Talking Tall Rounds: Infective Endocarditis
In this episode Bo Xu, MD, discusses multimodality imaging to evaluate suspected infective endocarditis. Haytham Elgharably, MD, provides an overview and reviews indications and optimal timing for cardiac surgery.
Talking Tall Rounds: Infective Endocarditis
Welcome to the Talking Tall Rounds series, brought to you by the Sydell and Arnold Miller Family Heart, Vascular and Thoracic Institute at Cleveland Clinic.
Haytham Elgharably, MD:
Good morning, everyone. Thanks for joining us today. The topic today is Management of Infective Endocarditis in Persons Who Inject Drugs to represent the clinical and ethical challenge for all the heart team. In today's agenda, we'll try to cover a different aspect of this problem, including preoperative imaging, medical management, role of surgery, transcatheter options, addiction treatment after and also postoperative rehab options available for us. We'll be starting with a case presentation by our chief resident Michael Javorski.
Michael Javorski, MD:
There's a broad spectrum of presentation and disease, so we'll go through three quick case scenarios. The first case is a 41-year-old female with a history of injection drug use and hepatitis C. She had a prior biologic tricuspid valve replacement in 2022. She presented one year later with infective endocarditis and septic arthritis and recent heroin use. She requested me to use methadone treatment for medication-assisted therapy. We wanted to get her on medication-assisted therapy before surgery, but we felt strongly about getting her on suboxone instead of methadone. She finally agreed to that, so we took her to the operating room and did a redo tricuspid valve replacement. This was the prosthetic tricuspid valve endocarditis pre-op. For follow-up, she was discharged last week on suboxone to a skilled nursing facility, and she'll complete 6 weeks of antibiotics.
The second case is a 31-year-old female with injection drug use. She had a history of infective endocarditis of the pulmonary valve, tricuspid valve in 2022. Her last injection drug use was a year ago and she'd been doing well on suboxone and goes to therapy. She presented with worsening heart failure symptoms, though, and tricuspid valve regurgitation, pulmonary regurgitation and CTEPH. We took her to the operating room to do a pulmonary valve replacement, tricuspid repair and a left PTE. Her valve disease is shown here, pulmonary valve insufficiency, and then tricuspid valve regurgitation. For her follow-up, she was discharged on post-up day 7 and she returned to clinic one month later doing well. Remains injection drug use-free on Subutex.
The third case is a 24-year-old male with a history of injection drug use with methamphetamine and a prior aortic valve replacement in 2021 for endocarditis. He was presented 2 years later with recurrent infective endocarditis with intermittent injection drug use, and he was declined surgery elsewhere. He was seen by our inpatient psych team and his addiction and anxiety disorder treatment started before surgery. We offered him a redo AVR with a homograft. This is his prosthetic valve endocarditis. His follow-up 3 months post-op, he's actually doing very well. He remains sober, he has a good family support system, and his ejection fraction is improving from surgery. Thank you.
Bo Xu, MD:
I'm going to focus on the role of multimodality cardiovascular imaging in right-sided and tricuspid valve endocarditis. Right-sided endocarditis is very serious, increasingly recognized, as we all know, with tricuspid valve being the most frequently involved. Risk factors include, as we know, intravenous drug use and the presence of cardiovascular devices and catheters.
This is a recent work we published. This work shows that for patients with endocarditis readmitted within 30 days, these patients were more likely to have hepatitis C, HIV, opioid abuse, cocaine abuse and multi-substance abuse. Commonly, it's important to highlight that a lot of these patients had coexisting valvular disease, and tricuspid valve is the most commonly involved, as shown here. In this invited editorial commenting on a multi-center study on the role of surgery in isolated tricuspid valve disease, we found what's interesting is that tricuspid valve disease due to endocarditis etiology had reduced survival compared to non-IE etiologies, in that these patients from this work had higher risk of reoperation, higher risk of late cardiac deaths, and therefore, I think this highlights the prognostic value of IE etiology, especially for patients with isolated tricuspid valve dysfunction.
