We report 3 cases of extraventricular ATTR amyloid deposits confirmed in surgically resected aortic valves and LA structures, all of which were unlikely to have ATTR amyloidosis infiltrating the ventricles as determined by multimodality evaluation and mid-term clinical course up to 5 years (summarized in Table 1). Our cases suggest the existence of extraventricular ATTR amyloid deposits localized in the aortic valve and LA structures.

AANP amyloid deposits in the atrium, often referred to as isolated atrial amyloidosis (IAA) [8, 9], have been reported in surgically resected appendages and autopsy studies [4, 10,11,12,13,14,15,16]; however, very few histologically reports of ATTR amyloid in the atrium are published [4, 10,11,12]. Further, most previous studies were limited to evaluation of the atrium and did not evaluate the ventricles or other cardiac structures. Thus, the relationship between ATTR amyloid deposits in the atrium and ATTR-CA infiltrating the ventricles, and the clinical impact of atrial ATTR amyloid deposits remains unclear.

Di Bella et al. reported in 2018 that atrial involvement in ATTR-CA is an early stage of ATTR-CA [7]. However, recent reports show contradictory results [2, 4]. Ichimata et al. [4], in line with Di Bella et al., reported in a study of 44 autopsy cases of ATTR-CA evaluating multiple cardiac regions that atria-predominant ATTR amyloid deposition may precede TTR amyloidosis of the ventricles; however, by using cluster analysis from retrospective insights. On the other hand, Hussain et al. [2] recently reported in a study of 580 patients with 99mTc-PYP scintigraphy that 58 patients had atrial uptake without ventricular uptake, while 105 patients had ventricular uptake without atrial uptake, indicating that atrial involvement may not necessarily precede ventricular involvement.

As ATTR amyloids are believed to be derived and formed from circulating transthyretin protein synthesized by the liver, ATTR amyloid deposits in the atrium may be a part of systemic disease and ATTR amyloid deposits could be found in other tissue sites after long-term follow up. However, our findings support the possibility that they could be localized atrial ATTR amyloid deposits. While the clinical significance of atrial ATTR amyloid deposits has not been fully elucidated and long-term close follow-up may be justified at this point, our 2 cases of surgically resected ATTR amyloid deposits in the atrium with mid-term prospective follow up, the first to be reported with multimodality evaluations, suggest the possibility of extraventricular ATTR amyloid deposits localized in the atrium.

While various types of amyloid deposits including light-chain (AL), amyloid A protein (AA) and Apo-AI have been reported in surgically resected aortic valves and autopsy studies, very few histological reports of ATTR amyloid in the aortic valve have been published [17,18,19], and their relationship with ATTR-CA infiltrating the ventricles is not fully understood. Recent studies have reported that ATTR-CA (infiltrating the ventricles) coexists in 11–16% of elderly patients with AS undergoing transcatheter aortic valve implantation [5, 6]. However, histological diagnosis is rarely confirmed in these patients, and the association among ATTR-CA, AS and aging has not been clarified.

Recently, Singal et al. studied 46 surgically resected aortic valves with concomitant ventricular biopsy and reported that 33 (71.7%) had aortic valvular amyloid deposits (including 19 ATTR derived); however, none had amyloid deposits in the ventricle, and they concluded that the majority were “isolated valvular ATTR amyloid deposits” [19]. In line with their report, our 2 surgical cases of ATTR amyloid deposits in the aortic valve with longer follow-up (up to 5 years) and multimodality evaluations support the possibility of localized ATTR amyloid deposits in the aortic valves.

Similar to the discussion with ATTR amyloid deposits in the atrium, deposition of ATTR amyloid is believed to be a systemic process, thus we could not rule out the possibility that valvular ATTR amyloid deposit is an early stage of ATTR-CA (infiltrating the ventricles) without long-term follow up. However, our cases suggest the possibility that they could be localized valvular ATTR amyloid deposits. Our findings also suggests that, to understand the pathophysiology of patients with concomitant AS and ATTR-CA, the existence of amyloid deposits localized in the valves should be taken into consideration.

While clinical implications of extraventricular ATTR amyloid deposits localized outside the ventricles are not fully elucidated, our cases are the first detailed report of surgically resected cases and with mid-term follow up (up to 5 years), which suggested the existence of extraventricular ATTR amyloid deposits localized in the aortic valve and the atrial structures. Diagnosis of amyloid deposits were confirmed by Congo red or direct fast scarlet staining, followed by immunohistochemistry staining for ATTR and AANP. Mass spectrometry or immunoelectron microscopy was unavailable in our patients, which might have provided further insights because of their proven superior sensitivity compared to immunohistochemistry [20, 21].

It could be hypothesized that our cases were localized amyloid deposits rather than an early ATTR-CA, and raises the possibility of localized “innocent” deposits not strictly related to systemic deposition as reported in literatures [10, 22,23,24]. However, non-cardiac manifestations or extracardiac ATTR amyloid deposits as in the case of carpal tunnel syndrome [25, 26] are believed to precede ATTR-CA by a period of years (although none of our 3 cases had history of carpal tunnel syndrome or lumber canal stenosis). Further long-term prospective studies will clarify the clinical implications of extraventricular ATTR amyloid deposits and help us understand whether they are useful for early diagnosis and treatment intervention of ATTR-CA [27].