A Novel Approach to Evaluate Amorphous-to-Crystalline Transformation of Active Pharmaceutical Ingredients in Solid Dispersion Using Time-Domain NMR

Abstract

The aim of this study was to demonstrate the usefulness of the time-domain NMR (TD-NMR) method to characterize the crystalline state of active pharmaceutical ingredients (APIs) containing a solid dispersion. In this study, indomethacin (IMC) was used as a model for poorly water-soluble API. Solid dispersions of IMC were prepared with polyvinylpyrrolidone (PVP) at different weight ratios. First, we measured the T₁ relaxation behavior of solid dispersions. From the result, the T₁ relaxation time (T₁) changed according to the API content; the T₁ tended to increase with increasing API content because the T₁ value of amorphous IMC was longer than that of PVP. Next, we tried to monitor the amorphous-to-crystalline transformation of IMC in the solid dispersion during the thermal stress test. In the case of solid dispersion containing 90% IMC, a clear prolongation of the T₁ could be observed during the thermal stress test. From the powder X-ray diffraction patterns, the change in T₁ relaxation behavior must be caused by the IMC transformation from amorphous to crystalline. From these findings, we were successful in monitoring the IMC amorphous-to-crystalline transformation by the changes in T₁ relaxation behavior. Our findings led us to conclude that TD-NMR is a novel approach for the evaluation of crystalline state of APIs in solid dispersions.

Introduction

In the field of drug development in the solid oral dosage form, the number of active pharmaceutical ingredients (APIs) that are poorly soluble in water is increasing.^1,2) Various approaches^3–8) to improving solubility have been applied to develop the formulations.⁹⁾ Amorphization using the solid dispersion technique is regarded as a promising pharmaceutical technology for improving the solubility of drugs that are poorly soluble in water.⁸⁾ Solid dispersions are defined as formulations in which the API is dispersed in an inert matrix such as polymers.¹⁰⁾ The reason for using polymers as a carrier is to stabilize the amorphous API (inhibition of crystal transformation). However, the stabilization mechanism is as yet not fully understood. For the development of solid-dispersion formulations, maintaining the amorphous state throughout the shelf-life period is still a huge challenge.¹¹⁾ Thus, it is important to investigate the crystalline state of APIs in solid dispersions in more detail.

For the investigation of the crystalline state of APIs, the powder X-ray diffraction (PXRD) method has been the most commonly used.¹⁰⁾ We recently applied the time-domain NMR (TD-NMR) method to characterize the crystalline state of APIs.¹²⁾ TD-NMR is a benchtop instrument aimed at measuring the ¹H-NMR relaxation. Instead of the analysis of precise molecular structures like solid-state NMR, TD-NMR enables measurement of T₁ and T₂ relaxation times (T₁ and T₂, respectively) of the samples much more rapidly and easily. Furthermore, it enables the evaluation of both solid and liquid samples.¹³⁾ However, at present, the application of TD-NMR in the pharmaceutical research field has been limited. To our knowledge, it has only been used to evaluate the polymorphism of APIs,¹⁴⁾ the molecular mobility of hydration water in APIs,¹⁵⁾ and the miscibility of solid dispersions with APIs and polymers.¹⁶⁾

In a previous study, we demonstrated the usefulness of the TD-NMR method in the characterization of the crystalline state of APIs including carbamazepine and indomethacin (IMC).¹²⁾ The study confirmed that T_1, the T₁ measured by TD-NMR, is an effective parameter for distinguishing between the crystalline and amorphous states; the T₁ values of crystalline APIs were much longer than those of amorphous APIs. The ability to identify the crystalline and amorphous forms remained effective even if/when APIs were mixed physically with a solid dispersion carrier (polyvinylpyrrolidone, PVP). The biphasic T₁ relaxation behaviors were observed from the physical mixtures; furthermore, the proportion of APIs to carrier in the physical mixtures was completely quantified by the binary analysis. Furthermore, we conducted continuous monitoring of physical mixtures during the thermal stress test, and then clarified that TD-NMR can detect the transformation to crystalline forms of APIs incorporated into physical mixtures. In contrast to the wide range of findings concerning physical mixture samples, there is scant data on solid dispersion. Although monophasic T₁ relaxation behavior was found from the solid dispersion consisting of 20% APIs and 80% PVP, further investigation is warranted. In particular, changes in the relaxation behavior that are accompanied by a change in composition and crystal transformation need to be investigated in details.

