Stephen Tindal, Director - Science and Technology, Catalent05.09.17
The molecules emerging from pharmaceutical discovery pipelines are, increasingly posing significant challenges for formulation development experts. It is difficult to achieve the required bioavailability, preferred dosage form and desired therapeutic profile while using a fast, efficient development process that does not consume large amounts of time, money and precious active pharmaceutical ingredient (API).
A suitable dosage form must be created before a developmental medicine can start its journey through the clinic and, ultimately, to market. While big pharma companies will often have the necessary capabilities in-house to achieve this, it may well be outside the scope of a small biotech, and assistance from an organization with expertise in formulating candidates with bioavailability issues will likely be required.
Acceleration of this process can often be achieved by applying a parallel screening platform to the API. Catalent’s OptiForm Solution Suite is one example. The platform offers a strategy for early small molecule screening that minimizes the consumption of API. It collects data to characterize the molecule and evaluates multiple different technologies in parallel, including lipid-based formulation, particle size reduction and solid dispersion—spray drying and hot melt extrusion—in a process that takes 12 weeks. The ultimate aim is to ensure the selection of optimal technology by making a decision based on actual data, rather than solely subjective appraisals such as prior experience.
The platform uses the Developability Classification System (DCS). This builds on the earlier biopharmaceutics classification system (BCS), where APIs are categorized according to the dosage solubility versus permeability ratio. DCS is particularly useful for segmenting those molecules that fall into BCS Class II, where the solubility is poor, but the permeability is good. In the DCS model, this class is further split into a classification IIa, where the problem is a slow dissolution rate and where particle size reduction or change in salt form can help, and IIb, where the difficulty is intrinsic limited solubility and where lipid formulation and solid dispersion should be considered.
Each protocol within the OptiForm Solution Suite system includes three phases. The first, ‘assess’ phase, involves the collection of detailed information about the molecule’s form, solubility and stability, as well as some tests to evaluate technology “fit”. The second ‘enhance’ phase will either aim to improve the salt form, or develop candidate formulations using existing, proven technologies in parallel. In the third ‘delivery’ phase, a number of candidate formulations, for which two-week stability studies have been performed, will be provided along with the data compiled during the study to underpin ongoing development. The protocol is designed for molecules right at the start of the formulation development process, but it can be customized. For example, if some of the data are already available, if there is insufficient API, or if there are still choices to be made about the lead candidate, salt form or polymorphic form. Small companies are able to tap into the resources of a development function that handles many more molecules, and has expertise in commercializing various dosage forms across a broad range of commercially proven technologies.
Case study: TML-002
UK-based biotech Trio Medicines came to Catalent for assistance with progressing a drug into the clinic, code-named TML-002. TML-002 is the acetyl derivative prodrug of TML-001, the major metabolite of a much-studied molecule that is well tolerated in both preclinical and clinical studies. Permission was therefore granted for a first-in-human study of TML-002, with no further toxicology studies. TML-002 has similar potency to the parent compound, but is more soluble at the lower pH of the small intestine.
However, its bioavailability proved low in healthy subjects, and hence the acetylated prodrug, TML-002, was created. Although this proved more bioavailable in first-in-human tests, it remained too low. This led to concerns about the prospects for Phase II clinical trials. Variable bioavailability results suggested that the unmodified or the micronized formulations of API in capsules, were suboptimal.
Trio engaged Catalent to use OptiForm Solution Suite to gain an understanding of the issues posed by the molecule and its formulation, and to find some way to increase its bioavailability. The drug is a small molecule. The reason for the poor bioavailability was not clear, but solubility in water was low, and the molecule falls into BCS class II. The target dose for each tablet was 25-50 mg, with the potential to extend the dosage in clinical trials up to 100 mg.
To start the assessment phase of the project, information on the molecule was collected to inform subsequent formulation development. As screens for excipient compatibility must be carried out before the dosage form has been selected, the API is screened across a standard set of excipients, looking for broad evidence of incompatibility based on the presence of important functional groups.
