Coronary Stent Infections

Abstract

Although rare, coronary artery stent infections are associated with a high mortality rate. Since the introduction of coronary stents in 1987, only 16 cases of infection have been reported. We report a new case in a 66-year-old woman who had undergone a difficult percutaneous coronary intervention procedure, during which 3 overlapping stents were implanted in the mid portion of the right coronary artery. Twenty-two days after the procedure, the patient died. Autopsy revealed the cause of death to be pericardial tamponade due to rupture of the right ventricular myocardium. The stented portion of the right coronary artery was enveloped by an abscess, and purulent material completely occluded the stents. Cultures of the myocardium were positive for methicillin-resistant Staphylococcus aureus.

We conducted a review of the literature on coronary artery stent infections. Data suggest that early-onset infections (<10 days after stent implantation) are potentially amenable to medical therapy alone, but late-onset infections (≥10 days after implantation) or major complications necessitate combined surgical and medical therapy. Medical therapy consists of broad-spectrum antibiotics. Surgical intervention includes stent removal if possible, and abscess drainage or perforation repair when indicated.

Coronary artery stents have been in use for more than 2 decades. Stent infections are rare, hard to treat, and potentially devastating. Our treatment of a patient with a coronary stent abscess prompted us to review the literature. From our experience and the evaluation of previously reported cases,^1–14 we propose an approach for diagnosing and treating these infections.

Case Report

A 66-year-old woman with an inferior ST-elevation myocardial infarction was admitted to our hospital. She underwent emergent angiography and ad hoc percutaneous coronary intervention (PCI), during which 2 drug-eluting stents and 1 bare-metal stent were placed within the mid portion of the right coronary artery in overlapping fashion. The PCI procedure was protracted and difficult. Opening the stenosis required 7 passes of wires and balloons through the femoral sheath; placing the 3 stents required an additional 12 passes. Postprocedural angiography showed no residual stenosis or evidence of dissection. After the sheath was removed, excessive bleeding at the insertion site was stopped by prolonged use of mechanical compression.

On hospital day 2, a peripheral intravenous catheter, which had been placed in the patient's left wrist by emergency medical services before admission, became infiltrated and was removed. On day 4, erythema and ecchymosis extended from the wrist to the upper arm. At the time of hospital discharge 2 days later, these symptoms had improved. The only antibiotic administered during hospitalization was ciprofloxacin to treat a klebsiella urinary tract infection.

The patient returned to the hospital 12 days after discharge (17 days after PCI). She had experienced general weakness and malaise for 3 days; subjective fever and rigors for 1 day; and midsternal chest pain, which had awakened her from sleep, 1 hour before presentation. Physical examination was unremarkable, and her temperature was 99 °F. The electrocardiogram showed only minor nonspecific ST-T wave changes.

Laboratory findings were notable for leukocytosis (17,100 white blood cells/mm³) and for elevated levels of creatinine (3.9 mg/dL). The initial chest radiograph showed no infiltrate or effusion, but subsequently the cardiac silhouette was enlarged. Blood cultures grew methicillin-resistant Staphylococcus aureus (MRSA) after 18 hours. The transesophageal echocardiogram (TEE) revealed nothing unusual, except for a small, hemodynamically insignificant pericardial effusion. A computed tomographic (CT) scan without contrast of her thorax, abdomen, and pelvis revealed a pericardial effusion.

Despite early appropriate antibiotic therapy, the patient's blood cultures remained positive, and her clinical condition continued to deteriorate. On hospital day 3, a pericardial friction rub was noted. On day 5, the patient experienced cardiac arrest with pulseless electrical activity and died. Autopsy revealed that the cause of death was pericardial tamponade, secondary to the rupture of necrotic right ventricular myocardium. An adjacent abscess (3.5 × 1.5 cm) enveloped the stents in the mid right coronary artery, and purulent material completely occluded the stented portion of the right coronary artery. Gram staining of the myocardium revealed clusters of gram-positive cocci consistent with S. aureus.

Discussion

The case of our patient demonstrated several facts about coronary stent infection, but many questions were unanswered. Therefore, we performed a PubMed search of the English-language medical literature, using these key words: infection, coronary artery, coronary intervention, and complication. In our analysis, we included all cases of infection documented by pathologic evidence or imaging studies. Definitive diagnosis of a stent infection was based on the presence of an abscess or inflammatory mass, or an aneurysm or pseudoaneurysm. Information obtained from each case included time of onset in relation to stent placement, presenting symptoms, duration of symptoms, organisms isolated, complications, usefulness of the echocardiogram and imaging studies, antibiotic and surgical treatment, and outcome.

Early-onset infections were defined as those occurring less than 10 days after stent placement, and late-onset infections, as those occurring 10 days or longer after stent placement. The designation of 10 days as the cutpoint between early- and late-onset infection was arbitrary. These definitions were based on evaluation and treatment of foreign-body infections in orthopedic surgery,^15,16 in which the earlier after a procedure and the fewer the days of symptoms before diagnosis of infection it was, the more likely that the infection could be cured.

Our literature search yielded a total of 16 patients with coronary stent infections,^1–14 and ours is the 17th. Table I lists their general characteristics. Most patients were age 50 or older and male, consistent with the characteristics of patients with coronary artery disease. Surprisingly, only a few patients had other diseases—such as hypertension, diabetes mellitus, and hyperlipidemia—that might have predisposed them to coronary artery disease. None of the patients were immunosuppressed, and only 2 had end-organ disease. Most cases (12 of 17) have been reported since 2000,4–10,12–14 during which time the use of stents has markedly increased.

