Article Text
Abstract
Objective To describe the characteristics of infective endocarditis (IE) in octogenarians and assess their prognosis.
Methods Patients with definite IE hospitalised at a referral centre between July 2008 and July 2013 were prospectively included. A total of 454 patients were divided into three groups: 230 patients under 65 years old, 173 patients between 65 and 80 years old, and 51 patients over 80 years old. The main end point was 1-year mortality.
Results One-year mortality was higher in the ≥80 years old group (37.3%) than in the <65 years old group (13%; p<0.001) and the 65–80 years old group (19.7%; p=0.009). Enterococci and Streptococcus gallolyticus were the more frequent micro-organisms. Embolism under antibiotic therapy (n=11 (21.6%), p=0.03) and renal failure (n=23 (51%), p=0.004) were more frequent in the ≥80 years old group. Among the ≥80 years old group, 38 patients had theoretical indication for surgery. Mortality was low (6.3%) in the 16 operated patients, but very high (72.7%) in the 22 patients not operated. Even if octogenarians were less often operated, their survival after surgery was excellent like younger patients (93.7%, 89.9% and 90.4%, respectively), whereas the absence of surgery was associated with very poor prognosis.
Conclusions IE in octogenarians is a different disease, with Enterococci as the most frequent micro-organisms and with higher mortality than younger patients. ESC recommendations for surgery are less implemented than in younger patients, yielding dramatic mortality in patients not operated despite a theoretical indication for surgery, while operated patients have an excellent prognosis. These results suggest that surgery is underused in octogenarians.
- octogenarians
- infective endocarditis
- elderly
- surgery
- guidelines
Statistics from Altmetric.com
Introduction
Infective endocarditis (IE) is an uncommon infection with an incidence of between 30 and 90 per million inhabitants per year.1 2 Despite improvements in diagnostic and therapeutic strategies, both inhospital2–5 and 1-year mortality remain unacceptably high.1 6
IE in the elderly represents an increasing proportion of IE patients2 7 8 and has been associated with less severe clinical symptoms and delayed diagnosis,5 8–10 but more complications and higher mortality.5 9 11 12
However, some recent studies gave contradictory results with no difference in clinical presentation and outcome.7 13 In addition, most studies focused on patients over 65 years old,5 7 9 10 14 whereas few works have studied IE among patients over 80 years old.8
Finally, the implementation of the guidelines in this specific population and the results of surgical therapy are not known.
The aims of our study were (1) to assess the prognosis of IE in octogenarians, as compared with younger patients; (2) to determine the clinical, echocardiographic, biological and microbiological profiles and the outcome of patients above 80 years old, in comparison with younger subjects; and (3) to evaluate the degree of implementation of the European Society of Cardiology (ESC) guidelines15 in these patients, as well as the influence of surgery on mortality.
Patients and methods
All patients presenting with IE and hospitalised in the Cardiology Department of La Timone Hospital, Marseille, France, between July 2008 and July 2013 were included in this prospective cohort study. We included all cases of definite IE according to the modified Duke criteria,16 on native or prosthetic valve, except patients under 18 years old and those who were pregnant. The patients were divided into three groups according to their age: under 65 years (<65 years group), between 65 and 80 (65–80 years group) and above 80 (≥80 years group).
Data registration
For each patient, several clinical, biological, microbiological and echocardiographic data were prospectively gathered. Medical treatment and surgical indications were based on the ESC guidelines,15 and then adapted to each patient case after discussion with the endocarditis team using a multidisciplinary approach. A full-body CT scan was most often done to each patient on admission. Time to surgery is specified as well the implementation or not of the ESC guidelines.15 Inhospital mortality and 1-year mortality were defined as all-cause mortality.
Data were collected during regular visits in the department by the cardiologist and the infectious diseases specialist at 1, 3, 6 and 12 months after the end of the treatment. Whenever a patient’s data were missing, we gathered them by contacting the patient’s doctor, or, if unavailable, the patient himself or his/her family. When none of these options enabled gathering of missing data, the municipal registers were checked to verify whether the patient’s death had not been declared. These data were collected and gathered during the months of June and July 2014.
Statistical analysis
Continuous variables were described by their mean and SD and the categorical variables by their size and percentages.
