(T1430-11-62) What Is That Peak in Hydrocodone Bitartrate USP Reference Standard?

Purpose: According to the United States Pharmacopeia's (USP) policy statement on Reference Standard Uses and Applications, USP Reference Standards (RS) are primarily intended for analytical, or laboratory applications outlined within the USP compendia. However, USP permits the utilization of a USP RS beyond its designated USP compendial uses, placing the responsibility on the user to ensure the appropriateness of the Reference (RS) for a non-USP application. During a method transfer to our laboratory of a new method for hydrocodone bitartrate tablets, we observed a small peak eluting before the main peak of the USP-RS for hydrocodone bitartrate (HCBT). Mass Spectrometry (MS) and Nuclear Magnetic resonance (NMR) and were used to understand the nature of this peak.

Methods: Chromatography was done on Waters®AcQuity system equipped with Waters®TUV Detector at UV 245nm using Luna 3 µm C18(2) 100 Å, LC Column 150 x 4.6. The mobile phase consisted of Phosphate buffer adjusted to pH 4  with phosphoric acid and mixed with methanol in 10:1 ratio in isocratic mode. USP RS of HCB lot # N2I034 was used in the method. NMR experiments were recorded on a Varian INOVA 400 MHz NMR spectrometer at room temperature. Chemical shifts are reported in parts per million. Mass spectra were obtained on an Advion CMS-L mass spectrometer.

Results: Figure 1a shows a typical chromatogram of HCBT with HCBT main peak at a retention time of 4.52 minutes and another unknown peak at a retention time 2.42 minutes. The two peaks look well separated and the asymmetry factor for the HCBT peak is 1.25 which is within the acceptable range A closer zoom into the same chromatogram, Figure 1b, reveals what is commonly referred to as a saddle or batman peak. In other words, the two peaks are not separated; instead, there is a non-baseline plateau between them. This batman peak indicates the presence of some equilibrium between the two peaks. Previous work by Brogle et al and Caldwell et al.[1,2], demonstrated that peak-fronting observed with oxycodone was due to the addition of water and methanol to the C-6 ketone via NMR and MS. We hypothesized that a similar phenomenon was occurring with hydrocodone. To confirm this, we performed both NMR and MS experiments. Proton NMR showed an additional peak at 4.07 ppm (1H), representing the α-hydrogen adjacent to the geminal-diol. This is downfield from the parent ketone (4.52 ppm), which is expected due to inductive effects. Additionally, a clear formation of a new peak at 96.8 ppm (¹³C) is indicative of a quaternary carbon (e.g., the hydrate), which aligns well with previous studies involving oxycodone, Figure 3 Tautomerization was not observed via NMR, as no new peaks were indicative of a carbon-carbon double bond. MS experiments were run in Methanol (MeOH)/0.1% formic acid to mimic HPLC conditions in which saddle peak was observed. The data show a clear formation of the geminal-diol (m/z=332), Figure 3a. Similarly, the formation of the geminal-diol (or hydrate) was observed when acetonitrile (ACN)/0.1 formic acid (m/z=319) was used along with an ACN-adduct (m/z=342), Figure 3b. The formation of ACN-adduct formation @ 341-342 in addition to the geminal-diol @ m/z=319 may explain why acetonitrile shows a broader HCBT peak than that observed in methanol. After the peak was identified as a gemdinal diol, the same conditions of the method were used at different column temperatures. The peak ratio was found to change with temperature, being higher at lower column temperatures and decreasing as temperature increases. The effect of different column temperature on the peak parameters is show in Table 1.

Conclusion: In this work we showed that a USP RS can show extraneous peaks not because of presence of impurity but because of interaction with mobile phase components leading to the formation of some forms of hydrates or other possible conformers. In the case of hydrocodone bitartrate there is a dynamic equilibrium between the geminal-diol peak and the main peak with a non baseline plateau region in between. This region may hide degradation products that may be found in a sample hydrocodone. The batman or saddle peak shape may not be observed in low sample concentrations and peak symmetry may not capture the peak shape distortion. This can affect the integration of the peak and may hide other components in the formulations. This work shows that caution should be taken when using USP-RS outside USP compendia, especially with peak shapes that might not be detected by system suitability parameters, e.g., peak asymmetry factor.

References: Broglé, K., Ornaf, R. M., Wu, D., & Palermo, P. J. (1999). Peak fronting in reversed-phase high-performance liquid chromatography: a study of the chromatographic behavior of oxycodone hydrochloride. Journal of pharmaceutical and biomedical analysis, 19(5), 669-678.
Caldwell G, Gauthier AD, Villani FJ, Maryanoff CA, Leo G. Carbonyl hydration of oxymorphone and oxycodone hydrochloride. Tet Lett., 1991, 32(31), 3763-3766

Acknowledgements: The authors would like to thank Heather Boyce, Lead Pharmacokineticist at the Office of Generic Drugs, US FDA, for her questions which inspired this work.


Typical chromatogram of hydrocodone bitartrate at a concentration of 100 mg/mL. The chromatogram shows a peak for hydrocodone at 4.52 minutes and an unknown peak at 2.42 minutes (a). A closer zoom into the chromatogram of HCBT (b).


Figure 2: NMR experiment: C13 NMR of hydrocodone bitartrate in D2O (a). The newly formed quaternary carbon is visible at 97.008 ppm but is in low concentration compared with parent hydrocodone. Extra peak values are present due to the increased zoom necessary to visualize new peaks.(b) Zoomed in snapshot of the top spectrum to show the presence of a small amount of hydrocodone geminol-diol (hydrocodone hydrate).


Table 1: MS spectrum of Hydrocodone in Methanol (MeOH)/0.1% formic acid (a) and MS spectrum of Hydrocodone in Acetonitrile (ACN)/0.1% formic acid (b)rtrate and it geminal diol under different column temperatures