| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2003;75:1565-1571
© 2003 The Society of Thoracic Surgeons
a Department of Cardiopulmonary Transplantation, Freeman Hospital, Newcastle-upon-Tyne, United Kingdom
Accepted for publication November 27, 2002.
* Address reprint requests to Dr Abid, Department of Cardiothoracic Surgery, Victoria Hospital Blackpool, Whinney Heys Rd, Blackpool, UK FY3 8NZ.
e-mail: qumarabid{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentReferences |
|---|
METHODS: A retrospective analysis of the data for patients who had undergone cardiopulmonary transplantation between May 1985 and December 2000 was undertaken. A total of 776 patients had either a median sternotomy or a transverse sternotomy through a clamshell incision. Transplantations were as follows: 591 heart (3 simultaneous heart and renal, and 1 heart and liver), 126 bilateral sequential lung, 57 heart–lung, 1 en bloc double-lung, and 1 heart and single-lung.
RESULTS: In all, 21 (2.7%) recipients had mediastinitis. Of these, 14 had heart, 3 heart–lung, and 4 bilateral lung transplantation. There were 18 median and 3 transverse sternotomies. There were 6 deaths (28.6%). Treatment consisted of antibiotics alone in 2 patients and subxiphisternal drainage in another 2 patients. The sternum was reopened in 17 (80.95%) patients, with debridement and primary closure alone in 5 of these 17 patients and additional irrigation in the other 12. Those who had resternotomy, debridement, and substernal irrigation had a better outcome when compared with the outcomes of other modes of treatment (1 death among 12 patients) (p = 0.06). Age, cardiopulmonary bypass time, body mass index, time to diagnosis, and treatment did not differ between those who survived and those who did not.
CONCLUSIONS: Early aggressive debridement with substernal irrigation is the best mode of treatment for patients with posttransplantation mediastinitis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Improvements in the surgical environment and in surgical and operating room protocols, as well as an increased awareness of the dangers of infection may reduce the incidence of mediastinitis after cardiac operations [3]. Nevertheless, the principles of management of established mediastinitis should include early diagnosis (before there is extensive tissue destruction) and prompt resternotomy with aggressive debridement and irrigation.
Predisposing or risk factors associated with posttransplant mediastinitis have been reported [7, 8]; however in these studies the numbers were small. We reviewed our experience over 15 years with 21 cases of mediastinitis after cardiopulmonary transplantation to determine the incidence, mortality, risk factors, and prognostic factors. We also analyzed the results of the various methods that we used and identified the best option.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Statistical analysis
Data was expressed as mean and standard deviation. The Mann-Whitney rank test was used to assess the statistical significance of continuous variables, and the Fisher exact test was used to find relationship between two categorical variables. The p value was considered to be significant when less than 0.05.
Immunosuppressive treatment
A standard immunosuppressive regimen was used in all patients. This consisted of preoperative oral cyclosporin A (Sandimmum, 2 to 6 mg/kg, depending on renal function) and azathioprine (Imuran 4 mg/kg) along with prednisolone in cardiac transplant recipients. Antithymocyte globulin (equine lymphoglobulin) antihuman lymphocyte immunoglobulin 3) was used postoperatively as a short induction course for lung transplant recipients, maintaining an absolute lymphocyte T-cell count of less than 50,000 x 103/mL. Intravenous cyclosporin was used twice daily for the first 5 days to maintain a blood trough level between 350 and 400, followed by oral cyclosporin to achieve similar levels. Azathioprine was administered with the goal of achieving an absolute white blood cell count of approximately 5,000 x 109/L.
