- Email: [email protected]
Outcomes of Open Repair of Mycotic Descending Thoracic and Thoracoabdominal Aortic Aneurysms
ADULT CARDIAC
Outcomes of Open Repair of Mycotic Descending Thoracic and Thoracoabdominal Aortic Aneurysms Christopher Lau, MD, Mario Gaudino, MD, Andreas R. de Biasi, MD, Monica Munjal, MS, and Leonard N. Girardi, MD Department of Cardiothoracic Surgery, Weill Cornell Medical College, New York, New York
Background. The purpose of this study was to evaluate the short- and intermediate-term outcomes of open repair of mycotic thoracic and thoracoabdominal aneurysms. Contemporary surgical and perioperative techniques were utilized. Methods. From November 1997 to May 2014, 14 consecutive patients underwent open repair of descending thoracic (n [ 9, 64.3%) and thoracoabdominal (n [ 5, 35.7%) mycotic aortic aneurysms. All procedures were performed through the left side of the chest. Infected tissue was completely debrided and excised. Aortic continuity was restored in situ with a Dacron prosthesis (Macquet Corp, Oakland, NJ). Soft tissue coverage of the prosthesis was performed when anatomy and patient condition permitted. Perioperative outcomes, intermediate-term survival, and reinfection rates were examined. Results. All patients presented with either aneurysmrelated symptoms or a clinical picture of sepsis. Diagnosis was confirmed utilizing computed tomography imaging. Mean age was 66 ± 13 years, 8 patients (57.1%) were male, and mean aneurysm size was 5.9 ± 1.3 cm. All patients were hypertensive, 3 (21.4%) had prior coronary revascularization, 7 (50%) had chronic pulmonary disease, 5 (35.7%) had diabetes mellitus, and 2 (14.3%) had end-stage renal disease requiring dialysis. Twelve patients (85.7%) had aneurysm-related pain, and 9 (64.3%) of them had contained rupture. Mean time from onset of illness to surgery was 36 days (range, 0 to 153). On preoperative blood cultures, 4 (28.6%) grew Staphylococcus aureus, 4 (28.6%) grew gram negative organisms, 2 (14.3%) grew mycobacterium, and 4 cultures (28.6%) had negative results. Empiric broad-spectrum antibiotics were initiated
on all patients and adjusted based on final cultures. A majority of patients underwent repair utilizing a clamp-and-sew technique (n [ 10, 71.4%); the remainder (n [ 4, 28.6%) required repair under profound hypothermic circulatory arrest. After radical debridement of the infected tissue, grafts were placed in the normal anatomic position; 6 (42.9%) patients had additional soft tissue coverage, 5 (35.7%) utilizing an omental flap and 1 (7.1%), a serratus muscle flap. There was 1 in-hospital death (7.1%) secondary to ischemic bowel. Four patients (28.6%) required tracheostomy, and 1 (7.1%) had recurrent nerve injury. None of the patients incurred spinal cord injury, stroke, or new onset renal failure requiring dialysis. After surgery, all patients were given 6 weeks of intravenous antibiotics. Lifelong suppression therapy was maintained with oral antibiotics. There were no episodes of prosthetic graft infection on follow-up. Univariate analysis revealed that New York Heart Association functional class, diabetes, and preoperative renal dysfunction were preoperative risk factors for major adverse events. Mean follow-up time was 26.5 months (median 8.2; range, 1 to 142). Actuarial 5-year survival was 71%. Conclusions. Open repair of mycotic descending thoracic and thoracoabdominal aortic aneurysms remains the gold standard of therapy. Aggressive intraoperative debridement with in situ prosthetic reconstruction permits a high rate of success in this very high risk cohort of patients. Lifelong antibiotic suppression therapy may prevent late prosthetic graft infection.
M
resection to rid the patient of infection while reestablishing aortic continuity with either a prosthetic graft or a homograft [1–6]. Owing to the rarity of MTAA, data on treatment are limited to small series and case reports. In addition, much of the existing literature combines MTAA with mycotic abdominal aortic aneurysms, which are managed quite differently and are associated with vastly different postoperative complications [2, 7]. The majority of recent literature comes from Asia, where the incidence of mycotic aneurysmal disease is much higher and the microbiology much different, with a predominance of Salmonella species being responsible for this potentially
ycotic thoracic aortic aneurysms (MTAA) are an unusual variant of aneurysmal disease, comprising less than 1% of all aortic aneurysms [1, 2]. These rare subtypes of aneurysm are quite lethal, with a high incidence of fatal rupture when medical therapy alone is attempted [1, 3]. Although there are no standard guidelines for treatment, the most common practice is to combine directed antibiotic therapy with surgical
Accepted for publication May 14, 2015. Address correspondence to Dr Lau, 525 E 68th St, Box 110, Ste M404, New York, NY 10065; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;100:1712–7) Ó 2015 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.05.067
lethal disease [8–12]. The treatment algorithm for MTAA has also undergone a paradigm shift in the last decade. Traditional therapy consists of open graft replacement of the infected aneurysm, oftentimes supported with muscle flap or omental coverage of the replaced segment. However, thoracic endovascular aneurysm repair (TEVAR) has now supplanted open surgery as the most common type of initial treatment strategy [9–11, 13–15]. TEVAR has reduced the incidence of fatal rupture of the presenting disease, as seen in a recent large multicenter study with a 91% 1-month survival [16]. However, longer-term outcomes with TEVAR as sole therapy are questionable when implanting prosthetic material into an infected field without debridement or excision of the diseased aorta. We sought to examine the outcomes of an aggressive open surgery treatment strategy for MTAA. A combination of traditional surgical debridement and in situ aortic replacement was utilized with a variety of contemporary end organ protection strategies. Short- and intermediateterm outcomes are examined, and a reproducible treatment algorithm is proposed.
Material and Methods This study was approved by the Weill Cornell Medical College Institutional Review Board. The need for individual patient consent was waived. We reviewed prospectively collected data from the Weill Cornell Medical College Department of Cardiothoracic Surgery aortic surgery database to identify patients who had repair of MTAA. Aneurysms were defined as mycotic if they were associated with clinical signs of infection, characteristic appearance of radiologic imaging or intraoperative examination, and positive cultures from either blood or aortic tissue. Patients presenting with infected aneurysms of the ascending aorta or arch that were approached through a median sternotomy were excluded. From November 1997 to May 2014, a total of 680 patients underwent open repair of either descending thoracic aneurysms or thoracoabdominal aortic aneurysms (TAAA). Fourteen were mycotic, with 9 descending thoracic aneurysms (9 of 14, 64%) and 5 TAAAs (5 of 14, 36%). TEVAR was not used for any patients with MTAA. All procedures were performed through either a left thoracotomy or left thoracoabdominal incision. Infected tissue was completely excised, and aortic continuity restored with a Dacron prosthesis (Macquet Corp, Oakland, NJ). Soft tissue coverage of the prosthesis was performed when anatomy and patient condition permitted. Standard methods for end organ and spinal cord protection were utilized, and have been described previously [17], including ambient cooling to 33 C to 34 C, cerebrospinal fluid drainage when clinically stable, and an open distal anastomotic technique. Because of the need for aggressive tissue debridement, intercostal arteries were not reimplanted in any patients. Postoperative mean arterial pressures were maintained at 85 to 90 mm Hg for 3 days using a combination of aggressive fluid resuscitation, norepinephrine, and
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
1713
vasopressin infusions. Patients were initially given broad-spectrum antibiotics followed by directed antibiotic coverage after identification and sensitivities of the inciting organisms were obtained. Intravenous antibiotics were administered for 6 weeks after surgery followed by life-long oral antibiotic suppression therapy. Patients surviving surgery were then followed with serial computed tomography imaging. Patients with either radiographic or clinical evidence to suggest possible reinfection were evaluated with nuclear imaging utilizing indium or positron emission tomography (PET) scan to confirm or negate the diagnosis, although no patients had reinfection of the prosthetic graft.
