ATS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Ann Thorac Surg 2008;86:761-768. doi:10.1016/j.athoracsur.2008.01.102
© 2008 The Society of Thoracic Surgeons
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Edward R. Nowicki
Gösta B. Pettersson
Nicholas G. Smedira
Eric E. Roselli
Eugene H. Blackstone
Bruce W. Lytle
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nowicki, E. R.
Right arrow Articles by Lytle, B. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nowicki, E. R.
Right arrow Articles by Lytle, B. W.
Related Collections
Right arrow Valve disease

Original Articles: Adult Cardiac

Aortic Allograft Valve Reoperation: Surgical Challenges and Patient Risks

Edward R. Nowicki, MD, MSa,*, Gösta B. Pettersson, MD, PhDa, Nicholas G. Smedira, MDa, Eric E. Roselli, MDa, Eugene H. Blackstone, MDa,b, Bruce W. Lytle, MDa

a Department of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, Ohio
b Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio

Accepted for publication January 28, 2008.

* Address correspondence to Dr Nowicki, Department of Thoracic and Cardiovascular Surgery, Cleveland Clinic, 9500 Euclid Ave, Mail Stop JJ-40, Cleveland, OH 44195 (Email: nowicke{at}ccf.org).

Presented at the Forty-fourth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28–30, 2008.


   Abstract
Top
 Abstract
Introduction
 Patients and Methods
 Results
 Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
 References
 Appendix Reference
 
Background: Aortic valve allograft reoperation is a challenge for the surgeon and a risk to the patient. We examined our experience to identify these challenges and risks.

Methods: From April 1987 to January 2006, 130 patients underwent first-time allograft-related reoperations. Prior implant was subcoronary 32 (25%), inclusion-root 28 (22%), and root 70 (53%). Reoperative indications were technical failure 11 (8.4%), endocarditis 31 (24%), and structural valve deterioration 88 (68%). Reoperative technique was allograft repair 7 (5.4%), simple valve replacement 80 (62%), and root replacement 43 (33%). Median follow-up was 3.1 years.

Results: Surgical challenges: 10 adverse intraoperative events occurred (7.7%), 3 allograft specific, with rescue from all. Reoperative procedure was highly dependent on original implantation technique (31 of 43 root replacements after previous roots) and reoperative indication (24 of 43 root replacements for endocarditis). Implanted valve prostheses were small for patient size, less so in intact native roots (previous subcoronary or inclusion root) than retained allografts (Z-value, –1.1 versus –1.6; p = 0.08), but allograft root re-replacement allowed normal-sized valves. Patient risks: Relevant postoperative morbidities included reoperation for bleeding 7 (5.4%), new pacemaker 6 (5.2%), stroke 1 (0.8%), and no myocardial infarction. Hospital mortality was 3.8% (5 of 130), 6.5% (2 of 31) for endocarditis. Late patient survival was substantially worse for endocarditis than for structural valve deterioration (60% versus 90% at 5 years, p = 0.0006). Five-year freedom from further reoperation was 94%.

Conclusions: Surgical challenges, even with endocarditis, can be surmounted and patient risks minimized by thoughtful preparation and appropriate reoperative procedure. An intact native root maximizes surgical options.


   Introduction
Top
 Abstract
 Introduction
Patients and Methods
 Results
 Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
 References
 Appendix Reference
 
Aortic allografts are used primarily for endocarditis with advanced pathology, most often as a full root replacement [1, 2]. Whatever the original indication, reoperation will eventually be needed. These reoperations, however, are perceived to be both challenging for the surgeon and high risk to the patient. Technical challenges involve avoiding adverse events common to any resternotomy, as well as those specific to allograft reoperation such as debriding degenerated, calcified tissue that may embolize and mobilizing and reimplanting coronary buttons that may compromise coronary circulation. Decision-making challenges include choice of replacement device and implant technique. The few reports available have not elaborated on these challenges, but nonetheless present favorable results [3–6]. Therefore, we reviewed our experience with aortic allograft valve reoperation to identify these technical and decision-making challenges and assess patient risks and long-term results.


   Patients and Methods
Top
 Abstract
 Introduction
 Patients and Methods
Results
 Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
 References
 Appendix Reference
 
Patients
From April 1987 to January 2006, 171 reoperations were performed at Cleveland Clinic on patients with a previously implanted aortic allograft. Of these, 130 undergoing first-time reoperation for allograft-related problems formed the study cohort (Table 1; Appendix Fig 1 *). The original allograft had been implanted at other institutions in 42 (32%), and 125 had been supplied by CryoLife, Inc.


View this table:
[in this window]
[in a new window]

  		Table 1 Patient Characteristics and Reoperative Details (n = 130)
 

Figure 1
View larger version (8K):
[in this window]
[in a new window]

  		Appendix Fig 1. Occurrence of first-time aortic allograft reoperations across time. Dots summarize raw data, and solid line within 68% confidence limits is a Loess fit to the raw data.

 
Medical records were reviewed to confirm and augment data collected routinely and concurrently with patient care. Use of these data for research was approved by the Institutional Review Board, with patient consent waived.

