Appendix: Common Deficiencies Related to Solution NMR in the Type-II DMFs

 APPENDIX 

 to

Characterization of Small-Molecule Drug Substances in Type II DMFs Supporting ANDAs

Part I: Solution Nuclear Magnetic Resonance Spectroscopy

In the February 2017 issue of Pharmaceutical Technology, the authors of "Characterization of Small-Molecule Drug Substances in Type II DMFs Supporting ANDAs, Part I: Solution Nuclear Magnetic Resonance Spectroscopy,” presented the following examples of seven subcategory deficiencies related to solution nuclear magnetic resonance (NMR) in the Type-II drug master files (DMFs):

No proton/carbon assignments

• You have only provided ¹H NMR and ¹³C NMR spectra without any interpretation.  Please provide a detailed structural elucidation to confirm the structure of drug substance. For ¹H NMR, please provide proton numbers, multiplicity, ³J_H-H coupling constant if applicable, and assignment to all the signals in a tabular format.

• Please provide ¹H NMR, ¹⁹F NMR, ¹³C NMR, ¹H-¹H COSY and ¹H-¹³C HMQC spectroscopic data to ambiguously confirm the proposed structure of the drug substance.

Incorrect or incomplete assignments

• You have provided ¹H NMR spectrum and data assignment for the drug substance.  However, your assignment on Multiplicity, Integration, and Position of proton are not in agreement with the proposed structure. Please provide ¹H-¹H COSY and ¹H-¹³C HMQC data for the drug substance to verify the elucidation of drug substance.

• There are several issues regarding the ¹H NMR elucidation of Impurity-D: (i) the 6-position is a quaternary carbon that has been assigned a proton with a peak of multiplet; (ii) two protons at 5-position: one has been assigned as a doublet (J = 8.3 Hz) while another was assigned as a singlet; and (iii) the assignment for the proton at 8-positon is missing. Please explain and correct them accordingly.

• Please explain the peak at δ 13.4 ppm on the ¹H NMR spectrum of the drug substance, which does not correspond to the proposed structure. Please clarify if the compound is free base or a salt.

No comparison to reference standards

• Please provide comparison of IR, ¹H NMR and ¹³C NMR spectra of the drug substance with the USP Reference Standard including peak assignments, good resolution and expansions of the NMR scans at appropriate chemical shifts. Both individual and overlay presentations are preferred.

• Considering the complexity of the drug substance with eight chiral centers, we recommend you to provide comparison of ¹H NMR (with expansion) between USP/EP Reference Standard and your in-house drug substance working standard to verify the proposed structure of drug substance. Overlaid spectra are preferred for better visualization.

• You did not assign the proton NMR peaks of working standard and production lot at 10.70, 4.39, 4.38, and ~1.3 ppm.  Please assign these peaks accordingly. In addition, we recommend that you compare ¹H NMR of your in-house working standard with that of a USP Reference Standard. For a better visualization, please provide overlaid spectra of your in-house working standard and the USP Reference Standard.

Spectra not matched with reference standards (RS)

• The ¹HNMR spectra of drug substance and USP Reference Standard are different in chemical shifts and splitting patterns from the range of 4.0 to 4.5 ppm. Please explain this discrepancy. In order to eliminate concentration effect on spectra resolution, you may consider providing a ¹H NMR spectrum of a mixed sample of drug substance and corresponding USP Reference Standard.

• The spectra of USP Reference Standard and your test sample are not comparable since the USP Reference Standard is a free base while your test sample is a hydrochloride salt. Please provide overlaid ¹H NMR spectra of USP Reference Standard and your test sample in the same form.

• The ¹H NMR spectra of USP Reference Standard in CDCl₃ and in-house working standard in CD₃OD show different chemical shift, splitting pattern and intensity at 1.0-1.8 ppm and 2.6-2.8 ppm, respectively. In order to eliminate the solvent impacts on the spectra, please provide the ¹H NMR spectra in the same deturated solvent. For better visualization, please provide overlaid ¹H NMR spectra of USP Reference Standard A and in-house working standard B for better visual comparison.

Inadequate stereochemistry assignments

• The characterization data you provided so far are not sufficient to confirm the trans fused geometry of the drug substance. Please provide additional evidences to unambiguously determine the trans fused geometry of drug substance A. The evidences may include, but not limited to, vicinal coupling constants ³J_H-H coupling constants, NOE experiment or NOESY/ROESY.

• The API has several chiral centers which are inherited from procured starting material for which no synthetic chemistry and characterization data was provided. Please discuss in detail how the stereochemistry is established for your API. If the characterization refers to the spectroscopic comparison to the USP Reference Standard, please explain what comparison data supports the stereochemical features.

• We acknowledge that all the chiral centers of the drug substance are from starting material. However, the potential epimerization of C-4 center under basic conditions will raise uncertainty in stereochemistry. To ensure the consistent quality of drug substance, please provide additional spectroscopic evidences, including but not limited to NOE, to unambiguously determine the absolute configuration of C-4 center of drug substance. The same requirement also applies for characterization of your in-house working standards for impurities A, B, D and E.

• We acknowledge that you have provided ¹H NMR and ¹³C NMR, IR, MS and UV spectra to support the structure of R-enantiomer. However, the provided information cannot differentiate R-enantiomer from S-enantiomer. Please provide additional evidences such as comparison of chiral HPLC retention time or specific optical rotation with that of innovator drug if applicable or Mosher ester analysis for confirming the absolute configuration of R-enantiomer.

Not well-resolved or overlapped spectra

• Regarding ¹H NMR of drug substance, the peaks of protons 4, 8, 5, and 7 from d3.10 to 4.5 ppm are too broad to show the splitting patterns. Please provide a well-resolved ¹HNMR spectrum with sharp resonances. You may lower analyte concentration, elevate temperature, use a different deuterated solvent, or transform it to a free base to achieve better resolution.  In addition, please provide ¹H-¹H COSY to confirm proton assignment.

• The proton NMR spectrum you provided does not allow for adequate evaluation of the structure due to the poor resolution and solvent overlap with critical proton peaks. Please consider using other deuterated solvents with unambiguous peak assignment and enlarge the area with clustered signals for evaluation.

Miscellaneous

• Please provide the batch number of your drug substance for characterization analysis.

• The carbon NMR spectra provided have very low signal to noise ratios. According to your response, the solubility of drug substance in DMSO is about 76 mg/mL, which is sufficient to obtain a good quality of carbon NMR spectra. Please re-run the experiment and provide an explanation for any differences observed between the lots.

• Please revise the field strength of ¹³C NMR. It should be 100 MHz, ¼ of that of 1H NMR (400 MHz).

• The ¹H NMR spectrum provided for your molecule is not readable.  Please provide a legible copy.