Echo Inconclusive? Here’s How to Navigate Endocarditis Imaging

Let me start this post with a question we encounter quite often in clinical practice:

Imagine a patient in whom you're suspecting infective endocarditis (IE). Maybe blood cultures are positive for a typical microorganism and the patient has a fever—but you still haven’t confirmed valve involvement according to Duke’s criteria (assuming the remaining minor criteria are negative). TTE and maybe even TEE are inconclusive, unavailable, or contraindicated—something that happens more often than we'd like. So, you start thinking about other imaging options: cardiac CT, cardiac MRI, and nuclear medicine studies.

An excellent review article on this topic (PMID: 32683888) offers a comprehensive summary of which imaging modalities to consider for IE in various clinical scenarios. The best approach is to break it down into native valve endocarditis (NVE) and prosthetic valve endocarditis (PVE). I’ll address CIED-related infections and other non-valvular hardware infections in a separate post.

Native Valve Endocarditis (NVE)

In most cases, NVE and its mechanical complications are diagnosed using echocardiography (TTE and TEE). This means CT and nuclear imaging modalities are traditionally underutilized in NVE. However, more patients now present with absolute contraindications to TEE, and these other imaging techniques are being used more frequently.

A key study by Koo et al. (PMID: 28329276) showed that gated cardiac CT has diagnostic accuracy comparable to TEE, though with slightly lower sensitivity for detecting vegetations (91% vs. 99%). Importantly, CT is excellent for identifying pseudoaneurysms, abscesses, and fistulae (PMID: 19179202). The 2023 ESC guidelines now include gated CT as a Class Ib recommendation when infection is suspected and TTE/TEE are non-diagnostic. (The 2015 AHA guidelines, still in circulation, unfortunately don’t include this—making them somewhat outdated in this regard.)

What about nuclear imaging for NVE?

In short, it's not recommended.

A prospective study by Granados et al. (PMID: 27261514) involving 80 patients (21 with NVE) found that FDG PET/CT was false-negative in all NVE cases!, yielding a shockingly low sensitivity of 31%. As for leukocyte (WBC) SPECT/CT, I’m not aware of solid data supporting its use in NVE, and given PET/CT's low sensitivity, it's unlikely we’ll see strong evidence for SPECT/CT either. (But please feel free to share any data I may have missed.)

Prosthetic Valve Endocarditis (PVE)

Here’s where advanced imaging really shines.

When echocardiography is inconclusive, guidelines—started with the ESC's 2015 update—support the use of alternative imaging. This is because prosthetic material creates acoustic shadowing, which hampers the visualization of vegetations and paravalvular complications. In PVE, infection may affect not just the prosthetic valve itself but also adjacent structures, particularly the paravalvular area.

Paravalvular extension occurs in up to half of PVE cases, and gated CT has even proven superior to TEE for detecting abscesses and mycotic aneurysms (PMID: 29555833). Despite some limitations—such as streak artifacts from prosthetic material or CIEDs—gated CT remains highly valuable and is a ESC Class Ib recommendation for evaluating both valvular and perivalvular lesions in suspected PVE.

What about FDG-PET/CT and WBC SPECT/CT for PVE?

Most of the supporting data comes from two meta-analyses:

• Mahmood et al. in 2017, (PMID: 28913626): Found that adding FDG PET/CT improved diagnostic accuracy, with a sensitivity of 80.5% and specificity of 73.1%.

• Wang et al. in 2020, (PMID: 32507019): Reported even higher sensitivity (86%) and specificity (84%).

This was enough to include FDG PET/CT as a Class Ib recommendation in the 2023 ESC guidelines for evaluating valvular (but not paravalvular) lesions in PVE.

What about leukocyte/WBC SPECT/CT?

Data is limited to case reports and retrospective studies. The European guidelines recommend WBC SPECT/CT as a Class IIa option, but only in complex cases where echocardiography is inconclusive and PET/CT is unavailable.

And what about MRI?

MRI still plays a role in the evaluation of embolic complications, particularly cerebral emboli (i.e., brain MRI). It may also help in rare cases of bacterial myocarditis secondary to IE (cardiac MRI). However, MRI has no established role in assessing valvular or paravalvular structures in IE.

Summary

When echocardiography is non-diagnostic or unavailable, alternative imaging modalities are critical in the diagnosis of IE:

Native Valve Endocarditis (NVE)

• Gated CT: Preferred modality; useful for complications like abscesses and pseudoaneurysms (Class Ib).

• FDG PET/CT & WBC SPECT/CT: No role due to poor sensitivity.

Prosthetic Valve Endocarditis (PVE)

• Gated CT: Highly useful for both valvular and paravalvular pathology (Class Ib).

• FDG PET/CT: Helpful for detecting valvular lesions (Class Ib).

• WBC SPECT/CT: Reserved for complex cases when PET/CT is unavailable (Class IIa).

Uff! That’s the end of this post. I hope it wasn’t too heavy on the eyes. I truly hope it helps in everyday clinical decision-making, especially when sometimes echocardiography leaves us with more questions than answers.

Please feel free to comment or (respectfully!) disagree if there are any inconsistencies or additional data worth discussing.

🎵 Musical Coda

A song from the 2000s in Colombia—still feels as fresh and joyful as the first time I heard it.