FormalPara Key Summary Points

Mural endocarditis is a rare but increasingly recognized health care-associated disease that requires a high index of suspicion and a multidisciplinary approach to avoid a relevant morbidity and mortality.

It is associated with intravascular catheter use and immunocompromised patients such as transplantation, cancer and hemodialysis patients.

Fungal etiology and large vegetations are common in mural endocarditis.

Introduction

Mural infective endocarditis (MIE) is a rare type of endovascular infection that involves the non-valvular endocardium [1,2,3,4,5,6]. Cases of MIE have been described in Candida and Aspergillus endocarditis [2, 7,8,9,10,11,12] as well as in infective endocarditis (IE) caused by other microorganisms more characteristic of valvular endocarditis (VIE) such as Staphylococcus aureus and Streptococcus. Others have described MIE in intravenous drug users or associated with catheter use [1, 5, 7, 13,14,15,16,17].

Since valves are not involved, diagnosis is often delayed and made only when complications such as systemic or pulmonary embolisms occur, conveying a significant morbidity and mortality. Prognosis of MIE is usually poor according to the literature [3, 5, 18].

Information about MIE comes from isolated case reports or short series of selected cases. The actual epidemiology, relevance and prognosis are unknown. Data on MIE management and outcome are lacking.

We present a series of MIEs from the prospective GAMES registry and a comprehensive review of the literature.

Methods

Study design

Between January 2008 and June 2018, consecutive patients with IE according to the Duke criteria were prospectively included in the “Spanish Collaboration on Endocarditis—Grupo de Apoyo al Manejo de la Endocarditis Infecciosa en España (GAMES)” registry, which is maintained by 35 Spanish hospitals. Multidisciplinary teams evaluated every case and completed standardized case report forms for each IE episode, including clinical, microbiologic and echocardiographic sections and follow-up data.

Literature search

Additionally, we performed a PubMed search that included the search term: “mural endocarditis.” We filtered case reports and case series and omitted cases that included valvular or intracardiac device involvement. Inappropriate and inaccessible posts were excluded. A standardized form adapted from that of the GAMES registry was filled in for each literature MIE with enough available information and included in a database. The list of cases included in the literature review is accessible in Supplementary material.

Definitions

We considered MIE cases with non-valvular endocardium involvement. Cases with both mural and valvular involvement were excluded from the analysis, as were cases of endovascular infection that could be considered endarteritis (involving great vessels, ductus arteriosus, etc.). We considered as non-mural endocarditis those cases of infective endocarditis that involved valvular endocardial surfaces [valvular endocarditis (VIE) or those with vegetations involving intra-cardiac devices (pacemakers, implantable cardioverter defibrillators) device-associated endocarditis (DIE)]. IE was defined according to the modified Duke criteria [19].

Age-adjusted Charlson Comorbidity Index was used to quantify comorbidities [20].

Place of acquisition of IE was defined following ICE recommendations [21].

The catheter was considered as the source of endocarditis when the etiologic microorganism was detected in both the catheter and the blood cultures or surgical samples (vegetation, valve or implantable device), or when, from a clinical standpoint, the treating physician considered the catheter was the source.

All patients were evaluated by cardiac surgeons to determine the need for surgery according to international indications [22]. The final decision on surgery was made in agreement with the multidisciplinary endocarditis team at each center.

Both in-hospital mortality (overall mortality rate during the hospital stay) and 1-year mortality were analyzed.

Data analysis

Cases with MIE were compared to cases with non-valvular endocarditis in the GAMES registry (both VIE and DIE). MIEs from the GAMES registry were compared to cases of MIE from the literature.

The quantitative variables were expressed as mean and standard deviation or median and interquartile range; qualitative variables were expressed as frequency and percentage. Continuous variables were compared using the t-test, and categorical variables were compared using the χ2 test or Fisher’s exact test when the χ2 test was not appropriate.

All statistical analyses were performed with SPSS version 25 software (IBM SPSS Statistics 22.0, Armonk, New York, NY: IBM Corp.).

Ethics

The study and the common case report form were approved by the local and national institutional review boards and ethics committees (Comité ético de Investigación Clínica Regional de la Comunidad de Madrid CEIC-R; EC 18/07; date 11/01/2008).

Results

During the study period 3767 patients were included. Of them, 27 MIE (0.7% of total endocarditis) and 3649 non-mural endocarditis [407 (11.1%) DIE and 3242 (88.2%) VIE] were registered in the GAMES database. Moreover, 36 cases from the literature review that provided enough information to analyze were selected.

Differences between mural and non-mural endocarditis in the GAMES registry

General characteristics of mural (Fig. 1) and non-mural endocarditis from the GAMES registry are displayed in Table 1.

