Prosthetic valve endocarditis

J

THORAC CARDIOVASC SURG

92:784-789, 1986

Prosthetic valve endocarditis The case for prompt surgical management Clinical and morphologic features are described in 27 patients with prosthetic valve endocarditis. The interval from valve replacement to onset of symptoms of prosthetic valve endocarditis was less than 2 months in 10 patients, longer than 2 months but less than 6 months in seven patients, and longer than 6 months in 10 patients. The most frequent infecting organism was Staphylococcus (11 patients). In nearly aUpatients, infection spread behind the site of attachment of the valve prosthesis and resulted in valve ring abscesses. Twenty-three of the 28 infected prostheses were partiaUy or almost completely detached, and in 15 patients the infection destroyed the entire valve anulus, burrowing to adjacent structures in six. Despite prolonged bactericidal antibiotic therapy, bacterial cultures of prosthetic valves removed at operation or autopsy were positive in 14 patients. Standard valve replacement was attempted in nine patients. All were hospital survivors, but two of these patients evidenced rapid postoperative valve dehiscence and required a complex surgical procedure at reoperation. The 14 other surgically treated patients had almost complete destruction of the annular root, and surgical repair was achieved by complex surgical techniques. There were five postoperative deaths, but nine patients survived with no further evidenceof infection (mean foUow-up 34 months). All patients with early prosthetic valve endocarditis who recovered underwent this type of operative technique. Total exclusion of the infected annular root, as described, may offer in patients with extensive endocarditic lesions the only possibility to eradicate the infection and to reduce the mortality.

Catherine Rocchiccioli, M.D., Jean Chastre, M.D., Yves Lecompte, M.D., Iradj Gandjbakhch, M.D., and Claude Gibert, M.D., Paris, France

]?osthetic valve endocarditis (PVE) remains a devastating complication of cardiovascular surgery, with an incidence of about 2%.1.3 The overall mortality described in recent reports is still very high, 78% in early PVE and 46% in late PVE.I.9 Prosthetic valve replacement has improved the prognosis over that obtained with antimicrobial therapy alone. I, 5·9 However, the optimal date and nature of the surgical procedure both remain to be determined. This report presents 27 cases of PVE in 26 severely ill patients. Because in many patients anatomic lesions were extremely important, so that it was difficult to seat the new prosthesis on the original anulus, complex From the Service de Reanimation Medicale (Hopital Bichat), the Centre Medico-Chirurigical de la Porte de Choisy, and the Departernent de Chirurgie Thoracique et Cardiovasculaire (Hopital Pitie-Salpctriere), Paris, France. Received for publication Feb. 28, 1985. Accepted for publication Jan. 2, 1986. Address for reprints: Dr. C. Gibert, Service de Reanimation Medicale, Hopital Bichat, 46 rue Henri Huchard, 75018 Paris, France.

784

surgical procedures were performed in 16. The purposes of this study were to summarize the results obtained in these patients and to better define the respective indications of the different surgical procedures. Patients and methods Study population. Between June 1, 1979, and May 31, 1983, 26 patients (18 male and eight female patients) were hospitalized in the intensive care unit of Bichat hospital for 27 episodes of infective endocarditis after cardiac valve replacement. One patient had two distinct episodes of PVE and his history was treated as two separate cases. PVE was diagnosed if at least two of the following three criteria were present: (1) a clinical illness compatible with bacterial endocarditis (two or more of these symptoms-fever; new or changing heart murmur; presence of petechiae, splinter hemorrhages, Roth spots, Janeway lesions, or Osler nodes; unexplained emboli; or splenomegaly); (2) two or more positive blood cultures for the same organism, either without an identifiable extracardiac source or, if associ-

