2.1 Chemical and reagent

Analytically pure (99.9%) reference standard of Avanafil (AV) was received as a gratis sample from Sunrise Remedies Pvt., Ltd. (Ahmedabad, India) and Om laboratory (Ahmedabad) along with a certificate of analysis. The marketed tablet formulation of Avanafil (Avana) was procured as a gratis sample from Sunrise Remedies Pvt., Ltd. (Ahmedabad, India). HPLC grade acetonitrile (ACN) and glacial acetic acid were procured from Merck, Mumbai, India. Ammonium acetate of HPLC grade was purchased from Finar Chemicals Pvt. Ltd. (Ahmedabad, India). HPLC grade water (triple distilled water) was prepared in the laboratory by distillation assembly. Hydrochloride acid, sodium hydroxide and hydrogen peroxide (30% w/v) used for stress degradation studies were of analytical grade procured from CDH chemicals, Delhi, India.

2.2 Apparatus and instrumentation

pH meter (Accurate scientific, Ahmedabad) was used to check the pH of all solutions. Other equipment used for studies were melting point apparatus (T060160, EIE Instruments Pvt. Ltd., Ahmedabad, India), precision analytical balance (CX 220, citizen, USA), sonicator (D-compact, EIE Instrument Pvt. Ltd., Ahmedabad, India) and variable micropipette (Axiva Sichem Biotech, Delhi). Precision water bath equipped with MV controller (EIE 406, EIE, Instrument Pvt. Ltd, Ahmedabad, India) was used for degradation studies in the solution state. Fourier transform infrared spectrometer (FTIR) (JASCO FT/IR − 6100 series, Jasco, Japan) was used for the identification of functional groups. The membrane filters 0.22 µm (Axiva sichem biotech, Delhi), and syringe filters 0.45 µm were used for filtration.

The Agilent 1260 Infinity HPLC (Agilent Technologies, USA) equipped with an autosampler, a quaternary pump, a UV detector, and a column compartment were used during the study. Jasco HPLC (Jasco analytical instruments, USA) equipped with PDA (Photodiode array) detector was used for peak purity calculation. LC column Nucleosil C18 (250 × 4.6 mm, 5 µm) was used for better separation. LC-MS/MS studies were carried out using MS (LTQ XL™ Linear ion trap mass spectrometer), in which the LC part consisted of autosampler HPLC and MS system consisted of LCQ Fleet Ion Trap LC/MSⁿ (Thermo Scientific, USA). The Xcalibur software used for the data processing. For MSⁿ experiments, the LTQ-XL-MS instrument (Thermo Scientific, USA) was used. The mass spectra were acquired and processed using Xcalibur software. The mass parameters used were mass range, 50–2000 amu; ion source (IS), 4500 v; declustering potential (DP), 20 V; entrance potential (EP), 10 V; nebulizer gas as air at a pressure of 25 psi and curtain gas as nitrogen at a pressure of 35 psi.

2.3 Chromatographic conditions

The AV stability indicating method was developed based on preliminary trials and drug physicochemical properties. Initially, ACN with water in different ratios was tried. The better separation achieved by the addition of buffer and the buffer pH should be selected in the range of ±2 of pKa of the drug. AV has pKa value 5.54 hence, the buffer pH was set between ±2 of pKa; as in that range nearly 100% of the drug is ionized or unionized which is suitable for better seperation. A further change in the organic phase ratio affects degradation peaks retention time and column temperature change can affect on elution time due to change in viscosity of the mobile phase with temerature. After several trials isocratic, novel and specific method was optimized; this fulfills all criteria according to ICH guideline. The chromatographic conditions were optimized using Hypersil, C18 (250 × 4.6 mm, 5 µm) column with a mobile phase composed of 10 mM ammonium acetate buffer (A) (pH 4.5, adjusted with acetic acid): ACN (B) (A: B, 60:40, v/v) at flow rate of 0.9 mL min⁻¹. The column temperature, injection volume, and detector wavelength were at 20 °C, 10 µL and 239 nm, respectively. The optimized and fully validated chromatographic conditions were MS compatible hence further subjected to LC-MS studies.

