Article pubs.acs.org/ac
Monolithic Precolumns as Efficient Tools for Guiding the Design of Nanoparticulate Drug-Delivery Formulations Christina Gatschelhofer,† Agnes Prasch,† Michael R. Buchmeiser,‡ Andreas Zimmer,§ Karin Wernig,§ Martin Griesbacher,§ Thomas R. Pieber,†,⊥ and Frank M. Sinner*,†,⊥ †
Joanneum Research, HEALTH, Elisabethstrasse 5, 8010 Graz, Austria Institute of Polymer Chemistry, University of Stuttgart, Pfaffenwaldring 55, 70550 Stuttgart, Germany § Institute of Pharmaceutical Sciences, Department of Pharmaceutical Technology, Karl-Franzens-University Graz, Humboldtstrasse 46, 8010 Graz, Austria ⊥ Division of Endocrinology and Metabolism, Department of Internal Medicine, Medical University of Graz, Auenbruggerplatz 15, 8036 Graz, Austria ‡
ABSTRACT: The development of nanomedicines for improved diagnosis and treatment of diseases is pushing current analytical methods to their limits. More efficient, quantitative high-throughput screening methods are needed to guide the optimization of promising nanoparticulate drug delivery formulations. In response to this need, we present herein a novel approach using monolithic separation media. The unique porosity of our capillary monolithic precolumns allows the direct injection and online removal of protamine−oligonucleotide nanoparticles (“proticles”) without column clogging, thus avoiding the need for timeconsuming off-line sample workup. Furthermore, ring-opening metathesis polymerization (ROMP)-derived monoliths show equivalent preconcentration efficiency for the target drug vasoactive intestinal peptide (VIP) as conventional particle-packed precolumns. The performance of the ROMP-derived monolithic precolumns was constant over at least 100 injections of crude proticle-containing and 300 injections of highly acidic samples. Applying a validated LC-MS/MS capillary monolithic column switching method, we demonstrate the rapid determination of both drug load and in vitro drug release kinetics of proticles within the critical first 2 h and investigate the stability of VIP-loaded proticles in aqueous storage medium intended for inhalation therapy. n the emerging field of nanomedicine, nanosized pharmaceuticals are being designed to improve drug bioavailability, tissue targeting, and controlled drug release. Nanomedicines of the first generation have meanwhile successfully entered clinical practice, and an increasing number of new generation nanomaterials are either in or on the verge of preclinical and clinical studies.1,2 Nanomedical research strongly depends on interdisciplinary collaborations to realize the desired properties of nanomaterials for their intended clinical use. Especially in the early stages of development, the combination of medical and pharmacological expertise with that of materials science and analytical chemistry is needed to create a rational design followed by iterative cycles of characterization and redesign. The physicochemical and biological characterization of nanoparticles (NPs) plays a pivotal role in their development.
I
© XXXX American Chemical Society
The diverse properties of NPs are driving advances in analytical techniques because they often require innovative methods of characterization. An overview of current methods for nano- and microparticle characterization based on physical as well as separation techniques, e.g., laser light scattering, membrane filtration, field-flow fractionation, capillary electrokinetic, and chromatographic methods, is given by Fedotov et al.3 Alongside these approaches to NP analysis, the use of monoliths also shows great promise in this field. Since their introduction in the late 1980s, monolithic separation media have been enjoying increasing popularity as Received: May 7, 2012 Accepted: August 3, 2012
A
dx.doi.org/10.1021/ac3012208 | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
■
Article
