Clinical Review (insufficient quality of evidence to enable a Clear Recommendation):
Some observational studies suggest that administration of antibiotics prior to bone biopsy or surgical management may modestly decrease yield of bone cultures for patients with osteomyelitis, including DFO and PJI. Thus, presuming other microbiological methods (e.g., blood cultures) have not already established a microbial etiology, it is reasonable to consider deferring antimicrobial therapy initiation until bone/joint microbiological samples are obtained for clinically stable patients. However, other studies are not concordant, and histopathology results are unlikely to be affected by prior short-term antibiotics. Decisions regarding the delay of empiric therapy therefore balance potential harm due to the risk of progression of life-threatening infection (e.g., sepsis) or impending spinal cord compression against the potential benefit of microbiological data.
There are limited data on the yield of bone biopsies in osteomyelitis and the effect of pre-biopsy antibiotics on pathogen recovery. Overall, studies are retrospective in nature, small, and primarily focus on vertebral osteomyelitis. In addition, the durations and/or spectrums of pre-biopsy antibiotics were variable. Nevertheless, several studies suggest that the diagnostic yield of biopsy may be diminished in cases of vertebral osteomyelitis with antecedent antibiotic use.
In a retrospective study of 72 patients with confirmed vertebral osteomyelitis, of whom 40 underwent 46 CT-guided biopsies, culture positivity was significantly lower among patients who had been treated with antibiotics in the previous 48 hours (23% vs. 60%, p = 0.013).(de Lucas, Gonzalez Mandly et al. 2009) Similarly, in a case series of 20 patients with vertebral osteomyelitis, 8 of 20 (40%) patients received antibiotics before the biopsy, with only 2 of 8 (25%) growing an organism after antibiotic use, in comparison to 6 out of 12 (50%) cases in which an organism was isolated without antibiotic use.(Rankine, Barron et al. 2004)
However, other studies suggest that pre-biopsy antibiotics may not necessarily impact pathogen recovery.(Marschall, Bhavan et al. 2011, Agarwal, Wo et al. 2016, McNamara, Dickerson et al. 2017, Lopes Floro, Munckhof et al. 2018, Wong, Tarr et al. 2021) For instance, in a retrospective cohort study of 150 adult inpatients with hematogenous vertebral osteomyelitis conducted by Marschall et al., the association of pre-biopsy antibiotics, which was defined as any antibiotic exposure within 14 days prior to biopsy, with negative culture results was not statistically significant (adjusted odds ratio (OR) 2.3; 95% CI, 0.8-6.2; p = 0.1).(Marschall, Bhavan et al. 2011)
Similarly, in a retrospective multicenter study of 104 patients, Wong et al. studied the effect of stopping antibiotics prior to biopsy and found that it had no significant effect on culture positivity when compared to patients with or without pre-biopsy antibiotics.(Wong, Tarr et al. 2021) Of note, the authors did not provide data on the precise time the antibiotics were stopped prior to biopsy (e.g., holding for 2 hours vs. 24 hours prior to biopsy might differ in result). Furthermore, subgroup analysis from Lopez Floro et al. found that when comparing patients who received a single dose of an antibiotic with patients who received longitudinal antibiotics prior to biopsy, patients who had multiple doses prior to biopsy had statistically significant lower culture positivity (p = 0.004).(Lopes Floro, Munckhof et al. 2018) Thus, single doses of antibiotics pre-biopsy may be less likely to affect culture results than multiple doses. In addition, the match between the antibiotic used prior to the biopsy and the sensitivity profile of the organism can also negatively affect the culture positivity.
In sum, although results from multiple studies are inconsistent and the definitions of pre-biopsy antibiotics are not well-defined, several studies suggest that the culture positivity yield of biopsy may be diminished in cases of vertebral osteomyelitis with antecedent antibiotic use. Furthermore, for other diseases, such as bacteremia, receipt of antibiotics prior to culture generally reduces culture yields,(Cheng, Stenstrom et al. 2019) and it is therefore likely that the sensitivity of bone cultures is also reduced by antecedent antibiotics. However, no study demonstrated reduced sensitivity of histopathology results for the diagnosis of osteomyelitis with prior antibiotic therapy.
