Bone-seeking radiotracers have revolved primarily around TTR-CA diagnosis while offering a low sensitivity for identifying AL-CA. Targeted amyloid PET imaging has the capability to identify all amyloid deposits regardless of the original protein precursor. Multiple radiotracers for amyloid PET have been investigated in Alzheimer’s disease; these tracers are believed to bind with the beta-pleated motif of the amyloid fibril and have shown promise in clinical research studies, as described in the North American and European expert consensus recommendations for cardiac amyloidosis [6,37]. PET imaging using semi-quantitative parameters (target-to-background ratio) or quantitative metrics (SUV or retention index) has been used in several studies and has demonstrated values for the detection of CA [38]. A meta-analysis including ¹¹C-PIB, ¹⁸F-florbetapir, and ¹⁸F-florbetaben PET studies showed that the combined sensitivity and specificity for the diagnosis of CA was 95% and 98%, respectively [39,40,41]. A recent study using ¹⁸F-flutemetamol PET/MRI or PET/CT showed it to be positive in only 2 of 12 patients with CA, but there was no difference in SUV_max and SUV_mean between CA and controls [42]. However, another pilot ¹⁸F-flutemetamol study demonstrated a higher number of favorable results (8 out of 9 patients with CA) [43]. Due to the contradictory results, further studies with large sample sizes and appropriate imaging protocols are needed in order better to define the accuracy of ¹⁸F-flutemetamol PET imaging in CA.

To date, only small preliminary studies in vivo utilizing these tracers have highlighted potential benefits in stratifying AL-CA from TTR-CA. The cardiac retention index of ¹⁸F-florbetapir tended to be higher in AL patients but was not significantly less in TTR-CA [38]. Subsequently, a similar finding was reported [41] without being able to separate the two CA subtypes by using dynamic PET with ¹⁸F-florbetaben in a 20 min scan. These initial findings were also re-observed on ¹¹C-PiB PET imaging [40,44,45]. More recently, a dual-center study demonstrated that ¹¹C-PiB PET imaging had 100% diagnostic accuracy of visual assessment in AL amyloidosis. Moreover, the myocardium to blood SUV ratio was significantly higher in AL than in ATTR patients [46]. However, since ¹¹C-PiB binding to cardiac amyloids has a larger variability in retention index [47], and an on-site cyclotron is required for the production of ¹¹C-labeled radiotracer, ¹⁸F-labeled PET tracers with a longer half-life (110 min) may overcome these drawbacks and may be more widely used.

¹⁸F-florbetaben PET imaging with a dynamic and delayed acquisition shows a significant increase in uptake in patients with AL compared to those with ATTR or without CA. However, in early studies, there was no difference between AL and TTR amyloidosis patients [48]. In addition, there was virtually no difference in the retention index between TTR and controls. These observations strongly suggested that a late scan obtained at least 30 min after ¹⁸F-florbetaben injection can reliably discriminate CA due to AL from either TTR or other mimicking conditions. Notably, because of the non-specific affinity to TTR-CA patients, the application of ¹⁸F-florbetaben and PET imaging for the identification of TTR in patients with suspected CA may not be recommended. Furthermore, a dynamic ¹⁸F-florbetaben PET study demonstrated that using kinetic model-fitting parameters allows for distinguishing between the two types of amyloid pathology and strengthens the diagnostic accuracy for AL-CA [49]. With the development and application of Deep Learning (DL) techniques in the field of medical imaging, it was reported that a simple DL model derived from cardiac ¹⁸F-florbetaben PET images acquired a few minutes after the injection could be used to help clinicians differentiate AL from TTR-CA [50]. Therefore, although the clinical manifestations of AL and ATTR amyloidosis overlap, ¹⁸F-florbetaben PET will likely open new avenues as the preferred imaging modality in patients with suspected AL amyloidosis.

Conventional serum biomarkers such as NT-proBNP and troponin have demonstrated prognostic value in patients with CA, as they indicate cardiac involvement of systemic amyloidosis. However, only a few studies have investigated the role of amyloid-targeting PET in this specific patient group. Recently, Lee SP et al. [40] reported that quantification of myocardial ¹¹C-PiB uptake by PET was associated with adverse outcomes independently among patients with AL-CA. Interestingly, the degree of myocardial ¹¹C-PiB uptake seems to be a better predictor for clinical events such as all-cause death or acute decompensated heart failure than LVEF or diastolic filling parameters (E/e’ ratio) assessed by echocardiography. Furthermore, Choi YJ et al. [51] showed that ¹¹C-PiB PET/CT was not only a strong independent predictor of one year overall survival, but also provided incremental prognostic benefits when combined with commonly used serum biomarkers such as troponin I, NT-proBNP, and dFLC in patients with AL-CA. Given the need for improved cardiac staging systems for prognostication in AL, future prospective studies, possibly multicenter studies, are warranted to discover whether PET/CT should be incorporated into risk stratification for AL amyloidosis patients.

Therapy response for AL amyloidosis is normally assessed by hematology (decreased light chains affected and normalization of bone marrow) and organ involvement. For CA, reduction in NT-proBNP by >30% and 300 ng/L (if baseline NT-proBNP ≥ 650 ng/L) is defined as a cardiac response [52]. However, there are drawbacks of solely using natriuretic peptides when they are influenced by other factors (e.g., renal dysfunction, atrial fibrillation and lower body mass index) which are common comorbidities in patients with AL-CA [53]. Thus, imaging biomarkers that reflect a cardiac response could be very useful in detecting changes in structure or function associated with biomarker variation following treatment.

A non-invasive method for detecting and quantifying cardiac amyloid depositions with ¹⁸F-florbetapir was evaluated in patients with AL-CA prior to and post-chemotherapy, and its serial utility in monitoring was also assessed [54]. In the three patients with complete hematological response, although no significant difference in cardiac uptake was found between the first and repeated images, there was a suggestion that treatment-naïve patients may have higher cardiac uptake. Given the prospective nature of this study, future studies are needed to include more significant numbers of patients and to assess the optimal time of repeat ¹⁸F-florbetapir imaging after treatment. Furthermore, another cardiac ¹⁸F-florbetapir PET study consistently showed that all AL patients in that cohort had demonstrable cardiac uptake, despite hematologic remission for more than 1 year [55]. The results offer new insight into the argument of “active” vs. “inactive” amyloid hearts [46], irrespective of serum FLC burden response. In addition to ¹⁸F-florbetapir, the use of ¹⁸F-florbetaben PET/CT for monitoring anti-inflammatory (AA), anti-myeloma (AL) and TTR-stabilizing (TTR) therapy were also evaluated [56]. The results showed that changes in myocardial tracer retention from baseline to follow-up corresponded well with treatment response.

In summary, the theoretical benefits of amyloid PET in assessing prognosis and monitoring treatment response allow more widespread use in CA. Furthermore, this method will shed much light on a currently opaque and ambiguous assessment process. Nevertheless, it is not yet possible to make reliable statements regarding the incorporation of amyloid PET into the CA-imaging algorithm due to limited data (mostly retrospective) [57], and there are no clinically available PET tracers authorized in the European Union or the United States. Currently, a larger prospective phase III trial evaluating the utility of ¹⁸F-florbetaben PET as a diagnostic method for AL-CA is underway (NCT05184088).