Andrew J. Dalley

Meropenem dosing in critically ill patients with sepsis and without renal dysfunction: intermittent bolus versus continuous administration? Monte Carlo dosing simulations and subcutaneous tissue distribution

OBJECTIVES To compare the plasma and subcutaneous tissue concentration-time profiles of meropenem administered by intermittent bolus dosing or continuous infusion to critically ill patients with sepsis and without renal dysfunction, and to use population pharmacokinetic modelling and Monte Carlo simulations to assess the cumulative fraction of response (CFR) against Gram-negative pathogens likely to be encountered in critical care units. PATIENTS AND METHODS We randomized 10 patients with sepsis to receive meropenem by intermittent bolus administration (n = 5; 1 g 8 hourly) or an equal dose administered by continuous infusion (n = 5). Serial subcutaneous tissue concentrations were determined using microdialysis and compared with plasma data for first-dose and steady-state pharmacokinetics. Population pharmacokinetic modelling of plasma data and Monte Carlo simulations were then undertaken with NONMEM. RESULTS It was found that continuous infusion maintains higher median trough concentrations, in both plasma (intermittent bolus 0 versus infusion 7 mg/L) and subcutaneous tissue (0 versus 4 mg/L). All simulated intermittent bolus, extended and continuous infusion dosing achieved 100% of pharmacodynamic targets against most Gram-negative pathogens. Superior obtainment of pharmacodynamic targets was achieved using administration by extended or continuous infusion against less susceptible Pseudomonas aeruginosa and Acinetobacter species. CONCLUSIONS This is the first study to compare the relative concentration-time data of bolus and continuous administration of meropenem at the subcutaneous tissue and plasma levels. We found that the administration of meropenem by continuous infusion maintains higher concentrations in subcutaneous tissue and plasma than by intermittent bolus dosing. Administration by extended or continuous infusion will achieve superior CFR against less-susceptible organisms in patients without renal dysfunction.

Piperacillin penetration into tissue of critically ill patients with sepsis - bolus versus continuous administration?

Objective:To describe a pharmacokinetic model of piperacillin concentrations in plasma and subcutaneous tissue when administered by bolus dosing and continuous infusion in critically ill patients with sepsis on days 1 and 2 of antibiotic therapy and to compare results against previous results for piperacillin from a cohort of patients with septic shock. Design:Prospective randomized controlled trial. Setting:Eighteen-bed intensive care unit at 918-bed tertiary referral hospital. Patients:Thirteen critically ill adult patients with known or suspected sepsis in whom the treating physician deemed piperacillin–tazobactam appropriate therapy were conveniently sampled. Interventions:Patients were randomized to receive different daily doses of piperacillin–tazobactam by bolus dosing or continuous infusion (continuous infusion—six patients; bolus dosing—seven patients). Serial plasma and tissue concentrations were determined on days 1 and 2 of treatment. Tissue concentrations of piperacillin were determined using a subcutaneously inserted microdialysis catheter. Separate pharmacokinetic models were developed for both bolus and continuous dosing. Measurements and Main Results:This is the first known article to report concurrent plasma and subcutaneous tissue concentrations of a β-lactam antibiotic administered by bolus and continuous dosing in critically ill patients with sepsis. With a 25% lower piperacillin dose administered to the continuous infusion group, the infusion group had statistically significantly higher median plasma concentrations than the bolus group on day 2 (16.6 vs. 4.9 mg/L; p = 0.007). There was a trend to higher median plasma concentrations on day 1 in the bolus dosing group (8.9 vs. 4.9 mg/L; p = 0.078). Median tissue concentrations were not statistically different on day 1 (infusion group 2.4 mg/L vs. bolus group 2.2 mg/L; p = 0.48) and day 2 (infusion group 5.2 mg/L vs. bolus group 0.8 mg/L; p = 0.45). A two-compartment pharmacokinetic model was found to describe the data best. Tissue pharmacodynamic targets were achieved more successfully with infusion dosing. Conclusions:Patients with sepsis do not seem to have the same level of impairment of tissue distribution as described for patients with septic shock. A 25% lower dose of piperacillin administered by continuous infusion seems to maintain higher trough concentrations compared with standard bolus dosing. It is likely that the clinical advantages of continuous infusion are most likely to be evident when treating pathogens with high minimum inhibitory concentration, although without therapeutic drug monitoring and subsequent dose adjustment, infusions may never achieve target concentrations of organisms with very high minimum inhibitory concentrations in a small number of patients.

