Stability and Beyond-Use Date of a Compounded Thioguanine Suspension

Abstract

Background: Thioguanine (TG) is available only in the form of 40 mg tablets in the United States, and the patient population in which TG is used comprises mostly children. Recognizing its importance as a therapeutic agent and limited stability data for its compounded preparation, the United States Pharmacopoeia has listed TG in its priority list of compounded preparations monographs. Objective: The goal of the present study was to generate stability data and establish a beyond-use date for compounded TG suspension. Methods: Suspensions were compounded using TG tablets and ORA-Plus and ORA-Sweet as vehicles. A robust high-performance liquid chromatography method was developed and validated. TG and guanine (G) in suspensions were quantified immediately after compounding and at regular intervals for 90 days. Physical stability of suspensions was evaluated by observation of organoleptic properties. Results: Results from the study indicate that average TG levels in suspensions remained above 90% of the starting concentration and G formation was less than 2.5% for 90 days. There was no statistically significant difference in the amount of TG degraded over 90 days between suspensions stored at room temperature and in refrigerated conditions. There was also no statistically significant difference in G concentration of suspensions between day 0 and day 90. Conclusion: TG suspensions are stable for 90 days when stored at room temperature or refrigerated conditions and the beyond-use date can be set to 90 days.

Background

Thioguanine (TG) was approved in the United States in 1966 for the treatment of acute and chronic myelogenous (nonlymphocytic) leukemias.^1,2 TG has also been used in the management of psoriasis, pediatric acute lymphoblastic leukemia, and inflammatory bowel disease.^3-5 More recent studies have confirmed that TG is safe and efficacious for the treatment of celiac disease in low doses.⁶

TG is a prodrug and is activated by hypoxanthine-guanine phosphoribosyl transferase (HGPRT). TG competes with guanine (G) and hypoxanthine for HGPRT and is converted to 6-thioguanosine monophosphate (TGMP). TGMP, in turn, interferes with the synthesis of G and other purine ribonucleotides by several mechanisms.^6,7 TGMP incorporation into DNA inhibits DNA replication via interstrand cross-linking, single-strand breaks, and sister chromatid exchange.^8,9

Extemporaneous compounding is routine in pharmacies, and it is essential that compounding pharmacists and practitioners have information on the formula, compounding directions, and stability of the preparation. TG is only available in the form of 40 mg tablets in the United States and the patient population in which TG is used comprises mostly children. Recognizing the importance of TG as a therapeutic agent and limited stability indicating data for compounded TG suspensions, the United States Pharmacopoeia (USP) has included TG in its priority list of compounded preparation monographs.¹⁰ These monographs provide information about the formulation, compounding instructions, storage, and beyond-use date (BUD) based on stability studies.¹¹

TG primarily degrades to G at room temperature and a few methods for TG quantitation have been published.^12-18 A 1983 study showed that TG suspension compounded with methylcellulose, wild cherry syrup, and simple syrup was stable for 84 days at room temperature. The study did not report pH measurements or G levels.¹³ A 2011 article detailed an LC-MS (liquid chromatography-mass spectrometry) method for quantification of TG and G in a suspension compounded from Lanvis tablets, which are only available in Canada and Europe.¹² Polonini and colleagues¹⁹ recently reported the stability of TG 2.5 mg/mL suspensions in Syrspend SF pH 4 using TG powder as the drug source. To our knowledge, there are no published reports for stability of TG suspensions compounded by using the currently available Tabloid tablets.

Objective

To develop and validate a high-performance liquid chromatography (HPLC) method for the analysis of TG and G from a compounded TG suspension and establish the BUD for the suspension.

Materials

Chemicals and Reagents

Pharmaceutical secondary reference standard TG (lot number SLBW6998) and G (lot number LRAC0250) were purchased from Sigma-Aldrich. Tabloid (Excella GmbH and Co) 40 mg TG tablets (lot numbers 811307 and 700515), ORA-Plus (Perrigo, lot number 7371150), ORA-Sweet (Perrigo, lot number 9237910), hydrochloric acid (HCl, Fisher Scientific, lot number 193303), sodium hydroxide pellets (NaOH, Fisher Scientific, lot number 068412), hydrogen peroxide 30% (Fisher Scientific, lot number 177290), dimethyl sulfoxide (DMSO, Acros, lot numbers B0518100 and B0511978, and Fisher Scientific, lot number 192803), and syringe filters (0.2 µm, Fisher Scientific) were used in the study. Standard buffer solutions (Traceable, lot numbers CC631032 and CC608021, and Fisher Scientific, lot number 11278) were used to calibrate the pH meter. Water for formulation and analysis was obtained from an in-house MilliQ Biopak system. An alcohol thermometer (Fisher Scientific) and SHT-85 temperature sensor (Sensiron) were used to record temperature. All chemicals were used as received.