Volume 92 Number 4 October 1986

ated with an identifiable extracardiac source, with evidence of valvular involvement by new or changing murmur, evidence of vegetations by echocardiogram, or confirmation at operation or autopsy; (3) histopathologic or microbiologic evidence of valvular infection at operation or autopsy. Twenty-six of the 27 episodes of PVE were eventually proved pathologically. Patients were divided into three groups, according to whether endocarditis occurred within 2 months (early PVE), later than 2 months but within 6 months (semi-late PVE), or later than 6 months (late PVE) postoperatively. Medical treatment. All patients received high intravenous doses of at least two synergistic antibiotics for a prolonged period. The choice of antibiotics was determined by microbial susceptibility, and the suitability of the association was confirmed in every case by determination of serum bactericidal titers. The effectiveness of antimicrobial therapy was judged by the fall of the patient's temperature and blood cultures. Surgical treatment. Emergency valve replacement was considered in all patients with congestive cardiac failure despite vasoactive drugs, evidence of conduction system disturbances, uncontrolled infection defined by positive blood cultures or temperature greater than 38.5° C (2 and 8 days, respectively, after the onset of medical treatment), and/or evidence of two or more systemic emboli. Anatomic lesions observed at operation or autopsy were assessed in each patient by using a semiquantitative scale, according to the presence or absence of annular ring abscesses, prosthetic valve dehiscence, and vegetations. The extent of annular ring abscesses was graded from I+ to 3+ in reference to their size, with 1+ being a localized abscess, 2+ an abscess involving almost the entire valve anulus, and 3+ an abscess extending to adjacent structures. Prosthetic valve dehiscence was also graded from 1+ to 3+, with I + being a prosthesis detached from the anulus by less than one third of its circumference, 2+ by more than one third but less than two thirds, and 3+ by more than two thirds. The same semiquantitative scale was used to assess the size of vegetations, with I + corresponding to small vegetations, 2+ to large vegetations but without valve obstruction, and 3+ to a prosthesis obstructed by vegetative material. After removal of the infected prosthesis and adequate excision of the necrotic tissues, one of the six different techniques described herein was used to insert a new prosthesis. In all instances, the bed of the infected root was treated with weak iodine solution. If the mitral ring anulus was not excessively damaged by infection, the valve was replaced in the original

Prosthetic valve endocarditis

785

Fig. 1. Diagram of intra-atrial insertion of a mitral prosthesis with a Dacron flange.

position by standard technique (mitral type I operative procedure). If there was extensive annular and subannular abscess around the mitral prosthesis, the new mitral valve with a Dacron flange was implanted in the left atrium 1.5 to 2 em above the mitral ring and sutured to the atrial wall (Fig. I), as described by Gandjbakhch and associates" (mitral type II operative procedure). In the case of aortic valve involvement, one of the following techniques was used to restore ventricularaortic continuity: (I) Standard valve replacement on the original anulus was performed (aortic type I operative technique). (2) The prosthesis was implanted on the physiologic anulus, but the sutures were tightened outside the heart. Depending on the site of the lesions, this was done by opening the right atrium or the infundibulum of the right ventricle and by passing the stitches through the aortic wall according to the technique described by Laborde, Lecompte, and Bex" (Fig. 2) (aortic type II operative technique). As in most of the patients in whom type I or II operative techniques were employed, infection extended behind the site of attachment of the original valve prosthesis and resulted in small ring abscesses; these cavities were carefully curetted and obliterated with buttressed sutures that were subsequently used to secure the new prosthesis." (3) When the entire circumference of the aortic ring was involved, reconstruction was accomplished with a composite graft incorporating a Bjork valve in a woven Dacron tube graft. The composite prosthesis was sutured to the base of the heart with horizontal pledgetsupported mattress sutures through the ventricular septum, ventricular wall muscle, and mitral valve. The

The Journal of

7 8 6 Rocchiccioli et al.

Thoracic and Cardiovascular Surgery

( Fig. 2. Diagram of operative repair using aortic type II technique. The stitches are tightened (A) in the right ventricle through the infundibular incision, (B) in the right atrium through the auricular incision, and (C) outside the aorta over strips of Teflon felt. The new prosthesis is only indicated by its flange.

