Abstract
Background
Lanthanum carbonate is used as a phosphate binder in patients with stage V chronic kidney disease (CKD). While well tolerated in clinical trials, with no toxicity reported as regards bone and liver metabolism, and cognitive function, concerns remain over possible toxicity. Published methods for the measurement of lanthanum ion in biological samples include aggressive and complicated sample preparation steps that are unsuitable for routine use. A simple method has been developed and validated for the measurement of serum lanthanum.
Method
A ThermoFisher Scientific XSERIES-II inductively coupled plasma-mass spectrometer was used to monitor 139La. Validation was undertaken using internal quality control solutions containing lanthanum ion (0.20, 0.70 and 4.00 μg/L). Lanthanum was measured in patients (number = 20) with CKD prescribed lanthanum carbonate (500–1500 mg/d) and patients undergoing haemodialysis not prescribed lanthanum carbonate (number = 20).
Results
Accuracy and imprecision were >95% and <5%, respectively. Calibration was linear (range 0.1–5 μg/L, R 2 = 0.99). The lower limit of quantification (LLoQ) was 0.1 μg/L lanthanum ion. In patients with CKD not prescribed lanthanum carbonate, serum lanthanum was below the LLoQ. Out of 20 CKD patients prescribed lanthanum carbonate, serum lanthanum was measurable in only 12 (range 0.11–0.60 μg/L lanthanum ion). There was no apparent relationship between dose and serum lanthanum in these patients.
Conclusions
A lack of relationship between the dose of lanthanum carbonate and the serum lanthanum concentration may have been due to poor adherence to the treatment regimen. However the concentrations measured were close to the LLoQ.
Introduction
Lanthanum is a metal of the lanthanide series that occurs naturally as radioactive (138La) and stable (139La) isotopes (0.01% and 99.99% abundance, respectively). Lanthanum is used in the optical and semiconductor industries, and within a research setting to assess the permeability of the blood:brain barrier. Lanthanum salts were not used clinically until the introduction of lanthanum carbonate hydrate [La2(CO3)3. xH2O (x = 4–5); Fosrenol, Shire] in the UK (2007) as an oral phosphate binder for patients with stage V chronic kidney disease (CKD).
Patients with CKD have hyperphosphataemia, which can result in pruritis, metastatic calcification and increased vascular ageing. Haemodialysis (4 h, thrice weekly) and a low phosphate diet are not adequate to control plasma phosphate in many patients and therefore oral phosphate binders are used widely. 1 Traditional phosphate binders are associated with adverse effects that limit their clinical application (aluminium hydroxide: encephalopathy; calcium-based binders: hypercalcaemia; sevelamer hydrochloride: gastrointestinal upset).
Although there have been no reports of adverse effects from lanthanum carbonate in clinical trials with respect to bone and liver metabolism, and to cognitive function, 2–6 there remain concerns as to (i) patient compliance to prescribed treatment and (ii) possible toxicity analogous to that seen in dialysis patients treated with aluminium hydroxide 1 and gastrointestinal side-effects. 7 We have thus developed a simple, inductively coupled plasma-mass spectrometric (ICP-MS) assay to measure lanthanum in serum samples from haemodialysis patients. An isotope of gadolinium, 157Gd, was used as the internal standard. As lanthanum is highly protein bound in human plasma (>99.7%), serum or plasma was considered a suitable sample for the measurement of lanthanum.
Methods
Materials and reagents
Nitric acid (Aristar grade) and Triton X-100 were from BDH Merck (Poole, UK). Pooled human serum was from Scipac (Sittingbourne, UK) and tested for the presence of 139La and 157Gd by ICP-MS. Water was purified by reverse osmosis and ion-exchange using an ELGA PURELAB Option-Q water purification system (Marlow, UK). All glass and plastic vessels were soaked in 5–10% v/v (volume/volume) aqueous nitric acid for at least two hours and rinsed three times using ultrapure deionized water (resistivity: 18.2 MΩ) prior to use.
Lanthanum nitrate [La(NO3)3] and gadolinium oxide (Gd2O3) were of Aristar grade (BDH Merck). Blood collection tubes assessed for possible lanthanum contamination were: SST II Advance (5.0 mL; Becton Dickinson [BD], Oxford, UK), trace element free serum Vacutainer (6.0 mL, BD), Lithium Heparin PST II (4.5 mL, BD), K3EDTA (tripotassium-ethylenediaminetetra acetic acid, 5.0 mL, Greiner-Bio, Stonehouse, UK) and serum tubes (6 mL, BD).