Instrumentation

HPLC analysis was performed on a Shimadzu LC system with LC-20AT pump, SIL-20 A injector, and SPD-M20A photo diode array detector. A reversed-phase C-18 column (Waters Atlantis T3, 150 mm × 2.1 mm, 3 µm particle diameter) and guard column (Waters VanGuard 5 mm × 2.1 mm) were used for separation. The mobile phase was water at a constant flow rate of 0.3 mL/minute. TG was detected by ultraviolet absorbance at 248 nm. The column was maintained at 30°C with a CH-30 column heater (Eppendorf). An Accumet XL60 pH meter (Fisher Scientific) and an Orion Ross 8104B combination electrode were used for pH determinations. A Mettler Toledo XS-205 balance was used for weighing. A Branson 1570 sonicator was employed throughout the study.

Methods

Compounding Extemporaneous Suspensions

Suspensions were compounded as follows: a suspending vehicle mixture of ORA-Plus:ORA-Sweet (1:2) was prepared. Fifteen tablets, each containing 40 mg of TG, were added to 1-ounce amber PET (polyethylene terephthalate) prescription bottles followed by addition of 7.5 mL of water. The bottle was vortexed for 1 minute and allowed to sit for 2 minutes. Vortexing and resting was repeated 5 times. Finally, the volume was made up to 30 mL with suspending vehicle mixture and the bottles were vortexed for another minute. Nine suspensions were prepared, 3 of which were stored at room temperature (22°C) protected from light, 3 suspensions were stored in refrigerated conditions (4°C) protected from light, and 3 were used on the same day for the homogeneity and robustness tests. The pH of all suspensions was determined using the Accumet XL60 pH meter. The pH meter was calibrated using standard buffer solutions of pH 4, 7, and 10. Suspensions stored in refrigerated conditions were allowed to equilibrate to room temperature before the pH measurements were made.

Preparation of Standard Samples

A standard stock solution was prepared by dissolving reference standard TG in DMSO. The target concentration (assay level) of TG for HPLC analysis was set to 50 µg/mL throughout the study. About 20 mg reference standard drug was added to a 10 mL volumetric flask and 7 mL of DMSO was added. The flask was sonicated for 10 minutes in a water bath maintained at 40°C. After allowing the flask to equilibrate to room temperature, the volume was made up to 10 mL with additional DMSO. A 2.5 mL aliquot of this standard stock solution was transferred to a 25 mL volumetric flask and made up to volume with water. Aliquots were then diluted with water in 10 mL volumetric flasks to final drug concentrations of 20, 40, 45, 50, 55, 60, 80, and 100 µg/mL.

The concentration of G in a compounded preparation of TG should not be more than 2.5% of the initial TG concentration.²⁰ The target concentration of TG for analysis was set to 50 µg/mL. Therefore, the limit of G in the compounded suspensions was 2.5% of 50 µg, that is, 1.25 µg/mL. To prepare a calibration curve for quantifying G, a stock solution was prepared by dissolving 20 mg G in 10 mL DMSO. Aliquots of stock solution were diluted with water in volumetric flasks to final concentrations of 1, 1.125, 1.25, 1.375, and 1.5 µg/mL.

Preparation of Assay Samples

A 0.5 mL aliquot of TG suspension was mixed with 7 mL of DMSO in a 10 mL volumetric flask. The flask was sonicated for 10 minutes in a water bath maintained at 40 °C. After allowing the flask to equilibrate to room temperature, the volume was made up to 10 mL with DMSO. The flask was shaken for 10 seconds and 1.25 mL of the solution was diluted with water to 25 mL in a volumetric flask. A 2 mL sample of the final solution was filtered through a 0.2-µm syringe filter for HPLC analysis.

Preparation of System Suitability Solution

System suitability solution was prepared by adding measured volumes of ORA-Plus (0.125 mL) and ORA-Sweet (0.25 mL) to 1 mL of TG reference standard stock solution (20 mg/mL) and the volume was made up to 10 mL with water. A 1.25 mL aliquot of the solution was diluted to 50 mL with water to get a final solution containing 50 µg/mL of TG. A 2 mL sample of the final solution was then filtered through a 0.2-µm syringe filter into a vial for HPLC analysis.

