| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Division of Cardiac Surgery, University of Maryland Medical Center, Baltimore, Maryland
c Emergency Department, University of Maryland Medical Center, Baltimore, Maryland
e Division of Epidemiology, University of Maryland Medical Center, Baltimore, Maryland
f Department of Neurology, University of Maryland Medical Center, Baltimore, Maryland
b Division of Infectious Diseases, The Oregon Health & Science University, Portland, Oregon
d Division of Cardiac Surgery, The Oregon Health & Science University, Portland, Oregon
Accepted for publication February 3, 2009.
* Address correspondence to Dr Gammie, Division of Cardiac Surgery, University of Maryland Medical Center, N4W94, 22 S Greene St, Baltimore, MD 21201 (Email: jgammie{at}smail.umaryland.edu).
Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5–8, 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: We reviewed 89 consecutive operations in 87 patients for native mitral valve IE at a single institution from 2002 to 2007. Operations occurred promptly after completion of preoperative studies. Independent risk factors for death were investigated using multivariable logistic regression.
Results: Mitral valve repair was accomplished in 56 of 89 patients (63%). Perioperative mortality was 4.4% (n = 4). Survival rates at 1 and 5 years were 89.9% (80 of 89) and 82.0% (73 of 90). There was a survival benefit for repair vs replacement at 1 (p = 0.03) and 5 years (p = 0.0017). Repair vs replacement (odds ratio [OR], 0.2; 95% confidence interval [CI], 0.06 to 0.72), diabetes (OR, 4.43; 95% CI, 1.18 to 16.66), and renal failure (OR, 3.65; 95% CI, 1.3 to 12.91) were independent risk factors for late mortality. Among 59 patients with active IE, preoperative head computed tomography (CT) showed 29 (49%) had abnormalities, including 12 (41%) with intracerebral hemorrhage. The median interval was 4 days from admission to operation. The rate of permanent postoperative stroke was 1.1% (1 of 89).
Conclusions: These results support early surgical therapy for mitral valve IE. Head CT abnormalities do not warrant delay of operation. Mitral valve repair was associated with a long-term survival advantage compared with valve replacement.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
The greatest opportunity to decrease the mortality and morbidity of IE may be the proper application of surgical therapy [2]. The decision to operate, timing of the operation, and type of operation performed remain individualized and are often based on anecdotal experience and individual clinician preferences; for example, recent publications both support [3] and refute [4] the value of surgical intervention. Only 5% of mitral valve operations in North America are for IE, making it difficult to accrue clinical experience from which decisions about the optimal application of surgical therapy can be derived [5].
We reviewed results of surgical therapy for mitral valve IE at a single institution that admits more than 110 patients each year with IE. We applied a uniform clinical approach that includes an emphasis on early operation and mitral valve repair rather than replacement. We hypothesized that (1) early operation is associated with optimal outcomes, (2) mitral valve repair is superior to replacement, and (3) most abnormalities seen on head computed tomography (CT) imaging should not delay surgical intervention.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
All patients were assigned a diagnosis of IE based on the modified Duke criteria [6]. The definition of active and treated IE was based on the STS National Adult Cardiac Surgery Database criteria: if the patient was being treated for endocarditis at the time of operation, the disease was considered active. If no antibiotics were being given at the time of operation other than prophylactic medication, the infection was considered treated.
Preoperative head, abdomen, and chest CT imaging was performed on all patients. Patients with risk factors for coronary artery disease underwent left heart catheterization without ventriculography. All patients underwent transesophageal echocardiography (TEE) before operation, and all patients received a predismissal echocardiogram. Indications for operation included any of the following: moderate or severe valvular regurgitation, clinical congestive heart failure, vegetations that were mobile or larger than 10 mm, persistent septicemia, fungal IE, or evidence of a single embolic event.
Operative Techniques
Median sternotomy was used for all patients with active IE. Right chest small thoracotomy incisions were used for selected patients (age < 75 years and body mass index < 30 kg/m2) with treated endocarditis. For sternotomy patients not requiring tricuspid intervention, single venous cannulation with vacuum-assisted drainage was always used [7]. Myocardial protection was with antegrade and retrograde cold blood cardioplegia. Exposure was through a traditional left atrial incision in the interatrial groove. The mitral valve was carefully examined, all infected and inflamed tissue sharply resected, and the feasibility of repair evaluated. Vegetations were routinely sent for microbiologic examination.
