Simple and rapid HPLC-UV methods for gabapentin quantification in human plasma and urine: applicability in pharmacokinetics and drug monitoring

Aim: We aimed to develop methods to determine gabapentin (GAB) in biological samples using high-performance liquid chromatography (HPLC) with application in pharmacokinetics and therapeutic drug monitoring. Methods: Simple, rapid and efficient HPLC-UV methods to quantify GAB in human plasma and urine were developed and validated for clinical analysis of GAB. The 1fluoro-2,4-dinitrobenzene (FDNB) was used as derivatization agent. For plasma samples, protein precipitation using acetonitrile was performed, before the derivatization reaction. Urine samples were cleaned-up by liquid-liquid extraction with dichloromethane:n-butanol (1:1, v/v) after derivatized. Amlodipine besilate was used as internal standard (IS). Results: Gabapentin and IS were resolved on LiChrospher® C18 RP column and a mixture of 50 mM sodium phosphate buffer (pH 3.9):methanol (27:73, v/v) as mobile phase, at 1.2 mL/min. The methods used small sample volumes, 100 and 50 μL of plasma and urine, respectively. Linearity was obtained in the interval of 0.2-14 μg/mL in plasma and 2-120 μg/mL in urine. Both methods showed to be selective, without carry-over effect, precise, accurate and stable in different conditions. GAB plasma concentration in patients receiving 600 to 3600 mg/day of GAB ranged between 0.40 to 11.94 μg/mL at steady-state. Conclusions: The methods described in this study were simple, rapid and fulfill all validation requirements. They were easily and successfully applied for population pharmacokinetics and therapeutic drug monitoring of GAB in patients with chronic pain.


INTRODUCTION
Pain is considered a public health problem 1 . In 2019, around 20% of adults reported chronic pain in the past 3 months in the U.S., according to the National Health Interview Survey 2 . In the Brazilian population the prevalence of chronic pain varied between 29.3 to 73.3% 3 . Neuropathic pain is one of the seven categories of chronic pain in the International Classification of Diseases (ICD-11). The number of people affected by neuropathic pain should increase even more due to the increase of elderly population, diabetes incidence and higher survival from cancer 4 .
Gabapentin (GAB) is recommended as first-line treatment for various neuropathic pain conditions (e.g. diabetic neuropathic pain, postherpetic neuralgia, central pain) 5 . For patients with hepatic dysfunction GAB is possibly the safest choice because it is not metabolized, does not bind to plasma proteins and is mainly eliminated unchanged in urine [6][7][8] . Although GAB is widely prescribed (in 2019, 69 million GAB prescription were dispensed in U.S.), its therapeutic drug monitoring (TDM) is not usual in clinical practice 9 . TDM can be applied when a reference target of plasma concentrations associated with drug efficacy and minimal risk of adverse events is available. In the context of drugs to treat neuropathic pain, TDM can be also a tool to monitor patient adherence, to help physicians decide the best treatment for the patient and to avoid unnecessary opioid prescription. Opioids are the third-line treatment for neuropathic pain, but show high risk of addiction 5,10 .

Standard and reagents solutions and quality control samples
Stock solutions of GAB for plasma and urine analysis were prepared in methanol at 100 and 1000 µg/mL, respectively. The work solutions were prepared at concentrations of 1, 2, 5, 10, 20, 30, 60 and 70 µg/mL in methanol for plasma. For urine the concentrations were 5, 10, 20, 30, 50, 75, 100, 150, 200 and 300 µg/mL in methanol. The standard solutions were stored at -20 ºC. Amlodipine besylate, used as internal standard (IS), was prepared at concentration of 200 µg/mL in water and stored at 4-8 ºC. The stock solution of 0.06 M FDNB was prepared in acetonitrile and stored at 4-8 ºC.

Method validation
Full validation was performed according to the "Guideline on Bioanalytical Method Validation" by European Medicines Agency 22 . The LLOQ was determined by analyzing samples spiked with 0.1 or 0.2 µg of GAB/mL plasma; and 2 or 4 µg of GAB/mL urine. Selectivity was performed with six blank plasma and twelve blank urine sources. Carryover effect was verified by injecting three samples of blank plasma/urine, one before and

Applicability of the method
The methods were applied to population pharmacokinetics (PopPK) and TDM of GAB in patients receiving the same daily dose of GAB for at least 1 week (steady-state) 23
Stability determinations indicated that samples were stable for up to 6 h at room temperature (short-term stability) and after three cycles of freeze (-70 ºC) and thaw (room temperature). Plasma samples were stable after stored for 120 days at -70 ºC. On the other hand, urine sample did not show stability when stored for 60 days at -70 ºC. The stock and working standard solutions of GAB were stable for 120 days when stored at -20 ºC (Table 2).

