From better patient selection to optimal approval rates and higher drug pricing.
A tremendous effort has to be made before clinical trials on humans are initiated for a potential drug. Nevertheless nine out of ten drugs entering phase 1 do not get approval.
The likelihood of bringing a drug entering Phase 1 to the market differs significantly in the various disease areas, from 26.1 % at the high end for hematology to 5.1% at the low end for oncology with an average of 9.6% for all disease areas.
Three out of four failing drug candidates fail due to lack of efficacy or safety.
Realizing an intelligent patient selection strategy as a fundamental component of each clinical trial is guaranteed to maximize the likelihood for approval of a new drug. This can be achieved with the use of biomarkers which help
Today, only a tiny percentage of clinical trials include the use of biomarkers. Although such studies rely only on single biomarkers the probability of success is still significantly increased.
OakLabs has gone a step further and moved from single biomarkers to complex biomarker signatures which explain a several times larger proportion of responses and therefore facilitate a reliable identification of responders. Based on our unique EVO algorithm which combines concepts from artificial intelligence and machine learning, our complex biomarker signatures build the basis for a superior improvement of the approval rate to even 68%.
We start with the development of complex biomarker signatures for patient selection based on OMICs data. Our scope of work comprises the acquisition of high throughput biological data such as RNA, protein or metabolites required for biomarker development via the molecular testing of each patient for presence of the responder signature as well as the development of a CE IVD including the required IEC 623044 compliant analysis software.
Other than current state of the art approaches, OakLabs' biomarker signature development starts during clinical research phase 1 or 2 and therefore circumvents serious translation risks from preclinics. Our streamlined processes and longstanding experience guarantee that the only rate-limiting step in a drug CDx co-approval is the drug and never the CDx.
By using machine learning algorithms, OakLabs achieves significant improvements of the accuracy of complex biomarker signatures. The representative ROC curve in fig. 3 illustrates the superiority of the protein type biomarker signature of eight analytes after OakLabs' support.
Out of the broad range of benefits, table A contrasts typical savings resulting from OakLabs' biomarker signature based patient selection compared to a conventional clinical study design as well as to patient selection based on a single biomarker. The performance parameters sensitivity and specificity refer to the colored data points of the ROC curves in figure 3.
|Patient selection||Conventional||Traditional Biomarker||OakLabs's Biomarker Signature|
|Sensitivity of responder selection||/||90%||95%|
|Specificity of responder selection||/||60%||94%|
|Overall number of patients to recruit||1,500||1,690||1,375|
|No. of patients to include based on biomarker test||/||990||945|
|No. of patients to exclude based on biomarker test||/||700||430|
|Total patient cost of phase 3||30,000,000 USD||23,000,000 USD||21,050,000 USD|
A companion drug that is capable of identifying strong responders of a specific therapy enables higher pricing and therefore fully compensates for the drawback of a reduced market share.
Table B summarizes the costs for phase 2 of a potential drug for a gastrointestinal therapeutic area. Based on the stated responder rate, the improved patient selection by the developed biomarker signature as well as the overall cost per patient in the clinical study, the following savings in phase 2 and 3 were calculated:
|Total cost of phase 2||16,000,000 USD|
|Number of patients||400 patients|
|Overall patient cost in phase 2||7,200,000 USD||Benefits|
|Increased responder rate with biomarker signature for patient selection||60 -> 95 %|
|No. of excluded patients based on biomarker signature||140 patients|
|Total budget saved||6,500,000 USD|
|Total time saved||8 months|
Though OakLabs' strategy and services require some inital effort, the resulting biomarker signatures guarantee significant savings in the whole subsequent progress and support higher drug prices on the market.
Benefits of OakLabs' biomarker signature
We have a proven track record and can look back at numerous successful CDx projects. Learn more about the reasons why pharma companies worldwide trust in OakLabs:
There is a broad range of data you can acquire from patients including including proteins, mRNA, miRNA, proteins, DNA methylation, metabolites or SNPs in various sample types like plasma, serum, urine, liquore, biopsies or facies. Which sample type and biomarker type is most suitable strongly depends on the individual clinical trial. The table below gives a brief overview of regular settings we support/accompany. Nonetheless, we profoundly evaluate each project individually and therefore ensure to proceed with the best strategy.
|Technology for signature development||Antibody Microarray Platform||DNA Microarray Platform||LCMS and GCMS|
|No. of detected analytes||> 900 proteins||> 30,000 mRNAs||> 1,000|
|Analytes||Broad set of different protein classes as well as relevant pathways in biomedical studies||Genome-wide gene expression analysis||~30% known and ~70% unknown metabolites (polor and lipid metabolites)|
|Compatible sample types (starting material)||Plasma, Serum, Urine, Liquor||PBMCs, Tissue samples||Plasma, Serum, Urine, Liquor, Facies|
|CDx translation technology||ELISA||RT-qPCR or micrarray-based||LCMS / GCMS / Other (depends on particular metabolites in biomarker signature)|
Rapid translation to CDx
broad equipment compatibility to run CDx in labs
|Rapid and cost-efficient translation to CDx
flexible technology depending on marker set size (RT-qPCR or microarray)
|Diverse technology options for translation to CDx|
We need samples from responders and non-responders to identify a responder signature. Therefore, we can start in any phase where patients are included in the clinical trial.
High throughput data of 20 responders and 20 non-responders are sufficient for a first evaluation whether a biomarker signature can be developed. Several hundred of patients are required for a robust biomarker signature towards CE IVD CDx.
OakLabs' innovative 9-year developed EVO algorithm manages all serious conflicts associated with biomarker signatures including overfitting, missing values and false classification. EVO's algorithms are based on artificial intelligence (AI) and machine learning and provide superior results.
OakLabs utilizes graphic cards (GPUs) to handle EVO's extreme computing demand.
EVO's biomarker signature is transferred to OakLabs' LDT development and FDA & IEC 62304 compliant software development.
With the help of a biomarker signature you will know in advance which patients will be affected by a treatment and which ones would even suffer from side effects. Thus, you can exclude a number of patients and save money. At a later stage, a companion drug that is capable of identifying strong responders of a specific therapy will enable a higher pricing.
This question cannot be answered easily and depends on the individual project. We are usually working with performance-based pricing components which will also differ in every development project. Please contact us to receive a first price estimation.
Please write us a short email to email@example.com expressing your interest and - if possible - also provide us with some initial information about your drug portfolio. One of our CDx experts will get in touch with you and discuss your needs and expectations in detail.
A biomarker refers to a measurable indicator of some biological state or condition.
One application of biomarkers is the prediction of a response to a therapeutic intervention.
CDx stands for Companion Diagnostics which basically means a test to predict patients who would benefit from a specific treatment.
It can be used as a companion to a therapeutic drug.
Companion diagnostics are co-developed with drugs to help selecting or excluding patient groups.
The prediction of the response is based on biomarkers.
While a single biomarker can only explain a small portion of the responses, a biomarker signature is capable to access the major proportion of the complex biological process. The use of a rapid and treatment-specific biomarker signature enables a new era of reliability in responder selection.
For most drugs it is recommendable to check for the presence of a responder signature either in the clinical trials phase or for market access strategies. However, it makes particular sense for drugs in competitive markets which do not have a standalone first-line status and are in need for strong sales arguments. In addition, identification of responders can be favorable for drugs with a long efficacy, because they enable a more suited treatment for the non-responders and prevent them from suffering and side effects.