Infective Endocarditis with Antineutrophil Cytoplasmic Antibody: Report of 13 Cases and Literature Review

Objective Chronic infections tend to induce the production of antineutrophil cytoplasmic antibody (ANCA). Infective endocarditis (IE) has been reported to exhibit positive ANCA tests and to mimic ANCA-associated vasculitis, which may lead to a misdiagnosis and inappropriate treatment. The aim of this study was to clarify whether there is any difference in the clinical features between ANCA-positive IE and ANCA-negative IE. Methods A retrospective study was carried out on 39 IE patients whose proteinase 3 (PR3)-ANCA and myeloperoxidase (MPO)-ANCA levels were measured. After dividing the patients into ANCA-positive and ANCA-negative IE, we compared their clinical features. Results we compared 13 ANCA-positive IE patients with 26 ANCA-negative IE patients. All 13 ANCA-positive IE patients were proteinase-3-ANCA positive. Compared with the ANCA-negative IE group, the prevalence of edema of the lower extremities, the serum lactate dehydrogenase (LDH) level and positive blood cultures rate were higher in ANCA-positive IE group, but there was no significant difference in other clinical features. Conclusion Therefore, if a patient presents with fever, arthralgia, skin rash and is ANCA-positive, appropriate steps should be taken to exclude infection (especially IE) before confirming the diagnosis of ANCA-associated vasculitis and embarking on long-term immunosuppressive therapy.

In this report, we describe 13 patients with IE who had positive findings upon testing for ANCA by an antigen-specific enzymelinked immunosorbent assay (ELISA). We then compared those findings with 26 ANCA-negative patients as well as cases reported in the literature.

Methods
This study was approved by the Ethics Review Board of Shanghai Jiaotong University (Shanghai, China). All patients including the guardians on the behalf of the minors participants provided written informed consent to be included in the study.

Patients
A total of 161 patients being treated at Shanghai Jiaotong University were diagnosed as having IE according to the modified Duke criteria [21] between January 2003 and June 2012. We have ruled out the patients with primary ANCA-associated-disease who occur super-infection or IE. Of the 161 patients with IE, only 39 individuals (21 males; age, 46.7613.5 (range, 17-75) years) had been tested for ANCA; the decision to measure ANCA had been made by the referring physicians. The remaining 122 patients who did not undergo ANCA measurement were excluded from this study. We classified the 39 patients as ''ANCA-positive IE'' or ''ANCA-negative IE'' and compared their clinical features. Investigations to exclude the possibility of drugs inducing ANCA were carried out for all patients.

Statistical analyses
Statistical analyses were carried out using SPSS software (SPSS, Chicago, IL, USA). Descriptive statistics are represented as the mean 6 standard deviation. Chi-square or Fisher's exact test were adopted to analyze all categorical variables. The Student's t-test was used to analyze continuous variables with a normal distribution. Survival curves were obtained using Kaplan-Meier's method. The significance was obtained by a log-rank test. Statistical significance was established at a two-tailed level of , 5%.