Statistics Data were stored using Microsoft Access (2013) software (Microsoft, Redmond, WA). Processing and analysis of the data were performed using IBM SPSS Statistics, version 20 (IBM Corp, Armonk, NY). Primary endpoints were inhospital mortality and follow-up death from any cause. Secondary endpoints were the incidence of major complications and the composite of major adverse events (inhospital death and major postoperative complications). Results are expressed as frequency and percentage for categoric variables and as mean and standard deviation for continuous variables. Data from the study population were compared using c2 for categoric variables and the Student t test for continuous variables. A p value of less than 0.05 was considered significant. Survival estimates were generated utilizing the Kaplan-Meier method.
Results Patient characteristics are listed in Table 1. The cohort of 14 patients consisted of 9 patients (64%) who had descending thoracic and 5 patients (36%) who had thoracoabdominal mycotic aneurysms. All patients presented with either aneurysm-related back pain or a clinical picture of sepsis. Diagnosis was confirmed utilizing computed tomography imaging. Mean age was 66 13 years, and mean aneurysm size was 5.9 1.3 cm. All patients were hypertensive. Twelve patients (85.7%) had aneurysm-related back pain, of whom 9 (64.3%) had contained rupture. Two patients (14.3%) had clinical sepsis and radiographic findings but no aneurysmrelated pain. On preoperative blood cultures, 4 (28.6%) grew Staphylococcus aureus, 4 (28.6%) gram negative organisms, and 2 (14.3%) Mycobacterium organisms. Four (28.6%) had negative culture results. Empiric broad-spectrum antibiotics were initiated on all patients and adjusted based on final cultures. Eleven patients received preoperative extended-spectrum penicillin (nafcillin, oxacillin, piperacillin/tazobactam, or ampicillin/sulbactam) for an average of 5 4 days. Vancomycin was added for 1 penicillin-allergic patient and standard four-drug antituberculosis therapy was given to 1 patient with Mycobacterium tuberculosis. Three patients (21%) were taken immediately to the operating room owing to hemodynamic instability and given their first dose of antibiotics
ADULT CARDIAC
Ann Thorac Surg 2015;100:1712–7
ADULT CARDIAC
1714
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
Ann Thorac Surg 2015;100:1712–7
Table 1. Preoperative Patient Demographics (n ¼ 14) Preoperative Demographics Age, years Male Fever White blood cells, 103/mL Aneurysm characteristics Aneurysm size, cm Symptoms Rupture Arterial occlusive disease Celiac axis Superior mesenteric artery Peripheral arterial disease Marfan syndrome Previous TEVAR Previous aneurysm surgery Comorbidities Ejection fraction, % Hypertension Prior coronary revascularization Percutaneous coronary intervention Coronary artery bypass graft Prior cerebrovascular accident Chronic pulmonary disease Diabetes mellitus Renal disease Chronic renal insufficiencya End-stage renal disease Current or former smoker a
Table 2. Bacteriology Values
66 13 8 (57.1) 6 (42.9) 12.2 4.4 5.9 1.3 12 (85.7) 9 (64) 1 1 5 1 2 2
(7.1) (7.1) (35.7) (7.1) (14.3) (14.3)
51 5 14 (100) 3 (21.4) 2 (14.3) 1 (7.1) 2 (14.3) 7 (50) 5 (35.7) 2 (14.3) 2 (14.3) 11 (78.6)
Creatinine 1.5 to 3 mg/dL.
Values are mean SD or n (%). TEVAR ¼ thoracic endovascular aneurysm repair.
before incision. The remaining bacteriology is outlined in Table 2. Patients whose blood cultures were positive preoperatively grew the same organism from the aortic wall cultures, except for 1 patient with M tuberculosis and 1 with Salmonella organisms, both of whom had negative results of operative cultures. Details of the operative procedures are listed in Table 3. A majority of patients underwent repair utilizing a clampand-sew technique (10 of 14, 71.4%). The remainder (4 of 14, 28.6%) had distal arch involvement and required repair under deep hypothermic circulatory arrest. Cerebrospinal fluid drain was used in 9 patients (64%). Of the remaining 5 patients, 3 were too unstable to permit placement and 2 had several failed attempts at placement. After radical debridement of the infected tissue, Dacron grafts were placed in situ in the normal anatomic position. Six patients (42.9%) had additional soft tissue coverage, 5 (35.7%) utilizing an omental flap and 1 (7.1%) a serratus muscle flap. There were no deaths within 30 days of the initial procedures. There was 1 (7.1%) inhospital death secondary to ischemic bowel. That patient had presented with
Bacteriology Citrobacter koseri Escherichia coli MSSA Mycobacterium tuberculosis Pseudomonas aeruginosa Salmonella enteritidis Staphylococcus epidermidis Streptococcus pneumoniae None
Preoperative Cultures 1 1 4 2
(7.1) (7.1) (28.6) (14.3)
Operative Cultures 1 1 4 1
(7.1) (7.1) (28.6) (7.1)
1 (7.1) 1 (7.1) 0 (0)
1 (7.1) 0 (0) 1 (7.1)
0 (0) 4 (28.6)
2 (14.3) 3 (21.4)
Values are n (%). MSSA ¼ methicillin-sensitive Staphylococcus aureus.
mesenteric vessel involvement, and despite revascularization, bowel necrosis developed and death occurred on postoperative day 45. Four patients (28.6%) required tracheostomy. One (7.1%) had temporary left recurrent nerve palsy. None of the patients incurred spinal cord injury, stroke, or new onset renal failure requiring dialysis. Univariate analysis revealed that New York Heart Association functional class, diabetes mellitus, and preoperative renal dysfunction were preoperative risk factors for major adverse events (Table 4). After surgery, all patients were given 6 weeks of intravenous antibiotics. Lifelong suppression therapy was maintained with oral antibiotics, tailored to sensitivities from either blood or intraoperative cultures. One patient was lost to follow-up. Mean follow-up time of the remaining 12 patients was 26.5 months. Three patients died during follow-up. One patient presented 2 months postoperatively with hemoptysis and
Table 3. Operative Details Operative Details Emergent operation Concomittant procedures Omental flap Serratus muscle flap Celiac bypass Superior mesenteric artery bypass Splenectomy Circulatory arrest Cardiopulmonary bypass time, minutes Circulatory arrest time, minutes Clamp-and-sew Cross-clamp time, minutes Cerebrospinal fluid drain PRBC transfusion, units Values are n (%) or mean SD. PRBC ¼ packed red blood cells.
