FACTORS INFLUENCING OUTCOME IN PEDIATRIC INFECTIVE ENDOCARDITIS

Infective endocarditis (IE) is a rare entity in children and predominantly affects patients with congenital heart defects (CHD). Our study assesses the association between the time to diagnosis of IE and the time to surgery and outcome. The mean period from initial symptom presentation to diagnosis confirmation was 38 days (3-180 days). 12 patients (60%) had an early diagnosis (< 30 days), while in 8 cases (40%) the diagnosis was delayed (>30 days). 75% of the patients with a delayed diagnosis had received prior antibiotic therapy vs. 58% of the children with an early diagnosis. All 3 children with lethal outcomes had received prior antibiotic therapy, in two of the cases the diagnosis of IE was early, and in one case it was delayed. In conclusion, empiric antibiotic therapy is associated with a delayed diagnosis of infective endocarditis, and the longer the time to surgery the greater the risk for unfavorable patient outcome


Introduction
Infective endocarditis (IE) is a rare entity in children -frequency 0.43-0.69cases in 100 000 patient-years [1,2,3] and predominantly affects patients with congenital heart defects (CHD) -native or surgically corrected.Diagnostic delays have a negative effect on patient outcomes, whereas timely surgery is associated with a better prognosis.

Aim
Our study aims to form a sample of patients aged < 18 years with CHD who were hospitalized and treated with IE for 14 years, and to assess how time to diagnosis and time to surgery influence their outcome.

Materials and methods
The study is retrospective and includes the data of 20 patients aged < 18 years with underlying CHD and IE treated at the National Heart Hospital -Sofia Pediatric Cardiology Department, 01.01.2009-31.12.2022.IE was diagnosed based on the modified Duke criteria [4,5,6]: major (isolation of typical microorganisms (Staphylococcus aureus, Enterococcus, Streptococcus viridans) from the blood culture and the visualization of valvular vegetations on the echocardiography) and minor (predisposing conditions, fever, and embolic phenomena -glomerulonephritis, Osler nodes, Roth spots, positive rheumatoid factor).The diagnosis of IE requires two major and one major and two additional criteria.We followed the patients for up to 14 years (9 months-14 years, mean 6.3).

Results
The demographic characteristics of the patients are presented in Table 1.
The clinical signs of IE are diverse (Table 2).The leading symptom is malaise in 19 patients (95%); the second most common symptom was fever in 18 children (90%).In the remaining two cases (10%), the course of the disease was subacute, without a significant rise in body temperature.Eight patients (40%) had heart failure.Five children (25%) had gastrointes- tinal symptoms, the same number of cases had neurological symptoms, four children (20%) had respiratory failure, and two patients had embolic phenomena.All patients had increased inflammatory markers -in eight children, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and white blood cells (WBC) were elevated; in four cases -ESR and CRP, in 2 children -CRP and WBC, in 2 -only CRP, in 2 -only ESR, in one -only WBC, and in one -ESR and WBC.
In 7 patients with CHD and IE (32%) the heart defect was not corrected (Table 3).
Endocarditis developed after surgery in the remaining 13 patients (72%) with IE and underlying CHD.Their characteristics are presented in Table 4.
The mean period from initial symptom presentation to diagnosis confirmation was 38 days (3-180 days).Three children (15%) had hospital-acquired, and the remaining 17 (85%) had community-acquired IE.The time to diagnosis was ≤ 30 days from symptom onset in 12 patients (60%), while in 8 cases (40%) the diagnosis was delayed (> 30 days).6 out of 8 patients (75%) with delayed diagnosis had received prior antibiotic therapy and 7 out of 12 children (58%) with early diagnosis had preceding antibacterial treatment.The blood cultures were positive in all cases with an early diagnosis whereas 20% of the patients with a late diagnosis had negative blood cultures.Out of the 4 patients without bacterial growth from blood cultures, 3 (75%) had received prior antibiotic treatment.
All patients were treated conservatively with antibiotics according to the guidelines [4].In 12 children (5 with unoperated CHD and 7 after surgery), operative treatment was undertaken.The indications for surgery were uncontrollable heart failure, uncontrollable infection, and high embolic risk.The mean period between the diagnosis of IE and surgery was 33 days (4-90 days).Valve replacement was needed in 7 out of 12 operated children (58%) -AVR in three cases, MVR in one case, and biological pulmonary valve in three cases (valved conduit in two children).
IE relapse was reported in one case.Lethal outcome within 3 months of IE diagnosis was documented in three children, corresponding to a mortality rate of 15%.All children with lethal outcomes had received prior antibiotic therapy, in two of the cases the diagnosis of IE was early, and in one case it was delayed (> 30 days from symptom onset).In two children with lethal outcomes, the causative microorganism was Staphylococcus aureus (methicillin-resistant in one case), and the infection was community-acquired, while in one case IE was caused by a hospital-acquired Klebsiella pneumoniae strain.Both children with S. aureus IE died before surgery was undertaken, while the patient with Klebsiella pneumoniae IE underwent surgery 30 days after IE diagnosis, and died in the early postoperative period.It is noteworthy however that the child had a severe aortic valve stenosis which may have contributed to the unfavourable outcome.

