The added diagnostic value of complementary gadoxetic acid-enhanced MRI to ¹⁸F-DOPA-PET/CT for liver staging in medullary thyroid carcinoma

Medullary thyroid carcinoma (MTC) accounts for 1–2% of all thyroid malignancies and causes up to 13% of all thyroid disease-related deaths [1, 2]. Sporadic occurrence encompasses 75% of all MTC cases while the remaining 25% are associated with hereditary tumour syndromes (e.g., multiple endocrine neoplasia (MEN) 2A and 2B). MTC-related lethality is mostly due to distant metastases and the median 10-year survival upon advanced stages of the disease is reported to be 10% [3, 4]. 13 to 15% of patients present with distant metastases at the time of diagnosis [5]. Medullary thyroid carcinoma is a malignant neuroendocrine tumour with the capability to take up amine precursors, such as dopamine, for decarboxylation (Amine Precursor Uptake and Decarboxylation system), thereby allowing the use of ¹⁸F-DOPA as a radiotracer for the detection of metastases. Particularly ¹⁸F-DOPA-PET/CT has been recognized as a highly sensitive and specific imaging modality for the detection of metastatic MTC [6, 7, 8, 9, 10].

The liver is the most frequently affected organ, with liver metastases in 45% of patients with advanced MTC [1]. However, liver staging in MTC remains challenging, as small metastases may remain ¹⁸F-DOPA-negative. A timely and comprehensive liver staging is of major importance to evaluate potential treatment options with a growing oncological toolbox including local ablative treatment, surgical, or systemic therapy [11, 12, 13, 14, 15, 16, 17, 18, 19]. Recent studies have shown that contrast-enhanced liver magnetic resonance imaging (MRI) is best suited for the detection of malignant liver lesions, particularly small metastases < 1.0 cm [20, 21, 22, 23, 24]. Gadoxetic acid is a contrast medium which specifically distributes into hepatocytes and the biliary tract system in a late, hepatobiliary phase. This allows for a differentiation of hepatocytes from neoplastic cells, which do not show a gadoxetic acid storage, thereby rendering gadoxetic acid a valuable contrast agent in patients with suspected hepatocellular carcinoma [25, 26, 27] or suspected liver metastases [28, 29, 30]. However, the value of contrast-enhanced liver MRI in patients with metastatic MTC was not yet investigated and the current guidelines do not recommend gadoxetic acid-enhanced MRI as routine liver staging in MTC. Therefore, the aim of the present study was to investigate the added diagnostic value of complementary gadoxetic acid-enhanced liver MRI to ¹⁸F-DOPA-PET/CT for liver staging in MTC.

This retrospective study was approved by the Institutional Review Board and the requirement for informed consent was waived. Written informed consent for the diagnostic ¹⁸F-DOPA PET/CT scan and the contrast-enhanced MRI scan was obtained from all patients prior to the examination.

Study population

Patients with histologically confirmed MTC who underwent gadoxetic acid-enhanced liver MRI and ¹⁸F-DOPA-PET/CT for whole-body tumour staging within 30 days between 2010 and 2016 were included in the analysis. Detailed inclusion and exclusion criteria are provided in Table 1, (Fig. 1) visualises the process of patient selection.

Table 1

Inclusion and exclusion criteria

Inclusion criteria	Exclusion criteria
Histologically confirmed MTC (surgical thyroidectomy or fine needle aspiration)	Unenhanced acquisition of the MRI scan (due to contraindications to gadolinium-based contrast agents, such as acute or chronic renal failure with estimated glomerular filtration rate < 30 mL/min or known hypersensitivity)
Whole-body 18F-DOPA PET/CT between 2010 and 2016	MRI contrast agent other than gadoxetic acid
Matching gadoxetic acid-enhanced liver MRI within 30 days	Unenhanced CT (due to contraindications for iodinated contrast agents, such as manifest hyperthyroidism, acute or chronic renal failure with estimated glomerular filtration rate < 30 mL/min or known hypersensitivity)
	Missing or inconclusive histological results confirming the diagnosis of MTC

