1. Introduction
3-Dimensional printing is set to be the next technology that will revolutionize the pharmaceutical industry in the coming years. In a time where personalized medicine is gaining more and more ground, additive manufacturing (AM) could be an interesting solution to tailor drugs to meet the personal needs of each patient. This state of the art technology has already gone beyond the stage of simple experimentation as the commercialization of Spritam
®, the first FDA-approved 3D printed pill, proves it [
1]. Among the diverse techniques of AM, fused deposition modeling (FDM) seems the most promising one, within the scientific community. Indeed, the production of solid oral dosage forms (SODFs) by FDM has been the subject of 72 papers from 2014 to 2019 [
2].
On the other hand, other 3D printing techniques remain not profoundly explored for pharmaceutical applications, for example, Selective Laser Sintering (SLS), which also proved to be very attracting [
3,
4] and might outweigh FDM in terms of precision and applied temperatures. This technique consists of the consolidation of powder particles with the energy provided by a laser. The process of SLS starts by the spreading of a thin layer of powder over the building area. Then, the laser scans the powder bed according to a specific pattern dictated by the pre-established design of the object, to fuse partially or completely the particles depending on the amount of transmitted energy. Next, the build platform lowered, and another layer of powder is spread over the previously sintered layer. This process repeats itself until complete achievement of the object (
Figure 1). The powder that has not been consolidated remains in place and serves as a support to the object during its building. Powder can also be recycled after sieving, making the SLS a very economical technique [
5]. One of the main advantages of the SLS for drug manufacturing along its high resolution is that the feedstock is powder, which is common to other pharmaceutical manufacturing processes. Hence, there is no need to pretreat the material like in FDM in which filaments need to be produced by Hot Melt Extrusion (HME) [
6].
To date, although only a few papers related to the production of SODFs by SLS have been issued [
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20], they demonstrated clearly that it is possible to manufacture oral medicines with a sintering machine using different pharmaceutical thermoplastic polymers (copovidone, cellulose derivatives and Eudragit
®). These contributions distinctly highlight the most important benefit that SLS could have in pharmaceutical manufacturing: its ability to create more or less porous forms by modulating the printing parameters and hence control the drug release from the printed SODFs.
It is important to note that although all the aforementioned work was conducted with SLS machines that use a blue diode laser, none of the evaluated polymers absorbed at the wavelength of the laser. Therefore, a colorant e.g., “Candurin
®” was added to enhance the absorbance and allow the sintering process [
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20]. Nonetheless, the majority of commercially available SLS printers use a different laser beam, which is the carbon dioxide (CO
2) laser beam (λ = 10.6 μm). This laser is relatively powerful and could be detrimental to active ingredients. However, incorporation of an absorbance enhancer could not be required since many biocompatible and biodegradable polymers absorb at the wavelength region of the CO
2 laser, as demonstrated by Salmoria et al. [
21]. This research team has printed drug delivery devices (DDDs) with a CO
2 laser using different polymers such as polycaprolactone [
22] and polyethylene [
23]. However, none of the studies conducted with the CO
2 laser on DDDs, evaluated drug stability [
4]. Hence, the implementation of CO
2 laser SLS for the production of oral dosage forms, would need to overcome the barrier of drug degradation. A way to achieve this is to use SLS printers that can modulate the laser power, a non-modifiable parameter in the commonly used Sintratec
® Kit printer.
