Optimal Treatment of Native and Prosthetic Aortic Valve Endocarditis

[1]

Native and prosthetic aortic valve infective endocarditis truly represents one of the most difficult surgical dilemmas that a cardiac surgeon encounters. The mortality and morbidity associated with endocarditis are rarely exceeded by other disease processes. Perhaps many of the problems that we have encountered with native and prosthetic valve endocarditis relate to our mindset in the methods of treatment and the approach we use in the surgical management of this disease. After reviewing the current literature and our endocarditis series, an approach will be presented that we believe is appropriate for the management of this entity.

Literature Review

The literature review consists of eight significant papers over the last 5 years dealing with native and prosthetic valve endocarditis. Several approaches and a general idea of the complexities involved in dealing with aortic valve endocarditis will be gleaned from this review. Beginning in 1992, Haydock and Barratt-Boyes et al.[2] presented a paper on 108 patients with aortic endocarditis. The importance of this paper relates to its strong recommendation for the use of the homograft as the valve of choice. Seventy-eight of these patients received freehand subcoronary homografts and 30 received a bioprosthesis. Patients diagnosed with endocarditis were treated medically for a mean of 4 weeks prior to valve replacement. The overall operative mortality was 18%, with a 17% mortality in the group receiving a homograft. Infection recurred in 13 patients. Twenty-two patients underwent reoperations, 5 (4.6%) of these for recurrent endocarditis and 17 (15.7%) for valve failure. Their actuarial 5-year survival was 64%. Then in 1994, there were four papers reviewing aortic valve endocarditis in both native and prosthetic valve disease.

Petrou presented a 23-year follow-up after using the homograft for prosthetic valve endocarditis and showed excellent results.[3] He reviewed 48 patients, 28 with an infected homograft, 15 with a mechanical valve, and 5 with a xenograft. Eighteen received a root replacement whereas 29 received a freehand subcoronary implantation. His operative mortality was 8.3%, at a mean followup of 4 years. Actuarial freedom from endocarditis at 10 years was 97%, with a 5-year actuarial survival rate of 97%. In all cases, exteriorization of abscess from the circulation was a key point. Additionally, success was attributed to the pliability of the homograft, which allows good coaptation of valve to viable tissue. Length of stay in the intensive care unit was a mean of 4.9 days, and mean overall length of stay was 23 days. 

Similarly, in 1994, Aranki had a 20-year follow-up of 200 patients with both prosthetic and native valve endocarditis.[4] Replacement valves were an even mix of mechanical and bioprosthetic, with only 12 receiving a homograft. The operative mortality for this series was 12.5% overall, with native valve endocarditis accounting for 7.5% and prosthetic valve endocarditis representing 22%. Recurrent endocarditis was evident in 24 of the 200 (12%), with 2 of these patients having a second recurrence. Recurrent endocarditis was early in 11 patients and late in 13. Actuarial survival at 10 years was about 55% for the combined group. Based on their results in operating on patients with healed endocarditis as compared to those operated on for active endocarditis, they advocate medical therapy for endocarditis, with operation being reserved for patients with a “pressing need.” They cite sepsis, large vegetations, arrhythmias, emboli, and uncontrolled congestive heart failure as indications for surgery.

The Toronto group,[5] in 1994, presented a series of 12 patients with endocarditis infecting a previously replaced aortic root. Six of these patients had a mechanical valve conduit; the other 6 had either a bioprosthetic conduit or a homograft as a root replacement. Mean length of time from diagnosis to operation was 4.7 days. Operative mortality was 8.3% with 2 late deaths. A significant point in their review was that all patients with biologic valves failed medical therapy and required surgery. Dr. David concluded that their excellent results were achieved through (1) extensive debridement to normal tissue, thereby totally exteriorizing the infection, (2) reinforcement of tissue with bovine or native pericardium, and (3) an anastomosis with no leaks. They also instituted two to three antibiotics routinely for a period of 6 weeks rather than one to two antibiotic agents.

After our original paper in 1993,[6] in which we described the use of the pulmonary autograft as a replacement valve for infective endocarditis, Joyce et al.,[7] in 1994, reported an additional 6 patients who had native or prosthetic valve endocarditis treated with pulmonary autograft replacement. They had a 0% mortality and 0% recurrence. They favored the Ross procedure due to the young age of their patients and to the difficulty in following patients with anticoagulation in their country. In the younger patients, the Ross procedure offered hope of a replacement valve that was permanent and would not require anticoagulation. Their conclusion stressed the importance of extensive and complete debridement and discussed features of the autograft that were highly suitable in the management of endocarditis.