There is an overall paucity of data on the yield of bone biopsies in DFO and the effect of pre-biopsy antibiotics on pathogen recovery. Studies are retrospective in nature and small in size, with confounding variables and lack of standardization which make it difficult to compare them. In addition, several observational studies that focus on the microbiologic accuracy of bone biopsy excluded patients who received antecedent antibiotics within two weeks prior to bone biopsy, leading to overall minimal data being available on this subject matter.
Among three observational studies including patients who received antibiotics within two weeks of biopsy, the proportion of culture positive percutaneous bone biopsies (PBBs) was high, ranging from 83% to 99%.(Lesens, Desbiez et al. 2011, Letertre-Gibert, Desbiez et al. 2017, Couturier, Chabaud et al. 2019) However, these high yields should be interpreted with caution because samples were often collected through the ulcer bed, and thus may include colonizing/contaminating microbes rather than true pathogens. In contrast, in a study including 75 biopsies of non-vertebral bones with clinical concern for osteomyelitis, among patients who received antibiotics within 24 hours of the biopsy, only 24% of cultures were positive.(Wu, Gorbachova et al. 2007) This compared to a positivity rate of 42% among patients who did not receive antibiotics within 24 hours of biopsy, suggesting that pre-procedural therapy lowered culture sensitivity.
In a meta-analysis by Schechter et al., the proportion of patients who received antibiotics within two weeks prior to percutaneous bone biopsy (PBB) for a diagnosis of DFO ranged between 32% and 53%.(Schechter, Ali et al. 2020) In their analysis, they found studies that excluded patients who received antibiotics ≤2 weeks before the PBB reported positive cultures in 56% to 87% of patients, with a pooled (95% CI) positive culture rate of 72% (59%-83%).(Senneville, Melliez et al. 2006, Senneville, Morant et al. 2009, Aslangul, M'Bemba et al. 2013, Ducloux, Tazi et al. 2016, Schechter, Ali et al. 2020) By comparison, studies that included patients who received antibiotics ≤2 weeks before the PBB report higher culture positivity from PBB, ranging from 83% to 99% with pooled (95% CI) positivity of 96% (84%-99%).(Lesens, Desbiez et al. 2011, Letertre-Gibert, Desbiez et al. 2017, Couturier, Chabaud et al. 2019)
In other studies, duration and/or spectrum of pre-biopsy antibiotics were unclear. Aragon-Sanchez et al. reviewed 185 patients with osteomyelitis from 2002 to 2007 hospitalized in a diabetic foot unit.(Aragon-Sanchez, Cabrera-Galvan et al. 2008) Patients initially treated for dry necrosis that became secondarily infected were excluded. Preoperative diagnosis of osteomyelitis was based on PTB test through the ulcer and a radiological study of the foot. All patients without penicillin allergies were given ampicillin-sulbactam, with the first dose at the time of anesthesia induction. Bone culture was collected during surgical intervention. One hundred and thirty-two patients (71.3%) received antibiotics prior to admission. Only 20 cultures were negative; 154 specimens yielded an organism. The authors concluded that negative cultures were not related to previous antibiotic treatment (p = 0.1). However, the potential impact of dose and duration were not clearly documented.
In a multicenter, small RCT of 40 patients, the authors suggested delaying antibiotic administration until the availability of culture results in DFO did not affect clinical failure rates.(Tone, Nguyen et al. 2015) This study included patients with DFO treated non-surgically. Empiric therapy, mostly amoxicillin-clavulanate, was prescribed while waiting for culture results if the treating physician considered it necessary, which was the case in 18 patients (45%). For the remaining patients for whom empiric therapy was not given, antibiotic therapy was initiated a median of 14 days (range 5 to 19) after the bone biopsy. Antibiotics were given orally for the full treatment course for 22 patients (55%) or IV therapy was used for 5-7 days, then followed by oral therapy in 18 patients (45%). Patients with or without empiric therapy had similar failure rates (6/18, 33% vs. 8/22, 37%, respectively; p = 0.8).
