Clinical Infectious Diseases

REVIEW ARTICLE

The Use of Long-term Antibiotics for Suppression

of Bacterial Infections
Molly Horne,1 Ian Woolley,1,2, and Jillian S. Y. Lau1,2,
1
Faculty of Medicine, Nursing, and Health Sciences, Monash University, Clayton, Victoria, Australia; and 2Monash Infectious Diseases, Monash Health, Clayton, Victoria, Australia

Suppressive antibiotic therapy is prescribed when a patient has an infection that is presumed to be incurable by a defined course of
therapy or source control. The cohort receiving suppressive antibiotic therapy is typically highly comorbid and the infections often
involve retained prosthetic material. In part due to a lack of clear guidelines regarding the use of suppressive antibiotics, and in part
due to the complex nature of the infections in question, patients are often prescribed suppressive antibiotics for extremely long, if

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

not indefinite, courses. The risks of prolonged antibiotic exposure in this context are not fully characterized, but they include
adverse drug effects ranging from mild to severe, the development of antibiotic-resistant organisms, and perturbations of the
gastrointestinal microbiome. In this narrative review we present the available evidence for the use of suppressive antibiotic
therapy in 4 common indications, examine the gaps in the current literature, and explore the known and potential risks of this
therapy. We also make suggestions for improving the quality of evidence in future studies, particularly by highlighting the need
for a standardized term to describe the use of long courses of antibiotics to suppress hard-to-treat infections.
Keywords. antibiotics; infection; prosthetic joint infection; antimicrobial stewardship; suppression.

Antibiotics are a familiar therapeutic agent for clinicians, and [5]—namely, periprosthetic joint infection (PJI), vascular graft
indeed for many patients themselves. Over 211 million antibi­ infection (VGI) and other vascular infections including infec­
otic prescriptions were written in the United States in 2021 [1]. tive endocarditis and mycotic aneurysm, cardiac-implantable
This review focuses on a subsection of antibiotic use: long-term electronic device infection (CIEDI), and osteomyelitis and oth­
antibiotics for infections thought to be incurable. Although this er orthopedic hardware–associated infections.
represents only a small proportion of overall antibiotic use, by The guidelines for these broad infections generally recom­
its nature it may represent an important opportunity for anti­ mend a combined medical and surgical approach to treatment
microbial stewardship. [6–8]. In some cases, definitive source control through surgery
A previous study by our group found that, of 3 broad indica­ is impossible and/or a patient may be too comorbid for surgical
tions for long-term antibiotics, the use of antibiotics to sup­ intervention. In such scenarios, the guidelines typically allude
press infections presumed to be incurable was supported by to SAT; however, currently, there are no specific recommenda­
the least robust evidence [2]. The other broad indications of tions for their use in any of the guidelines examined.
prophylactic and noninfective uses are outside the scope of A major challenge to curing infections involving prosthetic
this review. Long-term or indefinite antibiotics have been pro­ material is the ability for many bacteria to form biofilms.
posed as a treatment strategy in patients who are ill-suited to Bacterial biofilms are microscopic matrix that can adhere to
more invasive surgical methods [3]. This approach is not usu­ the surfaces of both living tissue and prosthetic hardware [9,
ally intended to cure the infection, but instead to improve 10] and can form rapidly following infection. Bacteria embed­
symptoms and prevent progression or relapse to a point of clin­ ded in biofilms on prosthetic material can evade detection
ical significance [4]. Our previous review identified 4 infection by conventional microscopy and culture [11]. The protective
groups that were treated most commonly with suppressive an­ features of a bacterial biofilm allow it to resist antimicrobial ac­
tibiotic therapy (SAT) in a major Australian hospital network tivity, making curative efforts to treat infections involving pros­
thetic hardware increasingly challenging [12]. Few antibacterial
agents can penetrate the biofilm; in fact, some beta-lactam an­
Received 14 September 2023; editorial decision 11 May 2024; published online 4 June 2024 tibiotics have been shown to induce the biofilm [13].
Correspondence: J. S. Y. Lau, Monash Infectious Diseases, Monash Medical Centre, 246
Clayton Road, Clayton 3168, Victoria, Australia ([email protected]).
An SAT may also be used for other miscellaneous infections.
Clinical Infectious Diseases® 2024;79(4):848–54 However, such indications are outside the scope of this review.
© The Author(s) 2024. Published by Oxford University Press on behalf of Infectious Diseases
Society of America.
This is an Open Access article distributed under the terms of the Creative Commons Attribution
METHODS
License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distri­
bution, and reproduction in any medium, provided the original work is properly cited.
Literature searches were conducted in November 2023. The da­
https://doi.org/10.1093/cid/ciae302 tabases searched were Ovid MEDLINE, PubMed, and Cochrane.

