Abstract
A high radiochemical purity (RCP) is recommended for radiopharmaceutical compounds used in the clinical practice of nuclear medicine. However, some preparations of 99mTc-sestamibi contain excess impurities (>6%). To understand the origin of these impurities, we investigated the effect of sodium nitrate on the RCP of sestamibi preparations by testing eluates from 3 commercially available 99mTc generators. Methods: The sestamibi kits (Stamicis) were reconstituted with 99mTc eluate from nitrate-containing wet-column (NCWC), nitrate-free wet-column (NFWC), and nitrate-free dry-column (NFDC) generators. Sodium nitrate was 0.05 mg/mL in eluates from the NCWC generators. The RCP was determined using aluminum oxide sheets as the stationary phase and absolute ethanol as the mobile phase. Succimer, tetrofosmin, oxidronate, exametazine, albumin nanocolloid, and soluble albumin were also tested for their RCP values with eluates from the 3 different 99mTc generators. Results: The RCP assessment of 99mTc-sestamibi was performed on 127 Stamicis preparations. Significantly lower RCP values were found for Stamicis kits prepared with the NCWC generator than for Stamicis prepared with the NFWC (P < 0.0001) and NFDC (P < 0.0001) generators. The number of Stamicis preparations with an RCP under 94% was greater with the NCWC generator (32 of 53 kits) than with the NFDC (2 of 51 kits) or NFWC (0 of 23 kits) generator. Furthermore, the addition of a 0.05 mg/mL concentration of nitrate in NFWC generator eluates significantly decreased the RCP of the Stamicis preparation. In the absence of nitrate in 99mTc eluate, no difference was observed between the RCP values of Stamicis kits prepared with the NFWC and NFDC generators. The 99mTc impurities generated by nitrates did not modify the quality of myocardial imaging (normal heart-to-lung ratio, 2.2), probably because these impurities are not in the heart field of view. No other tested 99mTc-radiopharmaceutical interfered with nitrates. Conclusion: We recommend using nitrate-free generator eluates in 99mTc-sestamibi preparations to improve the product quality and prevent unnecessary exposure of the patient to radiation.
Sestamibi labeled with 99mTc (99mTc-2-methoxyisobutylisonitril, or MIBI) is a lipophilic cationic complex originally developed as a myocardial perfusion imaging agent (1–4) and subsequently used for parathyroid (5) and tumor (6) imaging. Ninety-five percent of intracellular 99mTc-sestamibi is retained in the mitochondria. Its uptake involves a passive distribution across plasma and mitochondrial membranes, but its retention depends mainly on the large negative transmembrane potential of mitochondria (7). 99mTc-sestamibi is also a transport substance recognized by the multidrug resistance–related P glycoprotein, and its tumor cell accumulation is enhanced by inhibition of the efflux transport function. This property provided the basis for clinical studies that investigated the role of 99mTc-sestamibi in predicting the response to chemotherapy in patients with cancer.
The monovalent 99mTc-sestamibi complex consists of 1 atom of 99mTc in a +I oxidation state and 6 molecules of MIBI. According to the product package insert, MIBI is in the form of tetrakis (2-MIBI) copper (I) tetrafluoroborate, and stannous chloride dihydrate is the reducing agent of 99mTc. A high radiochemical purity (RCP) is recommended to prevent unnecessary exposure of the patient to radiation and to obtain good-quality images without hot spots due to the presence of radioactive impurities.
Several factors such as generator ingrowth time, 99mTc eluate age, activity amount, and dry- or wet-column generator have been shown to affect the RCP of 99mTc-sestamibi (8–11). Radiolysis of the water was proposed as a possible cause shared by all these factors (12). Effectively, the H2O radiolysis that results in the formation of strong oxidizing agents (
In this study, we showed that sodium nitrate present in eluates from commercially available generators decreased the RCP of 99mTc-sestamibi. More than half the RCP values fell below the limit recommended by the manufacturer. Furthermore, this effect was MIBI-specific because no sodium nitrate–induced RCP decrease was observed with other 99mTc-complexed radiopharmaceutical compounds.
