Compliance with ethical standards
The prospective study was approved by the institutional review board as well as the competent federal radiation protection authority and was performed in accordance with the ethical standards laid down in the Declaration of Helsinki. Informed consent was obtained individually from all participants included in the study.
Volunteers
The study was designed as a prospective, three-armed trial with volunteers exposed to [18F]FDG alone (study arm SA1), the SMF alone (SA2), and both agents combined (SA3). The null hypothesis was that there is no difference between the study arms. To obtain a power of 80% (Kruskal-Wallis test), given an effect greater than 1.5-fold of the inter-individual variance within the study arms, a minimum of 10 participants had to be investigated per arm.
Included were younger individuals not diagnosed with any disease that may alter DNA damage and repair [29]. Contraindications were an existing pregnancy, blood glucose concentrations of less than 80 or more than 180 mg/dl in case of an exposure to [18F]FDG, as well as active or passive implants or other objects of ferromagnetic or unknown material in case of an exposure to the SMF. The volunteers received a payment depending on the study arm [29].
Exposure scenarios, blood sampling, and incubation
As schematically shown in Fig. 1, the γH2AX response to a prolonged exposure of lymphocytes to [18F]FDG is markedly delayed compared to a quasi-instantaneous X-ray exposure and should thus be observed over a longer period. On the other hand, the time the volunteers had to lie still in the imaging systems to acquire activity concentration-time courses for dosimetry had to be limited. To balance these aspects, volunteers were exposed in vivo over 60 min. Nevertheless, in order to extend the duration of blood exposure to [18F]FDG as well as the time for the formation of exposure-induced γH2AX foci, blood was taken at the end of this period (sampling point SP3) and incubated in vitro up to 60 min, before lymphocytes were fixed. To determine the background of γH2AX foci, blood was also collected prior to the exposure phase (SP1, SP2). For each of the three study arms, the examination protocol was adapted as far as necessary to the respective exposure and measurement conditions, as illustrated by the flow charts in Fig. 2. Details are described in the following.
Volunteers were exposed to [18F]FDG alone (SA1) at a PET/CT system (Biograph mCT; Siemens Healthineers, Erlangen, Germany) or to a 3-T SMF alone (SA2) and in combination with the radiotracer (SA3) at a PET/MRI system (Biograph mMR; Siemens). Technical details of the systems can be found in [30].
Study arm 1
Following the acquisition of an X-ray topogram (tube voltage, 120 kV; tube current, 25 mAs) and administration of a body weight-adapted [18F]FDG activity (nominal, 4.3 MBq/kg) through an antecubital vein, emission data were detected in the list mode over 60 min from an abdominal body region. For quantitative evaluation of activity concentrations, attenuation-corrected PET images were reconstructed using low-dose CT data acquired after blood sampling at SP3 from the same body region. This time sequence avoids a substantial additional exposure of lymphocytes taken at SP3 to X-rays that would otherwise interfere with the effects of [18F]FDG and the SMF to be investigated. PET images were reconstructed for 55 non-equidistant time frames. For dosimetry, the average [18F]FDG activity concentration in blood was determined for each time frame by evaluating small circular regions of interest placed over the centre of the aorta on 11 transversal PET images in the middle of the axial scan region. Blood was collected before (SP1) and after the topogram (SP2) as well as 60 min post [18F]FDG injection (SP3).
Study arm 2
Blood was withdrawn from volunteers before (SP1) and after (SP3) exposure to the 3-T SMF of the PET/MRI system.
Study arm 3
MR images required for transmission correction of emission data were acquired for technical reasons at the beginning of the examination procedure using a T1-weighted Dixon VIBE sequence (specific absorption rate, < 2 W/kg). Apart from that, blood sampling as well as data and image postprocessing were identical to study arm 1.
