Use of in vitro models of haemofiltration and haemodiafiltration to estimate dosage regimens for critically ill patients prescribed cefpirome

Vancomycin Sequestration in ST Filters: An In Vitro Study

BACKGROUND

Sequestration of vancomycin in ST^® filters used in continuous renal therapy is a pending question. Direct vancomycin-ST^® interaction was assessed using the in vitro NeckEpur^® technology.

METHOD

ST150^® filter and Prismaflex dialyzer, Baxter-Gambro, were used. Two modes were assessed in duplicate: (i) continuous diafiltration (CDF): 4 L/h, (ii) continuous dialysis (CD): 2.5 L/h post-filtration.

RESULTS

The mean initial vancomycin concentration in the central compartment (CC) was 51.4 +/- 5.0 mg/L. The mean percentage eliminated from the CC over 6 h was 91 +/- 4%. The mean clearances from the CC by CDF and CD were 2.8 and 1.9 L/h, respectively. The mean clearances assessed using cumulative effluents were 4.4 and 2.2 L/h, respectively. The mean percentages of the initial dose eliminated in the effluents from the CC by CDF and CD were 114 and 108% with no detectable sequestration of vancomycin in both modes of elimination.

DISCUSSION

Significant sequestration adds a clearance to that provided by CDF and CD. The study provides multiple evidence from the CC, the filter, and the effluents of the lack of an increase in total clearance in comparison with the flow rates without significant sequestration in the ST^® filter comparing cumulative effluents to the initial dose in the CC.

CONCLUSIONS

There is no evidence ST^® filters directly sequestrate vancomycin.

Elimination of cefotaxime using polysulfone and polyacrylonitrile-derived filters: An in vitro assessment

Continuous renal replacement therapy (CCRT) efficiently eliminates cefotaxime. To our knowledge, there are no previous in vitro studies dealing with the disposition of cefotaxime. We studied the elimination of cefotaxime by two filters in a model mimicking a session of CRRT using the NeckEpur^® technology. The ST150^®-polyacrylonitrile filter with the Prismaflex, Baxter-Gambro, and the AV1000^®-polysulfone filter with the Multifiltrate Pro, Fresenius, were studied. Continuous filtration used a flowrate of 1 L/h in post-dilution only. Simulated blood flowrate was set at 200 mL/min. Routes of elimination were assessed using the NeckEpur^® technology. Cefotaxime concentrations were measured using ultra high-performance liquid chromatography, and tandem mass spectrometry. Two sessions were performed using the ST^® filter and three using the AV^® filter. Stability of cefotaxime during 6 h was assessed in triplicate with a mean variation of concentrations of 2.4 ± 1.5% at the end of the study. The mean measured initial concentration in the central compartment (CC) for the five sessions was 52.4 mg/L. The mean amount eliminated from the CC at the end of the sessions using the ST150^®-polyacrylonitrile and the AV1000^®-polysulfone filters were 72% and 73%, respectively. The clearances of cefotaxime from the central compartment (CC) were 1.1 and 1.2 L/h, respectively. The mean sieving coefficient were 0.99 and 0.99, respectively. The mean percentages of the amount eliminated from the CC by filtration/adsorption were 87/13% and 92/8%, respectively. Both adsorption percentages were below 15%. We conclude neither the ST150^®-polyacrylonitrile nor the AV1000^®-polysulfone filters result in clinically significant adsorption of cefotaxime.

Hemodynamic instability during connection to continuous kidney replacement therapy in critically ill pediatric patients

BACKGROUND

Emerging data suggest evidence of organ hypoperfusion during continuous kidney replacement therapy (CKRT). To facilitate kidney and global recovery, we must understand the hemodynamic risks associated with CKRT. We aimed to investigate frequency of hemodynamic instability and association with patient outcomes in pediatric CKRT.

METHODS

In a single-center study of CKRT patients between September 2016 and October 2018, we collected hemodynamic data using archived high-resolution physiologic data before and after connection. Primary outcome was hypotension defined as ≥ 20% decrease in baseline mean arterial pressure (MAP) for ≥ 2 consecutive minutes in the 60 min following connection. Secondary outcomes were tachycardia (≥ 20% increase in heart rate (HR)) and hemodynamic interventions.

RESULTS

Seventy-one patients median age 54 months (IQR 7-144), weight 16.7 kg (IQR 8-41), on hemodiafiltration had 304 filter connections, 4 (IQR 1-7) filters per patient; the median duration of CKRT was 9 days (IQR 3-20). The most common CKRT indication was AKI with fluid overload (48/71, 69%). There were 78 (27%) hypotension and 42 (14%) tachycardia events; cumulative duration of hypotension was 14 min IQR (3-31.75). Teams provided intervention in 17/304 (6%) of connections. Pediatric Logistic Organ Dysfunction 2 was the only independent predictor of hypotension (aOR 2.12 (CI 1.02-4.41)).