EXPERIMENTAL SECTION Chemicals and Reagents. The chemicals and suppliers for the monolith synthesis and NP preparation were as described previously.17,15 Acetonitrile (chromasolv. grade), ethyl vinyl ether (99%), formic acid (98%), and trifluoroacetic acid (TFA, 99.5%) were purchased from Sigma−Aldrich (Vienna, Austria). Dimethyl sulfoxide (seccosolv. grade) and Krebs−Henseleit buffer components (MgSO4, KH2PO4, NaCl, CaCl2·2H2O, NaHCO3; all p.a.) were obtained from Merck (Darmstadt, Germany). KCl (p.a. ACS) was purchased from ISO Roth (Karlsruhe, Germany). The Krebs−Henseleit buffer (pH 7.4) was prepared according to a standard operation procedure of Sigma−Aldrich. Met-oxidized VIP and Cy3-fluorescencelabeled VIP were both synthesized by piCHEM (Graz, Austria). Water was purified with a Milli-Q Plus System from Millipore (Vienna, Austria). Monolith Synthesis and Characterization. Surface modification of fused-silica capillaries (Agilent, Palo Alto, CA) was performed as described previously.17 Monolith synthesis by ROMP was carried out according to a published protocol19 using [2.2.1]bicyclohept-2-ene, 1,4,4a,5,8,8a-hexahydro-1,4,5,8-exo,endo-dimethanonaphthalene, toluene, and 2propanol for the preparation of capillary monolithic precolumns (10, 10, 10, and 70%; w/w) and monolithic analytical columns (20, 20, 10, and 50%; w/w). Unless stated otherwise, the precolumn dimensions were 320 μm × 3.5 cm and those of the analytical columns were 200 μm × 15 cm. Polymerization of the monolithic supports was carried out at room temperature for 30 min and was followed by end-capping with ethyl vinyl ether and dimethyl sulfoxide (20:80; v/v) in order to deactivate and remove the initiator, respectively, as previously described.20 Electron microscopy was carried out at the Center for Electron Microscopy, Graz, Austria. The column permeability k was calculated according to Darcy’s law, k = (uηL)/Δp where u is the linear flow velocity, η is the viscosity of the solvent, L is the length, and Δp is the pressure drop across the column. Nanoparticle Preparation. Protamine−oligonucleotide NPs (proticles) were prepared as reported previously.15 Briefly, aqueous solutions of protamine, oligonucleotides, and VIP were mixed in the mass ratios of 0.6:1:0 and 0.6:1:1 to obtain unloaded and VIP-loaded proticles, respectively, which selfassembled within the first few seconds of incubation. The size of the proticles was determined for three separately prepared batches by photon correlation spectroscopy using a Malvern Zetasizer 3000 HSA (Malvern, Herrenberg, Germany). LC-UV Equipment and Methods. Evaluation of the enrichment efficiency and chromatographic behavior of the intact proticles and the individual proticle components was performed on an Ultimate HPLC system comprising an autosampler with cooled tray, a column switching unit, a capillary LC-pump, and a UV detector (Dionex/LC-Packings, Amsterdam, The Netherlands). The UV detection-cell volume was 45 nL. Data acquisition was accomplished using Chromeleon software (Version 6.40). The injection volume was 10 μL and the elution flow rate was 10 μL/min. Enrichment Efficiency. Evaluation was performed for a ROMP-derived monolithic precolumn (320 μm × 40 mm) and a Zorbax C3 particle-packed capillary precolumn (300 μm × 5 mm, 5 μm particle diameter, custom-packed by Dionex, Vienna, Austria). The Zorbax C3 packing material was obtained from Agilent. Aqueous standard levels of VIP were prepared, covering a range of 5.4−107 μg/mL. The standards were
chromatographic tools. Comprehensive reviews of the monolithic concept, including a critical comparison of their characteristics and performance with those of traditionally packed LC-columns, as well as an overview of their current status and future potential, are available in the literature.4−8 With their defining characteristic (single block of continuous porous matrix rather than of many discrete particles) and bimodal structure (micrometer-sized flow-through pores and smaller diffusive pores), monoliths present a number of advantages over conventional particle-packed columns, including higher permeability as well as enhanced separation efficiency, especially for the analysis of macromolecules9 and even large molecular assemblies. Monoliths and monolithic cryogels have already been used to separate artificial10,11 and biogenic NPs (such as plasmid DNA, viruses, phages, and intact cells) from crude product solutions up to the preparative and industrial scale,12,13 showing that they are able to serve as inexpensive and efficient tools for NP purification as well as for quality assurance and safety control. Beside the separation of NPs, the tolerance of monolithic supports for highly complex, particle-containing matrixes presents the potential for further application in the field of NP development and characterization. Quantitative characterization of drug-loaded NPs in terms of drug loading and in vitro release kinetics are, to date, still rather