Research Article Received 4 November 2014,
Revised 14 February 2015,
Accepted 19 February 2015
Published online 27 May 2015 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3283
Radiolabelling rituximab with 99mTc in three steps procedure Charlotte Fontan,a,b* Christine Bezombes,b Anne Sophie Salabert,c Julien Costes,b Raphael Lopez,c Jean-Jacques Fournie,b Hervé Avet-Loiseau,b Yvon Coulais,d Pierre Payoux,c and Mathieu Tafanic Lymphomas are the most frequent haematological malignancy. In non-Hodgkin’s lymphomas (NHL), more than 90% of tumor cells express the cluster of differentiation (CD) 20 antigen. At the end of frontline therapy, the evaluation of remission is based on computed tomography (CT) and positron emission tomography coupled with computer tomography (PET/CT) with [18F]-fluorodeoxyglucose ([18F]FDG). Unfortunately, these techniques are not specific and cannot distinguish residual active tumor from inflammation. The aim of this study was to develop a specific radiotracer of NHL CD 20+ cells for clinical applications. The radiolabelling technique presented, based on the use of tricarbonyl compound, does not include an antibody reduction because this step could damage the protein. Actually, rituximab, an anti-CD 20 chimeric antibody used for the treatment of these NHL, was radiolabelled with IsolinkW 99mTc-tricarbonyl compound in a three-step procedure without using a specific antibody reducer. Radiolabelling yield was greater than 97%. In vitro experiments showed a conservation of antibody integrity. In vivo experiments using single-photon emission computed tomography/CT showed significant tumor targeting 24 h after injection of the radiotracer. It was consequently possible to develop an immunoradiolabelling method to specifically detect the residual disease. As this procedure is fast, reproducible and gentle, it will be possible to comply with Good Manufacturing Practices. Keywords: tricarbonyl compound; immuno imaging; SPECT/CT; rituximab; lymphoma
Introduction
274
In the US, about 500,000 people have lymphoma, and approximately 332,000 have non-Hodgkin’s lymphomas (NHLs). Every year, over 65,000 cases of NHL are diagnosed in America, and 85% are B-cell NHL.1 Treatment of these diseases has improved with the development of the monoclonal antibody rituximab, the first chimeric mouse/human immunoglobulin G1 (IgG1) directed against transmembrane antigen cluster of differentiation (CD) 20, which is expressed in more than 90% of B-cell NHLs.2 Diagnosis, evaluation of extension, response to treatment, or follow-up of NHLs are performed by modern imaging techniques such as positron emission tomography coupled with computed tomography (PET/CT). Hybrid [18 F]-fluorodeoxyglucose PET/CT is a method based on the fusion of an anatomical image from CT and a metabolic image produced by PET. Even though this method provides high sensitivity, it is not specific to NHLs and cannot be used for indolent lymphomas with slow growth and weak consumption of glucose. Moreover, it cannot distinguish between residual tumor and inflammation. Consequently, the development of a specific and sensitive compound for NHLs would to be of interest. Because of the widespread expression of CD 20 on NHL tumor cells, radiolabelling rituximab with a positron emitter3 or a single photon emitting isotope, such as Technetium-99 m (99mTc), constitutes an approach specific to the disease. Single photon emission computed tomography (SPECT) is a method of choice because of its sensitivity (10 14 moles of radionuclides detected), the depth of study without
J. Label Compd. Radiopharm 2015, 58 274–280
any limitation, and its resolution about 1 mm.4,5 Moreover, the availability of radionuclides like 99mTc (generator 99Mo/99mTc) and its convenient dosimetry (half-life = 6 h, gamma energy = 140 keV) make it an isotope of choice. During recent years, several methods based on radiolabelling rituximab with 99m Tc have been described.6–10 All of them include two reduction steps, one dedicated to the reduction of Technetium-99 m, the other specifically targeting the antibody, with either photoactivation9,10 or 2-mercaptoethanol (2-ME).6–8 This specific antibody reduction step creates thiol residues that facilitate technetium binding. However, this specific reduction could modify the protein structure and decrease antigen recognition. Since 2003, the Mallinckrodt Company has marketed IsolinkW containing tartrate, carbonate, and boranocarbonate, which decrease the oxidation number of native pertechnetate (99mTcO4 ; +VII) to +I. The moiety obtained is called the ‘technetium core’ [99mTc(CO)3(H2O)3] and is resistant to oxidation a
Radiopharmacy, University Hospital, Toulouse, France
b
The Cancer Research Center of Toulouse, Toulouse, France
c
UMR 825, INSERM, Toulouse, France
d
Nuclear medicine, University Hospital, Toulouse, France
*Correspondence to: Charlotte Fontan, Radiopharmacy, University Hospital, Toulouse, France. E-mail: [email protected]
Copyright © 2015 John Wiley & Sons, Ltd.
C. Fontan et al. by oxygen.11 The three molecules of water can be replaced by ligands containing N-heterocycle, thioether, thiolate functions,12 which are present in amino acids like histidine, cysteine, and methionine.11–13 Theoretically, radiolabelling of rituximab protein with the technetium core is possible without any additional antibody reduction step. We hypothesized that [99mTc(CO)3]rituximab would be stable and the properties of the high molecular weight (above 143.9 kDa) protein would not be altered. The purpose of this study was to radiolabel rituximab with Technetium-99 m by a method using a tricarbonyl compound without any additional reduction step, enabling to maintain the antibody’s structure and function. We evaluated the effectiveness of the radiolabelling by testing the radioligand in vitro and in vivo in mice xenografted with human Burkitt’s lymphoma.
Experimental Reagents and equipment
Step 1: Native buffer substitution In commercial form, MabtheraW contains several excipients, such as sodium citrate, polysorbate 80, sodium chloride, water for injection, sodium hydroxide, and hydrochloric acid. This native buffer had to be replaced because polysorbate 80 and sodium citrate can bind pertechnetate. Elution of rituximab (10 mg/mL, 1 mL) from the PD10 column is performed with nitrogen-purged saline solution (0.9%). The final concentration of rituximab was about 5 to 8 mg/mL.
Step 2: Preparation of
99m
Tc(CO)3(H2O)3
Technetium-99 m must be in the +I oxidation state to complex with rituximab. Sodium pertechnetate was reduced by the tricarbonyl compound IsolinkW. In comparison with classical radiolabelling methods, which used one kit of reducer for each radiolabelling,6,7 our method used a small amount of it. Indeed, IsolinkW powder (4 mg) was mixed with 99m TCO4 (200 μL, 1221 MBq). The mixture was heated (100 °C, 30 min). The pH of the solution was adjusted to 7 with acetic acid (0.2 M). Radiochemical purity was >95% (data not shown)
Step 3: Rituximab radiolabelling with
Reagents
99m
Tc(CO)3(H2O)3
99m
For radiotracer synthesis, MabtheraW was purchased from Roche, Inc. 99m (Switzerland, Bale). Pertechnetate [ TcO4 ] was eluted in saline 99 99m solution from a Mo/ Tc generator (IBA, Saclay, France), and IsolinkW was provided by Mallinckrodt Inc. (Petten, Netherlands). For cell culture and in vitro testing, we used RPMI 1640 (Gibco, Thermoscientific, France) supplemented with 10% fetal calf serum (Gibco, Thermoscientific, France), glutamine (2 mM, Gibco, Thermoscientific, France), streptomycin (10 μg/mL, Gibco, Thermoscientific, France), and penicillin (200 U/mL, Invitrogen, Life Technologies, France). Immunoreactivity was performed with phosphate-buffered saline (PBS, pH 7.4 Sigma Aldrich, Germany) and Bovine Serum Albumin (BSA, Rockland-inc). To reduce oxidation state of pertechnetate, IsolinkW was used. This kit contains 8.5 mg sodium tartrate, 2.85 mg Na2B4O7 10H2O, 7.15 mg of sodium carbonate, and 4.5 mg of sodium boranocarbonate in lyophilized form totalling 23 mg of powder. To establish a comparison with conventional methods, specific antibody reduction was performed using 2-ME (Sigma Aldrich, Germany).