First-dose and steady-state population pharmacokinetics and pharmacodynamics of piperacillin by continuous or intermittent dosing in critically ill patients with sepsis

The objectives of this study were (i) to compare the plasma concentration-time profiles for first-dose and steady-state piperacillin administered by intermittent or continuous dosing to critically ill patients with sepsis and (ii) to use population pharmacokinetics to perform Monte Carlo dosing simulations in order to assess the probability of target attainment (PTA) by minimum inhibitory concentration (MIC) for different piperacillin dosing regimens against bacterial pathogens commonly encountered in critical care units. Plasma samples were collected on Days 1 and 2 of therapy in 16 critically ill patients, with 8 patients receiving intermittent bolus dosing and 8 patients receiving continuous infusion of piperacillin (administered with tazobactam). A population pharmacokinetic model was developed using NONMEM, which found that a two-compartment population pharmacokinetic model best described the data. Total body weight was found to be correlated with drug clearance and was included in the final model. In addition, 2000 critically ill patients were simulated for pharmacodynamic evaluation of PTA by MIC [free (unbound) concentration maintained above the MIC for 50% of the dosing interval (50% f(T>MIC))] and it was found that continuous infusion maintained superior free piperacillin concentrations compared with bolus administration across the dosing interval. Dosing simulations showed that administration of 16g/day by continuous infusion vs. bolus dosing (4g every 6h) provided superior achievement of the pharmacodynamic endpoint (PTA by MIC) at 93% and 53%, respectively. These data suggest that administration of piperacillin by continuous infusion, with a loading dose, both for first dose and for subsequent dosing achieves superior pharmacodynamic targets compared with conventional bolus dosing.

Epithelial to mesenchymal transition (EMT) biomarkers--E-cadherin, beta-catenin, APC and Vimentin--in oral squamous cell carcinogenesis and transformation.

OBJECTIVES To investigate immunohistochemical (IHC) analysis of E-cadherin, β-catenin, APC and Vimentin for prediction of oral malignant transformation. MATERIALS AND METHODS Immunoreactivity for E-cadherin, β-catenin, APC and Vimentin were determined for 100 oral biopsies classified as normal, mild dysplasia, moderate-severe dysplasia or OSCC, using the IHC scoring or label index scoring systems. Co-expression of biomarkers and correlation with histopathological grading was analysed. Vimentin and E-cadherin results were confirmed by RT-PCR and further investigated in vitro using a novel organotypic cell invasion model based on human dermis. RESULTS A trend for decreased E-cadherin expression but increased Vimentin expression that correlated with increased disease severity was observed. Epithelial β-catenin localisation shifted from being membranous to cytoplasmic/nuclear with increased histopathological grade severity. Relative to normal, APC expression was decreased for mild dysplasia but increased for OSCC. Co-expression of β-catenin, APC and Vimentin (Spearman rank correlation) suggests interdependence of these molecules and involvement of the Wnt pathway in oral malignant transformation. Relative mRNA expression of E-cadherin for dysplasia and OSCC were less than 1% of normal tissue values, and mRNA expression of Vimentin was 3.7 times higher for OSCC than normal. After 63 days of organotypic culture neoplastic oral keratinocytes (PE/CA-PJ15) lost expression of E-cadherin and gained expression of Vimentin relative to their non-invasive counterparts in the epithelium. CONCLUSIONS Trends in the expression of EMT markers - E-cadherin, β-catenin, APC and Vimentin - suggest their involvement in oral carcinogenesis via Wnt pathway dysregulation. Aberrant expression of β-catenin, APC and Vimentin are potential markers of malignant transformation.

Keratinocytes contract human dermal extracellular matrix and reduce soluble fibronectin production by fibroblasts in a skin composite model

Composites of human de-epidermised acellular dermis and normal adult human keratinocytes and fibroblasts were examined for the ability of cells to contract these composites. Image analysis of the outline of the composites showed that, in this model, keratinocytes alone or in the presence of fibroblasts caused highly significant contraction (of the order of 25% by day 12). There was no significant contraction of the dermis with fibroblasts alone or in the absence of cells. The presence or absence of basement membrane antigens did not influence the effect of keratinocytes on dermal contraction. Analysis of the conditioned media from these composites showed that the greatest fibronectin production was seen with fibroblasts alone in the presence of basement membrane. Keratinocytes alone produced little fibronectin irrespective of the presence of the basement membrane. If keratinocytes were present with fibroblasts, however, then fibronectin production was significantly reduced both in the presence and absence of the basement membrane, indicating that keratinocytes modify dermal fibroblast extracellular matrix production. This study shows that while keratinocytes and fibroblasts are clearly influencing each others activity in this human skin composite model, under the circumstances we describe it is the keratinocyte and not the fibroblast which causes contraction of the human de-epidermised acellular dermis.