Fig. 1
figure 1

Simultaneous bi-plane imaging by transesophageal echo showing a large sessile heterogeneous mass (3.9 × 2.8 cm) attached to the lateral and inferior wall of the right atrium, with irregular contour. The tricuspid valve and inferior vena cava were not affected

Full size image
Table 1 Infective endocarditis in the GAMES registry: mural compared to valvular and device-associated endocarditis
Full size table

Compared to non-MIE, patients with MIE were significantly younger, and MIE was more prevalent in women. Patients with MIE presented a lower age-adjusted Charlson Comorbidity Index, were receiving hemodialysis more often and transplantation was more frequent as underlying disease.

Furthermore, in contrast with VIEs, which were mainly community-acquired (61.7%), the majority of MIEs were hospital-acquired (63%). DIEs were more common in a nosocomial setting and related to medical care.

A striking difference was found in the proportion of MIEs originating from the catheter, which was very high (59.3%) compared to non-MIEs with catheter source (8.8% and 9.7% in DIE and VIE, respectively) (p < 0.01).

The microorganisms most frequently implicated in endocarditis in all groups were Staphylococci, with a prevalence in MIEs of 29.6%, being 14.8% S. aureus and S. coagulase negative each, followed by Streptococci (25.9%). Streptococci were significantly more frequent in MIE and VIE (27.5%) than in DIE (8.1%). Significant differences were found between groups regarding Candida etiology (22.2% among MIE versus 2% in DIE and 1.2% in VIE, p < 0.01). Aspergillus was more common in MIE, though without statistical significance (3.7% among MIE vs. 0.2% in DIE and 0.3% in VIE, p = 0.144).

The vegetations were significantly larger in MIE, with a mean size of 22 mm compared to 12 and 10 mm in the DIE and VIE groups, respectively (p < 0.01).

Regarding complications associated with endocarditis, presence of intra-cardiac complications was significantly higher in VIE with 33.9% compared to 3.7% in MIE and in DIE each (p < 0.01). Likewise, systemic complications such as heart failure or central nervous system involvement were more frequent in VIE (p < 0.01 and p = 0.041); however, no differences were found in other complications such as acute renal failure, sepsis or septic shock.

Surgical indications were the same for all types of endocarditis and included: severe heart failure, septic shock, recurrent emboli, severe valve dysfunction, intracardiac complications (abscess), prosthetic valve infection, and large mobile vegetations or persistent bacteremia despite adequate antibiotic therapy. Non-MIEs more often had a surgical indication (65.3% in VIE and 82.8% in DIE vs. 33.3% in MIE, p < 0.01) and were operated on (42.6% in VIE and 77.1% in DIE vs. 18.5% in mural IE, p < 0.01).

No significant differences were found regarding in-hospital or 1-year mortality after the event.

Characteristics and differences between mural endocarditis from the GAMES registry and published cases of mural endocarditis

We retrieved 44 cases of MIE in the literature search, but only 36 had exclusive non-valvular involvement and enough information to analyze. General characteristics of MIE from GAMES and MIE published in the literature are given in Table 2. The column on the right side of the table summarizes the characteristics of both combined.

Table 2 Mural infective endocarditis: GAMES registry versus literature review
Full size table

The time span for the collection of cases from the literature (1979–2019) was much broader than the study period of cases obtained from the GAMES registry (2008–2018), including a high proportion of cases (33.3%) diagnosed before the beginning of the GAMES registry. Figure 2 shows an increasing incidence of MIE along time.

Fig. 2
figure 2

Distribution by calendar year of the cases of mural endocarditis from the literature (blue) and those from the GAMES registry (orange)

Full size image

Both series share some important characteristics such as a high proportion of fungal endocarditis and the presence of big vegetations. However, we found several relevant differences.

Patients in the literature group were younger than those in the GAMES series, with a mean age of 42.6 years versus 56.4 years (p < 0.001), had less frequent comorbidities as shown by a lower Charlson Index (1.3 literature vs. GAMES 4.3, p = 0.000). Of note, cases from the literature more frequently were users of intravenous drugs as a risk behavior for endocarditis compared to those from the GAMES registry (25% literature vs. 3.7% GAMES, p = 0.034). Acquisition of endocarditis in the community was more frequent in cases from the literature (69.4% vs. 25.9%) compared to those from GAMES, in which a nosocomial origin predominated (63% vs. 13.9%, p = 0,0001). Likewise, catheter as the source for endocarditis was more frequent in the GAMES series (59.3% vs. 13.9% in literature, p = 0.000). Regarding the etiology of endocarditis, S. aureus was more frequent in the literature group (50% vs. 14.8% in GAMES, p = 0.007) whereas coagulase-negative staphylococci were found in GAMES series but not in the literature series (14.8% vs. 0% in literature, p = 0.035), with significant differences. The most frequent location in the case of MIE in the literature was the left ventricle (47.2%), while in GAMES it was the right atrium (51.9%) (Fig. 3). Complications such as the presence of systemic embolisms and CNS involvement were more frequent in the literature group.