848 • CID 2024:79 (15 October) • Horne et al

The search strategy combined terms describing suppressive an­ failed treatment [39]. However, the trend of SAT patients
tibiotic therapy (prolonged OR long-term OR continuous OR with staphylococcal infections faring worse than those with in­
indefinite OR suppress* AND antimicrobial OR antibiotic), fections caused by other organisms is not universal. Both a 2020
with terms describing the indications described in our previous and 2022 study found that gram-negative infections were asso­
study: for PJI: PJI OR prosthe* OR joint OR arthroplasty OR re­ ciated with worse SAT outcomes than gram-positive infections
placement; for VGI: vascular OR graft OR endovascular OR en­ [28, 31]. The reason for this finding is not clear, although a re­
doprosthes* OR aort* OR endocar ditis OR mycotic; for CIEDI: view published in 2021 suggested that it may be due to the pau­
implant* OR cardiac OR electronic OR defibril* OR ventric* OR city of orally available antibiotics that target gram-negative
device; for OM/other implant infection: osteomyelitis OR om organisms [4].
OR bone OR spin* OR metalware. The ideal duration of SAT for PJI remains unclear. A 2020
Search results were limited to adult and human studies in study reported that the hazard rate of failing SAT did not sig­
English. Titles and abstracts were manually screened for rele­ nificantly decrease beyond a treatment duration of 12 months
vance to SAT. Studies that administered long courses of antibi­ [29]. Indeed, another study found that, while an SAT duration
otics with primary curative intent were excluded. of less than 90 days was associated with an increased risk of fail­

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

ure, longer treatment durations of 90–180, 180–365, and over
Periprosthetic Joint Infections 365 days were not associated with any change in failure rate
Staphylococcus aureus is most frequently described as the [19]. There is also anecdotal evidence in the literature of pa­
infecting organism in PJI in the literature, followed by tients stopping SAT after an extended duration and remaining
coagulase-negative Staphylococcus [14–30]. Beta-lactams and symptom free [16, 17, 26, 27, 39]. However, there are currently
tetracyclines are the most commonly prescribed antibiotics no definitive guidelines to determine a clinically relevant stop­
for SAT [15, 16, 18, 20, 22–27, 29, 31, 32]. ping point, nor a biomarker that may indicate that it is safe to
Historically, SAT for PJI was associated with poor outcomes cease SAT. Markers such as a normalized erythrocyte sedimen­
[14, 15], but success rates from the past decade have ranged be­ tation rate (ESR) have been used to guide SAT cessation or to
tween 60% and 93% [19–25, 27, 28, 31–34]. Many studies lack assess likely success of SAT as a therapeutic option [21]; how­
comparison groups, but in those that have a non-SAT control ever, there have been instances in the literature of patients with
population SAT is shown to reduce the risk of infection recur­ a normal ESR having infection recurrence between 3 days and
rence, although not always to a level of statistical significance 1 month after stopping SAT [16, 39]. Some authors have also
[29, 32, 35]. In particular, SAT used in the context of PJI suggested C-reactive protein (CRP) levels, serum bactericidal
with draining fistula was not found to significantly improve titers greater than 1:8, and radiolabeled leukocyte scintigraphy
the risk of implant retention [30]. as possible methods for monitoring response to SAT [16, 40],
Several published literature reviews explore the use of SAT in but these strategies have not been extensively documented in
PJI, often as a follow-on strategy after surgical debridement and the literature.
implant retention (DAIR) [4, 27, 36, 37]. These reviews typical­ As previously mentioned, the heterogeneity of patient popula­
ly find favorable success rates ranging from 66% to 75%, but tions between, and often within, studies makes it difficult to pro­
they acknowledge the difficulty in assessing the efficacy of vide recommendations for clinical practice. Acknowledging this
SAT due to the heterogeneity of the literature at present— limitation, and in light of a recent narrative review of SAT in PJI
namely, variations between medical and surgical management [36], we suggest the following practices when managing patients
prior to commencement of SAT, definitions of what therapy ac­ on SAT for PJI:
tually constitutes SAT, and characterization of treatment suc­
cess or failure. 1. Consider SAT cessation after 1 year of continuous therapy, if
In the SAT literature, patients with staphylococcal infections the inflammatory markers, particularly ESR have normalized.
(especially S. aureus) appear to have a worse prognosis than 2. Patients with staphylococcal infections should be treated
those without. Several studies have reported worse outcomes more conservatively, with longer duration of therapy, and
in patients with periprosthetic joint infections caused by S. au­ a more cautious approach to ceasing SAT, due to the appar­
reus compared with those caused by other bacteria [16, 17, 21, ent worse prognosis of staphylococcal infections in the
26, 29, 33, 38]. Although not directly comparable, other studies literature.
have found that SAT can be very effective in streptococcal PJI.
A 2019 study of patients with PJI caused by streptococci found Vascular Graft Infections
a 93% treatment success rate with SAT, compared to a 57% suc­ Most of the literature on SAT for VGI comes from case series
cess rate with patients not treated with SAT (P = .002) [32], al­ and reports [41–49]. These case reports are published presum­
though an older study with very small numbers found that 50% ably because they have unique etiology, not reflective of the ma­
(3/6) of patients with beta-hemolytic streptococcal infections jority of VGIs requiring suppressive therapy. These case reports

