Mitral valve surgery for mitral regurgitation caused by Libman-Sacks endocarditis: a report of four cases and a systematic review of the literature

A 49-year-old Caucasian man presented at our institution with SLE that had been diagnosed originally in August 1996. Manifestations of his disease included arthritis, a rash on sun-exposed skin, and skin lesions resembling urticaria. Laboratory findings are shown in Table 1. A skin biopsy revealed urticarial vasculitis. There was no evidence of cerebral or renal involevement. His therapy for SLE required long-term plaquenil and prednisone. In September 1997 the patient was admitted with progressive exertional dyspnoea, cardiac decompensation, and a blowing systolic murmur at the apex radiating to the left axilla. Transthoracic (TTE) and transesophageal echocardiography (TEE) revealed severe MR with thickened mitral valve leaflets and a small vegetation on the posterior mitral valve leaflet. Repeated blood cultures were negative and there was no other evidence of infective endocarditis. The patient was recompensated with diuretics and discharged. Echocardiographic follow-up over the following months revealed a rapid increase in left ventricular diameters and normal left ventricular (LV) function. Results of cardiac catherization are shown in Table 1. The patient underwent mitral valve repair in March 1998. Intraoperative inspection showed slightly thickened, but otherwise surprisingly normal leaflets. A small perforation was found in the P2 section of the posterior leaflet. Preoperatively a small vegetation was found near this location. Although rare and more often seen in infectious endocarditis, leaflet perforation in LS endocarditis has been reported before [16]. This patient's history did not reveal any documented thromboembolic events. A quadrangular resection of the P2 section of the posterior mitral valve leaflet was performed, followed by implantation of a 32 mm Carpentier-Edwards Classic annuloplasty ring. Microscopic examination of the excised mitral valve segment revealed myxoid degeneration and no evident signs of inflammation. Although evidence of LS endocarditis could not be found microscopically, the diagnosis was made based on the clinical features, laboratory findings, and echocardiographic appearance. The patient's recovery from surgery was uneventful, and he was discharged on the seventh postoperative day. Echocardiographic follow-up revealed stable slight MR from April 1998 through January 2009. When last seen in March 2009, the patient was doing well, except for a mild degree of dyspnoea.

A 56-year-old Caucasian man presented at our institution with severe SLE that had been diagnosed originally in July 2003. Manifestations of his disease included arthritis, pericarditis, and pleuritis without any evidence of skin, cerebral or renal involvement. Laboratory findings are shown in Table 1. His therapy for SLE required long-term prednisone, plaquenil and azathioprine. On routine examination in 2006 the patient appeared to have a blowing systolic murmur at the apex radiating to the left axilla. Transthoracic echocardiography (TTE) revealed mitral valve thickening with focal vegetations and severe MR. Repeated blood cultures were negative and there was no other evidence of infectious endocarditis. Results of cardiac catheterization are shown in Table 1. The patient underwent mitral valve replacement with a 31 mm St. Jude mechanical prosthesis in October 2007. The excised mitral valve was thickened and fibrotic with focal vegetations. Microscopic pathologic examination of the excised mitral valve revealed fibrosis, neovascularization, and vegetations with fibrin-platelet thrombi and evident inflammatory cell infiltration (Fig 2A,B). LS endocarditis of the mitral valve was confirmed. The patient's recovery from surgery was uneventful, and he was discharged on the seventh postoperative day. Echocardiographic follow-up revealed no MR. When last seen in April 2009, the patient was doing well.