Original Implant Technique
The original allograft had been implanted using (1) freehand subcoronary (SC), (2) inclusion root (INC), or (3) freestanding root replacement (RR) techniques (Table 1) [7]. An intact native root was considered to exist when there was minimal violation of the native aortic root (original SC or INC). Of patients having INC at original implant, 6 (21%) had at least one coronary button mobilized and implanted into the allograft sinus, rather than a wall-to-wall (allograft to native aorta) anastomosis.

Allograft-Related Indications for Reoperation
We recognized three indications for allograft reoperation as described by O'Brien and coworkers [6] (Table 1): technical failure (TF [Appendix Table *]), infectious endocarditis (IE), and structural valve deterioration (SVD [Fig 1]). Patients undergoing reoperation for TF had a shorter median time from original allograft implantation (3.4 months) than those with IE (9.4 months) or SVD (8.4 years); p < 0.001 (Fig 2).


View this table:
[in this window]
[in a new window]

  		Appendix Table 1 Individual Technical Failures and Their Management (n = 11)
 

Figure 1
View larger version (139K):
[in this window]
[in a new window]

  		Fig 1. Indications for aortic allograft valve reoperation. (A) Technical failure: pseudoaneurysm (PS) at proximal suture line circled. (B) Endocarditis: large vegetation (arrows) almost obstructing valve orifice. (C and D) Structural valve deterioration: arrows point to a perforated right coronary cusp (C) and heavy calcification of cusps (D). (L = left cusp; LCA = left main coronary artery; MV = anterior leaflet of mitral valve; NC = noncoronary cusp; R = right cusp; RCA = right coronary artery.)

 

Figure 2
View larger version (17K):
[in this window]
[in a new window]

  		Fig 2. Cumulative distribution of intervals from original implant to reoperation, by indication.

 
Ninety-four percent of patients (29 of 31) presenting with allograft IE had a history of endocarditis, but this was the case in only 55% of TF (6 of 11) and 26% of SVD cases (23 of 88). An infectious agent was identified in 84% of IE cases (26 of 31): bacterial in 16 (52%) and fungal in 10 (32%). Of the 10 with fungal IE, 4 had not had previous IE; 3 had previous bacterial endocarditis, and 3 had previous fungal endocarditis. The hemodynamic abnormality of SVD was aortic regurgitation or mixed regurgitation and stenosis in 91% (80 of 88) and pure aortic stenosis in 9% (8 of 88).

Reoperative Procedure
Three main reoperative techniques were employed: repair of original allograft, simple valve replacement within intact native root or retained allograft, and RR (Table 1; Appendix *). Repair procedures involved closure of three intracardiac fistulae, suture of two anastomotic leaks, insertion of an aortic allograft conduit at a failed distal allograft–aortic suture line, and right coronary artery bypass grafting for a stenotic button anastomosis (Appendix Table *).

Prosthesis size
Size of device implanted at reoperation was expressed as internal orifice diameter standardized to patient size (prosthesis–patient size), rather than label size, so that meaningful comparisons could be made across device categories [8]. This was expressed as Z-value, the number of standard deviations (SD) by which the prosthesis internal orifice diameter deviated from mean normal native aortic "anulus" diameter for the patient's body surface area [9].

Endpoints
Study endpoints were (1) adverse intraoperative events, (2) hospital mortality and morbidity, (3) time-related all-cause mortality, and (4) reoperation for aortic valve-related events. Adverse intraoperative events were defined as destabilizing injuries to mediastinal structures on reentry or during dissection, requiring immediate attention and repair, or intraoperative revision of the intended procedure to complete the operation. Hospital morbidities were defined according to the STS National Database (available at: http://www.ctsnet.org/file/rptDataSpecifications252_1_ForVendorsPGS.pdf). Vital status and aortic valve-related reoperations were identified by follow-up using an Institutional Review Board–approved questionnaire or approved telephone script with patient consent. A total of 515 patient-years of data were available for analysis, with mean follow-up among survivors of 4.3 ± 3.8 years (median, 3.1); 25% of survivors were followed more than 6 years and 10% more than 10 years.

Data Analysis
Descriptive statistics were summarized as frequencies and percentages for categorical variables and means and standard deviations for continuous variables. The Kaplan-Meier method was used for nonparametric estimates of time-related events [10]. A nonproportional hazards model was developed to identify the number of phases of instantaneous risk and to estimate their shaping parameters [11]; for additional details, see http://www.clevelandclinic.org/heartcenter/hazard.


   Results
Top
 Abstract
 Introduction
 Patients and Methods
 Results
Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
 References
 Appendix Reference
 
Surgical Challenges
Adverse intraoperative events
There were 10 adverse intraoperative events (n = 130, 7.7%; Appendix Fig 2 *). Frequency of their occurrence decreased somewhat across the experience (but p = 0.15). Of the 10, 7 were generic to reoperation (6 related to dissection injury to great vessels and 1 to replacement of an incorrectly sized replacement device), and 3 (30%) were specific to allograft reoperation. These latter included entering a known false aneurysm arising from left coronary artery button sites in 2 and repair of a disrupted coronary artery button anastomosis after reimplantation. All adverse intraoperative events were successfully managed.