Values 12 (85.7) 5 (35.7) 1 (7.1) 1 (7.1) 1 (7.1) 2 (14.3) 4 (28.6) 126.8 26.3 33.8 6.9 10 (71.4) 27.2 7.5 9 (64.3) 3.4 2.1
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
1715
Table 4. Risk Factors for Perioperative Major Adverse Event or Death Risk Factors Age, years Aneurysm size, cm Time from illness to surgery, days New York Heart Association class I II III Chronic pulmonary disease Diabetes mellitus Renal disease Chronic renal insufficiency Dialysis Previous cerebrovascular accident Smoking history Rupture Fever White blood cells, 103/uL Cultures MSSA Gram positive Gram negative Concomitant procedure Cerebrospinal fluid drainage Circulatory arrest Intercostal ischemic time, minutes a
Major Complication/ Death (n ¼ 2)
No Major Complication (n ¼ 12)
p Value
67 0 6.1 1.6 3.5 4.9
65 14.18 5.8 1.3 41.4 45.3
0.881 0.808 0.275
1 0 1 1 2
(50) (0) (50) (50) (100)
9 3 0 6 3
(75) (25) (0) (50) (25)
0 (0) 2 (100) 1 (50) 2 (100) 2 (100) 2 (100) 11.1 6.93
2 (16.7) 0 (0) 1 (8.3) 9 (75) 7 (58.3) 4 (33.3) 12.41 4.48
2 (100) 0 (0) 0 (0) 1 (50) 1 (50) 1 (50) 31.0 9.9
3 (25) 4 (33.3) 4 (33.3) 5 (41.7) 8 (66.7) 3 (25) 26.17 11.7
0.036a
1.000 0.040a 0.001a 0.119 0.425 0.225 0.078 0.723 0.241
0.825 0.649 0.469 0.594
Significant.
Values are mean SD or n (%). MSSA ¼ methicillin-sensitive Staphylococcus Aureus.
hyperkalemia degenerating into cardiac arrest, from which he did not recover. Autopsy revealed that the aortic repair was intact. One patient died in a rehabilitation facility at 3.5 months of unclear cause but with no clinical signs of sepsis. Another died at 142 months of unknown cause. No patients had a prosthetic graft infection. Actuarial 5-year survival after surgery was 71% (SD 2.0, 95% confidence interval: 4.6 to 12.5), as seen in Figure 1.
atherosclerotic ulcer. Aortic infection may also begin in the adventitial layer of the aorta as the vaso vasorum are seeded from infection in adjacent organs such as the lung,
Comment Mycotic aortic aneurysms are rare entities and comprise a minute proportion of all aneurysms, both abdominal and thoracic. One third originate in the thoracic aorta, and many theories have been put forth to explain the etiology of this highly lethal form of aortic pathology [1–3]. Bacteremia in the setting of a preexisting aneurysm is thought to be the most common cause of mycotic aortic degeneration, as seeding of either disrupted aortic intima or mural thrombus can lead to an aggressive local infection with rapid aortic expansion and rupture. Bacterial endocarditis may also contribute to the formation of MTAA as a result of emboli seeding a defect in the aortic intima at the site of a ruptured plaque or penetrating
Fig 1. Kaplan-Meier survival estimate. Solid line indicates survival function; hatch marks indicate censored. (Cum ¼ cumulative.)
ADULT CARDIAC
Ann Thorac Surg 2015;100:1712–7
ADULT CARDIAC
1716
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
esophagus, chest wall, or pleural cavity. Finally, cases of MTAA have been described in the setting of penetrating trauma [7]. Although the etiology of MTAA is similar across geographical distributions, the incidence and microbiology is quite different throughout the world, and the region of origin must be considered when confronted with a patient with a presumed MTAA. In the United States, MTAA remains quite rare, representing less than 1% of all aneurysm repairs even in high-volume aortic centers [2]. In the East, however, MTAA is much more common and is responsible for more than 3% of all aneurysm repairs in Asia [8]. The microbiology of MTAA is quite variable, and the location and ethnicity of the patient must be considered when choosing empiric antibiotic therapy. Similar to other series originating in the United States where gram-positive species predominate [1, 2, 7], our series found that 50% of our patients grew either streptococcal or staphylococcal species from either their blood or aortic cultures. Gram negative bacteria are not as prevalent here in the Western world—we found these species in fewer than one third of our patients. However gram negative species may be responsible for as much as three quarters of all MTAA in countries such as Taiwan [8]. In particular, Salmonella, a species thought to be a less virulent organism, is quite common in the Asian world [8, 9, 11]. Its virulence is underscored by 1 of our patients who presented with Salmonella bacteremia after a brief trip to Korea. He subsequently presented with a MTAA in the proximal descending aorta nearly 5 months after returning from his visit. In our series, preoperative blood cultures correlated well with intraoperative tissue cultures in the majority of patients. Only 1 patient had negative cultures in both the blood and tissue. For the other 13 patients, we were able to narrow the spectrum of antibiotic coverage based on levels of sensitivity from the blood or tissue cultures. Patients were given 6 weeks of intravenous antibiotics, with 9 of 14 patients receiving an extended-spectrum penicillin. After 6 weeks, they were switched to the narrowest spectrum oral antibiotic that demonstrated acceptable sensitivity levels on blood and tissue cultures—most commonly trimethoprim/sulfamethoxazole, cephalexin, doxycycline, or metronidazole. This life-long oral antibiotic regimen is thought to provide low-level suppression of bacterial growth. Given the indolent nature of some of the offending organisms such as Mycobacterium, it is not uncommon for patients with MTAA to present with negative cultures. Two of our patients (14.3%) never grew bacteria in either their preoperative blood or intraoperative aortic cultures—this despite a number of nuclear scans and PET scans suggesting the presence of infection. One could speculate that these 3 patients had a pathologic diagnosis more consistent with an inflammatory aneurysm [18]. However, the sensitivity and specificity of indium and PET scanning is quite accurate with respect to infection. Indium scan has been found to have a sensitivity of 73% and specificity of 87% [19]; PET has been found to have a sensitivity of 91% and specificity of
Ann Thorac Surg 2015;100:1712–7
64% for detecting aortic graft infections [20]. Not unlike subacute bacterial endocarditis, the administration of broad-spectrum antibiotics at the time of the initial presentation may obscure the ability of a microbiology laboratory to properly identify the organism responsible for the patient’s death. In these instances, securing the diagnosis can be quite difficult, and the entirety of the patient’s presentation (clinical picture, microbiology, imaging, nuclear studies, and laboratory values) must be examined to proceed with therapy and avoid catastrophic rupture. This conundrum is highlighted by the patients we present in this series. A majority presented to outside hospitals with prolonged prodromal illnesses. The mean time from the onset of illness to discovery of the MTAA was 36 days. That led to a distressingly high number, 64%, of our patients presenting to our institution with rupture. Although we cannot estimate the number of patients with MTAA who present with fatal rupture, we are gratified that the mortality in this series, 7%, is in line with our previously published series on ruptured thoracic aneurysms [21]. The mortality associated with open repair of MTAA is certainly not insignificant, but without question is superior to medical therapy alone. Numerous investigators have reported 30-day mortality in excess of 45% when patients with MTAA are treated with sole medical therapy [8, 14, 22, 23]. Surgical intervention is the primary therapy of choice for MTAA and should be considered for any patient diagnosed with MTAA. Weis-Muller and associates [7] published one of the larger series of mycotic aortic aneurysms treated with open surgical repair. Their reported 30-day mortality of 33% compares favorably with antibiotic therapy alone [7]. However, given that 50% of the patients in this series had infected abdominal aneurysms, it is difficult to extract the exact risk for those with MTAA. Although others have reported even lower mortality rates of 3% to 21% for open repair of mycotic aneurysms, all those reports are limited by the inclusion of both mycotic abdominal aortic and iliac artery aneurysms [2, 8]. One has to look to more contemporary series of endovascular MTAA repair to appreciate the efficacy of open surgery for this high-risk subset of thoracic aneurysms. Semba and colleagues [24] were the first to report on the successful treatment of MTAA with TEVAR. Although the initial results were acceptable, the number of patients treated (n ¼ 3) was quite small, and follow-up after the endovascular repair was not reported. Patel and associates [14] reported a more real world examination of the role of TEVAR for MTAA in their report from 2010. The perioperative mortality was 11.5% for the 27 patients treated in this fashion. Disappointingly, 18.5% had significant endoleaks, and an additional 14% of patients had late death due to ongoing sepsis. Freedom from reinfection was 81.2% at 2 years. Endovascular exclusion of an infected aneurysm with TEVAR may indeed temporarily protect extremely ill patients from fatal rupture while their multiple organ dysfunction resolves. This “staged” approach to MTAA repair, while a very expensive option, may be appropriate in this setting [25, 26].