Equipment
Tc(CO)3(H2O)3 (20 mg/mL, 50 μL) was added to rituximab (6,5 mg/mL, 100 μL), and the mixture was incubated (37 °C, 30 min). 99m The final concentration of Tc(CO)3rituximab was about 4 mg/mL, and the activity was about 250 MBq.
Specific rituximab reduction with 2-mercaptoethanol Rituximab (10 mg/mL,1 mL) was reduced with 2-ME in a molar ratio of 2-ME:rituximab = 10.000:16 or 2.000:17 for 30 min at room temperature, and the mixture was purified using Solid Phase Extraction with a mobile phase containing nitrogen-purged saline solution (0.9%). The final concentration was less than 1 mg/mL.
Quality control The quality checks were performed by HPLC with a size exclusion column. The eluent was NaCl 0.9% and the flow rate 2 mL/min. A UV detector operating at 280 nm was used to detect proteins containing amino acids with aromatic radicals. Another detector monitored 140 keV technetium-99 m gamma rays. After synthesis, in vivo stability towards metabolic enzymes was tested 99m by adding aliquots of [ Tc(CO)3]rituximab (4 mg/mL, 50 μL) to 100 μL fresh human plasma. They were placed in an incubator at 37 °C. The stability was evaluated with a radiochromatography after incubation times of 4 and 20 h.
Culture of cancer cells Daudi or Raji Burkitt’s lymphoma cells were cultured at 37 °C in 5% CO2 humidified atmosphere in a complete medium. Every 4 days, cells viability was measured with a cell counter.
In vitro experiments Immunoreactivity of [99mTc(CO)3]rituximab 99m
The immunoreactivity of [ Tc(CO)3]rituximab was measured by the Lindmo method.14 Four concentrations of Daudi lymphoma cells were 6 6 used (from 0.6 × 10 to 10 × 10 cells/mL). Before starting, the cells were washed twice with 1% BSA/PBS buffer. A constant concentration of 99m [ Tc(CO)3]rituximab (6 ng/mL) was added. Total volume was 1 mL in 1% BSA/PBS buffer. To test non-specific binding, a 1000-fold excess of unlabelled rituximab was added. Tubes were stirred for 2 h at 4 °C. After incubation, the tubes were centrifuged (1600 xg, 5 min). The cells were washed twice in 1% BSA/PBS, and the supernatant was removed. The radioactivity of the pellets was measured with a gamma counter. The
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org
275
During culture of lymphoma cell lines, viability was measured with a cell counter (AdamW, Alphametrix Rodermark). Native buffer was substituted thanks to a Solid Phase Extraction column (Sephadex PD 10, GE-Healthcare). To assess quality control, High Performance Liquid Chromatography (HPLC; ion chromatography system (ICS)) was used. It is a system equipped with a size-exclusion column (Cluzeau Info Labo (CIL) Stability Cluzeau 200 SEC-5 μm 300 × 8 mm) and two channels, one for radioactivity detection (NaI Cristal, GabiW Raytest, Germany) and the second for ultra violet (UV) detection at 280 nm (lambda 1010, ICS). Run acquisition in UV and gamma appeared on ICS Software. Time for runs was 10 or 15 min. 99m Tc(CO)3(H2O)3 were Stability in plasma and radiochemical purity of evaluated by Thin Layer Chromatography with MinigitaW, (Raytest, Germany) radioachromatograph using an AlugramW Thin Layer Chromatography (Servilab, France) strip. The mobile phase was a methanol/water mixture (90/10, v/v). The energy of the gamma radiation of technetium-99 m (140 keV) contained in the samples was measured with a gamma counter (Automatic Gamma counter 1470 Wizard 3 W, PerkinElmer, Massachusetts, USA). The interpretation of binding assay was performed by GraphPad Prism Version 6 Software (GraphPad Software, Inc, La Jolla, USA). In vivo tests were achieved using a nanoSPECT/CTW camera, Biospace, France. It consisted of four collimators, each containing nine pinholes. Imaging resolution was about 0.9 mm for SPECT and about 100 μm for CT. Image interpretation used the In VivoscopeW Software.
J. Label Compd. Radiopharm 2015, 58 274–280
Radiolabelling of the antibody
C. Fontan et al. experiment was carried out in triplicate. For each cell concentration, specific binding counts were obtained by subtracting non-specific binding from total binding counts. A graph representing total applied activity/specific binding in ordinate and reciprocal of concentration cells in abscissa was drawn. The 99m immunoreactivity of [ Tc(CO)3]rituximab was obtained by the reciprocal of the intercept on the ordinate of the linear regression line.
Binding of [99mTc(CO)3]rituximab 99m
The binding capacity of [ Tc(CO)3]rituximab was studied with a 6 constant concentration of Daudi cells (1 × 10 cells/mL of RPMI 1640 99m culture medium). Four concentrations of [ Tc(CO)3]rituximab were used (1.5 ng/mL to 4 ng/mL). Before starting, the cells were washed twice with RPMI buffer. For each concentration, two conditions were used to determine total and non-specific binding. In the latter case, a 1000-fold excess of unlabelled rituximab was added, and the tubes were incubated and centrifuged as mentioned earlier. The cells were washed twice in RPMI 1640 culture medium and the supernatant was removed. The radioactivity of the pellets was measured with a gamma counter. The 99m experiment was performed in triplicate. For each [ Tc(CO)3]rituximab concentration tested, specific binding counts were obtained by subtracting non-specific binding from total binding counts. Determination of Kd was obtained thanks to GraphPad Prism Version 6 Software using total, non-specific and specific binding.
Acquisitions on SPECT-CT were performed on three mice 5 and 24 h after i.v. injection of radiotracer. 99m Animals were sacrificed 24 h after i.v injection of [ Tc(CO)3] rituximab (130 μg, median activity = 8.14 MBq). For five mice, organs were weighed and the radioactivity in each organ counted with a gamma counter. Results are expressed as a percentage of the injected dose per gram of organ (%ID/g). To measure the real activity injected, we weighed the syringe before and after injection. Injected dose (expressed in counts per minutes (cpm)) was calculated by comparison 99m with a weighed sample of 10 μL of [ Tc(CO)3]rituximab placed in the gamma counter. Then, we compared it with the number of cpm in each organ studied.
Image acquisition and interpretation Protocol acquisition. For the CT, maximum voltage was 45 kV, and signal accumulation time per position of the generator was 500 ms. Number of projections was 180. For SPECT, whole body planar images registered 10,000 counts per second (cps) for each projection.
Interpretation of images In VivoscopeW Software CT reconstruction was fast (smoothing: 35%, resolution 67%, Iteration: 2*3). SPECT reconstruction used nih-fire2 as color scale.
Statistical analysis
In vivo experiment Experimental animals Mice were housed in a pathogen-free facility and handled in accordance with the procedures outlined in Council Directive 86/609/European Economic Community. The investigation respected the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH publication No. 85–23, revised 1996). Male severe combined immunodeficiency (SCID)-beige mice aged between 12 and 16 weeks were used. They were injected subcutaneously on the flank 6 6 with 6 × 10 Raji cells or 10 × 10 Daudi cells. The experiments were performed when tumors were detectable. For all experiments, the mice were anesthetized with isoflurane.
