in Iliac Artery Mycotic Aneurysm: The Role of Molecular Diagnostics

Introduction

Capnocytophaga canimorsus (C. canimorsus) is a gram-negative rod that is part of the oral flora of cats and dogs [1]. The prevalence of this infection has been estimated to be 0.5–4.1 cases per million [1,2]. C. canimorsus infection is usually acquired through animal bites or licks on skin wounds [1]. C. canimorsus is a fastidious, slow-growing bacterium that grows well on blood agar and in the presence of animal cells, but without cells to provide nutrition, fails to grow readily; thus, routine liquid and agar-based bacterial culture media may fail to isolate this organism [3]. In addition, contamination of bite wounds by skin flora can disrupt growth of C. canimorsus [4]. C. canimorsus and other fastidious organisms multiply slowly, taking days or weeks to grow in culture [5]. When clinically suspected, prolonged incubation or enrichment of culture media with rabbit blood with or without carbon dioxide-enriched atmosphere can improve the yield of organisms [6].

Though infections with C. canimorsus are rare, they can be severe, leading to complications like mycotic aneurysm [7,8]. Mycotic aneurysms, also known as infectious aneurysms, occur due to inflammation and degeneration of the arterial wall resulting from microbial infection [7]. Only 10 cases of vascular infection with C. canimorsus have been reported previously [7,8]. C. canimorsus infection can also lead to disseminated intravascular coagulation (DIC), endocarditis, meningitis, brain abscess, septicemia, and septic shock [1,3,9]. Mortality rates for C. canimorsus sepsis have been found to be as high as 30–50% [2,3]. Risk factors for severe infection include asplenia, alcoholism, and male sex [1]. However, up to 40% of infected patients have no known risk factors [2].

This report describes the case of a 67-year-old woman presenting with a C. canimorsus mycotic iliac artery aneurysm, requiring polymerase chain reaction (PCR) for microbial identification, 2 months after a dog bite.

Case Report

A 67-year-old woman with a history of mild uncomplicated alcohol use disorder presented to our clinic with 1 week of right lower-quadrant abdominal pain and nausea. She denied fevers, hematochezia, constipation, and diarrhea. She lived and worked in northwest Wisconsin. Her medical history also included aneurysmal disease of the left middle cerebral artery and anterior communicating artery, for which she underwent coil embolization, and is currently under surveillance of small right internal carotid artery and right middle cerebral artery aneurysms.

At the time of presentation, she had a blood pressure of 120/70 mmHg, heart rate of 104 bpm, SpO2 of 97% on room air, and temperature of 36.8°C. On physical examination, she showed moderate abdominal discomfort with escalation after palpation, no distension, and normal bowel sounds. Heart sounds were regular, with no murmurs appreciated. No abnormal skin findings were noted. Her physical exam was otherwise unremarkable. Laboratory tests were ordered and revealed an elevated C-reactive protein, at 107.1 mg/L; however, all components of a complete blood count were within normal limits. A computed tomography (CT) scan with contrast of the abdomen was performed, which revealed hydronephrosis and dilation of the right ureter from the kidney to the level of the pelvic inlet (Figure 1). The right iliac artery was dilated and eccentrically lobulated with thickened, heterogeneously-enhancing walls and inflammatory changes in the surrounding soft tissue, including fluid or edema present in the lower pelvis. These findings raised suspicion for mycotic aneurysm.

The patient was referred to vascular surgery and urology and was seen by both teams a few days following initial presentation and testing. Standard bacterial and candida peripheral blood cultures were then performed, which showed no growth after 5 days. A positron emission tomography (PET) scan was ordered by vascular surgery, which confirmed abnormal radiotracer uptake near the bifurcation of an aneurysmal right iliac artery (Figure 2). This further increased the likelihood of mycotic aneurysm.

Due to intense right flank pain from the hydronephrosis, a percutaneous nephrostomy tube was placed for decompression. A week after nephrostomy tube placement, the vascular and urologic surgical teams took the patient to surgery. Prior to surgery, antibiotics were withheld due to her clinical stability. Additionally, antibiotics were withheld in the hope that acquiring surgical tissue samples prior to antibiotic therapy would improve diagnostic yield of subsequent culture attempts, as previous culture was unsuccessful in isolating a causative pathogen, and allow for directed antibiotic therapy. A ureteral stent was placed to minimize risk of ureteral injury during dissection. Midline exploratory laparotomy revealed an intense desmoplastic reaction encasing the right ureter, iliac vein, and iliac arterial bifurcation. The distal common iliac artery demonstrated saccular aneurysmal degeneration causing external compression of the ureter (Figure 3). After the infrarenal aorta and iliac arteries were dissected free, an aorta-to-right external iliac artery bypass with cryopreserved cadaveric arterial allograft was performed and wrapped with omentum (Figure 3).

