Eyes Wide Shut: A Cohort Study Questioning the Role of Fundoscopy in Infective Endocarditis Diagnosis

Abstract In this retrospective/prospective study, we assessed the role of fundoscopy in 711 episodes with suspected infective endocarditis (IE); 238 (33%) had IE. Ocular embolic events (retinal emboli or chorioretinitis/endophthalmitis) and Roth spots were found in 37 (5%) and 34 (5%) episodes, respectively, but had no impact on IE diagnosis.

Prompt identification of infective endocarditis (IE) and its associated complications is the key to improving outcomes.Fundoscopy is considered part of the initial workup of patients with suspected IE [1] since it can detect embolic events and Roth spots (2% of patients with definite IE) [2], which are 2 of the minor Duke criteria (vascular and immunological phenomena) to improve the diagnostic process [3].
While fundoscopy is currently considered a standard component of the workup for patients [1], there are limited data on fundoscopy findings [2,[4][5][6], and its impact on diagnosis has not been previously assessed.Therefore, our objective was to delineate the frequency of fundoscopy findings among patients with suspected IE and evaluate their effect on diagnosis.

METHODS
This single-center study was conducted at the Lausanne University Hospital, Switzerland, from January 2014 to June 2023 (2014-2017, retrospective cohort; 2018 onward, prospective cohort).The Canton of Vaud Ethics Committee approved the study.
All adult patients (aged ≥18 years) with suspected IE (patients who had blood cultures drawn and an echocardiography performed specifically for the research of IE), fundoscopy performed and written consent (prospective cohort), or absence of refusal to use their data (retrospective cohort) were included.At our institution, fundoscopy by an ophthalmologist is part of the standard workup in patients with suspected IE.Episodes were classified as IE by the Endocarditis Team.
Demographic, clinical, microbiological, follow-up, and management data were collected.Using the 2023 Duke-International Society of Cardiovascular Infectious Diseases clinical criteria [3], we established 2 classifications based on clinical criteria: without and with the inclusion of fundoscopy findings.Using these 2 assessments, we calculated the reclassification rate (from rejected to possible and from possible to definite IE) for all patients with suspected IE and within the IE subgroups.
SPSS version 26.0 (SPSS, Chicago, IL) software was used for data analysis.Categorical variables were analyzed using the χ 2 or Fisher exact test, and continuous variables were analyzed using the Mann-Whitney U test.All statistical tests were 2-tailed, and P < .05 was considered statistically significant.
Ocular embolic events (OEEs) were observed in 37 (5%) episodes, with 24 having retinal embolic events and 13 having chorioretinitis/endophthalmitis.Of these, 25 (11%) were among IE episodes and 12 (5%) without IE (Table 1).In 23 (64%) episodes, the Duke vascular phenomena criterion had been met before fundoscopy.Roth spots were found in 34 (5%) episodes, with 22 (9%) in IE episodes and 12 (5%) without IE.Among them, the Duke immunological phenomena criterion had been fulfilled in 1 (3%) episode prior to the fundoscopy.Fundoscopy results (OEEs and Roth spots) led to the reclassification of 11 (2%) episodes from rejected to possible IE and 2 (0.3%) from possible to definite IE.However, all of these reclassified cases did not have IE based on the Endocarditis Team's evaluation.

DISCUSSION
In our study, fundoscopy was performed in 41% of patients with suspected IE and revealed OEEs (either retinal emboli or chorioretinitis/endophthalmitis) and Roth spots in a small proportion of patients (in 5% of episodes each).Previous studies on IE patients reported similar rates of OEEs (8%-19%) [4,5] but lower rates of Roth spots (2%) [2].
To the best of our knowledge, our study is the most comprehensive effort to explore the role of fundoscopy in diagnosing IE.However, the presence of OEEs and Roth spots had no impact on IE diagnosis, as the 13 patients reclassified according to Duke clinical criteria based on the fundoscopy findings did not have IE.Therefore, universal fundoscopy for diagnostic purposes in patients with suspected IE seems unnecessary [1].The lack of impact on diagnosis can be attributed, in part, to the nonspecific nature of Roth spots.Although commonly linked with subacute IE, they can be found in a diverse array of noninfectious diseases, such as collagen vascular disease, leukemia, hypertensive or diabetic retinopathy, preeclampsia, and anoxia, underscoring their limited diagnostic specificity [7].
Among patients with suspected IE, OEEs were linked to fungemia, reinforcing the recommendation for universal fundoscopy screening in candidemia cases [8].As expected, OEEs were more common among IE patients.In a large cohort of bloodstream infections and OEEs, IE was the second most common focus of infection (14%), following skin and soft tissue infections (17%) [9].In the present study among S. aureus bacteremia episodes, OEEs were found in 7%, consistent with previous findings (9%) [6].As expected, OEEs were associated with embolic events outside the eye and factors generally linked to systemic embolization, such as large vegetations (≥10 mm), S. aureus, persistent bacteremia/candidemia, and intracardiac abscess [10,11].
This study has several limitations.First, it was a single-center study, and some patients were retrospectively included.Nevertheless, it is the most extensive study to date on fundoscopy's impact on patients with suspected IE.Second, despite being part of the standard diagnostic workup, fundoscopy was performed in fewer than half of the patients with suspected IE.This can be explained by the fact that the decision to perform fundoscopy remained at the discretion of the physician.
In conclusion, this study has confirmed that systematic fundoscopy can detect ocular lesions in a nonnegligible proportion of patients with suspected IE.However, these fundoscopy findings did not influence the diagnosis of IE, suggesting that universal fundoscopy for diagnostic purposes is not justified.Nevertheless, in cases where patients exhibit ocular symptoms or have fungemia, fundoscopy is warranted.