Furthermore, the design and development of oral medicines by 3D printing techniques requires suitable material for this purpose. Printability of pharmaceutical polymers for FDM has been the subject of many articles in the last years. For example, it has been demonstrated that drug loading can affect the brittleness of the extruded filament and therefore may induce clogging during FDM printing [
24,
25]. As for SLS, except for a verification of the polymer’s absorbance at the laser’s wavelength [
7], the relationship between the properties of pharmaceutical polymers and sintering process remains not profoundly investigated. In order to be “sinterable”, the polymeric material should present suitable physicochemical properties. The adequate physical properties include good flowability and high packing density (compactness), which are mostly influenced by the granulometric and morphologic characteristics of the powder particles. In addition, the polymer should absorb at the laser’s wavelength which is mainly dependent on its chemical structure. Good flowability is desirable to achieve an effective powder deposition whereas compactness and absorbance are known to control the subsequent laser consolidation [
5,
26]. Although critical attributes for sintering are relatively well understood [
27], to our knowledge, there is no report of using pre-sintering methods to allow a fast screening of printable pharmaceutical polymers and formulations. Such tools could be beneficial for future pharmaceutical research to choose suitable powders for SLS and avoid trial-and-error methods. Far more importantly, study of printability could give more insight into the pharmaceutical materials’ sensitivity towards sintering and help manufacturers to improve the properties of already available thermoplastic pharmaceutical polymers that may be suitable for HME processes but not for SLS. Also, important to note that SLS, as well as other 3D printing techniques, show an important advantage in terms of personalizing oral dry forms and are therefore more intended for precision medicine than mass manufacturing. SLS is only at its beginning in the pharmaceutical landscape and more printability studies to implement the technology at a clinical scale and industrial scale will be required.
Kollidon
® VA64 (copovidone) is a water-soluble and thermoplastic copolymer composed of hydrophilic vinylpyrrolidone and lipophilic vinyl acetate. It is broadly used in pharmaceutical applications such as binder in the production of granules by wet granulation, dry binder in direct compression, film former in tablet coating, and polymeric matrix in HME [
28]. Kollidon
® VA64 was previously investigated as a polymer backbone in FDM and proved to be beneficial in terms of lowering the process temperatures and accelerating the drug release [
29]. It has also been tested on SLS and particularly for the production of orally disintegrating tablets (ODTs) [
8,
14,
15]. Moreover, Kollidon
® VA64 is available in two different grades depending on the particle size. This could be interesting for SLS since the particle size distribution is the main aspect to take into consideration, as mentioned above.
The aim of this work is to assess the suitability of a CO2 laser (λ = 10.6 μm) for selective laser sintering of oral dosage forms using Kollidon® VA64 as a polymeric carrier and Paracetamol as a model drug. Prior to sintering, native powders, as well as mixtures, were characterized in order to understand the thorough relationship between their physicochemical properties and printability. An SLS commercial powder “polyamide 12” was chosen as a reference material in order to interpret the results for Kollidon® VA64 and the powder mixtures. Then, printed solid oral dosage forms were characterized, and drug stability was studied by ultra-high performance liquid chromatography (UHPLC). Finally, the influence of drug loading on both the sintering process and the properties of the printed SODFs was assessed.
2. Materials and Methods
Kollidon® VA64 (KVA64) and Kollidon® VA64 Fine (KVA64F) were generously donated by BASF (Ludwigshafen, Germany). Duraform® polyamide 12 (PA12) was provided by 3D Systems (Santa Clarita, CA, USA) and used as a reference powder. Paracetamol crystal (PAR) and paracetamol crystal fine (PAR F) were purchased from Sequens (Porcheville, France).
2.1. Physicochemical Characterization of Powders
2.1.1. Scanning Electron Microscopy (SEM)
In order to study the particle morphology of powders, images of each of KVA64, KVA64F, PA12, PAR and PAR F were taken with a scanning electron microscope (4800 S, Hitachi, Tokyo, Japan) after platinum sputtering under vacuum before observation. The microscope was also used to take images of the printed SODFs (surface and vertical sections).
2.1.2. Preparation of Mixtures
Mixtures of the two grades of copovidone were prepared at three different proportions (
Table 1). Formulations based on KVA64 and paracetamol (PAR and PAR F) were also prepared at three different drug loadings (
Table 1). Mixing was conducted on a 3D shaker mixer Turbula
® T2F (WAB, Muttenz, Swizterland) at a speed of 49 rpm for 10 min.