laborious. Selective removal of NPs by means of ultracentrifugation, dialysis, ultrafiltration, or centrifugal ultrafiltration must be performed prior to analysis of the unbound or released amount of drug.14 This need stems, on the one hand, from potential interferences of NPs at the selected wavelength in direct UV or fluorescence spectrometric detection and, on the other hand, from immediate clogging of the separation columns due to the colloidal nature of NPs, which prevents direct injection of NP solutions in LC analysis. The use of off-line sample pretreatment, however, suffers from several drawbacks: First of all, separation of NPs to get neat drug solutions is generally a rather time-consuming process, which renders the whole characterization less suitable for highthroughput screenings. Furthermore, highly adsorptive drug analytes tend to adhere to filtration membranes and vial surfaces, leading to the loss of a significant proportion of the analyte. Thus, there is a pressing need for more efficient characterization methods for drug-loaded NPs. Here, we report on the applicability of monoliths for the direct, online removal of NPs and the simultaneous selective preconcentration of the target drug. A promising early stage protamine−oligonucleotide NP formulation (proticles) loaded with vasoactive intestinal peptide (VIP) designed as pulmonary depot formulation to treat primary pulmonary hypertension15,16 was used as a model NP for this purpose. The performance of ring-opening metathesis polymerization (ROMP)-derived capillary monoliths17,18 for the enrichment of VIP was first critically compared to that of a conventional particle-packed capillary precolumn. We then investigated the use of the monolithic precolumns for online partitioning of intact NPs from the free drug for the first time. Precolumn robustness after consecutive NP injections and matrix effects has been a major issue in LC-MS/MS method development and validation. The monolithic precolumns were then used for a straightforward and fast quantitative assessment of proticle encapsulation efficiency and in vitro release kinetics. Finally, the storage stability of VIP-loaded proticles was examined. B
dx.doi.org/10.1021/ac3012208 | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
concentration of each standard level, respectively. Results from all three days were pooled to calculate interday precision. The matrix effect was qualitatively evaluated by the postcolumn infusion method as proposed by Bonfiglio et al.:22 A standard solution of the target analyte (50 μg/mL VIP) was constantly infused via a syringe pump (at 1 μL/min) and a tee-connector into the LC stream parallel to the LC-MS/MS analysis of water as blank sample and aqueous nanoparticle suspensions containing different amounts of unloaded proticles (proticle matrixes 1 and 2 correspond to proticle masses of 90 and 32 μg/mL, respectively). Quantitative assessment of the matrix effect was performed according to Matuszewski et al.:23 VIP (10 μg/mL) was prepared as an aqueous solution as well as spiked into the proticle matrixes 1 and 2. The overall process efficiency was calculated as the percentage of the VIP response obtained in proticle suspensions relative to the response in neat solution. The robustness of monolithic precolumns was tested by the injection of acidic solutions (50% TFA) and proticle matrix 1. The precolumn performance was monitored by the analysis of quality control standards at regular intervals (level 1: 20 μg/mL VIP, level 2: 10 μg/mL VIP). Nanoparticle Characterization. For the quantitative assessment of the encapsulation efficiency and in vitro release kinetics of VIP-loaded proticles, proticle suspensions were prepared in water and Krebs−Henseleit buffer (pH 7.4), respectively, at an expected concentration of 20 μg/mL VIP. Proticle suspensions (three separately prepared proticle batches) were either directly injected onto the monolithic precolumns or centrifuged for 2 h at 4 °C and 20 000 g (Eppendorf centrifuge 5804 R, Eppendorf, Germany) prior to LC-MS/MS analysis of the supernatant. The stability of the VIP-loaded proticles was tested for the duration of five days, as follows: Three series of VIP-loaded proticles were prepared, and aqueous proticle suspensions were stored in Protein LoBind tubes (Eppendorf) at 4 °C. Samples were taken at daily intervals, mixed with an equal volume of pure TFA and analyzed.