Results are expressed as mean ± standard deviation. Statistical comparison of the data was performed using the t-test for comparison between two groups. A value of p < 0.05 was considered as significant.
Results and discussion Results Comparison of rituximab radiolabelling methods with and without antibody reduction As shown in Table 1, our radiolabelling method was faster and used smaller amounts of reagents than the method described
Table 1. Comparison of rituximab radiolabelling methods with and without antibody reduction Method WITH antibody reduction step
Method WITHOUT antibody reduction step
STEP 1: Specific antibody reduction MabtheraW incubated with molar excess of 2-ME : 10.000:1 or 2;000:1 30 min at room temperature
STEP 1: Substitution of MabtheraW native buffer By Solid Phase Extraction in nitrogen-purged saline (0.9%)
STEP 2: Purification and substitution of MabtheraW native buffer By Solid Phase Extraction in nitrogen-purged saline (0.9%)
STEP 2: Preparation of 99mTc(CO)3(H2O)3
STEP 3: Preparation of 99mTc(CO)3(H2O)3 One IsolinkW kit vial (23 mg) with 99mTcCO4 (1 mL) 40 s at 150 °C pH adjusted to 7
A part of IsolinkW kit vial (4 mg) + 99mTcCO4 (200 μL,1221 MBq) 30 min at 100 °C pH adjusted to 7 STEP 3: Rituximab radiolabelling Rituximab (0,65 mg, 150 μL) + 99mTc(CO)3(H2O)3 (50 μL) 30 min at 37 °C
STEP 4: Rituximab radiolabelling 3 h at 37 °C
276
NUMBER OF STEPS: 4 TIME OF RADIOLABELLING: 3 h 30 min
www.jlcr.org
NUMBER OF STEPS: 3 TIME OF RADIOLABELLING: 1 h
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 274–280
C. Fontan et al. previously.6,7 After substitution of native buffer, there were no excipients absorbing at 280 nm and antibody integrity seemed to be intact, which was not the case for the method with a specific reduction step (Figure 1). Efficiency of [99mTc(CO)3]rituximab radiolabelling without reduction Before radiolabelling of the antibody, we tested the purity of tricarbonyl compound with thin layer chromatography. Its Radiochemical Purity (PRC) was greater than 95% (data not shown). We did not use HPLC to identify tricarbonyl compound because it was not possible to detect it in the quality control conditions used. In Figure 2, retention time of radiolabelled rituximab (red peak) was 6 min, and a little peak appeared after the antibody. To identify it, we injected free pertechnetate, and its retention time (6.4 min) corresponded to this peak (black peak). After radiolabelling, no purification step was needed because [99mTc(CO)3]rituximab always had a radiochemical purity > 95%. Stability of [99mTc(CO)3]rituximab Using radiochromatography, radiolabelled antibody remained at the origin, whereas pertechnetate moved with the front and tricarbonyl compound migrated to a Rf = 0.6. Thus, after incubation in fresh human plasma at 37 °C, the stability of [99mTc(CO)3]rituximab was about 88% at 4 h and 86% at 20 h (green peak). The amount of free pertechnetate at 20 h was about 14% (red peak; Figure 3) without tricarbonyl compound.
Immunoreactivity of [99mTc(CO)3]rituximab The immunoreactivity of [99mTc(CO)3]rituximab was represented according to the Lineweaver–Burk method.14 The linear regression line shows an immunoreactive fraction of 96% (Figure 4). Binding of [99mTc(CO)3]rituximab The binding assay of [99mTc(CO)3] showed a Kd = 4.57 nM (Figure 5). Biodistribution of [99mTc(CO)3]rituximab After intravenous injection of [99mTc(CO)3]rituximab in mice xenografted subcutaneously with Daudi or Raji lymphoma, antibody accumulated slowly in the tumor (Figure 6). The tumor was detected 5 h after injection, but visualization and tumor/ blood ratio were significant only 24 h after the injection. Biodistribution studies 24 h after i.v. injection revealed uptake of the liver, kidneys, intestines, lungs, heart, tumor, spleen, stomach, and muscle (Table 2). The tumor/muscle ratio was 2.75 (p = 0.005).
Discussion We have described a new rituximab radiolabelling method in three steps: (i) first, substitution of the native buffer to obtain a solution of rituximab in NaCl 0.9%; (ii) preparation of a tricarbonyl compound with a small amount of IsolinkW; and (iii) incubation of 99mTc(CO)3(H2O)3 with rituximab. In a 1 h period,
Figure 1. Size exclusion UV HPLC chromatograms of rituximab before radiolabelling. (A) Chromatogram of MabtheraW in the native buffer. Peak 1 represents rituximab, peak 2 is excipients able to absorb at 280 nm. (B) Chromatogram of rituximab in NaCl 0.9% solution. There is only the peak 1 corresponding to rituximab, excipients which absorb have been removed. C and D chromatograms of rituximab in NaCl 0.9% solution after reduction by 2-ME. Peak 1 is rituximab, peak 2 represents antibody fragments created during the reduction, and peak 3 is excipients able to absorb at 280 nm.
277
J. Label Compd. Radiopharm 2015, 58 274–280
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org
C. Fontan et al.
99m
Figure 2. Radiochemical purity of [ Tc(CO)3]rituximab. Radio (top) and UV (bottom) chromatograms obtained by HPLC with a size-exclusion column. Retention time (Rt) 99m of [ Tc(CO)3]rituximab (red) is Rt = 6 minutes. Rt of pertechnetate (black) is Rt = 6 minutes and 40 seconds. Radiochemical purity of the radiolabeled antibody is 95%.
99m
Figure 5. [ Tc(CO)3]rituximab binding assay on Daudi lymphoma cells. Saturation curve obtained with four increasing concentrations shows a Kd = 4.57 nM. 99m
Figure 3. [ Tc(CO)3]rituximab stability in fresh human plasma at 20 h. The TLC was performed on an AlugramW strip at 20 h. It shows a stability of about 86% for radiolabelled antibody (green peak). The impurity is free pertechnetate and represents 14% (red peak).
99m
Figure 4. [ Tc(CO)3]rituximab immunoreactivity. Immunoreactive fraction was determined by linear extrapolation to infinite cell concentration. The reciprocal of intercept represents the immunoreactive fraction = 96%.
