In Vitro-In Vivo

1 downloads 0 Views 173KB Size Report
macodinámico-tiempo simulado, reforzando por lo tanto la experiencia clínica de su ineficacia observada por los .... ear regression between ka and kd, obtained .... cia.gub.uy/_Web/decretos/2007/01/266_15%. 2009%202006_00001.PDF>.
Latin American Journal of Pharmacy (formerly Acta Farmacéutica Bonaerense)

Short communication Received: October 24, 2013 Revised version: December 3, 2013 Accepted: December 6, 2013

Lat. Am. J. Pharm. 33 (1): 166-70 (2014)

Simulated Study of Therapeutic Equivalence for Flunitrazepam Tablets: In Vitro-In Vivo Correlation from Bibliographic-Experimental Data Laura MAGALLANES, Ismael OLMOS, Manuel IBARRA, Cecilia MALDONADO, Marta VÁZQUEZ & Pietro FAGIOLINO* Pharmaceutical Sciences Department – Faculty of Chemistry & Bioavailability and Bioequivalence Center for Medicine Evaluation (CEBIOBE), Universidad de la República – Uruguay.

SUMMARY. A simulated therapeutic equivalence scenario was created between two multisource drug products (tablets marketed in Uruguay) containing a poorly water-soluble compound, flunitrazepam, and the reference product (Rohypnol®). The simulated study was built from an in vitro - in vivo dissolutionpharmacokinetic C-level (IVIV) correlation and from an in vivo pharmacokinetic-pharmacodynamic (PKPD) correlation, using experimental and bibliographic data. In vitro parameters were experimentally obtained from the dissolution profile for each commercial brand at the three biorelevant dissolution media (pH 1.2, 4.5 and 6.8) whereas in vivo parameters were acquired from the bibliography. From IVIV correlation plasma drug concentration-time profiles were simulated. These concentrations were used to simulate the intensity of drug effect throughout time in order to predict relative clinical responses of each brand. It could be concluded that from the three brands of flunitrazepam marketed in Uruguay, one was revealed as having a very poor sedative effect based on its simulated low pharmacodynamic-time profile, reinforcing hence the ineffective clinical experience claimed by physicians. RESUMEN. Se ha creado un escenario de equivalencia terapéutica simulada entre dos productos farmacéuticos de origen múltiple (comprimidos comercializados en Uruguay) que contienen un compuesto poco soluble en agua, flunitrazepam, y el producto de referencia (Rohypnol®). El estudio simulado fue construido a partir de una correlación in vitro-in vivo disolución farmacocinética nivel-C ( IVIV ) y de una correlación in vivo farmacocinética-farmacodinámica (PKPD), utilizando datos experimentales y bibliográficos. Los parámetros in vitro fueron obtenidos experimentalmente a partir del perfil de disolución para cada marca comercial en tres medios de disolución biorrelevantes (pH 1,2, 4,5 y 6,8), mientras que los parámetros in vivo fueron adquiridos de la bibliografía. A partir de la correlación IVIV se simularon los perfiles de concentración-tiempo del fármaco en plasma. Estas concentraciones se utilizaron para simular la intensidad del efecto del fármaco a través del tiempo con el fin de predecir las respuestas clínicas relativas de cada marca. Se podría concluir que de las tres marcas de flunitrazepam que se comercializan en Uruguay, una demostró poseer muy mal efecto sedante en base a su bajo perfil farmacodinámico-tiempo simulado, reforzando por lo tanto la experiencia clínica de su ineficacia observada por los médicos.

INTRODUCTION Flunitrazepam is a 7-nitro-benzodiazepine drug, mainly used as a hypnotic and sedative night-time drug, and as anesthetics. As a hypnotic for insomnia, the usual dose is 1 to 2 mg at night, and its response is characterized by a very fast onset. Its pharmacokinetics was well characterized in human subjects after oral and intravenous administrations 1,2. According to the Biopharmaceutic Classification System, Flunitrazepam would be classified as Class II 3,4.