TML-002 was found to be somewhat unstable, especially in lipid solution. Data were collected on the solid state and particle size, with differential scanning calorimetry, thermogravimetric analysis, optical microscopy, dynamic vapor sorption and powder X-ray diffraction experiments also being carried out. A generic HPLC method was employed to determine assay and related substances, with traces being compared to samples of excipient alone.
Powder X-ray diffraction showed that the API was crystalline, with a melting point within the typical range (125-175°C); it was neither micronized nor hygroscopic. One minor observation was that there was some loss of volatile materials during the thermogravimetric analysis. The crystal form appeared to be fairly straightforward, with no reasons to anticipate any major handling issues.
Solution data were also collected which exhibited a pH dependent solubility profile that was consistent with the chemical structure. Solubility in fed and fasted state intestinal fluid was similar to that seen in a pH 6.5 buffer, which may indicate that co-administering the drug with food would not increase its bioavailability.
In many of the tests, any residual solid material was collected, and further X-ray crystallography experiments performed to look for evidence in a change in its crystallinity. This was done to confirm the crystallization characteristic of the molecule from a potentially new set of solvents not used in previous work.
With solubility in simulated intestinal fluid established, the drug’s DCS classification could be confirmed. The solubility limited absorbable dose—the parameter that determines whether it is Class IIa or IIb—was calculated at 367 mg. Furthermore, the cutoff from DCS IIa and I was calculated at 57mg. This means for a 100 mg dose, the molecule was DCS IIa but for a 50 mg dose, the drug fell just over the border into Class I. The permeability calculations did not take into account transporter systems such as efflux. Based on the poor oral bioavailability, it was concluded that prior studies had not been successful, either due to issues with wettability of micronized API or the API may be a substrate for efflux or other transporters.
Evaluating the options
With these data in hand, the next step was to evaluate potential formulations. There were three primary technologies that could be applied: particle size reduction, lipid-based formulation, and amorphous solid dispersion via hot melt extrusion. For particle size, since straightforward micronization using an air-jet mill was not previously effective, the API was co-milled with surfactant to improve solubility/wettability.
First, the potential excipients were checked to ensure they were acceptable and below any maximum daily intake level. Candidate formulations at drug loadings ranging from 2.5% to 10% were prepared to evaluate the different options.
The drug proved sufficiently soluble in four excipients as a lipid solution, and the solubility was reasonable in a further seven. This gave sufficient flexibility to develop several candidate lipid-phase formulations.
For solid dispersions, TML-002 proved miscible with six preliminary formulations in the screen, but these were limited to a 20% drug loading; all proved immiscible at 40% and 60%. Further development work in future may enable this range to be increased.
Trio elected to test all of the available lipid formulations, as well as two solid dispersions and one co-micronized sample. The lipid formulation showed a slight increase in decomposition products; it was proposed that future encapsulation within a softgel may be preferable, to prevent oxidation.
The solid dispersion indicated higher levels of degradation, and this is likely to be a result of thermal degradation from the manufacturing process. Co-micronized API meanwhile, was found to be stable in the solid state, an unsurprising finding based on the API data.
Delivery of material
Samples of four candidate formulations were provided to Trio for pharmacokinetic evaluation. These included two lipid-based formulations, one solid dispersion and one co-micronized sample, and were accompanied by a ‘risk ranking’ to help aid the selection of the best formulation (see Table 1).
The solid dispersion process had proved challenging and was not recommended. The co-micronized API scored lowest on the risk ranking, proving a good fit for DCS, simplicity of manufacture and both physical and chemical stability.
In pharmacokinetic studies, very little prodrug (the acetylated form) was detected in the blood, as it was quickly converted to the active metabolite. But administration of the prodrug improved the bioavailability of the major metabolite.
While the solid dispersion formulation did not improve the bioavailability of the active metabolite when tested in a healthy volunteer, the remaining three alternative formulations gave better results.
Trio’s assessment continues, and there may be potential to make additional incremental improvements to bioavailability before more extensive Phase I evaluation is carried out.
