Mycoplasma hominis Ascending Aortic Graft Infection Successfully Treated With Graft Preservation Using Negative Pressure Wound Therapy With Instillation and Dwell Time

Thoracic aortic graft infection is a complication of cardiac surgery that has an estimated overall occurrence rate of 0.5% to 3%.¹ Although rare, it is one of the most severe complications related to high mortality, morbidity, and costs.² There is no consensus regarding the best management of aortic graft infections. Arguably, the best solution would be a replacement with a homograft; however, this requires another major surgical intervention while facing a shortage of available homografts. Some authors recommend combining radical surgical debridement with negative pressure wound therapy (NPWT) and antibiotic therapy.^2-5 Most commonly, aortic graft infections occur in the first 3 months after surgery and result from direct contamination of graft intraoperatively or shortly thereafter from the skin.

Prosthetic vascular graft infections are related to biofilm-producing bacteria. Natural antibiotic resistance with bacteria in biofilm is a major obstacle for a successful treatment unless the prosthetic material is completely removed. Therefore, the treatment of graft-associated infection with the retention of prosthetic material should always include antibiotics that penetrate the biofilm.⁶

Hematogenous infections from a distant focus are rare.⁶ Intrathoracic vascular graft infections are most commonly caused by gram-positive cocci, followed by gram-negative rods,⁷ which also guide the empiric therapy in patients with prosthetic vascular graft infections. Other microorganisms are rare. Mycoplasma infections of the graft are very rare with only a few cases reported in literature.^8-11 The present authors report a case of ascending aortic graft infection with Mycoplasma hominis that was successfully treated with NPWT with antibiotic irrigation, surgical debridement, and systemic antibiotics.

A 63-year-old man with arterial hypertension was admitted to the Department of Cardiac Surgery due to ascending aortic aneurysm with concomitant severe aortic regurgitation. Aortic valve sparing surgery with the reimplantation of the coronary arteries¹² was performed using a prosthetic gelatin-sealed polyester vascular graft (Gelweave; Vascutek Terumo). Six days after the initial surgery, biological aortic valve replacement was necessary because of residual regurgitation of the aortic valve.

After the second procedure, the initial postoperative course was favorable. On post-op day (POD) 13 after the first surgery, his inflammatory markers were elevated (C-reactive protein [CRP] 107 mg/L, leukocytosis 15x109/L), and at that time without any clinical signs of infection. Echocardiography showed small pericardial effusion without other signs of graft or valve malfunction. The patient deteriorated the next day (POD 14), with signs of sepsis and spiking fever. Further increase in his inflammatory markers (CRP 202 mg/L, procalcitonin remained low 0.17 pg/mL) was observed. A computed tomography (CT) scan revealed periaortic fluid and a small pericardial effusion. (Figures 1, Figure 2, and Figure 3). There was no air or signs of an abscess. After obtaining blood cultures, empiric broad-spectrum antimicrobial therapy with imipenem/cilastatin and vancomycin was initiated, and the patient was admitted to the intensive care unit.

On POD 16, echocardiography showed significant pericardial effusion, while valve and graft function remained normal. A surgical reexploration was performed with no signs of intrathoracic infection detected. Samples of pericardial effusion and intrathoracic tissue were collected for microbiological analysis.

Despite the broad-spectrum antimicrobial therapy, the patient remained septic (CRP values remained above 250 mg/L), although the surgical wound showed no signs of infection. Four days later (POD 20), there was sternum instability; the patient underwent a reoperation. Intraoperatively, pus was evident in the pericardium. No signs of graft dehiscence were observed, although necrotic tissue was present around the graft. Sternal wires were removed, radical debridement of the perigraft region and thoracic cavity with lavage was performed, and NPWT with instillation (V.A.C. ULTA; 3M + KCI) was administered. Antiseptic fluid (0.02% polihexanide; Prontosan; B. Braun Medical AG) was instilled over the infected region for 10 minutes, after which continuous negative pressure (-125 mm Hg) was applied for 60 minutes. The same antibiotic therapy continued, because Gram stains of pericardial fluid were negative, and no growth was observed on standard microbiological cultures of pericardial effusion. Blood cultures remained sterile in the first 10 days of incubation. On POD 23, blood cultures were positive, and a microbiologist observed a growth of small colonies of bacteria without cell walls on blood cultures as well as in all the intraoperative tissue samples. Mycoplasma hominis was identified using real-time polymerase chain reaction.¹³