Against this background, the present study had the following objectives: (1) to determine whether the T₁ value of solid dispersions varies with the amount of API, and (2) to determine whether T₁ is changed by the transformation to the crystalline form. For sample preparation, solid dispersions with different proportions of IMC and PVP were prepared, and their T₁ relaxation behaviors were thoroughly investigated. We also monitored the transformation of amorphous APIs into the crystalline state in a solid dispersion. The findings lead us to conclude that TD-NMR is a novel approach for the evaluation of the crystalline state of APIs in solid dispersions.

Experimental

Materials

Crystalline IMC (minimum purity 98.0%) was purchased from TCI (Tokyo, Japan). PVP K30 (Kollidon^®30) was obtained from BASF (Ludwigshafen, Germany). Other chemicals used were of reagent grade.

Sample Preparation

Amorphous solid dispersions were prepared by the melt-quenching method¹⁷⁾ following the solvent evaporation method. Designated amounts of IMC and PVP, (20 : 80, 40 : 60, 60 : 40, 80 : 20, 85 : 15, 90 : 10, 95 : 5, 99 : 1 (w/w)), were dissolved in ethanol/dichloromethane, 1 : 1 (v/v), at a total concentration of 100 mg/mL, and the solvent was removed using an evaporator (Rotavapor^® R-134; Büchi, Flawil, Switzerland) at 40–50°C. The resulting samples were further treated by the melt-quenching method to ensure that all incorporated IMC would be amorphous. Namely, all the obtained samples were placed on a heated polytetrafluoroethylene sheet at ca. 160°C for a few minutes, and then the melts were quench-cooled by immersion in liquid nitrogen. The resulting samples were lightly ground in a mortar, then passed through a 500 µm sieve. As a comparative control, PVP and amorphous API were also prepared by the same method as for the solid dispersions.

PXRD Measurement

PXRD patterns of all samples were obtained using D8 DISCOVER (Bruker BioSpin Corp., Billerica, MA, U.S.A.) with Cu-Kα radiation (λ = 0.154 nm). The count scanning rate was 0.2 s/step and the scanning angle was in the range 2θ = 5–50°.

TD-NMR Measurement

The ¹H T₁ relaxation behaviors of the samples were measured by TD-NMR using a Bruker minispec mq20 (Bruker BioSpin Corp.) at a ¹H frequency of 20 MHz at 25°C. The solid echo sequence was used for the measurement, and recycle delays in the range 0.02–80 s were used. After acquisition of the free induction decay with different recycle delays, a saturation recovery curve of the T₁ relaxation was obtained based on the signal intensity at equilibrium, I₀. The T₁ was also calculated using the TD-NMR Analyze Software. The T₁ relaxation data were calculated according to Eq. (1)¹⁴⁾:

  

(1)

where I(t) and I₀ are the signal intensities at time t and equilibrium, respectively; t is the time delay interval used during the NMR relaxation measurement, and T₁ is the T₁ relaxation time, which is a time constant that represents the recovery speed from the I(t) component to the equilibrium I₀. Equation (2) is a mathematical transformation of Eq. (1) for linearizing.

  

(2)

Results

PXRD Characterization of APIs Incorporated into Solid Dispersions

Figure 1 shows the PXRD patterns of samples. The powdered IMC that was used for preparation of the amorphous powder showed a γ-pattern,¹⁸⁾ whereas the amorphous IMC presented a halo pattern. For the solid dispersions, all samples showed the halo pattern: thus, the IMC incorporated into the sample was proven to be in the amorphous state.

