Extracardiac ¹⁸F-florbetapir imaging in patients with systemic amyloidosis: more than hearts and minds

Amyloidoses are a group of diseases in which misfolded proteins are deposited in tissues/organs in a highly aggregated form as amyloid fibrils, leading to multisystem disease. About 30 such proteins have been reported to cause human disease, with the two most common types of systemic amyloidosis being transthyretin amyloidosis (ATTR) and acquired monoclonal immunoglobulin light-chain amyloidosis (AL). Age-related ATTR (ATTRwt) occurs due to wild transthyretin protein deposition in the heart, and is an increasingly recognized entity in the elderly. Systemic AL occurs due to an underlying plasma cell dyscrasia secreting an unstable immunoglobulin free light chain. AL is the commonest type of amyloidosis. Appropriate treatment is dependent on accurate identification of the amyloid fibril type and assessment of the disease extent (particularly in the heart) to allow a risk adapted treatment strategy [1, 2].

Imaging of amyloid deposits is challenging and progress has been very slow. Since cardiac involvement is the main cause of morbidity and mortality, focus has been on cardiac imaging. Various modalities including echocardiography (the gold standard for many years), cardiac magnetic resonance imaging (an emerging gold standard) [3, 4], ^99mTc-labelled 3,3-diphosphono-1,2-propanodicarboxylic acid (^99mTc-DPD) scintigraphy and ^99mTc-labelled pyrophosphate (^99mTc-PYP) scintigraphy [5] for ATTR cardiac amyloid imaging have become part of routine practice. However, the disease burden in amyloidosis is systemic and imaging other organ systems has not generally been possible. Standard cross-sectional imaging is widely used to document enlargement of an organ, but is not specific for amyloidosis. At our centre (and at the University of Groningen), ¹²³I-labelled serum amyloid P component (¹²³I-SAP) scintigraphy [6] has been routinely used for visceral amyloid imaging. However, the technical complexity of this procedure has limited its wider use outside the current two centres. ¹²³I-SAP scintigraphy is very sensitive and specific for evaluating amyloid in large solid organs [7], but it is not reliable for detecting cardiac involvement. Although SAP uptake is occasionally evident at other sites, the clinical significance of this is not known.

More recently PET tracers used in the diagnosis of Alzheimer’s disease have been used to study cardiac involvement in patients with amyloidosis. The most widely studied tracer is ¹¹C-Pittsburgh compound B [8‐12], but there is also published literature on the use of ¹⁸F-florbetapir [13‐15] and ¹⁸F-florbetaben [16]. Despite the different tracers and limited numbers of participants in each of the studies, the technique shows promise and provides good separation between patients confirmed as having cardiac amyloidosis and controls. Although the primary aim of these studies was to study cardiac involvement, several of the papers discussed findings in soft tissue areas outside the heart. Dorbala et al. [13] demonstrated intense ¹⁸F-florbetapir lung uptake in a patient with clinical manifestations and CT findings of lung amyloid. Osborne et al. [15] generated activity curves over the dome of the liver and found that while the uptake in ATTR patients was similar to that in controls, uptake in patients with AL was significantly increased. Ezawa et al. [12] found many sites of extracardiac uptake and a good clinical correlation between gastric uptake and clinical gastric dysmotillity.

We conducted a study investigating the dynamic uptake of ¹⁸F-florbetapir centred on the heart and tracer uptake in the remainder of the body 1 h after injection of tracer. This report addresses the uptake of tracer outside the heart.

Seventeen patients were enrolled in the study, of whom 15 had AL and two had ATTR. Two patients were treatment-naive, one patient was scanned 2 days after the first cycle of chemotherapy and one patient was scanned 9 days after the first cycle of chemotherapy. The patient demographics are summarized in Table 1. Two patients had repeat scans following chemotherapy treatment (one was scanned just under 7 months after completion of chemotherapy and the other was still undergoing chemotherapy but had achieved a complete response). One patient underwent repeat scanning to assess the stability of unexpected lung uptake. All patients showed uptake in the heart as reported previously by others [13‐15] and these patients will the subject of a separate article.