Based on the identified microorganism, the antibiotic therapy was changed to tigecycline 50 mg twice daily and levofloxacin 500 mg BID intravenously. Initially, the authors did not opt for doxycycline due to unpredicted gut absorption of oral therapy in the critically ill patient.¹⁴ After consultation with a clinical pharmacologist, the authors decided to install tigecycline in the thoracic cavity using NPWT with instillation and dwell time (NPWTi-d; V.A.C. VERAFLO Therapy; 3M + KCI). The standard reticulated open cell foam dressing (V.A.C. VERAFLO CLEANSE CHOICE Dressing; KCI + 3M) was used. The authors irrigated the thoracic cavity twice daily with a solution of 50 mg of tigecycline in a 5% glucose solution. After instillation, the solutions were allowed to dwell for 15 minutes, followed by 3 hours of -125 mm Hg continuous negative pressure. Dressing changes were performed every 2 days. No side effects of this therapy were observed, although the authors were aware of the potentially large reabsorption capacity of the thoracic cavity. Four additional surgical debridements were required during the next 10 days, but inflammatory markers began to decrease, and the patient became afebrile and stable shortly after the change of therapy.

The authors used NWPTi-d for 8 days. The patient was switched to oral monotherapy with doxycycline 100 mg twice daily after 23 days of parenteral therapy. The patient was discharged from the hospital after 12 weeks of therapy and continued therapy with doxycycline 100 mg BID for another 9 months. At the 12-month follow-up visit, he had no clinical problems, computed tomographic angiography of the thorax showed no signs of perigraft inflammation, and inflammatory markers were low; the authors decided to discontinue the antibiotic treatment. Two years later, the patient is clinically stable, with no recurrent episodes of infection and an improved quality of life. The patient continues to be managed on an outpatient basis.

Prosthetic graft infections are rare, but devastating surgical complications that diminish postoperative survival. Early postoperative aortic graft infections are usually caused by gram-positive cocci or gram-negative rods, whereas M hominis graft infections are extremely rare with only a few reports in the literature. M hominis can colonize the urogenital and respiratory tract, most commonly causing pelvic inflammatory disease or bacterial vaginosis in women. The bacteria are rarely associated with other infections; there are reports of infections after coronary bypass grafting or in heart transplant recipients.^8,9 The authors conjecture that the mechanism of infection in the present patient was a hematogenous spread of bacteria due to urinary catheterization, femoral catheter, or endotracheal intubation because no signs of sternal wound infection were initially present.

This case demonstrates a successful treatment of M hominis aortic graft infection with graft preservation using systemic and local antimicrobial therapy combined with surgical debridement and NPWTi-d. Negative pressure wound therapy is well-known in the treatment of mediastinitis after cardiac surgery^15-17; reports in the literature,^18,19 show that NWPTi-d better reduces the quantity of bacterial load locally than NWPT alone in the treatment of wound infections. Usually, antiseptic solutions have been used for irrigation, but reports also discuss the use of antimicrobial solutions for irrigation, such as vancomycin, tobramycin and gentamicin.²⁰ Although tigecycline and levofloxacin both have good tissue penetration,²¹ no published data exist on antibiotic penetration in perigraft tissue. Based on this assumption and due to our critically ill patient, we decided to add antibiotic tigecycline to the irrigation solution to achieve good local antibiotic concentration in the perigraft area. A staged treatment approach was used in accordance with the current literature^22,23 due to the authors’ intention to leave the infected material inside the thoracic cavity for as long as it is possible (recovery or infection progression).

A challenging case regarding early postoperative prosthetic aortic graft infection with an unusual pathogen M hominis inspired the authors' decision to irrigate the infected graft using NPWT with antibiotic instillation in addition to surgical debridement and systemic antibiotics. To the authors' knowledge, this is the first described successful treatment of an aortic graft infection due to M hominis without a surgical removal of the prosthetic material using antibiotic irrigation. It should be emphasized that a multidisciplinary approach played a crucial part in this patient’s successful treatment.

Authors: Miha Antonič, MD, PhD¹; Anže Djordjević, MD¹; Peter Jurič, MD¹; Maja Pirnat, MD²; and Nina Gorišek Miksić, MD, PhD³

Affiliations: ¹Department of Cardiac Surgery, University Medical Center Maribor, Maribor, Slovenia; ²Department of Radiology, University Medical Center Maribor, Maribor, Slovenia; and ³Department of Infectious Diseases, University Medical Center Maribor, Maribor, Slovenia

Correspondence: Anže Djordjević, MD, Department of Cardiac Surgery, University Medical Center Maribor, Slovenia, Ljubljanska 5, 2000 Maribor; anze.djordjevic@ukc-mb.si; anze.djordjevic@gmail.com

Disclosure: The authors disclose no financial or other conflicts of interest.