Figure 2
View larger version (8K):
[in this window]
[in a new window]

  		Appendix Fig 2. Occurrence of adverse intraoperative events across the experience. Dots summarize raw data, and solid line within 68% confidence limits is a Loess fit to the raw data.

 
Reoperative procedure
Original operative implant technique, indication for reoperation, and reoperative techniques were associated. The most common indication for simple valve replacement at reoperation was SVD (73 of 80, 91%). In contrast, the most common indication for RR at reoperation was IE (24 of 43, 56%); most (21 of 24, 88%) received an allograft root, and 3 a composite graft. Patients with an intact native aortic root were more likely to have simple valve replacement than RR (Fig 3A). In contrast, if original RR had been performed, 44% had a repeat RR. All allograft repairs were performed after original RR. Cases with TF or IE had a higher likelihood of reoperative RR than those with SVD (Fig 3B), but these were also more likely to have had RR at original implantation (Fig 3C). Average overall operating time, ischemic time, and total cardiopulmonary bypass time were longer for IE cases than for TF or SVD (Fig 4).


Figure 3
View larger version (22K):
[in this window]
[in a new window]

  		Fig 3. Associations among original allograft implant technique, indication for reoperation, and reoperative procedure. (A) Original allograft implant technique and reoperative procedure. (B) Indication for reoperation and reoperative procedure. (C) Indication for reoperation and original allograft implant technique. (SVD = structural valve deterioration.)

 

Figure 4
View larger version (25K):
[in this window]
[in a new window]

  		Fig 4. Association of operating times and indication. (CPB = cardiopulmonary bypass time; Overall = time from incision to closure; SVD = structural valve deterioration.)

 
Allograft-specific technical issues related to proximal coronary arteries or buttons occurred in 3 of 31 patients (9.7%) with original RR who had reoperative RR. In all 3, reimplantation of one or both coronary buttons was impossible: the first had an extensive periaortic allograft root abscess involving native coronary artery buttons, and the surgeon chose to ligate and bypass both left and right coronary arteries; the second, who had suture closure of the right coronary ostium and grafting of the right coronary artery at original allograft implant, required repeat bypass grafting to the right coronary artery; the third had ligation and grafting of both left and right coronary arteries at original implant, but did not require additional revascularization. Limited debridement of the buttons was performed in at least 7 patients without adverse effect at button reimplantation. None of the 60 patients who had original SC or INC implant had a coronary ostial problem.

Prosthesis type and size
Simple valve replacement was with a bioprosthesis or mechanical valve in 76 (95%), or second allograft valve in 4 (5%). The most common prosthesis implanted after RR was an allograft or autograft in 33 of 43 (77%), with a composite graft in 10 of 43 (23%; Table 1).

Use of human valves (allograft or autograft) for reoperation provided larger prosthesis–patient size compared with nonhuman replacement devices (Z-value 0.14 versus –1.3, p < 0.0001; Fig 5A). For SVD, simple valve replacement into an intact native root resulted in slightly larger prosthesis–patient size at the smaller end of the spectrum than replacement into a retained allograft root (Z-value –1.1 versus –1.6, p = 0.08; Fig 5B). Root replacement with a composite graft, however, did not provide larger sizes than simple valve replacement (Z-value –1.5 versus –1.3, p = 0.24).


Figure 5
View larger version (10K):
[in this window]
[in a new window]

  		Fig 5. Cumulative distribution of valve sizes, expressed as Z-value, at reoperation. (A) Human versus nonhuman valves. (B) Nonhuman prostheses, in cases of structural valve deterioration of original allograft, implanted into retained allograft root versus into native aortic root. (INC = inclusion root; RR = root replacement; SC = subcoronary.)

 
Patient Risks
Early mortality and morbidity
Hospital mortality was 3.8%, but was higher in TF and IE patients than in those with SVD (p = 0.01; Table 2). One patient with SVD whose allograft was debrided and retained suffered a stroke. Of 7 patients undergoing reoperation for bleeding, only 1 was related specifically to the site of valve replacement: bleeding at a reimplanted coronary button suture line. Of 6 requiring a new permanent pacemaker, indication for reoperation was IE in 3. The IE patients had more complications (10 [32%] versus 3 [27%] for TF and 8 [9.1%] for SVD) and longest hospital stay (p < 0.0001; Table 2). When the infectious agent was fungal, median hospital stay was double that of patients with bacterial infection (32 versus 16 days).


View this table:
[in this window]
[in a new window]

  		Table 2 Early Mortality and Morbidity
 
Late mortality
Survival was 96% at 30 days, and 91%, 82%, and 74% at 1, 5, and 10 years, respectively. It was substantially worse after reoperations for TF and IE than for SVD (p = 0.0006; Fig 6); survival after reoperation for SVD was only slightly less than that expected for an age-, sex-, and race-matched general population.


Figure 6
View larger version (14K):
[in this window]
[in a new window]

  		Fig 6. Survival according to indication for reoperation. Symbols represent deaths and vertical lines represent 68% confidence limits (CL) equivalent to ± 1 SE. Solid lines are parametric estimates and dash-dot line survival of an age-sex-race–matched United States population. (IE = infectious endocarditis; SVD = structural valve deterioration; TF = technical failure).