Over the series of pulmonary transplantations a regimen of steroid use was developed. All patients received steroids (methylprednisolone 500 mg IV) in theater before reperfusion, followed by 125 mg IV every 8 hours for the first 24 hours. However recipients in the initial one third of this series did not subsequently receive prednisolone postoperatively on a routine basis. Acute rejection episodes were treated with three pulse doses of methylprednisolone (10 mg/kg) followed by a rapidly reducing course of oral prednisolone from 1 mg/kg to 0.2 mg/kg over 10 days. Since early 1990s oral prednisolone was routinely administered, commencing at 1 mg/kg per day and reducing by 0.2 mg/kg every second day to a maintenance dose of 0.2 mg/kg per day. Acute rejection episodes were treated with three pulse doses of methylprednisolone (10 mg/kg). The steroids were weaned by 0.2 mg/kg every week to a maintenance level of 0.2 mg/kg per day over 1 month.
Antibiotic treatment
Routine prophylactic antibiotics for heart transplantation was flucloxacillin (2 g intravenously [IV]) at induction of anesthesia, followed by 500 mg IV every 6 hours for a total of 48 hours. Patients who were allergic to penicillin received clindamycin (600 mg IV) at induction and thereafter every 12 hours for 48 hours. In addition to the above antibiotics the lung transplant patients also received metronidazole 500 mg at induction followed by 500 mg IV every 8 hours until bronchoscopy (7 to 10 days). Patients with septic lung (positive sputum culture) were treated with aztreonam 2 g IV at induction followed by 2 g IV every 8 hours for 48 hours or as indicated by preoperative cultures and sensitivity. Prophylactic antimicrobial measures before surgical intervention for mediastinitis was guided by microbiological reports.
Operative protocol
All the operations were performed in a single center. The recipient criteria, perioperative management protocols, skin preparation, and draping were similar in all patients. The sternum was closed with interrupted stainless steel wires. The fascia and the subcutaneous and subcuticular tissues were closed with continuous absorbable sutures.
Definition of mediastinitis
We defined mediastinitis as the presence of pus or bacterial growth, or both, in mediastinal tissue or fluid sampled by aspiration or swab or during surgical exploration, either with or without sternal instability.
Surgical options used in the last 15 years
Subxiphisternal drainage
With the patient in the operating room under general anesthesia the subxiphisternum was opened. The substernal necrotic and infected tissue was debrided as far cephalic as possible using a combination of finger and blunt-instrumental dissection with suction. The resultant cavity was irrigated with 0.2% povidone-iodine and two substernal drains were left in situ. The incision was then formally closed.
Resternotomy, debridement, and primary closure
The mediastinum was reopened through the median or transverse sternotomy. All sternal wires and other suture materials were removed. Tissue and fluid samples were sent for microbiological culture and further antimicrobial sensitivity. All necrotic and infected tissue was excised. The bone edges were debrided with bone nibblers and cleaned by curettage.
Resternotomy, debridement, primary closure, and irrigation
After debridement to the point of obtaining clean viable tissue, the chest was closed after irrigation with aqueous iodine, or a suprasternal inflow port (catheter size 10F to 12F) was positioned for the inflow of a 0.2% povidone-iodine continuous irrigation solution. Two additional substernal drains (size 28F) were then positioned and exited through the subcostal/epigastric region as outflow tracts. The chest was then formally closed around the closed irrigation circuit; irrigation commenced 1 to 2 hours poststernotomy, once bleeding had stopped. The initial rate of infusion (100 mL/h for 12 hours) was then reduced (50 mL/h). The irrigation was stopped and drains were removed after three consecutive negative mediastinal fluid cultures. The mean duration of irrigation was 5 days (range 3 to 7 days).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Clinical presentation
The most common early features alerting us to the presence of mediastinitis were as follows (Fig 1):
(1) an abnormal white blood cell count (<4.0 or >14 x 109/L) in 20 patients (90%); (2) pyrexia (>37.5° C) in 15 patients (71%); retrosternal pain in 9 patients (40%); and (4) an oozing sternal wound in 5 patients (24%).
