Portal de Periódicos da CAPES

Falsely elevated vancomycin plasma concentrations sampled from central venous implantable catheters (portacaths)

2010; Wiley; Volume: 70; Issue: 5 Linguagem: Inglês

10.1111/j.1365-2125.2010.03749.x

1365-2125

Daniel F. B. Wright, Hesham S. Al‐Sallami, Pamela M. Jackson, David Reith,

Pneumonia and Respiratory Infections

Vancomycin is a glycopeptide antibiotic used to treat infections caused by gram-positive pathogens, including methicillin-resistant Staphylococcus aureus (MRSA). The monitoring of steady-state vancomycin plasma concentrations is recommended to reduce the risk of ototoxicity and nephrotoxicity and to ensure that target therapeutic plasma concentrations are achieved [1–3]. Most current recommendations suggest that vancomycin doses should be adjusted to achieve trough plasma concentrations from 5–20 mg l−1, depending on the severity of the infection and the pathogen being treated [1, 3]. Clinicians therefore rely on accurate plasma concentrations to aid dose adjustments and to ensure optimal patient care. We present two cases of spurious vancomycin plasma concentrations drawn from central venous implantable catheters (portacaths). Whereas a previous report described spurious vancomycin plasma concentrations drawn from a central catheter [4], this problem does not appear to have been described for vancomycin sampled from portacaths. A 39-year-old man was admitted to Christchurch Hospital in Christchurch, New Zealand with a severe infective exacerbation of bronchietasis. His medical history was notable for severe respiratory illness including childhood asthma, allergic bronchopulmonary aspergillosis, immuno-deficiency hyper-IgE syndrome, cor pulmonale and type-2 respiratory failure requiring non-invasive ventilation and home oxygen. Medicines on admission included budesonide/eformoterol turbuhaler, tiotropium inhaler, itraconazole, alendronate, calcium carbonate, salbutamol nebules, prednisone, calciferol and multivitamins. Sputum samples initially grew Pseudomonas aeruginosa and he was commenced on intravenous ceftazidime 1 g every 8 h and tobramycin 400 mg daily for 10 days. Repeat sputum samples on day 8 grew a multi-resistant Streptococcus pneumonia. The ceftazidime and tobramycin were stopped and intravenous vancomycin 750 mg (15 mg kg−1) every 12 h was commenced with a target trough plasma concentration of 5–10 mg ml−1. After 1 week of vancomycin therapy the patient was much improved, both clinically and radiologically. He was discharged after 21 days in the hospital. Serum creatinine concentrations remained stable throughout his hospital stay (0.07–0.09 mmol l−1). Table 1 summarizes the vancomycin plasma concentrations for case 1. All samples drawn from the patient's portacath resulted in vancomycin plasma concentration measurements >20 mg l−1, while both samples from a peripheral vein were within the therapeutic range of 5–10 mg l−1. Sample 4 was drawn from the portacath under the supervision of an i.v. nurse specialist, to ensure correct sampling technique. Samples 5 and 6 were drawn within 5 min of each other with the portacath sample resulting in a vancomycin plasma concentration nearly four times higher than the sample drawn from the peripheral vein. The portacath was a Vital-Port® (model 6012, Cook Vascular Incorporated, Vandergrift, PA USA) with an estimated volume of 0.62 ml. The port had been in situ for 22 months and had been used largely for blood sampling and antibiotic administration. As stated, in the previous month the patient had received ceftazidime (14 days), tobramycin (14 days) and vancomycin (4 days). The port was locked with heparin-saline 500 IU 5 ml−1 solution when not in use. It was assumed that all samples reported here were taken from the portacath according to local hospital policy. This can be confirmed for sample 4 (see Table 1) which was witnessed by the i.v. nurse specialist. The portacath sampling protocol entailed a 10 ml normal saline flush, followed by the withdrawal of 5 ml blood for discard. The blood sample was then withdrawn into a 4.5 ml lithium-heparin tube and the flush repeated with a further 10 ml of normal saline. The vancomycin assay was performed by the Toxicology Department at Canterbury Health Laboratories, Christchurch, New Zealand. Plasma concentrations were measured using a commercial reagent kit (Vanc