Discussion
Our data confirm the low frequency of IE in children -< 2 cases/year in our clinic.Gupta et al. report an incidence of IE of 0.43/100 000 children [2], Mahony et al.
-0.84/100 000 patients < 18 years [3].The mean age of the patients with IE in our cohort is 9.12 years, with a lower incidence of IE in children < 1 year (5%) and higher in older patients -25% among the age group 1-6 years and 35% among the patients > 7 years.The results of Gupta et al. are similar -in their study, 56.2% of the patients with IE were aged ≥ 11 years, while 15.4% were aged ≤ 1 year.In our patient cohort, IE was more frequent among males compared to females -60 vs. 40%, respectively.Luca et al. from Romania also report a higher frequency of IE by males -33 cases vs. 21 cases by females [7].In a study by Carceller et al. from Canada, 54% of the children with IE were males [8].
Our results confirm the critical role of an underlying heart disease in developing IE.This corresponds to the literature data reporting that the vast majority (50-70%) of the children with IE have CHD, which is the major predisposing factor [2,3].In the study of Mahony et  The risk for IE in children with CHD is 75 times higher than the general population risk [9,10].The probability of IE depends greatly on the CHD type, as the malformations associated with higher risk are cyanotic CHD, left-sided heart defects, and atrioventricular septal defect (AVSD) [11,12,13].Our data confirm these observations, as 35% of the children with CHD and IE have left-sided obstructive lesions, 25% -AVSD, and 15% -Fallot tetralogy.Regarding the localization of the vegetations, our results show that the left-sided heart structures are most often involved.These data are similar to the results of other authors (Luca et al., Ahmadi et al. [15].Our patients' most often reported symptoms are malaise, fever, and cardiovascular manifestations.The results from the meta-analysis of Day et al. are similar -all children had cardiovascular symptoms, 83% had fever, and 61% had malaise.It is worth mentioning that the diagnosis IE is often delayed, with a mean period between first symptom presentation and diagnosis confirmation of 43 days.The patients with an early diagnosis are those with acute symptoms who were timely referred to a pediatric cardiologist.The diagnosis is frequently delayed in cases with subacute or non-specific symptoms, and empiric antibiotic courses obscure symptoms. Regarding the etiology of IE, our data are similar to those of other authors.In 5 cases (26%), the blood cultures remained negative.This is the so-called culture-negative endocarditis (CNE), which is most often due to preceding antibiotic treatment or infections with difficult-to-culture microorganisms, whose incidence varies from 8 to 36% [16,17,18].For example, one of our patients with a negative blood culture had positive serology for Chlamydia, and three children had received antibiotic therapy before the blood culture was taken.
The leading bacterial pathogens in our cohort were the so-called "typical" microorganisms -most often Gram-positive bacteria.Four of the cases with IE were caused by Streptococci, and three were caused by Staphylococci.These data are similar to those reported by other authors [19].
Most patients with IE require surgery -all children with uncorrected CHD and almost two-thirds with operated CHD.In the study of Mahony et al., surgical treatment was necessary in 51% of the patients.
Hospital mortality from IE in our patient cohort is 15%.Carceller

Conclusion
Infective endocarditis is a disease associated with significant morbidity and mortality, which affects most often patients with an underlying heart disease (operated or unoperated) and whose diagnosis is based on positive blood cultures and the visualization of vegetations on the echocardiography.It is crucial to exclude IE in cases of persistent fever in a child with a known heart disease.The prerequisites for a favorable patient outcome are the timely diagnosis, the early initiation of effective antibiotic therapy, and, when necessary -surgical treatment.

Table 2 .Clinical symptoms in patients with IE
al., almost 60% of the children with IE had underlying heart disease, and in the study of Gupta et al. -53.5%.

Table 3 . Characteristics of the patients with IE and unoperated CHD
AVR -aortic valve replacement, CA -community-acquired, HA -hospital-acquired, MRSA -methicillin-resistant S. aureus, N/A -not applicable, VSD -ventricular septal defect