Table I. Distribution of patients with PVE by time of onset after operation and infecting organism Time of onset after operation Infecting organism Coagulase-negative staphylococci Staphylococcus aureus Streptococcit Enterococci Gram-negative bacilli Culture negative Total

Fig. 3. Diagram of operative repair using aortic type IV technique. Extensive annular and subannular abscess involving the origin of the coronary arteries is represented on the left. Reconstruction is achieved by a Dacron tube graft. The new prosthesis is implanted into the ascending aorta by transfixing stitches and the coronary circulation is reestablished by saphenous vein bypass grafts (right). PA, Pulmonary artery. CA, Coronary artery.

distal end of the graft was sutured into the lumen of the aorta just below the origin of the coronary arteries (aortic type III operative technique)." (4) When the infectious lesions of the aortic wall included the coronary artery origins, ventricular-aortic continuity was reestablished with a Dacron tube graft as in type III operative technique, but the upper margin of the tube was sewn to the transected aorta above the coronary artery ostia. The prosthetic valve was located either in the aorta itself by transfixing stitches or directly in the Dacron tube.

<2 mo 3

I

2-6 mo 1

4

I

>6 mo

2

Total

6"

0

1

0

0

0

2 2

3 2

2

2

1

5

0 10

3 7

3

6

10

27

* Five isolates were

methicillin resistant tExcludes enterococci.

The coronary artery ostia were closed and the coronary circulation was reestablished by saphenous vein bypass grafts from the aorta, distal to the prosthesis, to the coronary arteries (Fig. 3) (aortic type IV operative technique). 13, 14 In all instances, prosthetic valves removed at operation or autopsy were sent for pathologic examination and bacterial cultures. Mortality was considered postoperative when death occurred within 30 days of operation. Relapse of endocarditis was defined by the reappearance of the clinical syndrome described earlier, with positive blood cultures or valve cultures implicating the original organism. Long-term follow-up information was collected in June, 1985, by means of a questionnaire sent to each patient's cardiologist or by direct contact with the patient.

Volume 92 Number 4

Prosthetic valve endocarditis

October 1986

787

Table II. Anatomic findings and results of bacterial cultures of valves in 26 patients with PVE* Prosthesis infected

No. of patients

Aortic Mitral

17 II

Patients with vegetations] I+

5 5

I

2+ 4 6

I

Patients with PVDt

3+

I+

o o

1 2

I

2+

Patients with ARAt

3+

I+

12

4

5

2

6 2

I

I

2+

3+

I

Positive valve culture (No.)

6

4

8 7

o

5

Legend: PVD. Prosthetic valve dehiscence. ARA. Annular ring abscess. • An autopsy was not performed in one patient with early PYE. t For definition see text.

Results The patients' ages ranged from 8 to 71 years, averaging 42 years. The 27 patients had 33 prosthetic cardiac valves (25 mechanical valves and eight bioprostheses); 22 patients had one prosthetic valve and five had two or more prosthetic valves. There were 10 patients with early PVE, seven with semi-late PVE, and ten with late PVE. Factors predisposing to development of endocarditis were apparent in seven of the 10 patients with early PVE: intravenous catheters or thoracic drainage tubes (three patients), mediastinitis (three patients), and an infected scab on a foot (one patient). No source of infection was identified in 11 of the 17 patients with semi-late and late PVE. The six other cases of PVE were associated with dental or genitourinary manipulations. The infecting organisms are listed in Table I. Staphylococci caused seven of 10 cases of PVE during the initial 2 months and two of seven cases 6 months after operation. Patients with onset of PVE more than 6 months after valve replacement showed a different pattern of infecting organisms, with streptococci (n = 4) predominating. Surgical intervention was not performed in four patients, despite evidence of severe hemodynamic failure, because of active mediastinitis with aortic PVE in three and concomitant cerebrovascular accident in one. The other patients were operated on in addition to receiving antibiotic therapy. On 22 occasions, surgical intervention was performed while the patient was in severe congestive heart failure (New York Heart Association Class IV) despite medical treatment, and in six the infection was not controlled. High-grade conduction disturbances occurred in seven patients, all in the group with an infected aortic prosthesis. Four patients had two or more systemic emboli before operation. The duration of signs and symptoms of infective endocarditis from the onset of disease to operation or autopsy ranged from 5 to 90 days, with a mean delay of 28.5, 24, and 40 days for early, semi-late, and late PVE, respectively. Of the 30 prosthetic valves for which morphologic analysis was obtained, 28 were infected: the aortic