Blood was collected from the antecubital vein contralateral to the site of [18F]FDG administration via an 18- or 20-gauge indwelling cannula (Vasofix Safety IV; B. Braun Melsungen AG, Melsungen, Germany) into 7.5-ml lithium heparin monovettes (S-Monovette; Sarstedt AG & Co, Nümbrecht, Germany). Two monovettes were filled at SP1 and one at SP2. At SP3, blood was sequentially withdrawn into three monovettes, mixed together in 50-ml tubes (Falcon; Fisher Scientific GmbH, Schwerte, Germany) and then immediately portioned into five 15-ml centrifuge tubes (Falcon; Fisher Scientific GmbH). By this procedure, variations amongst sequentially collected blood were avoided as a source of error.
Blood samples were incubated in part at 37 °C over different periods as indicated in Fig. 2. Pre-exposure samples taken at SP1 were incubated over 0 and 60 min to detect and—if present—consider potential distortive effects of venipuncture and incubation on data analysis. Due to a relevant time delay between the potential induction of DSBs by the X-ray topogram (SA1) or MRI scan (SA3) and the formation of detectable γH2AX foci, blood samples withdrawn at SP2 were incubated over 60 min. To extend the exposure of lymphocytes to residual [18F]FDG activity in the samples and to assess a potential continuing influence of the SMF on the genotoxicity of [18F]FDG beyond the end of field exposure, samples taken at SP3 were incubated over 0, 15, 30, 45, and 60 min.
Hereinafter, the biological condition of lymphocyte DNA isolated from blood samples is characterised by the study arm (SA), the blood sampling point (SP), and the incubation time (IT).
Internal dosimetry
To estimate the total absorbed dose to blood as a surrogate for the dose to circulating lymphocytes, two contributions were considered: (i) the dose delivered in vivo to blood during the first 60 min p.i. by self-irradiation as well as cross-fire irradiation from other body tissues. (ii) The additional dose delivered subsequently in vitro during incubation, due to self-irradiation from the residual [18F]FDG activity in the samples.
In vivo exposure
The time-course of [18F]FDG activity in blood was estimated from measured activity concentration-time courses, using blood volumes determined from the body weight and height of the volunteers [31]. Both the self and cross-fire contributions to the total blood dose were estimated according to the MIRD approach [32]. In contrast to blood, the activities in other body tissues were not available for the volunteers and hence were estimated by using the model of Hays and Segall [33]. The required specific absorbed fractions (SAFs) were computed for the adult ICRP male and female reference voxel phantoms [34] and monoenergetic positron sources with the Monte Carlo code EGSnrc [35, 36]. Finally, the resulting monoenergetic data were convoluted with the β+-spectrum of 18F [37].
In vitro exposure
The exact geometries of the used 15-ml conical centrifuge tubes containing different predefined blood volumes were created using commercial software (Rhinoceros V5.0; Robert McNeel & Associates, Seattle, USA) in polygon-mesh format and voxelised. SAFs for the digitalised geometries were determined by Monte Carlo computations. For blood volumes deviating from the five predefined values, SAFs were estimated via a logarithmic interpolation of the simulated data. Doses delivered in vitro to the samples over incubation periods of 15, 30, 45, or 60 min were computed using the volunteer-specific residual activity concentrations in the aorta 60 min p.i as initial value.
Processing of blood specimen and γH2AX assay
Following incubation, blood samples were kept at 5 °C until isolation of peripheral blood leucocytes. This was performed by density gradient centrifugation (10 min, 1000g, 5 °C) using 12-ml separation tubes (Leucosep Tube; Greiner Bio-One GmbH, Frickenhausen, Germany) and separation medium (Histopaque-1077; Sigma Aldrich Chemie GmbH, Taufkirchen, Germany). After centrifugation, leucocytes were visible and transferred into 5 ml cell culture medium (RPMI 1640; Pan-Biotech GmbH, Aidenbach, Germany). Cell suspension was centrifuged again (10 min, 250g, 5 °C), the medium decanted, and cells fixed in 1 ml 2% paraformaldehyde (PFA; Sigma Aldrich)/phosphate buffered saline (Dulbecco’s PBS; Biochrom GmbH, Berlin, Germany) solution for 15 min at 5 °C before centrifugation (10 min, 250g, 5 °C). After decanting the solution, lymphocytes were stained with Tuerk’s solution (Merck, Darmstadt, Germany), counted in a counting chamber and concentrated to one million cells per ml in PBS. Cell suspensions were stored at 5 °C.