Key factors determining reparability were the amount and location of tissue destruction by the endocarditic process and the quality of the remaining tissue after débridement. If after débridement more than 50% to 60% of the posterior leaflet was absent, or if more than 10% to 20% of the free edge of the anterior leaflet was destroyed, repair was usually not possible. Large amounts of anterior leaflet destruction are commonly present (often associated with a perforation) and do not necessarily mandate replacement, because an autologous pericardial patch can be used to successfully reconstruct a large percentage of the body of the anterior leaflet. It is important that tissue remaining after débridement is of sufficient quality to hold the sutures upon which the repair is based.
We agree with Frater [8] that absolute microbiologic sterility of the remaining tissue is not essential, because IE without vegetations generally heals with antibiotic therapy alone, as will repaired valves. For isolated posterior leaflet lesions, a standard resectional approach (either quadrangular or triangular resection in combination with a flexible partial annuloplasty band) was used [9]. In cases in which an annular abscess was present (usually in the posterior atrioventricular groove), the abscess was débrided and the cavity was closed with fresh untreated autologous pericardium anchored to the ventricle and atrium with closely spaced bites of 4-0 monofilament suture.
In cases of anterior leaflet perforation, débridement of the perforation and reconstruction with fresh untreated autologous pericardial patch (using 6-0 monofilament running suture) was the preferred approach. If chordal support was damaged as a result of the infective process, reconstruction of the free edge of the anterior leaflet with Gore-Tex neochords (W. L. Gore, Flagstaff, AZ) was an important method of repair. A complete rigid ring was used in conjunction with anterior leaflet repairs. Although early in our experience we avoided an annuloplasty ring for simple leaflet resections, we now prefer to use a ring in all repairs. Commissural lesions were treated with sliding advancement plasties as described by Carpentier [9].
Postoperative Care
All patients undergoing mitral valve repair were treated with daily aspirin only.
Patient Follow-Up
The primary outcome was operative mortality, defined as in-hospital or 30-day mortality, whichever was greater. Late mortality was assessed using the Social Security National Death Index [10]. Long-term follow-up data were obtained using a cross-sectional (common closing date) method during a 4-month period through contact with patients or relatives and their primary care physicians. Information on the cause of death was obtained through physicians and official hospital records, as well as review of death certificates. Clinical follow-up was available for 81 of 87 patients (93%). Long-term results were assessed on basis of mortality and echocardiography. Recurrence of endocarditis and reoperations were investigated and documented.
Statistical Analysis
All-cause mortality was defined as the primary outcome measurement, and valvular dysfunction requiring reoperation and recurrent endocarditis were regarded as events. Kaplan-Meier analysis and log-rank testing was used to compare survival between mitral valve repair and mitral valve replacement using MedCalc 8.1 (MedCalc Software, Belgium). Multiple logistic modeling was performed on the end point of overall mortality with 14 clinical variables, including presence of active disease, age, concomitant aortic and tricuspid valve procedures, coronary artery bypass grafting (CABG) procedure, presence of congestive heart failure, and history of intravenous drug abuse, cerebrovascular accident, myocardial infarction, diabetes, renal failure, Staphylococcus aureus endocarditis, and methicillin-resistant S aureus (MRSA) endocarditis. A backwards step-wise strategy was used, with variables significant at p < 0.1 being included. Results were considered statistically significant at p < 0.05. All p values are two-tailed.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
|
|
Mitral valve repair was successful in 56 patients (62.2%). Repair rates for active (61%, 36 of 59) and treated IE (67%, 20 of 30) were similar. Overall 13 of 56 patients (23%) had isolated anterior leaflet repairs, 22 (9%) had posterior repair, and 21 (38%) had bileaflet repairs. Posterior leaflet repair techniques included triangular resection in 7, quadrangular resection in 9, quadrangular resection with sliding plasty in 2, and direct suture closure of perforation in 12. Anterior leaflet repairs included patch closure of perforations with autologous pericardial patches (not treated with glutaraldehyde) in 17 patients, and bovine pericardium in 2. Commisural repairs were undertaken in 14 (15.6%) of patients, and Gore-Tex neochords were used in 11 (12.2%). Four patients had patch closure of a posterior annular abscess. Some patients required multiple techniques to achieve repair. Complete rigid annuloplasty rings (Carpentier-Edwards Physio, Irvine, CA; n = 30), and partial flexible bands (Cosgrove-Edwards, n = 14) were used in 44 repairs (78.6%). Twelve patients received mitral valve repair without an annuloplasty ring. On predismissal echocardiography, 8 of these patients had none or trace mitral regurgitation (MR), and 4 (33%) had mild MR. At the time of the follow-up echocardiogram (average of 94 ± 221 days), 6 of these 12 patients (50%) had mild MR and the remaining 6 had none or trace MR. The remaining 34 patients (37.8 %) received a bioprosthetic (n = 24) or mechanical (n = 10) valve replacement.
