Right-sided infective endocarditis (RSIE) most commonly occurs in the setting of intravenous drug use (IVDU) and has taken on increasing surgical significance as the opioid crisis has driven increased abuse of intravenous heroin [ , ]. Surgery for RSIE, which has historically received less attention than left-sided infective endocarditis (LSIE), poses unique challenges because of the high rate of recidivism and recurrent endocarditis in IVD users [ , , ]. Surgical approaches to RSIE must balance competing priorities of maximizing efficacy through effective valve reconstruction and minimizing risk of IE recurrence by limiting implantation of prosthetic material. In this chapter, we will outline distinctive features of RSIE with respect to epidemiology, clinical presentation, and diagnosis and we will define indications for surgical intervention. We will also discuss the major surgical techniques for RSIE and important pre-, intra-, and postoperative considerations.
RSIE is significantly less common than LSIE, comprising only 5%–10% of total infective endocarditis cases [ , , ]. While RSIE can occur anywhere in the endocardium of the right heart, it displays a 90% predilection for the tricuspid valve with infections of cardiac implantable electronic devices (CIEDs) and the pulmonic valve rounding out the remaining 10% [ ]. The most common risk factor for RSIE is IVDU [ ]. In one study, 34.5% of RSIE patients had a history of IVDU [ ]. Furthermore, even among IVD users, HIV-positive IVD users suffer an odds ratio of 2.31 of developing IE compared to HIV-negative IVD users [ ]. CIEDs, such as pacemakers sand implantable cardioverter defibrillators, and central venous catheters serve as nidi for bacterial colonization and represent alternate mechanisms for the development of RSIE [ ]. Uncorrected congenital abnormalities of the right heart independently predispose patients to RSIE [ ]. With respect to causative microbiological agents, Staphylococcus aureus accounts for 60%–90% of RSIE, but the incidences of methicillin-resistant S. aureus , Pseudomonas aeruginosa , and polymicrobial RSIE are all rising [ , ]. In RSIE related to prosthetic valves and CIEDs, coagulase-negative Staphylococcus species are significant contributors [ , ].
Epidemiological concerns specific to surgery center around the increasing surgical incidence of RSIE and demographic differences between RSIE and LSIE patients. Overall, 5%–40% of RSIE cases require surgical intervention [ , ]. However, IVDU, particularly injection heroin, has more than doubled during the opioid epidemic and with this upsurge, the incidence of IE has increased significantly [ , ]. Additionally, CIEDs have become increasingly common, particularly in the elderly, which has prompted a similar increase in IE [ ]. These increases in the incidence of IE have led to a predictable corresponding rise in the number of required surgical interventions [ ]. In the span of just 6 years (2011–17), the surgical incidence of tricuspid endocarditis increased fivefold [ , ]. The demographic differences between RSIE and LSIE are also a reflection of the IVDU associated with RSIE. RSIE patients tend to be younger, male, and less burdened by comorbidities, reflecting the characteristics of the users of intravenous drugs [ , , ]. The high percentage of recidivism among IVD users (27%) also contributes to a greater percentage of recurrent endocarditis (12%–32%) in RSIE patients compared to LSIE patients [ , , , ]. Lastly, surgeons should be cognizant, both when evaluating imaging studies and performing repair and replacement operations, that 37.9% of RSIE cases are complicated by concomitant LSIE [ ].
RSIE typically presents with constitutional symptoms, chiefly fever, and respiratory symptoms resulting from septic pulmonary emboli [ , , ]. However, because the murmurs of RSIE are often subtle and nonspecific, diagnosis of RSIE can be delayed resulting in a later presentation with a more severe respiratory picture of pleural effusion, hemoptysis, pneumothorax, or pulmonary infarction, abscess, and empyema [ , ]. Cardiac manifestations of RSIE are a function of the severity of tricuspid valve involvement. Vegetations causing severe tricuspid regurgitation can lead to right atrial dilation with the potential for resultant supraventricular arrhythmias and progression to heart failure [ , ]. Systemic emboli with destination-dependent symptoms are possible if a patent foramen ovale is present but are much less common than in LSIE [ , ].