278
a solution of [99mTc(CO)3]rituximab was prepared, without any required purification step. Compared with other published methods containing two reduction steps,3,6–10 we were able to
www.jlcr.org
label rituximab quickly and easily in only three steps. We demonstrated that addition of 2-ME with molar ratio of 2-ME: rituximab between 10.000:16 and 2.000:17 created small antibody fragments, showing a partial destruction of the protein and probably a loss of its integrity. Although the addition of 99m Tc(CO)3 solution may induce antibody reduction because of the presence of boranocarbonate, the amount of this compound is smaller than in classical radiolabelling methods. Also, previous studies15,16 have shown a short half-life (about 10 min) of boranocarbonate in these conditions. Moreover, after the formation of the tricarbonyl compound, the only boroncontaining product was borate [BH3] , another reducing agent, which is unable to reduce disulfide bonds. An additional reduction step could be deleterious for the integrity of the protein because of the presence of disulfide bonds in the antibody structure. Rituximab is an Ig G1 where the disulfide bonds stabilize the structure,17 so the use of 2-ME could reduce the disulfide bonds in thiol groups and disorganize the structure and recognition of CD 20 positive cells by the antibody. Many publications use this additional step of reduction either to create a free thiol SH group from disulfide
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 274–280
C. Fontan et al.
99m
Figure 6. Lymphoma visualization in SCID Beige xenografted mice after injection of [ Tc(CO)3]rituximab. (A) Image obtained 5 h after IV injection of 7.4 MBq 99m 99m removed by [ Tc(CO)3]rituximab. Tumor is already detected (red arrow). (B) Image obtained 24 h after IV injection of 7.4 MBq [ Tc(CO)3]rituximab. Tumor is detected (red arrow) with a tumor/muscle ratio = 3.
bond6–10 or to create more accessible histidine and cysteine for complexation. Rituximab contains 26 histidines, which is the best amino acid for binding [99mTc(CO)3(H2O)3] because of its N-heterocyclic function that constitutes a bidentate or tridentate ligand.11,13 In the antibody, Fab fragments contain six histidines which cannot complex with tricarbonyl compound because of a restricted accessibility whereas in Fragment crystallizable (Fc) fragments, it was reported that several
histidine-rich regions are easily accessible.18 Even though a specific antibody reduction could increase the number of thiol groups or the number of accessible histidine regions, there is a high risk of altering the protein structure and decreasing antigen recognition. Our method showed that histidine located on Fc fragments is easily accessible and sufficient for stable tricarbonyl complexation (Figure 7).
Table 2. [99mTc(CO3)]rituximab biodistribution in SCID mice xenografted with human Raji lymphoma 24 h injection Organs
% ID/g of organs ± SD
J. Label Compd. Radiopharm 2015, 58 274–280
13.465 ± 1.491 7.722 ± 0.095 7.211 ± 5.401 6.370 ± 1.829 5.617 ± 2.276 2.549 ± 0.095 2.519 ± 0.266 2.134 ± 1.213 1.670 ± 0.563 0.616 ± 0.077
99m
Figure 7. Schematic representation of rituximab radiolabelling by Tc(CO)3. The complexation of the tricarbonyl compound could occur thanks to histidine present on rituximab Fc fragment.
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org
279
Blood Kidneys Spleen Liver Lungs Heart Intestine Stomach Tumor Muscle + bone
C. Fontan et al. Although it is not possible to determine exactly where the Tc(CO)3 binds to the chimeric antibody, the immunoreactivity of 96% and Kd = 4.57 nM close to that of rituximab (Kd = 5 nM)18 are better than those obtained with methods using specific antibody reduction step (Kd = 5–6 nM8or 8 and 3 nM7) and showed that the integrity of the protein is preserved. In vivo study showed a distribution in the liver, kidneys, intestines, lungs, heart, tumor, spleen, and stomach. The tumor was detectable 5 h post injection, but the best ratio was observed 24 h after injection because of the blood clearance of the radiolabelled antibody. Finally, the biodistribution of [99mTc(CO)3]rituximab can be influenced by the nature of the ligands: bidentate ligands increase liver and kidney incorporation of 99mTc(CO)3 complex whereas tridentate ligands do not.19,20 It can be assumed that the method used here to radiolabel rituximab with IsolinkW engages a bidentate ligand, such as histidine, not located in the N-terminal portion of the protein. Our work shows that the complexation reaction is faster (30 min vs. 3 h), and the complex created is stable. There is no need to assess purification because the radiolabelling yield is always greater than 95%. So the time of the synthesis is about 1 h versus 3.5 h for the classical method. 99m
Conclusion In conclusion, a simple, fast, and reproducible method has been developed to radiolabel rituximab using 99mTc(CO)3(H2O)3 without any additional reduction step. This method needs only three steps to radiolabel the antibody: (i) substitution of native buffer; (ii) generation of [99mTc(CO)3(H2O)3] under heating conditions (100 °C for 30 min); and (iii) antibody radiolabelling (30 min at 37 °C), with an excellent yield (>95%). This method respects Good Manufacturing Practices (GMP) standards because there are few steps, and the synthesis and quality controls of the radiolabelled protein are rapid. This procedure might be clinically developed, and a sensitive imaging technique like SPECT can be used for diagnosis, evaluation of extension, response to treatment, follow-up, and evaluation of residual disease in patients with NHLs, with high specificity.
Acknowledgements We thank the Non Invasive Exploration Department US006/CREFRE INSERM/UPS and especially Dr Carine Pestourie, who performed nanoSPECT/CT imaging experiments and participated in data analysis. MabtheraW was a gift from the Clinical Pharmacy and Oncologic Unit of Toulouse University Hospital, Purpan.
Conflict of Interest The authors did not report any conflict of interest.
References [1] Lymphoma Website. http://www.lymphoma.org [2013]. [2] MR Smith, Oncogene 2003, 22, 7359–7368. [3] T Olafsen, D Betting, VE Kenanova, FB Salazar, P Clarke, J Said, J. Nucl. Med. 2009, 50, 1500–1508. [4] M Baker, Nature 2010, 463(7283), 977–980. [5] J Condeelis, R Weissleder, Cold Spring Harb. Perspect. Biol. 2010, 2(12), a003848. [6] SJ Mather, D Ellison, J. Nucl. Med. 1990, 31(5), 692–697. [7] a)G Malviya, KL Anzola, E Podesta, B Lagana, C Del Mastro, RA Dierckx, F Scopinaro, A Signore, Mol. Imaging Biol. 2011; b)CR Dias, S Jeger, JA Osso Jr., C Muller, C De Pasquale, A Hohn, R Waibel, R Schibli, Nucl. Med. Biol. 2011, 38, 19–28. [8] CR Dias, S Jeger, JA Osso Jr., C Muller, C De Pasquale, A Hohn, et al. Nucl. Med. Biol. 2011, 38, 19–28. [9] T Gmeiner Stopar, J Fettich, S Zver, I Mlinaric-Rascan, S Hojker, A Socan, Nucl. Med. Commun. 2008, 29, 1059–1065. [10] TG Stopar, I Mlinaric-Rascan, J Fettich, S Hojker, SJ Mather, Eur. J. Nucl. Med. Mol. Imaging 2006, 33, 53–59. [11] A Egli, R Alberto, L Tannahill, R Schibli, U Abram, A Schaffland, J. Nucl. Med. 1999, 40, 1913–1917. [12] S Liu, Chem. Soc. Rev. 2004, 33, 445–461. [13] R Waibel, R Alberto, J Willuda, R Finnern, R Schibli, A Stichelberger, Nat. Biotechnol. 1999, 17, 897–901. [14] T. Lindmo, et al. J. Immunol. Methods 1984, 72, 77–89. [15] R Alberto, K Ortner, N Wheatley, R Schibli, AP Schubiger, J. Am. Chem. Soc. 2001, 123, 3135–3136. [16] TS Pitchumony, B Spingler, R Motterlini, R Alberto, Org. Biomol. Chem. 2010, 8, 4849–4854. [17] D. Baty, P Chames, Immuno. Anal. Biol. Spe. 2006, 21, 255–263. [18] DG Maloney, AJ Grillo-Lopez, CA White, D Bodkin, RJ Schilder, JA Neidhart, Blood 1997, 90, 2188–2195. [19] YS Kim, F Wang, S Liu, World J. Hepatol. 2010, 2, 21–31. [20] R Schibli, R La Bella, R Alberto, E Garcia-Garayoa, K Ortner, U Abram, Bioconjug. Chem. 2000, 11, 345–351.