Different brands are marketed in Uruguay, including the innovator: Rohypnol ® . Multisource drug products have not demonstrated to be bioequivalent with the original brand yet. This is a common situation in our country since pharmaceutical regulation of interchangeability has been issued in 2007 5. Then, non-documented speculation was made over the effectiveness of the copy brands. For this reason, it seemed to be interesting to construct a virtual comparison of bioavailabilities over a group of individuals

KEY WORDS: Flunitrazepam, In vitro-in vivo correlation simulated therapeutic equivalence. *

166

Author to whom correspondence should be addressed. E-mail: [email protected]

ISSN 0326-2383

Latin American Journal of Pharmacy - 33 (1) - 2014

that actually have participated in a bioequivalence study, receiving as reference product the one that is commercialized in our country by F.Hoffman–La Roche Ltd. as well. In order to do this, it would be necessary to establish an in vivo–in vitro (IVIV) correlation that supports relative bioavailabilities between products. The second step would be to transfer the pharmacokinetic data to clinical responses by using a pharmacokinetic–pharmacodynamic (PKPD) correlation established in the international scientific literature. Research groups are able to put into practice the universally accepted and standardized in vitro dissolution procedures according to WHO recommendations 6. So, in vitro dissolution performance of pharmaceuticals should not be affected wherever they were tested. Under this premise, locally obtained in vitro dissolution data will have a single in vivo pharmacokinetic response in subjects that underwent a previous bioequivalence study. In the case that these subjects shared similar characteristics with local population, regarding their ethnicity, habits, and environmental conditions, local bioequivalence studies could be subrogated and therapeutic / toxic response could be foreseen. The reference product would then play not only the role of being the formulation that allows a global comparison of multisource drug products, as it is currently expected through in vivo bioequivalence studies performed all over the world, but because of its pharmacokinetics and clinical response already assessed, it could also become the link for transferring in vitro dissolution data to clinical performances of multisource drug formulations. We think this approach could contribute significantly with the hard task our societies are carrying out in order to complete the bioequivalence requirements of multisource drug products, especially for those medicines having low therapeutic ranges and thereafter with a high risk for the population health. This work focuses on the field of benzodiazepines used as sedative and hypnotic agents as flunitrazepam. MATERIALS AND METHODS Formulation marketed in Uruguay and assayed in vitro The analyzed products were Rohypnol® 1 mg tablets (Roche International, Lot no. RJ0420, (Reference), Inervon® 2 mg tablets (Caillon & Hamonet S.A.C.I., Lot no 66) (Test 1) and Somnidual® 2 mg tablets (Laboratorio Athena S.A., Lot no. 98674) (Test 2).

In vitro dissolution testing Six units of each Test product and twelve units of the Reference product (two units per vessel) were assayed in Distek® dissolution system 2100C equipment. The in vitro dissolution was performed at the following media: Medium 1 (HCl/KCl pH 1.2 solution), medium 2 (acetate buffer pH 4.5), and medium 3 (phosphate buffer pH 6.8). The solutions were prepared according to United State Pharmacopoeia (USP) with 0.5% of Polysorbate 80 (Tween 80) v/v. The dissolution conditions were: USP Apparatus 2; temperature: 37 ± 0.5 °C; volume: 900 mL; stirring speed: 75 rpm; sampling times: 10, 20, 30, 40, and 60 min; volume of sample: not more than 2 mL each. The quantification of flunitrazepam concentration at each sampling time was carried out by a high performance liquid chromatographic method (HPLC). Dionex® system Ultimate 3000 Series equipment was used. Filtered samples were injected (20 µL) into the chromatograph under the following experimental conditions. Stationary phase: C18 (150 × 4.6 mm; 5 µm) column (Phenomenex® Luna); column oven temperature: 37 °C; mobile phase: phosphate buffer 45 mM pH 3.0 / Acetonitrile (60 / 40); flow rate: 1.5 mL/min; wavelength: 220 nm. Once the percentage of drug dissolved throughout time was determined, first order dissolution rate constants (kd) were calculated for each product at the three dissolution media. Bibliographic material Grahnén et al. 1 conducted an in vivo–randomized–balanced-single dose–crossover bioequivalence study between two brands of flunitrazepam with twenty Caucasian healthy subjects, establishing a relationship between the pharmacokinetics of flunitrazepam and its sedative effects. The Reference product corresponded with the one marketed in Uruguay. The Test product presented higher maximum plasma concentration (C max) and lower time-to-peak (T max ) values than the Reference. As flunitrazepam belongs to the BCS class II, in which the limiting factor for absorption is the dissolution rate, it would be possible to assume that the Test dissolved the drug more rapidly than the Reference. This in vivo Test/Reference performance ratio was similar to the in vitro observation at the three dissolution media for our local T1-product/Reference ratio. Then, our local T1 in vitro dissolution data was assumed to be displayed by the in vivo Test formulation assayed abroad. Hence, two formulations, Test or 167