|
Diagnostic investigations
For diagnostic investigations (Table 1),
computed tomography (CT) of the chest was helpful in 19 patients. The radiographic features of mediastinitis appeared as mediastinal widening, pneumomediastinum, obliteration of fat planes, localized fluid collections, multiloculated collections, and abscess formation. The most common diagnostic finding was a large substernal fluid collection (10 patients) with or without additional pleural or pericardial effusion. Small substernal fluid collection (n = 2) was the least diagnostic feature on CT. Pneumomediastinum, abscess formation, and obliteration of fat planes were the most definitive features of mediastinitis. Delayed or suboptimal intervention in these patients carried a very poor prognosis. Only 7 of the 21 patients demonstrated a pericardial effusion during transthoracic echocardiography, all of these patients had cardiac transplants. Chest roentgenography, which revealed a widened mediastinum, was helpful in strengthening a suspicion of mediastinitis in 1 patient only.
|
The most common infecting organism in cardiac transplant patients was Staphylococcus aureus. The infecting organism complicating pulmonary transplantation were variable (ie, Pseudomonas, Burkholderia cepacia, Staphylococcus epidermidis, Escherichia coli, and Klebsiella. Candida albicans was the cause of mediastinitis in 3 patients, 2 of whom died.
Risk factors
Immunosuppression by any cause is a well-known risk factor for mediastinitis in patients requiring median or transverse sternotomy. All patients in this study received steroid-based, triple immunosuppression therapy. Almost 50% (n = 10/21) of the patients had impaired renal function (creatinine >110 µmol/L). Two patients had undergone a previous sternotomy before transplantation, and another 2 had undergone reexploration for bleeding after transplantation.
Prognostic factors
Various factors affecting the outcome of the established mediastinitis were analyzed (Table 2).
The total numbers of patients were too small to have any statistical significance (p = ns for all variables). Of the 21 patients, 6 were in shock before any surgical intervention. The outcome was very poor in these patients, 5 of whom died (83.5% mortality). Of the 21 patients, 6 had positive blood cultures before any surgical intervention, and 4 of them died (66% mortality). Delayed diagnosis (mean delay 4.5 days) also carried a poor prognosis. A low white blood cell count (<4.0 x 109/L) was the presenting feature in 2 patients, both of whom died.
|
Treatment modes
Two patients had a 6-week course of antibiotics (first intravenous and then oral) as first-line treatment. In both of these patients the culture from the pericardial aspirate was positive; however, they had no other clinical feature of mediastinitis nor any radiological evidence of tissue destruction, and a decision was therefore made to treat them conservatively. Both of these patients survived their infections.
Subxiphisternal drainage was used in 2 patients. Of these patients 1 died; the other required subsequent resternotomy, debridement, and closed irrigation but still eventually died with multiorgan failure. The sternum was formally reopened in the other 17 patients (80.95%) with debridement and primary closure. Of these patients 12 had substernal continuous irrigation of 0.2% povidone-iodine, usually for 48 hours.
There were 6 deaths (28.6%). These included 3 HTx, 1 HLTx, and 2 BSLTx patients, all of whom patients experienced multiorgan failure. Both patients with subxiphisternal drainage died. Three of the 5 patients who underwent resternotomy debridement and primary closure without irrigation also died. In addition, 1 of the 12 patients requiring resternotomy, debridement, and primary closure with irrigation died (p = 0.06 vs other surgical treatment). This patient developed a fecal peritonitis due to colonic perforation postdebridement and irrigation that not diagnosed until the onset of multiorgan failure, which ultimately was fatal. Since 1996 when we adopted resternotomy, debridement, and primary closure with irrigation as our preferred mode of surgical intervention we have had no mortality in this group of patients. Only 1 of the survivors, who had undergone BSLTx for bronchiectasis, required further treatment in the form of pectoral muscle flaps. This patient had sternal osteomylitis and required extensive sternal debridement.
Cardiopulmonary bypass time and body mass index did not differ between survivors and nonsurvivors (Table 2). Delayed diagnosis, positive blood culture, and low white blood cell counts carried a poor prognosis.