Suppressive Antibiotic Therapy • CID 2024:79 (15 October) • 849

describe patients with VGI caused by organisms such as Brucella In contrast, a 2021 retrospective study of 69 patients with
melitensis [43], Edwardsiella tarda [44], and Capnocytophaga superficial driveline infections, of whom 43 received SAT,
canimorsus [45]. Beta-lactams are the most prescribed antibiotic found that SAT did not significantly reduce the risk of infection
agent for suppressive therapy in this context [42, 45, 50]. In sev­ relapse. Two earlier studies of patients receiving SAT for
eral studies, the antibiotics used for long-term suppression are LVAD-associated infections found similar failure rates of
not specified. 29% and 31% after an average treatment duration of 5.8 months
The success rate of SAT for VGI reported in retrospective and 1.6 years, respectively. Interestingly, the study that found a
studies is variable, although it appears that computed tomogra­ 29% failure rate in patients receiving SAT found that only 11%
phy (CT)/positron emission tomography (PET) imaging may of patients not receiving suppressive antibiotics experienced in­
have utility in determining the need for ongoing SAT. A 2021 fection relapse. It is possible that the SAT cohort in this study
retrospective cohort study investigated the outcomes of 32 pa­ had a substantially worse initial prognosis and were prescribed
tients with infective endocarditis who did not undergo an indi­ long-term antibiotics because they were considered less likely
cated surgical procedure. In this study 24 patients received SAT to be successfully treated.
for a median duration of 277 days and only 4 patients receiving In terms of survival rate, a 2017 study found that patients

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

SAT experienced infection relapse. The decision to cease SAT with CIEDI on SAT only survived for an average of 1.43 years.
in 9 patients was guided by CT/PET imaging results, and all Patients in this study did not have their CIEDs removed. In
of these patients remained symptom free during follow-up. contrast, a 2022 study found an overall 5-year survival rate of
The estimated overall survival rates in this study were 78% at 92% for patients with LVAD infection who had received SAT
1 year and 62% at 3 years. A 2023 report of 3 cases of patients until heart transplant. In this study, the patients receiving
with prosthetic valve infective endocarditis also found that SAT SAT for LVAD infections did not have worse post–heart trans­
could be safely ceased when guided by PET results [50]. plant outcomes compared with those with resolved or no
In terms of identifying patients at higher risk of failing LVAD infections. This retrospective study only investigated
SAT, it is unclear if surgical management prior to SAT alters patients who had received heart transplants, and accordingly,
prognosis. A study published in 2014 found that all 10 patients patients who were not candidates for transplant (due to comor­
managed with SAT without surgical intervention did not expe­ bidity or low likelihood of a positive outcome) were excluded,
rience reinfection during the first year of treatment, although 3 which may explain the unusually high SAT-associated success
patients had infection relapse after the first year of treatment. rate compared with previous studies.
An earlier study of patients receiving SAT but no surgical man­ Mortality is high for patients with LVAD infection on SAT.
agement, predominantly for VGI and prosthetic valve infec­ At present, we are unable to ascertain when and if SAT can be
tion, found that 51% of participants were still receiving SAT safely stopped, especially in patients with retained devices.
after 1 year. In contrast, a 2015 study of patients with mycotic Device retention is associated with high mortality, but SAT
aneurysm who had surgical management along with SAT may delay failure. In a highly comorbid cohort with limited
found that 71% of participants were still alive at 5 years. life expectancy, a longer relapse-free interval is a potentially sig­
Another study found a somewhat similar 5-year survival rate nificant consideration.
of 59% ± 8%. Patients in this study also had extensive surgical Considering this evidence, we suggest that the follow-up and
management; their aortic graft enteric erosions or fistulae were management of patients on SAT for CIEDIs be tailored to each
repaired with rifampin-soaked grafts and omental coverage be­ individual with close monitoring of biomarkers and surveil­
fore they were commenced on SAT. lance for relapse.

Cardiac Implantable Electronic Device Infections Osteomyelitis and Spinal Hardware Infections
The majority of the SAT literature for CIEDIs is focused on pa­ Apart from PJI, there is very limited evidence for the use of sup­
tients with left-ventricular assist device (LVAD) infections. The pressive antibiotics in osteomyelitis. A review published in
organisms identified in CIEDI requiring long-term antibiotic 2022 provided general recommendations for the diagnosis
suppression mirror those of PJI and VGI; staphylococcal species and treatment of acute and chronic osteomyelitis. This review
are the most common infecting organism. Interestingly, doxycy­ suggested SAT of greater than a 3-month duration may im­
cline was identified as the most common SAT agent more fre­ prove treatment outcomes in patients with implant-associated
quently than beta-lactam agents. osteomyelitis. In the reviewed literature, S. aureus was the most
The literature on suppressive antibiotics in CIEDI focuses on common infecting organism, and beta-lactams, doxycycline,
patients with ventricular assist devices [48]. Most of this litera­ and trimethoprim-sulfamethoxazole were the most used anti­
ture is in the form of small retrospective cohort studies. A 2020 biotic agents for SAT.
retrospective cohort study of 24 patients examining the use of Suppressive antibiotic therapy appears to be associated with
SAT for CIED driveline infections found a 50% success rate. good outcomes in early spinal implant–associated infection

