Endocarditis - injection drug users
- Staphylococcus aureus (MSSA or MRSA): most common
- Fever, malaise, chest/back pain, cough, dyspnea, arthralgia/myalgia, neurologic sx, wt loss, night sweats.
- Back pain could represent vertebral osteomyelitis, discitis and/or epidural abscess
- Endocarditis rates especially due to MRSA appear to be rising, commiserate with the U.S. opioid epidemic.
- Suspect endocarditis in any IDU with fever without an otherwise identifiable source.
- Pathogens: S. aureus- 60%, Streptococcal species- 20%, P. aeruginosa- 10%, Candida - 5%, S. epidermidis- 2%.
- Right-sided endocarditis, most common, tricuspid valve involved in 60% of cases.
- Duke Clinical Criteria: 2 Major OR 1 Major + 3 Minor OR 5 Minor.
- Major (microbiology):
- Major (valve):
- Echocardiography w/ vegetation definitive/oscillating mass
- New valve regurgitation
- Predisposing cardiac condition or IDU
- Fever ≥ 38°C (100.4°F)
- Vascular phenomenon (arterial emboli, mycotic aneurysm, intracerebral bleed, conjunctival hemorrhages, Janeway lesions)
- Immune phenomenon (glomerulonephritis, Osler nodes, Roth spots, positive rheumatoid factor
- Positive blood cultures not meeting above criteria
- Echocardiographic valve abnormality but not diagnostic for vegetation
- Patients with active substance abuse are often reluctant to stay in hospitals or nursing facilities to receive IV-based treatment.
- Short-course therapy (2 wks) with IV abx may be possible for some patients with uncomplicated TV endocarditis (see below).
- Oral therapy: rarely advocated but sometimes used depending on clinical circumstances when IV therapy not possible or patient leaves against medical advice, limited clinical data available.
- See specific pathogen module for additional information.
- S. aureus (MSSA, preferred):
- Standard course: oxacillin or nafcillin 2gm IV q4h x 6wks
- Non-severe PCN allergy: cefazolin 2gm IV q6h IV x 6 wks
- Standard course: oxacillin or nafcillin 2gm IV q4h x 6wks
- S. aureus (MRSA or severe PCN allergy):
- Streptococcus viridans
- PCN sensitive, MIC < 0.12:
- PCN MIC >0.12ug/ml-≤0.5µg/ml:
- PCN MIC > 0.5µg/ml (also including Abiotrophia defectiva and Granulicatella species, Gemella species): use enterococcal regimen.
- Ampicillin 2g IV q4h OR aqueous pen G 18-20 mil U/d IV continuous or divided into 4-6 doses x 4-6 wks plus gentamicin* 1 mg/kg q8h IV or IM x 4-6 wks
- Ampicillin 2g IV q4h OR aqueous pen G 18-20 mil U/d IV continuous or divided into 4-6 doses + ceftriaxone 2g IV q12H BOTH for 6 weeks
- Vancomycin 15 mg/kg q 12 h IV plus gentamicin* 1 mg/kg q8h IV or IM x 4-6 wks.
- HACEK organisms:
- P. aeruginosa (use in vitro data to guide, preferred):
- Piperacillin 4gm IV q4h or ceftazidime 2gm IV q8h + tobramycin 2.5mg/kg IV q8h IV (high dose, peak goal 15-20 mcg/ml) +x 4-6wks
- Alternatives: combination therapy
- Candida (preferred):
See PROSTHETIC VALVE endocarditis for PVIE regimens.
- Indications: severe heart failure, uncontrolled infection, persistent bacteremia despite abx, fungal endocarditis, unstable prosthetic valve, periannular extension.
- Tricuspid valve: may consider valvectomy or vegetectomy + valvuloplasty.
- Aortic or mitral valve: usually requires replacement.
- Issues: some cardiac surgeons are reluctant to operate for IE, requiring assurance there will be drug rehabilitation or refusing if second or third valve replacement.
- Available literature does not suggest a mortality difference in patients with endocarditis between IDU and non-IDU populations.
- Usual presentation: fever, chest x-ray with septic emboli, blood culture yields S. aureus, echocardiogram - tricuspid valve vegetations.
- Surgery: prognosis for prosthetic valve without drug rehabilitation is poor.
- For tricuspid valve endocarditis - valvectomy is an option.
- Concurrent HIV infection increases mortality rate when CD4 counts < 200.
- Trans-esophageal echocardiogram (TEE) recommended for patients with an initial negative trans-thoracic echocardiogram (TTE) if sufficient clinical suspicion remains that would inform treatment decision, those with inadequate TTE view or intracardiac complication on TTE.
Pathogen Specific Therapy
1st Line Agent
2nd Line Agent
Basis for recommendation
Comment: Current IE management guidelines by the American Heart Association
Comment: Persons who inject drugs have 16.3x more invasive MRSA infections than others. The rate of MRSA infections increased dramatically from 4.1% in 2011 to 9.2% in 2016. Infection types were frequently those associated with nonsterile injection drug use causing invasive MRSA infections, including endocarditis, osteomyelitis, and skin and soft tissue infections.
Comment: This study examined 13 studies including 1593 patients of which 341 were IDU-IE in the meta-analysis. IDU-IE patients more frequently had tricuspid valve infection, Staphylococcus infection, and heart failure before surgery. Meta-analysis revealed no statistically significant difference in 30-day post-surgical mortality or in-hospital mortality between the two groups.
Comment: One year mortality for patients who injected drugs and had a second bout of endocarditis was 36.3%. Staphylococcus aureus was the most common offending pathogen but occurred less commonly than in patient’s first bout of IE.