Our results suggest, however, that primary, aggressive open surgical debridement and in-situ replacement of the infected aorta is not only safe but durable and should remain the standard treatment for MTAA. In experienced hands, we believe that mortality less than 10% is achievable. Even in a grossly infected field, radical debridement and coverage of a Dacron graft with either muscle flap or omentum reliably relieved patients of their infectious burden while restoring aortic continuity. Even though our cohort included patients on hemodialysis and a majority presented with rupture, open repair was reproducible without an expensive intermediate stage of endovascular intervention. None of our patients had reinfection for a mean of 26 months after surgery. The 71% actuarial 5-year survival speaks well for the success of this more traditional approach to the high-risk patient with MTAA. The limitations of this study include its retrospective, observational nature and small sample size. Given the relative rarity of this disease, however, this report is not markedly different from other reports on MTAA. We also concede that results from a high-volume aortic center with significant experience in open thoracic aneurysm repair may not translate well to every instance in which a patient with MTAA may present. In conclusion, open repair of mycotic thoracic and thoracoabdominal aortic aneurysms remains the most effective therapy for this very high risk cohort of patients presenting with infected aneurysms. In experienced hands, with the utilization of contemporary end organ protection strategies, a gratifyingly high rate of salvage can be expected, even for patients presenting with rupture. In cases of gross contamination of the pleural or abdominal cavities, or in cases of extensive infection of surrounding structures, tissue coverage of the in situ prosthesis may enhance recovery and contribute to the low rate of reinfection seen with this approach. The role of lifelong antibiotic suppression therapy certainly warrants further investigation, but is not unreasonable given the fatal outcome of patients with ongoing infection who have been treated in a less aggressive fashion.
References 1. Usui A. Surgical management of infected thoracic aneurysms. Nagoya J Med Sci 2013;75:161–7. 2. Oderich GS, Panneton JM, Bower TC, et al. Infected aortic aneurysms: aggressive presentation, complicated early outcome but durable results. J Vasc Surg 2001;34:900–8. 3. Muller BT, Wegener OR, Grabitz K, Pillny M, Thomas L, Sandmann W. Mycotic aneurysm of the thoracic and abdominal aorta and iliac arteries: experience with anatomic and extra-anatomic repair in 33 cases. J Vasc Surg 2001;33: 106–13. 4. Bisdas T, Wilhelmi M, Haverich A, Teebken OE. Cryopreserved arterial homografts vs silver-coated Dacron grafts for abdominal aortic infections with intraoperative evidence of microorganisms. J Vasc Surg 2011;53:1274–81. 5. Harlander-Locke MP, Harmon LK, Lawrence PF, et al. The use of cryopreserved aortoiliac allograft for aortic reconstruction in the United States. J Vasc Surg 2014;59:669–74.
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
1717
6. Hsu RB, Chen RJ, Wang SS, Chu SH. Infected aortic aneurysms: clinical outcome and risk factor analysis. J Vasc Surg 2004;40:30–5. 7. Weis-M€ uller BT, Rascanu C, Sagban A, Grabitz K, Godehardt E, Sandmann W. Single-center experience with open surgical treatment of 36 infected aneurysms of the thoracic, thoracoabdominal, and abdominal aorta. Ann Vasc Surg 2011;25:1020–5. 8. Hsu RB, Tsay YG, Wang SS, Chu SH. Surgical treatment for primary infected aneurysm of the descending thoracic aorta, abdominal aorta, and iliac arteries. J Vasc Surg 2002;36: 746–50. 9. Huang YK, Ko PJ, Chen CL, et al. Therapeutic opinion on endovascular repair for mycotic aortic aneurysm. Ann Vasc Surg 2014;28:579–89. 10. Jia X, Dong YF, Liu XP, Xiong J, Zhang HP, Guo W. Open and endovascular repair of primary mycotic aortic aneurysms: a 10-year single-center experience. J Endovasc Ther 2013;20: 305–10. 11. Kan CD, Lee HL, Yang YJ. Outcome after endovascular stent graft treatment for mycotic aortic aneurysm: a systemic review. J Vasc Surg 2007;46:906–12. 12. Ting AC, Cheng SW, Ho P, Poon JT, Tsu JH. Surgical treatment of infected aneurysms and pseudoaneurysms of the thoracic and abdominal aorta. Am J Surg 2005;189:150–4. 13. Johnstone JK, Slaiby JM, Marcaccio EJ, Chong TT, GarciaToca M. Endovascular repair of mycotic aneurysm of the descending thoracic aorta. Ann Vasc Surg 2013;27: 23–8. 14. Patel HJ, Williams DM, Upchurch GR, Dasika NL, Eliason JL, Deeb GM. Thoracic aortic endovascular repair for mycotic aneurysms and fistulas. J Vasc Surg 2010;52(Suppl):37–40. 15. Sedivy P, Spacek M, El Samman K, et al. Endovascular treatment of infected aortic aneurysms. Eur J Vasc Endovasc Surg 2012;44:385–94. 16. S€ orelius K, Mani K, Bj€ orck M, et al. Endovascular treatment of mycotic aortic aneurysms: a European multicenter study. Circulation 2014;130:2136–42. 17. Girardi LN, Rabotnikov Y, Avgerinos DV. Preoperative percutaneous coronary intervention in patients undergoing open thoracoabdominal and descending thoracic aneurysm repair. J Thorac Cardiovasc Surg 2014;147:163–8. 18. Girardi LN, Coselli JS. Inflammatory aneurysm of the ascending aorta and aortic arch. Ann Thorac Surg 1997;64: 251–3. 19. Shahidi S, Eskil A, Lundof E, Klaerke A, Jensen BS. Detection of abdominal aortic graft infection: comparison of magnetic resonance imaging and indium-labeled white blood cell scanning. Ann Vasc Surg 2007;21:586–92. 20. Fukuchi K, Ishida Y, Higashi M, et al. Detection of aortic graft infection by fluorodeoxyglucose positron emission tomography: comparison with computed tomographic findings. J Vasc Surg 2005;42:919–25. 21. Girardi LN, Krieger KH, Altorki NK, Mack CA, Lee LY, Isom OW. Ruptured descending and thoracoabdominal aortic aneurysms. Ann Thorac Surg 2002;74:1066–70. 22. Chan FY, Crawford ES, Coselli JS, et al. In situ prosthetic graft replacement for mycotic aneurysm of the aorta. Ann Thorac Surg 1989;47:193–203. 23. Fillmore AJ, Valentine RJ. Surgical mortality in patients with infected aortic aneurysms. J Am Coll Surg 2003;196:435–41. 24. Semba CP, Sakai T, Slonim SM, et al. Mycotic aneurysms of the thoracic aorta: repair with use of endovascular stentgrafts. J Vasc Interv Radiol 1998;9:33–40. 25. Clough RE, Black SA, Lyons OT, et al. Is endovascular repair of mycotic aortic aneurysms a durable treatment option? Eur J Vasc Endovasc Surg 2009;37:407–12. 26. Fukunaga N, Hashimoto T, Ozu Y, et al. Successful treatment for infected aortic aneurysm using endovascular aneurysm repairs as a bridge to delayed open surgery. Ann Vasc Surg 2012;26:280.