One hundred microliters of cell suspension was spotted onto glass slides by cytospin centrifugation for 5 min at 54g. Slides were washed three times in fresh PBS containing 0.15% of a nonionic surfactant (TritonX-100; Sigma Aldrich) each time for 5 min, followed by three washing steps in blocking solution (1 g bovine serum albumin (BSA; Sigma Aldrich) mixed with 0.15 g glycine (Sigma Aldrich) in 100 ml PBS) each for 10 min. Seventy-five microliters blocking solution with anti-phosphohistone H2A.X (Ser139) rabbit mAb (Cell Signaling Technology Europe B.V., Frankfurt a.M., Germany) in the dilution 1:200 was transferred on each slide and incubated at 4 °C for at least 16 h. Slides were washed again for 5 min in PBS, for 10 min in PBS/Triton and for 5 min in PBS. Before incubating with the secondary antibody, an anti-rabbit IgG (H+L), F(ab')2 fragment conjugated to Alexa Fluor 555 fluorescent dye (Cell Signaling Technology Europe), in the dilution 1:1000 in blocking solution in a humid chamber for 45 min at room temperature, slides were treated with blocking solution for 7 min. After antibody binding, slides were washed twice in PBS/Triton for 5 min each, once in PBS for 10 min and once in PBS for 7 min. Cell nuclei were counterstained with Hoechst 33342 (Bisbenzimide H 33342 trihydrochloride; Sigma Aldrich) for 2 min and slides were washed again in PBS for 2 min twice. Finally, slides were covered by 16 μl antifade mounting medium (Vectashield; Vector Laboratories Inc., Burlingame, USA).
Search and image acquisition of cell nuclei on the slides was performed by automatic fluorescence microscopy using a scanning and imaging platform (Metafer 4, version V3.13.1; MetaSystems Hard & Software GmbH, Altlussheim, Germany) equipped with an objective (ZeissPlan-Neofluar 40×/0.75; Carl Zeiss Microscopy GmbH, Jena, Germany) yielding a 400-fold magnification. For foci analysis a Spectrum Orange bandpass filter (excitation: centre wavelength/bandwidth = 546/10 nm, emission 580/30 nm; Chroma 31003; Chroma Technology, Olching, Germany) and for counterstaining a DAPI bandpass filter (excitation 350/50 nm, emission 460/50 nm; Chroma 31000; Chroma Technology) was used. The threshold for foci intensity was set to 70% with a maximum gain of 500%. For lymphocytes exposed to very low doses and controls, cells with a high number of foci are not to be expected. Therefore, cells with more than five detected ‘foci’ were excluded from data analysis to eliminate potential artefacts from the automatic foci detection algorithm.
Statistical analysis
Data analysis was performed using SigmaPlot (version 13.0; Systat Software GmbH Erkrath, Germany). The central tendency and variability of data were described by the arithmetic mean and the standard deviation (SD), the distribution of data was represented by box-and-whisker plots. The Wilcoxon signed-rank and Mann-Whitney tests were run to test for significant differences between two groups of paired and unpaired samples, respectively. For three or more groups, for comparisons with repeated measures within study arms the Friedman test was used, for comparisons without repeated measures across study arms the Kruskal-Wallis test. In case of a significant result, Tukey’s or Dunns’s post hoc tests, respectively, were used to pairwise identify groups with significant differences. To test the null hypothesis of equal variances between groups, the robust Brown-Forsythe test was applied. All tests were performed as two-sided tests and p values of less than 0.05 were considered as significant. To quantify the effect of different exposures, excess rates, defined as difference between post- and pre-exposure rates, were computed.