850 • CID 2024:79 (15 October) • Horne et al

Table 1. Overview of the Current Suppressive Antibiotic Therapy Literature

Prospective Cohort Study Case-Control Study Retrospective Cohort Study Case Series Case Report
PJI 3 [17, 32] 1 [33] 18 [14, 18–25, 28, 29, 31, 35, 40] 6 [15, 16, 38, 39] 1
VGI 0 0 7 6 [41, 42, 47, 50] 7 [43–46, 49]
CIEDI 0 0 6 1 1 [48]
OM 0 0 3 2 1
Data are presented as number of studies with references in brackets.
Abbreviations: CIEDI, cardiac implantable electronic device infection; OM, osteomyelitis; PJI, prosthetic joint infection; VGI, vascular graft infection.

treated with debridement and retention. A retrospective cohort had debridement and implant retention after treatment has
review found that SAT improved 2-year survival in patients continued for 6 months to 1 year. As with other indications, pa­
with spinal-hardware infections treated with DAIR, compared tients should be closely followed in the first few months after
with those who had DAIR but did not receive SAT (80% vs SAT is stopped.

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

33%). The authors also found that SAT reduced the risk of A summary of the current SAT literature is presented in
treatment failure in patients with infection occurring within Table 1.
30 days of spinal implant, but not in late-onset infection.
Similarly, a 2016 case series found that patients with early-onset Adverse Effects Associated With Suppressive Antibiotic Therapy
multidrug-resistant spinal surgical site infections achieved clin­ Cross-sectional cohort studies typically found high levels of
ical and biochemical resolution of infection after debridement adverse effects among the SAT patient cohort. Two recent
and intravenous and long-term oral antibiotics. The patients in studies investigating SAT for variable indications found ad­
this series typically developed their infections within 3 weeks of verse effect rates of 41% and 52%. The most common adverse
their initial spinal surgery and underwent debridement within effects were gastrointestinal disturbances and asymptomatic
3 days of infection diagnosis. bacteriuria. Rates of adverse effects in retrospective studies
Longer durations of SAT do not appear to have increased are not always discussed in detail, and at times, a list of caus­
survival benefit. A retrospective cohort study investigating ative agents and a list of adverse effects are described, but the
the use of SAT in patients with a variety of orthopedic prosthe­ agents responsible for the particular adverse effects are not
ses (55% of patients had spinal hardware infections) found that elucidated. Including the antibiotic and the adverse effect at­
a 3-month duration of SAT was associated with treatment tributable to it would be beneficial for treating clinicians and
success, but a treatment duration of 6 months was not. patients receiving SAT.
Interestingly, a 2020 retrospective cohort study found that con­ A cross-sectional cohort study that assessed patient experi­
tinuation of suppressive antibiotics beyond 12 months was in­ ences of long-term antibiotic therapy, the majority of whom
versely associated with successful treatment. Indeed, a study were receiving suppressive therapy, found that, although 41%
found that 10 of 22 patients initially treated with SAT for a me­ experienced adverse effects, 72% of participants still reported
dian of 468 days had not experienced infection recurrence after complete adherence to their antibiotic therapy. This suggests
a median follow-up of 872 days. Of the remaining 12 patients, that the fear of severe infection recurrence compels patients
5 experienced failure while on treatment and were still receiv­ to continue their antibiotics despite the side effects, or that
ing SAT at the time of last follow-up. From these studies it is the side effects are not significantly bothersome. There are
possible that SAT improves patient survival in the first few many reports of patients switching or ceasing their SAT regime
months after infection, but it is unclear if continuing SAT be­ due to adverse effects; however, the impact on quality of life is
yond 6 months to 1 year is beneficial. generally not described [15, 19–26, 28, 29].
The literature on the use of suppressive antibiotics in other sub­ Adverse effects associated with SAT for various indications
types of osteomyelitis is even more limited than spinal-implant in­ are shown in Table 2.
fections. A retrospective cohort study published in 2015 examined
the use of SAT in patients with foot or long-bone osteomyelitis. Antimicrobial Resistance and the Microbiome in Suppressive Antibiotic
This study reported that 60% of patients had successful suppres­ Therapy
sion of their infections with long-term antibiotics. The duration of The potential for SAT to promote the development of antibiotic
SAT in the study ranged from 30 to 466 days. Most infections resistance has been identified as a concern in the literature as
in this study were polymicrobial, and most patients received early as 1998 [16]. Of the available literature, the most accurate
doxycycline. picture of antimicrobial resistance present in the SAT popula­
From this literature we make the following recommenda­ tion likely comes from a 2022 5-year longitudinal cohort study.
tion: SAT cessation may be considered in patients who have This study found that 56% of SAT patients were colonized with