ADULT CARDIAC
Ann Thorac Surg 2015;100:1712–7
Outcomes of Open Repair of Mycotic Descending Thoracic and Thoracoabdominal Aortic Aneurysms Christopher Lau, MD, Mario Gaudino, MD, Andreas R. de Biasi, MD, Monica Munjal, MS, and Leonard N. Girardi, MD Department of Cardiothoracic Surgery, Weill Cornell Medical College, New York, New York
Background. The purpose of this study was to evaluate the short- and intermediate-term outcomes of open repair of mycotic thoracic and thoracoabdominal aneurysms. Contemporary surgical and perioperative techniques were utilized. Methods. From November 1997 to May 2014, 14 consecutive patients underwent open repair of descending thoracic (n [ 9, 64.3%) and thoracoabdominal (n [ 5, 35.7%) mycotic aortic aneurysms. All procedures were performed through the left side of the chest. Infected tissue was completely debrided and excised. Aortic continuity was restored in situ with a Dacron prosthesis (Macquet Corp, Oakland, NJ). Soft tissue coverage of the prosthesis was performed when anatomy and patient condition permitted. Perioperative outcomes, intermediate-term survival, and reinfection rates were examined. Results. All patients presented with either aneurysmrelated symptoms or a clinical picture of sepsis. Diagnosis was confirmed utilizing computed tomography imaging. Mean age was 66 ± 13 years, 8 patients (57.1%) were male, and mean aneurysm size was 5.9 ± 1.3 cm. All patients were hypertensive, 3 (21.4%) had prior coronary revascularization, 7 (50%) had chronic pulmonary disease, 5 (35.7%) had diabetes mellitus, and 2 (14.3%) had end-stage renal disease requiring dialysis. Twelve patients (85.7%) had aneurysm-related pain, and 9 (64.3%) of them had contained rupture. Mean time from onset of illness to surgery was 36 days (range, 0 to 153). On preoperative blood cultures, 4 (28.6%) grew Staphylococcus aureus, 4 (28.6%) grew gram negative organisms, 2 (14.3%) grew mycobacterium, and 4 cultures (28.6%) had negative results. Empiric broad-spectrum antibiotics were initiated
on all patients and adjusted based on final cultures. A majority of patients underwent repair utilizing a clamp-and-sew technique (n [ 10, 71.4%); the remainder (n [ 4, 28.6%) required repair under profound hypothermic circulatory arrest. After radical debridement of the infected tissue, grafts were placed in the normal anatomic position; 6 (42.9%) patients had additional soft tissue coverage, 5 (35.7%) utilizing an omental flap and 1 (7.1%), a serratus muscle flap. There was 1 in-hospital death (7.1%) secondary to ischemic bowel. Four patients (28.6%) required tracheostomy, and 1 (7.1%) had recurrent nerve injury. None of the patients incurred spinal cord injury, stroke, or new onset renal failure requiring dialysis. After surgery, all patients were given 6 weeks of intravenous antibiotics. Lifelong suppression therapy was maintained with oral antibiotics. There were no episodes of prosthetic graft infection on follow-up. Univariate analysis revealed that New York Heart Association functional class, diabetes, and preoperative renal dysfunction were preoperative risk factors for major adverse events. Mean follow-up time was 26.5 months (median 8.2; range, 1 to 142). Actuarial 5-year survival was 71%. Conclusions. Open repair of mycotic descending thoracic and thoracoabdominal aortic aneurysms remains the gold standard of therapy. Aggressive intraoperative debridement with in situ prosthetic reconstruction permits a high rate of success in this very high risk cohort of patients. Lifelong antibiotic suppression therapy may prevent late prosthetic graft infection.
M
resection to rid the patient of infection while reestablishing aortic continuity with either a prosthetic graft or a homograft [1–6]. Owing to the rarity of MTAA, data on treatment are limited to small series and case reports. In addition, much of the existing literature combines MTAA with mycotic abdominal aortic aneurysms, which are managed quite differently and are associated with vastly different postoperative complications [2, 7]. The majority of recent literature comes from Asia, where the incidence of mycotic aneurysmal disease is much higher and the microbiology much different, with a predominance of Salmonella species being responsible for this potentially
ycotic thoracic aortic aneurysms (MTAA) are an unusual variant of aneurysmal disease, comprising less than 1% of all aortic aneurysms [1, 2]. These rare subtypes of aneurysm are quite lethal, with a high incidence of fatal rupture when medical therapy alone is attempted [1, 3]. Although there are no standard guidelines for treatment, the most common practice is to combine directed antibiotic therapy with surgical
Accepted for publication May 14, 2015. Address correspondence to Dr Lau, 525 E 68th St, Box 110, Ste M404, New York, NY 10065; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;100:1712–7) Ó 2015 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.05.067
lethal disease [8–12]. The treatment algorithm for MTAA has also undergone a paradigm shift in the last decade. Traditional therapy consists of open graft replacement of the infected aneurysm, oftentimes supported with muscle flap or omental coverage of the replaced segment. However, thoracic endovascular aneurysm repair (TEVAR) has now supplanted open surgery as the most common type of initial treatment strategy [9–11, 13–15]. TEVAR has reduced the incidence of fatal rupture of the presenting disease, as seen in a recent large multicenter study with a 91% 1-month survival [16]. However, longer-term outcomes with TEVAR as sole therapy are questionable when implanting prosthetic material into an infected field without debridement or excision of the diseased aorta. We sought to examine the outcomes of an aggressive open surgery treatment strategy for MTAA. A combination of traditional surgical debridement and in situ aortic replacement was utilized with a variety of contemporary end organ protection strategies. Short- and intermediateterm outcomes are examined, and a reproducible treatment algorithm is proposed.
Material and Methods This study was approved by the Weill Cornell Medical College Institutional Review Board. The need for individual patient consent was waived. We reviewed prospectively collected data from the Weill Cornell Medical College Department of Cardiothoracic Surgery aortic surgery database to identify patients who had repair of MTAA. Aneurysms were defined as mycotic if they were associated with clinical signs of infection, characteristic appearance of radiologic imaging or intraoperative examination, and positive cultures from either blood or aortic tissue. Patients presenting with infected aneurysms of the ascending aorta or arch that were approached through a median sternotomy were excluded. From November 1997 to May 2014, a total of 680 patients underwent open repair of either descending thoracic aneurysms or thoracoabdominal aortic aneurysms (TAAA). Fourteen were mycotic, with 9 descending thoracic aneurysms (9 of 14, 64%) and 5 TAAAs (5 of 14, 36%). TEVAR was not used for any patients with MTAA. All procedures were performed through either a left thoracotomy or left thoracoabdominal incision. Infected tissue was completely excised, and aortic continuity restored with a Dacron prosthesis (Macquet Corp, Oakland, NJ). Soft tissue coverage of the prosthesis was performed when anatomy and patient condition permitted. Standard methods for end organ and spinal cord protection were utilized, and have been described previously [17], including ambient cooling to 33 C to 34 C, cerebrospinal fluid drainage when clinically stable, and an open distal anastomotic technique. Because of the need for aggressive tissue debridement, intercostal arteries were not reimplanted in any patients. Postoperative mean arterial pressures were maintained at 85 to 90 mm Hg for 3 days using a combination of aggressive fluid resuscitation, norepinephrine, and
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
1713
vasopressin infusions. Patients were initially given broad-spectrum antibiotics followed by directed antibiotic coverage after identification and sensitivities of the inciting organisms were obtained. Intravenous antibiotics were administered for 6 weeks after surgery followed by life-long oral antibiotic suppression therapy. Patients surviving surgery were then followed with serial computed tomography imaging. Patients with either radiographic or clinical evidence to suggest possible reinfection were evaluated with nuclear imaging utilizing indium or positron emission tomography (PET) scan to confirm or negate the diagnosis, although no patients had reinfection of the prosthetic graft.