MATERIALS AND METHODS
Preparation of 99mTc-Sestamibi
Sodium 99mTc-pertechnetate (6–11 GBq in 3 mL), obtained from a nitrate-containing wet-column (NCWC) generator (Elumatic-III; Cisbio International), a nitrate-free wet-column (NFWC) generator (Ultra-TechneKow FM; Covidien plc), or a nitrate-free dry-column (NFDC) generator (Drytec; GE Healthcare), was added to a vial from a Stamicis or Cardiolite kit, as recommended by the manufacturer’s instructions (Cisbio International). 99mTc-sestamibi was always prepared with a second elution from a generator having an ingrowth time lower than 24 h, and the 99mTc-eluate was no more than 2 h old. After 5 upward–downward motions, the preparation was placed in a boiling water bath for 10 min. Then, the vial was removed from the water bath and allowed to cool for 15 min before being diluted with a sodium chloride solution (0.9%, w/v) to 6 mL for Cardiolite or 10 mL for Stamicis. The preparation was then ready for quality control and clinical purposes.
RCP Determination in 99mTc-Sestamibi Preparations
Thin-layer chromatography (TLC) was performed according to the instructions of the manufacturer. Briefly, 20 mL of absolute ethanol were placed into a cylindric chromatography tank (8 × 10 cm) and the atmosphere was allowed to saturate for at least 20 min. A drop of 3 μL of absolute ethanol was applied 1.5 cm from the bottom of Baker-flex (Avantor Performance Materials, Inc.) aluminum oxide sheets (2.5 × 7.5 cm, provided by Cisbio International), followed by a drop of 3 μL of 99mTc-sestamibi placed within the wetted region. The spot was allowed to dry in air, and then the TLC sheet was placed into the chromatography tank. When the solvent front reached 1 cm from the top, the TLC sheet was removed and dried. The distribution of 99mTc compounds on the aluminum oxide sheet was determined using a storage phosphor-imaging device (Cyclone; PerkinElmer), which gives an image of the radioactive products after activation of a phosphor screen for 3 min. Data acquisition was analyzed with Optiquan image analysis software (PerkinElmer). The mean position of a line, indicating the limit to discriminate between the radioimpurities (free 99mTc-pertechnetate and hydrolyzed-reduced 99mTc compounds staying at the origin) and 99mTc-sestamibi (migrating at the solvent front), was drawn at 4 cm from the bottom of the TLC sheet. The RCP of 99mTc-sestamibi was expressed as the ratio in percentage of radioactivity of the solvent front divided by total 99mTc of the TLC sheet. As recommended by the manufacturer, the RCP values should be at least 94% for clinical application. All values below 94% were measured in triplicate to improve the accuracy of the test.
This TLC method detected impurities A (free 99mTc-pertechnetate) and B (hydrolyzed-reduced 99mTc compounds) but not impurity C (99mTc-isonitrile complex with a methyl propene group), as mentioned in the European Pharmacopoeia (19). The other quality specifications (organoleptic characters, pH, or γ-spectrometry) were in conformity with the European Pharmacopoeia.
Addition of Sodium Nitrate to Generator Eluates
According to the manufacturer's package inserts, the sodium nitrate concentration is 0.05 mg/mL in NCWC generator eluates, and the other 2 generators, NFWC and NFDC, are eluted with a nitrate-free isotonic saline solution.
Sodium nitrate (BioUltra; Sigma), ranging from 0.05 to 0.5 mg/mL, was mixed with sodium 99mTc-pertechnetate (5–6 GBq in a 3-mL final volume) obtained from a NFWC 99Mo/99mTc generator. Then, the radioactivity together with nitrate was added to a vial from a Stamicis kit and the preparation was placed in a boiling water bath for 10 min. The vial was removed from the water bath and allowed to cool for 15 min before being diluted to 6 mL with a sodium chloride solution (0.9%, w/v). An aliquot of 0.1 mL was taken to test the RCP of the preparation in triplicate.
RCP Determination in 99mTc-Radiopharmaceutical Preparations
The RCP of the 99mTc-succimer (Renocis; Cisbio International) and 99mTc-human albumin nanocolloid preparations (Nanocoll; GE Healthcare) was determined using a narrow strip of chromatography paper (number 1, 2.5 × 12.5 cm; Whatman International Ltd.) as the stationary phase and acetone as the mobile phase.