Gabapentin quantification in patients with chronic pain
The method was successfully applied for the analysis of GAB in plasma samples at steadystate. Trough plasma concentrations ranged between 0.40 to 11.94 µg/mL in patients treated with 600 to 3600 mg/day, as reported previously 24 . Ten urine samples were collected in the intervals 0-12 h (n=4) and 0-8 h (n=6) after drug administration. The high interindividual variability observed in the amount of GAB eliminated unchanged in urine (Ae %) in the dose interval can be explained, at least in part, by the variable bioavailability of GAB (Figure 3).

DISCUSSION
HPLC has been the most used method to analyze GAB concentration in biological samples. Other methods as gas chromatography and capillary electrophoresis were also developed, but in a small number [25][26][27][28] . Plasma is the most frequent sample used, few methods use other biological matrices, as urine 11,12,16,25 . This study presents a full validation results of bioanalytical methods to quantify GAB in plasma and urine using HPLC-UV.
As GAB does not have a chromophore group, a derivatization reaction was required. Different derivatizing agents have been used to react with GAB for ultraviolet detection, such as 4-nitrophenylisothiocyanate, 1,2-naphthoquinone-4-sulphonic acid sodium salt (NQS); 2,4,6-trinitrobenzene sulfonic acid (TNBSA) and 1-fluoro-2,4-dinitrobenzene (FDNB) [11][12][13][14][15] . According to the literature, the reaction with FDNB is faster (10 minutes) than using other agents. Although some methods using FDNB are published, sample stability was not described 12,13 . Our results showed that samples are stable up to 24 h after the derivatization. A limitation of this study was not evaluating the urine sample stability after being stored at -70 ºC for a shorter period, since urine samples showed stability after 1 month at -84 ºC as reported by Yagi et al. 16 .
Many methods reported the recovery of GAB-derivatization agent complex 11,13,15,17,18,20,21 . However, the absolute recovery of GAB cannot be determined since no certificate standard of the product of derivatization is available 29 .
In classical pharmacokinetics studies, serial blood sampling is performed to obtain the plasma concentration versus time profile and pharmacokinetic parameters. Therefore, the use of small plasma volumes for the quantification of GAB is preferrable, since it requires small volumes of blood samples collected over time. The HPLC-UV method described in this study was validated using 200 µL of plasma, which is smaller than the used in previous methods (500 µL or 1 mL) 11,[13][14][15] . Furthermore, sample preparations are simple (protein precipitation and liquid-liquid extraction for plasma and urine samples, respectively) and lower volume of organic solvents are used if compared to methods with solid-phase extraction or liquid phase microextraction 11,12 . HPLC and UHPLC coupled to tandem mass spectrometry detector (MS/MS) methods have also been described [30][31][32][33][34][35][36][37][38] . These methods have several advantages from the analytical chemical point of view. However, even with the advance of chromatography those methods require expensive instrumentation which may not be available in many hospitals. In terms of detectability, lower values of LLOQ are not required due to the high plasma concentrations of GAB observed after therapeutic doses for chronic pain treatment.
The LLOQ determined for GAB plasma quantification (0.2 µg/mL) in this study was higher than the described in previous methods. This could be a limitation only for pharmacokinetic studies with single dose and longer sampling time. However, our results showed that the concentration ranges were adequate to determinate GAB plasma and urine concentration in all patients undergoing chronic treatment.
In conclusion, this study describes HPLC-UV bioanalytical methods to quantify GAB in human plasma and urine. Sample preparation was rapid to perform using protein precipitation and liquid-liquid extraction as a clean-up technique for plasma and urine, respectively. Samples were derivatized with FDNB, in a quick reaction, and showed stability up to 24 h after preparation. The methods are simple, selective, precise, and accurate and were successfully applied to PopPK and TDM of GAB in patients with chronic pain.