Suppressive Antibiotic Therapy • CID 2024:79 (15 October) • 851

Table 2. Adverse Effects Associated With Suppressive Antibiotic therapy

Cost/ Neurotoxicity/
Clostridioides Dosing GI Malaise/ Peripheral Skin Recurrent Renal
a
Allergy Cytopenia difficile Schedule Disturbanceb Weariness Neuropathy Disturbancec Infection Failure
Amoxicillin— … … 2 [29] … 1 [21] … … … … …
clavulanic
acid
Amoxicillin 1 [24] … … … 1 [24] 1 [21] … 1 [32] … 1 [24]
Cefadroxil … … … … … … … … 1 [15] …
Cefalexin … … … … … … … 1 [15] 1 [15] …
Dicloxacillin … … … … 2 [15, 16] … … 2 [15, 16] 1 [15] …
Imipenem … … … … … 1 [24] … … … …
Ciprofloxacin … 1 … 1 [28] 2 … … … … …
Moxifloxacin … … … … 1 [21] … … … … …
Ofloxacin … … … … … … 1 [24] … … …

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

Clindamycin … … 1 [32] … 2 [21] 1 [24] … … … …
Doxycycline … … 161 1 [28] 4 [24, 26, 28] … … 2 [24, 26] … …
Minocycline … … … … 1 [21] … … 3 [18, 21, 40] … …
TMP-SMX … 1 [24] 1 [20] … 2 [26] … 1 [24] 2 … 3 [20,
24, 28]
Linezolid … 2 … … … … … 1 [28] … 1 [28]
Metronidazole … … … … … … 1 … … …
Rifampin … … … … 1 … … … … …
Pristinamycin 1 [24] … … … … … … … … …
Data are presented as number of studies with references in brackets. Further adverse events are summarized in Supplementary Tables a and b.
Abbreviations: GI, gastrointestinal; TMP-SMX, trimethoprim-sulfamethoxazole.
a
Cytopenia: anemia/hematotoxicity/thrombocytopenia.
b
GI disturbance: digestive intolerance/loss of appetite/nausea/vomiting/severe nausea/inappetence/vomiting/diarrhea.
c
Skin disturbance: phototoxicity/photosensitization/pruritus/rash/skin discoloration/hyperpigmentation/tooth discoloration.

multi-resistant bacteria, 30% of which had developed sub­ Suggestions for Improving the Clinical Applicability of Future Suppressive
sequent to SAT commencement. Retrospective studies Antibiotic Therapy Research

across the 4 SAT indications found resistance rates between A common theme raised by many SAT review papers is that the
3% and 40% [15, 19, 22, 23, 31]. Resistance against cipro­ quality of the current primary evidence makes it difficult to
floxacin, trimethoprim-sulfamethoxazole [15], rifampicin provide recommendations for clinical practice. Some potential
[19], and tetracyclines has been described in various studies strategies for future research include the following:
[22, 23].
Another potential adverse effect of SAT is disruption of the 1. Stratifying analyses so that patients with different factors
intestinal microbiome—the ecosystem of microorganisms such as surgical or medical management prior to SAT are
(predominantly bacteria) that reside on and in the body surfac­ analyzed separately. Studies that examine a strictly defined
es. Impacts of short-term antibiotics on the gastrointestinal mi­ demographic, such as a specific age criterion or having mul­
crobiome have been described previously, whereas the tiple previous prosthetic-associated surgeries, often find
consequences of long-term suppressive antibiotics have not lower success rates [26].
been researched extensively [2]. A 2020 study found that mi­ 2. Where possible, selecting comparison groups of patients not
crobiome diversity was lower than expected in some partici­ receiving SAT. This may provide a more realistic view of the
pants receiving SAT, but not to a level of statistical efficacy of SAT. We acknowledge that the patients in the
significance. Although it was not demonstrated in this study, non-SAT group are potentially not fully comparable to those
the possibility that SAT promotes the development or enrich­ receiving SAT—if they were as comorbid as the SAT cohort,
ment of antimicrobial resistance within an individual’s micro­ they presumably would have been prescribed SAT as well.
biome is still of concern. The relevance of microbiome 3. Designing studies with larger cohort sizes to increase the
distortion because of SAT will be elucidated over time as the re­ statistical power of the results. Studies that have achieved
lationship between microbiome composition and health is re­ higher cohort sizes are generally multicentered or use na­
searched more extensively. Although not common, tionally available databases to identify patients [24, 29–31].
Clostridioides difficile infection as a result of SAT has been re­ 4. Including patients who did not receive a minimum duration
ported throughout the literature [16, 29, 31, 32]. of SAT, usually due to death from their infection or another

852 • CID 2024:79 (15 October) • Horne et al

 cause as a separate early failure group. The practice of ex­ patient. In frailer patients, with limited life expectancy in
cluding these patients entirely likely overestimates the suc­ whom an infection recurrence may prove fatal, it may be rea­
cess rate of SAT by constructing a more resilient cohort sonable to continue SAT indefinitely.
with a better initial prognosis for the analysis.
CONCLUSIONS
Due to the lack of clarity in the suppressive antibiotic litera­
Currently, a wide range of terminology is used, including
ture, clinicians and patients may err on the side of caution and
“long-term suppressive therapy,” “indefinite chronic suppres­
continue SAT indefinitely, exposing both the patient and the
sion,” and “indefinite oral antimicrobial suppression.” We rec­
community to the known and potential risks of this therapy.
ommend “suppressive antibiotic therapy” as a standardized
term that accurately and succinctly describes the practice of us­
GENERAL RECOMMENDATIONS FOR PRACTICE
ing long courses of antibiotics to indefinitely suppress bacterial
REGARDING SUPPRESSIVE ANTIBIOTIC THERAPY
infection, so that relevant data are not unintentionally excluded
Selecting Patients for Suppressive Antibiotic Therapy from systematic reviews. As antimicrobial resistance and the