RSIE presents several unique diagnostic challenges. Infectious endocarditis, including RSIE, can be diagnosed by pathological or clinical criteria [ ]. The Duke criteria, long considered the gold standard for diagnosis of IE, define as major clinical criteria (1) persistently positive blood cultures and (2) echocardiographic evidence of endocardial involvement [ ]. However, the Duke criteria were principally designed for LSIE and minor clinical criteria such as immunologic and vascular phenomena are less applicable to RSIE [ ]. Proposed modifications to the Duke criteria emphasizing the role of transesophageal echocardiography (TEE) have increased the capability of the criteria to detect RSIE [ ]. Transthoracic echocardiography (TTE) has traditionally been the imaging modality of choice for suspected RSIE because of the proximity of the transducer to anteriorly situated right heart structures [ ]. However, multiple studies have demonstrated comparable or improved sensitivity of TEE compared to TTE in the diagnosis of RSIE [ ]. The benefit of TEE relative to TTE is most clearly established in CIED-related RSIE where TTE provides especially poor sensitivity, especially of pacemaker leads situated on the roof of the right atrium [ ]. TEE has been shown to increase sensitivity to CIED-related RSIE by as much as 60% and is therefore mandated in such cases [ , , ]. Intracardiac echocardiography is useful in the setting of inconclusive TTE and TEE and offers a high sensitivity but comes at the cost of significantly increased invasiveness [ ]. Computerized tomography (CT) finds application in RSIE for detecting septic pulmonary infarcts and abscesses [ ].
The foundation of medical management of RSIE is intravenous antibiotic therapy [ ]. Blood cultures should be obtained as soon as clinical suspicion for IE arises and should be repeated at 30-minute intervals to a total of three sets [ ]. In suspected RSIE patients with hemodynamic instability, empiric IV antibiotic therapy should be initiated immediately following blood culture draws [ ]. Empiric coverage should always include S. aureus and can be further guided in the IVD user by the type of drug and solvent abused as well as the location of the infection [ , , ]. Following culture and sensitivity results, antibiotics should be tailored to the causative organism [ , ]. Appropriate antibiotic therapy is effective in clearing bacteremia in 70%–85% of cases of tricuspid valve infective endocarditis (TVIE) [ ]. However, additional intervention may be necessary to address satellite sources of bacteremia, for example, lung decortication to treat pleural abscesses or empyema resulting from septic pulmonary emboli [ ]. The authors therefore advocate a scaled invasiveness approach to treatment of RSIE, with valve surgery reserved for cases in which less invasive procedures have failed to eliminate bacteremia.
Indications for surgery
Surgery for RSIE may be indicated by failure of medical management, risk of future adverse events, deteriorating patient condition, or a combination thereof. In the case of persistent fever or bacteremia 5–7 days following initiation of appropriate antibiotic therapy, surgery is indicated to remove the source vegetation [ , , , ]. Surgery in this case has the additional benefits of disrupting biofilms and exposing residual live organisms to on-going antibiotic therapy and immune response [ ]. Similarly, surgery may be required to treat cases of RSIE caused by organisms, such as fungi, that are difficult to eradicate by antimicrobial medications alone [ , , ]. Size and mobility of vegetations are also important factors dictating surgical intervention. Vegetations larger than 2 cm with high mobility and corresponding risk for embolization should be surgically excised [ , , , ]. While septic embolization to the lungs is an undesirable complication of RSIE, the literature and author experience suggest that pulmonary emboli should not be an indication for surgery in and of themselves [ ]. That said, recurrent septic pulmonary emboli can lead to increased pulmonary vascular resistance, reducing the ability of the right heart to pump out excess volume from tricuspid regurgitation and exacerbating regurgitation-based heart failure that may already be developing [ ].
Heart failure from severe tricuspid regurgitation is an independent indicator for surgery, particularly when refractory to diuretics [ , , , ]. Perivalvular abscesses and destructive penetrating lesions which are more commonly seen in prosthetic RSIE also require surgical intervention [ , , ]. Furthermore, if endocarditis or associated inflammation spread into the conduction system of the right heart, heart block can necessitate surgery for both vegetation resection as well as potential pacemaker implantation [ , ]. When concomitant LSIE is present, typically the circumstances of the LSIE dictate the need for surgery [ , ]. Once the need for surgery has been established, the operation should be performed within 48 h as early surgery has been demonstrated to improve outcomes [ , ].
Once it has been decided that surgery is indicated for the treatment of RSIE, the appropriate surgical technique must be selected. Three main surgical approaches exist: valvectomy, valve repair, and valve replacement. Because the tricuspid valve is the most commonly involved structure in RSIE, this section will focus on surgical interventions for TVIE.