Prevalence of ANCA in IE patients with ANCA data
Of the 39 IE patients with ANCA data, 13 (13/39, 33.3%) patients (7 males) were ANCA-positive, the remaining 26 (26/39, 66.7%, 14 males) were negative for ANCA. The median age was 51.1612.4 years among ANCA-positive IE patients, and 44.5613.7 years among ANCA-negative IE patients. There was no significant difference in age or sex between the two groups. The baseline characteristics and predisposing conditions are shown in Table 1. Table 2 shows the clinical characteristics of ANCA-positive IE and ANCA-negative IE subjects. Fever was the most common symptom: it was observed in all ANCA-positive IE and 22 (84.6%) of ANCA-negative IE cases. Anemia was observed in 9 (69.2%) ANCA-positive IE and 17 (65.4%) ANCA-negative IE subjects. Splenomegaly was noted in 5 (38.5%) ANCA-positive IE and 10 (38.5%) ANCA-negative IE patients. Nephropathy was noted in 4 (30.8%) ANCA-positive IE and 7 (26.9%) in ANCA-negative IE cases, whereas 5 (38.5%) ANCA-positive IE and 4 (15.4%) ANCAnegative IE individuals complained of arthralgia. Edema of the lower extremities occurred in 5 (38.4%) ANCA-positive IE and 2 (7.7%) ANCA-negative IE cases. Rash was found in 2 (15.4%) patients and 1 (7.7%) patient had cerebral infarction among ANCA-positive IE subjects, whereas 2 (7.7%) patients had a rash and 2 (7.7%) had cerebral infarction in the ANCA-negative IE group. Two (15.4%) ANCA-positive IE patients showed finger clubbing but Janeway lesions and Osler nodes were uncommon. ANCA-positive IE cases had a higher prevalence of edema of the lower extremities. However, there was no significant difference in other symptoms between the two groups. Blood cultures were conducted in 37 patients of 39 cases and a causative microorganism was identified in 9 ANCA-positive IE patients (69.2%) and 6 ANCA-negative IE patients (25%). The prevalence of positive blood cultures was significantly higher in ANCA-positive IE subjects. Streptococcus was the leading microorganism in both groups. There were 10 microorganisms isolated from ANCA-positive patients: 9 (90%) Streptococcus spp. and 1 (10%) Enterococcus spp. Seven microorganisms were isolated from ANCA-negative IE subjects: 5 Streptococcus spp. (71.4%), 1 Enterococcus spp. (14.3%) and 1 Staphylococcus spp. (14.3%) ( Table 3).

Outcomes
Overall, 3 patients in the ANCA-positive IE group died in hospital, and 1 patient was lost to follow-up. One patient died of renal failure and 2 patients died of acute heart failure. No patient died during the observation period in the ANCA-negative IE group and 2 patients were lost to follow up. The survival rate was significantly lower in ANCA-positive IE (P = 0.012, Fig. 1). All remaining patients survived and recovered after therapy in which 6 ANCA-positive IE patients and 17 ANCA-negative IE patients underwent heart operation, and there was no significant difference between the two groups (P = 0.566).
A Medline search of English-language articles describing IE patients with ANCA positive provided 31 reports of such cases. The characteristics of the 31 cases from the literature review are described in Table 4, along with the characteristics of the 13 patients from the current study.