Table

m. Outcome of treatment Time of onset after operation

I 2-6 rno I >6 rno

Outcome

<2 rno

No. not operated No. requiring standard valve replacement No. reoperated No. discharged No. who died No. requiring complex surgical procedure No. discharged No. who died

3 I

I 3

0

I I

I 3

0

5

4* 9

2t

0

0

6

3

5

9 I 14

3 3

2

4 I

9 5

I

5

Total

J:I:

• All died in the hospital. tThese two patients underwent a complex surgical procedure at reoperation and finally recovered. :j:This patient died 24 months later of cardiac failure.

prosthesis in 17 patients and the mitral in 11. Sixteen of the 17 infected aortic valve prostheses and seven of the 11 infected mitral valve prostheses were partially or almost completely detached (Grade 2+ or 3+ prosthetic valve dehiscence in Table II). In contrast, six of the 11 infected mitral prostheses and only four of the 17 infected aortic prostheses were covered by large vegetative material. Ring abscesses were visible on gross examination of the anulus in 23 of the 28 infected prostheses. The ring abscess involved the entire valve anulus in 15 patients and only portions of the anulus in the other eight. In six of the 17 patients with an aortic valve prosthesis the infective process spread to adjacent structures; nevertheless, no intracardiac fistulous communication was observed. Patients with early PVE tended to have the more important lesions at operation or autopsy; seven of nine had extensive annular ring abscesses, compared to only eight of 17 patients with semi-late or late PVE. Despite prolonged (more than 5 days) bactericidal antibiotic therapy, bacterial cultures of prosthetic valves removed at operation or autopsy

7 8 8 Rocchiccioli et al.

were positive in 14 of the 26 patients (in seven of nine with early PVE, three of seven with semi-early PVE, and four of 10 with late PVE). Table III summarizes the outcome of treatment. Standard valve replacement, replacing the new prosthesis on the original anulus, was attempted in nine patients (on the aortic anulus in four and on the mitral anulus in five), despite the presence of Grade 1+ annular ring abscesses in five patients and a Grade 2+ abscess in one. However, two of these patients had rapid postoperative prosthetic valve dehiscence and required a complex surgical procedure at reoperation (aortic type IV procedure in one and mitral type II in the second); they finally recovered. Six of the remaining seven patients survived and were all well and symptom free with a mean follow-up of 49 months (range 28 to 70 months). The single death resulted from cardiac failure and was not related to relapse or to repair of the infected prosthesis. The 14 other patients who were operated on underwent a complex surgical procedure initially: aortic type II procedure in two, aortic type III procedure in five, aortic type IV procedure in four, and mitral type II procedure in four (in one patient, both the mitral and the aortic valves were replaced). In all these patients except one who underwent an aortic type II procedure, the anulus was almost completely destroyed (at least Grade 2+ aortic root abscess, Table II) by the infection, and in six the abscess spread through the adjacent structures. Nine of these 14 patients (64%) survived with no further evidence of infection, two after an aortic type II procedure, four after an aortic type III procedure, two after an aortic type IV procedure, and one after a mitral type II procedure. Follow-up was complete in all patients and averaged 34 months (range 13 to 58 months). Of the nine patients discharged from the hospital, all were alive at the time of this report, all with excellent results. Of the seven patients who had high-grade conduction disturbances preoperatively, all except one underwent complex aortic root reconstructive procedures. Five of these patients survived but two required permanent cardiac pacing for complete heart block. The overall mortality for the 23 patients who were treated surgically was 26% (6/23), with a mortality of 43% (3/7) for early, 17% 0/6) for semi-late, and 20% (2/10) for late PVE. Discussion

Despite major advances in cardiovascular surgical technique, postoperative care, and routine use of prophylactic antimicrobial agents, PVE continues to complicate the course of a small percentage of patients after cardiac