The project showed the importance of starting formulation development as early as possible in the development cycle, as issues with solubility can be complicated. Collecting more data resulted in the realization of issues that were only guessed at previously. Planned start dates of clinical trials are too often delayed because of formulation problems. It is wise to incorporate platform screening into every program where solubility is a limiting factor in achieving the required levels of drug in the blood.
A suitable dosage form must be created before a developmental medicine can start its journey through the clinic and, ultimately, to market. While big pharma companies will often have the necessary capabilities in-house to achieve this, it may well be outside the scope of a small biotech, and assistance from an organization with expertise in formulating candidates with bioavailability issues will likely be required.
Acceleration of this process can often be achieved by applying a parallel screening platform to the API. Catalent’s OptiForm Solution Suite is one example. The platform offers a strategy for early small molecule screening that minimizes the consumption of API. It collects data to characterize the molecule and evaluates multiple different technologies in parallel, including lipid-based formulation, particle size reduction and solid dispersion—spray drying and hot melt extrusion—in a process that takes 12 weeks. The ultimate aim is to ensure the selection of optimal technology by making a decision based on actual data, rather than solely subjective appraisals such as prior experience.
The platform uses the Developability Classification System (DCS). This builds on the earlier biopharmaceutics classification system (BCS), where APIs are categorized according to the dosage solubility versus permeability ratio. DCS is particularly useful for segmenting those molecules that fall into BCS Class II, where the solubility is poor, but the permeability is good. In the DCS model, this class is further split into a classification IIa, where the problem is a slow dissolution rate and where particle size reduction or change in salt form can help, and IIb, where the difficulty is intrinsic limited solubility and where lipid formulation and solid dispersion should be considered.
Each protocol within the OptiForm Solution Suite system includes three phases. The first, ‘assess’ phase, involves the collection of detailed information about the molecule’s form, solubility and stability, as well as some tests to evaluate technology “fit”. The second ‘enhance’ phase will either aim to improve the salt form, or develop candidate formulations using existing, proven technologies in parallel. In the third ‘delivery’ phase, a number of candidate formulations, for which two-week stability studies have been performed, will be provided along with the data compiled during the study to underpin ongoing development. The protocol is designed for molecules right at the start of the formulation development process, but it can be customized. For example, if some of the data are already available, if there is insufficient API, or if there are still choices to be made about the lead candidate, salt form or polymorphic form. Small companies are able to tap into the resources of a development function that handles many more molecules, and has expertise in commercializing various dosage forms across a broad range of commercially proven technologies.
Case study: TML-002
UK-based biotech Trio Medicines came to Catalent for assistance with progressing a drug into the clinic, code-named TML-002. TML-002 is the acetyl derivative prodrug of TML-001, the major metabolite of a much-studied molecule that is well tolerated in both preclinical and clinical studies. Permission was therefore granted for a first-in-human study of TML-002, with no further toxicology studies. TML-002 has similar potency to the parent compound, but is more soluble at the lower pH of the small intestine.
However, its bioavailability proved low in healthy subjects, and hence the acetylated prodrug, TML-002, was created. Although this proved more bioavailable in first-in-human tests, it remained too low. This led to concerns about the prospects for Phase II clinical trials. Variable bioavailability results suggested that the unmodified or the micronized formulations of API in capsules, were suboptimal.
Trio engaged Catalent to use OptiForm Solution Suite to gain an understanding of the issues posed by the molecule and its formulation, and to find some way to increase its bioavailability. The drug is a small molecule. The reason for the poor bioavailability was not clear, but solubility in water was low, and the molecule falls into BCS class II. The target dose for each tablet was 25-50 mg, with the potential to extend the dosage in clinical trials up to 100 mg.
To start the assessment phase of the project, information on the molecule was collected to inform subsequent formulation development. As screens for excipient compatibility must be carried out before the dosage form has been selected, the API is screened across a standard set of excipients, looking for broad evidence of incompatibility based on the presence of important functional groups.
TML-002 was found to be somewhat unstable, especially in lipid solution. Data were collected on the solid state and particle size, with differential scanning calorimetry, thermogravimetric analysis, optical microscopy, dynamic vapor sorption and powder X-ray diffraction experiments also being carried out. A generic HPLC method was employed to determine assay and related substances, with traces being compared to samples of excipient alone.