Method Validation

The assay for compounded TG suspensions was validated based on the USP technical proposal for compounded preparation monograph development.^21,22 System suitability of the HPLC system and the developed method were assessed by analyzing system suitability samples and calculating the resolution, tailing factor, and column efficiency (number of theoretical plates). Accuracy of the method was assessed by testing standard samples in triplicate at 80%, 100%, and 120% of target assay level and calculating percentage recovery of TG from these samples. Precision of the method was determined by calculating the % relative standard deviation (% RSD) for solutions of TG at 80%, 100%, and 120 % of target assay level.

Specificity of the method was assessed by analyzing TG samples subjected to forced degradation conditions. TG 2 mg/mL with 0.01 N HCl, TG 2 mg/mL with 0.01 N NaOH, and TG 2 mg/mL with 1% hydrogen peroxide were stored at room temperature in the dark. A fourth sample of TG 2 mg/mL with no additives was stored at 60°C and protected from light. Aliquots were periodically drawn from the samples and diluted for analysis as described for assay samples. The concentration of TG before the start and at the end of degradation study was calculated.

Other validation tests were intermediate precision (assay performance on different days and by different analysts), linearity, robustness to small variations in assay conditions, filter suitability, and intraday and interday stability of standards and assay solutions.

Three TG suspensions were used to assess homogeneity of the suspensions. From each suspension bottle, three 0.5 mL aliquots were sampled from the top, middle, and bottom part of the bottle. These aliquots were diluted for assay and analyzed in triplicate. The percent drug recovery at each sampling position (top, middle, and bottom) was calculated and results were compared.

Physical stability of the compounded suspensions was evaluated by inspecting them for color, odor, dispersibility, and caking. The pH of suspensions was measured on the day of preparation (day 0) and after 90 days of storage.

Chemical stability of TG suspension was assessed by analyzing TG and G concentrations in the compounded suspension on days 0, 7, 15, 30, 45, 60, and 90. The suspensions were shaken well before sampling. The aliquot from each suspension was analyzed in triplicate at each time point and the data averaged. The concentration of G formed in the suspension during the storage period was also calculated and recorded.

Room temperature throughout the study period was recorded using an SHT-85 temperature sensor and recorded on a laptop computer. The temperature inside the refrigerator was monitored by using an alcohol thermometer. The average temperature in each condition was determined at the end of study period.

Data were analyzed for mean and standard deviation using Microsoft Excel. Mean differences were compared using the Student’s t test.

Results

The assay separated TG from G and other suspension components, with resolution greater than 2 and more than 2000 theoretical plates. The peak response was linear with r² > 0.999 for the TG calibration curve in the range of 20 to 100 µg/mL as shown in Figure 1 (open circles). The assay was also linear in the range of 80% to 120% of the target assay concentration, with r² >0.995 and y-intercept <1.5% of the peak response at the target concentration. The residuals were randomly distributed and were less than 1.5% of peak response at the target concentration. The calibration curve for G was also linear with r² > 0.998 in the range of 1.050 to 1.565 µg/mL, as shown in Figure 1 (filled circles).

Figure 1.

Calibration curves for thioguanine (TG; open circles) and guanine (G; filled circles).

Resolution and number of theoretical plates for TG peak were greater than 2 and 2000, respectively, and tailing factor was less than 2. The recovery of TG from standard samples was between 98% and 102% and % RSD was less than 2.0% for the replicate injections at each assay level. Samples prepared by 3 analysts on 3 days showed % RSD of <3%. Recovery of TG from solution stability test samples was between 97% and 103% for up to 3 days after sample preparation. TG suspensions were homogeneous, with less than 3% difference between any 2 sampling regions. Assay validation results are summarized in Table 1.

Table 1.

Assay Validation Data.