2.4 Mass spectrometric conditions

LC-MS: The mass spectrometer conditions were optimized by a preliminary direct infusion of standard drug solution (10 µg mL⁻¹) into the ESI-MS system at a flow rate of 10 µL hr^-1, and all ion optics were optimized. General MS adjustments were set as follows: drying gas flow, 12 L min⁻¹; capillary voltage, 3500 V; desolvation temperature, 350 °C; nebulizer, 36 psi; ion change control (ICC) smart target, 150000; collision gas Ar at 20 Da collision energy (CE) [if requires increments of 5% until the intensity of the precursor ion is less than 5% of the intensity of the product ions] and max accumulation time, 150 ms used throughout the analysis. Initially, the precursor ion (m + 1) values were obtained using LC-MS studies with respect to the retention time of parent drug and it’s all D.P.s. Subsequently, the MS/MS study was performed for each D.P.s to determine product ion and its fragmentation pattern.

2.5 Stress studies for generation of degradation products

Stress degradation was carried out in different conditions like as hydrolysis (acidic, basic and neutral), oxidative, photolytic, and thermal according to the recommendations in ICH Q1A (R2). The stress studies were carried out by preparing the AV solution of 1 mg mL⁻¹ in the respective stressor. Acetonitrile (2 mL) was added to dissolve 10 mg drug and then the volume was made up to 10 mL with respective stressors. From that, 0.1 mL of sample was withdrawn at suitable time intervals and neutralized with acid/alkali and/or diluted up to 1 mL with ACN: H₂O (50:50, v/v) to get the concentration (100 µg mL⁻¹) then injected in HPLC.

Different degradation conditions as the molar concentration of stressor, temperature, and stress duration were selected based on initial trial results, which reveals the formation of primary D.P.s. Different degradation conditions Hydrolytic (acidic and alkaline) degradation study was carried out by refluxing the drug in 1 N concentrations of HCl and NaOH at different temperatures as 60, 70, and 80 °C, while neutral hydrolysis was performed in an equal mixture of water and ACN. For oxidative degradation study, the drug was subjected to 1%, 6%, and 10% peroxide at room temperature. Thermal stress degradation was carried out by exposing the solid drug as a thin layer in the petri-dish for 7 hr and as in solution form, an equal mixture of water and ACN was used and kept for 5 hr in an oven at 60 °C. Photolytic degradation study was carried out in the solid drug for 8 hr as well as drug solutions (an equal mixture of water and ACN) for 6 hr in direct sunlight. For Photodegradation comparison, a solid sample was placed in a photostability chamber at 1.2 million Lux hr for 6 hr. All stressed samples were kept in the amber color vial at the refrigerator until analysis.

The marketed formulation was also subjected to all degradation conditions. All stressed samples were collected, neutralized, and filtered through a 0.22 µm pore size membrane syringe filter before LC and LC-MS/MS analysis.

2.6 In silico studies

In silico study was carried out for the impurities for the prediction of physicochemical parameters like cLogP, Solubility (LogS), topological polar surface area (TPSA). Prediction of toxicity risk (in silico) was also carried out for the toxicity factors like mutagenicity, tumorigenicity, irritant, and reproductive effects using OSIRIS property explorer which uses Chou and Jurs algorithm, based on computed atom contributions. This software allows to underline properties with high risks of undesired effects such as mutagenicity, poor intestinal absorption, or drug conform behavior via different customization features. The OSIRIS property explorer calculates on-the-fly various drug-relevant properties whenever a structure is valid. Prediction results are valued and color-coded. Properties with high risks of undesired effects like mutagenicity or poor intestinal absorption are shown in red. Whereas a green color indicates drug-conform behavior. A working Java Runtime Environment (JRE) must be installed on the computer.