 
Subsequent aortic valve reoperations
Six patients had a subsequent aortic valve reoperation, with freedom from subsequent reoperation of 98%, 94%, and 89% at 1, 5, and 10 years, respectively. Four of these were for IE, 3 with endocarditis at first-time allograft reoperation, and all with a history of endocarditis before that. Two were for SVD. There were no reoperations for TF.


   Comment
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
Discussion
 Comment
Footnotes
 Acknowledgments
 References
 Appendix Reference
 
Our study demonstrates that first-time reoperation for allograft-related problems, despite being surgically challenging, can be performed with low risk to the patient. This finding supports the continued use of allografts for aortic valve replacement, particularly for endocarditis, and informs patients of future risks of eventual reoperation for allograft-related problems. However, allograft endocarditis complicates and lengthens reoperation and increases early and late mortality and morbidity. When a prior full root replacement has been done, options for implant type, size, and technique are more limited than when the native root is intact (subcoronary or inclusion root).

Surgical Challenges
Adverse intraoperative events
Common to any reoperation is potential heart and great vessel injury requiring immediate institution of cardiopulmonary bypass. In our study, most of these occurred during reentry or dissection before cardiopulmonary bypass [12–15]. Allograft-specific adverse events were uncommon. Awareness of potential allograft-specific problems, such as false aneurysms directly under the sternum from previous suture lines, with early preparation for immediate institution of cardiopulmonary bypass, aids successful management.

Reoperative procedure
Among indications for reoperation, as mentioned by O'Brien and colleagues [16], technical failure occurred earliest after original implantation. The common mode of failure was suture line dehiscence or improper suture placement leading to an intracardiac fistula or false aneurysm. For some, endocarditis affecting native tissues at original implant or dissolution of subanular allograft muscle may have played a role in causing tissue fragility at the proximal suture line. Where possible, with well-functioning allograft valves, we repaired defects due to technical failure of the primary implant without replacing the allograft.

Reoperation for allograft endocarditis was the most technically challenging indication. In a large proportion of these cases, the patients had a history of endocarditis. Many may represent unresolved infection in patients who may have a predilection for infection. However, it must be emphasized that because nearly a third of patients coming to reoperation had their original allograft operation elsewhere, we do not know the denominator, and thus the incidence, of endocarditis for our study. We also do not know the extent of debridement of infected tissue at original implant. Further, the possibility of contamination of the originally implanted allograft cannot be excluded.

An unexpected number of infections in this all-allograft series were fungal (32%), nearly half in patients without IE at original allograft valve replacement. This compares with 4.1% fungal infection reported for endocarditis of all types of prostheses [17]. Calderwood and colleagues [18] reported fungal infection in 4.3% of 116 prosthetic valve infections, none of which occurred in a previously implanted allograft. In 103 patients with prosthetic valve endocarditis (only two allografts), Sabik and associates [1] found fungal infection in 6.5%. Although reports of reoperation for allograft-related problems have included endocarditis as an indication, none has presented information about causative organism. These fungal infections confer further surgical challenge because of extent of tissue destruction and risk to the patient because of difficulty eradicating the infection.

At reoperation, aggressive, thorough debridement of infected tissue is mandatory [1]. After removing all infected tissue, there is little decision-making challenge for valve implant type or technique, because nearly always a root replacement will be required to reconstruct cardiac and aortic structures, particularly if the original implant was a root replacement. Our preference for root and left ventricular outflow tract reconstruction in destructive endocarditis remains an allograft.

Structural valve deterioration cases were least technically challenging, although a major concern is avoiding embolizing material during debridement of the native aorta and anulus or thromboendarterectomy of the calcified allograft. In the intact native root, removing allograft tissue and debriding the aortic anulus without violating it or surrounding structures was usually less challenging than preparing a calcified retained allograft root for simple valve replacement. Low occurrence of stroke in our patients indicates that both can be accomplished with little morbidity.

However, these cases pose more of a decision-making challenge regarding type of prosthetic replacement and implantation technique. Although others have reimplanted a second allograft in a subcoronary position [3], we rarely did; instead, we usually performed a simple valve replacement with a bioprosthesis or mechanical valve. Simple valve replacement within the retained allograft root, although with a smaller valve, was often possible, but future integrity of conduit tissue is unknown.

In patients with severe allograft root degeneration, repeat root replacement was undertaken. The technical challenge is to safely dissect the coronary buttons free, carefully detach them from the allograft, and reimplant them into either another allograft or prosthetic valved conduit (Bentall). Coronary buttons can be mobilized and directly reimplanted in most cases. Only 2 of 31 repeat RR patients (6.4%) in our series required coronary artery bypass grafting. Severe infection around the original allograft involving the main coronary arteries necessitated ligation and bypass grafting in both. This is in contrast to the report by Raanani and colleagues [19] that 16 of 31 patients (52%) having repeat root replacement required an extension of one or more coronary arteries by saphenous vein or synthetic graft to avoid undue tension at the anastomosis, but only 4 (13%) had an allograft reimplanted.

