Rifampicin-loaded 'flower-like' polymeric micelles for

Research Article For reprint orders, please contact: [email protected]

Rifampicin-loaded ‘flower-like’ polymeric micelles for enhanced oral bioavailability in an extemporaneous liquid fixed-dose combination with isoniazid Background: Coadministration of rifampicin (RIF)/isoniazid (INH) is clinically recommended to improve the treatment of tuberculosis. Under gastric conditions, RIF undergoes fast hydrolysis (a pathway hastened by INH) and oral bioavailability loss. Aim: We aimed to assess the chemical stabilization and the oral pharmacokinetics of RIF nanoencapsulated within poly(e-caprolactone)-b-PEG-b-poly(e-caprolactone) ‘flower-like’ polymeric micelles. Materials & methods: The chemical stability of RIF was evaluated in vitro under acid conditions with and without INH, and the oral pharmacokinetics of RIF-loaded micelles in rats was compared with those of a suspension coded by the US Pharmacopeia. Results: Nanoencapsulation decreased the degradation rate of RIF with respect to the free drug. Moreover, in vivo data showed a statistically significant increase of RIF oral bioavailability (up to 3.3-times) with respect to the free drug in the presence of INH. Conclusion: Overall results highlight the potential of this nanotechnology platform to develop an extemporaneous liquid RIF/INH fixed-dose combination suitable for pediatric administration. Original submitted 6 April 2013; Revised submitted 7 August 2013 KEYWORDS: extemporaneous liquid rifampicin/isoniazid fixed-dose combination n improved oral pharmacokinetics n poly(e-caprolactone)-b-PEG-b-poly(e-caprolactone) ‘flower-like’ polymeric micelle n pediatric tuberculosis n rifampicin chemical stabilization

Marcela A Moretton1,2, Christian Hocht3, Carlos Taira2,3 & Alejandro Sosnik*1,2,4 The Group of Biomaterials & Nanotechnology for Improved Medicines, Department of Pharmaceutical Technology, Faculty of Pharmacy & Biochemistry, University of Buenos Aires, Buenos Aires, Argentina 2 National Science Research Council, Buenos Aires, Argentina 3 Department of Pharmacology, Faculty of Pharmacy & Biochemistry, University of Buenos Aires, Argentina 4 Department of Materials Science & Engineering, Technion-Israel Institute of Technology, Haifa, Israel *Author for correspondence: [email protected] 1

Tuberculosis (TB) is the most deadly infection after HIV [1–3,101]. TB is regarded as a povertyrelated disease because it mainly affects poor countries [4]. In 1993 WHO declared a global sanitary emergency [5]. The first-line pharmacotherapy of nonresistant TB is comprised of two phases that last 6 months. Both phases demand the coadministration of rifampicin (RIF) and isoniazid (INH; isonicotinylhydrazine) by the oral route [6,7,102]. Nonresistant TB is curable but it accounts for more than 25% of preventable and 24% of all deaths worldwide [8,103]. Under gastric conditions, RIF undergoes hydrolysis to 3-formyl RIF SV (3-FRSV), a derivative without activity in vivo due to negligible gastrointestinal absorption [9]. This degradation pathway is hastened by soluble INH [10–12]. The coadministration of RIF/INH fixed-dose combinations (FDCs) is clinically recommended to prevent the development of resistance. The significant decrease of the oral bioavailability of RIF in most RIF/INH FDCs is usually disregarded [13]. WHO has raised awareness of this therapeutic drawback and advised the use of quality-assured FDCs of proven RIF bioavailability [14,104]. TB is among the ten main causes of childhood mortality, with an estimated annual toll of 130,000 deaths [15]. The lack of commercially available pediatric medicines that enable fine-tuning of dosage and swallowing in