The RCP of the 99mTc-tetrofosmin preparations (Myoview; GE Healthcare) was determined using an instant TLC plate impregnated with silica gel (ITLC/SG, 2.5 × 10 cm; Pall Corp.) as the stationary phase and 2-butanone as the mobile phase (20). Another method used a narrow strip of glass microfiber chromatography paper impregnated with silicic acid (2 × 11 cm; Varian SA) as the stationary phase and acetone/dichloromethane (65/35, v/v) as the mobile phase.
The RCP of the 99mTc-oxidronate preparations (TechneScan HDP; Covidien plc) was determined using 2 separated TLCs with ITLC/SG strips (2.5 × 10 cm; Pall Corp. or Varian SA) as the stationary phase and acetone and sodium acetate (1 M) as the mobile phases.
The RCP of the 99mTc-exametazime preparations (Cerestab; GE Healthcare) was determined using 2 separated TLCs with ITLC/SG strips (2 × 15 cm; Pall Corp.) as the stationary phase and 2-butanone and NaCl 0.9% as the mobile phases. Another method used the same mobile phases but chromatography paper (2 × 15 cm; Varian SA) instead of ITLC/SG as the stationary phase.
The RCP of the 99mTc-human soluble albumin preparations (Vasculocis; Cisbio International) was determined using a narrow strip of chromatography paper (number 1, 2.5 × 10 cm; Whatman International Ltd.) as the stationary phase and methanol/water (80/20, v/v) as the mobile phase.
Statistical Analysis
The nonparametric Kruskal–Wallis and Wilcoxon rank-sum tests were used for comparison of the RCP data among groups of different 99mTc-radiopharmaceutical preparations. The RCPs of 99mTc-sestamibi at different levels of nitrate concentrations and the heart-to-liver ratios were compared with the control using the Student t test. All P values were 2-sided, and the type I error was set at 5%. Statistical analyses were performed using SAS software, version 9.2 (SAS Institute).
RESULTS
The RCP assessment of 99mTc-sestamibi was performed on 127 Stamicis and 56 Cardiolite preparations. The Cardiolite kits were prepared with eluates from the NCWC generator (CardCis), and the Stamicis kits were prepared with eluates from the NCWC (StamCis), NFDC (StamGE), and NFWC (StamCov) generators. Table 1 presents sample characteristics, including the number of kits and the median, minimum, maximum, mean, and SD of the RCP values. Sample distribution appeared quite similar except for StamCis, which had more widely spread data, with an SD (3.55%) higher than that for CardCis (1.75%), StamGE (1.31%), or StamCov (0.94%). The statistical analysis indicated that the RCP values obtained with StamCis were significantly lower than those obtained with StamGE (P < 0.0001) or StamCov (P < 0.0001) (Fig. 1). Furthermore, the comparison of RCP values obtained with StamGE (only 2 RCP values of 51 kits were under 94%) and StamCov (no RCP value was under 94%) showed no statistical difference. On the other hand, the Cardiolite kits (16 RCP values of 56 kits were under 94%) were significantly less affected by eluates from the NCWC generators than were the Stamicis kits (32 RCP values of 53 kits were under 94%, P < 0.0001). Three to 5 different sestamibi batches were tested with each generator. The influence of different factors—such as the presence versus absence of nitrate, wet- versus dry-column generator, and kit type—on the RCP values is recapitulated in Table 2. Altogether, these results suggest a negative role for nitrates in the Stamicis preparations.
Sample Characteristics of RCP Measurements in Different 99mTc-Sestamibi Preparations
Distribution of RCP values of 99mTc-sestamibi preparations using 99mTc-pertechnetate from 3 different generators. Dotted line represents RCP limit (94%) recommended by manufacturer for clinical application. Number of RCP values under 94% is 16, 32, 2, and 0 for CardCis, StamCis, StamGE, and StamCov, respectively. CardCis = Cardiolite kit prepared with 99mTc eluate from NCWC generator; StamCis = Stamicis kit prepared with 99mTc eluate from NCWC generator; StamGE = Stamicis kit prepared with 99mTc eluate from NFDC generator; StamCov = Stamicis kit prepared with 99mTc eluate from NFWC generator; NS = not significant.