280 www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 274–280
Revised 14 February 2015,
Accepted 19 February 2015
Published online 27 May 2015 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3283
Radiolabelling rituximab with 99mTc in three steps procedure Charlotte Fontan,a,b* Christine Bezombes,b Anne Sophie Salabert,c Julien Costes,b Raphael Lopez,c Jean-Jacques Fournie,b Hervé Avet-Loiseau,b Yvon Coulais,d Pierre Payoux,c and Mathieu Tafanic Lymphomas are the most frequent haematological malignancy. In non-Hodgkin’s lymphomas (NHL), more than 90% of tumor cells express the cluster of differentiation (CD) 20 antigen. At the end of frontline therapy, the evaluation of remission is based on computed tomography (CT) and positron emission tomography coupled with computer tomography (PET/CT) with [18F]-fluorodeoxyglucose ([18F]FDG). Unfortunately, these techniques are not specific and cannot distinguish residual active tumor from inflammation. The aim of this study was to develop a specific radiotracer of NHL CD 20+ cells for clinical applications. The radiolabelling technique presented, based on the use of tricarbonyl compound, does not include an antibody reduction because this step could damage the protein. Actually, rituximab, an anti-CD 20 chimeric antibody used for the treatment of these NHL, was radiolabelled with IsolinkW 99mTc-tricarbonyl compound in a three-step procedure without using a specific antibody reducer. Radiolabelling yield was greater than 97%. In vitro experiments showed a conservation of antibody integrity. In vivo experiments using single-photon emission computed tomography/CT showed significant tumor targeting 24 h after injection of the radiotracer. It was consequently possible to develop an immunoradiolabelling method to specifically detect the residual disease. As this procedure is fast, reproducible and gentle, it will be possible to comply with Good Manufacturing Practices. Keywords: tricarbonyl compound; immuno imaging; SPECT/CT; rituximab; lymphoma
Introduction
274
In the US, about 500,000 people have lymphoma, and approximately 332,000 have non-Hodgkin’s lymphomas (NHLs). Every year, over 65,000 cases of NHL are diagnosed in America, and 85% are B-cell NHL.1 Treatment of these diseases has improved with the development of the monoclonal antibody rituximab, the first chimeric mouse/human immunoglobulin G1 (IgG1) directed against transmembrane antigen cluster of differentiation (CD) 20, which is expressed in more than 90% of B-cell NHLs.2 Diagnosis, evaluation of extension, response to treatment, or follow-up of NHLs are performed by modern imaging techniques such as positron emission tomography coupled with computed tomography (PET/CT). Hybrid [18 F]-fluorodeoxyglucose PET/CT is a method based on the fusion of an anatomical image from CT and a metabolic image produced by PET. Even though this method provides high sensitivity, it is not specific to NHLs and cannot be used for indolent lymphomas with slow growth and weak consumption of glucose. Moreover, it cannot distinguish between residual tumor and inflammation. Consequently, the development of a specific and sensitive compound for NHLs would to be of interest. Because of the widespread expression of CD 20 on NHL tumor cells, radiolabelling rituximab with a positron emitter3 or a single photon emitting isotope, such as Technetium-99 m (99mTc), constitutes an approach specific to the disease. Single photon emission computed tomography (SPECT) is a method of choice because of its sensitivity (10 14 moles of radionuclides detected), the depth of study without
J. Label Compd. Radiopharm 2015, 58 274–280
any limitation, and its resolution about 1 mm.4,5 Moreover, the availability of radionuclides like 99mTc (generator 99Mo/99mTc) and its convenient dosimetry (half-life = 6 h, gamma energy = 140 keV) make it an isotope of choice. During recent years, several methods based on radiolabelling rituximab with 99m Tc have been described.6–10 All of them include two reduction steps, one dedicated to the reduction of Technetium-99 m, the other specifically targeting the antibody, with either photoactivation9,10 or 2-mercaptoethanol (2-ME).6–8 This specific antibody reduction step creates thiol residues that facilitate technetium binding. However, this specific reduction could modify the protein structure and decrease antigen recognition. Since 2003, the Mallinckrodt Company has marketed IsolinkW containing tartrate, carbonate, and boranocarbonate, which decrease the oxidation number of native pertechnetate (99mTcO4 ; +VII) to +I. The moiety obtained is called the ‘technetium core’ [99mTc(CO)3(H2O)3] and is resistant to oxidation a
Radiopharmacy, University Hospital, Toulouse, France
b
The Cancer Research Center of Toulouse, Toulouse, France
c
UMR 825, INSERM, Toulouse, France
d
Nuclear medicine, University Hospital, Toulouse, France
*Correspondence to: Charlotte Fontan, Radiopharmacy, University Hospital, Toulouse, France. E-mail: [email protected]
Copyright © 2015 John Wiley & Sons, Ltd.
C. Fontan et al. by oxygen.11 The three molecules of water can be replaced by ligands containing N-heterocycle, thioether, thiolate functions,12 which are present in amino acids like histidine, cysteine, and methionine.11–13 Theoretically, radiolabelling of rituximab protein with the technetium core is possible without any additional antibody reduction step. We hypothesized that [99mTc(CO)3]rituximab would be stable and the properties of the high molecular weight (above 143.9 kDa) protein would not be altered. The purpose of this study was to radiolabel rituximab with Technetium-99 m by a method using a tricarbonyl compound without any additional reduction step, enabling to maintain the antibody’s structure and function. We evaluated the effectiveness of the radiolabelling by testing the radioligand in vitro and in vivo in mice xenografted with human Burkitt’s lymphoma.
Experimental Reagents and equipment
Step 1: Native buffer substitution In commercial form, MabtheraW contains several excipients, such as sodium citrate, polysorbate 80, sodium chloride, water for injection, sodium hydroxide, and hydrochloric acid. This native buffer had to be replaced because polysorbate 80 and sodium citrate can bind pertechnetate. Elution of rituximab (10 mg/mL, 1 mL) from the PD10 column is performed with nitrogen-purged saline solution (0.9%). The final concentration of rituximab was about 5 to 8 mg/mL.
Step 2: Preparation of
99m
Tc(CO)3(H2O)3
Technetium-99 m must be in the +I oxidation state to complex with rituximab. Sodium pertechnetate was reduced by the tricarbonyl compound IsolinkW. In comparison with classical radiolabelling methods, which used one kit of reducer for each radiolabelling,6,7 our method used a small amount of it. Indeed, IsolinkW powder (4 mg) was mixed with 99m TCO4 (200 μL, 1221 MBq). The mixture was heated (100 °C, 30 min). The pH of the solution was adjusted to 7 with acetic acid (0.2 M). Radiochemical purity was >95% (data not shown)
Step 3: Rituximab radiolabelling with
Reagents
99m
Tc(CO)3(H2O)3
99m
For radiotracer synthesis, MabtheraW was purchased from Roche, Inc. 99m (Switzerland, Bale). Pertechnetate [ TcO4 ] was eluted in saline 99 99m solution from a Mo/ Tc generator (IBA, Saclay, France), and IsolinkW was provided by Mallinckrodt Inc. (Petten, Netherlands). For cell culture and in vitro testing, we used RPMI 1640 (Gibco, Thermoscientific, France) supplemented with 10% fetal calf serum (Gibco, Thermoscientific, France), glutamine (2 mM, Gibco, Thermoscientific, France), streptomycin (10 μg/mL, Gibco, Thermoscientific, France), and penicillin (200 U/mL, Invitrogen, Life Technologies, France). Immunoreactivity was performed with phosphate-buffered saline (PBS, pH 7.4 Sigma Aldrich, Germany) and Bovine Serum Albumin (BSA, Rockland-inc). To reduce oxidation state of pertechnetate, IsolinkW was used. This kit contains 8.5 mg sodium tartrate, 2.85 mg Na2B4O7 10H2O, 7.15 mg of sodium carbonate, and 4.5 mg of sodium boranocarbonate in lyophilized form totalling 23 mg of powder. To establish a comparison with conventional methods, specific antibody reduction was performed using 2-ME (Sigma Aldrich, Germany).