poverty-related diseases is a remarkable hurdle for convenient therapy [16,17]. Single RIF and INH liquid formulations are commercially available in several countries. Conversely, liquid RIF/INH FDCs have not been developed. Macleods Pharmaceuticals (Mumbai, India) developed a series of double (RIF/INH) and triple (RIF/INH/pyrazinamide) FDC dispersible tablets that could be used to obtain extemporaneous suspensions for pediatric use [17,105]. These products are listed in the WHO List of Prequalified Medicinal Products; although oral bioavailability data are not currently available. Novel drugs are expected to shorten the course of the treatment and to be effective against resistant strains. At the same time, the development of innovative formulations of approved drugs that could also lead to breakthroughs in pharmacotherapy is in constant progress and it is becoming complementary to drug discovery [18]. For example, Choonara et al. have recently reported on the production of super-stable nanoparticles for the sustained release of anti-TB drugs [19]. Polymeric micelles are one of the most versatile nanocarriers to enhance the water solubility, the physicochemical stability and the bioavailability of poorly water soluble and instable drugs [20,21] and diverse administration routes, such as oral, parenteral, ocular and intranasal, have been explored [22–26].

doi:10.2217/NNM.13.154 © 2014 Future Medicine Ltd

Nanomedicine (Epub ahead of print)

part of

ISSN 1743-5889

Research Article

Moretton, Hocht, Taira & Sosnik

10.07°

200 nm

100 nm

-40.81°

Figure 1. Morphology of rifampicin-loaded poly(e-caprolactone)(4500) polymeric micelles. (A) Transmission electron microscopy micrograph of fresh 4% micelles and (B) tapping mode atomic force microscopy 2D micrograph of freeze-dried 1% micelles lyoprotected with hydroxylpropyl b-cyclodextrin 15%. The sample was diluted 1:20 before the analysis. Arrows indicate micelles.

The development of liquid pediatric FDCs is urgently needed to ease dose adjustment and swallowing and improve patient compliance [17]. At the same time, due to the detrimental RIF/INH interaction, liquid FDCs would need to be conceived as extemporaneous powders for redispersion where RIF is chemically stabilized. We previously described the encapsulation of RIF within ‘flower-like’ polymeric micelles of poly(e-caprolactone)-b -PEG-b -poly(ecaprolactone) (PCL-b-PEG-b-PCL) block copolymers and their physical stabilization by freeze drying [27,28]. The present work investigated the capacity of these nanocarriers to protect RIF from degradation under extreme acid conditions in the absence and presence of soluble INH and the pharmacokinetics of the drug in rats after oral administration. Overall results highlight the potential of this nanotechnology platform to develop an extemporaneous liquid RIF/INH FDC to treat childhood TB.

Materials & methods Materials PEG (molecular weight: 10 kg/mol, PEG10000), e-caprolactone (CL; monomer), tin (II) 2-ethylhexanoate (catalyst), RIF, INH, and solvents of analytical or HPLC grade were used as received. Hydroxypropyl-b-cyclodextrin (HPb-CD) was a gift of ISP Technologies Inc. (NJ, USA).

Two derivatives bearing terminal PCL blocks of average molecular weight of 3.7 (32 CL units/arm) and 4.5 kg/mol (40 CL units/arm) and theoretical molecular weights of 17.4 and 19.0 kg/mol, respectively, were used [27]. PCL3700-b-PEG10000-b-PEG3700 and PCL4500-b-PEG10000-b-PCL4500 copolymers are named PCL(3700) and PCL(4500), respectively. The hydrophilic–lipophilic balance of the copolymers was estimated by the ratio between the number of CL and ethylene oxide (EO) repeating units in the copolymer, designated the CL:EO ratio. Preparation of RIF-loaded PCL-b-PEG-bPCL polymeric micelles

RIF-loaded micelles of different copolymer concentrations (1, 4 and 6%) were prepared by means of the cosolvent/evaporation method [27]. RIF concentrations were determined by UV–visible spectrophotometry (Supplementa ry I nformation; see online at www.futuremedicine. com/doi/suppl/10.2217/nnm.13.154). Freeze drying of RIF-loaded micelles

RIF-loaded micelles were freeze-dried employing HPb-CD (10, 13, 15 and 20% w/v) as cryo-/ lyo-protectant [28]. Samples were reconstituted in distilled water (1 ml) before use (Supplementary I nformation ). Morphology of RIF-loaded micelles

Methods Copolymer synthesis

PCL-b-PEG-b-PCL copolymers were synthesized as described previously [27,29,30]. doi:10.2217/NNM.13.154


The morphology of RIF-loaded PCL(4500) micelles was studied by transmission electron microscopy and atomic force microscopy (Supplementary I nformation ). future science group

Rifampicin-loaded polymeric micelles for enhanced oral bioavailability

In vitro release of RIF

The analysis and fitting were conducted with SigmaPlot® software and Microsoft Excel® 2003 (both Microsoft, WA, USA).

