Apolipoprotein A-IV Cardiac Amyloidosis

Abstract

Cardiac amyloidosis is most commonly caused by transthyretin or immunoglobulin light-chain deposition, and contemporary diagnostic algorithms rely on monoclonal protein assessment and technetium-labeled bone scintigraphy to identify these subtypes. However, rarer forms of amyloid may not be detected by noninvasive testing. We present 2 cases of biopsy-proven apolipoprotein A-IV (ApoA-IV) cardiac amyloidosis illustrating complementary diagnostic pitfalls. The first patient demonstrated classical clinical and multimodality imaging features of cardiac amyloidosis despite negative monoclonal protein studies and pyrophosphate scintigraphy. Endomyocardial biopsy with mass spectrometry confirmed ApoA-IV. The second patient had abnormal serum-free light chains and bone marrow plasma cell involvement suggestive of light-chain amyloidosis, yet cardiac biopsy revealed ApoA-IV, preventing inappropriate chemotherapy. These cases highlight limitations of current nonbiopsy algorithms and emphasize that tissue diagnosis with proteomic typing remains essential when clinical suspicion is high or findings are discordant.

Visual Summary

Visual Summary.

Contemporary Nonbiopsy Algorithms Focused on ATTR and AL Amyloidosis May Lead to False Reassurance When Testing Is Negative or False Attribution When Monoclonal Gammopathy Is Present

When clinical suspicion is high or findings are discordant, endomyocardial biopsy with mass spectrometry is essential for definitive amyloid typing. AL = light-chain amyloidosis; ApoA-IV = apolipoprotein A-IV; CMR = cardiac magnetic resonance imaging; ECG = electrocardiogram; LV = left ventricular; MGUS = monoclonal gammopathy of undetermined significance; PYP = pyrophosphate.

Created in BioRender, Helmy I (2026), https://BioRender.com/amcb6t5.

Cardiac amyloidosis is an increasingly recognized cause of heart failure. Contemporary diagnostic algorithms focus on transthyretin (ATTR) and immunoglobulin light-chain amyloidosis (AL), which account for ∼98% of cardiac cases. These algorithms emphasize monoclonal protein testing followed by technetium-labeled bone scintigraphy, allowing many patients with ATTR to avoid biopsy.^1,2 However, these pathways may miss rarer forms. Apolipoprotein A-IV (ApoA-IV) amyloidosis is an uncommon systemic amyloid that can involve the myocardium and mimic ATTR or AL cardiomyopathy. We present 2 cases of biopsy-proven ApoA-IV cardiac amyloidosis that illustrate complementary diagnostic pitfalls: false reassurance from negative noninvasive testing and potential overtreatment in the setting of a coexisting monoclonal gammopathy.

Take-Home messages

• •Cardiac amyloidosis should remain a diagnostic consideration in patients with unexplained or disproportionate ventricular hypertrophy and characteristic imaging features.
• •Negative technetium-based scintigraphy and the absence of monoclonal proteins do not exclude non-ATTR non-AL cardiac amyloidosis.
• •The presence of monoclonal gammopathy or bone marrow plasma cell involvement, even when serum-free light chains are interpreted with renal adjustment, does not establish AL cardiac amyloidosis, and endomyocardial biopsy with mass spectrometry is essential for definitive amyloid typing when clinical suspicion is high or diagnostic data are discordant.

Patient 1

A 78-year-old man presents with progressive dyspnea on exertion, orthopnea, and leg swelling for 1 month.

His medical history included nonproteinuric chronic kidney disease (CKD) (baseline creatinine of 1.6 mg/dL, estimated glomerular filtration rate [eGFR] 43 mL/min/1.73 m²), well-controlled type II diabetes mellitus, and obstructive sleep apnea.

