Outsourced API Manufacture The top five mistakes when outsourcing early phase Active Pharmaceutical Ingredient manufacture

1. Criteria for selection of Outsource Partner

The alignment of the technical capabilities of the outsource partner should be appropriate for the phase of development of the asset. For initial campaigns supplying toxicology studies and Phase 1 clinical supplies, the supply route is usually unoptimised. As a result, the chemistry can require considerable development before it can be scaled up successfully. So, for early phase projects you generally need a strong technical capability at the contractor, along with the ability to innovate. For later phase projects (Phase 2b/Phase 3) where the route of manufacture has generally been fixed, the focus shifts to Contract Manufacturing Organisations (CMOs) whose strength lies in routine manufacturing; this may be at a low cost facility in India or China.

In order to meet the regulatory requirements for Phase 1 clinical studies, it is also necessary that the selected Outsource Partner has cGMP compliant Pilot Plant capability and desirable that they have a “kilo lab facility” (a laboratory or pilot plant with capability to produce API batches up to about 1kg). Ideally, the outsource partner should also be able to QA-release API batches. From a logistical and financial perspective, you also want to minimise the number of technical transfers that need to be carried out to deliver the manufactured API, so it is highly advantageous for the process R&D labs and the analytical labs to be co-located with the kilo lab. This ensures that the chemists who have developed the processes are able to assist the pilot plant operators during processing and to physically observe the critical stages first hand.

Building a strong relationship with your outsource partners is key – they are more likely to respond positively to any additional request if you take the time to build a culture of mutual trust and respect with them. RML Pharma Consulting Ltd has built up a strong relationship with a small and focused network of outsource partners based in Europe, India and the US, who possess these capabilities.

By using RML Pharma Consulting Ltd to manage your outsourcing requirements, you also benefit from their proven track record of delivery.

2. Starting materials not commercially available

RML Pharma Consulting Ltd has seen many examples of projects from research groups where the commercial availability of starting materials has not been identified as an issue early on. For research chemists, commercial availability of a material generally only extends to whether the compound is available from a catalogue company. Rightly so, as their needs and objectives are different to those of us in a development and/or manufacturing environment.
The downside is that when a candidate enters development, with a starting material that is not commercially available at larger scale, there will be a delay in delivering the next campaign. This delay could be in the region of 3-6 months if the material in question is required early in the synthetic sequence and/or uses complex chemistry in its manufacture. The criteria used for selecting the custom synthesis manufacturer needs to be carefully considered. This decision really shouldn’t be based on who is the cheapest, but rather who is the most likely to succeed in addressing unforeseen development challenges and has the best likelihood of delivering the target amount of material of the right quality and on time.

Having early visibility of the current supply route from the research chemists is invaluable in identifying potentially long-lead time starting materials early on and may prevent unnecessary delays later on. It will also enable sourcing activities to commence off the critical path to supply of drug.

3. Supply routes not scaleable or consistently reproducible

In addition to the problem outlined above relating to availability of starting materials, the initial supply route may only be able to deliver 20-50g of API. The next requirement could be 0.5 to 1kg followed shortly afterwards by 5kg! The initial supply route may be long, have low yielding steps and require chromatography to purify intermediates and/or final drug substance

Process safety and reaction calorimetry data generated prior to scale-up may also preclude parts of the existing chemistry from being scaled up, for example some nitration reactions. RML Pharma Consulting Ltd have seen a number of process intermediates that are thermally unstable under the reaction conditions used to prepare and isolate them. High yielding steps at research scale can quickly translate into poor yields once you attempt to scale up due to an inherent lack of robustness of the process.

For early phase projects, it doesn’t make sense to invest a lot of resource/money into understanding the design space and proven acceptable ranges for each reaction in the supply route. However, as drug development progresses, the reproducibility of a manufacturing process becomes increasingly important. You can and should evaluate the route for ‘potential’ scale-up issues well before you intend to scale up – for example unpredictable reactions which produce variable yields on even a gram scale. You then need to initiate a focused program of process development activity in order to maximise your chance of success in the next campaign. As a contingency, alternative route development work should also be considered.