A 28-year-old Caucasian woman was referred to our institution in October 2006 with arthralgias and intermittent haemoptysis. She had a missed abortion earlier that year, when she was nine weeks pregnant. Laboratory findings are shown in Table 1. The patient was diagnosed with primary APS. In November 2006 she presented with exertional dyspnoea and a blowing systolic murmur at the apex radiating to the left axilla. Transthoracic echocardiography (TTE) revealed mitral valve leaflet thickening with small vegetations on the edges of both leaflets (Fig 3A,B) and severe MR with backflow into the pulmonary veins (Fig 3C). Repeated blood cultures were negative and there was no other evidence of infectious endocarditis. Results of cardiac catheterization are shown in Table 1. The patient underwent mitral valve replacement in October 2007. Intraoperative inspection revealed thickened and fibrotic mitral valve leaflets with focal vegetations (Fig 3D,E). Therefore, mitral valve repair was not considered possible and the mitral valve was replaced with a 31 mm St. Jude mechanical prosthesis. Microscopic pathologic examination of the excised mitral valve revealed myxoid and hyaline degeneration, fibrosis, and vegetations with fibrin-platelet thrombi and evident inflammatory cell infiltration (Fig 2C,D). LS endocarditis of the mitral valve was confirmed. The patient's recovery from surgery was uneventful, and she was discharged on the seventh postoperative day. Echocardiographic follow-up revealed no MR. When last seen in June 2009, the patient was doing well.

A 22-year-old Hispanic woman with a history of hypothyreoidism was referred to our institution in October 2007 after a transient ischemic attack of the right cerebral hemisphere with temporary left hemiplegia. Routine trans-thoracic echocardiography revealed a tumor with a diameter of approximately 1 cm on the atrial side of the posterior mitral valve leaflet (Fig 4A,B) as the source of this thrombo-embolic event. In addition, a normal LV function and moderate (grade 2+) MR was found (Fig 4C). Repeated blood cultures were negative and there was no other evidence of infectious endocarditis. Based on her history and the echocardiographic appearance of the tumor the initial working diagnosis was papillary fibroelastoma. Laboratory findings are shown in Table 1. The patient was diagnosed with primary APS. Subsequently, LS endocarditis of the mitral valve was considered as an alternative diagnosis. To prevent future thrombo-embolic events the patient was accepted for mitral valve surgery. Cardiac catherization was not performed. The patient underwent mitral valve repair in March 2008. Intraoperative inspection showed a large verrucous tumor on the atrial side of the P2 section of the posterior mitral valve leaflet (Fig 4D). A quadrangular resection of the P2 section of the posterior mitral valve leaflet was performed (Fig 4E), followed by implantation of a 28 mm Cosgrove-Edwards annuloplasty ring. Microscopic examination of the excised mitral valve segment revealed myxoid degeneration and large vegetations with fibrin-platelet thrombi, but without an evident inflammatory infiltrate (Fig 2E,F). The initial working diagnosis of papillary fibroelastoma could not be confirmed on microscopic examination. A definite diagnosis of LS endocarditis was made. The patient's recovery from surgery was uneventful, and she was discharged on the seventh postoperative day. Echocardiographic follow-up after 1.5 years revealed no recurrence of MR. When last seen in September 2009, the patient was doing well.

At this point the exact pathogenesis of LS endocarditis is still unclear. The initial insult to the valve, which causes endothelial damage and elicits the pathogenetic sequence of events, has not yet been identified. However, LS endocarditis has been assumed to involve the formation of fibrin-platelet thrombi on the altered valve, the organization of which leads to valve fibrosis, edema, diffuse thickening, mild inflammatory changes, valve distortion, scarring, and subsequent valvular dysfunction [5, 7, 17‐19]. Both valve thickening and formation of vegetations represent different stages of the same pathological proces [5]. Immunologic injury has been postulated as a possible initiating insult, since immunofluorescent microscopy revealed deposition of immunoglobulins and complement on affected valves [5, 19]. Rather than playing a more direct pathogenetic role, aPLs are thought to promote thrombus formation on the endothelium of valves already compromised by immune complex deposition, leading to further valvular damage and inflammation [5, 8, 15, 17, 18, 20, 21].