He presented with tachypnea, hypoxemia, rales, elevated jugular venous pressure, and warm extremities with pitting edema. N-terminal pro-brain natriuretic peptide (NT-proBNP) level was elevated at 15,901 pg/mL. Electrocardiography showed a pseudo-infarct pattern and a first-degree atrioventricular block (Figure 1). Transthoracic echocardiography showed a preserved left ventricular ejection fraction (LVEF), moderate concentric left ventricular thickening, and mild right ventricular thickening (Figure 2A). He improved with parenteral diuretics and was discharged on empagliflozin, spironolactone, and furosemide.

Figure 1.

Electrocardiogram of Patient 1

This electrocardiogram demonstrates a first-degree atrioventricular block and a pseudo-infarct pattern. This tracing was obtained at a paper speed of 25 mm/s and at a voltage calibration of 10 mm/mV.

Figure 2.

Echocardiogram and Longitudinal Strain Pattern of Patient 1

(A) Parasternal long-axis still image view demonstrating moderate concentric left ventricular thickening. (B) Speckle-tracking longitudinal strain imaging shows markedly reduced peak-systolic global longitudinal strain (−9.5%) with apical sparing (apical longitudinal strain: 15.6%), corresponding to an apical-to-basal strain ratio of 2.26, a characteristic pattern associated with cardiac amyloidosis.

At the follow-up visit 1 month later, a transthoracic echocardiogram was repeated with strain imaging, demonstrating a severely reduced global longitudinal strain of −9.5% with apical sparing (apical-to-basal strain ratio of 2.26) (Figure 2B). This longitudinal strain pattern, the left ventricular thickening in the absence of loading conditions, and the relatively low voltages on his electrocardiogram raised suspicion for cardiac amyloidosis. A technetium-99m pyrophosphate (PYP) single-photon emission computed tomography scan showed no uptake (heart-to-contralateral lung ratio of 1.16 and 1.22 at 1 and 3 hours, respectively; visual grade zero). Serum and urine protein electrophoresis did not show a monoclonal band, which was confirmed on immunofixation. He had normal serum-free light-chain levels with a normal kappa-to-lambda ratio.

The clinical suspicion for cardiac amyloidosis remained high, so cardiac magnetic resonance imaging (CMR) was performed, showing diffuse subendocardial late gadolinium hyperenhancement (LGE) that is near transmural at the basal left ventricular walls and less so toward the apex that does not follow a coronary distribution (Figure 3). An endomyocardial biopsy was pursued, and histopathology confirmed cardiac amyloidosis with a positive Thioflavin S stain. Immunohistochemistry was performed, the transthyretin stain was negative, and both the kappa and lambda stains showed weak blushes (Figure 4). Mass spectrometry was then performed, which identified ApoA-IV amyloidosis (Supplementary Figure 1).

Figure 3.

Cardiac Magnetic Resonance Imaging of Patient 1

(A) Short-axis cine images demonstrating concentric left ventricular wall thickening. (B) Short-axis late gadolinium enhancement (LGE) image showing diffuse circumferential near-transmural enhancement throughout the left ventricle. (C) Four-chamber long-axis cine image confirming increased left ventricular wall thickness. (D) Four-chamber long-axis LGE image demonstrating diffuse near-transmural enhancement in a noncoronary distribution. The diffuse near-transmural pattern of enhancement in a nonischemic distribution, together with concentric hypertrophy, is characteristic of an infiltrative cardiomyopathy.

Figure 4.

Representative Images of Apolipoprotein A-IV Amyloidosis

Patient 1. (A) Nodular amorphous material was seen throughout the myocardium. (B) The material was fluorescent on a Thioflavin S stain and (C) stained blue-gray on a Masson Trichrome, in contrast to the collagen which stained a royal blue. Patient 2. (D) A solitary, large amyloid nodule was seen adjacent to the myocardium. (E) It displayed apple green birefringence upon polarization of the Congo Red stain. (F) No amyloid material was seen within the myocardium, as noted on this negative Congo Red stain showing only collagen (white) birefringence. (Original magnifications: A to E = 200×, B = 400×, F = 100×).

