Enabling the synthesis of clinical API: Onyx’s approach to defining phase 1 drug substance specifications

The establishment of appropriate specifications for the release of a new drug substance is a key part of obtaining regulatory approval for clinical studies and ultimately protecting the safety of study participants. Guidance exists from the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (commonly referred to as ICH) which outlines the basis of this process as applied to commercial products (eg. ICH Q6A). However, much of this guidance is not directly applicable to short term clinical studies with a limited number of participants, and where exposure to the candidate drug is low.
When considering the synthesis of drug substance for phase 1 specifically, there is often a limited understanding of critical process parameters (CPPs) and associated critical quality attributes (CQAs) and therefore a more pragmatic approach to specifications is required based on the limited batch data available at the time.

A typical early Phase drug substance specification

The specification represents a critical quality standard which ensures the drug substance produced is suitable for its intended use. A phase 1 drug substance specification will generally include a suite of basic tests and acceptance criteria as outlined below:

• Appearance testing – a basic description of colour and physical form and a physical inspection to confirm no visual contamination.
• Identity testing – including at least 2 of FT-IR, 1H-NMR, MS, HPLC retention time. NMR is typically included for early phase work due to the wealth of information it imparts. The methods chosen should be suitable for distinguishing related compounds, including salts and enantiomers. Solid form analysis (by XRPD / DSC) to identify polymorphic form is also typically included (including identification of solvates, hydrates and amorphous forms). A typical specification would simply indicate that the analysis conforms to reference.
• Impurities testing, covering the areas listed below, forms a key part of the specification and will be discussed later in greater detail.
  • Chemical purity by HPLC
  • Related substances by HPLC
  • Assay / potency
  • Residual solvents by (HS)GC, or in some cases 1H NMR.
  • Elemental analysis by ICP-MS (or OES).
  • Specified impurities by HPLC / GC / NMR / MS-SIM – this will typically also include testing for any potentially mutagenic impurities (PMIs).
• Assay value or potency – A specific, stability indicating procedure is required, typically either HPLC assay against a fully characterised reference standard, or quantitative NMR where this is not available (which is often the case during Phase 1 work). A limit of 95.0 – 103.0% is typically employed (on an anhydrous, solvent free basis). The expected value should generally be ±2% of the chemical purity figure given the accuracy of these methods.

Depending on the nature of the compound and route of synthesis employed, the following additional testing will then be considered on a project specific basis:

• Chiral purity – opposite enantiomer content typically determined by chiral HPLC with a limit of not more than 1.0%. For molecules possessing multiple chiral centres where a wide range of diastereomers are possible, an understanding of the chemistry will drive the analytical approach to ensure that any isomers that could be present in the API are adequately controlled.
• Water content – Karl Fischer titration, often with a report result or nominal 0.5% specification limit. Other specifications may be set where materials are known to be a hydrate (eg. ±20% of theoretical value), be hygroscopic or readily degraded by moisture.
• Residue on ignition – total inorganic content as determined by sulphated ash, often with a nominal 0.5% limit. Higher levels are expected where the drug substance is an inorganic salt.
• Determination of particle size distribution where required (eg. by laser diffraction, typically driven by formulation considerations).
• Determination of counter ion where appropriate – testing performed by 1H NMR, titration or IC as appropriate with a recommended early specification of ±10% of the theoretical value.
• Microbial – standard pharmocopeial limits apply for oral and include determination of Total Aerobic Microbial Count (TAMC), Total Yeast and Mold Count (TYMC) and absence of E. coli. For other formulations a more detailed specification may be required and must be discussed on a case-by-case basis.

Justification of specification is required for each procedure and acceptance criterion and should refer to any relevant batch data (which will initially be limited), pharmacopoeial standards, test data from any existing toxicological and clinical studies, and the results of accelerated and long-term stability studies. This justification will form part of the chemistry, manufacturing and controls (CMC) section of the relevant Investigational Medicinal Product Dossier (IMPD) or Investigational New Drug Application (IND) submission.

In many cases an initial phase 1 limit will be proposed which may be less stringent than that specified by the relevant ICH guidance. It is anticipated that these limits would be refined towards the commercial API acceptance criteria as the project progresses through subsequent clinical phases. Revisions to the specification should be made based on greater understanding of the process gained through additional work such as DoE and fate and purge studies, and additional raw material controls.