CONCLUSIONS

One in four and one in six pediatric CKRT filter connections were complicated by hypotension and tachycardia, respectively. Higher illness severity at CKRT initiation was independently associated with hypotension. Impact of CKRT-associated hemodynamic instability on global patient outcomes requires further targeted study. A higher resolution version of the Graphical abstract is available as Supplementary information.

Elimination of three doses of gentamicin over three consecutive days using a polyacrylonitrile-derived filter: An in vitro assessment

INTRODUCTION

Adsorption of gentamicin in a polyacrylonitrile filter was previously evidenced in a session lasting 6 h using the NeckEpur model. We extended the study over three consecutive days to mimic the 72-h life span of a filter.

METHODS

Prismaflex^® monitor and ST150^® filter were used in the continuous diafiltration (CDF) mode at a 2.5 L/h flowrate. The daily session started with a 6-h session of CDF. Thereafter, the 5-L central compartment was changed using a bag free of gentamicin to assess gentamicin release over the following 18 h. Experiments were repeated on Day 2 and stopped at the end of the 6-h session of CDF on Day 3. The experiment was performed in duplicate.

RESULTS

At a 2.5 L/h diafiltration flowrate, the mean daily clearances of gentamicin were 5.5, 4.0, and 3.3 L/h, respectively. The mean diafiltration and adsorption ratios in the daily elimination of gentamicin were 32/68%, 58/42%, and 88/12%, respectively. During days 1 and 2, the mean amount of gentamicin released from the ST150^® filter were 14 and 34 mg, respectively.

CONCLUSION

The pharmacokinetics of gentamicin over 3 days is strongly altered by adsorption in the same filter with a progressive decrease of elimination by adsorption, suggesting saturation of the filter. One limitation of our study results from the mode of administration using a bolus dose instead of an infusion over 30 min. Adsorption adds a clearance to those of diafiltration. The time-dependency of gentamicin clearance precludes using a constant dosage regimen over the filter's life span.

Diafiltration flowrate is a determinant of the extent of adsorption of amikacin in renal replacement therapy using the ST150®-AN69 filter: An in vitro study

Introduction:

In continuous renal replacement therapy, conduction and convection are controlled allowing prescribing dosage regimen improving survival. In contrast, adsorption is an uncontrolled property altering drug disposition. Whether adsorption depends on flowrates is unknown. We hypothesized an in vitro model may provide information in conditions mimicking continuous renal replacement therapy in humans.

Methods:

ST150®-AN69 filter and Prismaflex dialyzer, Baxter-Gambro were used. Simulated blood flowrate was set at 200 mL/min. The flowrates in the filtration (continuous filtration), dialysis (continuous dialysis), and diafiltration (continuous diafiltration) were 1500, 2500, and 4000 mL/h, respectively. Routes of elimination were assessed using NeckEpur® analysis.

Results:

The percentages of the total amount eliminated by continuous filtration, continuous dialysis, and continuous diafiltration were 82%, 86%, and 94%, respectively. Elimination by effluents and adsorption accounted for 42% ± 7% and 58% ± 5%, 57% ± 7% and 43% ± 6%, and 84% ± 6% and 16% ± 6% of amikacin elimination, respectively. There was a linear regression between flowrates and amikacin clearance: Y = 0.6 X ± 1.7 (R2 = 0.9782). Conversely, there was a linear inverse correlation between the magnitude of amikacin adsorption and flowrate: Y = –16.9 X ± 84.1 (R2 = 0.9976).

Conclusion:

Low flowrates resulted in predominant elimination by adsorption, accounting for 58% of the elimination of amikacin from the central compartment in the continuous filtration mode at 1500 mL/h of flowrate. Thereafter, the greater the flowrate, the lower the adsorption of amikacin in a linear manner. Flowrate is a major determinant of adsorption of amikacin. There was an about 17% decrease in the rate of adsorption per increase in the flowrate of 1 L/min.

Can we use an ex vivo continuous hemofiltration model to describe the adsorption and elimination of meropenem and piperacillin?

OBJECTIVES

To determine the adsorption and elimination characteristics of meropenem and piperacillin during simulated continuous renal replacement therapy (CRRT), and to compare the observed data from this ex vivo study with previous data from clinical studies.

METHOD

This was an experimental study utilizing a modified CRRT circuit and polysulfone membrane (1.2 m2), circulated with a blood-crystalloid mixture. Adsorption onto the CRRT circuit was tested over a 4-h period, and clearance was assessed separately using variable continuous hemofiltration settings.