Our predominant use of flexible allografts for repeat RR may explain the high proportion of successful direct coronary reimplantations. We have seen, as have others, that native aortic tissue in coronary buttons is little affected by the degenerative process in the surrounding allograft [5]. In the absence of endocarditis, accomplishing repeat root replacement had low morbidity; in only 1 case, reoperation was necessary for bleeding at a reimplanted coronary button site.

Prosthesis type and size
Our findings indicate that a quarter of patients in whom a nonhuman valve was placed into the intact native aorta had a prosthesis–patient size more than 2 SD below normal, and half were more than 1 SD below. This size discrepancy was somewhat greater when nonhuman prostheses were implanted in retained allograft roots or when composite grafts were used. Careful, complete debridement of restricting allograft anular scar tissue must be accomplished to implant a maximum-size valve.

Allograft valves have the largest effective orifice area and are preferred when it is desirable to minimize postoperative gradient. This decision-making challenge pits a less durable allograft, with better performance and no need for anticoagulation, against a more durable prosthesis that may require anticoagulation. Accelerated degeneration due to immunologic causes after repeat allograft implantation is claimed not to occur and should not be a consideration in the choice of another allograft at reoperation [3]. The alternative of using a stented or unstented nonhuman bioprosthesis in noninfected patients offers durability comparable to another allograft [20]. The decision can be made only by the surgeon and patient together before reoperation, weighing all considerations for infection resistance, need to avoid anticoagulation therapy, and maximum hemodynamic performance.

Implications for original allograft implant
Results of our study may have implications for original allograft implant technique. There has been a trend away from subcoronary implant methods to full root replacement over the last 10 years, with various opinions about preferred technique [6, 21–23]. The most common reasons relate to standardizing techniques to avoid early development of aortic regurgitation and possible accelerated allograft structural deterioration. However, long-term consequences of native root violation (full root replacement) at original operation are not entirely clear, but at reoperation, may include unfavorable prosthesis–patient size with simple nonhuman valve replacement in the retained allograft or higher risk with more complex repeat root replacement.

For patients with favorable anatomy, including noninvasive endocarditis, subcoronary or root inclusion technique at original allograft implantation may provide long-term benefit. Having an intact native aortic root increases options at reoperation and may enable a less complex reoperation, as well as implantation of a larger valve.

Patient Risks
Early mortality and morbidity
Although reoperation for allograft-related problems is surgically challenging, we found in our series, heavily weighted by endocarditis, an overall 30-day mortality of 4.0%. This compares favorably with that of others. Hasnat and colleagues [3] reported a 3.4% early mortality in Yacoub's series of 144 patients having a second allograft implant, mainly for tissue failure. In another series of 46 patients having reoperation to explant a failed allograft, Kumar and colleagues [5] reported a 30-day mortality of 8%.

Despite the complexity of reoperation, morbidity in our series was also low compared with that reported by others. For the Yacoub series, there were 33 complications in 144 patients (22%); for ours, 21 in 130 patients (16%). Others have not tabulated allograft-specific morbidity, whereas we documented 1 stroke, 1 reoperation for bleeding, and requirement for 6 new pacemakers in our series.

Late mortality and morbidity
Ten-year survival (74%) in our patients was slightly lower than in the Yacoub series (82%) [3], but this may have been due to more patients with IE in our series (24% versus 9.4%). For our patients with SVD as indication for reoperation, 10-year survival was more than 85% and nearly the same as that expected for a matched general population.

Indication for subsequent reoperations in our mixed-prosthesis group of patients was predominantly IE, as was also true of the early (first 5 years) results of Yacoub's pure second allograft series [3].

Limitations
This study represents a single experienced center's work. However, different approaches were taken by a number of surgeons, and techniques evolved over time. Also, because of a modest number of patients (130) and few events, we were not able to do analysis to identify risk factors, but were able to study each case in detail. Follow-up was limited to 10 years, so the fate of patients thereafter is unknown. We acknowledge a bias toward full root replacement, rather than a less destructive procedure at original allograft implant. Our inferences about choice of techniques at original allograft implantation may therefore be unjustified.

Conclusion
In conclusion, meeting the surgical challenges of safe reentry, dealing effectively with the pathologic anatomy, and choosing appropriate reoperative procedure and replacement prosthesis require thoughtful preparation and experienced hands to minimize early patient risk and maximize long-term results of allograft reoperation.


   Discussion
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
 Discussion
Comment
 Footnotes
 Acknowledgments
 References
 Appendix Reference
 
DR JOHN S. IKONOMIDIS (Charleston, SC): Having looked at your manuscript, you mentioned toward the end that a number of technical changes have evolved over time that have minimized morbidity with this procedure and intraoperative catastrophes. Could you possibly share with us any of those highlights or changes that you have made to make these operations safer?

DR NOWICKI: For the most part, the reentry problem is generic to any reoperation. Our strategy includes preoperative imaging analysis to find out where mediastinal and cardiac structures lie with respect to the sternum upon reentry. Extrathoracic cannulation, being prepared with the artery exposed and even with a sewn-on graft on the axillary artery in selected cases, is very important. We knew about two of the false aneurysms ahead of time, and they were easily managed when we immediately went on bypass. Dissection is limited to what you need to do. Of course, you are not going to mobilize the whole heart unless you are going to do coronary bypass grafting.