The release of RIF was assessed employing phosphate-buffered saline (pH 7.4) containing ascorbic acid (200 µg/ml) [31] as an external release medium (600 ml) under mechanical stirring (70 rpm) over 6.5 h, at 37°C (Supplementary Information ). The goal of this assay was to study the role played by the release of RIF in the degradation process. These pH conditions were selected to prevent RIF degradation during the assay and to simplify the analysis. The addition of HPb-CD as cryo-/lyo-protectant could alter the release of RIF. To study this effect, the release from RIFloaded micelles (4%) containing 20% HPb-CD was also assessed. Assays were carried out in triplicate and the results are expressed as the mean ± standard deviation of the mean. Release data were fitted to the Korsmeyer–Peppas model [32,33], considering micelles as spheres [27,34] (Supplementary Information).

Measurement of the micellar size & size distribution

The effect of RIF release on the size and size distribution (expressed by the polydispersion) of RIF-loaded micelles was studied by dynamic light scattering over 2 h, at 37°C (Supplementary I nformation). Chemical stability of RIF

Fresh and freeze-dried RIF-loaded micelles (4%) were incubated in acid medium in the absence and presence of soluble INH. For the latter assays, INH was solubilized in hydrochloric acid (HCl) 0.1 N and the RIF:INH weight ratio was maintained at 3:2, as used in previous studies [9] and recommended in clinics [10,11,14].

A

B 100 PCL(3700) 1% PCL(3700) 4% PCL(3700) 6% Free RIF

80

RIF cumulative release (%)


100

60 40 20

PCL(4500) 1% PCL(4500) 4% PCL(4500) 6%

80 60 40 20 0

0 0

2

4 Time (h)

6

0

8

2

4 Time (h)

6

8

D

C

100

100 PCL(3700) 4%



Research Article

PCL(3700) 6%

80 60 40 20

0

2

4 Time (h)

6

8

PCL(4500) 6%

80

PCL(4500) 4%/HPβ-CD 20%

60 40 20 0

0

PCL(4500) 4%

0

2

4 Time (h)

6

8

Figure 2. In vitro rifampicin release from poly(e-caprolactone)-b-PEG-b-poly(e-caprolactone) polymeric micelles containing different copolymer concentrations. (A) Fresh 1–6% PCL(3700) micelles containing 4.9 mg/ml of RIF, (B) fresh 1–6% PCL(4500) micelles containing 4.9 mg/ml of RIF, (C) fresh 4 and 6% PCL(3700) micelles containing 10.4 and 12.2 mg/ml of RIF, respectively, and (D) fresh and lyophilized 4% PCL(4500) micelles containing 10.4 and fresh 6% micelles containing 12.2 mg/ml of RIF. Each point represents the mean ± standard deviation of the mean of three independent experiments. HPb-CD: Hydroxylpropyl b-cyclodextrin; PCL: Poly(e-caprolactone); RIF: Rifampicin.