Flex®, Siemens Healthcare Diagnostics Inc., Newark, DE 19714, U.S.A). This is a particle enhanced turbidimetric inhibition immunoassay (PETINIA) which uses a Dimension® Xpand analyzer. According to the manufacturer, the with-in run coefficient of variation (CV) ranged from 5–20% for vancomycin concentrations from 4.6–30.0 µg ml−1[5]. A 2-year-old boy was admitted to Dunedin Public Hospital in Dunedin, New Zealand with fever, vomiting, cough, a temperature of 39.5°C, a heart rate of 170 beats min−1 and mild tachypnoea. He had a complex medical history including short gut syndrome secondary to gastroschisis, portal hypertension, liver disease related to total parenteral nutrition (TPN), gastroesophageal reflux disease and recurrent i.v. line infections. His white blood cell count was 8.0 × 109 l−1, C-reactive protein was 12 mg l−1 and his renal function was normal (serum creatinine 31 mmol l−1). His portacath site was erythematous. Intravenous flucloxacillin was commenced for a suspected cellulitis and portacath infection. Blood cultures grew a multi-resistant strain of Stapholococcus epidermidis so antibiotic therapy was changed to vancomycin. The cellulitis around the portacath site resolved after several days of therapy but the patient continued to spike fevers. A repeat blood culture from the portacath grew Candida krusei and intravenous amphotericin was commenced. The patient was subsequently transferred to another hospital for a fundoplication procedure and removal of the infected portacath. Table 1 summarizes the vancomycin plasma concentrations for case 2. Vancomycin was commenced at 171 mg (15 mg kg−1) every 8 h with a target trough plasma concentration of 10–20 mg l−1. A trough concentration just prior to the third dose was subtherapeutic at 3.1 mg l−1 (sampling site unknown) so the dose was increased to 250 mg (22 mg kg−1) every 6 h. A subsequent vancomycin trough plasma concentration, drawn from the patient's portacath after a single 250 mg dose, measured 20.0 mg l−1. Vancomycin was withheld for one dose and subsequently recommenced at 250 mg every 6 h. A repeat trough vancomycin plasma concentration drawn by finger-prick after three more doses measured 8.9 mg l−1. The portacath was a Vital Port® petit (Cook Vascular Incorporated, Vandergrift, PA USA) with a maximum volume of 0.35 ml. Dead space volume could not be determined. The port had been used regularly for TPN (40 ml h−1 for 12 h daily) and antibiotics, including intravenous vancomycin, vancomycin locks (for 3 weeks prior to admission), metronidazole (for ∼6 days), ceftriaxone (for ∼6 days) and flucloxacillin (approximately 2 days). Blood samples were collected after 3 ml of blood had been withdrawn from the port and discarded, as per Dunedin Hospital guidelines. Following the blood sample the port was flushed with 10 ml of normal saline using a pulsative technique. The vancomycin assay was performed by the Southern Community Laboratories, Dunedin. Plasma concentrations were measured using a commercial reagent kit (AxSYM Vancomycin II®, Abbott, Wiesbaden, Germany). This is a fluorescence polarization immunoassay using the Abbott AxSYM automated analyzer. According to the manufacturer, the with-in run coefficient of variation (CV) ranged from 2.61–4.25% for vancomycin concentrations from 7.0–75 µg ml−1 and the between day precision is reported as 0.63–0.88 (CV%) for the same concentrations [6]. In both cases presented above, the unexpectedly high vancomycin concentrations caused confusion amongst clinical staff with respect to vancomycin dosing and, in one case, lead to the temporary withdrawal of treatment in a complex paediatric patient. Subsequent blood samples drawn from a peripheral vein demonstrated that vancomycin plasma concentrations were actually substantially lower than those drawn from portacaths. There are conflicting, albeit sparse, reports concerning spurious vancomycin plasma concentrations drawn from central venous access devices (CVADs), and none specifically concerning portacaths. In one report [4], a 56-year-old man with sternal osteomyelitis was treated with vancomycin 1 g every 24 h. Steady-state vancomycin plasma concentrations drawn from a central catheter were grossly toxic, ranging from 62–120 µg ml−1, while three samples drawn from a peripheral vein were all within the therapeutic range (about 10 µg ml−1). By contrast, a