2.1.3. Laser Granulometry
Dry laser diffraction (Mastersizer 2.18, Malvern Instruments Ltd., Malvern, UK) was used to determine the mean particle size (D (4,3)) and the size distribution (span = ((D90 − D10))/D50) of the native powders as well as the mixtures. A jet pressure of 1.2 bar was used to deagglomerate the particles during laser measurement. Data treatment was realized using the software Mastersizer S 2.18 and choosing the analysis mode as polydisperse. For each powder, a sample of approximately 1 g was analyzed, and each measure was performed at least in triplicate.
2.1.4. Study of Flowability and Compactness
Compactness of the native powders and the mixtures was assessed by the measurement of bulk (
BD) and tapped density (
TD). The test was conducted following the method described in the European Pharmacopeia [
30] with a 250 mL graduated cylinder and a sample mass of 30 g. Hausner ratio (
HR) was then calculated to express the powder flowability, according to the following formula:
Powder flowability was also evaluated by the measurement of the angle of repose (AOR) using a granulate flow tester GTB (Erweka, Langen, Germany) according to the European Pharmacopeia guidelines [
31]. It was conducted by allowing a mass of 30 g of each powder positioned above a fixed diameter base to drain from a 200 mL funnel through a 15 mm nozzle. Stirring was fixed at the speed 4. The drained angle of repose was determined from the cone of powder formed on the base. Each measure was done in triplicate.
2.1.5. Fourier-Transform Infrared Spectroscopy (FTIR)
Infrared spectrophotometer Vector 22 FTIR (Bruker, Billerica, MA, USA) was employed to evaluate the absorbance of the polymers (KVA64 and PA12) as well as paracetamol at the wavelength of the printer’s laser. Absorbance was recorded from 4000 to 400 cm−1 at room temperature (approximately 25° C) and 64 scans were averaged at a resolution of 4 cm−1. Samples of 100 mg were prepared by blending 10 mg of the polymer or 1 mg of the drug with Q.S. (Quantum satis) of anhydrous potassium bromide (previously dried in the oven at 100 °C for 30 min) and compressing the mixture to form a disk. The FTIR spectrums were treated using the infrared software OPUS 6.5 (Bruker, Billerica, MA, USA).
2.2. Printing of SODFs:
3D model of a cylindrical dosage form (10 mm diameter and 3 mm height) was designed using an online CAD software OnShape® (Onshape, Boston, MA, USA) and exported as a STL file. Then, it was converted to a G-code with an open source software Slic3r® 1.2.9 before transferring it to the 3D SLS printer Sharebot® SnowWhite (Sharebot, Nibionno, Italy).
A mass of 300 g from each powder was loaded in the reservoir tanks and the building platform (100 × 100 × 100 mm3) to the brim. The air gaps formed in the deposited powder were eliminated by recalibrating the level of the tanks. An automatic recoater blade removed the surplus of powder on top of the building platform to create a flat surface. For the case of powders with poor flowability, powder filling, recalibrating, and recoating were repeated until the formation of a flat surface. For all the printings, the temperature mode was set at “powder temperature” which meant that the heaters were controlled by the temperature of the powder bed.
The optimized printing parameters for polyamide 12 were previously developed in our department. For the other printable powders, an optimized setting (
Table 2) for which all the SODFs were completely printed (with no missing layer), was achieved after preliminary tests. Heating temperature (°C), laser power (% of the maximum laser power) and scan speed (pps or points per second ≈ 0.05 mm/s) were machine parameters, whereas the layer thickness (mm) was entered in Slic3r
®.
Thirty-six SODFs were launched for printing per batch. The process started with the heating of the powder by infrared lamps (230 W) for thirty minutes. Afterward, a CO2 laser (14 W) sintered the successive powder layers according the 3D model of SODFs. The overall printing time depended mainly on the chosen scan speed and layer thickness. Finally, when printing was completed, the powder bed containing the printed SODFs was removed and sieved using a 250 μm sieve to eliminate the excess powder around the SODFs.
