The extracorporeal elimination of a pathogen or damage-associated molecular pattern via blood purification techniques is increasingly being used in patients with septic shock and other clinical conditions characterized by a life-threatening inflammatory response. The removal of these substances can be accomoplished by means of ultrafiltration or hemoadsorption. Independently from the blood putification technique used, they could also affect the clearance of antibacterial and antifungal agents with a potentially significant clinical impact.
1. Introduction
Even in the absence of randomized clinical trials (RCTs) fulfilling the Evidence-Based Medicine (EBM) criteria or recommendations provided by the recently issued guidelines of the Surviving Sepsis Campaign (SSC)
[1][2], the extracorporeal removal of pathogen or damage-associated molecular patterns (PAMPs and DAMPs, respectively) via techniques of blood purification (BP) is increasingly used to treat patients with septic shock or other clinical conditions characterized by an excessive inflammatory response
[3]. This non-EBM founded popularity can be ascribed primarily to the failure of other approaches aiming at the neutralization of these substances by means of specific antibodies or inhibitors; actually, the results of many RCTs performed to evaluate the effect of this approach did not confirm the promising results obtained in experimental and pre-clinical studies. Different factors could account for these findings, including (a) the heterogeneity of the underlying conditions and of the related frailties of patients commonly encountered in the real-world scenario; and (b) the array of septic mediators that constitute a network rather than a cascade, making the blockade of a single cytokine unable to neutralize those located downstream. Then, it was hypothesized that their extracorporeal removal using BP could constitute a valid alternative. Basically, this goal can be accomplished via (a) their removal from the bloodstream by ultrafiltration (UF) through a filter that has pores with a cut-off value to allow the passage of substances with a molecular weight (MW) of septic mediators (50–70 kD), or (b) their adhesion to the surface of a material able to adsorb them: hemoadsorption (HA)
[3][4].
Even if the UF and HA-based techniques share some similarities, such as the need for anticoagulation and a time-dependent decay of the clearance capabilities, the factors involved in the intensity of treatment are different and include (a) the volume of ultrafiltrate produced (Qf) per unit of time for the UF and (b) the volume of blood processed per unit of time the proxy of which is considered the blood flow (Qb).
Moreover, as both approaches are based on the interaction between the Qb and the filtering or sorbent material, it appears that a larger surface of contact or a longer duration of the procedure can be clinically relevant
[3]. As far as the UF is considered, it has been hypothesized that higher Qf values could be associated with an improved survival of septic shock patients treated with UF; however, despite the encouraging results of some studies
[5], a large RCT comparing elevated (70 mL/kg/h) with normal (35 mL/kg/h) Qf failed to confirm these findings
[6] and this approach has been substantially abandoned; however, patients treated with higher Qf demonstrated a significant loss of antibiotics
[7].
Whatever BP techniques are used, the possible influence on the pharmacokinetics (PK) of antibacterial, antifungal and antiviral agents constitutes a major concern. In fact, their removal is a recognized side-effect of UF, and a number of recommendations have been issued to adjust both the loading and maintenance doses
[8], but less is known about the possible effect of the HA-based techniques. Should this effect exist also for them, it could at least partially account for the conflicting results observed in clinical studies involving septic shock patients treated by this approach: actually, although the use of HA has been associated with a decrease of the catecholamine requests and a better outcome in some studies
[9][10][11][12][13], these findings have not been confirmed by other investigators
[14][15][16][17], and a recent meta-analysis involving 120 patients treated with HA failed to establish clear positive or detrimental effects
[18].
2. Principles of Hemoadsorption and Available Devices
HA can be performed either in a stand-alone mode or with Continuous Renal Replacement Therapy (CRRT) for patients with acute kidney injury (AKI) or with Extra Corporeal Membrane Oxygenation (ECMO); in the former case, it appears that the combination of the two techniques can exert additional effects on the removal of therapeutic agents. Three main HA techniques have been developed so far
[7][19]. The first approach is based on a cartridge containing multiple polystyrene fibers covered by immobilized polymixin (Toraymixin, Toray Industries, Tokyo, Japan) that binds the endotoxin molecules. Due to this characteristic, its use has been advocated in the treatment of septic shock caused only by Gram-negative germs.
The second consists of a filter containing a modified AN69 membrane (oXiris, Baxter, Meyzieu, France) able to adsorb endotoxin and remove by means of UF several septic mediators while providing CRRT.
The last uses a cartridge containing polystyrene and divinylbenzene microbeads (Cytosorb®, Cytosorbents Corporation, New Jersey, USA; Aferetica s.r.l., Bologna, Italy) with a large surface (~40,000 m2), where both hydrophobic pro- and anti-inflammatory mediators with a MW of 5–60 kD are absorbed. The efficacy of Cytosorb® is concentration-dependent, as substances present in large concentrations are removed more efficiently than those with lower levels. If needed, the Cytosorb® can also be used with CRRT/ECMO.
3. Practical Considerations
Many studies have been published on the effects of the Cytosorb
® and other HA-based techniques on septic shock patients. However, it is hard to draw definite conclusions since factors other than the BP can influence the outcome, including the timing of initiation, the intensity of the treatment, and the appropriateness of the antibiotic regimen. Actually, the use of HA could represent a double-edged sword. It appears that (i) different classes of anti-infective agents can be efficiently removed from the bloodstream; (ii) in many cases this effect is time-dependent, being maximal in the initial hours of the procedure when the binding sites are still largely unsaturated; and, more importantly, (iii) the combined effect of these two mechanism can determine a sub-optimal concentration of these substances just when their appropriate levels are keenly warranted. The SSC recommends appropriately prompt administration alongside a type of administration capable of maximizing the anti-infective effects (e.g., extended infusion for β-lactams) of antibiotics/antifungals
[2] as the cornerstone for treating sepsis and septic shock. This effect and the related consequences could be even more relevant should the cartridge be changed before its exhaustion to take advantage of its binding capabilities to septic mediators as suggested by some authors
[20].
Although precise indications are lacking, different approaches can be used to limit the risk and obtain a personalized antibiotic treatment. Although the bulk of the available data derives from experimental studies under conditions far from clinical, it is mandatory to know if, and in what amount, the used agent is removed (Table 1).
Table 1. Effect of HA with Cytosorb® on the PK of different anti-infective agents. The removal was considered significant when the concentration of the challenged agent dropped to sub-therapeutic levels.
Agent |
Mode |
Administration |
Removal |
Reference |
Clindamycin |
Associated with ECMO and CRRT |
Intermittent |
Not significant |
[21] |
Isavuconazole |
Associated with CRRT |
Intermittent |
Significant |
[22] |
Linezolid |
Associated with CRRT |
Intermittent |
Significant after 4 h |
[20] |
Meropenem |
Stand alone and associated with CRRT |
Intermittent |
Not significant |
[23] |
Teicoplanin + Vancomycin |
Stand Alone |
Intermittent |
Significant |
[24] |
Vancomycin |
Stand Alone |
Continuous Infusion |
Not significant |
[24] |
4. Conclusions
Different BP techniques are currently used to treat clinical conditions characterized by an overwhelming inflammatory reaction. It should be recognized that anti-infective agents can be removed along with the mediators; thus, the use of TDM and the adjustment of the dosing regimens of anti-infective drugs is warranted.
This entry is adapted from the peer-reviewed paper 10.3390/antibiotics11020180