POET

Published

Iversen K, et al. "Partial Oral versus Intravenous Antibiotic Treatment of Endocarditis". The New England Journal of Medicine. 2019. 380(5):415-424.
PubMed • Full text • PDF

Clinical Question

Among patients with left sided infective endocarditis from a strep species, E. faecalis, S. aureus, or coag-neg staph, is switching to oral antibiotics to complete therapy after at least 10 days IV safe and as effective as continued IV antibiotics?

Bottom Line

In patients with left-sided infective endocarditis from a strep species, E. faecalis, S. aureus (0% MRSA), or coag-neg staph, partial oral antibiotic treatment after stabilization and clinical response to IV treatment is non-inferior to treatment with only IV antibiotics. Please note that participants were highly selected (only 1% with IV drug use).

Major Points

Infective endocarditis (IE) is typically medically treated with prolonged bactericidal IV antibiotics. Prolonged IV therapy is expensive and requires complex outpatient care coordination or extended hospitalizations. Transitioning to PO antibiotics after initial stabilization with IV antibiotics would simplify management but this approach had not been evaluated in a large clinical trial.

The 2019 Partial Oral versus Intravenous Antibiotic Treatment of Endocarditis (POET) study randomized 400 Danish patients with left-sided IE known to be caused by typical organisms (a strep species, E. faecalis, S. aureus, or coag-neg staph,) to full IV antibiotic therapy or initial IV antibiotics followed by PO antibiotics. To avoid subtherapeutic antibiotic serum levels, only patients with “clinically normal gastrointestinal uptake” were enrolled. Oral antibiotic regimens each contained 2 highly bioavailable antibiotics from different drug classes, with different antimicrobial mechanisms and metabolization. The oral antibiotic regimen was non-inferior to the IV-only approach for the primary outcome all-cause mortality, unplanned cardiac surgery, clinically evident embolic events, or relapse of bacteria with primary pathogen (9.0% PO vs. 12.1% IV; non-inferiority P for difference in proportions=0.40).

This trial has several important limitations. The generalizability may be limited as the patients in the oral group treated as outpatients were seen frequently (2-3x a week), which may not reflect real-world monitoring, and only 5/400 (1%) patients were IV drug users. Participants thought to be poorly compliant were excluded. The trial did not exclude MRSA but none of the participants had this pathogen. Included organisms represent about 70-75% of all IE. This was the first large clinical trial examining the efficacy of oral antibiotic treatment for left-sided IE and seems to confirm what had been suggested by previous literature: that some patients with IE with working GI tracts capable of absorbing highly bioavailable antibiotics, can be partially treated with oral therapy.

Of note, the 2019 OVIVA study assessed the efficacy of oral antibiotics for bone and joint infection.

Guidelines

As of April 2019, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, open-label non-inferiority trial
• N=400 patients
  • PO treatment (n=201)
  • IV treatment (n=199)
• Setting: Danish patients referred to a cardiac center for suspected IE
• Enrollment: 2011-2017
• Mean follow-up: 6 months after ABx completed
• Analysis: Intention-to-treat
• Primary outcome: All-cause mortality, unplanned cardiac surgery, clinically evident embolic events, or relapse of bacteremia with the primary pathogen

Population

Inclusion Criteria

• ≥18 years old
• Left-sided endocarditis fulfilling the modified Duke criteria (native or prosthetic valve)
• 1 of these specific microorganism:
  • Streptococcus
  • Enterococcus faecalis
  • Staphylococcus aureus
  • Coagulase-negative staphylococci (CoNS)
• ≥ 10 days of appropriate IV antibiotic treatment overall
  • If valve surgery, ≥7 days of appropriate parenteral treatment after procedure
• T <38.0°C for >2 days
• A C-reactive protein (CRP) decrease to less than 25% of peak value OR <20 mg/L, and WBC count <15 x 10⁹/L during antibiotic treatment
• TTE and TEE performed within 48 hours of randomization

Exclusion Criteria

• Abscess formation demonstrated with transesophageal echocardiography
• BMI >40 kg/m²
• Concomitant infection requiring intravenous antibiotic therapy
• Inability to give informed consent to participation
• Suspicion of reduced absorption of oral treatment due to abdominal disorder
• Reduced compliance

Baseline Characteristics

Characteristics for PO treatment.