All 17 patients showed uptake at one or more extracardiac sites. Representative late half-body images of areas of ¹⁸F-florbetapir uptake are shown as anterior maximum intensity projections with corresponding ¹²³I-SAP images for comparison in Fig. 1 and transaxial fused PET/CT images in Fig. 2. Physiological uptake can be seen in the hepatobiliary system in all patients. Common sites of uptake were fat, parotids, tongue, masticating muscles, spleen and bone marrow. Increased uptake in extracardiac organs seen on the late half-body images is summarized for each patient in Table 2 (along with respective uptake on ¹²³I-SAP scintigraphy uptake in liver, spleen, kidneys and bone marrow). Since ¹⁸F-florbetapir is excreted through the liver, biliary duct and small bowel, uptake in these sites is not included.

This is the first study to provide a comprehensive description and analysis of extracardiac sites of ¹⁸F-florbetapir uptake on PET in patients with proven systemic amyloidosis. The most striking and consistent finding was spleen uptake of ¹⁸F-florbetapir. This is particularly important, as ¹²³I-SAP scintigraphy is the only specific method for spleen imaging and 70% of AL patients have spleen uptake on ¹²³I-SAP imaging [6]. Using ¹²³I-SAP as the gold standard, on dynamic ¹⁸F-florbetapir with a SRI cut-off value of 0.045, the sensitivity and specificity were 100%. On late half-body images sensitivity was 66.7% and specificity was 100%. Importantly, there is no clinical splenic involvement in ATTRwt and there was no spleen uptake of ¹⁸F-florbetapir in the two patients with ATTRwt. The higher sensitivity and specificity of ¹⁸F-florbetapir on dynamic than on late imaging suggests that earlier imaging (around 15 min after injection) would be best for imaging ¹⁸F-florbetapir uptake in the spleen. Since splenic involvement is a hallmark of AL, the presence of both cardiac and splenic involvement on ¹⁸F-florbetapir imaging could potentially be used as an indicator of AL. The combination is generally not seen in other types of amyloidosis except rare hereditary apolipoprotein A1 amyloidosis.

The other striking finding was marked lung uptake in three patients. There is currently no specific imaging technique for detecting lung involvement in patients with systemic AL – diffuse lung involvement remains one of the few areas that are very poorly studied and understood in AL. One of the patients with abnormal lung uptake had interstitial shadowing on high-resolution lung CT images and was suspected to have lung involvement. The significance of lung uptake merits further specific study particularly in patients with proven pulmonary amyloidosis, since a limitation of our study was the lack of histological proof of lung involvement. The altered haemodynamics in cardiac AL also raises the question about poor tracer clearance from the lungs in patients with heart failure since there are no data on this in other settings of heart failure. In their recent study, Ezawa et al. [12] considered ¹¹C-PIB lung uptake to be physiological in both patients and controls. The pattern and intensity of uptake in their series was very different from the one in our patient cohort. In our series, uptake was particularly intense and diffuse in two patients and more moderate and heterogeneous in the third patient. None of the other patients demonstrated any lung uptake. The difference in our findings is probably due to the use of different tracers. Further studies investigating the ability of amyloid tracers to detect amyloidosis lung involvement are needed to ascertain their exact role in patient management.

Uptake in fat was identified in more than half the patients which is not dissimilar to the proportion of patients with AL in whom amyloid deposition can be confirmed in abdominal fat aspirates. The uptake of ¹⁸F-florbetapir in liver was poorly correlated with ¹²³I-SAP scintigraphy and there was limited kidney uptake, suggesting limited utility of ¹⁸F-florbetapir for liver or kidney imaging.

Our results indicate that ¹⁸F-florbetapir is excreted from the gastric wall into the gastric lumen in most patients and is then cleared into the small bowel but at variable rates (possibly influenced by fasting status and diet). Early gastric uptake and clearance is likely to be physiological since similar stomach activity was also seen in a dosimetry study by Joshi et al. [19]. However, the late stomach wall uptake, noted in 24% of patients, is likely to have been pathological. Diagnosis of gastrointestinal involvement in AL is mainly dependent on symptoms (in the absence of pathology to explain those symptoms). ¹⁸F-Florbetapir may prove useful for imaging in such patients for diagnosis of gut amyloidosis. Ezawa et al. demonstrated a good correlation between gastric uptake and clinical gastric involvement (decreased gastric motility) in their series [12]. The differences between our findings and those of Ezawa et al. may be due to the different imaging protocols used. We performed a dynamic study over 60 min followed by a half-body acquisition, and Ezawa et al. scanned their patients 30 min after injection. There may also be differences in the results due to the use of different tracers.