MAGALLANES L., OLMOS I., IBARRA M., MALDONADO C., VÁZQUEZ M. & FAGIOLINO P.

T1 and Reference, were the minimum required to make IVIV correlation. As a pharmacodynamic parameter, Grahnén et al. 1 ranked the degree of sedation, according to the following five-grade scale: (0) no sedation, (1) slight sedation (subjects able to read) (2) moderate sedation (difficult to concentrate), (3) severe sedation (subjects unable to cope with any intellectual activity), and (4) sleeping (subjects not awake during sampling) Sedation in relation to plasma concentration was evaluated using the sigmoid Emax model (Hill equation) without an effect compartment according to Holford & Sheiner 7. Determination of in vivo pharmacokinetic parameters Pharmacokinetic analysis of the bibliographic concentration-time data previously mentioned was carried out considering a linear bi-compartmental model with first order absorption rate constant (ka). The ka values for reference and test were determined by the method of residuals. Search for in vivo (PKPD) – in vitro correlation The dissolution medium, where the experimental kdT1/kdR ratio was similar to the bibliographic calculated kaT/kaR ratio, was investigated. Once this similarity was determined, the linear regression between k a and k d, obtained from T or T1 and R values, was calculated and then used to infer ka for T2 from its correspond-

ing experimental kd. In this way, simulated plasma concentration profile and thereafter pharmacodynamic response throughout time for T2 was assessed. RESULTS AND DISCUSSION Mean dissolution profiles of formulations are shown in Fig. 1. T2 product dissolution is slower than reference (R) and T1 at the three dissolution media. Differences are more pronounced in media 2 and 3 (pHs 4.5 and 6.8) than in medium 1 (pH 1.2). As flunitrazepam presents low water solubility and high permeability, the main determinant of drug absorption would be the dissolution rate. In medium 1, simulating stomach pH, dissolution constant was calculated for the 0-60 min interval, since this time would be the maximum residence time in the stomach under fasting administration. In the case of medium 2 (duodenal pH) and medium 3 (jejunum pH), the dissolution constants were calculated in the time range of 0-30 minutes because more than 85% of dissolution was achieved for the reference in that time interval (Fig. 1). Calculated dissolution rate constants can be seen in Table 1. T1/R ratios of dissolution rate constants coming from the dissolution media 1 and 2, which represent the first portion of the gastrointestinal tract, were assayed considering these formulations as immediate release products. Table 2 shows that Kd(T1)/Kd(R) ratio obtained at pH

Figure 1. Dissolution profiles at pH 1.2 (medium 1), pH 4.5 (medium 2), and pH 6.8 (medium 3).

Product

kd (min–1) - pH 1.2 (0-60 min)

kd (min–1) - pH 4.5 (0-30 min)

kd (min–1) - pH 6.8 (0-30 min)

R T1 T2

5.33 7.01 3.62

4.94 6.36 2.23

4.56 4.55 1,98

Table 1. Dissolution rate constants (kd) determined at the three dissolution media.

168

Latin American Journal of Pharmacy - 33 (1) - 2014

A

B

Figure 2. Plasma concentration–time profiles actually determined (T1, R) or simulated (T2) for the three evaluated products, based on IVIV correlation found with medium 1: pH 1.2 (A) or with medium 2: pH 4.5 (B).