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

There are 2 broad categories of patients who may be started on levels of comorbidity in the population increase worldwide,
SAT. The first is in patients who have failed primary treatment the number of patients prescribed SAT will likely increase.
of their infection. Failure is characterized by persistence of clin­ Therefore, it is important that both patients and clinicians
ical infection or infectious biomarkers after the indicated sur­ have accurate information about the treatment strategy to pro­
gery and intravenous antibiotic course. mote safe patient-centered care.
The second group is patients who are not candidates for the
indicated surgical (with intravenous antibiotic) route. These Supplementary Data
patients are typically frail, highly comorbid, with limited life ex­ Supplementary materials are available at Clinical Infectious Diseases online.
Consisting of data provided by the authors to benefit the reader, the posted
pectancy. In these patients, the role of suppressive antibiotic
materials are not copyedited and are the sole responsibility of the authors, so
therapy is weighed against the risk of aggressive surgical inter­ questions or comments should be addressed to the corresponding author.
ventions, to provide more satisfactory patient outcomes.
The transition from a prolonged oral curative course to one Notes
with suppressive intent is not always clear; in our previous stud­ Financial support. The authors received no external funding for the re­
search, synthesis, and writing of this paper.
ies we have considered oral courses initially prescribed with an
Potential conflicts of interest. The authors report no conflicts of interest.
intended duration over 1 year to be suppressive therapy. All authors have submitted the ICMJE Form for Disclosure of Potential
Conflicts of Interest. Conflicts that the editors consider relevant to the con­
Monitoring Patients on Suppressive Antibiotic Therapy tent of the manuscript have been disclosed.
We recommend monitoring patients receiving SAT more fre­
References
quently in the initial year after commencing therapy, and
1. Centers for Disease Control and Prevention. Outpatient antibiotic prescriptions
then every 6 months or annually once the patient’s clinical —United States, 2021. Available at: https://archive.cdc.gov/www_cdc_gov/
and biochemical signs of infection have normalized. Evidence antibiotic-use/data/report-2021.html. Accessed 22 November 2022.
2. Lau JSY, Korman TM, Woolley I. Life-long antimicrobial therapy: where is the
suggests that patients are more likely to fail SAT early in the evidence? J Antimicrob Chemother 2018; 73:2601–12.
treatment course; thus, more frequent monitoring is warranted 3. Sandiford NA, Kendoff DO. Contemporary prosthetic joint infection manage­
ment: is there a role for suppressive antibiotic therapy? Arthroplasty Today
over this period [19, 29]. 2021; 10:133–4.
4. Cobo J, Escudero-Sanchez R. Suppressive antibiotic treatment in prosthetic joint
Duration of Suppressive Antibiotic Therapy infections: a perspective. Antibiotics (Basel) 2021; 10:743.
5. Lau JS, Kiss C, Roberts E, Horne K, Korman TM, Woolley I. Surveillance of life-
Some authors have demonstrated that continuing SAT beyond long antibiotics: a review of antibiotic prescribing practices in an Australian
1 year does not continue to reduce the risk of treatment failure. healthcare network. Ann Clin Microbiol Antimicrob 2017; 16:3.
6. Osmon DR, Berbari EF, Berendt AR, et al. Diagnosis and management of pros­
In our experience, cessation of SAT can be considered if the fol­ thetic joint infection: clinical practice guidelines by the Infectious Diseases
lowing criteria are met. The patient should have stable inflam­ Society of America. Clin Infect Dis 2013; 56:e1–25.
7. Chakfé N, Diener H, Lejay A, et al. Editor’s choice—European Society for
matory markers (eg, white blood cell count, ESR, and CRP)
Vascular Surgery (ESVS) 2020 clinical practice guidelines on the management
within normal reference ranges for at least 1 year, without clin­ of vascular graft and endograft infections. Eur J Vasc Endovasc Surg 2020; 59:
ical signs of infection such as ongoing pain or discharge from a 339–84.
8. Blomström-Lundqvist C, Traykov V, Erba PA, et al. European Heart Rhythm
prosthetic joint site. If appropriate baseline imaging exists, the Association (EHRA) international consensus document on how to prevent, diag­
infected site should be re-imaged to compare disease activity. nose, and treat cardiac implantable electronic device infections—endorsed by the
Heart Rhythm Society (HRS), the Asia Pacific Heart Rhythm Society (APHRS),
Finally, patient preference should be at the forefront of the de­ the Latin American Heart Rhythm Society (LAHRS), International Society For
cision to cease SAT. Patient factors to consider are how well the Cardiovascular Infectious Diseases (ISCVID) and the European Society of
Clinical Microbiology and Infectious Diseases (ESCMID) in collaboration with
patient has tolerated SAT, what is the patient’s life expectancy, the European Association for Cardio-Thoracic Surgery (EACTS). Europace
and how devastating infection recurrence would be for that 2020; 22:515–49.