In 1997, Dearani, at the Mayo Clinic, reported a 10year study involving 36 patients.[8] This paper stressed the benefit of the homograft valve as a replacement for complex endocarditis. Again, both native and prosthetic valve endocarditis was summarized. Replacement techniques included freehand subcoronary in 7 patients, inclusion cylinder in 19, and a root replacement in 10 of the patients. Their overall operative mortality was 13.8%, 10% in native valve endocarditis, and 18.8% in prosthetic valve endocarditis. With a mean follow-up of 2.6 years, there was no recurrence of endocarditis; however, 17% of the survivors had a III/IV aortic regurgitation on echocardiography. Their actuarial 5-year survival was 53%. Final recommendations admonished early operation for prosthetic valve disease and attempts to achieve a medical cure for native valve disease.

Another paper in 1997 by Gaudino from Rome, Italy reported an 8-year follow-up of 20 patients, with native valve endocarditis managed by mechanical valve replacement.[9] Their operative mortality was 0% with 15% late mortality. They had 0 recurrence of endocarditis in the survivors, with 3 (17.6%) of these patients having perivalvular leaks. They attributed their success to early operation, mean 27 days from diagnosis to surgery, and radical d6ridement. Autologous pericardium was used liberally in repairs. In conclusion, they submitted that the management of native valve endocarditis was different from that of prosthetic valve endocarditis, and that a mechanical valve replacement was not likely to have produced these outcomes in patients with prosthetic valve endocarditis.

Finally, Joyce and Oswalt presented a combined series of 70 patients in June of 1997.[10] These patients were all managed with the pulmonary autograft as the replacement valve with an operative mortality of 2.8%, a late mortality of 2.8%, and a reoperation rate of 2.8% at 7-year followup. The paper stressed the benefits of the pulmonary autograft as a root replacement because it allows for complete debridement and exteriorization of the infection. The pulmonary autograft is totally viable and pliable and has probable immunity from infection for two reasons: (1) permeation of the valve by systemic antibiotics administered preoperatively and (2) the natural resistance of the living tissue to reinfection. This valve is likely to be a permanent valve which requires no anticoagulation.

Update on the Use of the Pulmonary Autograft for Aortic Endocarditis in Austin, Texas

Since April of 1990, 35 consecutive patients with aortic endocarditis have been managed by root replacement of the aortic valve using the pulmonary autograft (the Ross procedure). There have been 29 males and 6 females, with a mean age of 38; 89% of these patients were in New York Heart Association’s class III or IV at the time of operation, and 69% had active vegetations with continuing positive blood cultures at the time of surgery. Our clinical summary of patients is shown in the table below. Mean length of time from diagnosis to operation was 6 days.

 

Clinical Summary of Patients
Age	Sex	Organism	Infection	Preoperative Complications
29	M	Enterococcus	Active	Stroke-speech
53	M	Streptococcus viridans	Active
34	F	S. viridans	Inactive
32	M	Unknown	Inactive	SBE 8-90/healed
58	M	S. viridans	Active
23	M	S. viridans	Active
32	M	Staphylococcus aureus Enterococcus	Active
39	M	S. viridans	Active
50	M	Enterococcus	Active
39	IM	S. viridans	Inactive	Stroke-speech
61	M	Streptococcus bovis	Active	>P-R Interval
59	M	S. viridans	Active
31	IM	S. viridans	Active	Previous freehand homograft
53	M	Unknown	Inactive
46	M	S. aureus	Active	Peripheral emboli
37	IM	S. aureus Candida albicans	Active
34	F	Unknown	Inactive
31	F	Grp B Streptococcus agalactlae	Active	Hypotensive, pulmonary edema, sepsis
38	M	S. aureus	Active	Sepsis
32	M	Haemophilus influenzae	Active
33	F	Brucellosis	Active	CHF
28	IM	S. viridans	Inactive	Cerebral emboli with seizures
27	M	S. aureus	Active
38	M	S. viridans	Active
19	M	S. viridans	Inactive
27	M	S. aureus S. viridans	Inactive
39	IM	S. viridans	Active
31	M	Enterococcus	Active
39	M	S. viridans	Active
54	M	Alpha streptococcus	Inactive
50	F	Enterococcus	Active	End-stage renal disease with dialysis
25	M	S. aureus	Active	Septic emboli-cerebral, renal, and peripheral
50	M	S. aureus	Active	End-stage renal disease with dialysis, multivegetations on mitral valve
36	F	S. virldans	Active
38	M	Unknown	Inactive	CHF, massive LVH, 4+ mitral regurgitation, ruptured chordae
SBE = subacute bacterial endocarditis; CHF = congestive heart failure, LVH = left ventricular hypertrophy.

