Development and Successful Validation of Simple and Fast TLC Spot Tests for Determination of Kryptofix® 2.2.2 and Tetrabutylammonium in 18F-Labeled Radiopharmaceuticals

Kryptofix® 2.2.2 (Kry) or tetrabutylammonium (TBA) are commonly used as phase transfer catalysts in 18F-radiopharmaceutical productions for positron emission tomography (PET). Due to their toxicity, quality control has to be performed before administration of the tracer to assure that limit concentration of residual reagent is not reached. Here, we describe the successful development and pharmaceutical validation (for specificity, accuracy and detection limit) of a simplified color spot test on TLC plates. We were able to prove its applicability as a general, time and resources saving, easy to handle and reliable method in daily routine analyzing 18F-tracer formulations for Kry (in [18F]FDG or [18F]FECh) or TBA contaminations (in [18F]FLT) with special regard to complex matrix compositions.


Introduction
Production of 18 F-radiopharmaceuticals is most commonly done by labeling the respective precursor with [ 18 F]fluoride via nucleophilic substitution. [ 18 F]Fluoride is a cyclotron produced radionuclide (nuclear reaction: 18 O(p,n) 18 F) from [ 18 O]water. To assure reactivity of the [ 18 F]fluoride, it is regularly dried azeotropically and phase transfer catalysts need to be added to enhance nucleophilicity of the anion. Most prominent example for this approach is the synthesis of 2-deoxy-2-[ 18 F]fluoro-D-glucose ([ 18 F]FDG) introduced by Hamacher et al. [1].
Kryptofix ® 2.2.2 (Kry), a crown ether, is one of the most widely used among the phase transfer reagents, another one is tetrabutylammonium (TBA; added e.g., as hydroxide, carbonate or hydrogen carbonate). Radiopharmaceutical formulations have to be analyzed for residual contaminations with these reagents before human application, due to their toxicity (e.g., Kry: LD 50 (i.v.) in rodents = 32-35 mg/kg [2] [3]), in US Pharmacopoeia (USP) limit is <50 µg/mL [4]. For TBA in 18 F-radiopharmaceuticals a limit of 2.6 mg/V is defined in Ph. Eur. (see same monographs as for Kry). Ph. Eur. suggests a color spot test on a thin layer chromatography (TLC) plate for Kry and a HPLC method for TBA in 18 F-radiopharmaceutical monographs (see above).
In the past, various procedures have been described in the literature for identification and quantification of Kry, with the aim of providing specific and fast methods. TLC procedures are the most widespread, as they do not require sophisticated instrumental equipment. A color spot test on a TLC plate as used in monographs with iodoplatinate [5] for staining was described by Mock et al. [6]. To increase specificity and avoid false positive (e.g., from other amines) or false negative results (from stabilizers) thin layer chromatography systems were developed [7][8][9][10]. But also gas chromatography (GC; [11]), high-performance liquid chromatography (HPLC; [12]) or liquid chromatography-tandem mass spectrometry (LC/MS/MS; [13]) and even NMR or IR spectroscopy [14] were applied. More recently, a very fast (<1 min) and highly sensitive (lower limit = 0.5 ng/mL) rapid-resolution liquid chromatography MS/MS coupled system for analysis of Kry was p ublished [15]. Still, these methods suffer from certain disadvantages. The latter (GC, LC, MS, NMR, IR) need high-quality instrumental equipment combined with a higher effort for validation. On the other hand, TLC tests are more time consuming due to development of plates or require the expensive reagent iodoplatinate. In addition, in our opinion to prove that in a certain radiopharmaceutical formulation the Kry (or TBA) content is below a defined limit-without the necessity of exact quantification, the te st should be as simple and quick as possible, but nevertheless meeting pharmaceutical requirements. This particularly holds true in view of the fact that all literature data agree that, when synthesis and purification were performed successfully, Kry or TBA residues were extremely low and far from limits. When any test for Kry (or TBA) is applied to a newly produced radiopharmaceutical, it will always need some kind of validation, no matter how specific the test proved to be so far. Matrix effects may play an important role, pH as well. The important factor is how much time and effort it will take in daily routine.
After first promising results with the TLC spot tests as will be described below for Kry and TBA, we set up validations regarding specificity, accuracy and detection limit for [ 18 F]FDG (Kry) and, to demonstrate a more general applicability, for 2-[ 18 F]fluoroethylcholine ([ 18 F]FECh; also Kry), where residual N,N-dimethylaminoethanol (DMAE) may interfere [16]. As the HPLC procedure for TBA analysis described in the monographs appeared to be challenging, the spot test was also to be validated for this reagent, with [ 18 F]FLT as example [17]. As an additional important parameter, the stability of the necessary standard solutions containing the limit co ncentrations of Kry or TBA should also be investigated.

General
All reagents were of highest purity available or of pharmaceutical grade, where necessary (article numbers in brackets) and were used as received.