Suppressive Antibiotic Therapy • CID 2024:79 (15 October) • 853

 9. Hall-Stoodley L, Stoodley P. Evolving concepts in biofilm infections. Cell 31. Escudero-Sanchez R, Senneville E, Digumber M, et al. Suppressive antibiotic ther­
Microbiol 2009; 11:1034–43. apy in prosthetic joint infections: a multicentre cohort study. Clin Microbiol
10. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of Infect 2020; 26:499–505.
persistent infections. Science 1999; 284:1318–22. 32. Renz N, Rakow A, Müller M, Perka C, Trampuz A. Long-term antimicrobial sup­
11. Atkins BL, Athanasou N, Deeks JJ, et al. Prospective evaluation of criteria for mi­ pression prevents treatment failure of streptococcal periprosthetic joint infection.
crobiological diagnosis of prosthetic-joint infection at revision arthroplasty. The Clin Microbiol Infect 2019; 79:236–44.
OSIRIS Collaborative Study Group. J Clin Microbiol 1998; 36:2932–9. 33. Siqueira MB, Saleh A, Klika AK, et al. Chronic suppression of periprosthetic joint
12. Sharma D, Misba L, Khan AU. Antibiotics versus biofilm: an emerging battle­ infections with oral antibiotics increases infection-free survivorship. J Bone Joint
ground in microbial communities. Antimicrob Resist Infect Control 2019; 8:76. Surg Am 2015; 97:1220–32.
13. Penesyan A, Paulsen IT, Gillings MR, Kjelleberg S, Manefield MJ. Secondary ef­ 34. Granger L, Mitchell P, Hutt J, Sandiford N. Is lifelong antibiotic suppression suc­
fects of antibiotics on microbial biofilms. Front Microbiol 2020; 11:2109. cessful in the management of prosthetic joint infection? Orthopaedic Proc 2018;
14. Johnson DP, Bannister GC. The outcome of infected arthroplasty of the knee. J 100-B(Suppl 17):7.
Bone Joint Surg Br 1986; 68:289–91. 35. Bryan AJ, Abdel MP, Sanders TL, Fitzgerald SF, Hanssen AD, Berry DJ. Irrigation
15. Tsukayama DT, Wicklund B, Gustilo RB. Suppressive antibiotic therapy in chron­ and debridement with component retention for acute infection after hip arthro­
ic prosthetic joint infections. Orthopedics 1991; 14:841–4. plasty: improved results with contemporary management. J Bone Joint Surg Am
16. Segreti J, Nelson JA, Trenholme GM. Prolonged suppressive antibiotic therapy for 2017; 99:2011–8.
infected orthopedic prostheses. Clin Infect Dis 1998; 27:711–3. 36. Cortes-Penfield N, Krsak M, Damioli L, et al. How we approach suppressive an­
17. Rao N, Crossett LS, Sinha RK, Le Frock JL. Long-term suppression of infection in tibiotic therapy (SAT) following debridement, antibiotics, and implant retention
total joint arthroplasty. Clin Orthop Relat Res. 2003;55–60. for prosthetic joint infection. Clin Infect Dis 2024; 78:188–98.

Downloaded from https://academic.oup.com/cid/article/79/4/848/7687610 by guest on 30 October 2024

