CASE REPORTS

From the Ca

Hospital, Mila

Reprint reque

of San Raffae

antonio.grima

0894-7317/$3

Copyright 201

doi:10.1016/j.

‘‘Grey Zone’’ Patterns of Unexplained Endocarditis:Still a Challenge for Clinical Decision Making

Antonio Grimaldi, MD, Maurizio Taramasso, MD, Francesco Maisano, MD, Giovanni La Canna, MD,Maria Grazia Pala, MD, Stefano Benussi, MD, Giorgio Vigano, MD, and Ottavio Alfieri, MD, Milan, Italy

The authors report two cases of unexplained active inflammatory endocarditis with totally different clinicalpresentations. The patients had undergone previous mitral repair surgery and were referred for multiple softmobile masses on the mitral ring without clinical or laboratory signs of endocarditis. Serologic screeningand blood culture results were negative, including those for specific fastidious bacteria, as well asimmunologic tests to rule out ‘‘nonbacterial thrombotic endocarditis.’’ Before new surgery, both patientswere treated with long-term antibiotic and anticoagulant therapy, with no significant changes in clinical settingand echocardiographic patterns. In neither case was it possible to characterize a specific microorganism: theintraoperative findings were highly evocative of active endocarditis with a macroscopic infiltration of the mitralring, and culture results from surgical material and valvular tissue were negative. (J Am Soc Echocardiogr2010;23:221.e1-221.e4.)

Keywords: Mitral valve repair, Acute endocarditis, Three-dimensional echocardiography

Infective endocarditis can usually be diagnosed by positive bloodculture results and the detection of vegetations on echocardiography.In one third of patients, blood culture results are negative as well aslaboratory signs evocative of infective endocarditis, and the differen-tial diagnosis between vegetations and thrombosis may be challeng-ing. Serologic screening and cultures from infected valve tissue areusually able to detect specific etiologies.

We report two unusual cases of patients with soft mobile masses onthe mitral valve (MV) after previous repair surgery and withoutclinical and laboratory signs of endocarditis. Before new surgery,both patients were treated with long-term antibiotic and anticoagu-lant therapy, with no change in echocardiographic patterns. In neithercase were we able to characterize a specific microorganism.

CASE REPORTS

Case 1

A 64-year-old man with a clinical history of prior MV repair under-went new surgical intervention because of recurrent transient ische-mic attacks due to vegetations on the MV. No report of surgicalfindings was available.

Six months before he was admitted to our hospital, echocardio-graphic follow-up showed new masses on the MV, so the patientreceived empiric antibiotic and anticoagulation therapy according tothe embolic sources detected. Clinical follow-up was characterized

rdiovascular and Thoracic Department, San Raffaele University

n, Italy.

sts: Antonio Grimaldi, MD, Cardiovascular and Thoracic Department

le University Hospital, via Olgettina, 60, 20132 Milan, Italy (E-mail:

ldi@hsr.it).

6.00

0 by the American Society of Echocardiography.

echo.2009.08.010

by recurrent ischemic events without laboratory signs of infection, not-withstanding the persistence of valvular masses. Three sets of bloodculture and thrombophilic and immunologic screening results werenormal, including coagulation factors, lipoprotein(a), anticardiolipinand antiphospholipid antibodies, homocysteinemia, lupus-like antico-agulant, protein C, protein S, antiheparin antibodies, b2 glycoprotein,and b2 lipoprotein. Serologic test results for Mycoplasma spp, Legionellaspp, Bartonella spp, Coxiella burnetii, Brucella spp, and Clamydia sppwere negative. Transesophageal echocardiography confirmed multi-ple masses 20 mm in length on the atrial side of the mitral ring, withsoft texture (Figure 1A, Video 1A) at high risk for embolism. No signif-icant mitral stenosis or regurgitation was observed (Figure 1B, Video1B). A 3-dimensional reconstruction showed the elective involvementof the mitral annulus, with a large thrombus-like vegetation arisingfrom the prosthetic ring (Figures 2A and 2B, Videos 2A and 2B).The patient then underwent MV replacement.

The histologic report described neoangiogenesis withinflammatory cell infiltrate. Gram’s method showed rare extracellularorganisms compatible with gram-negative bacteria; this finding wassuggestive, but not specific, of HACEK group (Haemophilus parain-fluenzae, H aphrophilus, and H paraphrophilus; Actinobacillus actinomyce-temcomitans; Cardiobacterium hominis; Eikenella corrodens; and Kingellaspp) microorganism infection.