Investigation of calmodulin and basic fibroblast growth factor (bFGF) in idiopathic myelofibrosis: Evidence for a role of extracellular calmodulin in fibroblast proliferation

The urinary concentration of calmodulin and basic fibroblast growth factor (bFGF) was determined in a total of 53 patients with various chronic myeloproliferative disorders (CMPD), including 22 patients with idiopathic myelofibrosis (IMF). Calmodulin excretion was significantly elevated in IMF (0.29 ± 0.04 μg/mmol creatinine) (P<0.001), when compared to polycythaemia vera (PV) (0.14 ± 0.02), essential thrombocythaemia (ET) (0.13 ± 0.04), chronic myeloid leukaemia (CML) (0.16 ± 0.02), unclassified myeloproliferative disorders (UMPD) (0.11 ± 0.02) and age‐matched controls (0.1 ± 0.02) (P<0.001). In contrast, bFGF was slightly elevated in all CMPD conditions when compared to age‐matched controls. A neutralizing antibody to calmodulin was demonstrated to significantly influence the in vitro proliferation of normal human fibroblasts, an effect dependent on both cell density and the presence of fetal calf serum (FCS). Essentially, the antibody reduced FCS‐induced proliferation of low‐density fibroblasts but had little or no inhibitory effect on high‐density fibroblasts in the absence of FCS. In addition, extracellular calmodulin was shown not to interact with known fibroblast mitogens, namely IFG‐1, EGF, bFGF and PDGF. We conclude that extracellular calmodulin should be considered, in addition to PDGF, TFG‐β and EGF, as a potential mitogen involved in the stromal reaction of idiopathic myelofibrosis.

Keratinocyte-driven contraction of reconstructed human skin.

We have previously reported that reconstructed human skin, using deepidermized acellular sterilized dermis and allogeneic keratinocytes and fibroblasts, significantly contracts in vitro. Contracture of split skin grafts in burns injuries remains a serious problem and this in vitro model provides an opportunity to study keratinocyte/mesenchymal cell interactions and cell interactions with extracted normal human dermis. The aim of this study was to investigate the nature of this in vitro contraction and explore several approaches to prevent or reduce contraction. Three different methodologies for sterilization of the dermal matrix were examined: glycerol, ethylene oxide and a combination of glycerol and ethylene oxide. While the nature of the sterilization technique influenced the extent of contraction and thinner dermal matrices contracted proportionately more than thicker matrices, in all cases contraction was driven by the keratinocytes with relatively little influence from the fibroblasts. The contraction of the underlying dermis did not represent any change in tissue mass but rather a reorganization of the dermis which was rapidly reversed (within minutes) when the epidermal layer was removed. Pharmacological approaches to block contraction showed forskolin and mannose‐6‐phosphate to be ineffective and ascorbic acid‐2‐phosphate to exacerbate contraction. However, Galardin, a matrix metalloproteinase inhibitor and keratinocyte conditioned media, both inhibited contraction.

Measurement of tissue cortisol levels in patients with severe burns: a preliminary investigation.