Instrumentation
A ThermoFisher Scientific XSERIES II quadrupole ICP-MS (Hemel Hempstead, UK) with a hexapole collision cell linked to a CETAC ASX-520 autosampler located within an ENC500 perspex enclosure was used. Analyses were performed in standard mode and m/z 139 and 157 were monitored for the measurement of lanthanum and gadolinium, respectively. The instrument settings used are given in Table 1.
Inductively coupled plasma-mass spectrometry settings
RF, radio frequency
Optimization of diluent
To optimize the concentrations of nitric acid and Triton X-100 in the diluent, eight separate aqueous diluents were prepared. Three diluents contained Triton X-100 at 0.1% (v/v) and varying concentrations of nitric acid (0.5%, 1% and 2% [v/v]), and a further five diluents contained nitric acid at 1% (v/v) and Triton X-100 at concentrations of 0%, 0.025%, 0.05%, 0.10% and 0.20% (v/v). These diluents were used to measure an aqueous solution of lanthanum nitrate containing 4.00 μg/L lanthanum ion.
Sample preparation
Samples were prepared for analysis in 10 mL polypropylene tubes (Teklab, County Durham, UK). Diluent (1.840 mL 1% [v/v] aqueous nitric acid containing 0.1% [v/v] Triton X-100) was added to the sample (135 μL) and internal standard solution (25 μL 1.00 g/L gadolinium oxide in 2% [v/v] aqueous nitric acid, final gadolinium ion concentration 5 μg/L) and vortex-mixed (10 s). Total sample volume was 2.0 mL.
Assay calibration
Calibration solutions (0.1, 0.2, 0.5, 1.0, 2.5 and 5.0 μg lanthanum ion/L) were prepared in 1% v/v aqueous nitric acid in 30 mL polypropylene sterile universal containers (VWR, Lutterworth, UK) by dilution of appropriate volumes of 1.00 g/L lanthanum nitrate in 2% (v/v) aqueous nitric acid using piston-driven air displacement pipettes (Anachem, Luton, UK). Analogous solutions were also prepared in pooled human serum and in ultrapure deionized water for the purposes of assay validation. The ratio of the signal intensity of 139La to 157Gd obtained on analysis of these solutions was plotted against the nominal lanthanum ion concentration to construct a calibration graph.
Internal quality control and external validation
To prepare internal quality control (IQC) material, 1.00 g/L lanthanum nitrate in 2% (v/v) aqueous nitric acid was added to pooled human serum using piston-driven air displacement pipettes, at low (0.20 μg/L), medium (0.70 μg/L) and high (4.00 μg/L) lanthanum ion concentrations. A portion of each solution was sent to the Centre for Analytical Sciences, University of Sheffield (CAS) for external validation. This material was used as IQC material for each batch analysis during method validation and when analysing patient samples. The acceptance criteria for each batch were ±20% of the value at each concentration measured by the CAS.
Accuracy and precision
For assessment of accuracy and precision, three separate solutions containing lanthanum ion at nominal concentrations of 0.20, 0.70 and 4.00 μg/L were prepared by adding appropriate volumes of 1.00 g/L aqueous lanthanum nitrate to 1% (v/v) aqueous nitric acid using piston-driven air displacement pipettes. These solutions were measured (n [number] = 10 at each concentration) to ascertain intra-assay accuracy and precision. They were also measured (n = 10) on five consecutive days to ascertain inter-assay accuracy and precision.
Recovery, selectivity and lower limit of quantitation
The recovery of lanthanum from aqueous solution was assessed by quintuplicate analyses of solutions of lanthanum nitrate prepared in ultrapure deionized water at concentrations of 0.125, 0.500 and 2.500 μg/L lanthanum ion, respectively.
Selectivity was ascertained by pooling serum from CKD patients not prescribed lanthanum carbonate and adding lanthanum nitrate at a concentration of 0.1 μg/L lanthanum ion to four portions.