Fig. 3
figure 3

Location of mural endocarditis. a: Septum. b: Vena cava. c: Tendinous chords. A GAMES, B literature

Full size image

No significant differences were found regarding surgical indication, proportion of cases that received surgery or mortality.

Discussion

Our results show that MIE represents 0.7% of endocarditis. It is often a complication of catheter use that appears in immunocompromised hospitalized patients, and fungal etiology is common. Surgery is performed less frequently than in non-MIE, although mortality is similar to that of other types of endocarditis.

Our series provides information about MIE prevalence among total cases of endocarditis. There is no prior published large series of MIE. MIE is a rare disease, which has become easier to diagnose with the improvement in cardiac imaging techniques, which can explain in part an increasing incidence (Fig. 2). In view of the lack of prior MIE series, we decided to perform a literature review. The period of the literature review encompasses a larger period than the GAMES registry and, as such, includes earlier cases. When analyzing cases from the literature review, publication bias must be taken into consideration. Information from the GAMES together with the literature review offers a broad picture of the spectrum of MIE.

The combined data show that underlying transplantation, neoplasia and hemodialysis are prevalent in patients with MIE. The greater use of catheters in this type of subjects and a high use of hospital care could explain this finding. However, in the literature review, we observed a high proportion of community-acquired cases with a different clinical phenotype: intravenous drug users with S. aureus endocarditis. The high prevalence of parenteral drug users in case reports collected in the literature review may be related to the period in which the cases were described, a time when this activity was more common than nowadays. This suggests a shift in the epidemiology of MIE along time, with a decrease in IVDUs in favor of nosocomial endocarditis in which catheters appear to be implicated in the pathogenesis of MIE. The same applies to the change in the location of the vegetations in MIE [23, 24]. In contrast to VIE, where valvular endocardium is extensively exposed to damage from the bloodstream, we hypothesize that mural endocardium requires external damage, for instance, by the catheter jet, to start the process of developing a vegetation as has been described in other cases with endocardial damage caused by a jet [4, 17, 25,26,27].

In MIE, vegetations were larger compared to non-MIE. Regarding large vegetations, MIE diagnosis in earlier cases from the literature series often came only after embolism had already occurred. As the valvular apparatus is spared, heart failure symptoms cannot be expected to unveil the presence of endocarditis. The delay in the diagnosis of MIE implies a longer time of growth for the vegetations. The diagnostic delay in older cases, in keeping with worse imaging techniques, can explain a higher prevalence of embolisms and CNS involvement in the literature cases together with publication bias. It could be expected that, due to the larger size of the vegetations, significant differences between MIE and non-MIE regarding persistent bacteremia and systemic complications of endocarditis such as embolisms would be encountered; however, we did not find such differences.

Regarding etiology, the most common pathogens in MIE were Staphylococci and Streptococci, but we found a striking prevalence of fungi compared to non-MIE [23]. Fungal endocarditis was represented mainly by Candida and, to a lesser extent, by Aspergillus [2, 7,8,9,10,11,12]. In earlier cases of endocarditis associated with intravenous drug use, Candida has been related to injection of brown heroin diluted in fresh lemon juice [28]. In addition, candidemia is also prevalent among catheter users, leading to the development of endocarditis in the long term. Fungal endocarditis also develops in immunocompromised hosts; in particular, Aspergillus spp. endocarditis has been associated with transplantation [29]. Fungal IE often cause large vegetations [8, 12, 30, 31].

Heart failure and valvular dysfunction are relevant indications for surgery. We consider that, since MIE does not involve heart valves, these complications are less frequent, which is a possible explanation for the surgery not being indicated in as many cases of MIE as it was in other types of endocarditis. Although cardiac complications of MIE were less frequent and surgery was less commonly indicated, mortality was very similar to that of non-MIE, which underlines the importance of MIE diagnosis and correct management.

Strengths and limitations

The present study analyzes a multicenter series from the entire Spanish territory and over a long period of time and is complemented with a review of the literature. We believe it presents a comprehensive image of MIE. Cases from the GAMES registry are collected prospectively, according to pre-established criteria, and evaluated by a multidisciplinary team, which allows putting the prevalence of this rare disease in perspective and provides a broad set of data to describe this population. Regarding series of cases from the literature, incidence or prevalence data cannot be obtained, as the total population with endocarditis is unknown, and publication bias has to be considered, so that information obtained from these cases must be evaluated with caution. Information about published cases is seldom complete. Nevertheless, that series covers a large period and is a thorough sample, which provides very valuable information to enrich the information about MIE.

Conclusion

Mural endocarditis is a rare but increasingly recognized and reported health care-associated disease that requires a high index of suspicion and a multidisciplinary approach to avoid the relevant morbidity and mortality.

Our results highlight the importance of this type of endocarditis in association with intravascular catheters and immunocompromised patients and, as such, with a potential for prevention.