The comparison of characteristics among the three groups was done using the non-parametric Kruskal-Wallis test for the continuous variables and the χ2 test or the exact Fisher’s test for the categorical variables, depending on the application conditions. When necessary we also compared the ≥80 years group vs the others (<65 years and 65–80 years groups) using Mann-Whitney U test and χ2 test or exact Fisher’s test for the categorical variables, depending on the application conditions.
The length of medical monitoring was calculated as the difference between the last check, or the date of death, and the date of initiation of the antibiotic treatment. The survival analysis was carried out up to 1 year with the Kaplan-Meier method; patients still alive after 1 year were considered censored. Tests were realised in bilateral situation. Results with a p value ≤0.05 were considered statistically significant. The statistical analysis was made with the R software (V.3.1.0).16
Results
Patient characteristics
Between July 2008 and July 2013, 454 patients were included in this study. Of these patients, 73.3% were male; 230 patients were in in the <65 years group, 173 in the 65–80 years group and 51 patients in the ≥80 years group. Patients above 80 years old ranged from 80 to 88.8 years. No patient older than 80 was lost to follow-up.
The epidemiological features, comorbidities, portal of entry, type of infection, clinical characteristics, and biological, microbiological and echocardiographic data are summarised in table 1.
Octogenarians suffering from IE had more frequent digestive or urinary portal of entry. Conversely, the frequency of dental and skin origin tended to decline with older age. There was no statistically significant difference between the age groups in terms of time until symptom onset, fever, heart murmur, heart failure or haemodynamic failure. Nevertheless, octogenarians presented with milder initial symptoms: extracardiac IE signs (Osler’s nodes, Roth’s spots and Janeway lesions) and embolic complications were less frequent with increasing age at the time of diagnosis. Total emboli were half as frequent (29.4% in the ≥80 years group compared with 50.9% in the <65 years group and 44.5% in the 65–80 years group; p=0.02), whether they were silent (p=0.06) or symptomatic (p=0.03), whereas the rate of major cerebral embolic events was not different. No mycotic aneurysm was encountered in the ≥80 years group. Conversely, spondylitis occurred in 10.1% of cases (data not shown). It was much more frequent among very old patients (23.5% in the ≥80 years group, 5.2% in the <65 years group and 12.6% in the 65–80 years; p<0.001). Enterococcus was identified in 50% of the cases (data not shown).
The most frequent isolated pathogens in the ≥80 years group were Enterococci (21.6% vs 10% in the <65 years group and 15.6% in the 65–80 years group; p=0.05) and Streptococcus gallolyticus (17.7% to 11.3% and 16.2%; p=NS). Conversely, the percentage of methicillin-susceptible Staphylococcus aureus tended to be less frequent with age (20.4% to 11.8% and 12.7%; p=0.07).
Mitral localisation was predominant in the older group (43.1% to 27.8% in the <65 years group and 38.7% in the 65–80 years group; p=0.02). No difference was observed between groups concerning echocardiographic data, except severe valvular regurgitation, which was more frequent in young subjects.
Prognosis
Inhospital mortality rates were 10.4% (<65 years group), 13.3% (65–80 years group) and 15.7%, (≥80 years group) (p=0.48). One-year overall mortality was, however, considerably higher in the octogenarian group (37.3%) than in the >65 years group (13%; p<0.001) and 65–80 years group (19.7%; p=0.009) (figure 1).
The causes of death are summarised in table 2. Deaths due to heart failure were predominant in the older group, both concerning inhospital mortality (50% vs 16.7% in the <65 years group and 21% in the 65–80 years group) and 1-year mortality (42.1% vs 16.7% and 14.7%, respectively). Conversely, deaths due to sepsis were less frequent in elderly patients (inhospital death: 12.5% vs 37.5% in the <65 years group and 30.4% in the 65–80 years group; and 1-year mortality: 15.8% vs 33.3% and 15.8%, respectively).
In the octogenarian group, among the 11 additional deaths occurring after hospitalisation, 4 were caused by heart failure, and all of them had theoretical indication for surgery and were not operated on (data not shown).