Influence of Different Factors on RCP (mean ± SD) of 99mTc-Sestamibi Preparations
To confirm nitrate effects, we tested whether the addition of sodium nitrate to 99mTc eluates from the NFWC generator modified the RCP of 99mTc-sestamibi. Clearly, an increase in nitrate concentrations decreased the RCP of 99mTc-sestamibi preparations from 96.77% for control to 67.77% for a 0.5 mg/mL concentration of nitrate (Fig. 2). Furthermore, a concentration level corresponding to that present in the NCWC generator eluates (0.05 mg/mL nitrate) led to a RCP (94.43%) lower than that of the control (P < 0.01). Myocardial perfusion images of a patient with a poor-RCP 99mTc-sestamibi preparation are presented in Figure 3. Despite the presence of impurities (17.2%), no image degradation was noted. The heart-to-lung ratio of count density (2.22) was relevant for image quality and was comparable to that initially described in healthy volunteers (3). Furthermore, the heart-to-liver ratio (0.61) in Figure 3 was in the range of those obtained for normal-heart scintigraphy with a good-quality 99mTc-sestamibi preparation (RCP value > 94%), 0.55 ± 0.10 (n = 8). The heart-to-liver ratio of normal-heart scintigraphic images with a poor-quality 99mTc-sestamibi preparation was 0.63 ± 0.13 (n = 7), slightly higher but not significantly different (P = 0.106) from ratios for 99mTc-sestamibi preparations with RCP values higher than 94%. The absence of abnormal organ uptake by the stomach or thyroid ruled out excess free pertechnetate in the Stamicis preparation (Fig. 3).
Effect of nitrate concentrations on RCP of Stamicis preparations. (A) TLC of 99mTc-sestamibi using Baker-flex aluminum oxide sheet as stationary phase and absolute ethanol as mobile phase. Dotted line indicates cutting position to determine percentages of 99mTc-sestamibi and radioimpurity fractions. (B) Nitrate-purity curve of RCP values obtained in A. Data are mean ± SD of triplicate RCP determinations. *P < 0.01. **P ≤ 0.001. A color version of this figure is available as a supplemental file at http://tech.snmjournals.org.
Myocardial perfusion images with 99mTc-sestamibi prepared with wet-column generator eluate containing nitrates (0.05 mg/mL). RCP of Stamicis preparation was 82.8%. Raw data and tomographic slices are presented in upper and lower image series, respectively. Stress images are displayed on top rows and rest images on bottom rows. Sternal region and hepatic dome were used as regions of interest to determine heart-to-lung and heart-to-liver ratios, respectively. Heart-to-lung ratio was 2.2, and heart-to-liver ratio was 0.61. A color version of this figure is available as a supplemental file at http://tech.snmjournals. org.
To investigate whether other radiopharmaceutical compounds were nitrate-sensitive, we measured the RCP of the 99mTc-succimer, 99mTc-oxidronate, 99mTc-tetrofosmin, 99mTc-exametazime, 99mTc-albumin nanocolloid, and 99mTc-soluble albumin preparations with the 3 commercially available generators (Table 3). A significant difference was observed only for 99mTc-tetrofosmin method 1 (P = 0.0002), and a tendency was observed for 99mTc-exametazime method 1 (P = 0.0532). Unexpectedly, the presence of nitrates in eluates from the NCWC generators increased rather than decreased the RCP of 99mTc-tetrofosmin and 99mTc-exametazime. However, the slight changes observed were without clinical consequence on image quality.
Influence of 99mTc Generator Eluates on RCP Values of Different Radiopharmaceuticals
DISCUSSION
The European Pharmacopoeia mentions 3 radiochemical impurities detected with 3 different methods (reversed-phase TLC, paper chromatography, and HPLC) for the 99mTc-sestamibi preparation (19). These reference methods are rather time-consuming and cumbersome without adding value for daily practice. We measured the RCP of 99mTc-sestamibi with a TLC method according to the manufacturer's instructions, which were easier to perform than the official methods proposed by the European Pharmacopoeia.