Open image in new window

In total, 207 liver lesions were detected and classified in 36 patients. The number of lesions detected with ¹⁸F-DOPA-PET/CT and gadoxetic acid-enhanced MRI is shown in Table 4. ¹⁸F-DOPA-PET/CT detected 149 lesions (72% of all lesions, 149/207). Six lesions (3%, 6/207) remained category 3 lesions based on the ¹⁸F-DOPA-PET/CT and could neither be classified as benign nor as malignant. In contrast, MRI detected 207 lesions. All lesions could be characterized as benign (category 1 or 2) or malignant (category 4 or 5). Differences between the number of detected lesions with ¹⁸F-DOPA-PET/CT and gadoxetic acid-enhanced MRI can be seen in (Fig. 2). MRI detected significantly more lesions than ¹⁸F-DOPA-PET/CT (p < 0.001) and significantly more lesions could be categorised as metastases (category 4 and 5, p = 0.001). MRI allowed for a definite lesion classification (category 1 or 5) in 92% of all lesions (190/207) with 51 category 1 lesions (25%, 51/207) and 139 category 5 lesions (67%, 139/207). In contrast, ¹⁸F-DOPA-PET/CT was not as conclusive in lesion categorisation as gadoxetic acid-enhanced MRI. Of the 149 lesions detected by ¹⁸F-DOPA-PET/CT, 76% could be classified as category 1 or 5 lesions (113/149). Figures 3, 4, and 5 show exemplary cases in which gadoxetic acid-enhanced MRI provided added diagnostic value. MRI detected significantly more category 5 lesions than ¹⁸F-DOPA-PET/CT (p < 0.001) (Fig. 6).

Table 4

Absolute liver lesion counts for ¹⁸F-DOPA-PET/CT and gadoxetic acid-enhanced MR

	Number of lesions
	18F-DOPA-PET/CT	Gadoxetic acid-enhanced MRI	Additional lesions detected by MRI
All lesion categories	149	207	58
Metastases (category 4, 5)	95	152	54
Cysts	33	37	1
Hemangiomas	15	18	3
Category 1	34	51
Category 2	14	4
Category 3	6	0
Category 4	16	13
Category 5	79	139
	Change in lesion category
	18F-DOPA-PET/CT-related	Gadoxetic-acid enhanced MRI-related
Category 2 to 1	0	10
Category 3 to 1	0	3
Category 3 to 5	0	3
Category 4 to 5	0	5

Open image in new window

Open image in new window

Open image in new window

Open image in new window

Open image in new window

Overall, gadoxetic acid-enhanced MRI detected 58 additional lesions. 54 (93%) of these ¹⁸F-DOPA-PET/CT-occult lesions were categorized as metastases. The additionally detected metastases were small (0.5 cm, interquartile range 0.4–0.7 cm).

On MRI, no definite lesion characterisation (categories 2–4) was rare (8% of all lesions, 17/207) whereas ¹⁸F-DOPA-PET/CT did not allow for a definite lesion categorisation in 36 lesions (24% of all lesions detected by ¹⁸F-DOPA-PET/CT, 36/149). Changes in lesion categorisation based on the MRI scan can be obtained from Table 4. Of note, all liver metastases could successfully be diagnosed using the multiphase contrast-enhanced sequences, the T2-weighted sequences, and the DWI. The hepatobiliary phase did not lead to the detection of additional metastases but added significantly to the level of confidence in lesion characterisation.

Of note, one patient in the investigated cohort did not show any metastatic lesions on ¹⁸F-DOPA-PET/CT but one metastasis (category 5 lesion) on multiparametric liver MRI. Of the 152 metastases detected on multiparametric MRI, 84 (55% of all detected metastases) demonstrated a diffusion restriction. However, a significant proportion (56/152, 37%) of the detected metastases demonstrated a brisk arterial rim enhancement and a cystic centre, resulting in T2-shine through (Fig. 7). ADC quantification was possible for 28 lesions with a diffusion restriction and clear, definable correlate on ADC maps. The median value of these lesions was 0.778 × 10^− 3 mm²/s with an interquartile range of 0.723 to 0.912 × 10^− 3 mm²/s. For a major proportion (56/84, 67%) of liver lesions, a clear delineation on ADC was not possible due to small lesion size and artifacts (e. g., subcapsular localisation resulting in susceptibility artifacts).

Open image in new window

The median follow-up interval for combined ¹⁸F-DOPA-PET/CT and MRI was 484 days with an interquartile range of 250 to 769 days.

In the present study, we investigated the added diagnostic value of gadoxetic acid-enhanced MRI to ¹⁸F-DOPA-PET/CT for liver staging in patients with MTC. Complementary gadoxetic acid-enhanced MRI allowed for the detection of significantly more lesions compared to ¹⁸F-DOPA-PET/CT alone. In addition, liver lesion categorisation as benign or malignant was more conclusive at a significantly higher frequency, especially for small lesions with a diameter < 1 cm.