Equipment
Tc(CO)3(H2O)3 (20 mg/mL, 50 μL) was added to rituximab (6,5 mg/mL, 100 μL), and the mixture was incubated (37 °C, 30 min). 99m The final concentration of Tc(CO)3rituximab was about 4 mg/mL, and the activity was about 250 MBq.
Specific rituximab reduction with 2-mercaptoethanol Rituximab (10 mg/mL,1 mL) was reduced with 2-ME in a molar ratio of 2-ME:rituximab = 10.000:16 or 2.000:17 for 30 min at room temperature, and the mixture was purified using Solid Phase Extraction with a mobile phase containing nitrogen-purged saline solution (0.9%). The final concentration was less than 1 mg/mL.
Quality control The quality checks were performed by HPLC with a size exclusion column. The eluent was NaCl 0.9% and the flow rate 2 mL/min. A UV detector operating at 280 nm was used to detect proteins containing amino acids with aromatic radicals. Another detector monitored 140 keV technetium-99 m gamma rays. After synthesis, in vivo stability towards metabolic enzymes was tested 99m by adding aliquots of [ Tc(CO)3]rituximab (4 mg/mL, 50 μL) to 100 μL fresh human plasma. They were placed in an incubator at 37 °C. The stability was evaluated with a radiochromatography after incubation times of 4 and 20 h.
Culture of cancer cells Daudi or Raji Burkitt’s lymphoma cells were cultured at 37 °C in 5% CO2 humidified atmosphere in a complete medium. Every 4 days, cells viability was measured with a cell counter.
In vitro experiments Immunoreactivity of [99mTc(CO)3]rituximab 99m
The immunoreactivity of [ Tc(CO)3]rituximab was measured by the Lindmo method.14 Four concentrations of Daudi lymphoma cells were 6 6 used (from 0.6 × 10 to 10 × 10 cells/mL). Before starting, the cells were washed twice with 1% BSA/PBS buffer. A constant concentration of 99m [ Tc(CO)3]rituximab (6 ng/mL) was added. Total volume was 1 mL in 1% BSA/PBS buffer. To test non-specific binding, a 1000-fold excess of unlabelled rituximab was added. Tubes were stirred for 2 h at 4 °C. After incubation, the tubes were centrifuged (1600 xg, 5 min). The cells were washed twice in 1% BSA/PBS, and the supernatant was removed. The radioactivity of the pellets was measured with a gamma counter. The
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org
275
During culture of lymphoma cell lines, viability was measured with a cell counter (AdamW, Alphametrix Rodermark). Native buffer was substituted thanks to a Solid Phase Extraction column (Sephadex PD 10, GE-Healthcare). To assess quality control, High Performance Liquid Chromatography (HPLC; ion chromatography system (ICS)) was used. It is a system equipped with a size-exclusion column (Cluzeau Info Labo (CIL) Stability Cluzeau 200 SEC-5 μm 300 × 8 mm) and two channels, one for radioactivity detection (NaI Cristal, GabiW Raytest, Germany) and the second for ultra violet (UV) detection at 280 nm (lambda 1010, ICS). Run acquisition in UV and gamma appeared on ICS Software. Time for runs was 10 or 15 min. 99m Tc(CO)3(H2O)3 were Stability in plasma and radiochemical purity of evaluated by Thin Layer Chromatography with MinigitaW, (Raytest, Germany) radioachromatograph using an AlugramW Thin Layer Chromatography (Servilab, France) strip. The mobile phase was a methanol/water mixture (90/10, v/v). The energy of the gamma radiation of technetium-99 m (140 keV) contained in the samples was measured with a gamma counter (Automatic Gamma counter 1470 Wizard 3 W, PerkinElmer, Massachusetts, USA). The interpretation of binding assay was performed by GraphPad Prism Version 6 Software (GraphPad Software, Inc, La Jolla, USA). In vivo tests were achieved using a nanoSPECT/CTW camera, Biospace, France. It consisted of four collimators, each containing nine pinholes. Imaging resolution was about 0.9 mm for SPECT and about 100 μm for CT. Image interpretation used the In VivoscopeW Software.
J. Label Compd. Radiopharm 2015, 58 274–280
Radiolabelling of the antibody
C. Fontan et al. experiment was carried out in triplicate. For each cell concentration, specific binding counts were obtained by subtracting non-specific binding from total binding counts. A graph representing total applied activity/specific binding in ordinate and reciprocal of concentration cells in abscissa was drawn. The 99m immunoreactivity of [ Tc(CO)3]rituximab was obtained by the reciprocal of the intercept on the ordinate of the linear regression line.
Binding of [99mTc(CO)3]rituximab 99m
The binding capacity of [ Tc(CO)3]rituximab was studied with a 6 constant concentration of Daudi cells (1 × 10 cells/mL of RPMI 1640 99m culture medium). Four concentrations of [ Tc(CO)3]rituximab were used (1.5 ng/mL to 4 ng/mL). Before starting, the cells were washed twice with RPMI buffer. For each concentration, two conditions were used to determine total and non-specific binding. In the latter case, a 1000-fold excess of unlabelled rituximab was added, and the tubes were incubated and centrifuged as mentioned earlier. The cells were washed twice in RPMI 1640 culture medium and the supernatant was removed. The radioactivity of the pellets was measured with a gamma counter. The 99m experiment was performed in triplicate. For each [ Tc(CO)3]rituximab concentration tested, specific binding counts were obtained by subtracting non-specific binding from total binding counts. Determination of Kd was obtained thanks to GraphPad Prism Version 6 Software using total, non-specific and specific binding.
Acquisitions on SPECT-CT were performed on three mice 5 and 24 h after i.v. injection of radiotracer. 99m Animals were sacrificed 24 h after i.v injection of [ Tc(CO)3] rituximab (130 μg, median activity = 8.14 MBq). For five mice, organs were weighed and the radioactivity in each organ counted with a gamma counter. Results are expressed as a percentage of the injected dose per gram of organ (%ID/g). To measure the real activity injected, we weighed the syringe before and after injection. Injected dose (expressed in counts per minutes (cpm)) was calculated by comparison 99m with a weighed sample of 10 μL of [ Tc(CO)3]rituximab placed in the gamma counter. Then, we compared it with the number of cpm in each organ studied.
Image acquisition and interpretation Protocol acquisition. For the CT, maximum voltage was 45 kV, and signal accumulation time per position of the generator was 500 ms. Number of projections was 180. For SPECT, whole body planar images registered 10,000 counts per second (cps) for each projection.
Interpretation of images In VivoscopeW Software CT reconstruction was fast (smoothing: 35%, resolution 67%, Iteration: 2*3). SPECT reconstruction used nih-fire2 as color scale.
Statistical analysis
In vivo experiment Experimental animals Mice were housed in a pathogen-free facility and handled in accordance with the procedures outlined in Council Directive 86/609/European Economic Community. The investigation respected the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH publication No. 85–23, revised 1996). Male severe combined immunodeficiency (SCID)-beige mice aged between 12 and 16 weeks were used. They were injected subcutaneously on the flank 6 6 with 6 × 10 Raji cells or 10 × 10 Daudi cells. The experiments were performed when tumors were detectable. For all experiments, the mice were anesthetized with isoflurane.