Although it may sound a little easier to debride an allograft root and then do a simple valve replacement, experience has taught us that resecting the whole allograft is possible and provides a fresh start. The caution is that you must have adequate-sized buttons for reimplantation after mobilization. This concern has implications for the original operation, because if you develop small buttons at the first root implantation, there will be little left to deal with on the second or third operation. Fortunately, the degenerative process is restricted to allograft wall and does not seem to involve autologous aorta buttons.

Nevertheless, creating generous buttons also at reoperation, which may include some allograft tissue, aids in handling them during mobilization. It is important to be able to move them around, keeping in mind the axis of the lumen so that you don't damage the arteries at this stage. At reimplantation, one can perform the final tailoring of buttons. We attribute careful mobilization of coronary buttons and extensive use of the flexible allograft as a root to lowering the risk of coronary complications at reoperation.

DR THORALF SUNDT (Rochester, MN): Professor Yacoub gave me the opportunity to review his experience some 15 years or so ago and he had the same comments about the calcification. The term that he used was he said that the calcification is noninvasive, and so if you just chip away the calcium it usually leaves some reasonable structures behind. But by the same token, I have been concerned about putting a new valve inside of an allograft root for fear that there will not be healing between the allograft root and the sewing ring of the prosthetic valve. Now, I saw that your reoperation rate was low, but do you have a sense of what is the incidence of paravalvular leak if you put a new valve inside of an allograft root? Thanks very much. I enjoyed the paper.

DR NOWICKI: Thank you for your comments. Yes, your 1995 paper with Dr Yacoub was quite informative as we started our review. There are several questions about the retained allograft. Among these is whether there will be aneurysmal change. We simply don't know about that, because our follow-up is only to 10 years. Another is your question about poor healing of valves implanted within the retained allograft. Among the 6 cases reoperated on after first-time reoperation, 4 were for endocarditis. All had a history of endocarditis dating back to the original implant, and 3 of the 4 had endocarditis as the indication for reoperation. We don't know what role possible paraprosthetic leak may have played in these patients. In noninfected cases, we have not seen paraprosthetic leak as an indication for valve-related reoperation.

DR SUNDT: You don't know about paravalvular leaks that have not been operated?

DR NOWICKI: No, I have no information about that, I am sorry to say.

DR JOHN D. PUSKAS (Atlanta, GA): Doctor Nowicki, tell us how this experience with reoperating on allografts has impacted the Cleveland Clinic's attitude toward using allografts in the first place? For whom do you use them and for whom do you not?

DR NOWICKI: This is the third paper from the Cleveland Clinic relating to reoperation, endocarditis, and use of aortic allografts. In 2002, Dr Lytle and colleagues demonstrated low hospital mortality and a high degree of freedom from recurrent infection in a complex group of 27 endocarditis patients with previous aortic root replacement (Bentall) or aortic valve replacement with supracoronary ascending aortic replacement. These results were credited to radical debridement and use of allografts as a root. In the second paper around the same time, Dr Sabik reported 103 cases of prosthetic valve endocarditis, all either mechanical or bioprosthesis, and again, receiving an aortic allograft at reoperation was associated with low operative mortality and low recurrence of infection.

My presentation deals with allograft reoperation, and these procedures were for advanced endocarditis, nearly all requiring allograft root replacement. And so, I don't think anything has changed regarding our use of allografts for infectious endocarditis. Certainly, our low operative mortality, low risk of stroke, and low coronary button complications associated with reoperation support our opinion that the allograft is a marvelous device, superior to or comparable with other bioprosthetic devices.


   Appendix
 
Details of Reoperative Procedure
Generic
Our general recommendations for reoperation, including preoperative evaluation, surgical plan, sternal reentry, dissection, operative considerations, and rescue strategies, have been described in detail by Roselli and colleagues [1]. However, several more specific conditions need emphasis. Heparinization and extrathoracic cannulation before sternal reentry is used, if a pseudoaneurysm lies immediately beneath the sternum.

If ascending aortic and arch surgery is required, right axillary artery and bicaval (for circulatory arrest and retrograde cerebral perfusion) are preferred. Cooling is continued for 20 minutes, 1 or more minutes per minute of expected circulatory arrest, until the electroencephalogram becomes isoelectric (most proximal arch procedures require less than 15 minutes of circulatory arrest). When aortic clamping is possible, we often still use bicaval cannulation and open placement of the retrograde cardioplegia cannula to optimize myocardial (right ventricular) protection (some use ice slush, in addition).

Another issue is aortic regurgitation, which is usually severe. Left ventricular (LV) venting through the left atrium or pulmonary artery (and at times, direct LV compression) is important to prevent LV distension and pulmonary congestion when the heart fibrillates. Finally, electrocautery is used primarily for hemostasis and not for dissection. This is particularly important when mobilizing the coronary buttons so as to avoid thermal injury to the coronary arteries, which may be unrecognized.