Patient 2

A 75-year-old man was referred for evaluation of progressive heart failure with reduced ejection fraction and suspected infiltrative cardiomyopathy.

His medical history included prior surgical aortic valve replacement with redo valve surgery and left internal mammary artery-left anterior descending bypass grafting, hypertension, and proteinuric CKD stage III (baseline creatinine of 2.0 mg/dL, eGFR 34 mL/min/1.73 m², urine microalbumin-to-creatinine ratio of 472.8 mg/g). He had been clinically stable for several years, with LVEF 45% to 50% on serial echocardiograms.

He presented with acute dyspnea, orthopnea, and fatigue. NT-proBNP was 11,057 pg/mL, and high-sensitivity troponin was persistently elevated at 138 ng/L without dynamic rise. Electrocardiography showed a first-degree atrioventricular block and a nonspecific intraventricular conduction defect with a QRS duration of 160 milliseconds (Figure 5). Chest radiography showed pulmonary edema and pleural effusions. Transthoracic echocardiography showed an LVEF ∼25% to 30%, moderate concentric left ventricular hypertrophy (LVH) (Figure 6A), and severe biatrial enlargement. Strain imaging showed a reduced global longitudinal strain (−6.0%) with disproportionate involvement of the anterolateral segments (Figure 6B). Coronary angiography demonstrated a chronic total occlusion of the mid-left anterior descending with a widely patent left internal mammary artery-left anterior descending graft and no obstructive disease elsewhere. His symptoms improved with diuresis, and he was started on guideline-directed medical therapy for heart failure with reduced ejection fraction.

Figure 5.

Electrocardiogram of Patient 2

This electrocardiogram demonstrates a first-degree atrioventricular block with a nonspecific intraventricular conduction defect with a QRS duration of 160 ms. This tracing was obtained at a paper speed of 25 mm/s and at a voltage calibration of 10 mm/mV.

Figure 6.

Echocardiogram and Longitudinal Strain Pattern of Patient 2

(A) Parasternal long-axis still image view demonstrating moderate concentric left ventricular thickening. (B) Speckle-tracking longitudinal strain imaging shows markedly reduced peak-systolic global longitudinal strain (−6.0%). Segmental analysis shows more pronounced impairment in the anterolateral segments compared with the inferior and septal walls. This distribution is atypical for cardiac amyloidosis, which characteristically demonstrates relative apical sparing.

Despite optimization of medical therapy, he experienced recurrent hospitalizations over the following year for acute decompensated heart failure. He had a significant burden of multifocal premature ventricular complexes and was initiated on amiodarone and mexiletine. He was discharged with a wearable cardioverter-defibrillator, with plans for cardiac resynchronization therapy with an implantable cardioverter-defibrillator.

Advanced heart failure evaluation raised concerns for an infiltrative cardiomyopathy given progressive systolic dysfunction, LVH, restrictive physiology, and a speckled-appearing myocardium. A PYP scan did not show any scintigraphic evidence of ATTR cardiac amyloidosis, with heart-to-contralateral lung ratios of 1.13 and 1.10 at 1 and 3 hours, respectively. Laboratory testing revealed a markedly elevated serum-free light-chain kappa level of 236.2 mg/L, with a kappa/lambda ratio of 14.23, although serum protein electrophoresis and immunofixation were negative for monoclonal protein. Subsequent urine protein electrophoresis detected monoclonal kappa-free light chains (Bence-Jones protein), raising suspicion for AL.

Bone marrow biopsy showed a plasma cell neoplasm accounting for ∼5% to 10% of marrow cellularity with amyloid deposition in rare small vessels. Fluorescence in situ hybridization demonstrated t(11; 14) translocation. A F-18 deoxyglucose positron emission tomography/computed tomography did not show a plasmacytoma. Given his frailty and cardiac and renal comorbidities, daratumumab monotherapy was planned for presumed systemic AL.