Concomitant cardiac surgical interventions were performed in 39 patients (43%), including aortic valve replacement in 16 (18%), comprising 13 bioprosthetic valves and 3 mechanical valves, and aortic valve repair in 2 (2%). Tricuspid valve reconstruction with a rigid annuloplasty ring was performed in 10 patients (11%), and 1 patient (1%) received a bioprosthetic replacement tricuspid valve. CABG was performed in 9 patients (10%).
Perioperative Outcomes
The in-hospital mortality was 4.5% (n = 4), with 3 deaths in the repair group and 1 in the replacement group. Two of the 4 patients who died underwent concomitant cardiac procedures (1 CABG and 1 aortic valve replacement). Causes of death after operation included 1 patient with bowel ischemia at 1 week and 1 patient with unexplained diffuse coagulopathy at 1 week. Operative deaths were due to profound vasodilatory shock in 2 patients with florid sepsis. Overall perioperative morbidity was 45% (40 of 89 patients), with 39% (22 of 56) after valve repair and 55% (18 of 34) after valve replacement (p = 0.15). Complications included reoperation for bleeding in 7 patients (8%). Renal failure developed in 11 patients (12%), with 6 (7%) requiring new dialysis. Sepsis was diagnosed in 3 patients, but no mediastinitis developed. Six patients required tracheostomy.
One new permanent postoperative cerebrovascular accident was documented. This occurred in a 31-year-old woman with a normal preoperative head CT who underwent a bioprosthetic mitral valve replacement for active S viridans IE. She presented with an acute-onset right-sided hemiplegia on postoperative day 7, despite therapeutic anticoagulation with warfarin (international normalized ratio, 2.2). Head CT and MRI demonstrated acute infarcts in an embolic distribution.
Predismissal echocardiography in patients who received mitral valve repair demonstrated none or trace MR in 41 (77.4%), mild MR in 11 (20.7 %), and moderate MR in 1 (1.9%). Recurrent IE was diagnosed in 2 patients, both after mitral valve repair. In 1 patient the initial operation involved repair of a posterior leaflet perforation associated with MSSA and cardiomyopathy. The perforation was closed primarily and an undersized complete rigid annuloplasty ring was inserted. Reoperation for severe MR was necessary 4 months after the initial repair, and partial dehiscence of the ring associated with vegetations (culture negative) was found. A new annuloplasty ring was inserted and the patient remains free of mitral regurgitation and infection 27 months postoperatively.
A second patient was diagnosed with MSSA mitral IE while receiving chemotherapy for lymphoma. The initial operation included resection of a large vegetation on the anterior leaflet and placement of a complete rigid annuloplasty ring. Predismissal echocardiography showed moderate MR. Candida albicans mitral IE and severe MR developed 20 weeks after the initial operation. At reoperation, partial posterior ring dehiscence associated with a large vegetation was observed. A new rigid annuloplasty ring was inserted, and 2 Gore-Tex neochords were placed to the anterior leaflet. Trace MR was noted on the predismissal and 2-month echocardiograms. He remains free of infection and clinically well 3 years later. No additional reoperations for failed mitral valve repair have been identified.