Case 2

A 63-year-old man with a clinical history of MV repair was admittedto our hospital. Eight months after surgery, echocardiographic follow-up showed multiple masses on the MV, without symptoms. Nohistory of fever or clinical embolism was reported, and no abnormal-ities of white blood cell count or inflammatory indices were detected.Repeated blood culture and serologic test results for fastidious bacte-ria were negative, as well as coagulative function and thrombophilicscreening. Immunologic tests ruled out any form of ‘‘nonbacterialthrombotic endocarditis’’ (NBTE). Transesophageal echocardiogra-phy showed multiple, movable masses 30 mm in length on the atrialside of the mitral prosthetic ring of ambiguous morphologic texture

221.e1

Figure 1 Floating soft masses on the atrial side of the prosthetic mitral ring (A) without significant valve insufficiency (B).

Figure 2 Transesophageal three-dimensional reconstruction of the MV showing the elective involvement of the annulus, with a largethrombus-like vegetation arising from prosthetic ring.

221.e2 Grimaldi et al Journal of the American Society of EchocardiographyFebruary 2010

related to vegetation or thrombus (Figure 3A, Video 3A). The entirering was involved in the inflammatory reaction and surrounded bysoft, nonfibrotic tissue (Figure 3B, Video 3A). The leaflets appearedto be thick, but no perforation or loss of tissue was observed, andsignificant mitral stenosis was absent (Figure 3C, Video 3B). Whole-body computed tomography ruled out subclinical systemicembolism. After 20 days of antibiotic therapy and heparin, the clinicalsetting and the echocardiographic pattern were unchanged, so thepatient underwent new surgical intervention.

The tissue removed was analyzed, and the histologic reportshowed no microorganisms; it only described a nonspecific fibrinoin-flammatory tissue and neutrophilic infiltration with minimal focalnecrosis. All bacterial culture results from the prosthetic ring materialand valvular tissue were negative.

DISCUSSION

Infective endocarditis can usually be evoked by fever, and thediagnosis is enhanced by the detection of vegetations and positiveblood culture results. In one third of patients with infective endocar-ditis, blood culture results are negative according to previous antibi-otic therapy or fastidious growing bacteria (negative blood cultureinfective endocarditis).1 However, modern techniques and culturesfrom infected valves are usually able to detect specific etiologies. If

there are no clinical features of endocarditis, the differential diagnosisbetween valvular vegetations and thrombosis may be challenging,and the potential risk of anticoagulant therapy in thrombus-like infec-tive masses is also well known.2,3 Serologic screening and culturesfrom the material collected during surgery usually support the diagno-sis.4,5 The cases described are particularly interesting and representa-tive of a ‘‘grey zone’’ population with discrepancies between clinicalpresentation and echocardiographic findings of embolic sourcesand a total absence of laboratory patterns evocative of ‘‘conventional’’endocarditis.

In both patients, serologic screening and blood culture results werenegative, including those for specific fastidious organisms requiringspecial tools (Coxiella, Legionella, the HACEK group, etc), as well asimmunologic tests to rule out NBTE.6 Immunologic disorders (anti-phospholipid syndrome, Libman-Sacks endocarditis) and variousconditions (hypercoagulable state, cancer, human immunodeficiencyvirus infection, old age) can be associated with potentially life threat-ening source of thromboembolism related to vegetations on thecardiac valves consisting of fibrin and platelet aggregates.7-10 Nospecific clinical features allow for the diagnosis of NBTE, but the asso-ciation of the underlying immunologic disease together with systemicemboli and cardiac valve sources is highly evocative of diagnosis.11 It isessential to differentiate infective endocarditis from NBTE, whereasrecurrent emboli should be considered a hallmark feature occurringin up to 50% of patients, with a tendency to embolize to the brain,kidneys, and spleen.6 Fresh vegetations are very easy friable and

Figure 4 Intraoperative findings of mitral ring endocarditis: large floppy vegetation in the first patient (A) and nonspecific inflammatoryreaction of the prosthetic ring in the second patient (B).

Figure 3 Masses attached to the atrial side of the mitral prosthetic ring of ambiguous morphologic texture (A). The entire prostheticring was involved in the inflammatory reaction (B), without structural damage or significant mitral stenosis (C).

Journal of the American Society of EchocardiographyVolume 23 Number 2

Grimaldi et al 221.e3

usually readily dislodge and embolize. Clinical signs, such cardiacmurmurs, fever, and leukocytosis, are usually absent, and immuno-logic assays should be strongly considered to detect the underlyingdisease. However, in the two patients examined here, all sets of bloodculture results and all immunologic screening results were negative, aswell as serologic tests for Mycoplasma spp, Legionella spp, Bartonellaspp, C burnetii, Brucella spp, and Clamydia spp. Moreover, whole-body computed tomography ruled out any subclinical splenic orsystemic embolism.

Because of the persistent high risk for embolization, theineffectiveness of both antimicrobial and anticoagulant therapy anddespite no evidence of valvular damage or heart failure, we acceptedthe endocarditis hypothesis and considered surgical intervention. Theresults of all serologic tests were also negative, and no specifictreatment was started to correct potential underlying immunologicdisorders.