An endomyocardial biopsy was performed concurrently with cardioverter-defibrillator implantation, and an unusual amorphous nodule was seen in an area of fibrosis adjacent to myocardium. No amyloid material was seen within the myocardium. The nodule stained positive for amyloid on both a Thioflavin S stain and a Congo red stain, but was not definitively amyloid based on the Masson trichrome and Movat stains. Electromicroscopy identified small patches of amyloid fibrils within the myocardium (Figure 4). Mass spectrometry was ultimately performed, which identified the amyloid as ApoA-IV. In light of the updated cardiac biopsy findings, the bone marrow biopsy was reinterpreted as monoclonal gammopathy of undetermined significance, and chemotherapy plans were withdrawn.

Discussion

Cardiac involvement represents a clinically important manifestation of systemic amyloidosis and may be underrecognized in contemporary heart failure populations. One prospective histopathology study estimated a 14% prevalence of cardiac amyloidosis in those presenting to a heart failure with preserved ejection fraction clinic.³ While ATTR and AL amyloidosis account for ∼98% of cardiac amyloid, unbiased proteomic analyses have demonstrated a wider spectrum of amyloid subtypes. A proteomics study of over 5,000 cardiac samples identified 9 amyloid subtypes, with ATTR accounting for 65%, AL for 33%, and ApoA-IV comprising 0.37% of cases.⁴ An earlier autopsy-based series similarly identified one ApoA-IV case among 108 cardiac amyloidosis cases.⁵

Several features suggest cardiac amyloidosis in a patient presenting with heart failure: LVH not explained by loading conditions such as hypertension or aortic stenosis, low-voltage QRS complexes discordant with LVH, presence of Q-waves on electrocardiography without prior infarcts, and persistent low-level troponin elevation.^1,6 Extracardiac manifestations such as polyneuropathy, bilateral carpal tunnel syndrome, bicep tendon rupture, and macroglossia may occur but are not universally present.¹ One small study showed that a relative apical longitudinal strain >1.0 on echocardiography in those with LVH was 93% sensitive and 82% specific in differentiating cardiac amyloidosis from controls.⁷ On CMR, cardiac amyloidosis is characterized by a diffuse and extensive subendocardial-to-transmural pattern of LGE, predominantly in the basal left ventricular segments, that does not confirm to a coronary perfusion territory. When combined with elevated T1 values (suggestive of increased extracellular volume), this pattern was 95% sensitive and 98% specific for the diagnosis of cardiac amyloidosis when compared with endomyocardial biopsy as the gold standard.⁸ As such, early CMR may be pursued in those with unexplained LVH to evaluate for specific cardiomyopathies, with endomyocardial biopsy likely to be diagnostic when characteristic findings are present.

The 2022 American Heart Association/American College of Cardiology/Heart Failure Society of America Guideline for the Management of Heart Failure suggests that those with a high clinical suspicion of cardiac amyloidosis should undergo monoclonal protein testing and pyrophosphate scintigraphy. In the absence of monoclonal proteins and with a negative pyrophosphate scan, the guidelines suggest that cardiac amyloidosis is unlikely, and no further workup should be pursued.^1,2 While this approach is effective for identifying the 2 most common subtypes of cardiac amyloidosis, namely ATTR and AL, it does not reliably exclude rarer forms of cardiac amyloid.

This diagnostic gap is illustrated in patient 1. Despite the classic clinical, echocardiographic, and CMR features of cardiac amyloidosis, both PYP scintigraphy and monoclonal protein evaluations were negative. The persistence of a high pre-test probability prompted endomyocardial biopsy, which ultimately revealed ApoA-IV amyloidosis by mass spectrometry. This case emphasizes the importance of tissue diagnosis when imaging and clinical features strongly suggest amyloidosis despite nondiagnostic noninvasive testing.