loaded onto the monolithic precolumn using 5% acetonitrile with 0.05% TFA at loading flow rates of 50, 100, and 200 μL/ min for 2, 1, and 0.5 min, respectively. VIP elution was achieved at 43% acetonitrile and 0.05% TFA at a flow rate of 10 μL/min, followed by equilibration of the precolumn at loading conditions for 2 min. The VIP signal was recorded at 210 nm. Five VIP standards covering a range of 2.7−21.4 μg/mL were used for calibration purposes. Chromatographic Behavior of Proticles. Aqueous suspensions of unloaded and VIP-loaded proticles were prepared (proticle mass 50 μg/mL). These suspensions, as well as neat protamine, oligonucleotide, and VIP standards (20−30 μg/ mL), were injected on the monolithic precolumn. Gradient elution was conducted from 5% to 72.5% acetonitrile (consistently 0.05% TFA) over 5 min and maintained for a further 5 min, followed by a decrease to start conditions over 1 min for column equilibration. Detection was carried out at 215 and 254 nm. LC-MS/MS Equipment and Methods. Method validation and characterization of proticles in terms of drug loading, in vitro release kinetics, and stability were performed on an Ultimate capillary HPLC system, as mentioned above, coupled to a Quantum TSQ Ultra AM (ThermoFinnigan). The system was controlled by Xcalibur-Software 1.4. For all studies, a sample volume of 5 μL was loaded onto the monolithic precolumn at a flow rate of 100 μL/min using 10% acetonitrile containing 0.05% TFA as loading solvent. After 2 min, the valve was switched to connect the precolumn to the analytical monolith. Elution of VIP was performed using a multistep gradient at a flow rate of 10 μL/min. The mobile phase consisted of solvent A (0.1% aqueous formic acid) and solvent B (0.1% formic acid in 80% acetonitrile). The gradient was programmed from 0 to 20% B within 0.1 min, maintained at 20% for 0.4 min, increased to 40% B over 0.4 min, maintained at 40% for 2.1 min, increased to 90% B over 2 min, maintained at 90% for 1.5 min, and decreased to 0% B over 0.5 min to equilibrate the column for 8 min. Meanwhile, at a gradient time of 8 min, the precolumn was switched back to loading position and was flushed with 80% acetonitrile for cleaning and the loading solvent for equilibration at a flow rate of 200 and 100 μL/min, respectively. Positive ESI mass spectrometry was performed using the following parameters: spray voltage 3.5 kV, capillary temperature 250 °C, sheath gas pressure 25 AU. VIP was detected as +5 charged molecular ion at m/z = 677.6 in pseudo-SRM mode applying a collision energy of 5 eV. Method Validation. Validation of the method was performed according to ICH guidelines for bioanalytical method validation.21 The performance criteria encompassed linearity, accuracy, and precision (intra- and interday). Additionally, the effect of matrix on the ion response of the analyte and the robustness of the monolithic precolumn were assessed. VIP stock solutions (400 μg/mL) were prepared in water and stored at −20 °C. Linearity was evaluated by a fivepoint calibration ranging from 4 to 20 μg/mL VIP, in five replicates. For the determination of accuracy and precision, three different validation standard levels (4, 10, and 20 μg/mL) were prepared and analyzed in three (4 and 10 μg/mL) and six (20 μg/mL) replicates. This was repeated over three days using freshly prepared standards. Accuracy and precision were calculated as the percent relative error (% RE) of the calculated concentration to the nominal concentration and as percent relative standard deviation (% RSD) of the calculated
■
RESULTS AND DISCUSSION Performance Comparison of Monolithic vs ParticlePacked Capillary Precolumns. Precolumns play a prominent role in capillary LC, as their purpose is not only to guard the analytical column from matrix components, which negatively affect separation performance over time; precolumns also permit online sample enrichment, which is essential for enhancing sensitivity for trace-level determinations. To evaluate the applicability of ROMP-derived monolithic supports17−19,24−26 for their intended use as precolumns in capillary LC, a head-to-head comparison with a conventional particle-packed capillary precolumn was conducted. A representative electron micrograph of ROMP-derived monolithic capillary precolumns that illustrates their morphology is shown in Figure 1. Based on the images of three separately prepared monoliths, an average microglobule diameter of 0.7 ± 0.1 μm was observed. The monolithic structure resulted in a rather low and linearly related back pressure, from 3 bar at a flow rate of 25 μL/min up to 31 bar at a flow of 200 μL/min. Moreover, the back pressures of the monolithic precolumns were comparable to those of Zorbax C3 packed precolumns that had the same inner diameter but were eight-times shorter. Given these results, calculation of the column permeability k of 5.6 ± 0.5 × 10−13 m2 and 6.1 ± 0.2 × 10−14 m2 for monolithic and particle-packed precolumns, C