Valvular LS lesions are microscopically characterized by fibrin deposits at various stages of fibroblastic organization, neovascularization, occasional haematoxylin bodies, and by a variable extent of inflammation with mononuclear cell infiltration [6, 7]. Valvular lesions change over time [10] and the end-stage or healed form of LS verrucous endocarditis is a fibrous plaque, sometimes with focal calcification [5]. If the lesions are extensive enough, their healing may be accompanied by marked scarring, thickening, and deformity of the valve [5].

Microscopically, the mitral valve vegetations seen in SLE are distinct from those seen in (primary) APS. A rather remarkable difference is the absence (or minimal extent) of inflammatory cell infiltration in (primary) APS [5, 22]. To emphasize this difference some authors prefer to use the term NBTE for the valve lesions seen in primary APS instead of the term LS endocarditis. However, others (including the authors of this study) prefer to use the term LS endocarditis, because the two underlying diseases are both auto-immune phenomena, are often interrelated (APS secondary to SLE), and probably share a (partially) similar pathologic pathway in causing valve lesions.

Due to its asymptomatic nature establishing a diagnosis of LS endocarditis can be rather difficult. This is further complicated by the fact that the condition can mimick intracardiac tumors [23‐25] and bacterial endocarditis ("pseudoinfective" endocarditis) [26] or may coexist with (superimposed) bacterial endocarditis (also known as "double-decker" endocarditis) [7, 26, 27].

The modified Duke criteria can be useful in helping differentiate between true infective endocarditis and LS endocarditis [28]. Helpful laboratory markers in distinguishing infective endocarditis from LS endocarditis are the white blood cell count (elevated in infective endocarditis and often decreased in LS endocarditis), C-reactive protein levels (elevated in infective endocarditis and relatively low in LS endocarditis), aPL levels (normal in infective endocarditis and moderate to high in LS endocarditis), and (repeated) blood cultures (positive in infective endocarditis and negative in LS endocarditis) [21, 29]. Echocardiographically, LS vegetations appear as valve masses of varying size and shape with irregular borders and echodensity, they are firmly attached to the valve surface and exhibit no independent motion [8]. Contrary to the vegetations of infective endocarditis, which typically exhibit independent motion [30].

As previously demonstrated [23‐25] and as we showed in patient 4, differentiation from intracardiac tumors can also be difficult. Although LS vegetations are usually typically sessile, wartlike, and small, varying from pinhead size to 3-4 mm [5], they can become rather large making them difficult to distinguish (echocardiographically) from a typical mitral valve tumor such as papillary fibroelastoma. On echocardiography papillary fibroelastoma usually arise via a pedicle from mitral valve tissue or adjacent endocardium, and have a characteristic frond-like appearance [31]. A remarkable feature that LS vegetations do not posses.

Recently, a prospective randomized controlled study showed that TEE was superior to TTE in diagnosing LS endocarditis [32]. Nevertheless, establishing the diagnosis remains challenging.

Corticosteroids do not prevent LS endocarditis, but they facilitate healing of LS lesions over time by decreasing the amount of inflammation [5, 33‐35]. However, they can increase fibrosis and scarring, ultimately worsening valvular damage and dysfunction [5, 33‐35]. Nonetheless, appropriate steroid therapy to control SLE disease activity is important.

The risk of thrombo-embolic events (mainly stroke and transient ischaemic attacks) is increased in LS endocarditis [8]. Current therapeutic guidelines for APS include thrombo-embolic prevention with long-term anticoagulation [5]. In addition, patients with LS endocarditis who have suffered a thrombo-embolic event are recommended to be on lifelong anticoagulation for prevention of future thrombo-embolic events [5]. Moreover, implantation of a mechanical valve requires anticoagulation and atrial fibrillation is frequently a concomitant condition necessitating anticoagulation in patients with severe MR. In other words, lifelong anticoagulation can often not be avoided in these patients.