Discussion
Detection of ANCA is highly specific for the diagnosis of AAV (e.g., anti-PR3 antibody for Granulomatosis with polyangiitis). However, various infections can result in a positive ANCA test (especially IE). Usually, IE associated with ANCA is rare, but 31 cases have been reported [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]: 16 for Streptococcus spp., 4 for Enterococcus spp, 4, for Bartonella spp., 2 for Staphylococcus spp., 1 for Neisseria spp., 1 for Propionibacterium spp., and 6 culture-negative species. IE with ANCA can mimic the manifestations of AAV such as skin purpura and glomerulonephritis. Eight cases of ANCAassociated IE misdiagnosed as AAV have been published, including our report [7,[10][11][12][13]15]. The differentiation between ANCA-associated IE and AAV may be difficult, but important differences in clinical presentation do exist between these entities. Chirinos et al. [12] described some pertinent differentiating features which may be more indicative of IE than AAV: organ involvement limited to skin and kidneys; positive blood cultures; abnormal levels of complement; immune deposits; and other autoantibodies (e.g., rheumatoid factor, antinuclear antibodies, cryoglobulins, anticardiolipin antibodies). AAV with valvular compromise involved almost exclusively the aortic valve and lead almost invariably to the need for valve replacement in the reported cases. Moreover, although discrete large vegetations seen on echocardiography suggest IE, small discrete vegetations have been rarely reported in AAV with valvula compromise. However, in their study age, sex, constitutional symptoms, ESRte and leukocytosis were not significantly different between the study groups. Bonaci-Nikolic et al. [15] compared 66 AAV patients with 17 PR3 and/or MPO-ANCA-positive patients with protracted infections, including 7 IE patients. They found that AAV patients more frequently had pulmonary and nervous system manifestations, while patients with infections more frequently had dual ANCA (high PR3, low MPO), aCL and anti-b2-GP I. Whereas there was no difference in frequency of lethality, renal failure and skin lesions in the two study groups. Haseyama, et al [4] suggested that in patients with high PR3 titers 50 U/ml in association with IE, it is initial to treated with immunosuppression. However, there have been 8 cases of IE with positive ANCA, including 2 cases in our report treated initially with immunosuppression or simultaneously with antibiotics because of delayed or equivocal diagnosis [7,10,11,13,15]. Four of these patients eventually died [10,15]. Chirinos et al. [12] reckoned that histological examination of compromised valves revealed myxomatoid or fibroid degeneration of the valve tissue rather than granulomatous inflammation in most cases, explaining the lack of response to immunosuppressive therapy. Damage occurring during the initial valvulitis could lead to further valve distortion despite remission of the valvulitis with  immunosuppressive therapy. Therefore, it is important that without convincing evidence of AAV and ruling out IE, there is no strong evidence to support immunosuppression. In this retrospective study, 39 patients underwent testing for ANCA by ELISA, among which 13 (33.3%) were ANCA-positive. The decision to measure ANCA had been made by referring doctors as ANCA had not been measured consecutively in our hospital during the period in question. Even if we consider the selection bias of our data that ANCA levels were measured in only 39 cases among 161 inpatients in our hospital, the prevalence of ANCA in IE may be considered to be high. However, prospective studies are needed to clarify the precise prevalence of ANCApositive IE. In addition, all 13 patients were PR3-ANCA-positive, a finding that is consistent with other reports. Of the 31 literature patients, 29 cases had PR3-ANCA, of which 6 cases had dual positive ANCA (PR3-ANCA and MPO-ANCA), whereas only 2 cases had MPO-ANCA. MPO-ANCA-positive IE, though rare, does exist.  All 13 PR3-ANCA associated patients had some manifestations mimicking AAV, such as fever, anemia, and splenomegaly. Streptococcal species were the most common pathogen, and cardiac-valve abnormalities were demonstrated in 10 patients, and all patients except 3 recovered with antibiotic therapy. We compared the clinical features of ANCA-positive IE patients and ANCA-negative IE patients: there was little difference between the two groups. These findings suggest that ANCA reflects the systemic inflammation rather than the activity of IE. As for laboratory findings, the white blood cell levels tended to be higher in ANCA-positive IE (P = 0.067), and LDH levels were higher in ANCA-positive IE. As LDH is a useful marker in activity of myocardial infarction, it is suggested that ANCA-positive IE was more active in cardiac lesions than ANCA-negative IE on admission. On the other hand, the rate of lower extremity edema was higher in ANCA-positive IE. We speculate that the edema was related to functional valve insufficiency. As this clinical feature is non-specific and the number of the patient was limited, their correlation needs more research. Interestingly, the prevalence of positive blood cultures was significantly higher in ANCA-positive patients than in ANCA-negative patients. This finding may suggest a link between infection and the development of ANCA. Negative blood cultures alone do not rule out the possibility of IE. Sugiyama et al. [14] reported a blood culture-negative IE patient who exhibited a positive ANCA test and simulated ANCA-associated vasculitis. Thus, when dealing with patients with negative blood cultures but positive ANCA tests, blood culture tests should be repeated. Three ANCA-positive IE cases died but all ANCA-negative patients recovered, and the survival rate was lower in ANCApositive IE in statistically. However, 2 of the 3 cases (50-year-old and 42-year-old women) were misdiagnosed as AAV and 1 (37year-old man) exhibited serious congenital heart disease. Hence, the relationship between ANCA-positive IE status and survival needs further investigation.
To date, there was no any report descried the pathogenicity of ANCA in IE. We showed that there were few differences in clinical features between ANCA-positive and ANCA-negative IE patients. Further studies are necessary to clarify the role of ANCA in patients with IE. However, ANCA-positive IE can mimic AAV. If a patient has been found to be ANCA-positive and systemic vasculitis is suspected, physicians should take appropriate steps to exclude infection (especially IE) before planning and implementing long-term immunosuppressive therapy.

Author Contributions
Conceived and designed the experiments: C. Yang. Performed the experiments: C. Ying DY. Analyzed the data: C. Ying DY. Contributed reagents/materials/analysis tools: HD. Wrote the paper: DY.