Statistics Data were stored using Microsoft Access (2013) software (Microsoft, Redmond, WA). Processing and analysis of the data were performed using IBM SPSS Statistics, version 20 (IBM Corp, Armonk, NY). Primary endpoints were inhospital mortality and follow-up death from any cause. Secondary endpoints were the incidence of major complications and the composite of major adverse events (inhospital death and major postoperative complications). Results are expressed as frequency and percentage for categoric variables and as mean and standard deviation for continuous variables. Data from the study population were compared using c2 for categoric variables and the Student t test for continuous variables. A p value of less than 0.05 was considered significant. Survival estimates were generated utilizing the Kaplan-Meier method.
Results Patient characteristics are listed in Table 1. The cohort of 14 patients consisted of 9 patients (64%) who had descending thoracic and 5 patients (36%) who had thoracoabdominal mycotic aneurysms. All patients presented with either aneurysm-related back pain or a clinical picture of sepsis. Diagnosis was confirmed utilizing computed tomography imaging. Mean age was 66 13 years, and mean aneurysm size was 5.9 1.3 cm. All patients were hypertensive. Twelve patients (85.7%) had aneurysm-related back pain, of whom 9 (64.3%) had contained rupture. Two patients (14.3%) had clinical sepsis and radiographic findings but no aneurysmrelated pain. On preoperative blood cultures, 4 (28.6%) grew Staphylococcus aureus, 4 (28.6%) gram negative organisms, and 2 (14.3%) Mycobacterium organisms. Four (28.6%) had negative culture results. Empiric broad-spectrum antibiotics were initiated on all patients and adjusted based on final cultures. Eleven patients received preoperative extended-spectrum penicillin (nafcillin, oxacillin, piperacillin/tazobactam, or ampicillin/sulbactam) for an average of 5 4 days. Vancomycin was added for 1 penicillin-allergic patient and standard four-drug antituberculosis therapy was given to 1 patient with Mycobacterium tuberculosis. Three patients (21%) were taken immediately to the operating room owing to hemodynamic instability and given their first dose of antibiotics
ADULT CARDIAC
Ann Thorac Surg 2015;100:1712–7
ADULT CARDIAC
1714
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
Ann Thorac Surg 2015;100:1712–7
Table 1. Preoperative Patient Demographics (n ¼ 14) Preoperative Demographics Age, years Male Fever White blood cells, 103/mL Aneurysm characteristics Aneurysm size, cm Symptoms Rupture Arterial occlusive disease Celiac axis Superior mesenteric artery Peripheral arterial disease Marfan syndrome Previous TEVAR Previous aneurysm surgery Comorbidities Ejection fraction, % Hypertension Prior coronary revascularization Percutaneous coronary intervention Coronary artery bypass graft Prior cerebrovascular accident Chronic pulmonary disease Diabetes mellitus Renal disease Chronic renal insufficiencya End-stage renal disease Current or former smoker a
Table 2. Bacteriology Values
66 13 8 (57.1) 6 (42.9) 12.2 4.4 5.9 1.3 12 (85.7) 9 (64) 1 1 5 1 2 2
(7.1) (7.1) (35.7) (7.1) (14.3) (14.3)
51 5 14 (100) 3 (21.4) 2 (14.3) 1 (7.1) 2 (14.3) 7 (50) 5 (35.7) 2 (14.3) 2 (14.3) 11 (78.6)
Creatinine 1.5 to 3 mg/dL.
Values are mean SD or n (%). TEVAR ¼ thoracic endovascular aneurysm repair.
before incision. The remaining bacteriology is outlined in Table 2. Patients whose blood cultures were positive preoperatively grew the same organism from the aortic wall cultures, except for 1 patient with M tuberculosis and 1 with Salmonella organisms, both of whom had negative results of operative cultures. Details of the operative procedures are listed in Table 3. A majority of patients underwent repair utilizing a clampand-sew technique (10 of 14, 71.4%). The remainder (4 of 14, 28.6%) had distal arch involvement and required repair under deep hypothermic circulatory arrest. Cerebrospinal fluid drain was used in 9 patients (64%). Of the remaining 5 patients, 3 were too unstable to permit placement and 2 had several failed attempts at placement. After radical debridement of the infected tissue, Dacron grafts were placed in situ in the normal anatomic position. Six patients (42.9%) had additional soft tissue coverage, 5 (35.7%) utilizing an omental flap and 1 (7.1%) a serratus muscle flap. There were no deaths within 30 days of the initial procedures. There was 1 (7.1%) inhospital death secondary to ischemic bowel. That patient had presented with
Bacteriology Citrobacter koseri Escherichia coli MSSA Mycobacterium tuberculosis Pseudomonas aeruginosa Salmonella enteritidis Staphylococcus epidermidis Streptococcus pneumoniae None
Preoperative Cultures 1 1 4 2
(7.1) (7.1) (28.6) (14.3)
Operative Cultures 1 1 4 1
(7.1) (7.1) (28.6) (7.1)
1 (7.1) 1 (7.1) 0 (0)
1 (7.1) 0 (0) 1 (7.1)
0 (0) 4 (28.6)
2 (14.3) 3 (21.4)
Values are n (%). MSSA ¼ methicillin-sensitive Staphylococcus aureus.
mesenteric vessel involvement, and despite revascularization, bowel necrosis developed and death occurred on postoperative day 45. Four patients (28.6%) required tracheostomy. One (7.1%) had temporary left recurrent nerve palsy. None of the patients incurred spinal cord injury, stroke, or new onset renal failure requiring dialysis. Univariate analysis revealed that New York Heart Association functional class, diabetes mellitus, and preoperative renal dysfunction were preoperative risk factors for major adverse events (Table 4). After surgery, all patients were given 6 weeks of intravenous antibiotics. Lifelong suppression therapy was maintained with oral antibiotics, tailored to sensitivities from either blood or intraoperative cultures. One patient was lost to follow-up. Mean follow-up time of the remaining 12 patients was 26.5 months. Three patients died during follow-up. One patient presented 2 months postoperatively with hemoptysis and
Table 3. Operative Details Operative Details Emergent operation Concomittant procedures Omental flap Serratus muscle flap Celiac bypass Superior mesenteric artery bypass Splenectomy Circulatory arrest Cardiopulmonary bypass time, minutes Circulatory arrest time, minutes Clamp-and-sew Cross-clamp time, minutes Cerebrospinal fluid drain PRBC transfusion, units Values are n (%) or mean SD. PRBC ¼ packed red blood cells.
Values 12 (85.7) 5 (35.7) 1 (7.1) 1 (7.1) 1 (7.1) 2 (14.3) 4 (28.6) 126.8 26.3 33.8 6.9 10 (71.4) 27.2 7.5 9 (64.3) 3.4 2.1
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
1715
Table 4. Risk Factors for Perioperative Major Adverse Event or Death Risk Factors Age, years Aneurysm size, cm Time from illness to surgery, days New York Heart Association class I II III Chronic pulmonary disease Diabetes mellitus Renal disease Chronic renal insufficiency Dialysis Previous cerebrovascular accident Smoking history Rupture Fever White blood cells, 103/uL Cultures MSSA Gram positive Gram negative Concomitant procedure Cerebrospinal fluid drainage Circulatory arrest Intercostal ischemic time, minutes a
Major Complication/ Death (n ¼ 2)
No Major Complication (n ¼ 12)
p Value
67 0 6.1 1.6 3.5 4.9
65 14.18 5.8 1.3 41.4 45.3
0.881 0.808 0.275
1 0 1 1 2
(50) (0) (50) (50) (100)
9 3 0 6 3
(75) (25) (0) (50) (25)
0 (0) 2 (100) 1 (50) 2 (100) 2 (100) 2 (100) 11.1 6.93
2 (16.7) 0 (0) 1 (8.3) 9 (75) 7 (58.3) 4 (33.3) 12.41 4.48
2 (100) 0 (0) 0 (0) 1 (50) 1 (50) 1 (50) 31.0 9.9
3 (25) 4 (33.3) 4 (33.3) 5 (41.7) 8 (66.7) 3 (25) 26.17 11.7
0.036a
1.000 0.040a 0.001a 0.119 0.425 0.225 0.078 0.723 0.241
0.825 0.649 0.469 0.594
Significant.