Table I. Characteristics of Patients with Coronary Stent Infection

The evaluation of a patient with a potential stent infection has several goals. These include making a clinical, radiographic, and microbiologic diagnosis; detecting complications; and determining optimal therapy. The clinical presentation of patients with stent infections was generally consistent. In particular, fever, chest pain, or both occurred in all patients. The details for each patient are reported in Table II.

Table II. Reports of Coronary Stent Infection, Grouped by Time of Onset after Procedure

Table II. (continued). Reports of Coronary Stent Infection, Grouped by Time of Onset after Procedure

The pathogenic organism was identified in all patients. Staphylococcus aureus, the most common organism, was found in 14 patients, all of whom had positive blood cultures. Six patients had methicillin-sensitive S. aureus (MSSA), and 5 had MRSA. In 2 patients with S. aureus, antibiotic sensitivity was not documented. Finally, in 2 patients, Pseudomonas aeruginosa was isolated from blood cultures, and in 1, coagulase-negative staphylococci were isolated from infected tissue.

We used the criteria proposed by Dieter¹⁷ to determine the diagnosis of coronary stent infection. Definitive diagnosis was made by autopsy or by examination of surgical material. For possible diagnosis, 3 of the following criteria must have been present: placement of a coronary stent within the previous 4 weeks; multiple repeat procedures performed through the same arterial sheath; the presence of bacteremia, significant fever, or leukocytosis with no other cause; acute coronary syndrome; or positive cardiac imaging.

Although positive blood cultures are diagnostic of a serious infection, they do not necessarily identify the anatomic source of the infection. Therefore, several radiographic tests were used to make the diagnosis of stent infection and complications. In all but 3 patients (one of whom was ours), the results of radiographic testing were positive before therapy was initiated. One patient who had been treated and cured with 4 weeks of parenteral antibiotics was found to have aneurysms at follow-up.⁷ Table III shows the usefulness of the different diagnostic imaging techniques. Surprisingly, the results of TEE were positive for stent infection in only 4 of 10 patients.5,6,9,11 In 1 patient,⁹ the TEE was initially negative, but repeat TEE revealed an aneurysm. All angiographic procedures, with the exception of 1 CT scan, showed a pseudoaneurysm or aneurysm, although in our patient both the chest CT and TEE were negative. Therefore, an angiographic study should be considered the procedure of choice, in conjunction with TEE, in making the primary diagnosis and detecting catastrophic complications. One patient¹⁴ also underwent cardiac magnetic resonance, and the results were positive. Whether this test, with further use, will consistently aid in the rapid diagnosis of stent infection remains to be seen.

Table III. Results of Diagnostic Methods

Expected findings in patients with coronary stents, which we refer to as vessel pathology, included local abscesses and simple pericarditis. These findings are rarely clinically significant. In contrast, major complications are unexpected, potentially catastrophic events, including pericardial empyema, purulent pericarditis, vessel perforation, destroyed vessel, and myocardial rupture.

Our review indicates that the development of a major complication, the time of onset of infection, and the type of therapy—medical therapy alone or combined with surgery—can dramatically affect outcome. Ten patients, including ours, had major complications, and 5 died (Tables II and IV).3,9,10,13 Early-onset infections appeared to be amenable to medical therapy alone; one half of such patients survived. The presence of major complications had little effect on mortality in these patients. Of 6 patients treated with combined medical and surgical therapy, 4 survived, 1 of whom had major complications and underwent stent removal. Survival with late-onset infection required surgical therapy. Three of the 4 surgical patients survived, but all patients who underwent medical therapy alone died. In cases where the length of medical therapy was documented, all survivors had received parenteral antibiotics for at least 4 weeks. The usefulness of subsequent oral antibiotic therapy could not be determined.

Table IV. Outcome Based on Time of Onset of Infection and Presence of Major Complications

Of note, all patients with MRSA, including ours, had late-onset infections, whereas all patients with MSSA had early-onset infections. There was no difference between the 2 organisms and the survival rate. Both patients with pseudomonas had early-onset infections.

Our patient presented 17 days after PCI, and thus was designated as having a late-onset infection. Her clinical course was similar to that in the previously published cases. She had multiple risk factors for stent infection, including advanced age, multiple skin punctures, a large number of catheters and balloons used during the procedure, manipulation of the sheath, and excessive bleeding. Her presentation with chest pain and fever, along with isolation of S. aureus from blood cultures, was characteristic. The TEE results were not helpful in making the diagnosis. Autopsy revealed necrotic myocardium with subsequent rupture into the pericardial sac, resulting in tamponade and death. Whether earlier use of other radiographic procedures, such as coronary angiography or CT angiography, might have changed her course, is unknown. However, she had a major complication and did not undergo surgery, both of which are associated with a poor prognosis.

Conclusion

On the basis of our case and literature review, we believe that the following suggestions can be helpful in the care of patients with coronary stent infections. In patients with fever and chest pain after stent implantation, stent infection should be suspected. Blood cultures should be drawn, and broad-spectrum antibiotics—to cover MRSA, MSSA, and gram-negative organisms—should be started. An angiographic procedure should be done as soon as possible. In late-onset infections, surgery should be considered early in the course, especially if major complications are apparent on the angiogram or TEE. In early-onset infections, medical therapy alone might prove successful, but surgery is indicated if any major complications are present or if medical therapy appears to be failing. If possible, the stent should be removed at the time of surgery. Antibiotic therapy should be administered parenterally for at least 4 weeks, or possibly longer, particularly when the stent is not removed at the time of surgery. The usefulness of subsequent therapy with oral antibiotics is not known.