3.1 LC separation of the drug and degradation products

The SIAM was developed and an optimized chromatographic condition (as described in Section 2.3) was used for hydrolysis, oxidative, photolysis, and thermal degradation conditions. A total sixteen D.P.s were found during different degradation conditions as shown in overlay chromatogram Fig. 1. The degradation results reveals that % degradation of AV found 21.71–25.62% in acidic condition; 20.00–25.00% in alkaline condition; 14.00–19.21% in neutral condition; 16.25–30.52% in oxidative hydrolysis; 22.39–28.96% in photolytic and 28.00–28.50% in thermal degradation. The drug (AV) found sensitive towards thermal, photolytic, oxidative, and hydrolytic conditions according to the AV stress degradation study (Table 1). The degradation was confirmed from the additional peaks achieved in the respective chromatogram and also by the reduction in the area of AV in each condition when compared with their area at zero time. The mass balance was calculated from the AV and all the D.P.s responses obtained after stress studies. Mass balance found greater than 90 for alkaline, photolytic, and thermal degradation conditions. While for acid and oxidative degradation it was 89 and 82 respectively, due to loss of some mass during degradation as seen in Table 1.

Close

Fig. 1

Overlay chromatogram of AV (100 µg/mL) in (a) acid hydrolysis, alkaline hydrolysis and peroxide (b) photolytic and thermal degradation in the optimized chromatographic condition.

Export to PPT

Amongst all the degradation conditions, the order of degradation behavior (based on average % degradation in each condition) observed was

Thermal > Photolytic > Oxidative > Hydrolytic (acid, alkaline and neutral)

The optimized LC method was compatible with LC-MS, hence subjected to LC-MS/MS study for AV degradation products. In different degradation conditions (acidic, alkaline, oxidative, photolytic, and thermal), various D.P.s were found. Nomenclature of formed D.P.s in different condition were given based on their Rt (retention time). Found D.P.s in LC chromatogram for the acid condition D.P. I, III, IV, VIII, XI, XIII and XIV; in alkaline condition D.P. I, III, IV, VIII, XI, XIII, XIV and XV; in neutral degradation of the drug formation of D.P.s I, III, IV, VIII, XIII and XIV; in oxidative degradation of the drug formation of D.P.s VI, XII, XIII, and XIV; in photolytic degradation D.P. II, V, VII, X, XIII, XIV and XVI and in thermal degradation D.P. II, V, IX, XIII, and XIV as shown in Fig. 1, Table 2. For additional confirmation of LC degradation peaks by LC-MS/MS peaks, RRT (relative retention time) and PDA (Photodiode array) data were used.

The degradation pattern of AV in photolytic and thermal conditions was found similar in liquid state and solid-state. Common D.P.s for acid and alkaline hydrolysis are I, III, IV, VIII, XI, XIII and XIV. Common D.P.s between acid, alkaline and neutral degradation are I, III, IV, VIII, XIII and XIV. At last for photolytic and thermal degradation found common D.P.s are II, V, XIII and XIV. Besides, AV XIII and AV XIV were common D.P.s in all conditions. The AV sixteen D.P.s were identified and characterized by LC-MS/MS and fragmentation pattern. The D.P.s structure were proposed based on MS/MS data and MSⁿ fragmentation.

3.2 Mass fragmentation behavior of AV

AV Fragmentation by MS/MS studies showed up to MS⁵. AV was analyzed in positive ion mode by direct flow injection using ACN: water (1:1, v/v) mixture as the solvent through MSⁿ spectra to determine the origin of each fragment. Line spectra of drugs, obtained in MS and MSⁿ studies are shown in Supplementary Fig S1. The best possible molecular formula and exact masses for the fragments were calculated using an elemental composition calculator. The same are listed in Table 2.