 

Surgical Technique

The Ross procedure has been described in other sources, but it is appropriate to discuss some of the surgical principles that should be employed during the management of aortic valve endocarditis. Complete debridement should be the order of the day, and this includes resection of all infected or marginally infected material to viable tissue. This often involves total excision of the annulus to the left ventricular outflow tract and the unroofing and exposure of all annular abscesses. Complete exteriorization of all infection away from the reconstructed circulation is important as it allows for healing of the ongoing infection. The autograft facilitates reconstruction by its pliability, allowing for viable tissue to be anastomosed accurately to viable tissue. The proximal suture line is reinforced with a strip of autologous pericardium, thereby placing no foreign material in the reconstruction of the root. All fistulas and other areas requiring repair (i.e., mitral valve or mitral valve annulus) are repaired with viable autologous pericardium or with extended right ventricular subannular muscle of the autograft taken as the autograft is harvested. Previously, we reported the harvesting of the pulmonary autograft prior to exposing the aortic root so that reconstruction of the right ventricular outflow tract with the homograft could be completed prior to exposure of the infected area. We now debride the infected area first. After the debridement, we can harvest more right ventricular muscle with the autograft if we have ascertained that additional tissue will be needed in the repair and reconstruction.

Results

Follow-up has been 77% at 7 years (mean 3 years). All 35 patients were cured of their infections. Antibiotics were administered for 3 to 6 weeks, depending on culture results it the time of surgery. If the culture was negative on tissue submitted at surgery, the shorter course was chosen, with subsequent cessation of antibiotics at 3 weeks. There was no operative mortality and there has been zero recurrence of infection on the aortic valve. One patient, at 48 months postoperatively, contracted endocarditis on his pulmonary homograft. This patient is HIV positive and a drug abuser. He was managed medically and cured of his recurrence. Postoperative complications have occurred in 20% (See table below). Postoperative bleeding occurred in 3 patients. One of these patients also required later reoperation for a false aneurysm at the proximal aortic anastomosis. Tissue from this area grew Candida, and the patient was placed on antifungal agents with no recurrence. The aortic valve was not involved. There have been 2 reoperations for valve failures; both of these patients had myxomatous changes in the autogr4ft valve leaflets at the time of mechanical replacement. Prior to their endocarditis, they both must have had myxomatous disease of their aortic valves, and this, likewise, affected the pulmonary valve. There have been 2 late deaths, I from myocardial infarction and I from congestive heart failure. Actuarial 5-year survival is 94%, and freedom from recurrent aortic endocarditis is 100%. Thromboembolic events have not been seen in late follow-up, and no patient- receives anticoagulation with the exception of the 2 patients who eventually received mechanical prostheses. The mean length of stay in the ICU was 5 days, and the hospital length of stay from procedure to discharge was 9 days.

 

Operative Morbidity
Renal	2
G1 reoperation	1
Cardiac arrest	1*
Tamponade	1
Reoperation bleed	3
Infection	1
Valve failure	2
Pulmonary	2
Pacemaker	1
*Resuscitated

 

Summary

Aortic endocarditis is a difficult disease process that is increasing rather than declining. One may conclude from the reviewed data that delaying surgery is best. The author would submit that retrospective studies subselect patients so that those that were managed with antibiotics and cured prior to their surgical repair naturally would do better than those requiring urgent operation. The complications that require our urgent intervention (i.e., emboli, heart block, fistula, and continued sepsis) often arise because of our delay in treating the disease process. Placing a patient on the appropriate antibiotics does not necessarily stop the advancement of the disease process. It is important for us to educate our Cardiology and Infectious Disease colleagues that the management of endocarditis should be medical antibiotic therapy only as long as there is a normally functioning aortic valve or as long as there are no complications pushing one to surgical intervention (i.e., emboli or persistent sepsis). Once the endocarditis has destroyed the valve so that it will require replacement, the therapy should then become surgical and the operation performed for the replacement of the diseased valve. This, in turn, will treat the active infection. It is the delay after the valve needs replacement that often creates a more difficult surgical experience.