• Demographics: Age 68 years, female sex 21%
• Comorbidities: DM 15%, renal failure 10% (HD 8%), COPD 4%, liver disease 3%, cancer 9%, IVDU 1%
• Lab results at randomization:
  • WBC: 7.2
  • C-reactive Protein: 19.9 mg/L
  • Creatinine: 141 micromol/L
• Pathogen:
  • Streptococcus: 46%
  • E. faecalis: 25%
  • Methicillin-Sensitive S. aureus (MSSA): 23%
  • Coag negative staph: 6%
  • Methicillin-Resistant S. aureus (MRSA): 0%
• Prosthesis/Implants:
  • Prosthetic heart valve: 27%
  • Pacemaker: 10%
    • Pacemaker endocarditis: 4.0% of entire group
• Valves involved:
  • Mitral-valve endocarditis: 35%
  • Aortic-valve endocarditis: 54%
  • Mitral + aortic-valve endocarditis: 10%
• Valve surgery during current disease course: 38%

Interventions

• Participants were randomized to a group:
  • PO treatment - With initial IV antibiotics
    • Completion of antibiotic therapy with a regimen of 2 antibiotics of different classes, based on sensitivities and MIC’s of cultured organisms. Serum levels of antibiotics in this arm were checked at Day 1 and 5 of therapy.
    • Most common regimens: Dicloxacillin and Rifampicin (RIF), Amoxicillin (AMX) and Moxifloxacin, AMX + RIF, AMX + Linezolid.
    • Specifics about the PO regimens appear in table S2 (page 17) of the supplementary appendix.
    • At patients' preference and with attending physician's discretion, patients in the oral group were discharged to follow-up in outpatient clinics 2-3 times a week until completion of therapy.
  • IV treatment
    • Completion of antibiotic therapy with continued IV antibiotics.
• Transesophageal echocardiogram (TEE) was done in all patients 1 week prior to completion of treatment. After completion of therapy, all patients were seen in clinic follow-up at 1 week, 1, 3, and 6-months.

Outcomes

Comparisons are PO therapy vs. IV therapy.

Primary Outcomes

All-cause mortality, unplanned cardiac surgery, clinically evident embolism, and relapse of bacteremia

9.0% vs. 12.1% (OR 0.72, 95% CI, 0.37 to 1.36)

3.1% difference (95% CI, -3.4 to 9.6, P non-inferiority=0.40)

Component Analysis

All cause mortality

3.5% vs. 6.5% (HR 0.53, 95% CI 0.21 to 1.32)

Unplanned cardiac surgery

3% vs. 3% (HR 0.99, 95% CI 0.32 to 3.07)

Embolic event

1.5% vs. 1.5% (HR 0.97, 95% CI 0.20 to 4.82)

Relapse of positive blood culture

2.5% vs. 2.5% (HR 0.97, 95% CI 0.28 to 3.33)

Post-hoc Analysis

Median length of stay in hospital after randomization

3 days vs. 19 days, P<0.001

Subgroup Analysis

There was no differences in the primary outcome by age, sex, diabetic status, kidney disease status, pathogen type, surgical management, prosthetic vs. native valve type, or involved valve.

Adverse Events

Adverse effects from antibiotic therapy

5% vs 6%, P=0.66

Details of this are in table S14 of the supplementary appendix.

Criticisms

• Only patients with left-sided IE were enrolled.
• 25-30% of patients with IE caused by bacteria other than the ones listed in the inclusion criteria were excluded.
• Few patients with IV drug use were enrolled (n=5/400)
• Patients were referred into the study by other physicians (therefore potentially creating referral bias).
• Patients in the oral group treated as outpatients were seen frequently (2-3x a week), which may not reflect real-world monitoring.
• No patients with methicillin-resistant Staphylococcus aureus (MRSA) IE were enrolled in this study
• Only 20% of the screened population was randomized

Funding

Trial was funded with grants from the Danish Heart Foundation, the Capital Regions Research Council, the Hartmann’s Foundation, Svend Aage Andersens Foundation, and the Novo Nordisk Foundation (grant).