Although few patients showed marrow uptake on ¹²³I-SAP scintigraphy in the current series, the agreement between the two tracers was striking. Bone marrow uptake of ¹⁸F-florbetapir occurs in normal controls [19] and hence its utility for bone marrow imaging remains unclear, although patients with higher bone marrow uptake are likely to have pathological bone involvement. Such uptake in a patient with amyloidosis may support a diagnosis of AL as marrow uptake is not seen (on ¹²³I-SAP scintigraphy) in any other amyloid type.

Three patients with AL had repeat ¹⁸F-florbetapir imaging. At the time of their initial imaging, one of these patients was treatment-naive and subsequently went on to have chemotherapy, one was receiving the first cycle of chemotherapy and one had achieved a complete response with chemotherapy. All three patients were in a haematological complete response at the time of repeat imaging. Despite achievement of a haematological response, there were no differences in the distribution and intensity of uptake between initial and repeat imaging. This may have been due to the relatively short time between the baseline and follow-up scans.

Imaging amyloid deposits outside the heart has remained challenging. Recently, most imaging methods such as cardiac MRI and ^99mTc-DPD/PYP have focused on cardiac amyloid detection. ¹²³I-SAP scintigraphy is used in routine clinical practice at the UK National Amyloidosis Centre for in vivo imaging of amyloid deposits and is the only method that is clinically validated for this purpose. The uptake of ¹²³I-SAP correlates with the amount of amyloid [6] with a 90% sensitivity in amyloid A amyloidosis and AL [20]. The main limitation of the technique is that it is available at only two centres worldwide (UK National Amyloidosis Centre and Groningen University) because of its technical complexity. The other limitation is that ¹²³I-SAP scintigraphy images the liver, spleen, kidneys and bone, but cannot image the hollow viscera, other soft tissues and heart. Recently, extracardiac uptake of PET tracers (¹⁸F-florbetapir, ¹⁸F-florbetaben and ¹¹C-PIB) has been described in a few case reports: in the larynx [21], cerebral amyloidoma [22], in the kidneys in ALECT2 type of amyloidosis [23] and in the brain [24]. The only systematic study of extracardiac uptake has recently been published [12]. Half-body images were acquired 30 min after ¹¹C-PiB injection and showed abnormal tracer uptake in the spleen, stomach, lachrymal glands, submandibular glands, sublingual glands, lymph nodes, brain, scalp, extraocular muscles, nasal mucosa, pharynx, tongue and nuchal muscles, but most patients were asymptomatic. Ezawa et al. also noted physiological tracer uptake in the urinary tract (kidney, renal pelvis, ureter and bladder) and enterohepatic circulatory system (liver, gallbladder, bile duct and small intestine). Our findings with ¹⁸F-florbetapir PET imaging are in broad agreement with the results of ¹¹C-PiB PET imaging. The main advantage of ¹⁸F-florbetapir over ¹¹C-PiB PET with regard to potential routine clinical implementation is its longer half-life, with no need for an on-site cyclotron and with commercially available solutions for supply and delivery already available in most of Europe and North America.

One of the main limitations of the study was that there was no biopsy proof for every site noted to have abnormal ¹⁸F-florbetapir uptake. This was not part of the protocol and, due to potential risks, not ethically acceptable in the absence of a specific clinical indication. The other limitation was that this study was limited to cardiac patients (mainly because the study was originally designed to study cardiac amyloidosis) and only patients with AL and ATTRwt were included. Investigation of patients with a wider range of amyloid types and organ involvement is needed to understand physiological and abnormal ¹⁸F-florbetapir uptake.

In summary, there are a limited number of techniques for imaging extracardiac amyloid deposits. ¹²³I-SAP, the gold standard for imaging systemic amyloidosis, has major limitations (it is only available in two centres worldwide, it requires the use of human plasma-derived SAP protein, and the preparation of the tracer is complex). ¹⁸F-Florbetapir (and other similar tracers) does not have the same limitations and has the great advantage that it is already commercially available worldwide and can be ordered by any appropriately equipped department. Scanning is easy with a half-body scan acquired in less than 30 min as early as 15 min after injection. PET cameras also have better resolution than gamma cameras, along with the possibility of acquiring dynamic scans on cross-sectional images. The combination of heart and spleen uptake of ¹⁸F-florbetapir may be a specific signature of AL. With its very high cardiac specificity that has already been reported and the additional imaging of extracardiac deposits, ¹⁸F-florbetapir has the potential to become the main tracer for the diagnosis of systemic amyloidosis. Its role in monitoring disease remains to be clarified, as we did not see any changes on repeat imaging following successful treatment. Large studies in unselected patients with AL are urgently needed to confirm these findings.