I want to focus on my expertise, my interest area, which is multimodality cardiovascular imaging of endocarditis. To me, there is a critical role of multimodality cardiovascular imaging, not only in diagnosis, but also in detection of complications and high-risk presentations, as well as guidance of treatment approaches, monitoring of treatment response, and prognostication. As shown, there are various imaging modalities available in our armamentarium and it should be used appropriately depending on the clinical scenario.
Transthoracic echo should always be the initial diagnostic modality. Tricuspid valve endocarditis can be associated with tricuspid valve stenosis and/or regurgitation, secondary to leaflet perforation and subvalvular involvement. The role of echo is mainly involving defining the size of the vegetations, assessing the degree of tricuspid valve stenosis and regurgitation, assessing prosthetic valve dysfunction if there is a prior tricuspid valve surgery, and also assessing cardiac implantable devices in CIED-IE, importantly not to miss assessing intracardiac shunts.
This is a transverse example of a published case that we had of a severe native tricuspid valve endocarditis as shown with the large vegetation, severe tricuspid valve regurgitation, RV dysfunction, as well as extensive pulmonary septic emboli. In comparison, this is another case that we have of native tricuspid valve endocarditis with the multi-lobulated vegetation, but relatively only mild to moderate tricuspid valve regurgitation and relatively preserved RV size and function, these factors are all relevant for potential surgical consideration. The role of transesophageal echo comes in because it has superior image quality and spatial resolution. Therefore, we should use it in investigation of prosthetic valve endocarditis as well as CIED-IE or complicated tricuspid valve endocarditis. This is another published example that we had of the role of TEE in tricuspid valve endocarditis. As shown here, there's a very large vegetation on the native posterior tricuspid valve leaflet, and this case really highlights the power of 3D imaging in determining the precise location of the vegetation.
Shown here is a different TEE example, this time of a prosthetic tricuspid valve endocarditis with very bulky vegetation and severe prosthetic tricuspid valve stenosis with elevated gradients. These are all factors relevant for consideration of surgical management. To add to the flavor, this is another gross example of a patient with a giant mobile tricuspid valve vegetation associated with flail tricuspid valve leaflets, and as you can see, torrential tricuspid valve regurgitation. So, if this is a patient who's a surgical candidate, this obviously should not be medically managed alone.
When it comes to intracardiac shunting, shown here on the left-hand side, we've got a very mobile interatrial septum with an ASD. In cases of tricuspid valve endocarditis with a large vegetation burden, the presence of such defects obviously poses embolic risks, and this has to be considered in terms of preoperative planning. Shown on the right-hand side, also commonly we see patent foramen ovale here with right to left shunting, which is also very important to be aware of.
We also need to be aware when complicated right-sided endocarditis is associated with left-sided endocarditis. As shown here on this example, there's a large vegetation on the device lead associated with three plus tricuspid regurgitation. There's, interestingly, concurrent perforation of the mitral valve with a large vegetation and four plus mitral valve regurgitation. Clearly, this is a very high-risk patient and, again, a candidate of surgical management will be indicated.
Another example of a very severe case of prosthetic tricuspid valve endocarditis, shown here is a huge prosthetic tricuspid valve vegetation with a very high infective burden and the whole valve on 3D imaging is coated. We showed, in our work published here recently, the contemporary TEE imaging with modern 3D imaging technology improved diagnostic performance of endocarditis, and this is predominantly driven by improved diagnostic performance in prosthetic valve endocarditis. This work shows the value of imaging for preoperative assessment and planning.
I want to highlight that the current ACC/AHA valvular heart disease guidelines have also acknowledged the role of cardiac CT and PET-CT in assessment of endocarditis in certain cases. We, actually, in our group, published the largest meta-analysis to date on this topic of PET imaging in endocarditis. We actually found that for this imaging modality, prosthetic valve endocarditis is clearly a niche area for PET-CT imaging as well as in certain cases of CIED-IE, whereas native valve endocarditis, it performs poorly. Therefore, we would argue that PET imaging has niche roles in prosthetic valve endocarditis and CIED-IE, and this imaging study should be performed and interpreted, obviously, at the center with experience and expertise.