Prosthetic valve endocarditis is a different disease process from that of native valve endocarditis primarily due to the organisms. The literature certainly supports that management with antibiotics is rarely going to be successful. Should echocardiographic data show that there is a perivalvular abscess or periprosthetic insufficiency, surgical replacement as soon as possible certainly improves the outcomes. Likewise, the literature is replete with suggestions that complete debridement is essential. The aortic root is a complex structure, and once the annulus is destroyed, root resection is likely to yield the best results. Once a complete and thorough debridement is done, the replacement valve choice probably becomes less critical. Of the choices of valves used in the repair following the debridement, the literature again would suggest that choosing a valve that completely exteriorizes the infection may yield better results. Mechanical and bioprosthetic valves yield more perivalvular leaks. The accurate fixation of a nonpliable valve in a diseased root is more difficult than reconstruction with a pliable valve. The techniques of freehand subcoronary or inclusion cylinder homograft valve replacement create potential spaces between the native diseased aorta and the graft. This does not allow for exteriorization of all abscesses and is a source of recurrent endocarditis. With simple endocarditis involving only the valve leaflets, it is likely that any valve selection will yield good results. However, in complex endocarditis involving the annulus with abscesses, best results appear to favor the root replacement. Root replacement allows for total debridement of all infected material and exteriorizes any remaining infection. One must remember that the aortic leaflets and annulus are a continuum of the entire root structure. Should any part of the annulus become infected or abscessed, complete debridement probably means resection of the root. In choosing a valve, one must consider the longevity and the availability of a valve. As the disease process primarily affects the younger age group, consideration of longevity is important but it is the control of the infection and repair of the insufficiency that should the primary concern over that of longevity of the valve.

The urgency with which one must perform a curative operation for the endocarditis also requires availability of valves. The aortic homograft is becoming more and more difficult to obtain, a fact that is a consideration in choosing this tissue. Mechanical valves obviously are always available, and pulmonary homografts to replace the harvested autograft continue to be readily available. However, in the not so distant future, there could be an insufficient supply of pulmonary homografts.

Finally, there are several key points in the management of aortic infective endocarditis, for either native or prosthetic valve endocarditis. Management of prosthetic valve endocarditis with antibiotics is likely to be unsuccessful, and early operation is recommended. Once the diseased valve becomes incompetent, surgical resection and replacement should be achieved as soon as possible regardless of the infective state. The surgical operation should employ complete debridement of all infected material and the unroofing and exteriorization of all abscessed areas. The replacement valve should be meticulously implanted, ensuring good coaptation of structures, and provide the best longevity against replacement. We strongly believe that the pulmonary autograft as a root replacement is the valve that best suits all of the essential characteristics needed to achieve a cure for aortic endocarditis.
The author appreciates the assistance of Suzi Nelson, R.N., and Sharon Koch, R.N., in the preparation of this manuscript.

 

References

1. Hufnagel CA. Surgical techniques ill the treatment of infected valvular prostheses. Baltimore: University Park Press. p. 143-60.
2. Hayclock D, Barratt-Boyes B, Macedo T, et al. Aortic valve replacement for active infectious endocarditis in 108 patients. J Thorac Cardiovasc Surg 1997; 113:285-91.
3. Petrou M, Wong K, Albertuci M, et al. Evaluation of unstented aortic homografts for the treatment of prosthetic aortic valve endocarditis. Circulation 1994;90(part 2): 198-204.
4. Aranki S, Santini F, Adams D, et al. Aortic valve endocarditis: determinants of early Survival – and late morbidity.Circulation 1994;90(part 2): 175-82.
5. Ralph-Edwards R, David T, Box J. Infective endocarditis in patients who had replacement of the aortic root. Ann Thorac Surg 1994;58:429-33.
6. Oswalt J, Dewan S. Management of aortic infective endocarditis by autograft valve replacement. J Heart Valve Dis 1993;2:380-4.
7. Joyce F, Tingleff J, Aagaard J, Pettersson G. The Ross operation in the treatment of native and prosthetic aortic valve endocarditis. J Heart Valve Dis 1994;3:371-6.
8. Dearani J, Orszulak T, Schaff H, et al. Results of allograft aortic valve replacement for complex endocarditis. J Thorac Cardiovasc Surg 1997; 113:285-91.
9. Gaudino M, DeFilippo C, Pennestri F, Possati G. The use of mechanical prosthesis in native aortic valve endocarditis. J Heart Valve Dis 1997;6:79-83.
10. Joyce F, Oswalt J, Tingleff J, et al. The Ross operation: treatment of choice for aortic endocarditis? Twenty Third Annual Meeting Western Surgical Association, June, 1997.