The lower limit of quantitation (LLoQ) was calculated by the replicate analysis (n = 10) of ‘blank’ human serum (Scipac) and taken as 10 times the standard deviation of these measurements. 8
Collection of patient samples
Blood was collected from renal dialysis patients prescribed lanthanum carbonate using a stainless steel butterfly needle into a 5-mL SST II Advance serum separator Vacutainer. Following collection, samples were centrifuged (10 min; SANYO Harrier, 18/80 refrigerated; 20°C, relative centrifugal force 3000
Assessment of possible sources of sample contamination
Expired blood bank whole blood was used for the assessment of blood collection tubes for lanthanum contamination. The bag into which the blood had been collected contained 100 mL of citrate phosphate dextrose (sodium chloride [150 mmol/L], adenine [1.25 mmol/L], anhydrous glucose [45.4 mmol/L] and mannitol [28.8 mmol/L]).
Whole blood was separated into three portions; lanthanum nitrate was added to two of the portions at nominal concentrations of 0.5 and 2.5 μg/L lanthanum ion, respectively. The third portion had no lanthanum nitrate added (‘0 μg/L’). Whole blood (4 mL) at each concentration was added to each type of blood collection tube in quintuplicate. Tubes were then placed on a roller mixer for one hour before centrifugation (10 min, SANYO Harrier, 18/80; 20°C; relative centrifugal force 3000
The possibility of lanthanum contamination arising from the stainless steel needle used for venepuncture was assessed by drawing 5 mL of 1% (v/v) aqueous nitric acid through a Precision glide stainless steel syringe needle (BD), attached to a 10 mL Plastipak syringe (BD). The volume of 5 mL was chosen to mimic the actual drawing of a blood sample.
Results
A diluent containing 1% (v/v) aqueous nitric acid containing 0.1% Triton X-100 gave maximal sensitivity (counts per second). Assay calibration was linear over the range 0.1–5.0 μg/L (R 2 > 0.99). As calibration graphs for solutions prepared in either 1% v/v aqueous nitric acid or pooled human serum were comparable (difference in response was <10% for each calibration standard), aqueous calibration solutions were used during the analyses of serum samples and for the method validation where appropriate.
Accuracy, precision and external validation
Intra- and inter-assay accuracy and precision was assessed by analysis of the aqueous lanthanum ion solutions and calculated as the percent difference between the measured and nominal concentrations. The results are given in Table 2.
Serum lanthanum assay
Intra- and inter-assay accuracy and precision assessed by analysis of the internal quality control solution (n = 10 at each concentration)
CV, coefficient of variation
The IQC solutions (nominal lanthanum ion concentrations 0.20, 0.70 and 4.00 μg/L) were analysed by CAS using quadrupole ICP-MS after dilution in 10% (v/v) aqueous tetramethylammonium hydroxide against calibration solutions prepared locally. The measured lanthanum ion concentrations were 0.24, 0.76 and 4.02 μg/L.
Recovery, selectivity and LLoQ
Analysis of aqueous lanthanum nitrate solutions against calibration solutions prepared in dilute nitric acid gave lanthanum recoveries of 93 (coefficient of variation [%CV] 3), 94 (CV 2) and 93 (CV 1)%, respectively. Recovery of lanthanum ion added to pooled serum was 100 (CV < 1)%. The LLoQ was estimated to be 0.1 μg/L lanthanum ion.
Possible sources of sample contamination
The concentration of lanthanum in 1% aqueous nitric acid that had been drawn through a stainless steel needle was found to be below the LLoQ. Lanthanum in blood collection tubes containing expired citrated blood-bank whole blood (lanthanum nitrate not added) was also below the LLoQ.
The recovery of lanthanum from expired citrated blood-bank whole blood (added lanthanum ion concentrations 0.57 and at 2.28 μg/L, respectively) added to blood collection tubes is shown in Table 3 (data analysed using Analyse-it for Microsoft Excel 2003). Blood collection tubes containing K3EDTA showed the poorest recovery of lanthanum ion at both concentrations.
Recovery of lanthanum ion added to expired blood-bank whole blood allowed to equilibrate (2 − 8°C) in different sample collection tubes
NS, non-significant; K3EDTA, tripotassium-ethylenediaminetetra acetic acid
Analysis of patient samples
Serum lanthanum in CKD patients (n = 20) undergoing haemodialysis, but not prescribed lanthanum carbonate, was below the LLoQ (0.1 μg/L lanthanum ion). Lanthanum was also measured in 42 samples from 20 CKD patients (11 men, 9 women aged 30–79 [median 55] y) prescribed 500–1500 (median 750) mg/d lanthanum carbonate. In eight of these patients (20 samples, dose 750–1500 mg/d) lanthanum was always below the LLoQ. However, in the remaining 12 patients (22 samples, dose 500–1500 mg/d) the serum lanthanum ion concentration was 0.11–0.60 (median 0.18) μg/L. There was no apparent relationship between dose and serum lanthanum concentration. Liver function tests in the patients studied (alanine aminotransferase, gamma-glutamyltransferase, alkaline phosphatase and total bilirubin) were all within the reference interval.