Surgical specimens, including the aneurysm wall, an adjacent lymph node, and surrounding tissue were collected for microscopic analysis, Gram stain, fungal and acid-fast bacilli (AFB) smear, and bacterial, fungal, and mycobacterial cultures. Histological examination of the iliac artery specimens showed an aneurysmal wall with thrombus (Figure 4), calcification within the iliac artery wall, and aneurysmal wall degeneration. A para-aortic lymph node showed no evidence of malignancy. Following surgical intervention, the patient was started on empiric intravenous ceftriaxone and vancomycin. Intravenous vancomycin was discontinued when bacterial cultures remained without growth on postoperative day 2. Intravenous ceftriaxone was continued pending results of testing, and the patient was discharged on postoperative day 6 with a plan for 6 weeks of intravenous antibiotic therapy. All cultures remained negative up to 42 days of incubation. Peripheral blood fungal and mycobacterial cultures drawn perioperatively were also negative after 30 and 42 days of incubation.

Due to high suspicion for bacterial infection, despite negative cultures with extended-duration incubation, broad-range bacterial 16s rRNA PCR plus product sequencing was requested for the arterial tissue specimen; 16s rRNA PCR is part of our standard protocol and is performed at the Mayo Clinic – Rochester laboratory. This test uses primers specific for the 16s rRNA gene, which is only found in bacteria and archaea, to amplify genetic material; in our case this was matched to C. canimorsus (Figure 5).

While the patient initially denied animal exposures, she later recalled receiving a dog bite 2 months prior to presentation. She completed 6 weeks of intravenous ceftriaxone and was transitioned to oral penicillin to complete 6 months of treatment until tissue graft epithelization. As she had a retained synthetic vascular graft with high risk of poor outcome if another surgery was needed, prolonged secondary prophylaxis with oral antibiotic therapy was initiated. Oral penicillin was chosen for this based on antimicrobial susceptibility data after completing a 6-week course for vascular graft infection.

11 days after surgery, a CT scan showed severe atheromatous disease of the aortoiliac vessels and a patent aorto right iliac bypass graft present. There was also low-attenuation fluid in the right retroperitoneum at this time. 2 months after surgery a CT scan showed stable postsurgical changes. The ureteral stent was removed 4 weeks after surgery, the nephrostomy tube was removed after 2 months without evidence of recurrent hydronephrosis, and the patient remained symptom free at 9-month follow-up.

Discussion

This case report illustrates the growing importance of molecular technologies in identifying serious infections with fastidious bacteria. Culture has long been regarded as the criterion standard for bacterial identification [5], but the fastidious nature of C. canimorsus and contamination of bite wounds can make this challenging [9]. In a previous study reviewing cases of C. canimorsus meningitis, blood cultures were positive in only 54% of cases [10]. In a review of previously reported C. canimorsus mycotic aneurysms, 3 of 8 cases required sequencing for diagnosis [7,8]. Despite high suspicion for infection, a microbiologic diagnosis was not achieved with culture despite multiple peripheral and surgical tissue specimens obtained prior to antibiotic administration and prolonged incubation. The history of dog bite was not forthcoming until after an infectious etiology was established.

Molecular technologies, such DNA-based sequencing, are emerging as important tools for identifying fastidious bacteria [5]. Broad-range bacterial 16s rRNA PCR plus sequencing uses primers to identify conserved sequences in the 16s rRNA gene, amplifies variable regions between bacterial genomes, and uses these amplified regions for identification [11]. Genus-specific identification can be made in up to 90% of cases, while species-specific identification is possible in 65–90% [11]. 16s rRNA PCR can also identify bacterial DNA after the initiation of antimicrobial therapy and assist in selecting effective antibacterial treatment that is optimally specific for the identified pathogen [11]. 16s rRNA PCR testing at our institution is done utilizing the dual priming technology for PCR amplification, as has been described by Kommedal et al [12].

Broad-range PCR is available as a test of choice within the Mayo Clinic health care system and is often used to identify organisms when traditional culture methods, even with prolonged incubation, fail in evaluation. 16s rRNA PCR is the preferred test for broad-range bacterial testing and reflex to next-generation sequencing, also available in-house in the Mayo Clinic lab, occurs based on an established protocol for sequentially escalating levels of testing in our central molecular labs; this is done for samples that do not yield accurate results on 16s rRNA sequencing or at the discretion of a clinical microbiologist in discussion with the treating infectious disease specialist.

Molecular-based testing is not without drawbacks. It is often not readily available, can be more expensive than conventional testing, and requires personnel with specialized lab training [5]. Results also depend on an institution’s access to existing molecular databases to match genetic sequences. Other methods like metagenomic sequencing are send out tests and are utilized per protocol only when in-house tests are non-contributory or unlikely to address the clinical concern. As we have access to extensive testing within our institution, routine sending of samples for commercial next-generation sequencing tests is not done; this helps limit cost to patients, a factor clinicians should carefully consider when selecting testing strategies.