Fig. 1. PXRD Patterns of (a) IMC-Containing Powders and (b) Solid Dispersions of IMC

T₁ Relaxation Behaviors of Solid Dispersions

This study measured the T₁ relaxation behaviors of amorphous solid dispersions. After the acquisition of NMR signal intensities with different recycle delays, these values were transformed according to Eq. (1) to calculate the T₁: the inverse of the slope corresponds to the T₁ of each sample. Figure 2a shows the T₁ relaxation behavior of crystalline and amorphous IMC, and of PVP, and Fig. 2b shows the T₁ relaxation behavior of solid dispersions. All samples showed a monophasic T₁ relaxation behavior. The T₁ values of the crystalline and amorphous IMC powders, and of PVP were 4.00, 1.07, and 0.27 s, respectively. A dose-dependent change was observed for the T₁ values of the solid dispersions: T₁ became longer with increasing API content (Fig. 2c); the T₁ values of solid dispersions containing 20, 40, 60, 80, 85, 90, 95, 99% IMC were 0.50, 0.74, 1.05, 1.13, 1.27, 1.21, 1.26, and 1.11 s, respectively. For the solid dispersions, their T₂ values were longer than the value for PVP, but shorter than the values for amorphous IMC: 0.013, 0.009, 0.014, and 0.008 ms for solid dispersions containing 90, 40%, amorphous IMC, and PVP, respectively (see Supplementary Materials, Fig. S1).

Fig. 2. (a) T₁ Relaxation of IMC-Containing Powders, (b) Solid Dispersions, and (c) T₁ Relaxation Time of Solid Dispersions According to IMC Content

To calculate T₁ relaxation times, the observed NMR signals were transformed (a, b) according to Eq. (2).

Continuous Monitoring of the IMC Amorphous-to-Crystalline Transformation in Solid Dispersion during the Thermal Stress Test

We evaluated the IMC amorphous-to-crystalline transformation in the solid dispersions by TD-NMR and PXRD. A thermal stress test was conducted at 60°C for this purpose. The T₁ relaxation behaviors before and after the thermal stress test were compared in detail by using solid dispersions containing 40 and 90% IMC as test samples (Figs. 3, 4a). For comparative control, the T₁ relaxation behavior of PVP and amorphous IMC was also monitored (Figs. 4b, c). For the solid dispersion containing 90% IMC, the slope of the plot in the graph was greatly extended with prolonged storage period, resulting in a longer T₁: values of 1.21, 1.50, 1.52, 1.73, and 1.88 s were obtained for the initial sample, and after storing the samples for 3, 7, 14, and 28 d, respectively. In contrast, the solid dispersion containing 40% IMC hardly showed any change in T₁ (Fig. 4a).

Fig. 3. T₁ Relaxation Behavior of Solid Dispersion Containing 90% IMC during Crystal Transformation of Amorphous IMC in the Solid Dispersions with PVP

During the thermal stress test, solid dispersion containing 90% IMC was stored at 60°C for (a) 3 d, (b) 7 d, (c) 14 d, and (d) 28 d.

Fig. 4. T₁ Relaxation Behavior during the Thermal Stress Test

T₁ Relaxation behaviors of solid dispersions measured at regular intervals, containing (a) 40% IMC, (b) amorphous IMC, and (c) PVP.

The PXRD pattern during the thermal stress test at 60°C up to 28 d is shown in Fig. 5. In accordance with the change in the T₁ relaxation behavior (Fig. 3a), the solid dispersion containing 90% IMC showed the crystalline peaks from the PXRD pattern during the test (Fig. 5a). In contrast, the solid dispersion with 40% IMC showed a halo pattern during the test (Fig. 5b). From the analysis of the PXRD patterns, the crystallinity degree of the solid dispersion containing 90% IMC increased steadily with prolonging the experiment. The amorphous IMC showed crystalline diffraction peaks after the test at 3 d (Fig. 5c), and PVP showed a halo pattern after the test (Fig. 5d).