Impurity testing

The impurity testing requirements for a particular API will be strongly process dependant and should be selected carefully to highlight any quality issues which may arise from the particular chemistry in use, especially those with serious consequences for product safety, or those which have already been observed during process development. Safe limits should be established based on sound toxicological evidence where available however this is often limited during early phase studies.
Control of organic impurities (related substances) forms one of the most important parts of the specification and is often the most difficult to set appropriately and maintain consistently. In some cases it may be possible to identify key impurity control points throughout the process and therefore take some analysis away from final product by introducing control at an earlier intermediate step. This gives greater opportunity to reprocess materials should an unexpected result be obtained for example by repeating a workup procedure or recrystallization.

A key factor in determining the specification for the first clinical material is the impurity profile obtained from the GLP toxicology batch. The toxicology batch should aim to incorporate all the main process impurities observed during the final stages of the synthesis (either process intermediates or impurities) at a level of 0.1 – 0.5% whilst maintaining a total purity of not less than 97.0%. This can often be achieved by limiting washes as part of a workup or filtration, recycling of the liquors when washing a filter cake, omitting a recrystallization step or by spiking. Time is well invested in getting the profile of the toxicology batch right so that known impurities can be qualified ahead of future campaigns.

Following the toxicology study limits for the known impurities can be set in consultation with a toxicologist and this is a key step in determining acceptance criteria for the clinical batch. The qualified impurity level can often be several times the level present in the toxicology batch. However, the specification limit for such impurities would generally be capped at not more than 1.0%. For any new unqualified impurities, a limit of not more than 0.15% (or NMT 1.0 mg/day whichever is lower) is proposed by ICH Q3A. As previously discussed, this limit is intended to apply to commercial products and it is often preferable to justify a higher limit for early phase work based on a paper by J. Harvey et al where a limit up to 0.7% is proposed (or 5 mg/day whichever is lower, study length <6 months). However, this approach would always involve detailed discussion with the wider project team and sponsor to ensure sufficient justification exists – with a limit up to 0.5% widely accepted at early phase. An overall chemical purity of not less than 98.0% would represent a specification with a good chance of acceptance by the relevant regulatory body.

Identification of impurities above a threshold is also expected as part of ICH Q3A (0.10%, or 1.0 mg per day whichever is lower). For early phase drugs this can be difficult, and a more pragmatic approach is required. If masses for impurities can be obtained by LCMS this can be combined with knowledge of the chemistry and potential mechanism of formation to derive likely structures. Whilst this does not offer conclusive structural proof it is deemed sufficient for the phase of development with the exception being impurities present at a high level where more in depth work would be required.

Specific impurities including those which may flag as potential mutagens (PMIs) will often have a dedicated test procedure and acceptance criteria derived from the ICH M7 guidance (which is applicable during all clinical phases). Due to the potentially low levels of such impurities these are often performed as a limit test giving a simple pass/fail result. The Onyx Scientific approach to managing PMIs is discussed in more detail in a dedicated blog post [link].

Residual Solvents

For common residual solvents well established limits exist and are documented as part of the ICH Q3C guidance and these limits are also normally adhered to during early phase work. Analysis often focuses on the solvents used in the last 3 stages of chemistry as these are most likely to appear in the final product. Problem solvents from the latter stages or final recrystallization can often be handled via an ICH “option 2” dose-based approach if required.

Elemental Impurities

Similarly for elemental impurities well established limits exist (ICH Q3D) which would be expected to be adopted for early phase work. Depending on the route of administration testing for a basic set of key elements would be performed (eg. class 1 and 2A for oral administration, additionally class 3 elements for parenteral and inhalation). It is preferable to adopt an “option 2” dose-based approach when setting elemental specifications as trace quantities present in raw materials and processing aids may otherwise cause an out of specification result especially when the sources of these materials are not yet well controlled. Where elemental impurities are deliberately introduced as part of the synthesis careful consideration should be given to the initial point of analysis so that remedial actions can be taken in case of an out of specification result (eg. repeat of workup or scavenging steps). By performing initial analysis at an earlier stage or as an in process check it is then possible to gain confidence of a pass before committing to final product analysis.

In Conclusion

The setting of specifications appropriate for early phase work requires interpretation of relevant guidance, batch data and toxicological information, as well as good judgement informed by detailed process and analytical knowledge. Overly restrictive specifications should be avoided in early phases where limited process knowledge exists and manufacturing may not yet have been demonstrated at full scale. Above all the specification should be considered as an active document that is constantly refined and tightened as further information becomes available and process knowledge expands.

^{¹J. Harvey et al. 2017. Management of organic impurities in small molecule medicinal products: Deriving safe limits for use in early development., Regul. Toxicol. Pharmacol. 84, 116-123, https://doi.org/10.1016/j.yrtph.2016.12.011}