4. Poorly defined physicochemical properties of API

There is universal recognition that the physicochemical properties of the API are very important to its behaviour during formulation and ultimately, its bioavailability in-vivo. Solubility is a key attribute and this can be controlled by choice of salt (counter-ion) of the API. Poorly soluble compounds can be difficult to purify by crystallisation and will, generally speaking, adversely affect the absorption characteristics of the drug in-vivo. In addition to this, poorly soluble compounds will probably require particle size reduction (milling, micronisation or nanomilling) in an attempt to improve the solubility and thus improve bioavailability. This will also have a negative impact on the cost-of-goods for the product, which is rarely considered a priority in early process development and at times only seriously addressed when too late and problems arise.

It’s not unusual for a project to have commenced API manufacturing at the outsource partner without a clearly defined process for preparing the final API. In some cases, the final salt has yet to be identified, let alone the most stable polymorph and this adds to the challenge for the outsource partner and sponsor. The best way to proceed is to discuss in partnership exactly what solid form of the API is appropriate. Once the parent API has been identified some thought to screening, characterising and selecting a crystalline, stable salt with the desired solubility profile needs to be considered.

The sooner that the final salt is identified, the sooner the polymorph screen to identify the most stable polymorph can commence. This is important as polymorphs of a given salt may also have differing solubilities, so it makes sense to ‘fix’ the polymorph in time to support the definitive toxicology studies ahead of Phase 1. The number of stable polymorphs can often be extensive so this can consume time and resource and needs to be considered in the overall timeframe for delivery of the campaign. However, investing early in this work can reduce long term costs and eliminate risk as the prospect of repeating toxicological studies is universally undesirable.

Once the salt and polymorph have been selected, the process for delivering the final API via a salt formation or recrystallisation can be developed. This discrete particle forming step will aim to control the physical attributes of the API (particle size, shape etc) thus providing the formulation team with API that will be reasonably representative of longer term supplies.

5. Poorly defined impurity qualification strategy

There are basically three ways to supply API for a clinical study, for the purpose of this article let us consider a Phase 1 clinical study (assessing safety and tolerability). The tactic to employ for a successful sequence of events is to ensure that the toxicological grade batch is less pure (i.e. has more impurities) than the clinical grade batch. Historically, often early batches of API can be of a very high purity. Thus, the risk of unqualified impurities is a real and present danger. (Note; regulatory authorities require that impurities are qualified at particular levels in the batch and thus are required to be submitted to preclinical testing before administration into humans.

From an impurity perspective, the lowest risk approach is to fund the clinical study from a single batch of API, which was also used to fund the definitive toxicology studies (14day/28day), which precede the clinical study. That way, you are guaranteed not to have any unqualified impurities in the clinical batch. In this case, you only need to release one batch of drug substance and submit the data for this batch alone in the regulatory submission. Less complexity equates to less time and money in the long run.

Another approach is to start with a single batch of toxicology grade API and reprocess a portion to fund the clinical study. Assuming that you have developed your reprocessing step successfully, you should end up with higher purity API than what you started out with – but not necessarily! This approach will require you to release two batches of drug substance, one for the toxicology study and the other for the clinical study. Batch analysis data for both batches will need to be included in the regulatory submission.

The highest risk approach is to provide two batches of API via two separate campaigns, which may also be separated by some considerable time (let’s call them campaign 1 and campaign 2). This scenario may arise if enabling funding for the larger clinical campaign occurs only once the results from the toxicology study reads out and this is not an uncommon situation in smaller pharmaceutical enterprises. Thus, the impurity profile for the clinical study was ‘fixed’ via the toxicology study and campaign 1 – these impurities are qualified at the particular levels observed in the batch.

Since the chemistry is at an early stage of development, there is usually insufficient knowledge about the robustness of the chemical process used and the relationship to observed impurities to be able to demonstrate control of the impurity profile. So, there is a very strong probability that campaign 2 will produce API with a different impurity profile or the same profile, but with different levels of the same impurities. This is the batch that is destined for the clinical study! You may be able to justify proceeding into the clinical study with unqualified impurities but the likely scenario will involve the batch being reprocessed (controlling physical form as well) in order to purify.

Other impurities that need to be considered, depending on the chemistry, include heavy metals, solvents and inorganics. In addition to these, for clinical batches you also have to consider potential genotoxic compounds that are either used in the synthesis of the API or are side-products formed during the process. The level of risk can be assessed by carrying out a ‘genotoxic risk assessment’ and if necessary, you may need to determine levels of these impurities to very low levels in the API.