Assessment of Therapeutic Approaches: Necrotizing Soft Tissue Infections and the Choice Between Vancomycin + Clindamycin and Linezolid

Patients and Methods

Results

A total of 400 patients were identified utilizing ICD-10 diagnosis codes for necrotizing fasciitis (M72.6), Fournier’s gangrene (N49.3), and gas gangrene (A48.0). Following review, 158 patients were excluded; most commonly because they did not receive ICU care on admission, died or pursued hospice care within 72 hours, received surgical debridement at an outside hospital, or did not have a final diagnosis of NSTI (Fig. 1). Overall, 242 patients were included in the final analysis with 199 patients in the clindamycin group and 43 patients in the linezolid group.

OPEN IN VIEWER

FIG. 1.

Exclusion flowchart.

Patients were approximately 50 years old, predominantly male (59.3%) in the clindamycin group versus predominantly female (65.1%) in the linezolid group (p = 0.004). The number of patients who received vasopressors in the study was 132 (55%). Median body mass index (31.8 kg/m² vs. 31.9 kg/m²) and sequential organ failure assessment (SOFA) scores, calculated using the first available value during the admission for each factor, (6 vs. 5) on admission were similar between patients in the clindamycin and linezolid cohorts. There was no statistically significant difference observed in cefepime usage between the two groups (p = 0.232); however, there was substantially greater usage of piperacillin–tazobactam in the clindamycin group (p = 0.002) and substantially greater usage of meropenem in the linezolid group (p = 0.006). Other baseline demographics were similar between the two groups (Table 1).

Discussion

In this retrospective single-center study evaluating clindamycin-based versus linezolid-based regimens for empiric treatment in patients with NSTI, there was no observed difference in the number of surgical debridements between groups. While the number of surgical debridements is not solely dependent on antibiotic therapy, looking at antibiotic efficacy from this angle provides a unique perspective that has not been studied. The quantity of debridements was used as a surrogate marker for the standard course of patients being treated for NSTI. Increasing SOFA score and growth of S. anginosus, S. constellatus, S. intermedius, or Bacteroides spp. were independently associated with the need for more debridements. To our knowledge, these findings have not been noted in the literature previously. Our findings are consistent with the study by Dorazio and colleagues which found no difference in 30-day mortality between patients receiving vancomycin + clindamycin versus linezolid as empiric treatment of NSTI. ¹² Similarly, Lehman and colleagues found that patients treated with linezolid received an average of one less day of anti-MRSA therapy than those receiving vancomycin + clindamycin. ¹³ Our study did not assess 30-day mortality or duration of anti-MRSA therapy; however, our findings add to existing literature as there is a paucity of data comparing the effect of empiric antibiotic regimens on the outcome of time to source control.

Secondary outcomes, including time to first surgical debridement, time to source control, duration of antibiotic agents after source control, AKI incidence, ICU LOS, or total duration of antibiotic agents, were also found to not have statistical significance. Dorazio and colleagues assessed similar secondary outcomes and did find the composite outcome of death, AKI, or CDI within 30 days to be more common in patients receiving vancomycin + clindamycin. ¹² Similarly, Lehman and colleagues found that patients receiving linezolid experienced a lower incidence of AKI; however, they excluded patients who received vancomycin for greater than 48 hours before switching to linezolid. ¹³ These findings differ from our study which did not observe a difference in rates of AKI or use a composite outcome. At the authors’ institution, the area under the concentration versus time curve (AUC)-based monitoring is utilized, as it is well known that the risk of AKI increases as troughs are maintained above 15–20 μg/mL. ¹⁴ Similarly, there are recent data to suggest an increased risk of AKI as the daily AUC exceeds 600–800 mg·h/L. ¹⁵ It is important to highlight that our institutional guideline mentions linezolid as a recommended option in patients at high risk for AKI, which may bias this result. Our finding of no substantial difference in AKI is strengthened by the fact that potential confounders, including the coadministration of other known nephrotoxic agents, were screened and there was found to be no difference between cohorts.

These findings add to the growing volume of literature describing similar outcomes for patients with NSTI, in that linezolid was found to have similar outcomes for the treatment of NSTI, compared with vancomycin plus clindamycin containing regimens. A strength of our study is that we attempted to replicate real-world practice. Furthermore, the primary outcome that we assessed expands upon existing literature by providing an additional surrogate measure of efficacy to support the efficacy of linezolid for the treatment of NSTI.

Limitations of this study include the single-center retrospective study design and the disproportionate sample sizes between the two groups. The small number of patients in the linezolid group may be because of provider hesitancy to change practice. In addition, susceptibilities were not within the scope of this review other than reporting methicillin-susceptible Staphylococcus aureus versus MRSA, and there could be some differences between resistance of clindamycin and linezolid, as well as other various agents being used empirically. Second, AKI measurement was obtained via diagnosis codes rather than risk, injury, failure, loss of kidney function, and end-stage renal disease or acute kidney injury network criteria. The presence of pre-existing renal disease was not collected at baseline, as all patients are eligible to receive either treatment arm. With the present study being retrospective in nature, there is potential for incomplete or unclear documentation that could result in missed or incorrect diagnosis of AKI. In addition, many patients in the study were receiving piperacillin–tazobactam, which was historically believed to cause nephrotoxicity because of its effect on serum creatinine. Recently published literature suggests that this elevation in serum creatinine may not be relevant to clinical outcomes such as renal replacement therapy and mortality and may demonstrate that the effects on serum creatinine represent pseudotoxicity. ¹⁶ Further literature is needed to validate these findings. Nonetheless, patients in the present study may have received a diagnosis of AKI on this basis, which could affect the overall assessment of AKI rates. Third, the exclusion of patients who did not receive ICU level of care at admission may not encompass patients who upgraded to the ICU following admission to a floor team during their hospitalization. Fourth, the present study contained a small number of patients with GAS infections from obtained cultures. Previous studies highlighting the impact of antitoxin effect from antimicrobial agents included a majority of patients with GAS infections; however, similar benefit would be assumed to be afforded for any organism capable of toxin-mediated infection, including viridans group Streptococci and S. aureus. ¹⁷ Lastly, similar to any institution with a large group of surgical providers, variations in operative practice and antimicrobial selection may exist.

Conclusions

Patients treated with linezolid-based regimens underwent a similar number of surgical debridements compared with patients who received clindamycin-based regimens. This study supports prior findings that linezolid, in combination with broad-spectrum gram-negative and anaerobic coverage, may be considered as an effective first-line option in patients with NSTI who undergo rapid surgical debridement.