Results are expressed as mean ± standard deviation. Statistical comparison of the data was performed using the t-test for comparison between two groups. A value of p < 0.05 was considered as significant.
Results and discussion Results Comparison of rituximab radiolabelling methods with and without antibody reduction As shown in Table 1, our radiolabelling method was faster and used smaller amounts of reagents than the method described
Table 1. Comparison of rituximab radiolabelling methods with and without antibody reduction Method WITH antibody reduction step
Method WITHOUT antibody reduction step
STEP 1: Specific antibody reduction MabtheraW incubated with molar excess of 2-ME : 10.000:1 or 2;000:1 30 min at room temperature
STEP 1: Substitution of MabtheraW native buffer By Solid Phase Extraction in nitrogen-purged saline (0.9%)
STEP 2: Purification and substitution of MabtheraW native buffer By Solid Phase Extraction in nitrogen-purged saline (0.9%)
STEP 2: Preparation of 99mTc(CO)3(H2O)3
STEP 3: Preparation of 99mTc(CO)3(H2O)3 One IsolinkW kit vial (23 mg) with 99mTcCO4 (1 mL) 40 s at 150 °C pH adjusted to 7
A part of IsolinkW kit vial (4 mg) + 99mTcCO4 (200 μL,1221 MBq) 30 min at 100 °C pH adjusted to 7 STEP 3: Rituximab radiolabelling Rituximab (0,65 mg, 150 μL) + 99mTc(CO)3(H2O)3 (50 μL) 30 min at 37 °C
STEP 4: Rituximab radiolabelling 3 h at 37 °C
276
NUMBER OF STEPS: 4 TIME OF RADIOLABELLING: 3 h 30 min
www.jlcr.org
NUMBER OF STEPS: 3 TIME OF RADIOLABELLING: 1 h
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 274–280
C. Fontan et al. previously.6,7 After substitution of native buffer, there were no excipients absorbing at 280 nm and antibody integrity seemed to be intact, which was not the case for the method with a specific reduction step (Figure 1). Efficiency of [99mTc(CO)3]rituximab radiolabelling without reduction Before radiolabelling of the antibody, we tested the purity of tricarbonyl compound with thin layer chromatography. Its Radiochemical Purity (PRC) was greater than 95% (data not shown). We did not use HPLC to identify tricarbonyl compound because it was not possible to detect it in the quality control conditions used. In Figure 2, retention time of radiolabelled rituximab (red peak) was 6 min, and a little peak appeared after the antibody. To identify it, we injected free pertechnetate, and its retention time (6.4 min) corresponded to this peak (black peak). After radiolabelling, no purification step was needed because [99mTc(CO)3]rituximab always had a radiochemical purity > 95%. Stability of [99mTc(CO)3]rituximab Using radiochromatography, radiolabelled antibody remained at the origin, whereas pertechnetate moved with the front and tricarbonyl compound migrated to a Rf = 0.6. Thus, after incubation in fresh human plasma at 37 °C, the stability of [99mTc(CO)3]rituximab was about 88% at 4 h and 86% at 20 h (green peak). The amount of free pertechnetate at 20 h was about 14% (red peak; Figure 3) without tricarbonyl compound.
Immunoreactivity of [99mTc(CO)3]rituximab The immunoreactivity of [99mTc(CO)3]rituximab was represented according to the Lineweaver–Burk method.14 The linear regression line shows an immunoreactive fraction of 96% (Figure 4). Binding of [99mTc(CO)3]rituximab The binding assay of [99mTc(CO)3] showed a Kd = 4.57 nM (Figure 5). Biodistribution of [99mTc(CO)3]rituximab After intravenous injection of [99mTc(CO)3]rituximab in mice xenografted subcutaneously with Daudi or Raji lymphoma, antibody accumulated slowly in the tumor (Figure 6). The tumor was detected 5 h after injection, but visualization and tumor/ blood ratio were significant only 24 h after the injection. Biodistribution studies 24 h after i.v. injection revealed uptake of the liver, kidneys, intestines, lungs, heart, tumor, spleen, stomach, and muscle (Table 2). The tumor/muscle ratio was 2.75 (p = 0.005).
Discussion We have described a new rituximab radiolabelling method in three steps: (i) first, substitution of the native buffer to obtain a solution of rituximab in NaCl 0.9%; (ii) preparation of a tricarbonyl compound with a small amount of IsolinkW; and (iii) incubation of 99mTc(CO)3(H2O)3 with rituximab. In a 1 h period,
Figure 1. Size exclusion UV HPLC chromatograms of rituximab before radiolabelling. (A) Chromatogram of MabtheraW in the native buffer. Peak 1 represents rituximab, peak 2 is excipients able to absorb at 280 nm. (B) Chromatogram of rituximab in NaCl 0.9% solution. There is only the peak 1 corresponding to rituximab, excipients which absorb have been removed. C and D chromatograms of rituximab in NaCl 0.9% solution after reduction by 2-ME. Peak 1 is rituximab, peak 2 represents antibody fragments created during the reduction, and peak 3 is excipients able to absorb at 280 nm.
277
J. Label Compd. Radiopharm 2015, 58 274–280
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org
C. Fontan et al.
99m
Figure 2. Radiochemical purity of [ Tc(CO)3]rituximab. Radio (top) and UV (bottom) chromatograms obtained by HPLC with a size-exclusion column. Retention time (Rt) 99m of [ Tc(CO)3]rituximab (red) is Rt = 6 minutes. Rt of pertechnetate (black) is Rt = 6 minutes and 40 seconds. Radiochemical purity of the radiolabeled antibody is 95%.
99m
Figure 5. [ Tc(CO)3]rituximab binding assay on Daudi lymphoma cells. Saturation curve obtained with four increasing concentrations shows a Kd = 4.57 nM. 99m
Figure 3. [ Tc(CO)3]rituximab stability in fresh human plasma at 20 h. The TLC was performed on an AlugramW strip at 20 h. It shows a stability of about 86% for radiolabelled antibody (green peak). The impurity is free pertechnetate and represents 14% (red peak).
99m
Figure 4. [ Tc(CO)3]rituximab immunoreactivity. Immunoreactive fraction was determined by linear extrapolation to infinite cell concentration. The reciprocal of intercept represents the immunoreactive fraction = 96%.
