Quantitative ³¹P-NMR for the Purity Determination of the Organophosphorus Compound Brigatinib and Its Method Validation

Abstract

The spectrum of ³¹P-NMR is fundamentally simpler than that of ¹H-NMR; consequently identifying the target signal(s) for quantitation is simpler using quantitative ³¹P-NMR (³¹P-qNMR) than using quantitative ¹H-NMR (¹H-qNMR), which has been already established as an absolute determination method. We have previously reported a ³¹P-qNMR method for the absolute determination of cyclophosphamide hydrate and sofosbuvir as water-soluble and water-insoluble organophosphorus compounds, respectively. This study introduces the purity determination of brigatinib (BR), an organophosphorus compound with limited water solubility, using ³¹P-qNMR at multiple laboratories. Phosphonoacetic acid (PAA) and 1,4-BTMSB-d₄ were selected as the reference standards (RSs) for ³¹P-qNMR and ¹H-qNMR, respectively. The qNMR solvents were chosen based on the solubilities of BR and the RSs for qNMR. CD₃OH was selected as the solvent for ³¹P-qNMR measurements to prevent the influence of deuterium exchange caused by the presence of exchangeable intramolecular protons of BR and PAA on the quantitative values, while CD₃OD was the solvent of choice for the ¹H-qNMR measurements to prevent the influence of water signals and the exchangeable intramolecular protons of BR and PAA. The mean purity of BR determined by ³¹P-qNMR was 97.94 ± 0.69%, which was in agreement with that determined by ¹H-qNMR (97.26 ± 0.71%), thus indicating the feasibility of purity determination of BR by ³¹P-qNMR. Therefore, the findings of this study may provide an effective method that is simpler than conventional ¹H-qNMR for the determination of organophosphorus compounds.

Introduction

Quantitative NMR (qNMR) is used in many fields as an absolute quantitative method to enable an accurate quantitation of analytes without the requirement of their own reference standards (RSs). Furthermore, qNMR is an analytical method that can yield results with traceability to the International System of Units (SI). In particular, ¹H-qNMR is a suitable method for evaluating RSs. We have previously determined the purity of reagents used as RSs for HPLC assays for crude drugs and Kampo formulations in the Japanese Pharmacopoeia (JP) via an accurate qNMR with an internal reference substance (AQARI).^1–6) The results revealed that 24 reagents that were evaluated by ¹H-qNMR were listed as RSs for HPLC assays of 50 crude drugs and Kampo formula extract monographs up to the publication of JP 18th Edition Supplement I.⁶⁾ Simultaneously, we established an optimized and reproducible ¹H-qNMR sample preparation method suitable for various properties of target compounds including hygroscopic substances.^7–9) The international standardization of qNMR was also initiated based on a proposal from Japan. Following the publication of the ISO standard 24583 : 2022 “Quantitative nuclear magnetic resonance spectroscopy—Purity determination of organic compounds used for foods and food products—General requirements for ¹H NMR internal standard method” in December 2022, the requirements for purity determination by qNMR have been agreed upon internationally.¹⁰⁾

However, the selection of quantitative signals is occasionally difficult in ¹H-qNMR, which can be attributed to their overlap and complexity depending on the purity and structure of the target compound. We thus hypothesized that the measurement of other nuclei might be a solution to this problem. The ³¹P nucleus was chosen owing to its superior sensitivity and high probability of being present in pharmaceuticals. Selecting a quantitative signal in ³¹P-qNMR is simple since most pharmaceuticals contain only one phosphorus nucleus or two phosphorous nuclei in their molecule. Furthermore, the shape of the ³¹P-qNMR main signal is simple. The ¹H-qNMR method has been already established, and the conditions for performing it have been described in the JP.^1,6) To the best of our knowledge, ³¹P-qNMR has been rarely reported in the literature.