Although elderly patients had less embolic complications on admission, their evolution was more frequently complicated by embolic events under treatment (21.6% in the ≥80 years group vs 9.2% in the <65 years group and 15% in the 65–80 years group; p=0.03). Acute renal failure was also much more frequent among this population (51% vs 27.1% in the <65 years group and 32.4% in the 65–80 years group; p=0.004).
Surgical therapy and implementation of guidelines
Figures 2 and 3 show the relationship between the degree of implementation of the ESC guidelines and mortality in < 80-year-old and ≥80-year-old patients. IE, respectively. When applying the ESC guidelines, surgery was theoretically indicated in 79.1%, 78% and 74.5%, but was ultimately performed in 82.4%, 69.6% and 42.1% in the <65 years, 65–80 years and ≥80 years groups, respectively. When octogenarians were not operated, it was mostly for medical reasons: high comorbidities, high operative risk or multidisciplinary decision (86.4% in the ≥80 years group vs 50% in the <65 years group and 75.6% in the 65–80 years group; p<0.001)
Table 3 represents the relationship between use of surgery and mortality in the 3 age categories of patients. One-year mortality was globally higher in octogenarian group than in younger patients, regardless of surgical indication. Moreover, when a patient above 80 had a surgical indication, the patient’s 1-year death probability was 10 times lower when surgery was performed (6.3%) than when it was not (72.7%). Thus, when operated, octogenarians had mortality similar to the other groups (10.1% in the <65 years group, 9.6% in the 65–80 years group and 6.3% in the ≥80 years group). Conversely, octogenarian patients who were not operated presented a twice probability of death than younger patients (72.7% were in the ≥80 years group vs 36.4% in the <65 years group).
One-year survival according to the Charlson Comorbidity Index or EuroSCORE II is represented in online supplementary table 1. No significant difference was observed in terms of comorbidities between octogenarian patients with theoretical indication for surgery who had been denied for surgery and those operated on (see online supplementary table 2).
Supplementary Material
Discussion
To the best of our knowledge, this study represents the largest cohort of IE ever studied among subjects older than 80. It appears that octogenarian patients had milder symptoms, fewer septic embolisms and more spondylitis than younger patients. Moreover, they presented considerably more complications as well as more severe prognosis than younger patients. Non-implementation of ESC guidelines was more frequent in younger patients. Finally, surgical therapy was underused in octogenarians, but when performed is associated with excellent outcome.
Patient characteristics
In accordance with the published data, the clinical context varied with age: contrary to younger patients who frequently presented predisposing factors (congenital heart disease or intravenous drug abuse), the oldest patients do not have a specific profile. Indeed, cardiovascular and general comorbidities increased with ageing, like in the general population.11 13 14
The IE clinical presentation among octogenarians was not significantly different in our study as compared with the younger patients, although some published studies suggested that IE was more insidious and benign in older subjects with less haemodynamic failure8 and less extracardiac signs.12
The prevalence of spondylitis associated with IE had rarely been reported. It occurred in 10.1% of cases in our cohort, when it usually varied from 3.7% to 15%.17–20 Although former studies did not find a link between age and spondylitis associated with IE, a strong association between those two pathologies had been highlighted in our cohort. Indeed 23.5% of our patients above 80 have presented a spondylitis vs 5.2% of those under 65. Enterococcus was identified in half of the cases. This micro-organism was not usually frequently observed in isolated spondylitis, but it had been described as playing a major role when the latter was associated with IE.17 19 20 The older age of our patients compared with previous series accounts for those results. Furthermore, IE and spondylitis prevalence both increased with ageing and IE due to Enterococcus being higher in octogenarians. Finally, a full-body CT scan is frequently performed in our centre and this may also explain our results. We suggest that the risk for spondylitis needs to be carefully considered in very old patients, especially those presenting an Enterococcus infection.
The higher frequency of digestive and urinary portal of entry in old patients as opposed to the dental and skin entry for younger patients had considerably altered the infectious microbiological profile in octogenarians. As a result, a greater proportion of Enterococcus and S gallolyticus and less Staphylococcus and S viridans were observed. These results were consistent with previous publications.8 11 12 14 Likewise, Olmos et al 21 recently described a greater proportion of IE due to Enterococcus among diabetic patients, regardless of their age. Knowing the fact that in our cohort more than a quarter of the octogenarian patients suffered from diabetes, this could account for the greater proportion of Enterococcus.