Economic considerations
The patients who had undergone resternotomy, extensive debridement, and closed irrigation had a shorter stay in the intensive care unit (3 days, range 1 to 5 days) and a shorter hospital stay (28 days, range 21 to 42 days). The majority of the patients treated with other surgical modalities (5 of 7 patients) died, after prolonged intensive care unit stays (13 days, range 7 to 21 days). Two patients who survived had a hospital stays of 39 days and 46 days. The strategy of early exploration, extensive wound debridement, and closed irrigation was associated with a shorter intensive care unit stay and shorter hospital stay and hence is cost-effective.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Obesity, internal mammary artery use, prolonged intensive care stay (>5 days), reexploration, diabetes, New York Heart Association congestive cardiac failure class, previous cardiac surgery and duration of cardiopulmonary bypass time are well-reported risk factors in the setting of conventional cardiac surgery [9, 11]. However, as a reflection of our strict recipient selection criteria, none of the patients in the study group had diabetes, obesity, chronic obstructive airway disease, or concomitant internal mammary artery harvesting. All of the patients, by virtue of their immunosuppressive drug regimen, were at high risk for mediastinitis. In addition, in 4 patients a intraaortic balloon pump was used pretransplantation to support the failing myocardium. Two patients had previous cardiac surgery, and 2 patients underwent reexploration for postoperative bleeding.
With immunosuppressive drugs (particularly steroids) the classic symptoms and signs of acute mediastinitis (sternal pain, fever, leukocytosis, and sternal discharge with or without sternal instability) may be subdued or nonexistent. Persistent fever and leukocytosis were the most common early clinical findings alerting us to the possible presence of mediastinitis. However these signs are frequent after cardiac surgical procedures and may be indicative of an occult sepsis. The fever may mistakenly be attributed to a reaction to medication or to the result of an opportunistic infection. The leukocytosis may be due to steroid administration (or leukopenia induced by drugs such as azathioprine). High indices of suspicion coupled with appropriate investigations are therefore crucial during the postoperative course of these patients.
Positive cultures (wound, blood, or aspirated pericardial effusion) should be aggressively sought for preoperative identification of a causative organism. Chest roentgenography was unhelpful in diagnosing mediastinitis. However, CT scan was helpful not only in suggesting the diagnosis (especially when the importance of clinical parameters of infection were unclear) but also in assessing the possible extent of mediastinal involvement. Our CT diagnosis of mediastinitis, in keeping with reports by others [12, 13], was based on the presence of mediastinal air and fluid collections, with or without adjacent abnormalities such as soft tissue edema or sternal separation. Nonetheless, such findings may also be found in healthy patients after median sternotomy [10]. The sensitivity of CT for diagnosing mediastinitis has been reported to be 67% with a specificity of 83% [14]. In our experience CT was helpful in establishing the diagnosis in 90% of patients (diagnostic in 57% and supportive of clinical suspicion in another 33%). When the diagnosis is still in doubt, gallium-67 scintigraphy may offer additional information to localize the site of retrosternal infection [12, 15]. We did not need to use this investigative modality in our patients.
The degree of mediastinal involvement, although often difficult to assess clinically, may range from a relatively localized to a deep-seated and imminently life-threatening infection, with or without osteomyelitis. Consequently the treatment will vary according to the perceived mediastinal involvement and extent of tissue necrosis, which in turn will be influenced by the rapidity of diagnosis. Suppurative mediastinitis is undoubtedly a surgical problem; however, in keeping with our own experience and that of others [7], a few patients with lesser degrees of infection have been amenable to prolonged antibiotic therapy [1]. However, conservative measures of this kind, which are considered to be associated with unacceptable mortality and morbidity [1], have largely been abandoned in favor of more aggressive surgical intervention. Similarly the use of sternal debridement, dressing changes, and closure by secondary intention, which is associated with a mortality of approximately 30% [16], has been abandoned.