I also want to add the flavor of multi-detector cardiac CT. Shown here is a patient who underwent prior tricuspid valve repair. As can be observed, there is a lobulated pocket of free-flowing contrast or blood between the tricuspid valve annulus and the aortic annulus. This is compatible with abscess or fistula formation. Again, CT scan very well assess periannular complications, as we demonstrated before.
We previously published on the diagnostic performance of TEE versus multi-detector CT in endocarditis. Again, this is one of the largest bodies of work on this topic. We showed that, as we know, TEE clearly performs superiorly to CT for vegetations as well as leaflet perforations. However, in terms of periannular complications, CT actually showed a very positive trend towards better sensitivity. And also, for prosthetic valve endocarditis, the pooled specificities for vegetations are significantly higher for multi-detector CT.
We also have published on the role of multimodality imaging, listing the relative strength and weaknesses of each modality ranging from transthoracic echo, transesophageal echo, to multidetector CT and PET imaging for tricuspid valve endocarditis, it's really important to be aware of your local expertise and to understand the strengths and limitations when applying a multimodality imaging approach.
In summary, echocardiography is the key first-line imaging modality for right-sided IE and tricuspid valve IE. TEE imaging should be considered if transthoracic imaging is non-diagnostic or if there are certain high-risk features. Contemporary TEE imaging with 3D imaging is incremental in diagnostic evaluation and we must remember that PET-CT and MDCT adjuvant advanced imaging modalities for certain niche cases. I would argue that multimodality cardiovascular imaging approach is increasingly important for the management of tricuspid valve endocarditis. Thank you very much.
Haytham Elgharably, MD:
My focus to talk about today is actually relapse after cardiac surgery, which I believe is the Achilles heel for the outcomes of these patients. Because you can do a perfect operation, we can set them up with antibiotics plan and a rehab plan for addiction, but then if they relapse, they may overdose or die or have recurrent endocarditis.
Our traditional approach to manage these patients, us and other centers, is to treat the endocarditis first, and then we have a discussion with the patient before surgery that they commit the addiction rehab plan we have to offer them after.
We'll take a look at the national outcomes or trend with that approach that has been going since 2002 to 2018 or '20. This first study is from our Endocarditis Center group published in 2019. They look at hospitalizations for patient with endocarditis associated with ejection drug use, and you can see in the blue color there, the Midwest state has actually the highest annual percentage change of these hospitalizations during that time period of the study. This is a different one for publish in JACC around the same time they look for readmission rate. From 2010 to 2015, you can see readmission rate has doubled in these patients with endocarditis associated with injection drug use. On the right side of the slide, these are the most common factors causing readmission. We can see drug relapse or abuse is the third most important factor for readmission within the first 3 months after cardiac surgery.
This is from STS surgical database. From 2011 to 2018, around more than 10,000 patients had cardiac valve surgery for endocarditis associated with injection drug use. And obviously we can see the increased trend during that time period of the study.
Now, important findings. They have that operative mortality and significant morbidity were significantly higher for patients who present for redo valve surgery obviously relapse after a prior valve surgery. This study that Dr. Gordon has presented, Dr. Shrestha's study published in 2015, we looked at the endocarditis cases done at the Cleveland Clinic from '07 to 2012: 41 patients had endocarditis surgery for injection drug use. And as Dr. Gordon pointed out, the first three to six months post-op rehab, patients who had history of injection drug use have 10 times higher risk for death or reoperation compared to patients without drug use, which probably related also to relapse.
If we capture all these observations from the prior study starting at 2002 into 2018, we can obviously see increasing trend of surgery for injection endocarditis either hospitalization readmission or surgical databases. We also observe higher death and recurrence of endocarditis. Higher death in these patients. They're all young, no significant comorbidities, so the most expected common cause of death is probably relapse. The same for recurrent endocarditis. Also, we observe worse outcomes for patients who are presented for re-do surgery. Probably these surgeries are also more complex, and that's also related to relapse. The third thing that was pointed out in the JACC study, that increased readmission during to drug relapse.