Values are mean SD or n (%). MSSA ¼ methicillin-sensitive Staphylococcus Aureus.
hyperkalemia degenerating into cardiac arrest, from which he did not recover. Autopsy revealed that the aortic repair was intact. One patient died in a rehabilitation facility at 3.5 months of unclear cause but with no clinical signs of sepsis. Another died at 142 months of unknown cause. No patients had a prosthetic graft infection. Actuarial 5-year survival after surgery was 71% (SD 2.0, 95% confidence interval: 4.6 to 12.5), as seen in Figure 1.
atherosclerotic ulcer. Aortic infection may also begin in the adventitial layer of the aorta as the vaso vasorum are seeded from infection in adjacent organs such as the lung,
Comment Mycotic aortic aneurysms are rare entities and comprise a minute proportion of all aneurysms, both abdominal and thoracic. One third originate in the thoracic aorta, and many theories have been put forth to explain the etiology of this highly lethal form of aortic pathology [1–3]. Bacteremia in the setting of a preexisting aneurysm is thought to be the most common cause of mycotic aortic degeneration, as seeding of either disrupted aortic intima or mural thrombus can lead to an aggressive local infection with rapid aortic expansion and rupture. Bacterial endocarditis may also contribute to the formation of MTAA as a result of emboli seeding a defect in the aortic intima at the site of a ruptured plaque or penetrating
Fig 1. Kaplan-Meier survival estimate. Solid line indicates survival function; hatch marks indicate censored. (Cum ¼ cumulative.)
ADULT CARDIAC
Ann Thorac Surg 2015;100:1712–7
ADULT CARDIAC
1716
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
esophagus, chest wall, or pleural cavity. Finally, cases of MTAA have been described in the setting of penetrating trauma [7]. Although the etiology of MTAA is similar across geographical distributions, the incidence and microbiology is quite different throughout the world, and the region of origin must be considered when confronted with a patient with a presumed MTAA. In the United States, MTAA remains quite rare, representing less than 1% of all aneurysm repairs even in high-volume aortic centers [2]. In the East, however, MTAA is much more common and is responsible for more than 3% of all aneurysm repairs in Asia [8]. The microbiology of MTAA is quite variable, and the location and ethnicity of the patient must be considered when choosing empiric antibiotic therapy. Similar to other series originating in the United States where gram-positive species predominate [1, 2, 7], our series found that 50% of our patients grew either streptococcal or staphylococcal species from either their blood or aortic cultures. Gram negative bacteria are not as prevalent here in the Western world—we found these species in fewer than one third of our patients. However gram negative species may be responsible for as much as three quarters of all MTAA in countries such as Taiwan [8]. In particular, Salmonella, a species thought to be a less virulent organism, is quite common in the Asian world [8, 9, 11]. Its virulence is underscored by 1 of our patients who presented with Salmonella bacteremia after a brief trip to Korea. He subsequently presented with a MTAA in the proximal descending aorta nearly 5 months after returning from his visit. In our series, preoperative blood cultures correlated well with intraoperative tissue cultures in the majority of patients. Only 1 patient had negative cultures in both the blood and tissue. For the other 13 patients, we were able to narrow the spectrum of antibiotic coverage based on levels of sensitivity from the blood or tissue cultures. Patients were given 6 weeks of intravenous antibiotics, with 9 of 14 patients receiving an extended-spectrum penicillin. After 6 weeks, they were switched to the narrowest spectrum oral antibiotic that demonstrated acceptable sensitivity levels on blood and tissue cultures—most commonly trimethoprim/sulfamethoxazole, cephalexin, doxycycline, or metronidazole. This life-long oral antibiotic regimen is thought to provide low-level suppression of bacterial growth. Given the indolent nature of some of the offending organisms such as Mycobacterium, it is not uncommon for patients with MTAA to present with negative cultures. Two of our patients (14.3%) never grew bacteria in either their preoperative blood or intraoperative aortic cultures—this despite a number of nuclear scans and PET scans suggesting the presence of infection. One could speculate that these 3 patients had a pathologic diagnosis more consistent with an inflammatory aneurysm [18]. However, the sensitivity and specificity of indium and PET scanning is quite accurate with respect to infection. Indium scan has been found to have a sensitivity of 73% and specificity of 87% [19]; PET has been found to have a sensitivity of 91% and specificity of
Ann Thorac Surg 2015;100:1712–7
64% for detecting aortic graft infections [20]. Not unlike subacute bacterial endocarditis, the administration of broad-spectrum antibiotics at the time of the initial presentation may obscure the ability of a microbiology laboratory to properly identify the organism responsible for the patient’s death. In these instances, securing the diagnosis can be quite difficult, and the entirety of the patient’s presentation (clinical picture, microbiology, imaging, nuclear studies, and laboratory values) must be examined to proceed with therapy and avoid catastrophic rupture. This conundrum is highlighted by the patients we present in this series. A majority presented to outside hospitals with prolonged prodromal illnesses. The mean time from the onset of illness to discovery of the MTAA was 36 days. That led to a distressingly high number, 64%, of our patients presenting to our institution with rupture. Although we cannot estimate the number of patients with MTAA who present with fatal rupture, we are gratified that the mortality in this series, 7%, is in line with our previously published series on ruptured thoracic aneurysms [21]. The mortality associated with open repair of MTAA is certainly not insignificant, but without question is superior to medical therapy alone. Numerous investigators have reported 30-day mortality in excess of 45% when patients with MTAA are treated with sole medical therapy [8, 14, 22, 23]. Surgical intervention is the primary therapy of choice for MTAA and should be considered for any patient diagnosed with MTAA. Weis-Muller and associates [7] published one of the larger series of mycotic aortic aneurysms treated with open surgical repair. Their reported 30-day mortality of 33% compares favorably with antibiotic therapy alone [7]. However, given that 50% of the patients in this series had infected abdominal aneurysms, it is difficult to extract the exact risk for those with MTAA. Although others have reported even lower mortality rates of 3% to 21% for open repair of mycotic aneurysms, all those reports are limited by the inclusion of both mycotic abdominal aortic and iliac artery aneurysms [2, 8]. One has to look to more contemporary series of endovascular MTAA repair to appreciate the efficacy of open surgery for this high-risk subset of thoracic aneurysms. Semba and colleagues [24] were the first to report on the successful treatment of MTAA with TEVAR. Although the initial results were acceptable, the number of patients treated (n ¼ 3) was quite small, and follow-up after the endovascular repair was not reported. Patel and associates [14] reported a more real world examination of the role of TEVAR for MTAA in their report from 2010. The perioperative mortality was 11.5% for the 27 patients treated in this fashion. Disappointingly, 18.5% had significant endoleaks, and an additional 14% of patients had late death due to ongoing sepsis. Freedom from reinfection was 81.2% at 2 years. Endovascular exclusion of an infected aneurysm with TEVAR may indeed temporarily protect extremely ill patients from fatal rupture while their multiple organ dysfunction resolves. This “staged” approach to MTAA repair, while a very expensive option, may be appropriate in this setting [25, 26].