dx.doi.org/10.1021/ac3012208 | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Moreover, the packing quality of the resulting precolumns was highly variable and, most significantly, frequent column changes were necessary because of a drop in column performance. This phenomenon may be related not only to the packing quality per se but also to the fact that particle-packed columns in the capillary format are more susceptible to deterioration in packing density caused by sudden and repeated pressure changes during use and, moreover, are more likely to become clogged by submicroscopic particulate matter in the solvents or sample than their standard counterparts.27 In this regard, ROMP-derived monoliths have superior properties for miniaturization relative to other types of monolithic materials, i.e., straightforward and reproducible in situ preparation in capillary columns, high mechanical stability of the polymeric material as well as of the covalent links to the capillary inner surface, and a porous structure that tolerates complex, colloidal sample matrixes. To demonstrate the advantages of monoliths and to gain insight into the chromatographic behavior of intact NPs on ROMP-derived monolithic supports, injection of various proticle suspensions differing in particle size was conducted. The primary, unloaded complex exhibited a hydrodynamic diameter of about 120 nm, which increased on addition of VIP and Cy3-conjugated VIP to 162 and 206 ± 12 nm, respectively (n = 3, each). As illustrated in Figure 2, all these proticle
Figure 1. Electron micrographs of a ROMP-derived monolithic capillary precolumn (320 μm i.d.). The scale bar corresponds to 200 μm (10 μm in the inset).
respectively, further confirms the higher permeability of the monolith. The enrichment efficiency of both types of precolumns for the target drug analyte was tested by loading increasing amounts of VIP onto the precolumns and by monitoring the UV-signal response in a subsequent elution step. The effect of loading flow rate on their enrichment capabilities was also examined. Data processing was performed using a five-point calibration in the range of 2.7 to 21.4 μg/mL VIP, which gave a linear response of r2 ≥ 0.990 in all cases. The results obtained with the monolithic and particle-packed precolumns are summarized in Table 1. The two precolumn types showed similar enrichment efficiencies for the target drug. Inadequate performance was observed for both when a loading flow rate of 50 μL/min was applied. Under these conditions, efficiencies were ≤95% for almost all VIP standard levels tested, even lacking preconcentration of small amounts of analyte. In contrast, VIP loading at higher flow rates yielded a maximum loading capacity of up to 200−250 μg VIP on both the monolithic and the particle-packed precolumn. Despite the comparable performances of the two columns, the experiments did reveal the limitations of the particle-packed capillary precolumns. Their commercial availability is currently still restricted to certain packing materials; i.e., the desired Zorbax C3-based capillary precolumn had to be custom prepared.
Figure 2. Chromatographic behavior of (A) VIP-loaded, (B) Cy3-VIPloaded, and (C) unloaded proticles on a ROMP-derived monolithic precolumn (320 μm × 3.5 cm). Chromatographic conditions: gradient elution from 5% to 72.5% acetonitrile (consistently 0.05% TFA) within 5 min; flow 10 μL/min.
Table 1. Determination of Enrichment Efficiency (EE) of Monolithic and Particle-Packed Precolumns at Different Loading Flow Rates loading flow rate, 50 μL/min precolumn
a
loading flow rate, 200 μL/min
calcd concn (μg/mL)
EE (%)
calcd concn (μg/mL)
EE (%)
calcd concna (μg/mL)
EE (%)
5.4 10.7 21.4 26.8 53.5 107 5.4 10.7 21.4 26.8 53.5 107
3.82 9.62 18.64 19.18 18.43 21.41 4.22 10.85 20.91 21.23 31.64 42.58
70.8 89.9 87.1 71.6 34.5 20.0 78.2 101.4 97.7 79.2 59.1 39.8
5.34 11.35 21.40 24.60 28.94 28.19 5.28 10.91 21.80 24.64 35.82 43.35
98.8 106.0 100.0 91.8 54.1 26.3 97.7 101.9 101.9 91.9 67.0 40.5
5.87 11.01 20.77 24.96 26.10 26.13 5.23 10.96 20.92 22.74 36.90 47.71
108.7 102.9 97.1 93.1 48.8 24.4 96.9 102.4 97.8 84.9 69.0 44.6
particle-packed
a
loading flow rate, 100 μL/min
standard concn (μg/mL)
monolithic
a
Mean of double injection. D
dx.doi.org/10.1021/ac3012208 | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
formulations passed directly through the precolumn within the injection peak ranging from 0.75 to 2.25 min, at a flow rate of 10 μL/min. Although proticles showed no retention on the monolithic phase, various passing times were observed. These do not correlate with the proticles’ average hydrodynamic diameters and were furthermore detected even within single proticle formulation types with a rather small size distribution. Whether these are shape-dependent effects will be further investigated. The retention profiles of the individual NP components, i.e., protamine, oligonucleotides, and VIP, were subsequently examined. As shown in Figure 3B, protamine also
Table 2. Validation Results for Accuracy and Intraday Precision of Vasoactive Intestinal Peptide day 1 2 3 1 2 3 1 2 3 a
specified concn (μg/mL)
mean calcd concn (μg/mL)
accuracy (% RE)
intraday precision (% RSD)
higha
20.0
mediumb
10.0
19.6 19.0 18.9 9.2 9.8 10.6 3.8 4.0 4.1
−1.9 −5.3 −5.3 −8.2 −1.9 6.2 −5.1 1.2 1.5
2.9 1.5 2.1 1.5 1.1 2.2 0.3 2.9 1.0
level
lowb
4.0
n = 6. bn = 3.