When recurrent embolic events are the predominant clinical fea-ture, almost immediate surgery is needed according to persistent sour-ces totally refractory to empiric therapy and unchanged pattern atechocardiographic follow-up; in this case, as in the first patient, theneed to control the ischemic events is the main determinant of clinicaldecision making: the patient will be sent to surgery with or withouta precise diagnosis of embolic substrate and despite the risk ofanticoagulation in a supposed active endocarditis.

In the first patient, the masses were confirmed by inspection(Figure 4A), and the results of microorganism cultures, includingscreening for HACEK organisms and Koch Bacillus, were negative.

The histologic report described microabscesses with inflammatorycell infiltrate and neoangiogenesis. Truant and Ziehl-Neelsen colora-tion results were negative. Gram’s method showed rare extracellularorganisms compatible with gram-negative bacteria but not specific forHACEK infection. At this moment, 3 months after surgery, thepatient is asymptomatic, and no specific microorganisms have beencharacterized.

Also in the second patient treated, the intraoperative findingswere highly evocative of active endocarditis, with a macroscopic mas-sive involvement of the prosthetic annulus (Figure 4B), so thevalve was replaced. The results of bacterial cultures from the pros-thetic ring material and valvular tissue were negative, and the histo-logic examination did not reveal bacterial growth describinga fibrinoinflammatory and neutrophilic infiltration with focal necrosis.

CONCLUSIONS

The two cases examined are examples of active inflammatoryprosthetic endocarditis with a nonconcordant pattern between clini-cal presentation (totally different even in the presence of the sameprocess), laboratory findings (negative in both cases), andmacroscopic disease (as detected on surgical inspection).

The origins of embolic sources in both patients were related toactive massive inflammatory infiltrate of the prosthetic ring with veg-etating masses of unexplained nature, and cultures from the materialcollected during surgery did not support a clear, specific origin. This

221.e4 Grimaldi et al Journal of the American Society of EchocardiographyFebruary 2010

population may represent a ‘‘grey’’ zone whereby differentcomponents interact with uncertain background: high embolic riskrelated to friable and active vegetations on cardiac valves, poor sys-temic response with massive inflammatory local reaction, and a totallynonspecific serologic pattern. As for the two cases described, the clin-ical decision making will be made on a case-by-case basis with regardto the balance between the need to control the clinical symptoms andthe risk of delaying the diagnosis of a potential underlying disease.

REFERENCES

1. Tariq M, Alam M, Munir G, Khan MA, Smego RA Jr. Infective endocarditis:a five-year experience at a tertiary care hospital in Pakistan. Int J Infect Dis2004;8:163-70.

2. Baddour LM, Wilson WR, Bayer AS, Fowler VG Jr, Bolger AF, Levison ME,et al. Infective endocarditis: diagnosis, antimicrobial therapy, and manage-ment of complications: a statement for healthcare professionals from theCommittee on Rheumatic Fever, Endocarditis, and Kawasaki Disease,Council on Cardiovascular Disease in the Young, and the Councils onClinical Cardiology, Stroke, and Cardiovascular Surgery and Anesthesia,American Heart Association: endorsed by the Infectious Diseases Societyof America. Circulation 2005;111:e394-434.

3. Tornos P, Almirante B, Mirabet S, Permanyer G, Pahissa A, Soler-Soler J.Infective endocarditis due to Staphylococcus aureus: deleterious effectof anticoagulant therapy. Arch Intern Med 1999;159:473-5.

4. Brouqui P, Raoult D. New insight into the diagnosis of fastidious bacterialendocarditis. FEMS Immunol Med Microbiol 2006;47:1-13.

5. Naber CK, Erbel R. Infective endocarditis with negative blood cultures. IntJ Antimicrob Agents 2007;30(suppl):S32-6.

6. Lopez JA, Ross RS, Fishbein MC. Nonbacterial thrombotic endocarditis.A review. Am Heart J 1987;113:773-84.

7. Raoult D. Afebrile blood culture-negative endocarditis. Ann Intern Med1999;131:144-6.

8. Moyssakis I, Tektonidou MG, Vasilliou VA, Samarkos M, Votteas V,Moutsopoulos HM. Libman-Sacks endocarditis in systemic lupus erythe-matosus: prevalence, associations, and evolution. Am J Med 2007;120:636-42.

9. Hojnik M, George J, Ziporen L, Shoenfeld Y. Heart valve involvement(Libman-Sacks endocarditis) in the antiphospholipid syndrome.Circulation 1996;93:1579-87.

10. el-Shami K. Griffiths E, Streiff M. Nonbacterial thrombotic endocarditis incancer patients: pathogenesis, diagnosis, and treatment. Oncologist 2007;12:518-23.

11. McKay DG, Wahler GH Jr. Disseminated thrombosis in colon cancer.Cancer 1955;8:970-8.