In most patients hemodynamically important valvular dysfunction can be controlled with conservative treatment (i.e immunosuppression, anticoagulation, endocarditis profylaxis, and specific heart failure treatment including ACE-inhibitors, beta-blockers, diuretics) [5, 36, 37]. However, if severe symptomatic valvular dysfunction persists mitral valve surgery may be required.

In contemporary cardiac surgery mitral valve repair has become the mainstay of surgical treatment for most causes of MR. Particularly, in the last two decades there has been a gradual shift from mitral valve replacement to mitral valve repair for MR caused by a broad range of etiologies. Several general advantages of mitral valve repair over replacement include a lower operative mortality rate, higher survival rates, better maintainance of left ventricular function, a lower risk of endocarditis, a lower risk of thrombo-embolic complications, less use of lifelong anticoagulation, and lower costs [38, 39].

In recent years mitral valve repair for significant MR due to LS endocarditis has been described in 10 patients (Table 2) [16, 23, 24, 40‐45]. Unfortunately the exact surgical repair techniques were not described in 4 of these patients [16, 42, 44, 45]. In this report we added two mitral valve repair cases to the literature with a precise description of mitral valve pathology and mitral valve repair techniques. In both patients (one with SLE and one with primary APS) intraoperative macroscopic examination revealed only localized abnormalities with otherwise relatively normal leaflets. Therefore mitral valve repair was considered a good surgical option in these two patients. Echocardiographic and clinical follow-up of both patients after 11 and 1.5 years, respectively, showed excellent results and no recurrence of MR. To our knowledge, this is the longest follow-up ever described after mitral valve repair for MR caused by LS endocarditis.

Some studies suggest that results of mitral valve replacement (MVR) are usually superior to repair for LS endocarditis [41, 42]. According to these studies, severe (ongoing) calcification and fibrosis lead to rapid recurrence of MR after repair with a subsequent reoperation and MVR [41, 42]. In our systematic review we found two cases of MVR after initial repair for MR due to LS endocarditis [41, 42]. In the first case severe mitral stenosis developed 6 months after mitral valve repair due to ongoing calcification [41] and in the second case severe MR recurred 29 months after repair [42].

Taken together, we believe mitral valve repair for LS endocarditis of the mitral valve can be justified in specific patients. If SLE and/or APS has been relatively stable (with immunosuppresive therapy) in a young patient, if intraoperative macroscopic examination reveals relatively normal leaflets with only localized abnormalities, and if repair seems feasible, then mitral valve repair is in our opinion certainly justified and probably the preferred surgical option. Especially in young females, who are likely to become pregnant in the near future, long-term anticoagulation is preferably avoided by mitral valve repair. However, as previously mentioned, anticoagulation is often still necessary in LS endocarditis and APS to prevent future thrombo-embolic events. In that case mitral valve replacement with preservation of the subvalvular apparatus may be a good surgical alternative.

In case of mitral valve replacement, prosthetic valve selection is highly individualized based on age and other factors. In patients with bleeding abnormalities the superiority of bioprosthetic valves over mechanical valves is clear. However, mechanical valve replacement is recommended in patients at low risk with anticoagulation and at high risk for bioprosthetic valve calcification. Although succesful placement of porcine bioprostheses in patients with LS endocarditis has been reported (Table 2), complications can arise. In our systematic review we found one case in which a bioprosthetic porcine valve had to be replaced, because it was affected by (rapid) calcification and valvulitis and subsequent perforation [46] and another case in which massive bioprosthetic thrombosis occurred [47]. Again stressing the importance of anticoagulation in this disease. In addition, (SLE associated) renal failure can accelerate bioprosthetic degeneration as a consequence of abnormal calcium and phosphate metabolism [40]. Therefore, a mechanical prosthesis may provide better results than a bioprosthesis for MR caused by LS endocarditis, even though mechanical prostheses carry a higher risk of thrombo-embolic complications.

Evidently, no definite consensus has been reached at this point as to whether or not these valves should be replaced or repaired and whether a mechanical prosthesis is more advantageous than a bioprosthesis.