2.3. Characterization of the Printed SODFs
2.3.1. Differential Scanning Calorimetry (DSC)
DSC was used to determine the melting point (or the glass transition temperature) and the solid state of polymers, drug, physical mixtures and printed SODFs. Accurately weighed samples (5–10 mg) were placed in sealed aluminum pans and heated from 25 °C to 200 °C at 10 °C/min with a DSC 4000 (Perkin Elmer, Waltham, MA, USA). A heat-cool-heat cycle method was conducted to remove the thermal history of copovidone. Nitrogen was used as a purge gas with a flow rate of 20 mL/min. Data collection and analysis were conducted using Pyris Manager software (Perkin Elmer, Waltham, MA, USA).
2.3.2. X-ray Powder Diffraction (XRPD)
The solid state of the polymers, drug, physical mixtures and printed SODFs was characterized using a Bruker D8 Advance diffractometer (Bruker, Billerica, MA, USA) and the monochromatic Cu Kα1 radiation (λα = 1.5406 Å, 40 kV and 40 mA). The angular range of data recording was 2–80° 2θ, with a stepwise size of 0.02° and a speed of 0.1 s counting time per step, using LYNXEYE detector 1D (Bruker, Billerica, MA, USA).
2.3.3. Weight, Dimensions and Mechanical Strength of the Printed SODFs
Weight of the SODFs was determined using a precision electronic balance Adventurer® (OHAUS, Parsippany, NJ, USA). Physical dimensions (height and diameter) and hardness were measured using a Sotax Multitest 50FT (Sotax AG, Basel, Switzerland). Measurements were carried out on 10 SODFs per printing batch and results were expressed as the mean value ± standard deviation.
2.3.4. Disintegration Time of the Printed SODFs
Disintegration tests were performed on a disintegration apparatus (Sotax DT50, Sotax AG, Basel, Switzerland) with distilled water at 37 °C according to the European Pharmacopeia guidelines [
32]. For each printing batch, six SODFs were tested simultaneously. The disintegration time was reached when no residues were present on the bottom of the test basket. Results were reported as the mean value ± standard deviation.
2.3.5. Drug Content of the Printed SODFs
For each formulation, three individual SODFs were dissolved in 100 mL of distilled water. Samples of the solutions were then diluted and the drug concentration was determined by ultra-high performance liquid chromatography (UHPLC, Thermofisher Scientific, Waltham, MA, USA) using a UHPLC-DAD system. It consisted of a Thermo Scientific™ Dionex™ UltiMate™ 3000 BioRS equipped with a WPS-3000TBRS autosampler and a TCC-3000RS column compartment set at 35 °C. The system was operated using Chromeleon 7 software (Thermofisher Scientific, Waltham, MA, USA). An Accucore C18 column (2.6 µm, 100 × 2.1 mm2) combined with a security guard ultra-cartridge (Phenomenex Inc., Torrance, CA, USA) was used. An isocratic binary solvent system was utilized, consisting of water/formic acid (1%, v/v) as solvent A and acetonitrile/formic acid (1%, v/v) as solvent B (90%A, 10%B). The flow rate of the mobile phase was 1.5 mL/minute, and the injection volume was 50 μL. Quantitative analysis of paracetamol in the SODFs was carried out using an external standard method. The calibration curve was constructed using 5 different standard levels in the concentration range 1–20 mg/L. The peak of paracetamol was monitored at 244 nm.
