effects
of gadoxetate disodium injection on cardiac and pulmonary function. This includes
determination of the exact time point of the onset of breath-hold failure after injection
of gadoxetate disodium, e. g. by fluoroscopic imaging of the lung during contrast
injection or continuous image acquisition during MRI examinations. The results might
explain the exact mechanism that triggers breath-hold failure and consecutive TSM.
Second, future studies also need to further investigate the previously assessed and
other, currently unknown, potential risk factors for TSM. Knowing such risk factors
may help to identify patients that should not undergo gadoxetate disodium-based, but
instead extracellular gadolinium-based contrast-enhanced liver MRI protocols.
Third, it is important to assess if and how TSM hampers the diagnostic accuracy of
the entire liver MRI examination [15]. Arterial enhancement is a cardinal imaging feature for diagnosis of HCC and other
arterially enhancing lesions [19]. Nevertheless, some of these lesions have other specific imaging features that are
displayed in other sequences of a comprehensive liver MRI examination, thereby potentially
providing a confident diagnosis despite TSA in the arterial imaging phase.
A thorough assessment of a diagnosis-limiting effect of TSM requires a prospective
study that ideally fulfills two major prerequisites: first, an intra-patient comparison
is required, with both, an extracellular gadolinium-based contrast agent and gadoxetate
disodium. Only this direct comparison will allow assessment of whether arterially
enhancing lesions (i. e. HCC) are detected with the extracellular contrast agent and
if these lesions are then missed or equally well detected with gadoxetate disodium
in cases of TSM. Second, all identified lesions require histopathological confirmation
or another robust reference standard for validation.
Since only 5 – 22 % of all patients experience TSM, large prospective studies are
required to answer the question of whether the overall diagnostic accuracy of gadoxetate
disodium-enhanced liver MRI is significantly hampered by TSM. Also, the patient collective
that may potentially be affected has to be clearly defined. Diagnostic accuracy in
patients with hypervascular lesions such as HCC, FNH, adenomas and hypervascular metastases
(renal carcinomas, neuroendocrine tumors) may be more affected, since arterial enhancement
is a cardinal imaging feature of these lesions, which might be hampered by TSM. Diagnostic
accuracy in patients with suspected colorectal liver metastases may not be affected
at all, since most sensitive detection of liver metastases relies on the hepatobiliary
phase of gadoxetate disodium, which is not affected by TSM [4]
[19].