Fig. 5. PXRD Patterns Measured during the Thermal Stress Test at Regular Intervals

(a) Solid dispersions containing 90% IMC, (b) solid dispersions containing 40% IMC, (c) amorphous IMC, and (d) PVP.

Discussion

In a previous study, we examined the T₁ relaxation behavior of mixtures of APIs and PVP and found that biphasic relaxation patterns were observed from the physical mixtures.¹²⁾ We note that the weight ratios of PVP and API were quantified by binary analysis of their T₁ saturation recovery curves. In addition to the detailed investigation on the mixtures, the previous study examined the T₁ relaxation behavior of a solid dispersion containing 20% API. In that experiment, the test solid dispersion showed a monophasic relaxation behavior, which was quite different from the relaxation behavior of physical mixtures. Several research groups have conducted T₁ relaxation studies to investigate the miscibility of components in dispersion systems^16,19) and found that monophasic and biphasic relaxation patterns are related to the domain size of the sample. Based on these findings, our previous study concluded that on the nanoscale the APIs and PVP were completely mixed with each other in the solid dispersions and then formed homogeneous phases.

The present study investigated the T₁ relaxation behavior of solid dispersion in more detail. First, we prepared solid dispersions with different IMC contents and then investigated the T₁ relaxation behavior. As shown in Fig. 2, the T₁ relaxation behavior of all the amorphous solid dispersions showed a monophasic pattern regardless of the IMC content, indicating that IMC and PVP were fully mixed with each other and then formed homogeneous phases. Furthermore, the T₁ values changed with the IMC content; there was a trend of extending the T₁ with increasing IMC content. This seems reasonable because the T₁ value for amorphous IMC is longer than that for PVP. Aso et al. investigated the T₁ and T₁ρ relaxation behaviors of solid dispersions with API content and polymer species using TD-NMR,¹⁶⁾ and succeeded in evaluating the miscibility of the solid dispersion with different API content. The present study agrees well with their findings. A similar change was also observed in the T₂ relaxation behavior; with increasing IMC content, the T₂ values were extended. This seems reasonable because the T₂ value for amorphous IMC is longer than that for PVP.

In the next phase of this study, we attempted to monitor the crystal transformation of IMC in solid dispersion during the thermal stress test (Fig. 3). In the previous study, the test solid dispersion contained 20% APIs. The amorphous APIs incorporated into the test solid dispersion were so stable that the crystalline transformation could not occur under the thermal stress test at 60°C for 28 d. By taking this result into account, the present study tested the solid dispersions containing 40 and 90% amorphous IMC as stable and unstable samples. For the solid dispersion containing 40% IMC, no change in either T₁ relaxation behavior or PXRD pattern was observed after the thermal stress test. For the solid dispersion containing 90% IMC, a clear prolongation of the T₁ value could be observed during the thermal stress test (Fig. 3). From the PXRD patterns, the change in T₁ relaxation behavior must be caused by the transformation of amorphous IMC into crystalline form. Furthermore, the results suggest a possibility that PVP inhibits the crystal transformation not only in terms of amount, but also in terms of rate. That is, because the crystal transformation of the solid dispersion containing 90% IMC was clearly slow when compared to amorphous IMC: the gradual change in the T₁ relaxation behavior and PXRD pattern during the thermal stress test was quite different from that of amorphous IMC. These results were probably caused by the following mechanism. In the case of the solid dispersion containing 40% IMC, the whole IMC amount fully interacts with PVP. Meanwhile, in the case of the solid dispersion containing 90% IMC, some portion of IMC could not interact with PVP and formed the amorphous IMC-enriched domain; the amorphous IMC-enriched domain predominantly turned into crystalline form by the thermal stress test. We also note that the crystal transformation rate in the solid dispersion was much slower than that of the pure amorphous IMC. From these findings, we confirmed that the measurement of T₁ relaxation behavior by TD-NMR was effective in monitoring the amorphous-to-crystalline transformation occurring in the solid dispersion.