Solutions for the Validation of Kryptofix 2.2.2 in [ 18 F]FECh
As matrix buffer PBS was used (pH 7.4). When necessary, pH was adjusted to pH 6.5 using a 10% solution of NaH 2 PO 4

TLC Procedure for Analysis of Kry
Spots of 2 µL samples were applied on silica plates by means of an Eppendorf micro pipette. It proved to be mandatory that the tip slightly touched the plate and allowed the liquid to drain slowly and penetrate the silica material without any active pressure from the pipette. Plates were immediately placed on a holder inside a glass chamber for 10 min above iodine (ca. 5 g) covering the complete bottom of the vessel, homogenously saturating the whole chamber with iodine vapor. Plates were photographed for documentation (PowerShot SX110 IS, macro mode; jpg files, Canon, Krefeld, Germany) and visually analyzed. All experiments were n = 3.

TLC Procedure for Analysis of TBA
Spots of 2 µL samples were applied on silica plates by means of an Eppendorf micro pipette, in analogy to Kry analysis, in addition it was essential that only two spots per plate were applied. Otherwise time was to long between first and last spot application and results were invalid. Spots needed to be ca. 10 mm from the edges of the plate. After drying of spots with a cool air stream (hair dryer) 10 µL of the MeOH/NH 4 OH solution were applied on each spot. Plates were then placed in the same iodine containing chamber as for Kry, for exactly 1 min. Afterwards, plates were photographed immediately for documentation and visually analyzed (before discoloration). All experiments were n = 3.

Results and Discussion
TLC spot tests were validated as quality control methods to assure that residual contaminations of Kry ([ 18 F]FDG, [ 18 F]FECh) or TBA ([ 18 F]FLT) in 18 F-radiopharmaceutical formulations were below the limits required by e.g., Ph. Eur. Objectives of the validation experiments were specificity, accuracy and detection limit. Moreover, stability of standard solutions for routine quality control was evaluated. Validations had to be performed individually for each radiopharmaceutical, taking into account that determination of either Kry or TBA may be matrix dependent. All experiments were performed in triplicate.

Validation of Kry in [ 18 F]FDG Formulation
Concentrations of 100 mg/L of Kry were used either in matrix or [ 18 F]FDG solutions, in accordance with requirements of current Ph. Eur. [3]. The limit defined in the Pharmacopoeia is 2.2 mg/V, i.e., per patient. The maximum volume in our case that could theoretically be injected in one patient is 20 mL, resulting in a calculated limit concentration of 110 mg/L. Our specification of <100 mg/L is more narrowly defined. For determination of the detection limit of the method additional solutions with concentrations from 3.1 to 125 mg/L were used. After staining with iodine, background on the plates showed an orange, light brown color. Spots of solutions containing Kry were dark brown with a diameter of 1-2 mm. In first experiments, the standard solution was prepared in WFI instead of citrate buffer. But it could be shown that as result spots were darker and of a smaller diameter. Therefore, Kry standard solutions are always prepared using buffer, to have most comparable conditions. Also, in preliminary tests it was shown that the intensity of the color of the Kry spot was dependent on the pH value of the test solution. At pH 5.0, intensity was lower than at higher pH. Therefore, buffers of standard solutions were adjusted to pH 5.0, representing the most unfavorable case. In general, content of residual Kry in [ 18 F]FDG solutions (>100 productions) was far below the limit of 100 mg/L. Only in cases where pH of a solution was 6.0 or higher and the spot of the batch sample was of similar or higher intensity than the standard solution (100 mg/L; pH 5.0) there would be uncertainty. Then, the standard solution would need to be adjusted to the pH of this batch sample and the test to be repeated, to prove that Kry concentration of the batch solution was below the specified limit.
To test for the specificity of the method, [ 18 F]FDG solutions and plain matrix buffers (pH 5.0 and 6.0) were investigated. With matrix buffers no coloration of the spots was visible, with [ 18 F]FDG a very weak, diffuse, light brown spot developed that was completely different from Kry spots from standard solutions (Figure 1a,b). Comparison of standard solutions of 100 mg/L of Kry in either [ 18 F]FDG solution or matrix buffer (pH 6.0) showed small, dark spots of comparable intensity (Figure 1b). Third, the two series of standard solutions (pH 5.0 and pH 6.0) with Kry concentrations of 3.1, 6.25, 12.5, 25, 50, 75, 100 and 125 mg/L were investigated (Figure 1c). Within both series spots differing in concentrations by a factor of 2 could be distinguished. As already shown in the preliminary tests, spots from a certain concentration at pH 6.0 corresponded to spots from the series at pH 5.0 with twice the concentration. visible, whereas spots 1 and 2 showed same shape and intensity. The results from these validations met the defined specifications from the respective validation plan, proving the applicability of the described TLC method as test for residual Kry in [ 18 F]FDG formulations, regarding specificity, accuracy and detection limit. For routine quality control it was decided to use a standard solution containing 100 mg/L of Kry in matrix buffer pH 5.0, which then is tested against the individual [ 18 F]FDG batch. This standard solution is prepared as large batch and stored in portions of 1 mL in small glass vials at <−15 °C. In frame of the validation the stability of the standard solution was evaluated after 2 weeks, 1 month, 3 months and 6 months by comparison with a freshly mixed solution. 3 different batches of standard solution were prepared. These stability studies demonstrated that spots from stored solutions showed no difference to spots from fresh solutions. Consequently, the shelf life of the Kry standard solution was specified to be 6 months.