Long-Term Outcomes
One-year mortality among all patients with mitral valve IE was 12.3% (n = 11). One-year (94.6% vs 82.4%) and 5-year (91.1% vs 70.6%) survivals were significantly better for those patients undergoing mitral valve repair compared with replacement (p = 0.004; Fig 1). Repair vs replacement (odds ratio [OR], 0.21; 95% confidence interval [CI], 0.06 to 0.72), diabetes (OR, 4.43; 95% CI, 1.18 to 16.66), and preoperative renal failure (OR, 3.65; 95% CI, 1.3 to 12.91) were independent risk factors for late mortality.
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Early operation (median time from admission to operation of 4 days) did not compromise neurologic recovery in these patients. We believe that advantages of early operation include prevention of additional thromboembolic events, minimization of ongoing valvular destruction with an increased likelihood of mitral valve repair compared with replacement, shortened hospitalization, and earlier relief of hemodynamic compromise caused by mitral regurgitation.
Timing of operation for left-sided IE associated with severe valvular insufficiency remains a controversial topic. A common approach embraces the notion that "surgery is best delayed until antibiotics are completed to reduce the risk of perioperative complications and early prosthetic valve endocarditis" [11]. A recent retrospective, propensity-matched study from a quaternary care center examined the effect of surgical intervention on outcomes for 546 patients with left-sided IE and found that an operation had no survival benefit [4]. Among the key limitations of that report were that it was not contemporary (1980 to 1998) and that the mortality rate in the surgically treated group was high at 27%. More recent propensity-matched studies suggest a strong benefit to early operation for early-term [12] and long-term survival [3]. In the absence of a prospective randomized trial, we believe that surgical series such as ours and others that report operative mortality rates of between 3% and 12% [13–15], when compared with reported outcomes from medically-treated IE, support an aggressive operative approach for patients with mitral valve IE [12].
Although most predictors of long-term survival in this experience were related to patient characteristics that are not modifiable, the process of mitral valve repair was highly associated with improved long-term survival. Global application of mitral valve repair for patients with IE is low, with only one-third of patients undergoing repair in the STS Adult Cardiac Surgery Database compared with nearly half of patients with other indications for mitral valve intervention [5].
This series and others from centers with focused interest on mitral valve repair suggests that repair rates might be improved [9, 15–17]. In addition to the generally recognized advantages of mitral valve repair of superior ventricular function, lower operative mortality rate, low long-term rates of thromboembolism, and reoperation, mitral valve repair is likely associated with lower rates of reinfection compared with replacement. Published rates of prosthetic valve endocarditis after mitral valve replacement for IE are 8% to 27% [16, 18], whereas reported rates of repaired valve endocarditis are as low as 0.04% per year [19]. Although we experienced repaired valve IE in 2 of 56 cases (3.4%) both were amenable to repeat repair and neither was associated with death. In our experience, repair was durable, with excellent freedom from reoperation and minimal mitral regurgitation on follow-up echocardiography. Increased surgeon adoption of mitral valve repair for IE is an important care process that will improve long-term outcomes among patients in whom repair is amenable.
The safety of cardiopulmonary bypass (CPB) and cardiac operation in patients with acute neurologic injury has been debated [20–22]. It has been proposed that CPB can exacerbate neurologic deficits due to the risk of heparinization and subsequent hemorrhagic conversion. Hypotension during CPB might also worsen cerebral ischemia and increase parenchymal edema.
One report recommends delaying cardiac operation for IE for 2 weeks if evidence of cerebral embolism is present on head CT and for 4 weeks if hemorrhage is present [22]. Our experience is at odds with this recommendation. Although 50% of patients with active IE in this series had abnormalities on head CT, we pursued early operation (median length of time from admission to operation of 4 days) and noted only one permanent postoperative stroke. There were no cases of hemorrhagic conversion or of progression of brain injury. Based on our experience, we do not delay mitral valve intervention because of subarachnoid blood, cerebral infarction, or parenchymal hemorrhage that does not exceed 1 to 2 cm in size.