Prognosis
Inhospital mortality rates were not significantly different among the age groups, although 1-year mortality rate was significantly higher in octogenarians. The first explanation may be the lower number of inhospital events (8 deaths) as compared with 1-year mortality (19 deaths), making statistical significance more difficult to reach. Another explanation could be the progressive left ventricular deterioration caused by valvular damage and not adequately treated by surgery. Of note, among the 11 additional deaths occurring after hospitalisation, 4 were caused by heart failure, and all of them had theoretical indication for surgery and were not operated on.
Overall mortality in our study was lower than in previous studies1 5 7 (table 4). As previously reported,6 this may be related to our multidisciplinary approach, a strict implementation of guidelines and a high rate of surgery.
López-Wolf et al 8 recently reported a slightly higher mortality rate (20% vs 15.7%) than in our series, but 1-year mortality was not reported in this series.
In our study, 1-year mortality rate among octogenarians was considerably higher than among the younger patients, contrary to inhospital mortality rate. This finding of a very high late mortality rate, mostly caused by heart failure, may be related to the low operative rate in our population. Patients above 80 were frailer, present with more comorbidities and are facing higher risks of complications.
Surgical therapy and implementation of ESC guidelines
The very old patients had considerably more complications and poorer outcomes than others, and this may be related to the lower rate of surgical management in these patients, although theoretical indication rate was the same, confirming the poor implementation of ESC guidelines in old patients.5 7 14
Our data confirm that surgery is of major importance in IE.22 23 Operated patients showed an excellent 1-year survival rate: more than 90%, even in the oldest patients (93.6%). Other smaller series13 recently drew similar conclusions.
More than age, frailty—however hard to quantify—seems of utmost importance. Numerous scores have emerged over the past years to evaluate this geriatric concept.24 25 These scores most often include neurocognitive, thymic and nutritive evaluations, dependence on daily life, social background, and comorbidities. Some of the studies have recently proven their utility for cardiovascular diseases management26 and prognosis evaluation before cardiac surgery.27–29 We recommend a more global patient evaluation and a greater cooperation with geriatricians to improve IE management in the elderly population.
Limitations
Our study presents some limitations.
First of all, it is based on a monocentric cohort performed in the cardiological department of a referral care centre for cardiac surgery and IE, with possible selection bias.
The low mortality rate in our centre, already described in previous publications,6 make difficult the extrapolation of the results.
Finally, our study also suffers from being a non-randomised study, with associated limitations and selection bias affecting comparisons between outcomes. Among patients in whom surgery is indicated, baseline characteristics, comorbidities and operative risks may be significantly different between patients treated medically and those who underwent surgery. These differences in treatment groups make it difficult to assess the impact of surgery on outcomes. Thus, additional studies are needed to confirm the benefits of surgery in elderly patients with high operative risk.
Conclusion
IE in octogenarians is a different disease, with more spondylitis, less embolism and Enterococci as the most frequently identified micro-organisms. Implementation of ESC guidelines is poor and surgery is underused, although operated octogenarian patients present with an excellent outcome.
Age alone should probably not be regarded as a contraindication to surgery. A better implementation of ESC guidelines, a more frequent use of surgery and a greater cooperation between cardiologists and geriatricians would help in improving the management of octogenarian patients with IE.
Key messages
What is already known on this subject?
Infective endocarditis (IE) in the elderly represents an increasing proportion of IE patients and has been associated with less severe clinical symptoms and delayed diagnosis, but more complications and higher mortality.
Most published studies focused on patients over 65 years old, whereas few works have studied IE among patients over 80 years old.
IE is becoming more and more frequent in very old patients (>80 years old).
What might this study add?
It represents the largest series ever published on the topic and give new information on IE in this specific population, particularly about its prognosis and management.
In addition, this study will give new insights concerning the degree of implementation of the ESC guidelines in this particular group of patients, as well as the influence of surgery on mortality.
How might this impact on clinical practice?
Our study will deeply impact the management of IE in octogenarians. A better implementation of ESC guidelines, a more frequent use of surgery and a greater cooperation between cardiologists and geriatricians will help improve the management of octogenarian patients with IE, the incidence of which is likely to increase considerably in the coming years.
References
Footnotes
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.