future science group

www.futuremedicine.com

doi:10.2217/NNM.13.154

0.12 0.18 0.14 0.14 0.12 0.10 0.15 0.08 0.07 0.18 0.20

0.75 0.45 0.53 0.77 0.73 1.00 0.60 0.67 0.96 0.46 0.67

0.9787 0.9772 0.9405 0.9274 0.9690 0.9803 0.9789 0.9854 0.9241 0.9806 0.9115

Saturated free RIF solutions (2.6 mg/ml) were used as control. Samples were diluted (1:10 and 1:40) in HCl 0.1 N and incubated at 37°C under continuous magnetic stirring (100 rpm) for 3 h [22,35]. These pH conditions and sampling intervals were selected to mimic the gastric transit time and acid environment where RIF degradation is maximal. The addition of HPb-CD could modify the chemical stability of RIF with respect to fresh samples. Thus, freeze-dried samples were resuspended in the original volume of distilled water and diluted with HCl 0.1 N. At different time points, samples (500 µl) were diluted with phosphate buffer saline of pH 7.0 to quench the acid degradation process and analyzed by reverse-phase HPLC (see below). RIF follows first-order degradation kinetics [10]. Thus, the degradation constant under the different conditions was determined by the following equation: Fresh micelles. Micelles lyophilized with 20% hydroxylpropyl b-cyclodextrin. § Analysis by the Korsmeyer–Peppas model was conducted for Mt/M ∞ (fraction of drug released at a given time) ≤0.6. k: Kinetic constant; ko: Zero-order release constant; k1: First-order release constant; n: Release exponent; PCL: Poly(e-caprolactone).

Lyophilized PCL(4500) ‡

PCL(4500)†

‡

10.4 12.2 10.4

10.4 12.2 4.9

doi:10.2217/NNM.13.154

†

0.10 0.07 0.06 0.22 0.10 0.21 0.09 0.05 0.08 0.06 0.15 PCL(3700)†

1 4 6 4 6 1 4 6 4 6 4

4.9

8.27 5.00 4.07 8.03 8.94 13.22 5.86 5.62 7.59 7.11 14.51

0.9674 0.9599 0.9655 0.8688 0.9629 0.9915 0.9953 0.9768 0.8282 0.9996 0.8825

6.81 4.63 4.39 11.04 6.88 11.61 5.83 3.95 6.40 4.64 8.79

0.9710 0.9905 0.9480 0.9532 0.9588 0.9896 0.9982 0.9423 0.9235 0.9187 0.8451

0.10 0.07 0.05 0.10 0.11 0.17 0.08 0.06 0.09 0.09 0.21

0.9436 0.9682 0.9562 0.8666 0.9642 0.9785 0.9924 0.9833 0.8315 0.9984 0.8791

0.9788 0.9951 0.9180 0.8372 0.9609 0.9417 0.9943 0.9577 0.9253 0.9428 0.9132

R2adjusted n k k1 (h-1)

R2adjusted

Moretton, Hocht, Taira & Sosnik

R2adjusted k1 (h-1) R2adjusted ko (% h)

Korsmeyer–Peppas model § First order0–6.5 h First order0–3.5 h Zero order0–6.5 h

Copolymer Copolymer Rifampicin Zero order0–3.5 h concentration concentration ko (% h) R2adjusted (%) (mg/ml)

Table 1. Curve-fitting analysis of rifampicin release data from poly(e-caprolactone)-b-PEG-b-poly(e-caprolactone) polymeric micelles over 3.5 and 6.5 h.

Research Article


In (C f ) = In (C ) - K.dt 0

where Co is the initial concentration of RIF and Cf the concentration at a certain time, K d the apparent degradation constant (min-1) and t the time expressed in minutes. The percentage of RIF degraded at 3 h, D3h (%), was also calculated. This time is clinically relevant because it is the maximum gastric transit of RIF. Assays were carried out in triplicate and the results are expressed as the mean ± standard deviation of the mean. HPLC method for chemical stability assays

Analyses were carried out using an adapted reverse-phase HPLC validated method with a UV detector (l = 254 nm) (S upplementa ry I nformation ) [11]. Oral pharmacokinetics of RIF

The goal of this study was to evaluate the effect of RIF solubilization and chemical stabilization within the micelles on its bioavailability. The oral pharmacokinetics of RIF was assessed in male Wistar rats (weight of 300–350 g) in the absence and presence of INH (RIF:INH weight ratio 3:2) (Supplementary Information). HPLC method for oral pharmacokinetics assays

RIF concentrations in plasma were determined by reverse-phase HPLC with UV (l = 330 nm) (Supplementary I nformation ). future science group


25.4 (2.8)

37.6 (2.0)