In a tricuspid valvectomy, the tricuspid valve leaflets and chordae tendinae as well as all associated vegetations are excised [ ]. Valvectomy can be performed as a destination procedure or as a bridge to valve replacement following a period of IVDU abstinence [ ]. The chief advantage of valvectomy over valve repair is that valvectomy eliminates the valve as a future nidus for endocarditis. Nor does valvectomy create a new prosthetic nidus for reinfection as does valve replacement. The benefit of reducing the surface area at high-risk for recurrent infection is most apparent in IVD users given their high rate of recidivism and recurrent endocarditis [ , ]. When compared to valve replacement in a recent metaanalysis, valvectomy did reduce the incidence of recurrent endocarditis, but the trend was not significant [ ]. Valvectomy has the additional advantage of being significantly less likely than repair and replacement to cause heart block [ ]. Furthermore, valvectomy eliminates the requirement for postoperative anticoagulation therapy to which IVD users may be nonadherent [ , ].
The advantages of valvectomy are offset by its complication profile. Predictably, valvectomy recipients are likely to experience tricuspid regurgitation [ ]. However, heart failure only occurs postoperatively in 27% of patients, which some authors attribute to cardiac compensation that may occur in response to long-standing endocarditis-related tricuspid regurgitation [ , ]. In the setting of pulmonary hypertension, however, increased pulmonary vascular resistance leads to intolerable tricuspid regurgitation, and therefore, these patients are not well suited for valvectomy [ , ]. Valvectomy recipients experience between 12% and 13% early mortality which, despite valvectomy being older and less technically advanced, is not significantly increased compared to valve repair and replacement [ , ].
In TVIE patients requiring surgery, valve repair is the recommended operative approach [ , , , ]. Vegectomy without repair is utilized when possible to minimize implantation of prosthetic material which has the potential to promote thrombosis and/or reinfection [ ]. However, when debridement of vegetations requires removal of enough tissue to cause significant regurgitation, repair is indicated [ , ]. Tricuspid valve repair has three key components: (1) patch placement, (2) annuloplasty, and (3) chordae tendinae reconstruction [ ]. Multiple variations of each component exist, with the location and extent of tissue removal during vegectomy dictating which variations are selected [ ].
Autologous pericardial patch repair is first line for vegetations affecting a single leaflet of the tricuspid valve, particularly those that are small and can be circumferentially excised [ , ]. Patch size should be large enough so that no sutures are placed within compromised tissue [ ]. For the same purpose, debridement should, when possible, aim to preserve a margin of healthy leaflet [ ].
Manipulation of the annulus may be required for more extensive vegetative damage of the tricuspid valve. When near-complete resection of the posterior leaflet is necessary, Kay bicuspidization (annuloplasty), with patch placement as necessary, offers a nonprosthetic alternative to ring annuloplasty [ , , ]. In Kay bicuspidization, the annulus corresponding to the posterior leaflet is sutured closed, reshaping the annulus to allow cooptation of the anterior and septal leaflets during systole [ , , , ]. If instead the anterior or septal leaflet is resected, De Vega annuloplasty can be employed [ , , , ]. In the De Vega procedure, a purse string suture is run counterclockwise through the tricuspid annulus from the postero-septal commissure to the antero-septal commissure [ , , , ]. Tightening of the purse strings narrows the annulus allowing cooptation of the remaining leaflets [ , ]. Lastly with respect to annuloplasty, a prosthetic ring can be sutured into the annulus, which restores the shape of the annulus and, in conjunction with patch repair, reduces regurgitation [ ]. Concern over the potential for increased infection in prosthetic ring annuloplasty has not been realized in comparative studies with suture annuloplasty [ , ].
Chordae tendinae reconstruction in tricuspid valve repair helps to more closely approximate native valve function. Artificial chordae, also known as neochords, were designed to replicate the functionality of the chordae tendinae [ ]. Composed of expanded polytetrafluoroethylene, neochord sutures are anchored in the papillary muscles of the right ventricle and secured to patches placed during repair of the tricuspid leaflets [ , , ]. A saline infusion test is performed to simulate systole and valve closure, thereby allowing the surgeons to determine the appropriate length for the neochords [ ]. In a cohort of 12 patients with severe tricuspid valve endocarditis, Tarola and colleagues demonstrated effective regurgitation reduction using patch repair and neochords [ ].
The high rate of recurrent endocarditis in IVD users and the significant mortality associated with prosthetic valve endocarditis underscore the advantages of avoiding prostheses in RSIE surgery [ , , , ]. However, endocarditis damage to the tricuspid valve may be so severe in some cases that repair is not possible and replacement is required [ ]. Replacement can also be considered for patients in whom repair has failed [ ]. Additionally, tricuspid valve repairs rapidly become highly complex, requiring significant surgical experience and expertise that may not be available at all institutions [ ]. Thus, although rarely ideal, replacement plays a significant role in surgical management of tricuspid valve endocarditis.
