Jan De Backer, CEO of Antwerp-based FluidDA, calculates that the combination of biomedical imaging and computational fluid dynamics offered by his company has the potential to reduce development costs for OINDPs by hundreds of millions of dollars. According to De Backer, FluidDA’s functional imaging methods provide significantly more sensitive parameters for measuring the effects of respiratory drugs than those currently in use, reducing the numbers of patients needed for clinical trials. The techniques also allow accurate modeling of deposition patterns for device optimization with minimal need for patient involvement.
Using high resolution computed tomography (HRCT), the company can build 3-dimensional, patient-specific models of airways, typically down to the 7th generation. Using standard HRCT machines, with the settings tweaked to minimize radiation exposure (down to 1-2 millisieverts), technicians perform two static scans pre-dosing, one at inspiration and one at expiration, as a baseline measure. After inhalation of the drug, at least one more scan is performed.
Combining the images allows the company to measure outcome parameters such as changes in volume and airway resistance that provide direct indications of bronchodilation. Where weakly correlated parameters such as FEV1 or scores on the Saint George Respiratory Questionnaire (SGRQ) result in the need for extremely large numbers of patients in order to achieve statistical significance, imaging has the sensitivity to generate statistically significant results with data from a relatively small number of subjects.
“It’s a misconception that spirometry is very cheap,” notes De Backer. While imaging might cost slightly more on a per patient basis – approximately €15-20,000 per patient, depending on the type of study – the greater sensitivity of the parameters allows a 10-20 fold reduction in the number of patients necessary. Even if spirometry cost only half as much as imaging per patient, a trial involving thousands of patients would still greatly exceed the cost of an imaging study involving ten times fewer patients.
Having access to 3-D airway models also would allow companies to select patients for clinical studies based on airway anatomy. Patients with extremely narrow upper airways, for example, or COPD patients with large portions of the lung affected by emphysema, may not inhale effective doses of drug. De Backer suggests that, “If you can take those patients out of the equation at the beginning, you can get a good look at the efficacy of the compound,” then later evaluate how to best deliver the drug to all patients.
In addition to saving money on clinical trials, providing a common parameter for measuring both animal and human outcomes has the potential to save money by easing the transition from pre-clinical to clinical studies. FluidDA has to date conducted imaging studies on rats, guinea pigs, and primates, generating the same types of airway models as in humans, allowing measurements of volume and airway resistance changes.
“It’s easier to have the same type of outcome parameter that can help you assess how it translates from the pre-clinical to the clinical stage,” notes De Backer. Using FEV1 for human studies is one thing, but, he points out, “it’s difficult to get rats to do spirometry.”
Once the company has created the patient-specific airway models, whether in animals or in humans, it can use computational fluid dynamics (CFD) to simulate deposition of inhaled medications, a technique that FluidDA has validated through scintigraphic studies. Several of the companies engineers, including De Backer, have backgrounds in aerospace engineering and have been able to apply CFD techniques to the more complex task of airway modeling.
Because the company funded its earliest studies on its own, FluidDA owns the rights to many of the airway models it has created and can provide deposition modeling services to companies that might not want to conduct trials of their own. The deposition simulations can be used for optimizing formulation and/or inhaler design without the need for any additional patient participation.
Theoretically, De Backer says, calculating the amount of drug deposited in the lung for each patient using CFD with the patient-specific airway models should give at least as accurate results as the more labor-intensive PK/PD testing necessary when using spirometry, and initial data have shown good correlation between the two methods. The company is currently working to confirm that correlation. As a post-processing step, the CFD requires little extra work, and it is relatively easy to include data on airway resistance and deposition for all of the study patients when calculating an effective lung dose.
FluidDA recently submitted a letter to the FDA indicating their intent to begin discussions with the regulatory agency and is beginning the same process with the European regulators. In the meantime, says De Backer, the company is staying very active in the scientific community and has published data from a number of studies funded by companies such as GSK, Novartis, and Chiesi as it works towards approval that would allow use of this type of data in marketing applications.
For companies interested in conducting trials using imaging techniques, FluidDA offers all of the necessary CRO services, from writing the study protocol, getting ethical approval, and initiating the trial through analyzing the data and submitting reports. The time necessary for such studies ranges from 3 months to a year, with a typical 30-patient study taking 6-8 months.
The company is also getting ready to launch a new service that would allow clinicians to upload patient scans for analysis. It only makes sense, De Backer comments, for doctors to determine whether the patient is receiving the proper dose and to evaluate the effectiveness of their patient’s inhaled drug regimens. The service should be available at a few centers in Belgium, where the company is headquartered, within a few months and will roll out gradually as the logistics are worked out.