Allograft Specific
Indication for reoperation, technique of original allograft implantation, and condition of the allograft are important in the approach to and conduct of the reoperation. The aorta is transected at the previous anastomosis between the allograft and ascending aorta. Original implantation technique and pathology are confirmed and the options for reoperation decided. If the allograft is infected, radical removal of all allograft and devitalized tissue is essential; if the infection is fungal, this radical approach is mandatory. In our experience, an allograft root replacement is nearly always necessary under these conditions.

For a degenerated noninfected allograft, which is invariably calcified and dysfunctional, peeling the old allograft out of an intact native root is an attractive option, followed by standard aortic valve replacement.

However, after original root replacement, the dilemma is whether to retain the allograft conduit, debriding the intima and simply replacing the valve, or to remove entirely the allograft root, performing a repeat root replacement. The first option seems less complex, but requires valve implantation in a narrow diseased conduit, placing sutures through partly calcified allograft tissue.

The second option is more complex, but is actually easier after some practice. The first step is to redevelop the coronary buttons by cutting in the allograft wall outside of the previous anastomotic suture line, maintaining the U-shape for handling. Next, scissors or a knife is used to dissect allograft tissue free, staying in the plane between allograft and native normal tissue. Thick inflammatory scar tissue is resected as necessary to improve access to the root and LV outflow tract. Old suture material and felt are good landmarks. As dissection proceeds, underlying original anatomy and native LV outflow tract disclose themselves. Avoiding violation of the aortic anulus, but cleanly removing all scar and foreign material, is critical to reimplanting a maximum-size valve.

After freeing the allograft root, the main coronary arteries are mobilized by carefully dissecting through the thick scar tissue, keeping track of location and direction of the coronary artery lumen. Adequate coronary button mobility is necessary for safe effective reimplantation and unimpaired coronary blood flow. Finally, the residual allograft tissue and any intimal hyperplasia involving the coronary ostia are removed from the buttons before reimplantation.

Choice of valved conduit is a matter of personal preference, although we have found that the greater flexibility of an allograft root compared with that of a composite graft (Bentall) makes reestablishing normal coronary continuity easier and more predictable.


   Comment
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
Discussion
 Comment
Footnotes
 Acknowledgments
 References
 Appendix Reference
 
In recent years, as we have gained more experience with reoperation for allograft root structural valve deterioration, we have tended to remove the entire allograft and start anew with a repeat allograft root replacement (sometimes a composite valve graft), rather than to debride retained allograft tissue in order to perform simple valve replacement.

The wisdom of decisions at first-time allograft reoperation in some younger patients will, no doubt, be tested at second-time reoperation.


   Acknowledgments
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
References
 Appendix Reference
 
The authors thank Angela York for database management, Tanya Ashinhurst and Trinity Bell for medical record review, Capri Spencer for follow-up, Jingyuan Feng for survival analysis, and Tess Parry for editorial assistance. Gösta B. Pettersson is supported in part by the Peter and Elizabeth C. Tower and Family Endowed Chair in Cardiothoracic Research, James and Sharon Kennedy, the Slosburg Family Charitable Trust, and Stephen and Saundra Spencer. Eugene H. Blackstone is supported in part by the Kenneth Gee and Paula Shaw, PhD, Chair in Heart Research.


   Footnotes
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
 Discussion
 Comment
 Footnotes
Acknowledgments
 References
 Appendix Reference
 
*The Appendix is available online only. To access it, please visit: http://ats.ctsnetjournals.org and search for the article by Nowicki, Vol. 86, pages 761–768.e1–2.

* See note at end of article. Back


   References
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
 References
Appendix Reference
 