Our results are supported by a study that investigated ultrasound, CT, whole-body and liver MRI, and ¹⁸F-FDG-PET/CT for the detection of tumour recurrence and metastases in MTC patients with elevated serum calcitonin levels after initial treatment [32]. They found MRI to detect more liver metastases than the other imaging modalities, with MRI being the only modality to detect very small metastases (millimeter range) in two patients. The combination of MRI and CT detected significantly more metastases than ¹⁸F-FDG-PET/CT. However, the authors do not detail on the type of intravenous contrast agent and the comprehensive liver MRI protocol did not include DWI sequences. Our study extends the literature as it provides evidence for the use of multi-phase liver MRI with an intravenous, hepatobiliary contrast agent for liver staging in MTC including a late, hepatobiliary phase and diffusion-weighted images for the detection of liver metastases and for appropriate characterisation of liver lesions.

In line with our results, previous studies have shown a significant benefit of multi-phase contrast enhanced MRI for the detection of metastatic liver disease. In a head-to-head comparison of somatostatin receptor scintigraphy, multi-phase CT, and MRI in patients with neuroendocrine tumours for liver staging, MRI with intravenous Gd-DOTA administration allowed for the detection of significantly more metastases than CT and scintigraphy [33]. In addition, it was shown, that for the detection of liver metastases in patients with neuroendocrine tumours, the use of liver MRI using hepatobiliary contrast agent gadoxetic acid yields significant additional diagnostic value to nuclear imaging techniques [22]. Recent studies have also demonstrated that gadoxetic acid-enhanced MRI is superior for the detection of liver metastases to MRI with non-specific contrast agent or CT with contrast enhancement [29, 34]. In our cohort, DWI was an important co-feature for metastasis detection and characterisation. However, a significant proportion of MTC liver metastases showed a brisk arterial rim enhancement and a cystic centre, resulting in T2-shine through. As a consequence, our results suggest that DWI alone is not sufficient for liver metastasis detection and characterisation in MTC. Therefore, MTC liver staging should be performed based on all sequences of the comprehensive liver MRI protocol.

In our institution, ¹⁸F-DOPA-PET/CT is performed for whole-body tumour staging in MTC since numerous studies demonstrated superiority to ¹⁸F-FDG hybrid imaging [6, 8, 9, 10, 35]. The detection of liver metastases in MTC patients, however, is difficult due to the typically small size with correspondingly low ¹⁸F-DOPA uptake or size below the spatial resolution of PET imaging. Taken together, there is no existing consensus on imaging modalities for liver staging in patients with MTC. Our results strongly suggest that hepatobiliary contrast-enhanced liver MRI should be performed as routine part of whole-body tumour staging protocols in MTC. The hepatobiliary phase enabled by the use of the hepatocyte-specific contrast agent added significantly to the level of confidence in lesion characterisation. Future studies should further investigate the impact of multiphase liver MRI on clinical patient management, with regard to the various treatment options for metastatic liver disease as well as clinical endpoints such as progression-free and overall patient survival. Even in patients with multifocal liver metastases, the exact number and localisation of individual metastases may play a major role for therapy guidance, for instance to determine resectablility or eligibility for locoregional treatment options.

We acknowledge several limitations of the study. First, MRI was performed on two different 1.5 Tesla MRI scanners. Although imaging protocols were standardised, slight differences in image acquisition with potential impact on image reading cannot be fully excluded. Second, ¹⁸F-DOPA-PET/CT and gadoxetic acid enhanced MRI were performed within 30 days. Despite this relatively short interval, it is possible that patients with highly aggressive tumours experienced rapid tumor progression with new liver metastases between the two scans. However, the actual median time interval between gadoxetic acid-enhanced MRI and ¹⁸F-DOPA-PET/CT was 3 days (interquartile range 0–12 days), making short-term tumour progression unlikely. Third, a reading bias may be present since reading of one modality potentially influences the sensitivity of the reader for lesion detection on images of the other modality. To limit this bias, the reading order was randomised. Fourth, follow-up imaging was used as reference standard as no patient from the investigated population underwent liver biopsy. Fifth, the CT part of the ¹⁸F-DOPA-PET/CT scan was only performed in portal venous phase. Appending an unenhanced, an arterial, and a late dynamic phase may further improve the diagnostic performance of ¹⁸F-DOPA-PET/CT for the detection of (hypervascular) liver metastases. However, acquisition of additional CT phases increases radiation exposure and potentially requires a higher injection volume of iodinated contrast agent, which limits the applicability especially in young patients and chronic kidney disease.

Gadoxetic acid-enhanced liver MRI significantly increases the detection rate for liver metastases in patients with MTC, particularly of small lesions with a diameter < 1 cm. In addition, gadoxetic acid-enhanced MRI allows for a definite lesion categorisation when ¹⁸F-DOPA-PET/CT remains inconclusive, with potential impact on clinical patient management and therapy guidance. Our results provide evidence for the routine use of gadoxetic acid-enhanced MRI as part of comprehensive staging protocols in MTC patients.