We have previously reported absolute quantification by ³¹P-qNMR of cyclophosphamide hydrate (CP, listed in the JP) and sofosbuvir (SOF), which are water soluble and insoluble organophosphorus compounds, respectively.^11,12) This study investigated the absolute determination of brigatinib (BR), an organophosphorus anti-cancer drug with poor water solubility (Fig. 1) at ten different laboratories. The optimal conditions for the ³¹P-qNMR measurement were determined. The BR purity determined by ³¹P-qNMR was compared with that obtained by an established ¹H-qNMR method. The parameters employed to establish the absolute determination method by ³¹P-qNMR are also described.

Fig. 1. Structures of an Organophosphorus Compound, Brigatinib (BR), Quantitative NMR (qNMR) Certified Reference Materials Traceable to the SI Reference Standards of ¹H- (1,4-BTMSB-d₄) and ³¹P-NMR (Phosphonoacetic Acid (PAA))

Experimental

Table 1 presents the conditions (including sample preparation) employed in the ¹H-qNMR and ³¹P-qNMR measurements at the 10 laboratories (A–J). The chemicals and conditions employed for the sample preparation in each laboratory are provided in Table S1 (Supplementary Materials), while the conditions employed by each laboratory for the ¹H-qNMR and ³¹P-qNMR measurements are summarized in Tables S2 and S3 (Supplementary Materials), respectively. Further details on the experimental section including sample preparation (NMR validation test, moisture adsorption and desorption analysis, humidity conditions), equipment, and qNMR conditions can be found in the Supplementary Materials.

Table 1. Sample Preparation and Employed Conditions for the ¹H-qNMR and ³¹P-qNMR Measurements at the 10 Laboratories

Contents	No.	Conditions	1H	31P
NMR instrument	1	Spectrometer frequency*	400–700 MHz	162–243 MHz (1H 400–600 MHz)
qNMR sample preparation	2	Deuterated solvent	CD3OD	CD3OH
	3	Reference standard for qNMR	1,4-BTMSB-d4	PAA
	4	Hygroscopicity of analyte and reference standard for qNMR	[BR] Yes (Hygroscopic at a relative humidity >30%)	[BR] Yes (Hygroscopic at a relative humidity >30%)
			[1,4-BTMSB-d4] No	[PAA] Yes (Deliquesced at a relative humidity >50%)
	5	Solution stability of analyte and reference standard for qNMR	Stable up to 23 h	Stable up to 23 h
qNMR measurement	6	Relaxation times (T1, T2)	Not measured	[BR] T1: 0.97–1.33 s, T2: 0.45–0.63 s
				[PAA] T1: 0.94–1.05 s, T2: 0.17–0.18 s
	7	Acquisition time	4 s	2.1–4 s
	8	Relaxation delay time*	60 s	30 s
	9	Spectral width*(spectral center and spectral width at actual measurement)	20–25 ppm	100–110 ppm
	10	Pulse offset	5 ppm	30–30.5 ppm
	11	Integration width	Depends on each signal	Depends on each signal
	12	Signal shape, coupling	—	Including 31P-13C coupling
	13	Pulse angle*	90°	90°
	14	Scan times*	8–32 (SN > 100)	64–128 (SN > 100)
	15	Digital resolution*	0.25 Hz	Set at 2.5 times or more than half the width (0.25–0.47 Hz)
	16	Spinning*	No	No
	17	Decoupling*	Yes (13C)	Yes (1H)
	18	Decouple sequence	MPF8, MPF9	Waltz, Waltz 16, Waltz 65 (No-NOE)
	19	Measurement temperature *	24–30 °C	24–30 °C
	20	Measurement time	30–120 min (n = 3)	120–240 min (n = 3)

*Described in JP18.

Results and Discussion

Solvent and RSs for ¹H-qNMR and ³¹P-qNMR

In our previous studies and owing to the chemical shift and solubility of CP, D₂O was used as the solvent, while either KH₂PO₄ or O-phosphorylethanolamine was used as the RS for ³¹P-qNMR. Consequently, the choice of the RS and solvent is of utmost important in ³¹P-qNMR measurements.¹¹⁾ In the case of the SOF, phosphonoacetic acid (PAA) was used as the RS for ³¹P-qNMR in the protic solvent CD₃OD. The purity of the SOF determined by ³¹P-qNMR was 1.6% higher than that determined by ¹H-qNMR, thus indicating that using a protic solvent, such as CD₃OD, is inappropriate for ³¹P-qNMR because of deuterium exchange with PAA, which results in a small integrated intensity. Consequently, the purity of SOF was determined in an aprotic solvent dimethyl sulfoxide (DMSO)-d₆. When a target organophosphorus compound or RS with exchangeable protons in the molecule is used for purity determination by ³¹P-qNMR, an aprotic organic solvent such as DMSO-d₆ enables avoiding deuterium exchange, which affects the results.¹²⁾