interferences during the ionization process. Here the postcolumn infusion method22 was used to qualitatively identify chromatographic regions most likely to experience matrix effects caused by incomplete online removal of proticles. Figure 4 shows the comparison of the constantly infused VIP response between the injection of a blank solution and NP suspensions containing different amounts of unloaded proticles using the postcolumn infusion technique. No significant influence of the proticles on the analytical responses around the retention time of VIP at 7 min, or indeed during the entire chromatographic run, was observed. Further experiments were conducted to evaluate whether the presence of NPs influences VIP enrichment on, or VIP elution from, the precolumn. Quantitative determination of the overall process efficiency (sum of matrix effect and recovery) was carried out as proposed by Matuszewski et al.23 Excellent results (within 95−105%) were obtained by comparing the VIP responses in neat solution standards with the responses in VIPspiked proticle suspensions of three separately prepared and analyzed sample batches. The efficiencies were independent of the proticle amounts, with 99.5 ± 4.4 and 99.6 ± 3.3% obtained for proticle matrixes 1 and 2, respectively. Furthermore, we investigated the robustness of the monolithic precolumns in order to test their performance in analyzing VIP levels in crude particle-containing and highly acidic samples. To monitor alterations in analytical performance, consecutive injections of aqueous proticle suspensions and acidic solutions were performed while monitoring the back pressure (Figure 5). Relative to the constant response seen on the injection of neat aqueous standards, consecutive injections of proticle matrix 1 led to a gradual increase of about 0.08 bar per injection. The increase was greater over the last 20 proticle injections, resulting in an overall increase from 33 to 43 bar over the whole testing period of 100 proticle injections. Note that the proticle matrix applied in these experiments was 2.5fold higher than in the NP characterization studies in order to enhance potential effects. Consecutive injections of acidic samples composed of 50% TFA yielded no significant changes in back pressure up to 200 injections. In the final 100 injections, however, a steady increase of 3 bar per injection was observed. The implications of these observations on the chromatographic and analytical performance of our method were examined by regularly analyzing two levels of quality control standards. The data of chromatographic peak parameters in terms of retention time, peak width at half height and symmetry, and the accuracy of the results are representatively listed for quality control
Figure 3. Retention profile of the individual proticle components (A) VIP, (B) protamine, and (C) oligonucleotides on a ROMP-derived monolithic precolumn (320 μm × 3.5 cm). Chromatographic conditions: gradient elution from 5% to 72.5% acetonitrile (consistently 0.05% TFA) within 5 min; flow 10 μL/min.
exhibited no retention, with its UV signal partly suppressed by the negative injection response, whereas the retained amounts of oligonucleotides and VIP were well separated (Figure 3A,C), even on the monolithic precolumn, with a steep gradient. LC-MS/MS Method Development and Validation. A capillary LC-MS/MS column switching method was developed and optimized for the selective enrichment, retention, and detection of VIP. Moreover, the method was custom-designed for in vitro release studies to cover the range of a 20−100% release of VIP and was critically tested, according to ICH guidelines for bioanalytical method validation in terms of accuracy and precision, by setting acceptance criteria to ±10% and