1

2

www.futuremedicine.com

8.9

4.0

14.3 18.4 10.0

4.8 11.0 19.3 14.9 13.1

100.0 100.0 100.0 8.9 9.7 95.2 89.0 80.7 85.1 86.9

– – – 91.1 90.3

%

142.4 (5.4)

88.8 (9.5) 14.1

13.8

293.4 (5.1) 85.7 316.4 (3.7) 81.6 684.9 (57.4) 90.0

170.3 (3.7) 427.3 (18.2) 506.6 (68.2) 457.3 (19.8) 234.6 (5.3)

– – – 170.9 (3.1) 171.9 (7.4)

Dh ; nm (SD)

Peak 2

Peak 3

801.5 (46.0) 860.3 (61.3)

– – –

– – – – –

– – – – –

Dh ; nm (SD)

PCL(3700)

77.0

82.2

– – –

– – – – –

– – – – –

%

0.686 (0.021)

0.712 (0.017)

0.540 (0.021) 0.541 (0.005) 0.691 (0.010)

0.150 (0.017) 0.172 (0.002) 0.174 (0.010) 0.424 (0.061) 0.360 (0.046) 0.361 (0.047) 0.557 (0.009) 0.616 (0.054) 0.610 (0.040) 0.492 (0.031)

PDI (SD)

Dh: Hydrodynamic diameter; PCL: Poly(e-caprolactone); PDI: Polydispersion; RIF: Rifampicin; SD: Standard deviation of the mean.

12.2

6

39.6 (2.1) 49.2 (3.3) 35.4 (7.9)

1 2 0

4.9

6

10.4

28.4 (4.6) 39.9 (4.8) 39.2 (1.4) 58.0 (7.3) 32.5 (4.0)

2 0 1 2 0

4.9

4

4

227.0 (8.2) 84.1 (1.2) 75.9 (0.8) 27.2 (5.6) 32.9 (3.8)

0 1 2 0 1

4.9

1

%

Peak 1

Dh ; nm (SD)

Time (h)

Copolymer RIF concentration payload (%w/v) (mg/ml)

35.2 (4.7) 38.9 (7.5) 47.1 (3.6) 47.6 (5.9) 113.8 (13.0) 50.5 (8.2) 35.6 (1.2) 1289.3 (124.6) 1264.3 (84.3) 213.8 (39.7)

169.9 (7.1) 66.6 (5.4) 72.7 (1.0) 25.4 (4.8) 27.6 (3.0)

14.4

100.0

7.9 6.5 100.0

10.4 9.1 11.5 12.4 5.1

100.0 100.0 100.0 7.7 7.1

%

Peak 1 Dh ; nm (SD)

89.6 90.9 88.5 87.6 94.9

– – – 92.3 92.9

%

1464.3 (98.8)

–

85.6

–

365.0 (29.8) 92.1 327.9 (9.0) 93.5 – –

187.2 (14.8) 361.3 (27.9) 481.0 (14.8) 450.6 (26.5) 858.9 (36.1)

– – – 176.0 (11.2) 173.3 (8.4)

Dh ; nm (SD)

Peak 2

PCL(4500)

0.627 (0.054)

0.209 (0.017)

0.472 (0.008) 0.314 (0.007) 0.154 (0.024)

0.321 (0.040) 0.473 (0.014) 0.585 (0.038) 0.562 (0.041) 0.338 (0.037)

0.158 (0.016) 0.193 (0.111) 0.137 (0.009) 0.337 (0.061) 0.284 (0.010)

PDI (SD)

Table 2. Micellar size and size distribution of rifampicin-loaded poly(e-caprolactone)-b-PEG-b-poly(e-caprolactone) micelles over time, at 37ºC (n = 3).