These tests are also limited in that they cannot distinguish between viable and non-viable bacteria, possibly giving false-positive results, and cannot directly test for antimicrobial susceptibilities [11]. However, clinicians should be familiar with DNA-based strategies and collaborate with their institutional experts in cases that would benefit. Although 16s RNA PCR cannot distinguish between viable and non-viable bacteria, in this case the clinical picture of the patient suggested active infection with C. canimorsus. The focus of abnormal radiotracer uptake in the iliac vessel on PET scan, elevated CRP, operative findings suggesting inflammation, and surgical pathology of the vessel showing aneurysmal wall with thrombus and inflammation all support that the C. canimorsus detected by 16s rRNA PCR was an active infection. Correct diagnosis is important to guide antimicrobial therapy and source control.

In surgery, it is important to resect the infected tissue, and vascular reconstruction with autogenous or non-autogenous biological graft is recommended to reduce risk of re-infection [13]. Rifampin-soaked Dacron graft with omental coverage has been utilized with encouraging results [14]. From an infection control standpoint, it has been advised to resect all infected tissue and use biological grafts to reduce risk of re-infection [13], with the use of rifampin-soaked prosthetic grafts when allograft material is not available [14].

Preoperative imaging can assist with achieving correct diagnosis and planning an appropriate surgical strategy. Findings suggestive of a mycotic aneurysm, not seen in non-infective atherosclerotic aneurysms, are: a) saccular, eccentric, or multi-lobulated aneurysm; b) soft tissue inflammation/mass around the vessel; c) intramural air or air around the vessel; d) perivascular fluid collection; e) thickening of the aneurysm wall (also present in inflammatory aneurysms); and f) high uptake in fluorodeoxyglucose-positron emission tomography (FDGPET) and gallium scanning [15].

To date, only 10 previous cases of vascular infection with C. canimorsus have been reported [7,8]. 9 cases were reported in a review by Etemady-Delamy et al and 1 case was described by Berndsen et al [7,8]. 8 of these cases were reported as mycotic aneurysm, as the diagnosis in our case was classified, and 2 as vasculitis caused by C. canimorsus infection [7,8]. 5 of these 8 previously reported cases of C. canimorsus mycotic aneurysm were diagnosed using conventional culture and the remaining 3, like the case detailed in this report, required sequencing for diagnosis [7,8]. Most cases involve adults over 50 years of age with history of exposure to cats or dogs [7,8]. Our patient falls within this age category, at 67 years old, and had also received a dog bite prior to presentation, as elaborated above. 4 of the previous patients were female and 6 were male [7,8]. The most commonly reported vessel involvement was the abdominal aorta [7,8], with only 1 case other than ours involving the iliac artery [8]. There are no guidelines for antimicrobial susceptibility testing of Capnocytophaga spp. Broad-spectrum antibiotics are often utilized for empiric therapy until organism identification is achieved; successful use of ceftriaxone, piperacillin/tazobactam, ciprofloxacin, and carbapenems have been reported [7,8] and in most reported cases, as in this case, patients underwent surgery [7,8].

Infection of tissue generally occurs through 1 of 3 mechanisms: bacteremia in which hematogenous spread brings a microbe to new tissue, direct inoculation of the pathogen into the tissue, and extension into new tissue from nearby infection [7]. It has been suggested that vascular infections, such as mycotic aneurysms, occur through hematogenous spread with involvement of the aorta or its branches due to pre-existing vascular defects [8]. This defect could be a pre-existing pathology such as atherosclerosis or aneurysmal changes, or prosthetic material like a vascular graft. The relatively common occurrence of both atherosclerosis and aneurysms in the aorta may explain why this location is often affected. In our case, the patient had known pre-existing cerebral aneurysms, and microscopic tissue analysis found calcific changes, raising the possibility of a pre-existing atherosclerotic iliac aneurysm secondarily infected through hematogenous spread of C. canimorsus following her dog bite.

Conclusions

C. canimorsus is a bacterial organism and oral commensal of cats and dogs; it can cause life-threatening infections. Infection with C. canimorsus can be difficult to diagnose due to laboratory testing limits and the organism’s fastidious growth requirements. This report presents the case of a 67-year-old woman with an iliac artery mycotic aneurysm 2 months following a dog bite. C. canimorsus was identified as the causative infectious organism through 16s rRNA PCR. This case emphasizes that use of molecular diagnostic techniques such as 16s rRNA PCR can be essential to providing accurate diagnoses and guiding prompt treatment in similar situations. Additionally, a thorough history is essential for assuring clinical reasoning is sound, especially in cases involving rare infections. Clinicians should keep C. canimorsus infection in mind when risk factors such as dog or cat bite are present, and when standard microbiological identification techniques cannot provide diagnostic answers.