small study [7] compared vancomycin concentrations taken from central venous catheters with those drawn from peripheral veins in 21 patients. The samples were taken within 5 min of each other and no significant difference was found. Reports concerning the sampling of plasma concentrations of other antibiotics from CVADs are equally conflicting. Some studies [8–10] have shown that aminoglycoside concentrations drawn from CVADs are falsely elevated, while others do not [7, 11]. Similarly, reports of falsely elevated ciclosporin plasma concentrations drawn from central venous catheters [12, 13] have not been consistently replicated elsewhere [7, 14]. These conflicting results are difficult to interpret, but may relate to differences in the catheters studied (e.g. presence of dead space), bioabsorption or to variable sampling techniques. Few studies have examined the effect of pooling on drug concentrations drawn from CVADs. The theory is that some of the drug administered via the port will remain trapped in the catheter and cause a falsely elevated plasma concentration. One study has suggested that this can be rectified by simply using a larger (i.e. 20 ml) volume of flush solution [15]. While this is an appealing solution, limited evidence from the published literature does not support this practice. For example, in a study of vancomycin sampling discussed above [7], a flush volume of 5 ml was sufficient to achieve concordance between centrally and peripherally sampled drug, while a case report [4], also discussed above, noted grossly elevated vancomycin concentrations drawn from a central line despite flushing with 50 ml of solution. This suggests that increasing the flush volume may not solve the problem with regards to aberrant vancomycin concentrations sampled from CVADs. Few reports have examined the bioabsorption of vancomycin to catheter constituents. A study [16] of antibiotic-lock stability in central venous catheters suggested significant adherence of vancomycin to catheter surfaces. Vancomycin 10 mg ml−1 was placed in a polyurethane central venous catheter for 72 h. A loss of 30% was found and attributed to adherence of vancomycin to the luminal surface of the catheter. Another study [17] measured vancomycin concentrations in Gram-positive bacterial biofilm within polyurethane central venous catheters. The catheters were removed after an in vivo dwell time of 1–14 days and perfused in vitro with vancomcyin 1 g in 250 ml of 5% dextrose for 2 h. The biofilm was recovered and each lumen was found to contain between 1.2 and 80.9 mg (mean 31.6 mg) of vancomycin absorbed by the biofilm. It is unclear what influence these findings would have on vancomycin concentrations drawn from CVADs. Initiatively, bioabsorption would be expected to cause reduced vancomycin concentrations, rather than the elevated results observed. However, it is possible that drug leaching from biofilm in the presence of blood in the catheter lumen (but presumably not saline or heparin lock) could result in elevated concentrations, although this is speculative and does not appear to have been studied. A simple clinical solution to the problem of spurious vancomycin concentrations from portacaths, and perhaps other CVADs, might be to draw all drug samples from peripheral veins. However, many patients with portacaths have limited venous access, or are patients in whom repeated venepuncture is undesirable, e.g. paediatrics or those with compromised immunity. In these patients, a portacath is often inserted for the express purpose of drawing blood samples. Another solution might be to increase the flush volume before and after blood sampling. However as noted above, this practice has not been found to consistently reconcile peripheral and central drug concentrations. Spurious vancomycin plasma concentrations drawn from portacaths are clearly an important clinical problem. Falsely elevated concentrations have the propensity to cause incorrect dosage adjustments and to contribute to poor patient outcomes. Further research on sampling techniques, including the influence of flush volumes, is therefore warranted, as are studies examining the in vitro fluid dynamics of portacaths. These studies may clarify the influence of drug pooling and biofilm on drug sampling from central venous access devices. There are no competing interests to declare.