Research Article

doi:10.2217/NNM.13.154

doi:10.2217/NNM.13.154

–

–

PCL(3700)

PCL(4500)

0.21 0.053

0.1

0.053

0.1 0.4

0.21

0.053

0.1 0.4

0.21

0.058

0.1 0.4

0.23

–

mM

0.4

–

%

Copolymer final concentration upon dilution +† +† +† +† +† +† +† +† +†

INH content

0.0031 (0.0001) 0.0041 (0.0001) 0.0008 § (0.0001) 0.0018 § (0.0001) 0.0015§ (0.0002) 0.0023 § (0.0002) 0.0007§ (0.0001) 0.0018 § (0.0001) 0.0016 § (0.0001) 0.0024 § (0.0003) 0.0008 §,†† (0.0001) 0.0021§,†† (0.0001) 0.0021§ (0.0001) 0.0028 §,†† (0.0001) 0.0011§,†† (0.0001) 0.0021§,†† (0.0003) 0.0022§ (0.0001) 0.0042 (0.0001)

Kd ; min-1 (SD)

0.9769 (0.0023) 0.9933 (0.0018) 0.9312 (0.0062) 0.9799 (0.0175) 0.9549 (0.0271) 0.9685 (0.0152) 0.9800 (0.0285) 0.9502 (0.0057) 0.9452 (0.0484) 0.9531 (0.0646) 0.9776 (0.0274) 0.9668 (0.0220) 0.9807 (0.0175) 0.9954 (0.0059) 0.9868 (0.0078) 0.9560 (0.0411) 0.9736 (0.0094) 0.9507 (0.0373)

R2 (SD)

Results are expressed as mean (SD of the mean); n = 3. † Sample containing INH in a RIF:INH weight ratio of 3:2. ‡ Final RIF concentration after the corresponding 1:10 dilution. § Kd parameter of RIF-loaded micellar dispersions is significantly lower than that of free RIF controls. ¶ Final RIF concentration after the corresponding 1:40 dilution. # HB b -CD was incorporated as cryo-/lyo-protectant before the lyophilization at a concentration of 15%. †† No statistically significant difference was observed for the Kd of lyophilized and fresh RIF-loaded micellar dispersions. ‡‡ HB b -CD was incorporated as cryo-/lyo-protectant before the lyophilization at a concentration of 20%. D3h: Percentage of RIF degradation at time point 3 h; HP b -CD: Hydroxylpropyl b -cyclodextrin; INH: Isoniazid; Kd: Degradation constant; RIF: Rifampicin; SD: Standard deviation of the mean.

Lyophilized 20 PCL(4500) ‡‡

Lyophilized 15 PCL(4500) #

–

Control

0.25 (0.03) 0.24 (0.04) 1.08‡ (0.04) 1.11‡ (0.06) 0.26¶ (0.01) 0.28 ¶ (0.01) 1.15‡ (0.02) 1.12‡ (0.04) 0.27¶ (0.01) 0.28 ¶ (0.01) 0.98‡ (0.05) 0.95‡ (0.05) 0.27¶ (0.01) 0.27¶ (0.00) 0.98‡ (0.04) 1.07‡ (0.08) 0.24¶ (0.01) 0.25¶ (0.01)

HPb-CD RIF final concentration content (%) upon dilution; mg/ml (SD)

Sample

42.8 (1.5) 52.2 (0.1) 14.2 (1.1) 27.7 (1.8) 22.2 (4.0) 33.9 (2.1) 12.4 (1.8) 27.7 (1.3) 24.1 (0.8) 35.1 (2.3) 13.4 (0.1) 31.5 (1.7) 31.5 (0.1) 39.6 (0.1) 17.2 (1.1) 32.0 (4.3) 32.1 (0.9) 52.6 (0.6)

D3h ; % (SD)

Table 3. Rifampicin degradation parameters under gastric-like conditions for fresh and cryoprotected/lyophilized rifampicin-loaded 4% poly(e-caprolactone)-b-PEG-b-poly(e-caprolactone) polymeric micelles upon 1:10 and 1:40 dilution in the presence and absence of isoniazid.

Research Article Moretton, Hocht, Taira & Sosnik




Evaluation on in vivo data

Fr ^%h = AUCmic / AUCsusp # 100

where AUCmic and AUCsusp are the AUC0–24h of micellar dispersion and suspension, respectively. Statistics

Statistical analysis was performed by one-way analysis of variance using Microsoft Excel 2003 software. The results were considered statistically significant if p