In this study, we analyzed the effects of sodium nitrate on 99mTc-sestamibi synthesis. First, the use of nitrate-containing generator eluates gave unacceptable RCPs for the Stamicis preparations when compared with RCPs from preparations reconstituted with nitrate-free generator eluates. Second, the addition of nitrates at the concentration present in the NCWC generators significantly decreased the RCP of a nitrate-free Stamicis preparation. Third, the Cardiolite kits showed a failure rate that was substantial but lower than that for the Stamicis kits when the sestamibi preparations were reconstituted with a nitrate-containing generator eluate. Fourth, no significant difference was observed between the RCP values of Stamicis kits prepared with 99mTc eluates from a dry-column generator, compared with those obtained from a wet-column generator (Fig. 1, StamGE and StamCov), suggesting that water radiolysis was not involved in the formation of Stamicis impurities. Although the NFWC generators are delivered as dry-column generators, they are considered wet-column generators because excess physiologic saline solution stays in the column after the first elution. Taken together, our results demonstrate that the presence of nitrates in 99mTc eluates is the main factor responsible for the RCP failure in the preparation of Stamicis kits.
Sodium nitrate is an oxidizing agent present in 99mTc eluates to protect pertechnetate reduction caused by the electrons emitted by 99Mo and to react with strong reducing agents, such as aqueous electron or hydrogen radicals, provided by the water radiolysis of wet-column generators. In the Stamicis kit preparations, sodium nitrate might also oxidize Sn2+ to Sn4+ and hinder the full reduction of pertechnetate ions, preventing the formation of a complex between reduced 99mTc (+I oxidation state) and 6 molecules of MIBI. According to Varelis et al. (10), the major radiochemical impurity in 99mTc-sestamibi preparations from fractionated Cardiolite kits was found on HPLC to be a 99mTc-isonitrile complex or a derivative different from sodium 99mTc-pertechnetate, 99mTc-cysteine, and 99mTc-citrate. The last 2 compounds could conceivably be produced by chelation of 99mTc with cysteine hydrochloride and sodium citrate present in the Cardiolite and Stamicis kits (21,22). Furthermore, rat scintigraphic imaging with the Cardiolite impurity (99mTc-isonitrile complex) showed no heart uptake and rapid clearance through the kidneys. This finding agrees with scintigraphic images of the human heart, for which radioimpurities do not interfere with image quality, suggesting that the behavior of both the impurities described by Varelis et al. (10) and those in our study is similar.
The significant difference in RCP values between the Cardiolite preparations and its generic medicine, the Stamicis preparation, is not easy to explain. However, the use of high-dissolved-oxygen standard saline has been described as an artifact in the stability of radiopharmaceuticals because dissolved oxygen promotes formation of peroxide and hydroperoxy radicals (23). Thus, a greater dilution volume with standard saline for the Stamicis preparations compared with Cardiolite might be an explanation for the higher RCP failure rate observed with Stamicis.
Why is 99mTc-sestamibi the only radiopharmaceutical compound interfering with nitrates in its labeling reaction? One possible reason would be the low oxidation state of technetium (+I) in the 99mTc-sestamibi complex formed from the Tc (VII) precursor and the great number of potential ligands that would stabilize the intermediate oxidation states, including the isonitrile, chloride, mannitol, cysteine, citrate, and oxo ligands (24). The pathway to the synthesis of 99mTc-sestamibi involves 7 intermediates and at least 2 steps. The first step is the lowering of the oxidation state of technetium by the initial coordination of cysteine, and the second step is the direct coordination of isonitrile. It is the intermediates in this second step that require thermal energy to produce 99mTc-sestamibi. In comparison, the oxidation states of technetium in other radiopharmaceutical compounds are IV or V, the number of intermediates is lower, and the synthesis needs no heating. In the 99mTc-sestamibi preparation, sodium nitrate probably blocks an intermediate complex, but determination of the exact chemical nature of the impurities requires further investigation.
CONCLUSION
We showed that sodium nitrate present in eluates of NCWC generators increased the RCP failure rate of 99mTc-sestamibi preparations. This effect was independent of the generator characteristics—wet or dry column—and did not affect other radiopharmaceutical compounds. The radiochemical impurities generated by nitrates did not modify the interpretation of myocardial perfusion scintigraphy, probably because these impurities are not in the heart field of view. However, we recommend using nitrate-free generator eluates in the 99mTc-sestamibi preparation to improve the product quality and to reduce unnecessary exposure of the patient to radiation.
Acknowledgments
We are grateful to all the nuclear medicine technologists and medical staff from the Department of Nuclear Medicine of Poitiers University Hospital for their expert assistance. No potential conflict of interest relevant to this article was reported.
Footnotes
Published online Jul. 7, 2012.
REFERENCES
- Received for publication November 23, 2011.
- Accepted for publication March 13, 2012.