Analytical interference of intravascular contrast agents with clinical laboratory tests: a joint guideline by the ESUR Contrast Media Safety Committee and the Preanalytical Phase Working Group of the EFLM Science Committee

Iodine-based contrast media

The effect of ICM on clinical assays has not been studied systematically or extensively. Depending on the method and ICM used, interference may be clinically relevant [3]. M-protein analysis is paramount in the diagnosis and monitoring of monoclonal gammopathy [4]. Several studies report the interference of ICM on the spectrophotometric detection of monoclonal protein analysis in urine and blood by capillary zone electrophoresis with spectrophotometric detection (CZE-UV) [5]. ICM absorbs UV light at a similar wavelength as the peptide bonds in m-proteins, thereby mimicking the presence of (M-) proteins in the commonly used CZE analysis with UV detection. In contrast, Capaldo and co-workers [6] demonstrated that the opposite may also occur, i.e. masking of an M-protein peak. In the M-protein analysis by CZE-UV, a duplication in the beta-2 fraction, which was at first assigned to ICM (iomeprol) interference and the beta-1 fraction, did not display any M-protein peak. These specific cases demonstrate that ICMs may cause incorrect detection of an M-protein, resulting in unnecessary diagnostics and/or treatment or on the other hand missing an M-protein thereby delaying treatment.

Otnes and co-workers investigated the analytical interference of two specific ICM, iodixanol and iomeprol [2] in vitro. They reported in the high, but clinically relevant, concentration range of the ICM, either a positive bias (colorimetric calcium assay) or a negative bias, i.e. colorimetric iron, magnesium, and zinc assay as well as in the direct potentiometric sodium assay. Other assays did not show any interference with both ICM.

ICM can affect immunoassays differently, depending on the manufacturer. Such an example is when evaluating cardiac troponin-I in patients undergoing coronary angiography. When evaluating two different assays, the Opus Magnum (Behring Diagnostics) and the Access (Beckman Coulter, Inc.) using 12 different ICM, Lin et al. [7] showed that the outcome of the Opus system was affected when performed directly after the coronary angiography procedure, but not after 30 min in patients with normal kidney function. In patients with reduced kidney function, the interference lasted longer. The access assay did not show any interference.

An interference by iohexol on endocrine immunoassays was observed by Loh and co-workers in in-vitro experiments [8]. They reported that soon after contrast administration, iohexol may affect follicle-stimulating hormone (FSH), luteinizing hormone (LH), plasma renin activity (PRA) and thyroid stimulating hormone (TSH) measurements by different manufacturers, either over- or underestimating the true value. The interference on immunoassays may be explained by either the presence of an unidentified antigenic site on the contrast medium molecule blocking or cross-reacting with antibodies of the immunoassay, dilutional effects due to the high osmolar aspects of iohexol and/or, as described before, due to spectrophotometric aspects of the ICM, interfering with UV-detection. No other ICM were studied by Loh et al. and most of the interference effects were seen only at very high iohexol concentrations, which is very uncommon in clinical practice.

Next to the photometric aspects of ICM, the analysis of the specific gravity of urine uses the refractive index. A higher refractive index due to the presence of the ICM in urine may produce false results [9], [10], [11].

Besides interference in laboratory testing, sample integrity and quality may be impacted [12].

Since the density of blood is altered due to the presence of ICM in the blood, the gel cell separator characteristics may be altered, resulting in incorrect plasma or serum collection [13], [14], [15] and thereby causing mechanical problems by clogging sample needles in the routine platforms. Table 1 shows demonstrated ICM interference on clinical laboratory tests.