Surgical intervention should include the fundamentals of approaching any surgical infection, including adequate debridement, proper cleaning, and appropriate use of antimicrobial agentss. In our experience we have found minimal intervention, in the form of subxiphisternal drainage, to be inadequate. Both patients treated in this way had multiorgan failure. One patient died before any further intervention. The other patient required further intervention in the form of resternotomy, debridement, and closed irrigation; however, the existing multiorgan failure did not resolve and this patient eventually died. The amount of tissue accessible for debridement by subxiphoid approach is limited. Any residual infected tissue may maintain a state of sepsis, which may then lead to multiorgan failure. It is possible that earlier intervention by subxiphisternal drainage might have prevented these deaths.
At its most extreme, surgical intervention may include complete sternectomy and plastic reconstruction with muscle or omental flaps [17, 18]. However, with prompt diagnosis and rapid intervention (and thus minimization of tissue destruction and general physiological deterioration) we have found such extreme measures to be largely unnecessary.
Our preferred intervention (as used in 17 of the 21 patients) consists of a formal resternotomy, debridement with peroperative irrigation, and primary sternal closure. In addition to this, in 12 of 17 patients, we used closed mediastinal irrigation, as first described by Shumacker and Mandelbaum [19], with 0.2% povidone-iodine. With our regimen of irrigation we have not found iodine toxicity to be a problem [20]. Both of the patients who underwent subxiphoid debridement died. In addition, 3 of 5 patients who underwent resternotomy, debridement, and primary closure without irrigation died. However, only 1 of 12 patients in the group of irrigation with resternotomy and debridement died. This patient developed fecal peritonitis, which was not diagnosed until the onset of multiorgan failure, to which he succumbed despite all possible supportive measures. Early diagnosis, extensive surgical debridement, and irrigation saved 92% of the patients with posttransplant mediastinitis in whom this approach was applied.
If eradication of the mediastinitis is not achieved after early diagnosis, debridement, and closed irrigation, we then (as in 1 case) consider a plastic reconstructive procedure such as transposition of muscle or omentum. Indeed a number of retrospective studies have compared closed irrigation with reconstructive procedures [21, 22] and have found similar short-term and long-term results for both treatment modalities. The report by Scully and colleagues [22] concluded that the techniques were equally effective in the treatment of mediastinitis after cardiac surgery. With such evidence, coupled with our own experience, we find it difficult to justify use of plastic reconstruction procedures, as advocated by some [8], or extensive mobilization of omentum [23, 24] to treat mediastinitis as a primary mode of therapy, provided that reexploration is undertaken early. However, in patients who need extensive sternal debridement, plastic reconstruction in the form of muscle flaps can be justified. We agree with Stolf and colleagues [25] that promptness in diagnosis and early intervention is the key to a successful outcome in these patients. The early intervention may preclude the need for extensive surgical debridement and may bring about improved survival, with a reduction in the duration of intensive care unit and hospital stay. Fungal and polymicrobial mediastinitis carries a very poor prognosis (66% mortality) in comparison with Staphylococcus aureus (11.1% mortality); therefore a more aggressive approach is needed in these patients.
Although the patient numbers in our study are too small to allow any definite conclusions, our experience suggests that a management policy encouraging early diagnosis followed by an aggressive resort to early resternotomy, debridement, and closed irrigation is one that could be usefully extended to all patients with mediastinitis after sternotomy. This approach decreases intensive care unit and hospital stay and is therefore cost-effective. Delayed diagnosis, positive blood culture, shock, fungal or polymicrobial mediastinitis, and low white blood cell count carry a poor prognosis.
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. S. George, E. J. Birks, S. Haj-Yahia, C. T. Bowles, A. Hall, A. Khaghani, and M. Petrou Acinetobacter Mediastinitis in a Heart Transplant Patient Ann. Thorac. Surg., August 1, 2006; 82(2): 715 - 716. [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 |