The Journal of Thoracic and Cardiovascular Surgery

valve replacement. Further advances in the treatment of this infection will depend on a better understanding of its pathophysiology and improvements in surgical management. Among our patients with PVE, the distribution of infecting organisms with time after operation showed some differences from previous studies. In the past, Staphylococcus aureus, gram-negative bacilli, and coagulase-negative staphylococci were the most frequent pathogens causing PVE with onset in the first 2 months after operation. Streptococci were the predominant cause of PVE after 2 months.!" In the present series, bacteria considered likely to have caused infection acquired outside of the hospital (such as streptococci or enterococci) did not appear prominently at 2 months after operation but rather only after 6 months had elapsed after valve implantation. However, although early and semi-late PVE had a nearly identical bacteriological spectrum in our study, their prognoses were quite different: The overall mortality was 60% for early PVE and only 28% for semi-late PVE. We consider it very important to point out these differences so that surgical management may be discussed with respect to the onset of PVE rather than the causative agent." In our study, the mean delay between the onset of endocarditis and operation or autopsy was 34 days, the duration of illness exceeding 60 days in only three patients. Nevertheless, in nearly all patients, infection spread behind the site of attachment of the valve prosthesis, resulting in valve ring abscesses. Twentythree of the 28 infected prostheses were partially or almost completely detached, and in 15 cases the infection destroyed the entire circumference of the valve, burrowing to adjacent structures in six. Moreover, prolonged bactericidal antibiotic therapy failed to sterilize the infected prosthesis in 14 patients. All of these facts should be kept in mind before cardiac valve replacement is postponed in an attempt to control infection preoperatively with antimicrobial therapy. Only prompt valve replacement may prevent annular ring abscess formation and extension of the infection into vital and/or inaccessible structures, lessen the possibility of paravalvular leaks or thrombotic stenosis, avoid major vessel embolization, and reduce the operative risk, which is mainly correlated with the degree of congestive heart failure at operation. J.5, 7, 8 In severe PVE with destruction of the anulus by abscesses, conventional valve replacement may not be technically possible or may predispose to rapid postoperative valve dehiscence, as in two of the nine patients in the present study who underwent a standard surgical procedure." Several innovative techniques have been

Volume 92 Number 4 October 1986

used to repair aortic root defects resulting from bacterial endocarditis. Initial successful treatment of aortic valve endocarditis and aortic root abscess was achieved by Danielson, Titus, and DufShane," who translocated the new prosthesis at the ascending aorta with supravalvular coronary artery vein grafts. Others avoided the weakened aortic anulus by interposition of a composite prosthetic valve-Dacron tube graft or a preserved homograft aortic valve.14, 18, 19 The advantages of these techniques are that they exclude the weakened and infected root from the high intravascular pressures and allow the new prosthesis to be sutured in continuity with healthy tissues. Long-term success with composite prosthetic valve-graft reconstruction will depend on the incidence of graft infection. Past experience has shown that valve replacement in the presence of active endocarditis is possible without contamination of the prosthesis." Prevention of this complication should depend on thorough debridement of the infected tissue and postoperative antimicrobial therapy. In the present series, complex surgical procedures were considered when there was extensive destruction of the anulus with massive abscess excavations or in case of postoperative failure, that is, recurrent endocarditis or postoperative prosthetic valve dehiscence. There were five postoperative deaths among the 16 patients who were treated by such techniques, but 11 patients survived with no further evidence of infection, with a mean follow-up of 34 months. All patients with early PVE who recovered underwent this type of operative technique. In conclusion, we believe that early surgical intervention must be considered in most cases of PVE before clinical evidence of severe hemodynamic deterioration. In patients with extensive lesions of endocarditis, a radical surgical approach may offer the only possibility to stabilize cardiac hemodynamics, to eradicate the infection, and to reduce the mortality.