The surgical technique for tricuspid valve replacement is essentially conserved regardless of whether the replacement valve is bioprosthetic or mechanical [ ]. Prior to insertion of the prosthesis, native valve leaflets, including any vegetations, should be excised in order to both remove the source of endocarditis and prevent future right ventricular outflow obstruction (RVOT) [ ]. During excision, a rim of septal leaflet should be preserved [ , ]. Preservation of the septal leaflet rim allows sutures to be placed in purely leaflet tissue at the apex of the triangle of Koch, thereby reducing the potential for sutures to cause AV block [ ]. Any remaining chordae tendinae should be transected close to their papillary muscle attachments [ ]. Appropriate sizing and orientation of the prosthesis are critical to prevent RVOT and/or damage to the interventricular septum by the replacement valve [ , ]. As an additional antiseptic measure, the prosthesis may be soaked in and the annulus swabbed with antibiotic solution [ ].
Considerations of physiology, complications, and compliance inform the choice between prosthetic and mechanical replacement valves for tricuspid endocarditis. Compared to the left heart, the right heart operates at lower pressure, lower flow velocities, and lower levels of antithrombotic prostacyclins [ , ]. Collectively, these factors contribute to concern for increased risk of mechanical valve thrombosis in the right heart [ , ]. Additional thrombosis concern stems from potentially poor compliance with postoperative anticoagulation therapy in the IVD user population [ ]. In corroboration of these concerns, a recent metaanalysis demonstrated increased frequency of thrombotic events in mechanical valve recipients [ ]. However, the significance of increased thrombosis in mechanical prostheses has been challenged because it does not lead to an increased rate of 5-year valve failure or reoperation compared to bioprosthesis and because thrombosis in this context can be successfully medically treated [ , ]. Also pertinent to the IVDU population, no significant difference has been found in recurrent endocarditis rates between bioprosthetic and mechanical tricuspid valve prostheses [ ]. With respect to survival, a recent metaanalysis found no significant difference in short- or long-term mortality between bioprosthetic and mechanical tricuspid valve replacement. [ ] One additional consideration is that the design of mechanical valves does not allow for pacing leads to enter the right ventricle and thus bioprostheses are a better selection for patients with conditions that predispose them to future pacemaker implantation [ ].
Repair versus replacement
As the two predominant surgical approaches to tricuspid valve infectious endocarditis, valve repair and valve replacement warrant a comparison with respect to key outcomes variables. In patients requiring surgical intervention for RSIE, repair was performed in 49%–59% of cases and replacement in 41%–46% of cases [ , ]. In these RSIE cases, no significant difference has been found between repair and replacement in perioperative mortality (6% vs. 11%) or late survival or in the rate of major complications (17% vs. 25%) [ , ]. However, moderate-to-severe postoperative regurgitation was more likely in repair recipients [ ]. The increased rate of regurgitation in repair is offset by a significantly decreased likelihood of recurrent endocarditis, reoperation, heart block, and pacemaker implantation compared to replacement [ , ].
Arrested heart versus beating heart
Tricuspid valve repair and valve replacement are performed under cardiopulmonary bypass [ ]. Both repair and replacement can be done on either a beating heart or a heart arrested by aortic cross-clamp, even in the setting of a patent foramen ovale [ ]. The beating heart technique has the advantages of reducing myocardial ischemia and reperfusion injury and allowing immediate detection and removal of sutures that cause AV block [ , ]. Benefits of the arrested heart technique revolve principally around improved valve visualization and a stationary operative field [ ]. The authors prioritize ischemia minimization and so prefer to perform both repair and replacement under beating heart conditions. With respect to outcomes, beating heart and arrested heart tricuspid repair/replacement have similar operative and short-term mortality [ ]. Intraoperatively, most studies found no significant difference in time on cardiopulmonary bypass or operation time overall [ ]. Heart block was slightly more common with the arrested heart technique as was the need for a permanent pacemaker [ ].
Cardiac implantable electronic device–related infective endocarditis
CIED-related infective endocarditis is most commonly due to infection of permanent pacemakers or implantable cardioverter-defibrillators [ ]. CIED infection describes a microorganism colonization limited to the device generator pocket [ , ]. CIED infection is distinct from, but may progress to, CIED-related endocarditis which implies vegetations on device leads and/or the tricuspid valve [ ]. This is a pertinent distinction as surgical management is different and more complex for CIED-related endocarditis [ , ]. While CIED infection is generally amenable to percutaneous transvenous lead extraction, CIED-related endocarditis is addressed surgically in order to perform concomitant tricuspid valve repair and to reduce the risk of vegetation embolization to the lungs [ , ]. CIED-related infective endocarditis dictates complete removal of all components of the implanted device in order to prevent retained equipment from serving as a nidus for recurrent endocarditis [ ]. Interestingly, because only one-third to one-half of patients will require a procedure to replace the explanted device, the operation for CIED-related infective endocarditis is an opportunity to reevaluate the cardiac condition that originally precipitated device implantation [ , ].