Limitations of this experience include incomplete follow-up, which was compromised by noncompliance among some (intravenous drug abusers) of our patients. There remains a wide spectrum of referral patterns among internists and infectious disease specialists at our institution, and it is possible that there was a selection bias in referral for operation. Patients undergoing mitral valve replacement had greater degrees of valvular destruction, and it is possible that this may have been a marker of more advanced disease.
In conclusion, we have found that early operation and aggressive application of mitral valve repair was associated with low rates of perioperative and long-term mortality among patients with mitral valve IE. Our current approach to patients with mitral valve IE includes routine head, chest, and abdominal CT scanning, transesophageal echocardiography, and left-sided heart catheterization without ventriculography, if warranted. The presence of a splenic abscess (compared with an infarct) mandates splenectomy or percutaneous drainage before the cardiac operation [23]. Unless the patient is comatose, we do not delay operation because of neurologic symptoms or abnormalities on brain imaging, with the exception of large (> 1- to 2-cm) hematomas. The presence of a hematoma warrants neurosurgical consultation and consideration of cerebral angiography to rule out a mycotic aneurysm.
Mitral valve repair is the favored intraoperative approach. Repair is associated with improved long-term survival compared with replacement is durable and can be maximized by methodical valve assessment and use of a variety of repair strategies.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This series provides us with several opportunities for discussion. First of all, your repair rate of 63% in a very difficult group of patients is to be commended. I actually had the opportunity to re-review our Michigan series, and we are also in the mid-60s, and this is essentially or nearly the same as the STS repair rate now that your lead author, Dr Gammie, will talk to us about in a little while as well, and I think that is commendable. Your mortality, also in the 4% range in this most challenging group of patients, shows your commitment not only to analyzing the appropriate patients to operate upon, perhaps your industry in the decision making about the timing of it is also part of this. And finally, although you are symbolized by a somewhat cumbersome terrapin, your decision to operate earlier may make repair perhaps a more likely approach.
I have three questions. Can you elaborate, as you noted in your manuscript, on the transition to using more rings? Because initially, you had 12 patients that did not have rings in some of the earliest parts of your repair series. A second question, was the repair rate different for those that had the active endocarditis that you operated upon, which is about two-thirds of your patients, as compared to those who were fully treated? And third, are there any CT [computed tomography] findings or neurologic findings that preclude or have you delay operative intervention? It was an excellent presentation. Thank you.
DR SHANG: Thank you for the questions. In response to the first question, you are certainly right that the practices have changed over the years. Initially we had sought to save people from the placement of an annuloplasty ring, but we later found that on intraoperative saline testing and TEE [transesophageal echocardiography] after we come off of bypass that we have less than perfect results that way. It is now our thought and our practice that a complete annuloplasty ring, a slightly undersized one, be placed to maximize coaptation, and we found much better results that way. And to make an additional point, I know that there are people that are hesitant to place annuloplasty rings in the setting of endocarditis due to worry of infection, but we feel that both the literature and our experience show that reinfection rates are very low.
As far as to the second question, the repair rates in both our active and our treated populations were roughly identical: 69% of treated endocarditis patients were successfully repaired vs 63% of the active patients.
With regards to things and CT abnormalities that would preclude us from surgery, we would be hesitant to offer an operation to patients that were profoundly neurologically devastated. Similarly, if on head CT they had evidence of a large bleed, for example, greater than 1–2 cm in diameter, then at that time we would probably obtain a neurosurgical consult and that might sway us from an operation.
DR KEVIN D. ACCOLA (Orlando, FL): Very nice presentation. I would like to expand on the last point that Dr Prager made with his last question, because I think that is a controversial point of your presentation. Obviously, if you have a large neurological defect you are going to not operate. I am concerned about stroke in evolution in these patients as they may be quite stable from a cardiac standpoint. What part of your diagnostic and treatment algorithm does the patient's current hemodynamic status play in your decision making process? In other words, if you have a quite stable patient and a neurological deficit, do you wait to operate? Our neurologists are quite nervous when we want to take a patient to the operating room with any type of CT changes suggestive of a possible evolving deficit.
