1. Introduction
Infective endocarditis (IE) is a rare disease, but its incidence seems to be increasing, especially in older adults [
1]. This seems to be related to more frequent healthcare contacts and a higher prevalence of cardiovascular implantable devices and prosthetic heart valves [
2]. Healthcare-associated and nosocomial IE are frequently due to multi-resistant pathogens [
3]. This fact, and the high prevalence of comorbidities in the elderly [
4], explains why IE in advanced age has a poor prognosis [
5] and frequent functional impairment [
6]. This is particularly true for frail elderly in patients with previous comorbidities [
7].
IE management frequently differs according to age group, with a low rate of cardiac surgery in octogenarians [
8]. This fact is explained, in part, to the increase in surgical risk seen in advanced ages [
9]. In patients undergoing surgery for IE, mortality increases significantly with age and perioperative mortality is about 20% in patients > 75 years [
10]. In some cases, as in mitral valve surgery and multiple valve interventions, surgical mortality is even higher [
10].
However, the underuse of cardiac surgery is associated with adverse outcomes in older patients with IE [
11]. It is unclear if this is mainly due to patient selection and to the influence of comorbidities and functional status in the prognosis but recent data suggest that surgery is underused in elderly patients [
12]. As elderly patients are increasingly afflicted with IE, the decision whether to perform surgery or not is very relevant. Surgery should not be denied on the basis of age alone. Shared decision making and experienced multidisciplinary teams’ evaluation are essential in these complex patients, and we need new data regarding the benefit of IE surgery in the elderly.
The purpose of our study is to describe the current profile of IE in octogenarians and to analyze the prognostic impact of baseline comorbidities in this population.
3. Results
A total of 726 patients ≥ 80 years with definite endocarditis were included, 357 (49%) with CCI ≥ 3 and 369 (51%) with CCI < 3 (
Table 1). Compared with the group of low comorbidity, octogenarians with CCI ≥ 3 were more frequently male, had a higher rate of
Enterococcus spp. and lower rate of gram-negative bacilli, and had a higher prevalence of nosocomial/health-care related IE. Surgery indication, surgery, and hospital survival were less common in patients with high comorbidity.
A total of 443 patients had surgical indication, but surgery was only performed in 176 (39.7%) patients, mainly due to a high risk profile with high estimated surgical risk (
Table 2). The impact on the surgical performance of baseline conditions of the patients with indication for cardiac surgery is shown in
Table 3. Compared with patients treated with surgery, those with surgical indication and conservative management were older (83.7 ± 3.4 vs. 82.2 ± 2.3 years,
p < 0.001), had more common mitral valve location (137 (51.3%) vs. 52 (29.5%),
p < 0.001) and presented more frequently CCI ≥ 3 points (140 (52.4%) vs. 66 (37.5%),
p = 0.002). Patients with surgical indication treated conservatively had higher mortality than those treated with surgery (in-hospital mortality: 147 (55.1%) vs. 55 (31.3%),
p < 0.001), (1-year mortality: 172 (64.4%) vs. 68 (38.6%),
p < 0.001).
In the whole sample, a total of 265 patients (36.6%) died during hospital admission. The effect of baseline conditions in in-hospital mortality is shown in
Table 4. CCI was an independent predictor of in-hospital mortality (
Table 5).
During 1-year follow-up, 338 (45.5%) patients died. CCI was an independent predictor of in-hospital mortality (
Table 6 and
Figure 1).
4. Discussion
Our main finding is that half of octogenarians with IE had high comorbidity and that a CCI ≥ 3 was a strong predictor of mortality. Our data also suggest that the underperformance of cardiac surgery in this group of patients might have a role in their poor prognosis.
Octogenarians represent a heterogeneous group but usually present an elevated prevalence of predisposing IE events such as previous interventions and recurring health care contacts [
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19]. These factors increase the risk of bacteremia with resistant microorganisms [
20]. This is particularly true for older patients with comorbidities, as our data show.
The most frequent microorganisms in our sample were
Enterococcus spp., especially in octogenarians with high comorbidity, and
Streptococcus spp. in accordance with previous findings [
21]. Concomitant diseases like cancer, and associated interventions might explain some of the etiology differences according to the presence of comorbidity.
Enterococcus spp., especially
Enterococcus faecalis, is an important cause of IE. This etiology is frequently related to digestive tract conditions, cardiac devices implantation, and vascular access. All these risk factors are more common in the elderly, for instance, the average age at the time of diagnosis for colon cancer is 70 years. The prevalence of enterococcal IE is increasing in recent decades and this is mainly due to population ageing, the increasing number of health care-associated interventions, and microbiological resistance [
22]. In previous studies,
Enterococcus spp. is already the main cause of IE in the elderly [
23]
IE patients with advanced age that present are less likely to be operated than younger patients, despite the described lower mortality in patients treated with surgery [
11,
24,
25]. Although older adults have a high risk of in-hospital mortality after surgery, after discharge their mid-term outcomes are similar to the ones seen in younger populations [
11,
26]. Surgery is associated with lower incidence of adverse events irrespective of age, but it is usually underused in older patients [
12].
ESC guidelines recommend assessing comorbidities and operative risk to guide the decision in patients with surgical indication [
16]. Comorbidities were strongly connected with in-hospital mortality and with conservative management, in concordance with previous studies [
1,
5,
6,
27]. Surgical risk assessment might be done with non-specific risk scores like EuroSCORE II [
28] or Society of Thoracic Surgeons (STS) [
29], although specific scores for patients with IE have been developed, such as Prosthetic valve, Age ≥ 70, Large intra-cardiac destruction,
Staphylococcus spp, Urgent surgery, Sex [female], EuroSCORE (PALSUSE) [
14], RISK-E score [
30], AEPEI score [
31] or EndoSCORE [
32]. In any case, most of these scores do not consider important geriatric factors such as global comorbidity, frailty, malnutrition and functionality [
33].
The Elderly IE study [
6] described the functional impact of IE in older patients with IE, suggesting that the management of older patients with IE should include more than antibiotics and surgical decisions. The prevention and treatment of recurrent complications such as delirium, malnutrition, functional decline and drug adverse effects should also be considered. EI teams seem to improve early diagnosis and survival [
34]. Incorporating geriatricians and geriatric-expertise cardiologists in these teams could facilitate individualized management of octogenarians [
35].
In-hospital mortality in our octogenarian cohort was very high (36.5%). A recent analysis of the Swedish Registry of Infective Endocarditis (SRIE) found that patients ≥ 85 years had an in-hospital mortality of 23% [
12]. The high prevalence of comorbidity in our patients might have contributed to this difference. Moreover, in-hospital mortality rates above 20% have been reported in younger patients in previous studies, as in patients > 65 years [
3] or >75 years [
10]. In addition, recent data published in elderly patients with IE have shown even higher in-hospital mortality than that seen in our cohort [
23].
Our study has some limitations. Some recordings of clinical or diagnostic characteristics might be influenced by interobserver variability. In addition, although most GAMES centers have a Cardiac Surgery Department, this is not true for all hospitals. Finally, our data form did not include scales of geriatric syndromes such as frailty and malnutrition. However, our cohort of octogenarians is one of the largest reported and all centers used the same clearly established protocol.