278
a solution of [99mTc(CO)3]rituximab was prepared, without any required purification step. Compared with other published methods containing two reduction steps,3,6–10 we were able to
www.jlcr.org
label rituximab quickly and easily in only three steps. We demonstrated that addition of 2-ME with molar ratio of 2-ME: rituximab between 10.000:16 and 2.000:17 created small antibody fragments, showing a partial destruction of the protein and probably a loss of its integrity. Although the addition of 99m Tc(CO)3 solution may induce antibody reduction because of the presence of boranocarbonate, the amount of this compound is smaller than in classical radiolabelling methods. Also, previous studies15,16 have shown a short half-life (about 10 min) of boranocarbonate in these conditions. Moreover, after the formation of the tricarbonyl compound, the only boroncontaining product was borate [BH3] , another reducing agent, which is unable to reduce disulfide bonds. An additional reduction step could be deleterious for the integrity of the protein because of the presence of disulfide bonds in the antibody structure. Rituximab is an Ig G1 where the disulfide bonds stabilize the structure,17 so the use of 2-ME could reduce the disulfide bonds in thiol groups and disorganize the structure and recognition of CD 20 positive cells by the antibody. Many publications use this additional step of reduction either to create a free thiol SH group from disulfide
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 274–280
C. Fontan et al.
99m
Figure 6. Lymphoma visualization in SCID Beige xenografted mice after injection of [ Tc(CO)3]rituximab. (A) Image obtained 5 h after IV injection of 7.4 MBq 99m 99m removed by [ Tc(CO)3]rituximab. Tumor is already detected (red arrow). (B) Image obtained 24 h after IV injection of 7.4 MBq [ Tc(CO)3]rituximab. Tumor is detected (red arrow) with a tumor/muscle ratio = 3.
bond6–10 or to create more accessible histidine and cysteine for complexation. Rituximab contains 26 histidines, which is the best amino acid for binding [99mTc(CO)3(H2O)3] because of its N-heterocyclic function that constitutes a bidentate or tridentate ligand.11,13 In the antibody, Fab fragments contain six histidines which cannot complex with tricarbonyl compound because of a restricted accessibility whereas in Fragment crystallizable (Fc) fragments, it was reported that several
histidine-rich regions are easily accessible.18 Even though a specific antibody reduction could increase the number of thiol groups or the number of accessible histidine regions, there is a high risk of altering the protein structure and decreasing antigen recognition. Our method showed that histidine located on Fc fragments is easily accessible and sufficient for stable tricarbonyl complexation (Figure 7).
Table 2. [99mTc(CO3)]rituximab biodistribution in SCID mice xenografted with human Raji lymphoma 24 h injection Organs
% ID/g of organs ± SD
J. Label Compd. Radiopharm 2015, 58 274–280
13.465 ± 1.491 7.722 ± 0.095 7.211 ± 5.401 6.370 ± 1.829 5.617 ± 2.276 2.549 ± 0.095 2.519 ± 0.266 2.134 ± 1.213 1.670 ± 0.563 0.616 ± 0.077
99m
Figure 7. Schematic representation of rituximab radiolabelling by Tc(CO)3. The complexation of the tricarbonyl compound could occur thanks to histidine present on rituximab Fc fragment.
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org
279
Blood Kidneys Spleen Liver Lungs Heart Intestine Stomach Tumor Muscle + bone
C. Fontan et al. Although it is not possible to determine exactly where the Tc(CO)3 binds to the chimeric antibody, the immunoreactivity of 96% and Kd = 4.57 nM close to that of rituximab (Kd = 5 nM)18 are better than those obtained with methods using specific antibody reduction step (Kd = 5–6 nM8or 8 and 3 nM7) and showed that the integrity of the protein is preserved. In vivo study showed a distribution in the liver, kidneys, intestines, lungs, heart, tumor, spleen, and stomach. The tumor was detectable 5 h post injection, but the best ratio was observed 24 h after injection because of the blood clearance of the radiolabelled antibody. Finally, the biodistribution of [99mTc(CO)3]rituximab can be influenced by the nature of the ligands: bidentate ligands increase liver and kidney incorporation of 99mTc(CO)3 complex whereas tridentate ligands do not.19,20 It can be assumed that the method used here to radiolabel rituximab with IsolinkW engages a bidentate ligand, such as histidine, not located in the N-terminal portion of the protein. Our work shows that the complexation reaction is faster (30 min vs. 3 h), and the complex created is stable. There is no need to assess purification because the radiolabelling yield is always greater than 95%. So the time of the synthesis is about 1 h versus 3.5 h for the classical method. 99m
Conclusion In conclusion, a simple, fast, and reproducible method has been developed to radiolabel rituximab using 99mTc(CO)3(H2O)3 without any additional reduction step. This method needs only three steps to radiolabel the antibody: (i) substitution of native buffer; (ii) generation of [99mTc(CO)3(H2O)3] under heating conditions (100 °C for 30 min); and (iii) antibody radiolabelling (30 min at 37 °C), with an excellent yield (>95%). This method respects Good Manufacturing Practices (GMP) standards because there are few steps, and the synthesis and quality controls of the radiolabelled protein are rapid. This procedure might be clinically developed, and a sensitive imaging technique like SPECT can be used for diagnosis, evaluation of extension, response to treatment, follow-up, and evaluation of residual disease in patients with NHLs, with high specificity.
Acknowledgements We thank the Non Invasive Exploration Department US006/CREFRE INSERM/UPS and especially Dr Carine Pestourie, who performed nanoSPECT/CT imaging experiments and participated in data analysis. MabtheraW was a gift from the Clinical Pharmacy and Oncologic Unit of Toulouse University Hospital, Purpan.
Conflict of Interest The authors did not report any conflict of interest.
References [1] Lymphoma Website. http://www.lymphoma.org [2013]. [2] MR Smith, Oncogene 2003, 22, 7359–7368. [3] T Olafsen, D Betting, VE Kenanova, FB Salazar, P Clarke, J Said, J. Nucl. Med. 2009, 50, 1500–1508. [4] M Baker, Nature 2010, 463(7283), 977–980. [5] J Condeelis, R Weissleder, Cold Spring Harb. Perspect. Biol. 2010, 2(12), a003848. [6] SJ Mather, D Ellison, J. Nucl. Med. 1990, 31(5), 692–697. [7] a)G Malviya, KL Anzola, E Podesta, B Lagana, C Del Mastro, RA Dierckx, F Scopinaro, A Signore, Mol. Imaging Biol. 2011; b)CR Dias, S Jeger, JA Osso Jr., C Muller, C De Pasquale, A Hohn, R Waibel, R Schibli, Nucl. Med. Biol. 2011, 38, 19–28. [8] CR Dias, S Jeger, JA Osso Jr., C Muller, C De Pasquale, A Hohn, et al. Nucl. Med. Biol. 2011, 38, 19–28. [9] T Gmeiner Stopar, J Fettich, S Zver, I Mlinaric-Rascan, S Hojker, A Socan, Nucl. Med. Commun. 2008, 29, 1059–1065. [10] TG Stopar, I Mlinaric-Rascan, J Fettich, S Hojker, SJ Mather, Eur. J. Nucl. Med. Mol. Imaging 2006, 33, 53–59. [11] A Egli, R Alberto, L Tannahill, R Schibli, U Abram, A Schaffland, J. Nucl. Med. 1999, 40, 1913–1917. [12] S Liu, Chem. Soc. Rev. 2004, 33, 445–461. [13] R Waibel, R Alberto, J Willuda, R Finnern, R Schibli, A Stichelberger, Nat. Biotechnol. 1999, 17, 897–901. [14] T. Lindmo, et al. J. Immunol. Methods 1984, 72, 77–89. [15] R Alberto, K Ortner, N Wheatley, R Schibli, AP Schubiger, J. Am. Chem. Soc. 2001, 123, 3135–3136. [16] TS Pitchumony, B Spingler, R Motterlini, R Alberto, Org. Biomol. Chem. 2010, 8, 4849–4854. [17] D. Baty, P Chames, Immuno. Anal. Biol. Spe. 2006, 21, 255–263. [18] DG Maloney, AJ Grillo-Lopez, CA White, D Bodkin, RJ Schilder, JA Neidhart, Blood 1997, 90, 2188–2195. [19] YS Kim, F Wang, S Liu, World J. Hepatol. 2010, 2, 21–31. [20] R Schibli, R La Bella, R Alberto, E Garcia-Garayoa, K Ortner, U Abram, Bioconjug. Chem. 2000, 11, 345–351.
280 www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 274–280