Based on our previous results, a suitable RS for ³¹P-qNMR is the one that (1) is soluble in the same solvent as the target compound and (2) does not possess a chemical shift significantly different from that of the target compound. In the case of BR, PAA, an SI-traceable certified reference material (CRM), was the RS of choice for ³¹P-qNMR, because of its solubility characteristics and the ³¹P phosphorus chemical shift, while 1,4-BTMSB-d₄, an SI-traceable CRM, was the RS of choice for ¹H-qNMR owing to its solubility.

The solvents used for the measurements were then analyzed. Since BR is not soluble in DMSO-d₆, an aprotic solvent, CD₃OD was investigated as a possible solvent for the ³¹P-qNMR measurements. However, since BR and PAA have exchangeable protons in their molecules, deuterium exchange by CD₃OD was expected to affect the purity determined by ³¹P-qNMR. To avoid this effect, CD₃OH was selected as the solvent for ³¹P-qNMR measurements. The ³¹P chemical shift was referenced to PAA (δP: approximately 13.0 ppm in CD₃OH) at 0 ppm. The chemical shift of BR was as follows: δP of approximately 48.0 ppm (35.0 ppm in this study) in CD₃OH. On the other hand, CD₃OD was the solvent of choice for the ¹H-qNMR measurements because the water signal in CD₃OH may affect the quantitative signal.

Relaxation Delay Time for the ³¹P-qNMR Measurements

For quantitative NMR experiments, a relaxation delay time exceeding seven times that of T₁ is required for stabilization.¹³⁾ The T₁ values of the ³¹P signals of BR and PAA in CD₃OH were in the range of 0.97–1.33 and 0.94–1.05 s, respectively. Therefore, the relaxation delay time for ³¹P-qNMR spectroscopy was set to be at least 30 s. To obtain accurate integral values, the acquisition time was set so that the digital resolution was less than 1/2.5 times the half-width (2.1–4 s).

Hygroscopicities of BR and PAA

Prior to the qNMR measurements, the hygroscopicity of each of BR and PAA was determined by moisture adsorption and desorption analysis. In the relative humidity range of 0–80%, the weight of BR slightly varied (1% or less) at a relative humidity exceeding 30%. Therefore, BR weighing was performed in the relative humidity range of 30–50%. Furthermore, PAA weighing and sample preparation were performed at a relative humidity below 50% since it deliquesced and its weight drastically increased at a relative humidity exceeding 50%.

Stability of the Sample Solution for ³¹P-qNMR Measurements

The stability of the BR purity determined by ³¹P-qNMR was then monitored up to 23 h after preparation. Since the purity remained relatively stable up to 23 h after preparation (data not shown), the data obtained within 23 h of sample preparation was acceptable.

¹H- and ³¹P-qNMR Measurements in a Protic Solvent (CD₃OD and CD₃OH)

The ¹H-qNMR spectrum of BR in CD₃OD is shown in Fig. 2A. The enlarged spectra are shown in Fig. 2B, and the purity of BR determined by ¹H-qNMR is provided in Table 2. In the ¹H-qNMR spectrum of the mixture of BR and 1,4-BTMSB-d₄ in a protic solvent (CD₃OD), a broad impurity signal for water was observed at approximately δH 4.60 ppm (Fig. 2A). Considering the impurity, the multiplicity of coupling, and signal overlap in the 400 MHz NMR, the singlet signal of H-b (δH 7.78 ppm) of BR was selected as the quantitative signal due to its simplicity, sufficient separation, and height (Figs. 2A, 2B). The mean purity value of BR was 97.26 ± 0.71% as determined by ¹H-qNMR in the nine laboratories using 1,4-BTMSB-d₄ as the RS (Table 2).