Discussion
Methods have been described for measuring lanthanum in bovine and in human whole blood by quadrupole ICP-MS. However, these methods require aggressive sample preparation, making them prone to contamination from the reagents used. 9,10 Hence the simplified method described here for measuring plasma/serum lanthanum was developed. Currently, there are neither external quality control materials nor external quality assessment schemes for lanthanum in biological fluids, hence the external validation arranged for the IQC solutions.
That the concentration of lanthanum in 1% aqueous nitric acid that had been drawn through a stainless steel needle was below the LLoQ is consistent with the findings of Rodushkin and Odman. 11 In their studies using sector field ICP-MS, lanthanum and gadolinium (limits of detection: 0.42 and 1.2 ng/L, respectively) were not detected in aqueous nitric acid (0.05 mol/L) that had been drawn through a disposable Vacutainer stainless steel needle (BD). Therefore, contamination from the needle during collection may not be of concern when measurement of lanthanum is required.
Rodushkin and Odman 11 found lanthanum ion (0.38–4.2 μg/L) in five different commercially available blood collection tubes after adding dilute nitric acid (0.05 mol/L) to the tubes and allowing to stand overnight. However, use of dilute nitric acid does not provide an ideal comparison to body fluids since it does not contain the many ligands present in biological fluids that have an affinity for trace elements, such as amino acids and peptides. 12
Blood collection tubes containing K3EDTA showed the poorest recovery of lanthanum at both concentrations in this study. In contrast, Rodushkin and Odman 11 reported that dilute nitric acid (0.05 mol/L) collected in tubes containing K3EDTA showed increases in lanthanum ion concentration on standing. This disparity in findings could be accounted for by differences in tube manufacturer and the difference in matrix used to assess contamination (i.e. dilute nitric acid compared with whole blood). Indeed, Rodushkin and Odman 11 found that lanthanum and gadolinium concentrations measured in blood collected tubes from different manufacturers varied by >50%. For the measurement of serum/plasma lanthanum ion, whole blood should be collected into tubes containing lithium heparin or serum separator tubes; blood tubes containing EDTA should be avoided.
Serum lanthanum in healthy volunteers, measured by ICP-MS, has been reported to be 0.06 μg/L. 10 Subsequently, Joy and Finn 13 reported plasma lanthanum concentrations in haemodialysis patients (n = 126) not prescribed lanthanum carbonate to be <0.03 μg/L. These findings are consistent with those reported here: the serum lanthanum concentration in haemodialysis patients not prescribed lanthanum carbonate was <0.1 μg/L. As to plasma lanthanum in patients prescribed lanthanum carbonate, concentrations in the range of 0.07–3.29 μg/L after 24 weeks of treatment at doses of 900–2500 mg/d have been reported. 14 Lanthanum measured in samples from patients prescribed lanthanum carbonate in phase II and III studies rarely exceeded 1.0 μg/L (dose up to 2500 mg/d). 15 These findings are comparable to those obtained in this survey.
Damment and Pennick 15 did report a relationship between lanthanum dose and the plasma lanthanum concentration, a relationship that was not apparent in the patients studied here. Impaired liver function and hence reduced lanthanum excretion could account for the variation in serum lanthanum concentrations observed in the present survey, as could drug–drug interactions. However, liver function tests were normal and lanthanum is not a substrate for enzymes such as the cytochrome P450 mono-oxygenase system. A more likely cause for the lack of a relationship between prescribed dose and serum lanthanum in the present study is variable adherence.
Conclusions
The method described here has proved capable of measuring lanthanum in serum from patients prescribed lanthanum carbonate. However, the serum lanthanum concentrations measured were very low in relation to the prescribed dose. Moreover, there were was no clear relationship between dose of lanthanum carbonate and the serum lanthanum concentration measured, which may have been due to poor adherence to the prescribed regimen.
DECLARATIONS