POET II reopens the question of how long infective endocarditis needs treatment
Infective endocarditis has always forced clinicians to balance two competing risks: overtreating patients with prolonged intravenous antibiotics or stopping too early and allowing infection to return. POET II pushes that debate into sharper focus. The trial’s central finding is stark: stopping antibiotics once patients are clinically stable shortened treatment duration without increasing death or the need for surgery, yet relapses of endocarditis or bacteraemia rose to nearly three times the rate seen with standard treatment.
That trade-off is not a minor statistical wrinkle. It goes directly to the biology of endocarditis, where infected vegetations can shelter bacteria from antimicrobial exposure and where apparent clinical improvement may precede microbiological eradication by days or even weeks.
Response-guided stopping moved beyond the oral-switch model
The earlier POET paradigm made many clinicians more comfortable with partial oral therapy in carefully selected stable patients. POET II goes further by asking whether antibiotics can be stopped altogether once a patient reaches clinical stability. This is a different therapeutic proposition, and it deserves to be judged on its own terms rather than as a simple extension of oral step-down.
The appeal is obvious. Endocarditis treatment is long, burdensome, and often tied to inpatient or closely supervised outpatient care. Reducing total exposure could lower catheter-related complications, medication toxicity, and the practical strain of prolonged treatment. In a disease that often affects older patients with multiple comorbidities, any strategy that trims unnecessary days of therapy has immediate bedside relevance.
Yet endocarditis is not pneumonia, and it is not a urinary tract infection. Vegetations, abscesses, prosthetic material, and embolic complications create sanctuaries where bacterial clearance can lag behind symptom control. A patient who looks clinically stable may still carry enough residual infectious burden to seed recurrence once treatment is withdrawn.
The relapse signal defines the study’s clinical caution
The most consequential result is the relapse signal. Even as stopping therapy did not increase mortality or surgery, the rates of relapse of endocarditis or bacteraemia were materially higher in the response-guided arm. For practicing cardiologists, infectious disease specialists, and endocarditis teams, this matters more than any convenience benefit. Relapse is not a benign endpoint. It can mean renewed bacteremia, repeat imaging, re-hospitalization, re-exposure to broad-spectrum therapy, and another period of uncertainty over whether the original infection was truly controlled.
The finding also reinforces a basic principle of endocarditis care: clinical stability is necessary, but it may not be sufficient, if the goal is durable microbiological cure. The host response can settle before the infected nidus is fully sterilized. In that setting, shortening therapy may expose the residual edge of infection that standard-duration treatment was still suppressing.
Why a tailored strategy may still have a place
Even with the relapse penalty, POET II does not close the door on individualized treatment duration. It instead sharpens the conditions under which tailored therapy might be defensible. The safety of stopping early will depend on rigorous patient selection, careful microbiological and clinical reassessment, and multidisciplinary follow-up that can detect recurrence before catastrophic complications develop.
There is also a plausible rationale for why a response-guided approach might work in some patients but not others. Endocarditis is not one disease in the microbiological sense. Organism virulence, valve involvement, prosthetic versus native infection, residual vegetations, and the presence of metastatic infection all influence bacterial persistence. A stable patient with a rapidly clearing bloodstream infection and no structural complication may have a very different relapse risk from a patient with more complex disease biology, even if both appear clinically improved at the same bedside assessment.
The challenge is that current stability criteria may not yet be sharp enough to discriminate between those two states with enough precision for routine practice.
What the trial means for everyday endocarditis management
For now, POET II argues against treating clinical improvement as a green light to stop antibiotics broadly across the endocarditis population. Standard-duration therapy still retains a strong logic when the objective is to minimize relapse. The new data do, however, create pressure to refine treatment duration rather than accept a single fixed course for every patient.
This is where endocarditis care is likely to evolve: not toward blanket shortening, but toward better risk stratification. Biomarkers, imaging, microbiological response, valve pathology, and clinical trajectory may eventually be combined to identify the subgroup in whom stopping earlier is truly safe. Until then, any attempt to abbreviate therapy should remain tightly controlled and reserved for settings where endocarditis teams can provide close surveillance and rapid rescue if bacteremia returns.
The clinical tension is familiar, but POET II gives it sharper edges. Fewer antibiotic days came with fewer treatment burdens, but also with more relapse. In infective endocarditis, that is not a marginal compromise; it is the central decision.