Gadolinium-based contrast agents

Since their clinical approval and introduction in 1988, GBCAs have been administered in 750 million standard doses (Bayer Healthcare, estimated from multiple internal and external sources). Several interferences on laboratory tests have been described, ranging from commonly used laboratory tests [12] to more specialized laboratory tests [18]. The probably most clinically relevant interference is the interference of GBCAs on serum calcium measurement by specific colorimetric assays. Gadodiamide [19], [20], [21], [22] and gadoversetamide [7] are the GBCAs most frequently reported to interfere. Those are no longer on the market, but it should be noted that the interference has been shown to occur irrespective of the molecular configuration of the chelate (i.e. linear or macrocyclic and ionic or non-ionic) [20], although the largest interference was observed on GBCAs with a linear molecular configuration of the ligand [23]. This interference has not been observed with other serum calcium measurement tests, e.g. Ca-specific electrode, atomic absorption or mass spectrometry.

In an in vitro study, Proctor and co-workers [24] investigated the analytical interference of four GBCAs on multiple analytes and multiple analysers. They demonstrated that depending on the specific GBCA a positive and negative analytical interference is observed, which is most prominent in Angiotensin Converting Enzyme (ACE), calcium, iron, total iron binding capacity (TIBC), magnesium and zinc. Mechanistically, all the affected analytes are either endogenous divalent cations or somehow use divalent cations in the reaction of the laboratory test. Gd³⁺ can interact with the analyte of interest (e.g. trans-metallation), thereby potentially interrupting the analytical process, or in colorimetric assays by binding with the chromophore [23]. In an in-vitro experiment, Otnes and co-workers demonstrated a similar interference by the GBCAs gadoxetate disodium, gadoterate meglumine, and gadobutrol on iron and zinc (negative bias) assays. Other 29 clinical tests did not display any clinically relevant interference by these GBCAs [2].

In the field of trace elements and heavy metals, inductively coupled plasma mass spectrometry (ICP-MS) is the golden standard. Gd³⁺ may interfere also with this technique in multiple ways, i.e. space-charge effects, interference in the mass spectrometric analysis by double-charged ions and polyatomic interference [18]. The latter can be circumvented by applying the correct analytical technique. Especially the analysis of selenium by ICP-MS may be complicated by the presence of ¹⁵⁶Gd due to similar mass-to-charge ratios. Gd ions may also interfere with the ionization process, suppressing ions of analytes, e.g. trace elements or (toxic) heavy metals and internal standards used.

An increase in urinary Zn and Cu concentration was seen, especially with gadodiamide [25]. This increase is, as the authors hypothesized, possibly related to in vivo transmetallation and not to a true analytical interference, and was therefore excluded.

Table 2 shows described GBCA interference on clinical laboratory tests.

Clearance of contrast media

Most studies on iodine-based contrast media (ICM) employ an open, 2-compartment model for pharmacokinetic analyses. The first compartment is the plasma in which the molecules are being diluted and the second compartment is the extracellular volume, excluding the brain (due to the blood-brain barrier). The plasma concentration decays by distribution of the ICM from plasma to the extracellular volume (distribution phase, rate constant α), and by elimination of the CM from plasma to urine by renal excretion (elimination phase, rate constant β).

Biodistribution studies have suggested that an open 3-compartment model may better fit the pharmacokinetic data of GBCA. The second and third compartments are the extravascular extracellular spaces of rapidly and slowly equilibrating tissues (storage compartment (of unknown exact composition)). Apart from the distribution phase and the rapid (renal) elimination phase, there is a slow residual excretion phase that is species-independent and whose rate constant γ is closely related to the thermodynamic stability of the specific GBCA molecule [27, 28].

Contrast media are eliminated through glomerular filtration. In addition, the liver-specific GBCAs have partial hepatic excretion of up to 50 % of the intravascular administered dose. With a normal glomerular filtration rate (GFR), 90 mL/min/1.73 m², the half-life in plasma is about 2 h, roughly for both ICM and GBCA, although in normal renal function, half-life is on average somewhat shorter for GBCA. In patients with advanced renal function loss with a GFR<30 mL/min/1.73 m², the half-life may increase up to 30 h [28]. Near-complete elimination to 1.5 % of the original concentration occurs after six elimination half-lives. Thus, to avoid interference from contrast media, sampling should be delayed as outlined in Table 3, depending on the renal function of the patient.