18. Marculescu CE, Berbari EF, Hanssen AD, et al. Outcome of prosthetic joint infec­ 37. Malahias M-A, Gu A, Harris EC, et al. The role of long-term antibiotic suppres­
tions treated with debridement and retention of components. Clin Infect Dis sion in the management of peri-prosthetic joint infections treated with debride­
2006; 42:471–8. ment, antibiotics, and implant retention: a systematic review. J Arthroplasty
19. Peel TN, Buising KL, Dowsey MM, et al. Outcome of debridement and retention
2020; 35:1154–60.
in prosthetic joint infections by methicillin-resistant staphylococci, with special
38. Salmons HI, Bettencourt JW, Wyles CC, Osmon DR, Berry DJ, Abdel MP.
reference to rifampin and fusidic acid combination therapy. Antimicrob Agents
Irrigation and debridement with chronic antibiotic suppression for the manage­
Chemother 2013; 57:350–5.
ment of acutely infected aseptic revision total joint arthroplasties. J Arthroplasty
20. Prendki V, Zeller V, Passeron D, et al. Outcome of patients over 80 years of age on
2023; 38:2710–2715.e2.
prolonged suppressive antibiotic therapy for at least 6 months for prosthetic joint
39. Goulet JA, Pellicci PM, Brause BD, Salvati EM. Prolonged suppression of infec­
infection. Int J Infect Dis 2014; 29:184–9.
tion in total hip arthroplasty. J Arthroplasty 1988; 3:109–16.
21. Wouthuyzen-Bakker M, Nijman JM, Kampinga GA, van Assen S, Jutte PC.
40. Ceccarelli G, Perciballi B, Russo A, et al. Chronic suppressive antibiotic treatment
Efficacy of antibiotic suppressive therapy in patients with a prosthetic joint infec­
for staphylococcal bone and joint implant-related infections. Antibiotics (Basel)
tion. J Bone Joint Infect 2017; 2:77–83.
2023; 12:937.
22. Pradier M, Nguyen S, Robineau O, et al. Suppressive antibiotic therapy with oral
41. Pistolese G, Ippoliti A, Tuccimei I, Lorido A. Conservative treatment of aortic
doxycycline for Staphylococcus aureus prosthetic joint infection: a retrospective
graft infection. Eur J Vasc Endovasc Surg 1997; 14:47–52.
study of 39 patients. Int J Antimicrob Agents. 2017; 50:447–52.
42. Roy D, Grove DI. Efficacy of long-term antibiotic suppressive therapy in proven
23. Pradier M, Robineau O, Boucher A, et al. Suppressive antibiotic therapy with oral
tetracyclines for prosthetic joint infections: a retrospective study of 78 patients. or suspected infected abdominal aortic grafts. J Infect 2000; 40:184–7.
43. Bakhos CT, Gangadharan SP, Snyder GM, Wong MT, Hagberg RC. Management
Infection 2018; 46:39–47.
24. Prendki V, Ferry T, Sergent P, et al. Prolonged suppressive antibiotic therapy for of aortic brucellosis with infection of a descending thoracic aortic stent graft. Ann
prosthetic joint infection in the elderly: a national multicentre cohort study. Eur J Thorac Surg 2010; 89:2038–40.
Clin Microbiol Infect Dis 2017; 36:1577–85. 44. Ebisawa KF, Nishimura S, Yamamoto S, Ohji G, Iwata K. Mycotic aneurysm
25. Weston JT, Watts CD, Mabry TM, Hanssen AD, Berry DJ, Abdel MP. Irrigation caused by Edwardsiella tarda successfully treated with stenting and suppressive
and debridement with chronic antibiotic suppression for the management of in­ antibiotic therapy: a case report and systematic review. Ann Clin Microbiol
fected total knee arthroplasty: a contemporary analysis. Bone Joint J 2018; 100-B: Antimicrob 2018; 17:21.
1471–6. 45. Jalava-Karvinen P, Grönroos JO, Tuunanen H, Kemppainen J, Oksi J, Hohenthal
26. Leijtens B, Weerwag L, Schreurs BW, Kullberg BJ, Rijnen W. Clinical outcome of U. Capnocytophaga canimorsus: a rare case of conservatively treated prosthetic
antibiotic suppressive therapy in patients with a prosthetic joint infection after hip valve endocarditis. APMIS 2018; 126:453–6.
replacement. J Bone Joint Infect 2019; 4:268–76. 46. Alqallaf A, Alhashim A, Alajmi M, Alsaqobi A, Al-Adsani W. Coxiella burnetii
27. Sandiford NA, Hutt JR, Kendoff DO, Mitchell PA, Citak M, Granger L. Prolonged endocarditis in a patient with systemic lupus erythematosus: a case report of a di­
suppressive antibiotic therapy is successful in the management of prosthetic joint agnostic challenge. Am J Case Rep 2020; 21:e926699.
infection. Eur J Orthop Surg Traumatol 2020; 30:313–21. 47. Lechner AM, Pretsch I, Hoppe U, Seitelberger R, Dinges C. Successful long-term
28. Burr RG, Eikani C, Adams WH, Hopkinson WJ, Brown NM. Predictors of success antibiotic suppressive therapy in a case of prosthetic valve endocarditis and a case
with chronic antibiotic suppression for prosthetic joint infections. J Arthroplasty of extensive aortic and subclavian graft infection. Infection 2020; 48:133–6.
2022; 37:S983–8. 48. Radcliffe C, Grant M. Over 870 days of successful antibiotic suppression therapy
29. Shah NB, Hersh BL, Kreger A, et al. Benefits and adverse events associated with for VRE-infected left ventricular assist device. J Card Surg 2020; 35:1746–8.
extended antibiotic use in total knee arthroplasty periprosthetic joint infection. 49. Donabedian H. Long-term suppression of Salmonella aortitis with an oral antibi­
Clin Infect Dis 2020; 70:559–65. otic. Arch Intern Med 1989; 149:1452.
30. Lensen KJDF, Escudero-Sanchez R, Cobo J, et al. The efficacy of suppressive an­ 50. Valsecchi P, Calia M, Marvulli LN, et al. The possible role of serum bactericidal
tibiotic treatment in patients managed non-operatively for periprosthetic joint in­ titres in long-term suppressive antibiotic treatment for infective endocarditis: re­
fection and a draining sinus. J Bone Joint Infect 2021; 6:313–9. port of three cases. Infect Dis (Lond) 2023; 55:514–8.

854 • CID 2024:79 (15 October) • Horne et al