Prosthetic valve endocarditis

6

7

8

9

10

II

12

.13

14

15 16

17

We are indebted to Miss C. Brun and Miss M. Kail for assistance in the preparation of the manuscript. REFERENCES

2 3

4

5

Dismukes WE, Karchmer A W, Buckley MJ, Austen WG, Swartz MN: Prosthetic valve endocarditis. Analysis of 38 cases. Circulation 48:365-377, 1973 Masur H, Johnson WD: Prosthetic valve endocarditis. J THoRAc CARDIOVASC SURG 80:31-37, 1980 Mayer KH, Schoenbaum SC: Evaluation and management of prosthetic valve endocarditis. Prog Cardiovasc Dis 25:43-54, 1982 Karchmer AW, Dismukes WE, Buckley MJ, Austen WG: Late prosthetic valve endocarditis. Clinical features influencing therapy. Am J Moo 64: 199-206, 1978 Kay PH, Oldershaw PJ, Lincoln JCR, Lennox SC, Paneth

18

19

20

789

M: The management of prosthetic valve endocarditis. J Cardiovasc Surg 24: 127-131, 1983 Saffle JR, Gardner P, Schoenbaum SC, Wild W: Prosthetic valve endocarditis. The case for prompt valve replacement. J THoRAc CARDIOVASC SURG 77:416-420, 1977 Wilson WR, Danielson GK, Giuliani ER, Geraci JE: Prosthetic valve endocarditis. Mayo Clin Proc 57: 155-161, 1982 Wilson WR, Jaumin PM, Danielson GK, Giuliani ER, Washington JA II, Geraci JE: Prosthetic valve endocarditis. Ann Intern Moo 82:751-756, 1975 Raychaudhury T, Cameron EWJ, Walbaum PR: Surgical management of prosthetic valve endocarditis. J THORAC CARDIOVASC SURG 86: 112-114, 1983 Gandjbakhch I, Laskar M, Pavie A, Mesnildrey P, Cabrol C: Implantation intra-atriale de la valve mitrale. Presse Moo 12: 1723-1724, 1983 Laborde F, Lecompte Y, Bex JP: Remplacement valvulaire aortique en cas de destruction infectieuse de l'anneau aortique. Nouvelle technique. Presse Med 14: 157-159, 1985 Symbas PN, Vlasis SE, Zacharopoulos L, Lutz JF: Acute endocarditis. Surgical treatment of aortic regurgitation and aortico-Ieft ventricular discontinuity. J THoRAc CARDIOVASC SURG 84:291-296, 1982 Gandjbakhch I, Villemot JP, Barrat J, Pavie A, Guiraudon G, Mattei MF, Cabrol C, Cabrol A: Traitement chirurgical des endocardites sur protheses valvulaires aortiques. Ann Chir 35: 158-162, 1981 Reitz BA, Stinson EB, Watson DC, Baumgartner WA, Jamieson SW: Translocation of the aortic valve for prosthetic valve endocarditis. J THoRAc CARDIOVASC SURG 81:212-218,1981 Moore-Gillon J, Eykyn SJ, Phillips I: Prosthetic valve endocarditis. Br Moo J 287:739-741, 1984 Buckley MJ, Mundth ED, Daggett WM, Austen WG: Surgical management of the complication of sepsis involving the aortic valve, aortic root and ascending aorta. Ann Thorac Surg 12:391-399, 1971 Danielson GK, Titus JL, DuShane JW: Successful treatment of aortic valve endocarditis and aortic root abscesses by insertion of prosthetic valve in ascending aorta and placement of bypass grafts to coronary arteries. J THORAC CARDIOVASC SURG 67:443-449, 1974 Frantz PT, Murray GF, Wilcox BR: Surgical management of left ventricular-aortic discontinuity complicating bacterial endocarditis. Ann Thorac Surg 29: 1-7, 1980 Lau JKH, Robles A, Cherian A, Ross DN: Surgical treatment of prosthetic valve endocarditis. Aortic root replacement using a homograft. J THoRAc CARDIOVASC SURG 87:712-716, 1984 Wilson WR, Danielson GK, Giuliani ER, Washington JA II, Jaumin PM, Geraci JE: Valve replacement in patients with active infective endocarditis. Circulation 58:585-588, 1978