Our results suggest, however, that primary, aggressive open surgical debridement and in-situ replacement of the infected aorta is not only safe but durable and should remain the standard treatment for MTAA. In experienced hands, we believe that mortality less than 10% is achievable. Even in a grossly infected field, radical debridement and coverage of a Dacron graft with either muscle flap or omentum reliably relieved patients of their infectious burden while restoring aortic continuity. Even though our cohort included patients on hemodialysis and a majority presented with rupture, open repair was reproducible without an expensive intermediate stage of endovascular intervention. None of our patients had reinfection for a mean of 26 months after surgery. The 71% actuarial 5-year survival speaks well for the success of this more traditional approach to the high-risk patient with MTAA. The limitations of this study include its retrospective, observational nature and small sample size. Given the relative rarity of this disease, however, this report is not markedly different from other reports on MTAA. We also concede that results from a high-volume aortic center with significant experience in open thoracic aneurysm repair may not translate well to every instance in which a patient with MTAA may present. In conclusion, open repair of mycotic thoracic and thoracoabdominal aortic aneurysms remains the most effective therapy for this very high risk cohort of patients presenting with infected aneurysms. In experienced hands, with the utilization of contemporary end organ protection strategies, a gratifyingly high rate of salvage can be expected, even for patients presenting with rupture. In cases of gross contamination of the pleural or abdominal cavities, or in cases of extensive infection of surrounding structures, tissue coverage of the in situ prosthesis may enhance recovery and contribute to the low rate of reinfection seen with this approach. The role of lifelong antibiotic suppression therapy certainly warrants further investigation, but is not unreasonable given the fatal outcome of patients with ongoing infection who have been treated in a less aggressive fashion.
References 1. Usui A. Surgical management of infected thoracic aneurysms. Nagoya J Med Sci 2013;75:161–7. 2. Oderich GS, Panneton JM, Bower TC, et al. Infected aortic aneurysms: aggressive presentation, complicated early outcome but durable results. J Vasc Surg 2001;34:900–8. 3. Muller BT, Wegener OR, Grabitz K, Pillny M, Thomas L, Sandmann W. Mycotic aneurysm of the thoracic and abdominal aorta and iliac arteries: experience with anatomic and extra-anatomic repair in 33 cases. J Vasc Surg 2001;33: 106–13. 4. Bisdas T, Wilhelmi M, Haverich A, Teebken OE. Cryopreserved arterial homografts vs silver-coated Dacron grafts for abdominal aortic infections with intraoperative evidence of microorganisms. J Vasc Surg 2011;53:1274–81. 5. Harlander-Locke MP, Harmon LK, Lawrence PF, et al. The use of cryopreserved aortoiliac allograft for aortic reconstruction in the United States. J Vasc Surg 2014;59:669–74.
LAU ET AL OPEN REPAIR OF MYCOTIC AORTIC ANEURYSMS
1717
6. Hsu RB, Chen RJ, Wang SS, Chu SH. Infected aortic aneurysms: clinical outcome and risk factor analysis. J Vasc Surg 2004;40:30–5. 7. Weis-M€ uller BT, Rascanu C, Sagban A, Grabitz K, Godehardt E, Sandmann W. Single-center experience with open surgical treatment of 36 infected aneurysms of the thoracic, thoracoabdominal, and abdominal aorta. Ann Vasc Surg 2011;25:1020–5. 8. Hsu RB, Tsay YG, Wang SS, Chu SH. Surgical treatment for primary infected aneurysm of the descending thoracic aorta, abdominal aorta, and iliac arteries. J Vasc Surg 2002;36: 746–50. 9. Huang YK, Ko PJ, Chen CL, et al. Therapeutic opinion on endovascular repair for mycotic aortic aneurysm. Ann Vasc Surg 2014;28:579–89. 10. Jia X, Dong YF, Liu XP, Xiong J, Zhang HP, Guo W. Open and endovascular repair of primary mycotic aortic aneurysms: a 10-year single-center experience. J Endovasc Ther 2013;20: 305–10. 11. Kan CD, Lee HL, Yang YJ. Outcome after endovascular stent graft treatment for mycotic aortic aneurysm: a systemic review. J Vasc Surg 2007;46:906–12. 12. Ting AC, Cheng SW, Ho P, Poon JT, Tsu JH. Surgical treatment of infected aneurysms and pseudoaneurysms of the thoracic and abdominal aorta. Am J Surg 2005;189:150–4. 13. Johnstone JK, Slaiby JM, Marcaccio EJ, Chong TT, GarciaToca M. Endovascular repair of mycotic aneurysm of the descending thoracic aorta. Ann Vasc Surg 2013;27: 23–8. 14. Patel HJ, Williams DM, Upchurch GR, Dasika NL, Eliason JL, Deeb GM. Thoracic aortic endovascular repair for mycotic aneurysms and fistulas. J Vasc Surg 2010;52(Suppl):37–40. 15. Sedivy P, Spacek M, El Samman K, et al. Endovascular treatment of infected aortic aneurysms. Eur J Vasc Endovasc Surg 2012;44:385–94. 16. S€ orelius K, Mani K, Bj€ orck M, et al. Endovascular treatment of mycotic aortic aneurysms: a European multicenter study. Circulation 2014;130:2136–42. 17. Girardi LN, Rabotnikov Y, Avgerinos DV. Preoperative percutaneous coronary intervention in patients undergoing open thoracoabdominal and descending thoracic aneurysm repair. J Thorac Cardiovasc Surg 2014;147:163–8. 18. Girardi LN, Coselli JS. Inflammatory aneurysm of the ascending aorta and aortic arch. Ann Thorac Surg 1997;64: 251–3. 19. Shahidi S, Eskil A, Lundof E, Klaerke A, Jensen BS. Detection of abdominal aortic graft infection: comparison of magnetic resonance imaging and indium-labeled white blood cell scanning. Ann Vasc Surg 2007;21:586–92. 20. Fukuchi K, Ishida Y, Higashi M, et al. Detection of aortic graft infection by fluorodeoxyglucose positron emission tomography: comparison with computed tomographic findings. J Vasc Surg 2005;42:919–25. 21. Girardi LN, Krieger KH, Altorki NK, Mack CA, Lee LY, Isom OW. Ruptured descending and thoracoabdominal aortic aneurysms. Ann Thorac Surg 2002;74:1066–70. 22. Chan FY, Crawford ES, Coselli JS, et al. In situ prosthetic graft replacement for mycotic aneurysm of the aorta. Ann Thorac Surg 1989;47:193–203. 23. Fillmore AJ, Valentine RJ. Surgical mortality in patients with infected aortic aneurysms. J Am Coll Surg 2003;196:435–41. 24. Semba CP, Sakai T, Slonim SM, et al. Mycotic aneurysms of the thoracic aorta: repair with use of endovascular stentgrafts. J Vasc Interv Radiol 1998;9:33–40. 25. Clough RE, Black SA, Lyons OT, et al. Is endovascular repair of mycotic aortic aneurysms a durable treatment option? Eur J Vasc Endovasc Surg 2009;37:407–12. 26. Fukunaga N, Hashimoto T, Ozu Y, et al. Successful treatment for infected aortic aneurysm using endovascular aneurysm repairs as a bridge to delayed open surgery. Ann Vasc Surg 2012;26:280.
ADULT CARDIAC
Ann Thorac Surg 2015;100:1712–7