Conclusion

In the initial phase of this study, we investigated the T₁ relaxation behavior of solid dispersion with change in the amount of API. T₁ was consistently prolonged with increasing amount of API. Every solid dispersion showed a monophasic relaxation behavior, indicating that APIs and PVP were completely mixed and then formed a homogeneous phase structure. Furthermore, we succeeded in monitoring the transformation of the amorphous to crystalline form by T₁ relaxation using TD-NMR. The prolongation of T₁ was accompanied by the crystal transformation of APIs in the sample. To our knowledge, the present study is the first report on the evaluation of the amorphous-to-crystalline transformation of API in solid dispersion by TD-NMR. This study demonstrated a proof-of-concept that TD-NMR is effective in evaluating the crystalline state of API incorporated into solid dispersion; hence, the value of TD-NMR for the development of solid dosage forms is expected to be greatly enhanced.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Number JP16K08192 and JSPS Core-to-Core Program, B. Asia-Africa Science Platforms.

Conflict of Interest

The authors declare that they have no financial or competing interests concerning this manuscript. The Department of Pharmaceutical Technology, University of Toyama, is an endowed department supported by an unrestricted grant from the Nichi-Iko Pharmaceutical Co. (Toyama, Japan).

Supplementary Materials

The online version of this article contains supplementary materials.

References

• 1) Lipinski C. A., Lombardo F., Dominy B. W., Feeney P. J., Adv. Drug Deliv. Rev., 23, 3–25 (1997).
• 2) Zhang M. Q., Wilkinson B., Curr. Opin. Biotechnol., 18, 478–488 (2007).
• 3) Berge S. M., Bighley L. D., Monkhouse D. C., J. Pharm. Sci., 66, 1–19 (1977).
• 4) Rabinow B. E., Nat. Rev. Drug Discov., 3, 785–796 (2004).
• 5) Lubach J. W., Xu D., Segmuller B. E., Munson E. J., J. Pharm. Sci., 96, 777–787 (2007).
• 6) Loftsson T., Duchêne D., Int. J. Pharm., 329, 1–11 (2007).
• 7) Rautio J., Kumpulainen H., Heimbach T., Oliyai R., Oh D., Järvinen T., Savolainen J., Nat. Rev. Drug Discov., 7, 255–270 (2008).
• 8) Vasconcelos T., Sarmento B., Costa P., Drug Discov. Today, 12, 1068–1075 (2007).
• 9) Kawabata Y., Wada K., Nakatani M., Yamada S., Onoue S., Int. J. Pharm., 420, 1–10 (2011).
• 10) Kawakami K., Adv. Drug Deliv. Rev., 64, 480–495 (2012).
• 11) Laitinen R., Löbmann K., Strachan C. J., Grohganz H., Rades T., Int. J. Pharm., 453, 65–79 (2013).
• 12) Okada K., Hirai D., Kumada S., Kosugi A., Hayashi Y., Onuki Y., J. Pharm. Sci., 108, 451–456 (2019).
• 13) Metz H., Mäder K., Int. J. Pharm., 364, 170–175 (2008).
• 14) Stueber D., Jehle S., J. Pharm. Sci., 106, 1828–1838 (2017).
• 15) Schumacher S. U., Rothenhäusler B., Willmann A., Thun J., Moog R., Kuentz M., J. Pharm. Biomed. Anal., 137, 96–103 (2017).
• 16) Aso Y., Yoshioka S., Miyazaki T., Kawanishi T., Tanaka K., Kitamura S., Takakura A., Hayashi T., Muranushi N., Chem. Pharm. Bull., 55, 1227–1231 (2007).
• 17) Strachan C. J., Taday P. F., Newnham D. A., Gordon K. C., Zeitler J. A., Pepper M., Rades T., J. Pharm. Sci., 94, 837–846 (2005).
• 18) Surwase S. A., Boetker J. P., Saville D., Boyd B. J., Gordon K. C., Peltonen L., Strachan C. J., Mol. Pharm., 10, 4472–4480 (2013).
• 19) Yuan X., Sperger D., Munson E. J., Mol. Pharm., 11, 329–337 (2014).