DR SHANG: It is true that there is some nervousness and some controversy over this topic. As far as our data, we only had one incidence of postoperative stroke, and that was on a patient with a normal CT and a normal MRI preoperatively. Furthermore, we didn't have any incidence of hemorrhagic conversion of any of the infarcts or bleeds that were previously identified.
DR CULLEN MORRIS (Athens, GA): Great presentation and outstanding results. Just a couple of questions. One, all these patients you said got chest, abdomen, and pelvic CTs before they went to the operating room. What did you do with the patient that you found an abscess in, say a liver abscess or a spleen abscess? And then, also, do you think that the patients that did better with mitral valve repair just represents a patient that was sort of an easier operation vs the more difficult operation where you had to cut everything out and just replace it?
DR SHANG: Excellent questions. On preoperative head, chest, and abdomen CT, there were several patients that were identified to have infarcts and embolic events that were not in the head, for example, in the spleen and in the liver, as you mentioned. Those patients were treated with antibiotics. And, in fact, the presence of these embolic events on some occasions drove us forward to consider an earlier operation to minimize those events down the road.
DR GAMMIE: I just want to add to that. If we find a splenic abscess, we will take care of that beforehand. We know from the literature that the mortality rate is a 100% if you do operate on the valve, but that is a fairly uncommon scenario. We would treat a splenic abscess with a laparoscopic splenectomy or a percutaneous drainage.
Kevin, I appreciate your question, and we have sort of taken at some level a damn the torpedoes approach. We feel that waiting to operate on these patients has a clear downside potential for further embolization, further valve destruction. You saw that we had a lot of Staph aureus patients and there is ongoing valve destruction. And so we really just went forward and operated on these patients once their workup was complete, and I think our data support that approach. If, obviously, we had a large-volume hemorrhage, we would get neurosurgery involved. We generally would do an angiogram on those patients to rule out a mycotic aneurysm. But if there was hemodynamic instability and severe mitral regurgitation and some degree of a bleed and/or infarct, we would go ahead and operate on those patients, and I think this supports that approach.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Rossi, A. Gallo, R. Joseph De Silva, and R. Sayeed What is the optimal timing for surgery in infective endocarditis with cerebrovascular complications? Interact CardioVasc Thorac Surg, January 1, 2012; 14(1): 72 - 80. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Funakoshi, S. Kaji, A. Yamamuro, T. Tani, M. Kinoshita, Y. Okada, and Y. Furukawa Impact of early surgery in the active phase on long-term outcomes in left-sided native valve infective endocarditis J. Thorac. Cardiovasc. Surg., October 1, 2011; 142(4): 836 - 842.e1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Shin, M. Mori, K. Yamabe, and S. Iwanaga Reconstruction of the free margin of the anterior mitral leaflet with autologous pericardium in active infected endocarditis Eur J Cardiothorac Surg, May 1, 2011; 39(5): 784 - 785. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Omoto, T. Tedoriya, M. Oi, N. Nagai, T. Miyauchi, and N. Ishikawa Significance of mitral valve repair for active-phase infective endocarditis Asian Cardiovasc Thorac Ann, April 1, 2011; 19(2): 149 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Gaca, S. Sheng, M. A. Daneshmand, S. O'Brien, J. S. Rankin, J. M. Brennan, G. C. Hughes, D. D. Glower, J. S. Gammie, and P. K. Smith Outcomes for endocarditis surgery in North America: A simplified risk scoring system. J. Thorac. Cardiovasc. Surg., January 1, 2011; 141(1): 98 - 106.e2. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. F. Bolling, S. Li, S. M. O'Brien, J. M. Brennan, R. L. Prager, and J. S. Gammie Predictors of Mitral Valve Repair: Clinical and Surgeon Factors Ann. Thorac. Surg., December 1, 2010; 90(6): 1904 - 1912. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Pektok, J. Sierra, M. Cikirikcioglu, H. Muller, P. O. Myers, and A. Kalangos Midterm Results of Valve Repair With a Biodegradable Annuloplasty Ring for Acute Endocarditis Ann. Thorac. Surg., April 1, 2010; 89(4): 1180 - 1185. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