Fig. 2. Characteristic ¹H-qNMR Spectrum of BR with 1,4-BTMSB-d₄ in CD₃OD (A), the Enlarged Spectra of BR (B)

Table 2. BR Purity (%) Determined by ¹H-qNMR in Nine Laboratories

Position		Laboratory									Total average and S.D. (%)
		A	B	D	E	F	G	H	I	J
7.78 ppm	Average (%)	97.34	95.93	96.84	97.18	97.75	97.37	97.15	98.59	97.17	97.26
	S.D. (%)	0.08	0.16	0.32	0.06	0.15	0.19	0.09	0.27	0.15	0.71

S.D.: Standard deviation

In the ³¹P-qNMR spectrum of BR (Fig. 3), the main signal of BR at 35.0 ppm was used as the quantitative signal. The ³¹P-¹³C coupling of BR (¹J_³¹P-¹³C coupling of PO-Phenyl = 97 Hz and ¹J_³¹P-¹³C coupling of PO-(CH₃)₂ = 72 Hz) was observed (Fig. 3, signals A and B). Therefore, the quantitative values were calculated over the integration range, including the ³¹P-¹³C coupling of BR. Additionally, ¹³C satellite signals were observed in PAA (¹J_³¹P-¹³C coupling, 120 Hz) (Fig. 3, signal C), and the coupling signals were integrated, including the PAA signal. The mean purity value of BR by ³¹P-qNMR was 97.94 ± 0.69% in the 10 laboratories (Table 3), which was in agreement with that determined by ¹H-qNMR (97.26 ± 0.71%; Table 2). These results indicated that even when the aprotic solvent cannot be selected based on the solubility of the analyte and RS for qNMR, quantification is possible by selecting deuterated protic solvents with different degrees of deuteration, such as CD₃OD and CD₃OH, for ¹H-qNMR and ³¹P-qNMR, respectively.

Fig. 3. Characteristic ³¹P-qNMR Spectrum of BR with PAA in CD₃OH

Signals A: ¹J_³¹P-¹³C coupling of PO-Phenyl (97 Hz) of BR and B: ¹J_³¹P-¹³C coupling of PO-(CH₃)₂ (72 Hz) of BR. C: ¹J_³¹P-¹³C coupling of PO-CH₂ (120 Hz) of PAA.

Table 3. BR Purity Determined by ³¹P-qNMR in the 10 Laboratories

Position		Laboratory										Total average and S.D. (%)
		A	B	C	D	E	F	G	H	I	J
35.0 ppm	Average (%)	97.71	97.73	97.87	99.09	98.02	98.40	97.42	98.11	98.54	96.54	97.94
	S.D. (%)	0.62	0.28	0.19	0.25	0.51	0.37	0.24	0.28	0.52	0.34	0.69

S.D.: Standard deviation

Conclusion

This study reported the absolute determination of the non-water-soluble organophosphorus compound BR using ³¹P-qNMR. Based on the solubilities of BR and the RSs, CD₃OH and CD₃OD were selected as the solvents for the ³¹P-qNMR and ¹H-qNMR measurements, respectively. The data revealed that the mean BR purity determined by ³¹P-qNMR (97.94 ± 0.69%) was in agreement with that determined by ¹H-qNMR (97.26 ± 0.71%), thus indicating the possibility to determine BR using ³¹P-qNMR. It was thus concluded that, similar to ¹H-qNMR, ³¹P-qNMR could be employed for the absolute determination of organophosphorus compounds in the presence of an appropriate solvent and RS. Therefore, the findings of this study may provide an effective method that is simpler than conventional ¹H-qNMR for the purity determination of organophosphorus compounds. We are currently investigating other qNMR techniques, such as ¹⁹F-qNMR, for absolute determination of organic fluorine compounds.

Acknowledgments

This study was supported by the Japan Agency for Medical Research and Development (AMED) (Grant number: JP22mk0101220).