The MS/MS spectrum of AV is as shown in Supplementary Fig S1. AV having a molecular formula of C₂₃H₂₆ClN₇O₃ (m/z 483.95). The fragment of AV (m/z 483) with m/z 484.22 was formed as an M + 1 peak, the fragment m/z 485.97 formed due to M + 2 peak and the fragment with m/z 506.14 was formed as a sodium adduct since its mass was approximately 1 Da (Da), 2 Da (Da) and 23 Da (Da) higher than the molecular ion peak. The mass peak of AV shows chloro pattern (1:3) in spectra which confirms the presence of chlorine in structure. MS/MS spectra shows that AV m/z 484.22 further fragment in to m/z 375.09, m/z 357.10, m/z 342, m/z 329.14 and m/z 155.12. From the MS/MS spectra, a postulated fragmentation pattern was derived as shown in Fig. 2. From drug AV, MS² fragment m/z 375.09 was formed due to the loss of C₅H₇N₃ (Diamino pyridine). Further, it fragments into MS³ m/z 357.1 due to loss of H₂O (water) molecule and MS⁴ m/z 342.00 was formed due to removal of the methyl group. Then AV fragments into MS⁵ m/z 329.14 due to loss of - COH. At last due to cleavage of a single bond from m/z 329.14 structure it converted in to m/z 155.12. The MS/MS spectra reveals that m/z 155.12 is common to the last fragment of AV, having a molecular formula of C₈H₈ClO and chemical name of 2-chloro-1-methoxy-4-methyl-benzene. The best possible molecular formula and exact masses for the fragments were calculated as shown in Table 2.

Close

Fig. 2

Proposed fragmentation pattern of AV.

Export to PPT

3.3 Characterization of the impurities and degradation products using mass data

All the D.P.s were characterized based on LC-MS/MS. The line spectra of all the D.P.s were taken in ESI positive (ES+) mode. Accurate mass data of the molecular ions, their fragments, and accurate masses were listed in Table 2. Degradation peaks generated in all conditions were numbered as AV I to AV XVI and structures were predicted as shown in Fig. 3. They were characterized and identified from mass data. The data obtained in LC-MS and MS/MS studies were systematically utilized for the structure elucidation of D.P.s of AV as described below.

Close

Fig. 3

Structure of AV and proposed structures of its degradation products.

Export to PPT

3.4 Postulated mechanistic pathway for the formation of D.P.s

Based on the data obtained from the line spectra of LC-MS/MS studies of each degradation product, the proposed degradation pathway of AV in different degradation conditions were established. The degradation pathways for AV in (i) acid, (ii) alkaline, (iii) photolytic, (iv) thermal and (v) oxidative degradation conditions were shown as Fig. 4, Fig. 5, Fig. 6, Fig. 7 and Fig. 8.

Close

Fig. 4

Degradation mechanism of AV in acidic condition.

Export to PPT

Close

Fig. 5

Degradation mechanism of AV in alkaline condition.

Export to PPT

Close

Fig. 6

Degradation mechanism of AV in photolytic condition.

Export to PPT

Close

Fig. 7

Degradation mechanism of AV in thermal condition.

Export to PPT

Close

Fig. 8

Degradation mechanism of AV in oxidative condition.

Export to PPT

In acid hydrolysis, AV XIV form due to cleavage of –CH bond between carbonyl amide and pyrimidine ring from the structure of AV. Formation of AV VIII from the structure of AV explained by nucleophilic substitution reaction and formation of aryl carboxylic acid from aryl amide with loss of amino pyrimidine. From the AV XIV structure loss of ‘CONH’ functional group leads to the formation of the AV XI degradation product. AV IV degradation product form due to dechlorination from the structure of AV XI. AV III degradation product form in the presence of acid hydrolysis by removal of pyrrolidine ring by – OH (hydroxy) functional group. Formation of AV I and AV XIII can be explained by dechlorination from the AV structure. Both degradation product having the same molecular mass but position of - OH functional group is different and further fragments are different, confirms that AV XIII is R isomeric form of AV I. (Fig. 4)

In the case of alkaline hydrolysis, the formation of AV I from AV can be explained by the removal of chlorine functional group. The formation of AV VIII is explained by nucleophilic substitution reaction due to OH^– ion abstracts the proton followed by hydrolysis. AV XIV form due to hydrolysis of AV and removal of the pyrimidine ring. The formation of AV XI is explained by the removal of ‘CONH’ functional group from the structure of AV and AV IV is due to dechlorination from the structure XIV. AV XV is identified as acetonitrile adduct formation with AV (Fig. 5).