  1. Sabik JF, Lytle BW, Blackstone EH, Marullo AG, Pettersson GB, Cosgrove DM. Aortic root replacement with cryopreserved allograft for prosthetic valve endocarditis Ann Thorac Surg 2002;74:650-659.[Abstract/Free Full Text]
  2. Yankah AC, Pasic M, Klose H, Siniawski H, Weng Y, Hetzer R. Homograft reconstruction of the aortic root for endocarditis with periannular abscess: a 17-year study Eur J Cardiothorac Surg 2005;28:69-75.[Abstract/Free Full Text]
  3. Hasnat K, Birks EJ, Liddicoat J, et al. Patient outcome and valve performance following a second aortic valve homograft replacement Circulation 1999;100:II42-II47.[Medline]
  4. Byrne JG, Karavas AN, Aklog L, et al. Aortic valve reoperation after homograft or autograft replacement J Heart Valve Dis 2001;10:451-457.[Medline]
  5. Kumar P, Athanasiou T, Ali A, et al. Re-do aortic valve replacement: does a previous homograft influence the operative outcome? J Heart Valve Dis 2004;13:904-913.[Medline]
  6. O'Brien MF, Harrocks S, Stafford EG, et al. The homograft aortic valve: a 29-year, 99.3% follow up of 1,022 valve replacements J Heart Valve Dis 2001;10:334-345.[Medline]
  7. Kouchoukos NT, Blackstone EH, Doty DB, Hanley FL, Karp RB. Cardiac surgeryPhiladelphia: Churchill Livingstone; 2003. pp. 571-574.
  8. Blackstone EH, Cosgrove DM, Jamieson WR, et al. Prosthesis size and long-term survival after aortic valve replacement J Thorac Cardiovasc Surg 2003;126:783-796.[Abstract/Free Full Text]
  9. Capps SB, Elkins RC, Fronk DM. Body surface area as a predictor of aortic and pulmonary valve diameter J Thorac Cardiovasc Surg 2000;119:975-982.[Abstract/Free Full Text]
  10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations J Am Stat Assoc 1958;53:457-481.
  11. Blackstone EH, Naftel DC, Turner Jr ME. The decomposition of time-varying hazard into phases, each incorporating a separate stream of concomitant information J Am Stat Assoc 1986;81:615-624.
  12. Follis FM, Pett Jr SB, Miller KB, Wong RS, Temes RT, Wernly JA. Catastrophic hemorrhage on sternal reentry: still a dreaded complication? Ann Thorac Surg 1999;68:2215-2219.[Abstract/Free Full Text]
  13. Dobell AR, Jain AK. Catastrophic hemorrhage during redo sternotomy Ann Thorac Surg 1984;37:273-278.[Abstract]
  14. Lytle BW, Cosgrove DM, Taylor PC, et al. Reoperations for valve surgery: perioperative mortality and determinants of risk for 1,000 patients, 1958-1984 Ann Thorac Surg 1986;42:632-643.[Abstract]
  15. Roselli EE, Pettersson GB, Blackstone EH, et al. Adverse events during reoperative cardiac surgery: frequency, characterization, and rescue J Thorac Cardiovasc Surg 2008;135:316-323.[Abstract/Free Full Text]
  16. O'Brien MF, Harrocks S, Clarke A, Garlick B, Barnett AG. Experiences with redo aortic valve surgery J Card Surg 2002;17:35-39.[Medline]
  17. Wang A, Athan E, Pappas PA, et al. Contemporary clinical profile and outcome of prosthetic valve endocarditis JAMA 2007;297:1354-1361.[Abstract/Free Full Text]
  18. Calderwood SB, Swinski LA, Waternaux CM, Karchmer AW, Buckley MJ. Risk factors for the development of prosthetic valve endocarditis Circulation 1985;72:31-37.[Abstract/Free Full Text]
  19. Raanani E, David TE, Dellgren G, Armstrong S, Ivanov J, Feindel CM. Redo aortic root replacement: experience with 31 patients Ann Thorac Surg 2001;71:1460-1463.[Abstract/Free Full Text]
  20. Smedira NG, Blackstone EH, Roselli EE, Laffey CC, Cosgrove DM. Are allografts the biologic valve of choice for aortic valve replacement in nonelderly patients?. Comparison of explantation for structural valve deterioration of allograft and pericardial prostheses. J Thorac Cardiovasc Surg 2006;131:558-564.[Abstract/Free Full Text]
  21. O'Brien MF. Allograft aortic root replacement: standardization and simplification of technique Ann Thorac Surg 1995;60(Suppl):92-94.
  22. Doty JR, Salazar JD, Liddicoat JR, Flores JH, Doty DB. Aortic valve replacement with cryopreserved aortic allograft: ten-year experience J Thorac Cardiovasc Surg 1998;115:371-380.[Abstract/Free Full Text]
  23. Lund O, Chandrasekaran V, Grocott-Mason R, et al. Primary aortic valve replacement with allografts over twenty-five years: valve-related and procedure-related determinants of outcome J Thorac Cardiovasc Surg 1999;117:77-91.[Abstract/Free Full Text]

   Appendix Reference
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Comment
 Discussion
 Comment
 Footnotes
 Acknowledgments
 References
 Appendix Reference
 

1. Roselli EE, Pettersson GB, Blackstone EH, et al. Adverse events during reoperative cardiac surgery: frequency, characterization, and rescue J Thorac Cardiovasc Surg 2008;135:316-323.[Abstract/Free Full Text]




This article has been cited by other articles:


Home page
 J. Thorac. Cardiovasc. Surg.Home page
M. Schmoeckel, P. Boekstegers, K. Nikolaou, and B. Reichart
First successful transapical aortic valve implantation after aortic allograft replacement
J. Thorac. Cardiovasc. Surg., October 1, 2009; 138(4): 1016 - 1017.
[Full Text] [PDF]

Home page
 ESC Textbook of Cardiovascular MedicineHome page
A. Vahanian, B. Iung, L. Piérard, R. Dion, and J. Pepper
CHAPTER 21 Valvular Heart Disease
ESC Textbook of Cardiovascular Medicine, January 1, 2009; 2(1): med-9780199566990-chapter - med-9780199566990-chapter.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Edward R. Nowicki
Gösta B. Pettersson
Nicholas G. Smedira
Eric E. Roselli
Eugene H. Blackstone
Bruce W. Lytle
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nowicki, E. R.
Right arrow Articles by Lytle, B. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nowicki, E. R.
Right arrow Articles by Lytle, B. W.
Related Collections
Right arrow Valve disease

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
ANN THORAC SURG ASIAN CARDIOVASC THORAC ANN EUR J CARDIOTHORAC SURG
J THORAC CARDIOVASC SURG ICVTS ALL CTSNet JOURNALS