Validation of Kry in [ 18 F]FECh Formulation
Validation was performed in a similar way as for [ 18 F]FDG, regarding specificity, accuracy, detection limit and stability of standard solutions. Kry standard solutions had the same concentration of 100 mg/L as limit, as there is no monograph for [ 18 F]FECh yet available. Calculating from our batch volume of 13 mL that might be injected in a patient, a limit concentration of 169 mg/L would result  For defining the detection limit the series of Kry concentrations (Figure 2c) were analyzed. The spot from the Kry concentration of 12.5 mg/L was clearly visible, detection limit was therefore well below the specified nominal value of -≤50 mg/L‖. Compared to the [ 18 F]FDG validation, detection limit was higher for Kry in [ 18 F]FECh solutions.
Accuracy of the method was again proven by comparing spots of 100 mg/L Kry in matrix buffer (1); [ 18 F]FECh solution, drying, addition of 100 mg/L Kry in buffer (2) and as spot 3 [ 18 F]FECh solution was applied. Experiments were again performed at pH 6.5 and 7.4, showing no pH dependence. In both cases spots 3 showed minimum coloration, spots 1 and 2 were of same intensity (Figure 2d). Overall, validation was performed successfully, clearly demonstrating the feasibility of this TLC spot test.
Stability of the standard solution to be used in routine quality control (100 mg/L Kry in PBS matrix buffer, pH 7.4), stored in 1 mL portions at <−15 °C was proven to be 6 months (three different batches; performance as in [ 18 F]FDG validation).  To meet Ph. Eur. requirements for the limit of TBA impurities in 18 F-radiopharmaceuticals (2.6 mg/V) the general standard solution contained 100 mg/L TBA, as the maximum volume in our case is 21 mL, i.e., the calculated limit from Ph. Eur. would be 124 mg/L.

General Discussion
Overall, the presented and validated TLC spot test proved its capability for determination of residual Kry in [ 18 F]FDG and [ 18 F]FECh formulations with the perspective to become a general method. Indeed, validations have to be carried out for each tracer formulation and its individual matrix, as organic impurities (e.g., amines) might give false positive or stabilizers false negative results. So far, our results never showed disturbing interferences of major concern. No problems arose from neither ethanol nor residual DMAE (in [ 18 F]FECh solutions) or pH dependencies of spot intensities. Latter effect would lead to the necessity of pH adjustments of standard solutions for valid results of each batch in daily routine, but on the one hand residual Kry and also TBA concentrations close to the specified limits were never observed, normally they are not detectable-and detection limits are low. On the other hand, in the investigated radiopharmaceuticals (all in buffered solutions) alterations in pH values were minimal (Table 1).  Whereas the test for Kry proved its robustness, the test for TBA reacted more sensitively to variations and instructions for performance of the test had to be followed strictly, like fast administration on plate, drying, time of development etc. Only two spots could be administered on one plate, to minimize time before iodine staining, otherwise results were erroneous. Only spots on one plate were comparable, which made validation laborious, but gives no problem in routine analysis, when only sample and standard solution are applied. Detection limit was higher for TBA than for Kry and concentrations of TBA spots that were to be compared needed to differ by 60 mg/L to give valid results. Still, these aspects are no problem for routine analysis, as TBA concentrations in [ 18 F]FLT formulations from routine production were found to be very low (see Figure 3a), and then determination of a concentration of -<100 mg/L‖ is reliably assured ( Table 1).
This TLC spot test is designed to determine required limits to be met and not to absolutely quantify the concentration of Kry or TBA present, although this was possible for [ 18 F]FDG or [ 18 F]FECh. Nevertheless, in routine quality control for chemical impurities the specification -below limit‖ (e.g., -<100 mg/L‖) is absolutely sufficient and in conformity with Pharmacopoeia norms.

Conclusions
A TLC spot test for determination of residual Kry or TBA in 18 F-radiopharmaceutical formulations was successfully developed and formally validated for specificity, accuracy and detection limit for [ 18 F]FDG, [ 18 F]FECh (Kry) and [ 18 F]FLT (TBA). Interference with various matrix effects was negligible; the test is adequate, easy to handle, very fast and resources saving (no expensive reagents or HPLC system necessary). Now, it is already implemented in routine quality control of these three tracers in our institution, in case of [ 18 F]FDG it was formally accepted by authorities within our marketing license. With these positive results, the test is now under validation for further 18 F-radiopharmaceuticals.

Conflicts of Interest
M.M. is an employee of ABX. All other authors declare no conflict of interest.