A distinct but equally important pitfall is highlighted in patient 2. In this case, the presence of markedly abnormal serum-free light chains, urinary monoclonal light chains, and bone marrow plasma cell involvement initially suggested systemic AL amyloidosis with possible cardiac involvement. However, endomyocardial biopsy demonstrated amyloid deposits that were ultimately typed as ApoA-IV by mass spectrometry. This finding led to reclassification of the plasma cell disorder to a monoclonal gammopathy of undetermined significance and avoided the inappropriate initiation of cytotoxic chemotherapy. This case emphasizes that the presence of a plasma cell dyscrasia does not establish the amyloid subtype responsible for cardiac involvement, and that reliance on extracardiac tissue or serologic findings alone may lead to misdiagnosis or overtreatment.

ApoA-IV is a protein synthesized in the hypothalamus, liver, and intestines; secreted in response to lipid intake; and works to regulate satiety and glucose homeostasis.⁹ ApoA-IV amyloidosis commonly involves the kidneys, heart, and gastrointestinal tract, and very rarely involves the skin and lungs.⁴ In contrast to other forms of amyloidosis, renal involvement in ApoA-IV amyloidosis is characterized by predominantly medullary deposition, minimal proteinuria, and a slowly progressive course of kidney disease.¹⁰ Renal involvement is particularly relevant when interpreting serum-free light chains. In CKD, the kappa/lambda ratio is more informative than the absolute free light chain values because reduced GFR disproportionately impairs renal clearance of kappa-free light chains, resulting in a higher ratio even in the absence of a monoclonal process. However, even when the eGFR is <30 mL/min/1.73 m², the validated renal-adjusted upper limit of normal of the kappa-to-lambda ratio is 3.30.^1,11 In our second patient, the ratio of 14.23 remained well above this range, so AL amyloidosis remained a diagnostic concern prior to definitive testing. Furthermore, the markedly elevated NT-proBNP values in both patients may have been influenced in part by CKD, although severe NT-proBNP elevation in cardiac amyloidosis may also reflect substantial myocardial involvement and fibrosis.

The diagnosis of Apo-AIV cardiac amyloidosis is challenging. It shares almost all of the clinical and CMR findings of the other cardiac amyloidosis subtypes. However, myocardial strain analysis on echocardiography may not consistently demonstrate the characteristic apical sparing pattern, as illustrated in our series where patient 2 showed a nonapical sparing pattern. It also does not commonly demonstrate uptake on technetium-based scintigraphy. A study of 14 cases of ApoA-IV cardiac amyloidosis showed no uptake of technetium-labeled 3,3-diphosphono-1,2-propanodicaboxylic acid in 13 patients, and only grade 1 uptake in one patient.¹² Nine of those patients underwent CMR, all of whom had the characteristic findings of increased maximal wall thickness, abnormal tissue characterization with LGE, and an increased extracellular volume.¹² The diagnosis of this subtype of cardiac amyloidosis is also difficult to make histologically: routine immunohistochemistry may not contain antibodies specific to ApoA-IV, and its specificity is also limited by background staining. Immunoelectron microscopy also relies on antibody assays that contain the subtype in question and has limited accessibility. Only mass spectrometry–based proteomics would allow for an unbiased identification of amyloid protein, which would be needed to establish the diagnosis in the context of appropriate clinical and histomorphologic features.¹³

To date, there are no directed therapies for ApoA-IV cardiac amyloidosis. A case series of 13 patients with ApoA-IV cardiac amyloidosis (determined by mass spectrometry) showed that all were alive at a median 79.2-month follow-up time, suggesting a favorable prognosis.¹⁴ Consistent with these observations, patient 1 has remained clinically stable on guideline-directed medical therapy without recurrent hospitalizations 51 months after presentation, while patient 2 avoided unnecessary chemotherapy and continues to be managed with advanced heart failure and device-based therapies.

Funding Support and Author Disclosures

The authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Appendix

Supplementary Figure 1

xxx