By Michael H. Crawford, MD, Editor
SYNOPSIS: A single institution case series of cancer-associated thrombotic endocarditis has shown that it most frequently presents as a systemic embolism and is detected largely by transesophageal echocardiography, which displays mobile masses attached to thickened mitral and aortic valves.
SOURCE: Kurmann RD, Klarich KW, Wysokinska E, et al. Echocardiographic findings in cancer-associated non-bacterial thrombotic endocarditis: Clinical series of 111 patients from a single institution. Eur Heart J 2024;25:1255-1263
Marantic endocarditis, or non-bacterial thrombotic endocarditis (NBTE), associated with cancer has been described in autopsy series and small clinical series using echocardiography. Thus, this series of more than 100 cases of cancer-associated NBTE from the Mayo Clinic is of interest.
Using their database from 2002 to 2022, investigators sought patients with cancer-associated NBTE diagnosed by echocardiography, with negative blood cultures and serum markers for systemic lupus. There were 111 such patients identified (mean age 64 years, 68% women) with NBTE diagnosed by transesophageal echocardiography (TEE) in 102 patients and transthoracic echocardiography (TTE) in nine patients. In the 47 patients who had both studies, only 22 of the TTE studies detected NBTE (47%).
Stroke was the most common reason for ordering an echocardiogram (53%). Autopsy findings in 15 cases and surgical specimens from two cases confirmed the diagnosis of NBTE. Interestingly, 55% of the patients had a history of venous thromboembolism. The mitral valve most commonly was affected by NBTE (62%), followed by the aortic valve (50%), the tricuspid valve (7%), and the pulmonic valve (1%). Two valves were affected in 16% and one patient each had three or four valve involvement.
Systemic emboli frequently were found (92%), with stroke being the only embolic complication in 75%. Mitral valve NBTE more frequently affected the upstream (atrial) side of the valve (90% vs. 49%) and tricuspid valve NBTE more frequently affected the downstream (ventricular) side (75% vs. 38%). The largest NBTE masses were found on the tricuspid valve, and valve leaflet thickening frequently was observed on the mitral and aortic valves. On the mitral valve, NBTE mainly affected the closing margins but not the commissures. Only six patients with mitral valve NBTE had moderate to severe regurgitation and five had severe regurgitation.
In those with aortic valve NBTE, seven patients had moderate to severe regurgitation and five patients had severe regurgitation. However, patients with gynecological or urogenital cancer were overrepresented in those with significant valve regurgitation. NBTE size and location were not influenced by cancer type. The authors concluded that cancer-associated NBTE is manifested mainly as mobile masses attached to thickened aortic or mitral valves. Tricuspid valve location is much less common, but it hosts the largest NBTE masses.
Commentary
In my institution, everyone with a stroke or systemic thromboembolism gets a TTE with saline to look for an atrial septal defect or patent foramen ovale. Apparently, at the Mayo Clinic, if you have evidence of a thromboembolism associated with active cancer, you get a TEE, since almost all subjects in their report had one and very few had a TTE. Their data support the wisdom of this practice, because in the group that had both studies, TTE had only a 47% sensitivity for detecting valve masses. The specificity of TEE is not addressed because almost all the patients (85%) had clinical evidence of thromboembolism. Thus, the authors urged us to prioritize TEE in cancer patients with signs of thromboembolism.
They also remarked on the characteristics of the masses. Almost all (88%) were “vegetation-like” and more often were broad-based and less frequently stalk-like. Thus, they were less likely to be mobile masses, but since the incidence of thromboembolism was high, this could be because the mobile part of the NBTE had left the scene. Most were between 0.5 cm and 1.0 cm in diameter, with the largest being an aortic mass that measured 2.7 cm long. Larger-sized NBTE masses did correlate to the presence of thromboembolism.
Although the morphology of the masses largely agreed with previous smaller autopsy and echocardiography reports, the researchers noted that on the aortic valve, the masses usually were at the tips of the leaflets, whereas on the mitral valve they often were along the closing margins, especially on the anterior leaflet. However, significant mitral stenosis was observed only in one patient, and moderate or more regurgitation was unusual and largely observed in those with gynecologic or urologic cancer.
There are limitations to this study. It is retrospective and relies on routine clinical documentation. There was no overreading of the echocardiograms. Also, embolism may alter the characteristics of the NBTE masses, so the details reported may not be what would be observed in cases without evidence of thromboembolism. In addition, there are no data on treatment or outcomes reported. However, a companion publication fills in some of these details.1 For example, 94 patients (82%) had brain emboli, 11 patients had spleen emboli, 10 patients had renal emboli, six patients had emboli in the coronary arteries, and four patients had emboli in the extremities. Anticoagulation was administered to 104 patients, most with low molecular heparin, and there were 14 patients with major bleeding. Finally, mortality was very high (78%). Thus, whether the detection of cancer-associated NBTE alters the clinical course of the patient is unclear.
REFERENCE
- Patrzalek P, Wysokinski WE, Kurmann RD, et al. Cancer-associated non-bacterial thrombotic endocarditis — clinical series from a single institution. Am J Hematol 2024;99:596-605.