INJECTION DRUG USERS ENDOCARDITIS

Endocarditis in injection drug users
The incidence of endocarditis in the injection drug user is estimated to be 40 times that of the general population, or approximately 1 case in 500 injection drug users. Unlike the general population, endocarditis in injection drug users is typically right-sided (50 to 76 percent), with 40 to 69 percent of cases involving the tricuspid valve, 20 to 30 percent involving the mitral and aortic valves, and 5 to 42 percent involving multiple valves. The pathogenesis of right-sided endocarditis in injection drug users is thought to be dependent on four factors: (1) endothelial damage; (2) specific interactions between microorganisms and valvular location; (3) degree of bacterial load; and (4) host immune status. Patients with injection drug user-related endocarditis usually have no evidence of prior valve damage, a common risk factor for endocarditis in non-injection drug users. It is thought that the injection drug user contributes to endocardial damage via mechanical and ischemic mechanisms. Mechanical damage is caused by both direct bombardment with particulate matter (talc or other diluents) and by increased right-sided pressure gradients and turbulence as a consequence of drug-induced pulmonary hypertension. The tricuspid valve is particularly susceptible to mechanical damage because it is the first valve exposed to these substances.

Additional endothelial damage may be caused by chronic valvular inflammation due to frequent antigenic exposure. An increased risk of endocarditis in intravenous cocaine users is attributed to a the greater frequency of injections (increased mechanical damage), less chance of needle sterilization because cocaine does not need to be heated in order to go into solution, and the greater likelihood of needle sharing. An alternative explanation is that the vasoconstrictive properties of cocaine may cause interstitial or endothelial damage which predisposes to infection.

Specific interactions between valvular location and infecting organisms may account for the prevalence of right-sided Staphylococcus aureus and Pseudomonas aeruginosa endocarditis. S. aureus is the most common pathogen in isolated tricuspid valve endocarditis, accounting for 50 to 60 percent of cases among injection drug users. S. aureus possesses unique surface proteins that enable it to adhere to host tissue. In addition, injection drug users are postulated to have a greater expression of matrix molecules that bind to the microbial surface components recognizing adhesive matrix molecules found on injected particulate matter. This increased affinity as well as S. aureus' protective fibrin coat enhances vegetation formation and prevents host clearance. Pseudomonas endocarditis is more prevalent in injection drug users because of the frequent use of tap or toilet water in the preparation of the street drugs pentazocine and tripelennamine.

Bacterial load is related to the risk of endocarditis in a "dose-response" fashion. The fact that cocaine, which has a shorter half-life than heroin and requires more frequent dosing, is associated with more cases of endocarditis may be related to increased mechanical damage and greater frequencies of bacteremia. Immunologic dysregulation with hypergammaglobulinemia, high levels of circulating immune complexes, exaggerated lymphocytosis as well as atypical lymphocytosis is thought to play a role in the increased risk of endocarditis. In injection drug users with HIV, low CD4+ T-cell counts increase the risk for endocarditis sixfold.

Presenting signs and symptoms in injection drug users with endocarditis include fever, cardiac murmur (>50% of patients), cough, pleuritic chest pain, and hemoptysis. Right-sided murmurs, which vary with respiration, are typically pathologic and more specific for the diagnosis. In those with right-sided endocarditis and septic pulmonary emboli, pulmonary complaints, infiltrates on chest radiographs, and moderate hypoxia have been described in more than one-third of patients, and may mislead the physician to identify the lung as the primary source of infection. Pyuria and hematuria are ascribed to glomerulonephritis, embolic renal infarction, and perinephric abscess.

Blood cultures will be positive in more than 98 percent of injection drug user-related endocarditis patients if three to five sets are obtained. True culture-negative endocarditis in the setting of high clinical suspicion and careful laboratory procedure is rare. S. aureus has been isolated from blood cultures in more than 50 percent, and up to one-third of these isolates, particularly those from large urban areas, are methicillin resistant.10 Streptococcus is the second most frequently reported isolate, particularly S. viridans. Unlike S. aureus, streptococci are more likely to involve left-sided structures.

Other less-commonly isolated organisms include enterococci and gram-negative bacteria, particularly P. aeruginosa, Serratia marcescens, and Klebsiella pneumoniae. Atypical organisms usually reflect local environmental pathogens or drug-injecting habits. For example, licking needles prior to injection has been implicated in Eikenella corrodens, Haemophilus parainfluenzae, Bacteroides sp., and Neisseria sp. infections. Up to 20 percent of injection drug user-related endocarditis is polymicrobial in nature. Although rare, fungal endocarditis has been reported, with Candida parapsilosis accounting for over half of all isolates.

Diagnosis generally requires microbial isolation from a blood culture and/or the ability to demonstrate typical lesions on echocardiography. The classic findings of embolic phenomena, Janeway lesions and Roth spots, are usually not observed unless the infection is advanced. A complete blood count (CBC), chest radiograph, and urinalysis will usually demonstrate abnormalities, although no single result is specific for endocarditis. Typical radiographic findings include infiltrates consistent with septic emboli, pneumonia, or congestive heart failure.

Transthoracic echocardiography (TEE) is the most sensitive imaging modality for demonstrating vegetations, myocardial and ring abscesses, and tricuspid valve involvement in injection drug user-related endocarditis and will reveal diagnostic cardiac lesions in up to 80 percent of patients. While the accuracy of TEE is clear, the timing of the study is not. Some authors advocate early imaging to confirm the diagnosis, while others advocate reserving imaging on those with positive blood cultures, to determine response to therapy and plan treatment.

Attempts to develop criteria that prospectively identify endocarditis in the injection drug user with reasonable certainty have failed. At least two sets of blood cultures should be obtained in patients with suspected endocarditis, followed by hospital admission. The need for empirical antibiotic therapy should be determined by the patient's clinical stability and ability to wait for initial blood culture results. Initial treatment should be directed against S. aureus and Streptococcus spp., with consideration of local sensitivities and pathogens. Vancomycin or nafcillin and gentamicin are often initial therapy. The addition of an aminoglycoside has been shown to shorten the duration of bacteremia and duration of treatment in patients with S. aureus infections. Although 4 to 6 weeks of antibiotic therapy is standard, excellent cure rates have been achieved with only 2 weeks of an isoxazolyl-penicillin (e.g., cloxacillin) and an aminoglycoside in patients with right-sided endocarditis caused by sensitive S. aureus.

Complications of endocarditis include pump failure, dysrhythmias, and pulmonary and systemic emboli. Mortality is related to vegetation location and size, response to antibiotics, and therapeutic compliance. Left-sided endocarditis is more likely to be complicated by left-sided heart failure, septic cerebral and systemic emboli, and the need for surgery. Overall mortality for injection drug user-related left-sided endocarditis is 14 to 21 percent. In patients with right-sided endocarditis, the noteworthy complication is septic pulmonary emboli in 30 to 60 percent. Most complications related to right-sided endocarditis may be managed medically, and the prognosis is excellent with overall mortality rates 2 to 7 percent.