Comment: A minimally invasive approach may hold some promise with further study in assisting control of infection in patients with TV disease.
Comment: Study from Cleveland Clinic found a 10x risk of death or reoperation in the 3-6 month period after cardiac surgery in those who use injection drugs. Available followup after 6 mos shows a much smaller risk.
Comment: Authors review 7 trials examining oral therapy for bacterial endocarditis. The largest study used ciprofloxacin and rifampin for right-sided disease with equivalent results to traditional IV therapy (Heldman 1996) but still was not sufficiently powered. Other smaller studies offer varying quality and outcomes.
Comment: Review of S. aureus endocarditis in 133 patients including 53 IDU's. Mortality of right sided endocarditis in IDUs was 3.7% vs. 82% in cases associated with IV lines.
Comment: Review of 247 cases of endocarditis in San Francisco, 74% were IDUs, most heroin users. OR for IDUs vs. others: S. aureus 5.5, E. faecalis 0.2. tricuspid valve 4.4, mitral valve 0.4.
Comment: Retrospective study of IE due to MSSA who received empiric vancomycin versus beta-lactam or beta-lactam + vanc. Vancomycin patients switched to beta-lactam therapy once susceptibility results became available had outcomes inferior to those for patients treated with a beta-lactam from the initiation of therapy. Based on the observed results, empirical therapy with both a beta-lactam and an anti-methicillin-resistant S. aureus agent should be considered for serious S. aureus infections
Comment: IDU's accounted for 167/1797 (9.2%) in global collaborative study.
Comment: Incidence of endocarditis with HIV and IDU risk decreased from 20.5/1000 person years in the pre-HAART era to 6.6/1000 person years in the HAART era. Most common pathogen - S. aureus - 69% including MRSA in 28%. At one year 52% were dead and 16% had a recurrence.
Comment: Prospective trial of S. aureusendocarditis showed daptomycin (6mg/kg/day) was "not interior" to vancomycin therapy.
Comment: Author note the epidemic spread of MRSA in injection drug users across Europe and N. America. Nasal carriage is a risk for infection and IDU's have high risks of colonization.
Comment: Prospective study of 2,529 injection drug users x 10 yrs showed incidence of endocarditis was 7/1000 pt yrs which is about 100-fold higher than for non-ID users. Major pathogen - S. aureus in 76%, Strep 13% & Staph unspecified 7%.
Comment: Author presents 3 cases including two with HIV.
Comment: The authors report a prospective study of cloxacillin vs vancomycin x 14d for S. aureusendocarditis in IDUs. All patients also received aminoglycosides. Cloxacillin proved superior to vancomycin with cure rates of 100% & 60%, respectively.
Comment: The authors conclude that there is no unifying hypothesis to explain this clearly established association.
Comment: This is one of the few published prospective controlled trials of treatment of endocarditis. The attempt was to find an adequate oral agent for IDU with S. aureusendocarditis. The authors showed cipro + rifampin x 4wks was as good as nafcillin + gent x 4wks; the side effects clearly favored the oral regimen. The practical use of this oral regimen has been confounded by the increasing resistance of S. aureus to fluoroquinolones and the issue of reliability/compliance of this population to an oral regimen.
Comment: The authors review 144 cases of IE in injection drug users including 45 with HIV S. aureus accounted for 65% of cases. Mortality was higher with CD4 count <200 (but this is from the pre-HAART era).
Comment: Duke criteria as basis for diagnosis of endocarditis.
Comment: Injection drug users with S. aureus tricuspid valve endocarditis were treated with cloxacillin 2gm q 4h IV + amikacin 7.5mg/kg q 12h x 2 wks. Among 72, 67 were cured, 4 needed longer course & 1 died in the hospital. This is the largest study of the 2 wk regimen.
Comment: The authors from Detroit with what appears to be the greatest experience with tricuspid valve surgery in IDUs with endocarditis review 61 cases. 1)Tricuspid valve replacement 100% mortality due to inability to control infection, reinfection due to IDU, or inability to comply with anticoagulation; 2) Tricuspid valvectomy (55 pts): acute mortality 11%; 9/10 who died 6mo-13 yrs later died from complications of injection drug use.
Comment: The distribution of valves in 67 addicts with endocarditis was tricuspid in only 27 (40%) - low than usually reported. Pathogens were S. aureus - 58% & strep species 25%.
Comment: The authors present an alternative to tricuspid valve removal for refractory endocarditis in addicts. Vegetectomy with valvuloplasty is now frequently preferred.
Comment: A famous prospective study that compared Nafcillin x 4wks + gent x 2 wks vs. nafcillin alone in non-addicts with S. aureusendocarditis. Combination treatment showed more rapid sterilization of blood cultures, but no difference in outcome by clinical parameters. The authors concluded gentamicin did not add significantly; nevertheless, most authorities (and the AHA Council on endocarditis) advocate 3-5days of an aminoglycoside.
Comment: The authors provide a follow-up of patients who underwent tricuspid valvectomy. Most tolerated the lack of a tricuspid valve in the early post-op period, but many eventually required valve replacement at a second procedure due to refractory right heart failure.
Comment: The authors draw from their experience with P. aeruginosaendocarditis in addicts in Detroit. Their recommendations are provided here - high dose tobramycin (8mg/kg/d) to achieve peak levels of 15-20ug/ml combined with an anti-pseudomonal penicillin or ceftazidime.
Official website of the Johns Hopkins Antibiotic (ABX), HIV, Diabetes, and Psychiatry Guides, powered by Unbound Medicine. Johns Hopkins Guide App for iOS, iPhone, iPad, and Android included. Complete Product Information.