So, it's clear that drug relapse after cardiac surgery is a critical issue and we wanted to look at this in our own data, examine the incidence of relapse after cardiac surgery, the risk factors, and also outcomes in that context. We looked at our own patients from 2010 to 2020. We found more than 200 patients had undergone surgery for endocarditis with injection drug use during that time. The average age was 36 years old, and the mean follow up was around 3 years.
In that diagram we can see simply the fate of these patients who are under one operation. In the far right we still see good surgical outcome, so less than 2 percent hospitalization risk considering the complexity of these cases. At first glance, 25 percent had relapsed, and they didn't receive an offer for under reoperation. We have 26 patients who had reoperations, and out of these, 64 reoperations were due to drug relapse, 5 percent for suspicious drug relapse. So, we can say 70 percent, almost, of the reoperation was due to drug relapse. This drug relapse has a trend, which goes along, actually, with the studies that have been published before during the first year. There is an increased risk of drug relapse in the initial phase and then it plateaus out after the first year, which goes along with readmission or hospitalization risk in the first 3 to 6 months as showed up in other studies we just showed.
Survival is not that great. At 5 years, 64 percent died of these young patients after surgery. I think this is the most important critical figure we have of the study. I want to point out, first, that we did a lot of effort related by Mike Javorski here and the other residents to get follow up data from these patients. It's very difficult to call and catch these patients. So, at 5 years, our last follow-up was around a quarter of these patients. But with the available data we have, we can see here in the graph, after cardiac surgery at 5 years, almost 50 percent relapse, almost 30 percent died without the known of relapse, but it's most expected, again to be overdosed, and then the green line shows the patient who's alive without any documentation of relapse, which is maybe below 20 percent. In our analysis, we have found that the risk factors for relapse were younger age, injection heroin use and lower education level below high school diplomas.
At our Endocarditis Center, our effort has been focused on a multidisciplinary team approach for these patients. So, we get, obviously, cardiology, ID, us, psych addiction, social worker to see each of these patients before surgery and we formulate together this plan, surgery. Then we have the antibiotic plans, and then we have the addiction rehab plan. Where are they going to go after surgery? We have also developed a close follow-up program like MOSAIC, which Betsy is going to be talking about in a little bit.
The conclusion of these approaches that we have low operative mortality for these cases, but still, as we showed in the prior figure, we have higher relapse and late death rates among this young population. We are thinking together now is there's room for improvement, if there's something we need to do different to approach these patients. One approach that Dr. Gordon pointed out is a bridge to decision. So, we have patients who come in with active drug use but may not need urgent surgery, especially with patients on the right side, endocarditis. If we can formulate a plan to treat their addiction first, get it under control, and then do the surgery if they can wait, maybe that can increase the success rate and prevent relapse.
In this approach, possible candidates that we discussed together are isolated tricuspid valve endocarditis, obviously documented history of injection drug use, but they don't need urgent surgery. There is no indication for urgent surgery. And obviously these, it goes case by case with our multidisciplinary team discussion. So, they don't have ongoing sepsis that can be controlled by antibiotics, there is no right heart failure from the tricuspid regurgitation, there is no hardware that needs to come out, there is no large vegetation that poses risk for large PEs, and there is no PFO or left-sided endocarditis.
AngioVac is one of the options that Dr. Ghobrial will be talking about that may give us some time if we plan to do an upfront addiction rehab first, but eventually this patient is still going to come to do surgery. The last point, obviously, that the patient needs to agree for our rehab plan.
We would like to take this opportunity to share this with you guys, because you're going to be seeing this in the chart more and more as our group has been discussing these with patients. Thanks.
Thank you for listening. We hope you enjoyed the podcast. Like what you heard? Visit Tall Rounds online at clevelandclinic.org/tallrounds and subscribe for free access to more education on the go.
A Cleveland Clinic podcast exploring heart, vascular and thoracic topics of interest to healthcare providers: medical and surgical treatments, diagnostic testing, medical conditions, and research, technology and practice issues.