IntroductionThe assessment of adrenal function in critically ill patients is problematic, and there is evidence to suggest that measurement of tissue glucocorticoid activity may be more useful than estimation of plasma cortisol concentrations. Interstitial cortisol concentrations of cortisol represent the available pool of glucocorticoids able to enter the cell and bind to the glucocorticoid receptor. However the concentrations of plasma cortisol may not accurately reflect interstitial concentrations. We elected to perform a preliminary study into the feasibility of measuring interstitial cortisol by microdialysis, and to investigate the relationship between total plasma cortisol, free plasma cortisol and interstitial cortisol in patients with severe burns.MethodsA prospective observational study carried out in a tertiary intensive care unit. Ten adult patients with a mean total burn surface area of 48% were studied. Interstitial cortisol was measured by microdialysis from patient-matched burnt and non-burnt tissue and compared with that of 3 healthy volunteers. Plasma sampling for estimations of total and free cortisol concentrations was performed concurrently.ResultsIn the burn patients, mean total plasma and free plasma cortisol concentrations were 8.8 +/- 3.9, and 1.7 +/- 1.1 mcg/dL, (p < 0.001), respectively. Mean subcutaneous microdialysis cortisol concentrations in the burn and non-burn tissue were 0.80 +/- 0.31 vs 0.74 +/- 0.41 mcg/dL (p = 0.8), respectively, and were significantly elevated over the mean subcutaneous microdialysis cortisol concentrations in the healthy volunteers. There was no significant correlation between total plasma or free plasma and microdialysis cortisol concentrations. Plasma free cortisol was better correlated with total burn surface area than total cortisol.ConclusionsIn this preliminary study, interstitial cortisol concentrations measured by microdialysis in burnt and non-burnt skin from patients with severe thermal injury are significantly elevated over those from healthy volunteers. Plasma estimations of cortisol do not correlate with the microdialysis levels, raising the possibility that plasma cortisol may be an unreliable guide to tissue cortisol activity.

A simple in vitro model for investigating epithelial/mesenchymal interactions: keratinocyte inhibition of fibroblast proliferation and fibronectin synthesis

Hypertrophic scarring and graft contracture are major causes of morbidity after burn injuries. It is well established that application of a split‐thickness skin graft reduces scarring and contraction, and cultured epithelial autografts have a similar effect. To investigate the influence of keratinocytes on fibroblast proliferation and fibronectin synthesis, we used an in vitro separated co‐culture model in which epithelial sheets were cultured above fibroblast monolayers without physical contact. We also investigated the response of fibroblasts to keratinocyte‐conditioned medium (KCM) obtained from confluent and subconfluent keratinocyte monolayers. Both cultured epithelial sheets, composed of adherent fully confluent keratinocytes, and their conditioned medium, reduced fibroblast proliferation. However, KCM from subconfluent keratinocytes stimulated fibroblast proliferation at low concentrations while inhibiting it at higher concentrations, indicating that keratinocytes can produce both mitogenic and growth‐inhibiting factors for fibroblasts. KCM, but not epithelial sheet co‐culture, also inhibited fibroblast fibronectin synthesis. This indicates regulation of fibroblast phenotype by soluble factors released by the keratinocyte and also suggests that there is a dialogue between keratinocytes and fibroblasts with respect to fibronectin production. We conclude that this separated co‐culture model is a simple way to study epithelial/mesenchymal communication particularly with respect to the role of the fibroblast in wound healing.

Organotypic culture of normal, dysplastic and squamous cell carcinoma‐derived oral cell lines reveals loss of spatial regulation of CD44 and p75NTR in malignancy

Oral squamous cell carcinomas (OSCC) often arise from dysplastic lesions. The role of cancer stem cells in tumour initiation is widely accepted, yet the potential existence of pre-cancerous stem cells in dysplastic tissue has received little attention. Cell lines from oral diseases ranging in severity from dysplasia to malignancy provide opportunity to investigate the involvement of stem cells in malignant progression from dysplasia. Stem cells are functionally defined by their ability to generate hierarchical tissue structures in consortium with spatial regulation. Organotypic cultures readily display tissue hierarchy in vitro; hence, in this study, we compared hierarchical expression of stem cell-associated markers in dermis-based organotypic cultures of oral epithelial cells from normal tissue (OKF6-TERT2), mild dysplasia (DOK), severe dysplasia (POE-9n) and OSCC (PE/CA P J15). Expression of CD44, p75(NTR), CD24 and ALDH was studied in monolayers by flow cytometry and in organotypic cultures by immunohistochemistry. Spatial regulation of CD44 and p75(NTR) was evident for organotypic cultures of normal (OKF6-TERT2) and dysplasia (DOK and POE-9n) but was lacking for OSCC (PE/CA PJ15)-derived cells. Spatial regulation of CD24 was not evident. All monolayer cultures exhibited CD44, p75(NTR), CD24 antigens and ALDH activity (ALDEFLUOR(®) assay), with a trend towards loss of population heterogeneity that mirrored disease severity. In monolayer, increased FOXA1 and decreased FOXA2 expression correlated with disease severity, but OCT3/4, Sox2 and NANOG did not. We conclude that dermis-based organotypic cultures give opportunity to investigate the mechanisms that underlie loss of spatial regulation of stem cell markers seen with OSCC-derived cells.