1.2 and pH 4.5 media is similar to ka(T) / ka(R) ratio. Fig. 2 shows real plasma concentration profile for formulation T or T1 and R and the simulated profile for T2, when the absorption rate constant was interpolated from the linear regression obtained at pH 1.2 (A) or at pH 4.5 (B). Product T2 displayed a considerably lower Cmax value in comparison to the other two products when IVIV correlation was obtained from dissolution medium 2 (Fig 2-B). Using the simulated concentrations for T2 and real concentrations for T1 and R, pharmacodynamic profiles were built using the Hill equation, with the parameters obtained from the literature (Table 3). Mean values of Emax, EC50, and S correspond to the maximum sedation effect, plasma concentration to obtain the half of Emax, and the exponent of EC50 obtained in those healthy subjects that took part in the bioequivalence study 1.

Fig. 3 shows the foreseen sedative effect of flunitrazepam throughout time for the three products. The result obtained for T2 is very different depending on the dissolution medium (pH 1.2 or pH 4.5) considered for the IVIV correlation. This could be expected for drugs having a high slope for the effect-concentration relationships. Regardless of the dissolution medium used for the IVIV correlation, the intensity of sedation reached by T2 is low enough to have serious clinical consequence. A survey carried out with the medical staff of the Vilardebó Psychiatric Hospital revealed that no sedation effect was observed in patients using T2 product. Then, the assayed in vitro dissolution medium 2, pH 4.5 with 0.5% of Tween 80, seemed to be highly relevant to correlate with the in vivo pharmacodynamic/clinical performance of formulations containing flunitrazepam as active ingredient.

In vivo

pH = 1.2

pH = 4.5

pH = 6.8

ka(T) / ka(R)

1.32

-------

-------

-------

kd(T1) / kd(R)

-------

1.32

1.29

0.997

Table 2. Bibliographic T/R ka ratio, and experimental T1/R kd ratio.

Parameter

Emax

S

EC50 (ng/mL)

Value

4.00

11.4

6.75

Table 3. Literature parameters for the Hill equation explaining PK-PD correlation.

169

MAGALLANES L., OLMOS I., IBARRA M., MALDONADO C., VÁZQUEZ M. & FAGIOLINO P.

A

B

Figure 3. Effect (sedation intensity) vs. time profile projected for the evaluated products. (A) Dissolution medium 1: pH 1.2 for IVIV correlation. (B) Dissolution medium 2: pH 4.5 for IVIV correlation.

CONCLUSION AND PERSPECTIVE A virtual in vivo biopharmaceutic inequivalence between 2 formulations was detected by means of an in vivo–in vitro pharmacokineticpharmacodynamic multiple-correlated experimental and bibliographic mixed procedure. This work was an attempt to search a scientific explanation of the ineffectiveness clinically detected for one formulation containing flunitrazepam in the Uruguayan market. The strategy, used here, of finding a possible scenario of bioinequivalence among multisource drug products aimed us to continue the investigation in order to formalize a National Project in order to convey that all our pharmaceutical products containing active ingredients with a medium-tohigh health population risk could be in vivo virtually evaluated, and thereafter, if some bioinequivalence evidence was suspected, a real bioequivalence study could be put in practice to definitively assess their interchangeability. Acknowledgements. The authors are grateful to Dr. Nikoletta Fotaki (Department of Pharmacy and Pharmacology, University of Bath, United Kingdom) for her highly valuable suggestions.

170

REFERENCES

1. Grahnén A., P.Wennerlund, B.Dahlstrröm & S.Å. Eckernäs (1991) Br. J. Clin. Pharmac. 31: 89-92. 2. Boxenbaum H.G., H.N. Posmander, T. Macasieb, K.A. Geitner, R.E. Weinfeld, J.D. Moore, et al. (1978) J. Pharmacokinet. Biopharm. 6: 283-93. 3. Cano J.P., M. Soliva, D. Hartmann, W.H. Ziegler & R. Amrein (1997) Arzneim. Forsch. 27: 2383-8. 4. Benet L.Z., F. Broccatelli & T.I. Oprea (2011) AAPS J. 13: 519-47. 5. República Oriental del Uruguay (2007) Decreto del Poder Ejecutivo Nº 12/007 del 12 de Enero de 2007. Available at: . 6. World Health Organization (2006) WHO 14th Report, Technical Report series No 937, Annex 7: Multisource (generic) pharmaceutical products: guideline on registration requirements to establish interchangeability. Geneva, pp. 34790. 7. Holford, N.H. & L.B. Sheiner (1981) Clin. Pharmacokinet. 6: 429-53.