In the case of photolytic degradation condition formation of AV related substance is identified as AV XVI, due to the presence of free radical and hydrolysis. AV X formation can explained by demethylation from the AV structure. The formation of AV VII can explained by the opening of the pyrrolidine ring in the presence of free radical and further attachment of the hydroxyl functional group on it, which involved a radical cation mechanism. AV V and AV II form due to attachment of - OH functional group in the AV structure (Fig. 6).

In the case of thermal hydrolysis, AV IX formation can explained by the removal of dimethoxy benzyl from the AV structure due to energy and bond cleavage (Fig. 7).

In oxidative degradation formation of AV XII is due to oxidation by peroxide and AV VI is due to the addition of chlorine functional group in AV structure (Fig. 8).

3.5 In silico studies

In silico physicochemical parameters and toxicity risk assessment study (prediction) was carried out to get an idea about the physicochemical properties and any probable toxicity due to the impurities. Factors of toxicity risk assessment like mutagenicity, tumorigenicity, irritant and reproductive effects (Table 3) relies on a computed set of structural fragment that give rise to toxicity alerts in case they are encountered in the structure of impurity. These fragment lists were created by rigorously shredding all compounds of the RTECS database known to be active in a certain toxicity class (e.g. mutagenicity). During the shredding, any impurity was first cut at every rotatable bond leading to a set of core fragments. These in turn were used to reconstruct all possible bigger fragments being a substructure of the original impurity. Afterward, a substructure search process determined the occurrence frequency of any fragment (core and constructed fragments) within all compounds of that toxicity class. It also determined these fragment's frequencies within the structures of more than 3000 traded drugs. Toxicity risk alerts are an indication that the impurity may be harmful concerning the risk category specified (toxic, moderately toxic, and non-toxic). Results of toxicity prediction study revealed that no impurity is at any risk of toxicity. Prediction of physicochemical properties revealed that all the physicochemical parameters of impurities were within the acceptable limit.

3.6 Method validation (ICH, 2005b)

The developed stability-indicating method was validated as per ICH guidelines Q2 (R1) and results shown as per Table 4 (i). The system suitability of the developed method was checked for 10 µg mL⁻¹ by evaluating % RSD values of peak area, asymmetry, theoretical plate, and retention time of five standard replicates. % RSD values were found less than 2, for all parameters as shown in Table 4 (ii). The regression coefficient was found to be 0.9970 for AV in the range of 10–70 µg mL⁻¹ which confirms the linearity of the method. All degradation product peaks having resolution (>2) and peak purity from 0.990 − 0.999 indicating peaks were well separated. LOD and LOQ values were found to be 0.36 µg mL⁻¹ and 1.15 µg mL⁻¹, respectively. The interday and intraday precision was evaluated at three different concentrations, i.e. 10, 30, and 50 µg mL⁻¹ (each in triplicate) on consecutive three days and the same day respectively. The % RSD values for interday and intraday was found to be 1.680–1.800 and 0.326–0.697, respectively. Accuracy of AV was performed by using a standard sample at three different levels 80%, 100%, and 120% (54, 60, 66 µg mL⁻¹) respectively. % Recovery ± RSD was found to be 99.280 ± 0.239–101.420 ± 0.092. % Assay ± RSD was found to be (99.430–102.040) ± 0.997. Hence the developed method can be further used for estimation of AV in tablet formulation without the interference of excipients. The similar forced degradation conditions were applied for marketed formulations and found alike method sensitivity and specificity towards APIs, degradation products, and impurities.

3.7 Marketed formulation degradation

The developed LC-MS/MS method was applied for marketed tablet formulation (AVANA). Avana tablet was subjected to acid, alkaline, photolytic, thermal, and oxidative degradation. Overlay chromatogram of AV and its marketed formulation in different stress condition (Supplementary scheme S7) shows a similar pattern to that of API. The peaks were also confirmed through LC-MS/MS and RRT. Hence developed method is useful for the determination of drug from the formulation and can extend up to toxicity study.