Validation tests	Results
System suitability	Resolution		Theoretical plates		Tailing factor		% RSD
	6.90		5186		1.2		0.31
Accuracy	% recovery at 80%		% recovery at 100%		% recovery at 120%		Repeatability % RSD
	99.44		98.65		98.74		0.33
Intermediate precision	% RSD for 9 different combinations of variations
	2.1%
Specificity	Degradant formed		% TG degradation with hydrogen peroxide treatment		Peak purity index		Co-elution
	Guanine		14%		>0.99, indicating spectrally pure peak		No co-elution with degradants
Linearity	Equation of line		r2 for the line		Residuals		Intercept
	Y = 13135x + 94224		0.998		Randomly distributed, less than 1.5% of peak response at target concentration		Less than 1.5% of peak response at target concentration
Robustness (compare with system suitability results)	Effect of flow rate variation				Effect of temperature variation
	Resolution	Theoretical plates	Tailing factor	% RSD	Resolution	Theoretical plates	Tailing factor	% RSD
	6.36	4160	1.13	0.29	6.41	4252	1.13	0.10
Solution stability	Intraday stability				Interday stability
	Average % recovery of TG over 24 hours		Recovery %RSD		Average % recovery of TG over 3 days		Recovery % RSD
	99.55		0.36		99.07		0.78
Filter suitability	% variation in TG recovery between centrifuged and filtered standard sample				%variation in TG recovery between centrifuged and filtered suspension sample
	−1.25				0.22

Abbreviations: RSD, relative standard deviation; TG, thioguanine.

TG suspensions that were intentionally stressed with 0.01 N HCl at room temperature, 0.01 N NaOH at room temperature, or storage at 60°C showed minimal degradation to G after 5 days (data not shown). Reaction with 1% hydrogen peroxide, on the other hand, resulted in significant TG degradation by approximately 14% to form G. Representative chromatograms are shown in Figure 2. Peaks at 1.8 and 2.2 minutes were confirmed to be DMSO and hydrogen peroxide, respectively. No coelution or interference between TG, G, and the excipients was observed in the chromatograms. Diode array based spectral peak purity analysis confirmed that the TG peak in the sample stressed by hydrogen peroxide was spectrally pure, with peak purity index greater than 0.99.

Figure 2.

Representative chromatograms of thioguanine reference standard, thioguanine suspension, thioguanine forced degradation by hydrogen peroxide, and guanine reference standard.

The pH of all suspensions was measured at day 0 and day 90. The pH of the suspensions on the day they were compounded averaged to 4.54 ± 0.26. After a 90-day storage period, the pH of suspensions averaged to 4.61 ± 0.24.

The compounded suspensions were assessed for physical stability by evaluating their organoleptic properties. The suspensions retained their original white color and syrupy odor and were easy to redisperse throughout the storage period. While microbiological assays were not performed, organoleptic evaluation revealed no indication of microbial growth in any suspension.

The results of the chemical stability testing of the compounded TG suspensions stored at room temperature and in refrigerated conditions are shown in Table 2. The TG concentration remaining in suspensions ranged from 94.1 ± 0.51% to 106.5 ± 1.61% for room temperature samples and 94.8 ± 0.80% to 105.1 ± 1.58% for refrigerated samples. The average concentration of TG in all 6 suspensions was 18.22 mg/mL at day 0 and 17.62 mg/mL on day 90.

Table 2.

Chemical Stability of Compounded Thioguanine Suspensions.

Day	Day 0 concentration (mg/mL) and thioguanine remaining in compounded suspensions (%)
	Room temperature (22 °C)			Refrigerated (4 °C)
	1	2	3	4	5	6
0	18.5 ± 0.16	17.9 ± 0.45	18.2 ± 0.14	18.5 ± 0.13	18.5 ± 0.26	17.7 ± 0.73
7	100.5 ± 3.37	99.8 ± 1.54	105.3 ± 1.64	101.3 ± 1.35	100.0 ± 1.15	102.1 ± 1.61
14	97.9 ± 0.76	97.9 ± 1.10	94.1 ± 0.51	97.4 ± 1.00	97.9 ± 1.26	102.1 ± 1.17
30	98.6 ± 0.90	99.7 ± 1.10	97.1 ± 1.84	97.3 ± 2.47	93.2 ± 2.07	97.6 ± 1.51
45	105.6 ± 1.04	106.5 ± 1.61	105.7 ± 1.67	100.9 ± 3.25	105.3 ± 3.96	105.1 ± 1.58
60	103.6 ± 3.54	103.4 ± 3.05	102.8 ± 1.09	100.6 ± 1.38	99.2 ± 0.59	103.7 ± 0.63
90	95.8 ± 0.78	96.6 ± 2.28	99.1 ± 0.68	94.8 ± 0.80	95.7 ± 0.93	98.4 ± 1.05

An independent samples t test revealed no statistically significant difference in the TG concentration between the suspensions stored at room temperature and in refrigerated conditions over a storage period of 90 days (P > .05). The concentration of G in all suspensions was less than 2.5% of the starting TG concentration at the end of the storage period. An independent samples t test revealed no statistically significant difference in the G levels between day 0 and day 90. (P > .05).