Echocardiography, both transthoracic and transesophageal, is critical for operative planning. Skilled use of echocardiography can locate vegetations, anticipate structure involvement, identify the presence of a patent foramen ovale, and provide preliminary guidance as to whether repair is possible based on the extent of damage to the valve [ , ]. If echocardiography demonstrates mild RSIE, a minimally invasive approach can be considered [ ]. In light of the prevalence of IVDU in RSIE, it is also common practice at some institutions to require preoperative counseling wherein the patient must sign a contract promising to abstain from future IVDU and to enroll in an addiction therapy program [ ]. This practice, although controversial, highlights the importance of addiction therapy as an adjunct to medical and surgical treatment in the management of RSIE [ , , ]. It is also recommended that the patient receive aggressive preoperative diuresis to control tricuspid regurgitation and it is imperative that the patient be receiving antibiotics at the time of operation, ideally initiated at least 1 week prior to surgery [ , ]. Specific to the case of CIED-related endocarditis, preoperative consideration must be given to interim replacement of device functionality after the device is extracted [ , ].
A full sternotomy is the approach of choice for endocarditis surgery, especially if involvement of adjacent structures or concomitant left-sided lesions are suspected. As with any valvular surgery, intraoperative TEE is a necessity for assessing repair competency and paravalvular leak [ ]. In addition, intraoperative TEE can be used to search for valvular vegetations that may not have been identified preoperatively [ ]. Careful and thorough debridement is required to remove infected and necrotic tissue without damaging viable tissue or the conduction system [ ]. However, because damage to the conduction system is possible, all patients should receive temporary pacing wires. If there is any sign of heart block during the procedure, permanent epicardial leads should be implanted with a tunnel to a subcutaneous pocket where an electrophysiologist can place a pacemaker postoperatively as needed [ ].
It is important to anticipate and plan for hemodynamic swings that can occur as debridement of infected tissue releases bacterial toxins into the circulation. The authors recommend having high doses of vasopressors quickly accessible to combat septic shock. Debridement can also shower the lungs with septic emboli making it difficult to wean the patient off of cardiopulmonary bypass. For this reason, the authors prepare cannulation sites in the right jugular and right femoral veins in case emergent venous–venous extracorporeal membrane oxygenation (V–V ECMO) is required. After debridement is completed and before reconstruction of the valve, copious irrigation, replacement of all instruments, and glove exchange are necessary to ensure sterile conditions.
Heart block is a major complication of RSIE surgery and one that must be monitored closely postoperatively. Inflammation from surgical trauma can cause swelling of the annulus which may progress to heart block within 48 h following surgery. This type of heart block will not necessarily present during surgery and hence justifies the placement of temporary pacing leads in all patients. It is common for IVD-addicted patients to experience hypertension and tachycardia postoperatively from the combination of pain and reduced efficacy of narcotics due to opioid tolerance. Beta blockers should not be used to control hypertension due to the risk of exacerbating heart block. Other agents that act at the AV node, such as amiodarone, should be avoided as well. Standard postoperative antibiotic therapy lasts for 6-week and should be tailored based on culture and sensitivity results. For patients who received surgery for fungal endocarditis, most authors recommend life-long suppression therapy [ ].
RSIE has traditionally received less attention than LSIE because it composes a comparatively small percentage of IE cases. However, increasing rates of IVDU and cardiac device implantation have imparted greater significance to surgical management of RSIE. Surgery for RSIE is indicated following failure of antibiotics to eliminate bacteremia, echocardiographic evidence of large vegetations, or heart failure resulting from severe tricuspid regurgitation. Concerns regarding recurrent endocarditis dictate a surgical approach conscious to minimize the use of prosthetic materials. For this reason, valve repair is performed in nearly 60% of RSIE operations. Replacement may be the only option for severe valve destruction and, while the increased risk of recurrent endocarditis is realized, survival following replacement is comparable to survival following repair. Surgical management of RSIE utilizes TEE to guide planning in the preoperative period, vasopressive and mechanical support to control anticipated hemodynamic swings in the intraoperative period, and proactive temporary pacing to manage potential development of heart block in the postoperative period.