Conflict of Interest

Takanori Komatsu, Katsuo Asakura, Takako Suematsu and Hitomi Muto are employees of JEOL Ltd., Mika Murabayashi is an employee of Takeda Pharmaceutical Co., Ltd., Kengo Kobayashi and Taeko Shinozaki are employees of Daiichi Sankyo Co., Ltd., Yoshinori Fujimine is an employee of Otsuka Pharmaceutical Co., Ltd., Katsuya Ofuji and Hitoshi Shimizu are employees of Chugai Pharmaceutical Co., Ltd., Takashi Hasebe, Yumi Asai and Eri Ena are employees of Eisai Co., Ltd., Kohei Kiyota and Kazuhiro Fujita are employees of Shionogi & Co., Ltd., Yoshinobu Makino is an employee of Juzen Chemical Corp., Toru Miura and Yasuhiro Muto are employees of FUJIFILM Wako Pure Chemical Corporation, Takashi Goto and Masu Yasuda are employees of Nippon Shinyaku Co., Ltd., and Tomohiko Ueda is an employee of Sumitomo Pharma Co., Ltd., Ai Kohama is an employee of Pharmaceutical and Medical Device Regulatory Science Society of Japan (PMRJ). Nahoko Uchiyama, Junko Hosoe, Naoki Sugimoto, Kyoko Ishizuki, Tatsuo Koide and Yukihiro Goda have no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) The Ministry of Health, Labour and Welfare, Japan. “The Japanese Pharmacopoeia Eighteen Edition (2021).”: ‹https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000066597.html›, cited 2 September, 2023.
• 2) Hosoe J., Sugimoto N., Goda Y., Pharmaceutical and Medical Device Regulatory Science, 41, 960–970 (2010).
• 3) Hosoe J., Sugimoto N., Suematsu T., Yamada Y., Hayakawa M., Katsuhara T., Nishimura H., Goda Y., Pharmaceutical and Medical Device Regulatory Science, 43, 182–193 (2012).
• 4) Hosoe J., Sugimoto N., Suematsu T., Yamada Y., Miura T., Hayakawa M., Suzuki H., Katsuhara T., Nishimura H., Kikuchi Y., Yamashita T., Goda Y., Pharmaceutical and Medical Device Regulatory Science, 45, 243–250 (2014).
• 5) Goda Y., Pharmaceutical and Medical Device Regulatory Science, 48, 615–619 (2017).
• 6) The Ministry of Health, Labour and Welfare, Japan. “The Japanese Pharmacopoeia Eighteen Edition Supplement I (2022).”: ‹https://www.mhlw.go.jp/content/11120000/001111831.pdf›, cited 2 September, 2023.
• 7) Uchiyama N., Hosoe J., Miura T., Sugimoto N., Ishizuki K., Yamada Y., Iwamoto Y., Suematsu T., Komatsu T., Maruyama T., Igarashi Y., Higano T., Shimada N., Goda Y., Yakugaku Zasshi, 140, 1063–1069 (2020).
• 8) Uchiyama N., Hosoe J., Miura T., Sugimoto N., Ishizuki K., Yamada Y., Iwamoto Y., Suematsu T., Komatsu T., Maruyama T., Igarashi Y., Higano T., Shimada N., Goda Y., Chem. Pharm. Bull., 69, 26–31 (2021).
• 9) Uchiyama N., Hosoe J., Sugimoto N., et al., Chem. Pharm. Bull., 69, 118–123 (2021).
• 10) ISO 24583:2022 Quantitative nuclear magnetic resonance spectroscopy—Purity determination of organic compounds used for foods and food products—General requirements for ¹H-NMR internal standard method: ‹https://www.iso.org/standard/78997.html›, issued, 19 December, 2022.
• 11) Uchiyama N., Hosoe J., Sugimoto N., et al., Chem. Pharm. Bull., 69, 630–638 (2021).
• 12) Uchiyama N., Kiyota K., Hosoe J., et al., Chem. Pharm. Bull., 70, 892–900 (2022).
• 13) qNMR primary guide. “A guide to quantitative analysis for beginners-from basics to practice,” qNMR primary guide working group, Kyoritsu Publication, Tokyo, Japan, 2015, pp. 96–97.