In contrast, a retrospective multicenter study from France reported that bone culture-based antibiotic therapy was associated with higher remission rates [OR 4.8 (95% CI, 1-22.7), p = 0.04] in patients with non-surgically treated diabetic foot osteomyelitis.(Senneville, Lombart et al. 2008) Therefore, while the data are mixed, when possible, it may be desirable to delay initiation of antibiotic therapy for stable patients until bone or deep tissue culture can be obtained.
Observational studies have shown preoperative therapeutic antibiotics are associated with a decrease in intraoperative culture positivity in patients with PJI. In a retrospective, case-control study, 135 patients with culture-negative PJI were matched to 135 patients with culture positive PJI.(Malekzadeh, Osmon et al. 2010) The investigators reported that 64% of patients with culture-negative PJI and 25% of patients with culture-positive PJI received antibiotics within three months before the diagnosis of culture-negative PJI (OR 4.1; 95% CI 2.3-7.2). The median duration of prior antibiotic treatment for culture-negative PJI was 35 days vs. 18 days for culture positive patients. Cefazolin and ciprofloxacin were the most used antimicrobials (16% and 15% respectively). The study found that patients with culture negative PJI were more likely to have received antibiotics within three months of their diagnosis (OR 4.1; 95% CI, 2.3-7.2), suggesting that pre-culture therapy might reduce sensitivity.
Similarly, in a study of 182 patients with late PJI after total knee arthroplasty, in which 65 patients received antibiotics prior to aspiration, the authors found that patients with pre-aspiration antibiotic administration are more likely to have negative culture than those without antecedent antibiotics (26.4% vs. 12.9%; RR 2; 95% CI, 1.1-3.9; p = 0.046).(Shahi, Deirmengian et al. 2015) Overall, patients who received pre-aspiration antibiotics also had lower values for serologic and synovial markers for PJI.
Moreover, from a prospective trial of 331 patients with total knee or hip prostheses, in which 79 had PJI and 252 had aseptic failure, Trampuz et al. concluded that preoperative administration of antibiotics lowers the positive yield of both tissue and sonicate-fluid cultures from patients with PJI.(Trampuz, Piper et al. 2007) The tissue culture sensitivity decreased from 76.9% to 47.8% to 41.2% as the antimicrobial free interval before surgery decreased from greater than 14 days to 4-14 days to less than 3 days (p < 0.001). The same effect was observed in sonicated-fluid culture where culture positivity decreased from 82.1% to 87% to 58.8% as the antimicrobial-free duration trended lower, from greater than 14 days to 4-14 days to less than 3 days prior to surgery (p = 0.1). Of note, 9 of the 31 patients with negative tissue culture had antibiotic stopped greater than 14 days prior to surgery and 7 out of 9 patients had negative sonicate-fluid cultures.
In contrast, two recent retrospective studies suggested antimicrobial therapy prior to surgery may not negatively impact intraoperative culture positivity.(Stephan, Thurmer et al. 2021, Watanabe, Kobayashi et al. 2021) Therefore, future studies are needed to determine the optimal duration and/or the effect of withholding antimicrobial therapy prior to obtaining meaningful culture results for patients with PJI.
On balance, although the data are mixed, multiple studies across all types of osteomyelitis have suggested a modest decrease in biopsy culture positivity with prior antibiotics. However, none of these studies have suggested a reduction in positivity of histopathology. Hence the risk of prior therapy is to reduce confirmation of microbial etiology, limiting ability to tailor antimicrobial therapy. If other microbiological methods identify the etiologic pathogen, these considerations become superfluous. Furthermore, for patients who are clinically unstable or have serious or life or limb-threatening infections in addition to suspected osteomyelitis, the risk:benefit of waiting for biopsy to initiate therapy may not be favorable, in which case empiric therapy should be administered without delay.