Discussion

The assay method was validated and all test results conform to the acceptance criteria set by USP for official compounded formulation monographs. System suitability tests were performed with solutions prepared from both reference standard TG powder and suspension samples. The syringe filters used throughout the study were demonstrated not to interfere with TG quantitation. Standard and assay samples were stable for up to 3 days after sample preparation, allowing for delayed analysis if necessary. Forced degradation studies demonstrated excellent resolution between TG and other components. Thus, the assay was shown to be robust and stability-indicating.

The suspension formulation employed Ora-Plus and Ora-Sweet, which are widely available and commonly used for compounding of extemporaneous formulations. These vehicles were chosen to represent common compounding practice and to provide a standardized formulation such that pharmacists may be assured of the compatibility and BUD of TG suspensions they compound. The proportions of 1 part water, 1 part Ora-Plus, and 2 parts Ora-Sweet were chosen based on prior experience with olmesartan medoxomil compounded suspension.²³ The results cannot be extrapolated to suspensions prepared with other vehicles due to the potential effects of different formulation components.

The suspension formulation is prepared in the dispensing bottle, eliminating the potential for drug loss in compounding process. Looking ahead to future implementation of USP <800>, the compounding process was conducted in a closed system and represents low risk to pharmacy personnel. The TG concentration was homogeneous throughout the compounded suspensions. Shaking the suspension well before measuring ensures the patient will receive an accurate and consistent dose of drug.

TG suspensions degraded by less than 10% for 90 days after compounding whether stored at room temperature or in the refrigerator. These findings are in agreement with Poloninin and colleagues,¹⁹ who found TG 2.5 mg/mL suspension in Syrspend SF pH 4 to be stable for 90 days at both room temperature and refrigerated conditions.¹⁹ The average change in suspension pH over this time period was less than 0.1 units. The TG suspensions were, therefore, chemically and physically stable for 90 days. While microbiological assays were not conducted, organoleptic evaluation revealed no evidence of increased microbial load over the study period. This was expected as both the vehicles (Ora-Plus and Ora-Sweet) contain preservatives.

As shown in Table 2, the average TG concentration at the time of compounding was 18.22 mg/mL, while the expected concentration was 20 mg/mL. The lower day 0 concentration was not due to compounding errors, as whole intact Tabloid tablets were placed directly in the dispensing bottle. The source of the TG deficit was investigated by comparing the content of 3 tablets against the standard curve prepared with TG reference standard (results not shown). Assay showed that the suspension concentration reflects the potency of the Tabloid tablets from which they were prepared. The tablets were stored in an air-conditioned room in the original container and closed tightly after each use and were used within the labeled expiration date.

It is known that TG absorbs ultraviolet A light to produce singlet oxygen species (¹O₂), leading to a variety of oxidation products. A study demonstrated the formation of guanine sulfonate (GSO₃) and G as degradation products of TG and ¹O₂ via low- and high-energy barrier pathways, respectively.²⁴ Study suspensions were stored in amber prescription bottles to protect the product from light, and the bottles were stored in a dark cabinet, minimizing the practical relevance of these reactions. Exposure to light was therefore not included as part of forced degradation studies.

The main limitation of the study was lack of microbial testing. The water activity of the suspensions was measured and found to exceed 0.9 (data not shown). It is generally recommended that aqueous formulations with water activity greater than 0.9 include a suitable preservative to inhibit microbial growth throughout the product shelf life.²⁵ The vehicles used in this study, Ora-Sweet and Ora-Plus, both contain preservatives. Organoleptic evaluation did not detect signs of microbial growth in TG suspensions, but this is not proof of microbial stability. Only rigorous microbial testing can validate the effectiveness of preservatives in a formulation. In vivo studies were not conducted, so inferences regarding bioavailability of the suspension cannot be made.

Conclusion

TG suspensions compounded using Tabloid TG tablets and ORA-Plus and ORA-Sweet as vehicles retained their original color and odor and were easy to redisperse. The concentration of TG in all suspensions was more than 90% of the starting concentration and that of G was less than 2.5% of the TG concentration throughout the 90-day storage period at 22°C and 4°C. The suspensions did not show formation of any degradant other than G. Therefore, the BUD for TG suspensions compounded using ORA-Plus and ORA-Sweet as vehicles may be set to 90 days, whether stored at room temperature or refrigerated conditions.