2.3.6. Size Exclusion Chromatography—Multi Angle Light Scattering (SEC-MALS)
The degradation of copovidone during the SLS process was assessed by analyzing both the raw polymer and the sintered KVA64 placebo SODF on SEC-MALS (Thermofisher Scientific, Waltham, MA, USA). The experiments were performed at 35 °C on a Thermo Scientific Ultimate 3000 module equipped with a OHpak SBG Shodex column guard (50 × 6 mm
2) and a SB-805-HQ Shodex column (300 × 8 mm
2) connected in series in association with a miniDawn Treos laser light scattering detector having a 658-nm laser (Wyatt Technology Corp., Santa Barbara, CA, USA) and with a RID-6A refractive index monitor (Shimadzu Corp., Kyoto, Japan). The eluent used was composed of a mixture of 0.15 M phosphate buffer and 1 M NaCl at pH = 7.4. The eluent was filtered using Durapore membrane filters of 0.1 μm cut-off. Incremental refractive index (dn/dc) value of 0.15 was used, as found in the literature [
33]. The polymer samples (100 μL injection volume at a concentration of 1 g.L
−1) were eluted at a 1 mL.min
−1 flow rate. The data were analyzed using the Astra software (Wyatt Technology Corp., Santa Barbara, CA, USA, v6.1.1.17).
2.3.7. Drug Release of the Printed SODFs
A dissolution test was carried out for SODFs containing paracetamol with a Pharma Test DT70 dissolution tester (Pharma Test Apparatebau AG, Hainburg, Germany) using a paddle apparatus (European Pharmacopeia) [
34]. For each formulation, three SODFs were randomly selected and individually placed in the dissolution vessels, each containing 900 mL of 0.1 M HCl (sink condition) and stirred at 100 rpm and 37 ± 0.5 °C. Samples were withdrawn automatically each 2 min and analyzed using a continuous flow through system attached to an 8 cell UV/Vis spectrophotometer Specord 250 (Analytik Jena, Jena, Germany) at a wavelength of 268 nm. Results were expressed as mean values with standard deviation.
4. Conclusions
In this study, the production of solid oral dosage forms with copovidone and paracetamol by SLS using a CO2 laser was demonstrated for the first time. The ability of KVA64 to absorb at the laser’s wavelength (10.6 µm) make it suitable for SLS, and the addition of an absorbance enhancer was not necessary. Furthermore, UHPLC analysis confirmed that no drug degradation occurred during sintering despite the relatively high power of the laser. This opens a new area of research in the use of this type of printer for the preparation of SODFs. However, more thermosensitive drugs could be affected by the CO2 laser and their degradation should be evaluated in further studies.
Flowability was found to be critical for the process and was mainly dependent on the morphology and granulometry of the particles. Hence, in the preparation of formulations for SLS, not only the grade of polymer (KVA64) have to be chosen correctly but also the grade of the API (PAR). Mixtures of KVA64 and PAR presented lower compactness compared to the reference material (PA12), which resulted in mediocre mechanical properties. However, if high density is usually preferable for printed parts intended for engineering, presence of porosity is more interesting for pharmaceutical applications especially for modulation of drug release. The percentage of drug was proven to have an impact on the sintering process by lowering the heating temperature of the powder due to the plasticizer effect of paracetamol. Different drug loadings also influenced the SODF properties, especially drug release.
Overall, Kollidon® VA64 has potential in 3D printing techniques, and this aptitude could be considerably boosted for SLS when powder particles are matching the morphological and rheological requirements for the technology: adequate particle shape, size distribution, and most importantly, a good flowability. This confirms that critical quality attributes of raw materials need to be rethought with the advent of new pharmaceutical production processes like additive manufacturing. In order to facilitate the establishment of the SLS technology in the pharmaceutical landscape, future studies would be encouraged to explore further the material-process relationship and to optimize the feedstock’s printability with physical modifications. Nevertheless, this study suggests some predictive tools for the “sinterability” of polymeric excipients: measurement of the absorbance at the laser’s wavelength, evaluation of the compactness using bulk density and study of flowability by calculation of Hausner ratio and angle of repose. This demarche is interesting since no GMP certified SLS machine nor pharmaceutical grade feedstock are commercially available.