Powder X-ray diffraction showed that the API was crystalline, with a melting point within the typical range (125-175°C); it was neither micronized nor hygroscopic. One minor observation was that there was some loss of volatile materials during the thermogravimetric analysis. The crystal form appeared to be fairly straightforward, with no reasons to anticipate any major handling issues.
Solution data were also collected which exhibited a pH dependent solubility profile that was consistent with the chemical structure. Solubility in fed and fasted state intestinal fluid was similar to that seen in a pH 6.5 buffer, which may indicate that co-administering the drug with food would not increase its bioavailability.
In many of the tests, any residual solid material was collected, and further X-ray crystallography experiments performed to look for evidence in a change in its crystallinity. This was done to confirm the crystallization characteristic of the molecule from a potentially new set of solvents not used in previous work.
With solubility in simulated intestinal fluid established, the drug’s DCS classification could be confirmed. The solubility limited absorbable dose—the parameter that determines whether it is Class IIa or IIb—was calculated at 367 mg. Furthermore, the cutoff from DCS IIa and I was calculated at 57mg. This means for a 100 mg dose, the molecule was DCS IIa but for a 50 mg dose, the drug fell just over the border into Class I. The permeability calculations did not take into account transporter systems such as efflux. Based on the poor oral bioavailability, it was concluded that prior studies had not been successful, either due to issues with wettability of micronized API or the API may be a substrate for efflux or other transporters.
Evaluating the options
With these data in hand, the next step was to evaluate potential formulations. There were three primary technologies that could be applied: particle size reduction, lipid-based formulation, and amorphous solid dispersion via hot melt extrusion. For particle size, since straightforward micronization using an air-jet mill was not previously effective, the API was co-milled with surfactant to improve solubility/wettability.
First, the potential excipients were checked to ensure they were acceptable and below any maximum daily intake level. Candidate formulations at drug loadings ranging from 2.5% to 10% were prepared to evaluate the different options.
The drug proved sufficiently soluble in four excipients as a lipid solution, and the solubility was reasonable in a further seven. This gave sufficient flexibility to develop several candidate lipid-phase formulations.
For solid dispersions, TML-002 proved miscible with six preliminary formulations in the screen, but these were limited to a 20% drug loading; all proved immiscible at 40% and 60%. Further development work in future may enable this range to be increased.
Trio elected to test all of the available lipid formulations, as well as two solid dispersions and one co-micronized sample. The lipid formulation showed a slight increase in decomposition products; it was proposed that future encapsulation within a softgel may be preferable, to prevent oxidation.
The solid dispersion indicated higher levels of degradation, and this is likely to be a result of thermal degradation from the manufacturing process. Co-micronized API meanwhile, was found to be stable in the solid state, an unsurprising finding based on the API data.
Delivery of material
Samples of four candidate formulations were provided to Trio for pharmacokinetic evaluation. These included two lipid-based formulations, one solid dispersion and one co-micronized sample, and were accompanied by a ‘risk ranking’ to help aid the selection of the best formulation (see Table 1).
The solid dispersion process had proved challenging and was not recommended. The co-micronized API scored lowest on the risk ranking, proving a good fit for DCS, simplicity of manufacture and both physical and chemical stability.
In pharmacokinetic studies, very little prodrug (the acetylated form) was detected in the blood, as it was quickly converted to the active metabolite. But administration of the prodrug improved the bioavailability of the major metabolite.
While the solid dispersion formulation did not improve the bioavailability of the active metabolite when tested in a healthy volunteer, the remaining three alternative formulations gave better results.
Trio’s assessment continues, and there may be potential to make additional incremental improvements to bioavailability before more extensive Phase I evaluation is carried out.
The project showed the importance of starting formulation development as early as possible in the development cycle, as issues with solubility can be complicated. Collecting more data resulted in the realization of issues that were only guessed at previously. Planned start dates of clinical trials are too often delayed because of formulation problems. It is wise to incorporate platform screening into every program where solubility is a limiting factor in achieving the required levels of drug in the blood.