RESULTS

A rapid 9% reduction in circulating meropenem and piperacillin concentrations was observed at approximately 0.5 and 1.0 h for each antibiotic, respectively. The post-dilution setting was associated with a significantly higher sieving coefficient (Sc) and filter clearance (CLfilter) (mean ± SD) (Sc 1.14 ± 0.10 versus 1.06 ± 0.04; CLfilter 19.05 ± 1.63 versus 17.59 ± 0.62 ml/min, P values < 0.05) for meropenem. No significant differences were observed for piperacillin pharmacokinetics. Clinically comparable Sc data were observed between data obtained from the ex vivo study and data from previous clinical studies, for both antibiotics.

CONCLUSIONS

Meropenem and piperacillin appear to be rapidly adsorbed into the CRRT circuit, and the delivery site of fluid replacement significantly influences meropenem pharmacokinetics. However, these findings are likely to be clinically insignificant and not affect dosing requirements. This ex vivo method could be a surrogate for future clinical pharmacokinetic studies of CRRT. Further research is required to explore the applicability of the ex vivo method to further characterize antibiotic pharmacokinetics during CRRT.

The challenges of multiple organ dysfunction syndrome and extra-corporeal circuits for drug delivery in critically ill patients

The multiple organ dysfunction syndrome (MODS) is characterized by more than one organ system failing, especially during critical illness. MODS is the leading cause of morbidity and mortality in current ICU practice; moreover, multiple organ dysfunction, especially liver and kidneys, may significantly affect the pharmacokinetics (PKs) of different drugs that are currently administered in critically ill patients. These PK alterations may either result in insufficient drug concentrations to achieve the desired effects or in blood and tissue accumulation, with the development of serious adverse events. The use of extra-corporeal circuits, such as extracorporeal membrane oxygenation (ECMO) and continuous renal replacement therapy (CRRT), may further contribute to PKs changes in this patients' population. In this review, we have described the main PK changes occurring in all these conditions and how drug concentrations may potentially be affected. The lack of prospective studies on large cohorts of patients makes impossible any specific recommendation on drug regimen adjustment in ICU patients. Nevertheless, the clinicians should be aware of these abnormalities in order to better understand some unexpected therapeutic issues occurring in such patients.

Clearance of moxifloxacin during continuous haemofiltration (CVVHF) in vitro

BACKGROUND/AIMS

The clearance of moxifloxacin is reported to be unaltered in the presence of renal insufficiency. There is little information about the clearance of intravenous moxifloxacin in renal replacement therapies during intensive care. The aim of this study was to determine the clearance of moxifloxacin during continuous veno-venous haemofiltration (CVVHF) in vitro.

METHODS

The elimination of moxifloxacin (reservoir with 600 mL of washed human erythrocytes, 100 mL of NaHCO3 and various amounts of Ringer solution and human albumin to give a total volume of 1000 mL, pH 7.35 +/- 0.5; haematocrit 41 +/- 2) during CVVHF in vitro with two filter conditions (during priming, after priming), three protein concentrations (human albumin: 0 g/L, 20 g/L, 40 g/L) and two filtration velocities [(i) standard condition: blood flow at 100 mL/min and turnover of 2 L/h; (ii) blood flow at 50 mL/min and turnover of 1 L/h] were investigated.

RESULTS

A new filter needs 20 min of priming before moxifloxacin reaches a steady relative filtration rate. The sieving coefficient with 0 g/L albumin was 1.07, with 20 g/L 0.90 and with 40 g/L 0.80. Under standard filtration conditions (i) the renal clearance was between 26.7 and 35.7 mL/min, and under the altered conditions (ii) it was 15.2 mL/min.

CONCLUSION

During CVVHF in vitro we found filtration clearances of moxifloxacin of the same order as its renal clearance in healthy subjects. The high sieving coefficient, nearly independent of blood protein concentration, would suggest that moxifloxacin is filtered almost as freely as creatinine. These results do not indicate a need for dose adjustment under appropriate haemofiltration conditions and normal hepatic function.

The effect of adsorption, filter material and point of dilution on antibiotic elimination by haemofiltration an in vitro study of levofloxacin

We studied an in vitro model of continuous venous-venous haemofiltration (CVVH), into which levofloxacin 100 mg was infused, to determine levofloxacin adsorption and to determine the effect of filter material and point of dilution (pre- or post-filter) on sieving coefficient. Mean (standard deviation; S.D.) adsorption was 18.7 (5.3) mg for the polyamide filter and 40.2 (2.0) mg for the polyacrylonitrile (PAN) filter (P < 0.001). Post-dilution resulted in a minor, but statistically significant, decrease in sieving coefficient (pre-dilution 0.96 (S.D. 0.10), post-dilution 0.88 (S.D. 0.11) with the PAN filter. These data indicate that the variability in published values for levofloxacin sieving coefficient are not due to variation in point of dilution or membrane type (PAN or polyamide). Significant adsorption of levofloxacin onto PAN filters occurs.